Discourse Studies (Sage Benchmarks Discourse Studies)

SAGE BENCHMARKS IN DISCOURSE STUDIES DISCOURSE STUDIES VOLUME I EDITED BY TEUN A . VAN DIJK Introduction and editori...

Author: Dr. Teun A. van Dijk

504 downloads 4713 Views 127MB Size Report

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

DOWNLOAD PDF

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.


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


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


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


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


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


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


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.


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


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'


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.


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


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 ; —


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


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.


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


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


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


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 ' ârlaire & 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


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


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


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)


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


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


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 abit 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


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 imû™^ 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


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 "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


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 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 -] [ =! laughs]. 7 Father: [ [>]. 24 Father: [] [=! laughing]. 31 Father Oh [=! laughs] []. 37 Matthew: [ [>] 53 Matthew: [ [>] [>| 2 Ruti: <ma [//] ex haya ba-gan> [ )< 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? ] 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: [ [>] Mother: <ma at medaberetxx> [] What are you saying, things like this . . . dvarim kaele + ... Father: []. Father: [«, 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 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 " = 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


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


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


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


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


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


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


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


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 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

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


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


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


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*> ..„„:.. 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 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 thuh 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 s. Gwen: AH character 'n (0.3) beautiful. Lesley: Ye:s. (0.7) Gwen: B't J_ suppose she must 'v come t' the end of'er (.) tether 'n just walked out then. The pause following Lesley's minimal agreement/acknowledgement suggests a slight hiatus, associated with the fact that her previous assessment was insufficient to close the topic. Again, her production of a figurative summary may be a move to bring about the topic closure that was attempted or imminent in the turns before. 12. Lesley initially responds in 6 with a low, brief chuckle, in which she seems to recognize .-1

ni

,*H

..

—

•

-

"

122

DISCOURSE STUDIES

might otherwise have produced in her prior response slot. Her doing so simul taneously with Steven's "Yeah" indicates, of course, that it was independent of, and unprompted by, Steven's turn. (On the possible relations between puns and sum maries/story endings, see Sacks 1992:2.419-30.) 13. The imperative mood here, "take this with a dollop 'v salt", suggests that Robbie is telling Lesley that she (Lesley) should not take all her (Robbie's) previous complaints too seriously - that she's not too disheartened by teaching at the school. There are other idiomatic aspects of Robbie's summary turn, particularly her use of the idiomatic (but not, perhaps, figurative) "getting everything done in the day." (from 10) 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. Lesley: Yes: that's ri:ght,= Note that her self-repair - she began with "getting all," after which "all" appears to be replaced with "everything" - may be akin to the instances illustrated in exx. 2-4: self-repairs in which speakers begin with a non-idiomatic version, but subsequently change it to an idiomatic formulation. 14. This shift in the focus of the topic (again, not marked as a change to a new topic) ex ploits the opportunity afforded by the figurative expression to develop the topic in a stepwise fashion. We use "exploit" here to draw attention to the possibility that speakers may use (whether intentionally or deliberately is not in point here) the op portunities that figurative idioms afford, not only to summarize and exit from a prior topic, but also to lead them, in an apparently topically connected fashion, to a "next" topic, something they wish to talk about. In such cases, the figurative expres sion has a bivalent character, summarizing a prior topic and simultaneously pointing forward to another matter. A case in which the pivotal work of the figurative expres sion seems especially to result from the speaker exploiting its bivalent property is the following. [NB:11:2:5] (Speakers have been talking about the assassination of Robert Kennedy a few days before, to which Nancy refers when she says "everybody is talking about it") Nancy: Yeah it's bgen a rough week an everbuddy is (.) you know (0.2) Emma: Mm hm Nancy talking about it en everbuddy: course L. don't know whether it's that or just that we're js:t (.) completely bo:gging down at work,h .hhhhmh (•) Nancy: E [r whatta WIH: WITH ME; with my finals? hhhh Emma: [Oh: well e_v r y buddy's sa::d Nancy: hhuh uh:[:: Emma: [Oh ho:w'd you do. with yer finals. Nancy manages what is really a two-stage idiomatic exit from the topic of the assassination. Initially, she summarizes it as having been a "rough week" (which is idiomatic, if not quite figurative). Then she offers, as a possible reason for it having

DREW A N D HOLT

123

FIGURES OF SPEECH

bo:gging down at work". (Note that the transition from her first reference to "everybody," through a next "everybody," and then to "we," seems to enable Nancy to disengage from talking about "the population at large/in the city," and to move to focus instead on her colleagues at work, "we".) The figurative "bogged down" at work is then pivotal, by allowing her the opportunity to suggest another factor that has contributed to its having been a rough week - a factor associated with another kind of "work", namely the finals examination she has taken for a course at a local university. This touches off (as it might well have been designed to do) an enquiry from Emma about how she has done in her finals. So this idiomatic utterance, and in particular the figurative "bogged down" at work, has been pivotal in the transition from talking about the assassination to talking about Nancy's course and the examin ation. In this way, the bivalent properties of figurative summaries may perhaps be exploited as a device for managing, rather smoothly, the move to a next topic (and perhaps even a particularly favored topic). 15. The matter of securing the other's agreement, affiliation, or sympathy may be of some interactional moment, since figurative idioms are used so recurrentiy in com plaint sequences, where a complainant can have some interest in whether or not his/her recipient will affiliate with him/her (see Drew & Holt 1988). 16. Note that this last figurative expression conveys a position which perhaps draws closer to Claire's, since the more the better has elements of Claire's suggestion that going ahead with the show and sale might depend on a sufficient number of people being interested - while nevertheless allowing that the thing may "snowball."

References Abeille, Anne (1995). The flexibility of French idioms: A representation with lexicalized tree adjoining grammar. In Everaert et al. (eds.), 15-42. Atkinson, J. Maxwell, & Heritage, John (1984), eds. Structures of social action: Studies in conversation analysis. Cambridge & New York: Cambridge University Press. Black, Max (1972). The labyrinth of language. Harmondsworth, UK: Penguin. Bobrow, Samuel. A., & Bell, Susan M. (1973). On catching on to idiomatic expressions. Memory and Cognition 1:343-46. Bolinger, Dwight (1976). Meaning and memory. Forum Linguisticum 1:1-14. Brown, Gillian, & Yule, George (1983). Discourse analysis. Cambridge & New York: Cambridge University Press. Button, Graham (1990). On varieties of closings. In George Psathas (ed.), Interaction competence, 93-148. Lanham, MD: University Press of America. Chafe, Wallace (1968). Idiomaticity as an anomaly in the Chomskyan paradigm. Foundations ofLanguage 4:109-25. Drew, Paul, & Heritage, John (1992), eds. Talk at work: Interaction in institutional settings. Cambridge & New York: Cambridge University Press. , & Holt, Elizabeth (1988). Complainable matters: The use of idiomatic expressions in making complaints. Social Problems 35:398-417. Estill, Robert B., & Kemper, Susan (1982). Interpreting idioms.yoarna/ofPsycholinguistic Research 11:559-68. Everaert, Martin; Van der Linden, Erik-Tan: S c h e n V Anrlrf

I n extract 13, A n n e is countering a neighbour's complaint about various kinds of household noise, including that from h e r television set. T h e object of the v e r b moan, however, is 'tjust gv'ryday thimgs' (line 6). T h e television noise is b e i n g used as a n instance of t h e category 'everyday things', a n d A n n e nicely formulates it as b o t h routine a n d reciprocal: 'she e n h e a r m'telly, T I ' E A R ' E R itelly' (lines 5-6). T h e sense is conveyed, therefore, of a neighbour w h o is dis posed to complain about non-complainables (in the s a m e sequence A n n e also mentions complaints about footsteps o n t h e stairs). Extracts 14 a n d 15 contain the s a m e vague moan-object: 'moa:nin' about somethin' Tglse' (extract 14, line 6), a n d 'J/mfianing about som'ing ejse I can't i 1 J /—J._„„J. I K i;„ K_frt 'Snmethinsr e l s e ' is a n i c e l y chosen Q t !

EDWARDS

MOANING, WHINGING AND LAUGHING

177

adds to that sense. Again, the impression is of an essentially intransitive kind of moaning, an activity d o n e by, a n d indexical of, the complainer, rather t h a n something caused b y a specifically complainable event or circumstance. T h e phrase 'start mo_a:nin' (extract 14, lines 5-6) works in the same w a y as ' m o a n i n g on', implying a sustained activity b e i n g embarked u p o n , given whatever pretext, b y the complainer. T h r o u g h the choice a n d application of lexical descriptions, then, speakers can characterize the activity of complaining in ways that either support or u n d e r m i n e its causation, constructiveness, a n d legitimacy. M u c h of the everyday work d o n e in countering complaints works u p their subjective, ir rational, dispositional causation. 1

'

i

i

Discussion

T h e idea that m a k i n g a complaint m a y require the m a n a g e m e n t of speakerindexical matters such as motive, disposition, or subjective investment, was raised by H a r v e y Sacks with regard to direct complaints, in a consideration of w h a t next turns might d o . ... if y o u say 'You interrupted m e ' a n d I say 'You're always complaining', t h e n the n e x t m o v e might turn o n a discussion, n o t of m y interruption, b u t of your complaining ... A n d if y o u d o n ' t want your complaining to b e the topic, then y o u m a y h a v e to avoid making things which are formulatable as 'complaints'. A n d in that regard, if s o m e b o d y does something routinely which is capable of being complained about, a n d y o u complain about it each time, you're in a good position to b e treated as complaining. (Sacks, 1992: vol. 1, 637-8) Sacks pointed to two features that we have also focused o n here: the speakerindexical dangers of complaining, a n d the w a y i n w h i c h r e c u r r e n t actions work in practical reasoning as a basis for ascribing dispositions ('you complain about it each time, y o u ' r e in a good position to b e treated as complaining': see Edwards, 1995). Sacks's idea that 'complaining' might therefore b e something to avoid being characterized as doing, raises the issue of the analytical status, a n d formal identifiability, of the speech category 'complaint'. Complaining will sometimes b e overt a n d obvious, but it is also likely to b e a subde business in which speaker/ complainers work against the indexical category of dispositional m o a n e r , while providing object-descriptions that permit the complainable nature of those objects to be h e a r d a n d taken u p b y recipients. It is a feature of the ironic a n d jocular delivery of complaints examined in this study, a n d of the choice of displaced objects, t h a t s p e a k e r / c o m p l a i n e r s m a n a g e to a c c o m p l i s h s o m e ' s e r i o u s ' complaining within an ironic delivery. T h e tendency of recipients to acknowledge the seriousness of a n ironically delivered complaint, even w h e n laughing along w i t h it m n v Ko in

-~— —rj.1—

1

'

*

178

DISCOURSE STUDIES

a n d displacement bring off. T h e r e is a tendency in m a k i n g complaints, as with invitation refusals and 'dispreferred' actions' generally, to project oneself as doing it reluctandy, or only t h r o u g h necessity. O n e feature of indirect complaints is that they are often performed at some distance from the offence itself. T h e y are often reports of past events, past complainables. This gives rise to a kind of duality between the situation pertaining w h e n the offence took place, a n d the current situation of talking about it, in which the initial complaint has to b e re-animated. Subjectivity announcements, a n d the detailed activity of complaining, attend to this notion of what kind of grievance it is/was, h o w the complainer is n o w dealing with it, and h o w the cur rent complaint-recipient should hear it as relevant to them, h e r e a n d now. Past complainables seem to b e routinely d o n e ironically, whether humorously or as something n o w coped and dealt with. This is reminiscent of a thousand holiday, family a n d peer group disaster stories, everyday 'war stories' in which all kinds of things that went w r o n g get to b e the stuff of tall tales, funny stories, and accounts of triumphing over adversity a n d seeing the funny side of it all. T h e r e is a surprising frequency in the data used for this study, of some kind of ironic take b e i n g d o n e . T h e laughter a n d irony are probably there, n o t because people find their misfortunes all that funny, but rather, as recipient-designed manage m e n t of attitude, coping, a n d (as Sacks anticipated) general avoidance of being seen as a complainer. I r o n y a n d h u m o u r , in enacting a complaint, are m e m b e r s ' m e t h o d s for dis playing a n d m a n a g i n g the speaker/complainer's stance or attitude, such that the complaint recipient is n o t placed in the potentially awkward position of having to w o r k out h o w to respond. But h u m o u r is also d o n e as part of the act of com plaining, n o t m e r e l y as a contrast to it, n o r to suggest that the complaint is weak or unfounded. It can even b e part of getting a complaint taken seriously, precisely b y signalling that the complainer is n o t disposed to m a k e too m u c h of it. This m a y b e a basis for serious uptakes of h u m o r o u s deliveries, where the h u m o u r is n o t taken to b e inviting the hearer to laugh, n o r showing that it is nothing serious. Rather, it is doing w h a t Sacks suggested might b e necessary, avoiding finding yourself 'in a g o o d position to b e treated as complaining'. Institutional relevancies c o m e into play. I n the n e i g h b o u r h o o d mediation data, which figured briefly in the final section of the analysis, complaining is n o t only frequent, and of endemic relevance to the setting, but there is also an orientation to the setting, to the n e e d for mediation or police action (cf. Meehan, 1989), and to the role of mediator as a questioning, sympathetic, b u t non-partisan complaint recipient. A complaint in such a setting also has to b e m a d e severe e n o u g h to b e worth resorting to that kind of setting, requiring those kinds of resources. This is reminiscent of couple counselling, in which couples work up the extremity and intractabiHty of their mutual complaints during initial troublestelling (Edwards, 2000b), in response to the question of w h a t brings t h e m there. I n the m u n d a n e telephone calls, that provided the bulk of materials for this study, a complaint can b e something m u c h m o r e casual and underplayed, and a* oil *Via L-inrl o f tViino r o n i i i r i n c r n r n r i p n t t n c r tct formal mediation DroCedllTeS. 3

1

EDWARDS

MOANING, WHINGING A N D LAUGHING

179

they are the site for complaints to have to b e broached, worked u p , and performed 'from scratch', w h e r e there is n o institutional or pre-established relevancy or requirement for it, or for its severity or seriousness. As D r e w notes, 'we do not complain to just anyone: we choose w h o to complain to a n d what kinds of complaints might appropriately b e m a d e to which kinds of recipients' (1998: 323-4). T h a t kind of situated, recipient orientation remains a topic in n e e d of investigation. T h e p h e n o m e n o n of 'displacement' m a y b e a w a y of avoiding talking about personal feelings o n sensitive matters, such as h o w Lesley or J e a n - C l a u d e felt, how hurt they are, h o w vulnerable it feels to h a v e someone b r e a k into (or walk out of) your h o m e like that. Recall that Lesley's ' m u d d y cushions' complaint (extracts 4 and 8) was p r e c e d e d b y a delicate inquiry b y Moira, into something Moira did not n a m e , a n d that Lesley would p e r h a p s 'rather not talk about'. Psychological explanations might b e offered for avoiding sensitive a n d painful matters, including the Freudian, psychodynamic possibility, w h e r e the term 'displacement' is also used, of some kind of avoidance or repression of deep, painful feelings. H o w e v e r plausible such explanations m a y be, the discursivepsychological analysis offered here is different, in three ways: (1) we are examining it as interaction; (2) it makes sense in interactional terms; a n d (3) the psychology is presumptive, a n d m a y b e a case of providing a psychodynamic or mentalistic explanation for s o m e t h i n g w h o s e n a t u r e is to b e found in the observable dynamics of social interaction (Billig, 1999; Edwards and Potter, 1992; Wooffitt, 1992). It is n o t so m u c h a matter of disproving psychodynamic explanations as obviating them, or at least of analysing social interaction in its own performative, occasioned, a n d sequentially organized terms. T h e major focus of this study has b e e n the subjective side of complaints as a performed a n d m a n a g e d feature of h o w they are m a d e . I n m a k i n g a n d report ing complaints, speakers display some kind of stance or attitude, b o t h to the complaint-object a n d to the current action of telling. Complainers w o r k to ob jectify complaints (Pomerantz, 1986), and also use displays a n d formulations of emotional investment, or ironic stance, in ways that m a y either enhance or distract from a complaint's seriousness a n d consequentiality. It is not only a matter of softening complaints b y playing t h e m down or laughing through them, nor a matter of choosing between irony a n d seriousness. Rather, it is a matter of how complaining is effectively d o n e , in which speaker-indexical features, and potential recipient uptakes, h a v e to b e managed.

Notes 1. 'Safe' compliments and complaints are those that have no negative implications for other people present, who might be characterizable under the same membership category used for the compliment/complaint. 2. Edwards (1997, 1999) examines the uses of presupposed and 'given' information in factual rhetoric; Sacks notes specifically with reeard to r n m n l a W c riw '~

180

DISCOURSE STUDIES

3. Identifying something as a complaint may depend on contextual features of the talk: 'routinely a piece of praise plus "but" plus something else, tells you that the something else is a "complaint," where it isn't obvious, often, that it is a complaint. That is to say, isolated, it wouldn't be a complaint' (Sacks, 1992: 359). 4. 'Overdetermined' descriptions are those where instead of, say, 'I had a drink', the speaker reports that she 'took the drink and drank out of it' (Drew, 1998: 318). 5. Drew (1998: 311) also notes that 'reporting... their emotional response - their sense of grievance - enables complainants to characterize how far the other's behavior has caused offense.' This echoes discussions of the everyday rhetoric of emotion talk, where emotional reactions may be opposed to rational judgements, making an emotiondriven report unreliable; but they can also be used to signal the spontaneity and genu ineness of whatever version of events they are attached to (Edwards, 1997, 1999). 6. A jumble sale is an informal, occasional, usually charitable sale of cheap bric-a-brac that people donate to raise cash for some kind of community project. 7. Characterizing a complaint as expectable from the complainant, i.e. as dispositional, can be a way of undermining its factual basis or seriousness. See Edwards (1995, 1997) on the relations between 'script formulations' and dispositional characterizations of actors, and the relation of that to factual reporting. 8. The specific placing of the laughter may signal sensitivity to this then-versus-now 'risk' of indirect complaints. Examples include: 'I thought well I'm gon' tell Jayce that, ehhheh heh-heh he-e.uh: T.eh.eh t.hhhhh' (extract 6); 'I nearly said sh'd 'ee've asked your p(h)erm(h}issio(h)on' (extract 7); and 'you couldn't'v: phoned't a better Tti:me,hheh Theh' (extract 1). 9. Something you 'must say' contrasts with something you might want to say. Lesley is complaining about fingerprint dust left by the police. She is complaining while im plying reluctance to do so, or an orientation to the delicacy of doing so. Again, this is a small token of subjectivity or 'attitude' management, of managing speaker-indexical implications of saying something. 10. One feature that can make a complaint a candidate 'moan' is the absence of orien tation to constructive next actions or solutions. Meehan notes, in a study of telephone calls to the police, how 'complaints' (in the sense of reported disturbances, offences, etc.) are normatively oriented to subsequent police action: 'they will try to formulate events in police-worthy form' (1989: 117). 11. Membership Categorization Device: see, for example, Hester and Eglin (1997), Sacks (1992). 12. Thanks to Elizabeth Stokoe for making some neighbourhood mediation data avail able. Some of the discussion in this article, on the nature of complaints and what they are designed to manage, was helped by examining a wide range of materials that it is impossible to include within the scope of a single article. 13. 'Dispreferred actions' in conversation analysis are actions such as refusing invitations and declining offers, where speakers orient in various systematic ways to the nor mative expectation that such things will be accepted (Atkinson and Heritage, 1984).

References Aarkhus, M. and Aldrich, A. (2002) 'Crafting Communication Activity: Understanding

EDWARDS

MOANING, WHINGING A N D LAUGHING

181

Antaki, C. and Wetherell, M. (1999) 'Show Concessions', Discourse Studies 1: 7-27. Atkinson, J.M. and Heritage, J. (eds) (1984) Structures of Social Action: Studies in Conversation Analysis, pp. 1-15. Cambridge: Cambridge University Press. Austin, J.L. (1962) How to Do Things with Words. Oxford: Clarendon Press. Billig, M. (1999) Freudian Repression: Conversation Creating the Unconscious. Cambridge: Cambridge University Press. Dersley, I. and Wootton, A. (2000) 'Complaint Sequences within Antagonistic Argument', Research on Language and Social Interaction 33: 375-106. Drew, P. (1987) 'Po-faced Receipts of Teases', Linguistics 25: 219-53. Drew, P. (1998) 'Complaints about Transgressions and Misconduct', Research on Language and Social Interaction 31:295-325. Edwards, D. (1994) 'Script Formulations: A Study of Event Descriptions in Conversation', Journal ofLanguage and Social Psychology 13: 211-47. Edwards, D. (1995) 'Two to Tango: Script Formulations, Dispositions, and Rhetorical Symmetry in Relationship Troubles Talk', Research on Language and Social Interaction 28: 319-50. Edwards, D. (1997) Discourse and Cognition. London: Sage. Edwards, D. (1999) 'Emotion Discourse', Culture and Psychology 5: 271-91. Edwards, D. (2000a) 'Extreme Case Formulations: Softeners, Investment, and Doing Nonliteral', Research on Language and SocialInteraction 33:347-73. Edwards, D. (2000b) 'Couples, Conflict and Counselling', paper presented at the 7th International Pragmatics Conference, Budapest, 9-14 July. Edwards, D. and Potter, J. (1992) Discursive Psychology. London: Sage. Edwards, D. and Potter,J. (2005) 'Discursive Psychology, Mental States and Descriptions', in H. te Molder andJ. Potter (eds) Conversation and Cognition, pp. 241-59. Cambridge: Cambridge University Press. Geertz, C. (1988) Works and Lives: The Anthropologist as Author. Oxford: Polity. Glenn, P.J. (1995) 'Laughing At and Laughing With: Negotiations of Participant Alignments through Conversational Laughter', in P. ten Have and G. Psathas (eds) Situated Order: Studies in the Social Organization of Talk and Embodied Activities, pp. 43-56. Washington, DC: University Press of America. Goffman, E. (1979) 'Footing', Semiotica25: 1-29. Haakana, M. (2001) 'Laughter as a Patient's Resource: Dealing with Delicate Aspects of Medical Interaction', Text 21: 187-219. Hester, S. and Eglin, P. (eds) (1997) Culture in Action: Studies in Membership Categorization Analysis. Washington, DC: University Press of America. Jefferson, G. (1984) 'On the Organization of Laughter in Talk about Troubles', in J.M. Atkinson and J. Heritage (eds) Structures ofSocial Action: Studies in Conversation Analysis, pp. 346-69. Cambridge: Cambridge University Press. Jefferson, G., Sacks, H. and Schegloff, E.A. (1987) 'Notes on Laughter in Pursuit of Intimacy', in G. Button andJ.R.E. Lee (eds) Talk and Social Organization, pp. 152-205. Clevedon: Multilingual Matters. Lakoff, G. (1987) Women, Fire and Dangerous Things: What Categories Reveal about the Mind Chicago, IL: University of Chicago Press. Levinson, S.C. (1979) 'Activity Types and Language', Linguistics 17: 356-99. Levinson, S.C. (1983) Pragmatics. Cambridge: Cambridge University Press. Meehan, AJ. (1989) 'Assessing the "Police-worthiness" of Citizens' Complaints to the Police: Accountability and the Negotiation of "Facts"', in D.T. Helm, W.T. Anderson. A.I.

Mooli-."

A W

'

> * "»

-

182

DISCOURSE STUDIES

Pomerantz, A.M. (1984) 'Agreeing and Disagreeing with Assessments: Some Features of Preferred/Dispreferred Turn Shapes', inJ.M. Atkinson andJ. Heritage (eds) Structures of Social Action: Studies in Conversation Analysis, pp. 57-101. Cambridge: Cambridge University Press. Pomerantz, A. (1986) 'Extreme Case Formulations: A Way of Legitimizing Claims', Human Studies 9: 219-29. Potter, J. (1996) Representing Reality: Discourse, Rhetoric, and Social Construction. London: Sage. Potter,J. and Hepburn, A. (2003) '"I'm a Bit Concerned": Early Actions and Psychological Constructions in a Child Protection Helpline', Research on Language and Social Inter action 36: 197-240. Sacks, H. (1992) Lectures on Conversation, vols 1 and 2, ed. G.Jefferson. Oxford: Basil Blackwell. Schegloff, E.A. (1988) 'Presequences and Indirection: Applying Speech Act Theory to Ordinary Conversation', Journal ofPragmatics 12:55-62. SearleJ.R. (1969) Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Stokoe, E.H. (2003) 'Mothers, Single Women and Sluts: Gender, Morality and Member ship Categorisation in Neighbour Disputes', Feminism and Psychology 13: 317-44. Whalen, M.R. and Zimmerman, D.H. (1990) 'Describing Trouble: Practical Epistemology in Citizen Calls to the Police', Language in Society 19: 465-92. Woofiitt, R.C. (1992) Telling Tales of the Unexpected: The Organization of Factual Discourse. London: Harvester Wheatsheaf.

23 Trial Discourse and Judicial Decision-Making: Constraining the Boundaries of Gendered Identities Susan Ehrlich

Introduction

R

ecent formulations of the relationship between language and gender, following Butler (1990), h a v e emphasized the performative aspect of gender. U n d e r this account, language is one important m e a n s b y which gender - a n ongoing social process - is enacted or constituted; g e n d e r is some thing individuals do - in part through linguistic choices - as opposed to something individuals are or have (West a n d Z i m m e r m a n , 1987). While the theorizing of gender as 'performative' has encouraged language a n d gender researchers to focus o n the agency a n d creativity of social actors in the constitution of gender, to m y m i n d there has b e e n less emphasis placed o n another aspect of Butler's framework - the 'rigid regulatory frame' (Butler, 1990) within which gendered identities are p r o d u c e d - that is, the limits a n d constraints o n speakers' agency in constructing such identities. This emphasis o n the 'performative' aspect of Butler's work, rather than o n h e r discussions of the regulatory n o r m s that define and police normative constructions of gender, m a y arise because, as C a m e r o n (1997) suggests, philosophical accounts of Buder's 'rigid regulatory frame' often remain very abstract. For C a m e r o n (1997:31), too often in feminist philosophical discussions, ' g e n d e r . . . floats free of the social contexts a n d activities in which it will always b e . . . e m b e d d e d , ' obscuring the fact that the routine enactment of gender is often, p e r h a p s always, subject to w h a t she calls the 'institutional coerciveness' of social situations. I n other words, cultural n o r m s (i.e. Buder's rigid regulatory frame) m a k e certain performances of gender seem appropriate

184

DISCOURSE STUDIES

a n d intelligible; in Buder's (p. 49) words, they 'congeal over time to p r o d u c e the a p p e a r a n c e . . . of a natural kind of being.' These same cultural n o r m s r e n d e r other performances of gender inappropriate a n d unintelligible a n d at times sub ject to social a n d physical penalties a n d sanctions (e.g. homophobia, gay-bashing, the 'fixing' of intersexed infants). By examining data from a public institution, a C a n a d i a n sexual assault trial, I demonstrate h o w culturally d o m i n a n t notions of male a n d female sexuality can impose constraints on the formation of participants' g e n d e r e d identities. Moreover, I show h o w the g e n d e r e d identities ascribed a n d assigned to individ uals can depart from the identities individuals intend to claim or adopt. Significant to a n investigation of 'talk' in trial contexts is the fact that inter pretations a n d understandings of the talk are discernable n o t only in the talk's local discourse (i.e. in the trial), b u t also in the non-local assessments a n d judg m e n t s of non-speaking recipients (i.e. juries a n d / o r judges). T h u s , while the C r o w n attorney (i.e. the lawyer representing the state) in the sexual assault trial 'talk' described later represented the complainant as resistingher perpetrator of sexual assault to the extent possible, judges at b o t h the trial level a n d the appeal level represented the complainant as participating in consensual sex with the accused. Put somewhat differently, the g e n d e r e d sense-making framework, or discourse, that the judges imposed o n the events in question departed quite dra matically from the g e n d e r e d sense-making framework, or discourse, invoked b y the C r o w n attorney in h e r questioning of the complainant. T h r o u g h her questioning of the complainant, the C r o w n attorney allowed a feminist under standing of sexual violence to emerge: one that acknowledged the unequal power dynamics that can shape a n d restrict w o m e n ' s behavior in the face of m e n ' s sexual violence. Yet such an understanding of the complainant's behavior failed to find its w a y into the judicial decisions of the trial j u d g e or the appeal judge. Indeed, the fact that it was not recognized in these judicial decisions is illustrative, I suggest, of the way that dominant, androcentric discourses (i.e. Butler's rigid regulatory frame) can r e n d e r certain performances of g e n d e r as unintelligible a n d incoherent. T h u s , this chapter considers the notion of 'the female voice' to the extent that it demonstrates the w a y a feminist perspective, w h e n manifest in a public context, can b e distorted or r e n d e r e d invisible b y the androcentric discourses that often dominate in these contexts. 1

T h e Participant Structure of Trial D i s c o u r s e I n describing trial practices, H a l e a n d Gibbons (1999:203) m a k e a useful distinc tion between 'two intersecting planes of reality' in the courtroom: the reality of the courtroom itself - what they call the 'courtroom reality' - and the reality that comprises the events u n d e r investigation in the courtroom - w h a t they call the 'external reality.' In the court's representation of this 'external reality,' visual images (e.g. photographs, diagrams) a n d physical entities (e.g. weapons, clothing) are "

• "

1

1

'-'•ii

/ m n n . ono\

„\

r

±u t n

EHRIICH

TRIAL DISCOURSE AND JUDICIAL DECISION-MAKING

185

is . . . through testimonial evidence which consists of descriptions of the events b y witnesses - versions of the second reality presented through language.' T h a t is, within the context of legal adjudication processes, language is the primary means b y which witnesses a n d lawyers convey information about the events that are the subject of a court's deliberations. A n d , while m u c h of the courtroom language representing this 'external reality' ostensibly occurs between lawyers and witnesses - that is, it is dyadic - the participant structure of the speech event of the trial is in fact m o r e complex than this. Given that the p r i m a r y target of courtroom interactions between lawyers a n d witnesses is a third-party, over hearing recipient - ajudge a n d / o r j u r y - trial talk is m o r e accurately characterized as multi-party (Drew, 1985; Cotterill, 2003). O n e w a y of conceptualizing this multi-party structure is b y appealing to Goffman's (1981) notion of the 'gathering.' According to Goffman (1981:136), 'interactional facts' h a v e to b e considered in relation to 'the full physical arena in which persons present are in sight and sound of o n e another' - what Goffman labeled a 'gathering.' Put somewhat differendy, Goffman (1981) argued that the two-person, face-to-face, s p e a k e r hearer m o d e l is too crude a construct to account for significant aspects of talkin-interaction given that speakers will alter h o w they speak a n d / o r what they say 'by virtue of conducting their talk in visual a n d aural range of non-participants' (Goffman, 1981:136). These 'non-participants,' according to Goffman, can include unratified recipients such as bystanders or eavesdroppers as well as ratified re cipients. Within the courtroom context, judges a n d juries constitute ratified recipients, a n d for Levinson (1988), w h o elaborates o n Goffman's framework, they are the 'indirect target' of trial talk. For m y purposes here, what is important is Goffman's recognition that partici pants w h o are n o t actively a n d directiy participating in a n interaction m a y n o n e theless influence m a n y of its properties. While Goffman talks expliciuy about how speakers m a y alter their linguistic identities u n d e r the influence of unratified indirect recipients of talk (e.g. bystanders), in this p a p e r I a m interested in h o w ratified indirect recipients of talk, that is judges in trial contexts, m a y influence and control the (gendered) meanings a n d interpretations assigned to speakers' linguistic identities. I shall argue that, in order to understand the complexity of gendered identities and, in particular, the way they can b e constrained b y cultural norms (i.e. Butler's rigid regulatory frame), we n e e d to b e attentive to the work of indirect recipients within a n interaction or, in Goffman's (1981: 136) words, to 'the full physical arena in which persons present are in sight a n d sound of one another.'

Data The data analyzed h e r e c o m e from a C a n a d i a n criminal trial involving sexual assault: H e r Majesty the Q u e e n v. E w a n c h u k , 1995. T h e case involved a sexual assault that took place during a j o b interview between the accused (also rpfiprrprl rn ac frlio AafanArtnA

A

1

~—— -

, / 1

r

1.

-

1

'

"

18(>

DISCOURSE STUDIES

T h e accused was a carpenter a n d wished to hire individuals w h o would sell his work for him. While the complainant suggested that h e r j o b interview b e held in a mall, the accused expressed a preference for m o r e privacy a n d p r o p o s e d instead that the interview take place in his van. T h e interview was conducted in a polite, business-like fashion, according to the complainant's testimony. D u r i n g the interview the complainant left the door of the v a n o p e n because she was hesitant about discussing the j o b offer in his vehicle. After the j o b interview, the accused invited the complainant to see some of his work in his trailer (i.e. caravan) which was attached to the van. Again, according to the complainant's testimony, she purposely left the trailer door o p e n out of fear of being alone with the accused in a confined space, b u t the accused ignored her efforts a n d closed a n d locked the door. T h e accused initiated a n u m b e r of incidents with the complainant that involved sexual touching, with each incident b e c o m i n g progressively m o r e intimate than the previous. T h e complainant said that she complied with m a n y of his requests out of fear that any resistance would p r o m p t the accused to be come m o r e violent. However, w h e n his touching progressed to the complainant's breast, she used her elbows to push him away and said 'no.' T h e accused resumed his sexual touching and began to massage the complainant's inner thigh and pelvic area, at which point the complainant again said ' n o . ' T h e accused r e s u m e d his advances b y grinding his pelvis into hers, touching h e r vaginal area a n d placing his penis o n the complainant's pelvic area u n d e r h e r shorts. H e stopped after the complainant said ' n o ' a third time. After these incidents, the accused opened the door of the trailer at the complainant's request a n d the complainant left the trailer. She later charged the accused with sexual assault. T h e trial j u d g e acquitted the accused of sexual assault relying on the defense of 'implied consent' and the Alberta Court of A p p e a l u p h e l d this acquittal. T h e S u p r e m e Court of C a n a d a overturned this acquittal and entered a conviction of sexual assault for the accused. I n what follows, I analyze data from the 1995 trial, the 1995 decision of the trial judge a n d the 1998 decision of the Alberta Court of Appeal. I n particular, I contrast two sets of data from this case: excerpts from direct recipients' talk, that is question-answer sequences between the Crown attorney and the complainant, and excerpts from indirect recipients' discourse, that is the j u d g e s ' decisions. Specifically, looking b e y o n d the face-to-face aspects of communication to the discourse of indirect recipients allows us to consider the regulatory n o r m s - that is, Buder's rigid regulatory frame - a n d the constraints they can impose o n identities.

Q u e s t i o n - A n s w e r S e q u e n c e s i n Trial D i s c o u r s e Like other types of institutional discourse, courtroom discourse has b e e n the subject of m u c h research over the past two decades - research that, a m o n g other things, has highlighted its asymmetrical character. A s others h a v e n o t e d about courtroom discourse (e.g. Atkinson a n d D r e w , 1979; Walker, 1987; Conley and

EHRL1CH


187

the right to initiate a n d allocate turns b y asking questions of witnesses b u t the reverse is n o t true. A n d , such differential participation rights, it has b e e n argued, bestows considerable conversational p o w e r a n d control u p o n the participant who is sanctioned to ask questions (Hutchby a n d Wooffitt, 1998). For example, in discussing doctor-patient interaction, D r e w a n d Heritage (1992) note that the question-answer pattern that characterizes most such interactions n o t only allows doctors to gather information from patients, b u t can also result in doctors dir ecting a n d controlling talk: introducing topics, changing topics, a n d selectively formulating and reformulating the terms in which patients' problems are expressed. Similar claims h a v e b e e n m a d e about lawyers in their role as questioners in the courtroom. However, given the adversarialnatare of courtroom discourse within the Anglo-American c o m m o n law system a n d the legally sanctioned p o w e r of lawyers to ask questions, lawyers m a y exercise even m o r e conversational control than doctors in their respective institutional settings. A d v e r s a r i a l dispute resolution, of w h i c h trials are a n o t a b l e e x a m p l e , requires that two parties c o m e together formally, usually with representation (e.g. lawyers), to present their (probably different) versions of the dispute to a third p a r t y (e.g. j u d g e , jury, tribunal) w h o hears the evidence, applies the ap propriate laws or regulations, and determines the guilt or innocence of the parties. Lawyers h a v e as their task, then, convincing the adjudicating b o d y that their (i.e. their client's) version of events is the most credible, or in C a p p s ' a n d O c h s ' (1995) t e r m s , ' t h e official story.' A p a r t from m a k i n g o p e n i n g a n d closing arguments, however, lawyers do not themselves testify. Thus, through the posing of questions, lawyers must elicit from witnesses testimony that will build a credible version of events in support of their o w n clients' interests in addition to testimony that will challenge, weaken, a n d / o r cast doubt o n the opposing parties' version of events. Atkinson a n d D r e w (1979: 70) n o t e that while trial discourse is con ducted p r e d o m i n a n d y through a series of question-answer sequences, other actions are accomplished in the form of such questions a n d answers. For example, questions m a y b e designed to accuse witnesses, to challenge or u n d e r m i n e the truth of w h a t they are saying, or, in direct examination, to presuppose the truth and adequacy of what they are saying. Direct Examination of the Complainant In contrast to the adversarial, combative nature of cross-examination, direct examination, that is the questioning of o n e ' s o w n witness, h a s b e e n char acterized b y b o t h legal practitioners a n d scholars as supportive a n d cooperative (e.g. W o o d b u r y , 1984; Barry, 1991; M a l e y , 1994). A c c o r d i n g to Cotterill (2003: 129), direct examination 'represents a n initial, dominant narrative state ment, which is then r e s p o n d e d to, challenged a n d sometimes subverted in crossexamination questioning.' Because the emphasis in direct examination is o n developing n e w information, o p e n - e n d e d questions t e n d to b e m o r e frequent in direct examination. Moreover, W o o d b u r y (1984: 211) suggests that open-

188

DISCOURSE STUDIES

third-paxty recipients (e.g. j u d g e a n d / o r jury) that lawyers are trusting of their witnesses. Indeed, Harris (2001: 68) argues that lawyers must exercise a high degree of control over all witnesses, but that the strategies required to d o this differ d e p e n d i n g o n whether the witness is 'friendly' or 'hostile.' Because t h e v a r y i n g v e r s i o n s of events that e m e r g e in trial discourse are determined to a large extent b y the questions that lawyers ask of witnesses (e.g. their confrolling of witnesses' topics, their selective reformulations of witnesses' prior answers, etc.), Cotterill (2003: 149) argues that courtroom narratives are best characterized as 'dual-authored texts,' 'with the emphasis o n the voice of the lawyer as the primary a n d authoritative teller.' D u e to the 'dual-authored' n a t u r e of c o u r t r o o m narratives, in what follows I analyze q u e s t i o n - a n s w e r sequences between the C r o w n attorney a n d the complainant using the analytic framework of Conversation Analysis, broadly conceived. Conversation Analysis is fundamentally concerned with the sequential analysis of utterances, that is ' h o w utterances are designed to tie with, or 'fit' to, prior utterances a n d h o w a n utterance has significant implications for w h a t kinds of utterances should come n e x t ' (Wooffitt, 2 0 0 1 : 54). I a d o p t this approach to question-answer sequences in the direct examination of the complainant because the nature of the questions asked b y the C r o w n attorney has significant implications for what kinds of u t t e r a n c e s c o m e next, specifically the k i n d of t e s t i m o n y t h e c o m p l a i n a n t produces a n d the kind of discursive identity she projects. I n the direct examination of the complainant in the Ewanchuk case, the C r o w n attorney typically b e g a n b y asking a b r o a d Wh-question (e.g. ' W h a t h a p p e n e d then?'), that is a Wh-question that allowed the complainant to construct a n extended narrative. I n response to this b r o a d Wh-question, the complainant would provide an answer that described a n event or a series of events. I m m e d i ately following such an answer, the lawyer would ask a m o r e narrow Wh-question, that is a Why-question that attempted to elicit m o r e specific information: the plaintiff's motivation for performing a particular action that she h a d described. For example, in (1) below, the C r o w n attorney begins b y asking a b r o a d Whq u e s t i o n a b o u t t h e e v e n t s t h a t t r a n s p i r e d o n c e t h e plaintiff r e a c h e d the defendant's van. T h e plaintiff's answer is followed b y a m o r e narrow Wh-question - a Why-question inquiring about the plaintiff's motivation for suggesting that she a n d the defendant talk inside the mall, as o p p o s e d to in his van. W h a t emerges in the answer to the Why-question is the complainant's desire to talk to the defendant in a public place. 2

Oj Was he inside the van or trailer when you first got there? A: I believe he was inside the van, but - he might have stepped out to meet me. Qj What happened once you got there? A: I asked him if we could go inside the mall, have a cup of coffee and talk about whatever. —» Qj Why did you want to go inside the mall to talk? A: Because it was - it was a public place. I mean, we could go in and sit down

EHRLICH


189

According to Woodbury (1984:211), the narrow Wh-questions that follow broad Wh-questions in direct examination serve a narrative function. Since it is im portant that lawyers and witnesses co-construct 'a coherent and maximally detailed account' for the sake of third-party recipients, Woodbury maintains that narrow Wh-questions allow witnesses to elaborate on details that contribute to the coherence of the narrative. But, what is it that determines the kind of details that will make a narrative cohere? Why, for example, does the Crown attorney in the Ewanchuk case follow up her broad Wh-questions with Whyquestions that probe the plaintiff's reasons for performing certain actions in the series of events under question? It is m y contention that the narratives elicited from the complainant through the Crown attorney's questions cohered as feminist narratives. That is, the Crown attorney allowed a feminist understanding of sexual violence to emerge in the complainant's testimony. In particular, it became clear through the questionanswer sequences of the Crown attorney and the complainant that the threat of men's sexual violence can create an asymmetrical power relationship between men and w o m e n which, in turn, can shape and restrict women's efforts to resist their perpetrators. Consider examples (l)-(5) below which all come from the direct examination of the complainant (I have repeated example (1) below), and all contain a Why-question from the Crown attorney that asks why the complainant has performed a particular action. The answers elicited by these questions (italicized below) reveal the strategic nature of the complainant's actions; that is, the particular actions the Crown attorney asks about are re presented b y the complainant as ways she attempted to discourage the accused's sexual advances. Example 1 Oj Was he inside the van or trailer when you first got there? A: I believe he was inside the van, but - he might have stepped out to meet me. Oj What happened once you got there? A: I asked him if we could go inside the mall, have a cup of coffee and talk about whatever. —» Q: Why did you want to go inside the mall to talk? A: Because it was - it was a publicplace. I mean, we could go in and sit down somewhere and talk. Example 2 Oj What happened then? A: He said, Why don't we just talk inside the van here. And he sat into his driver's seat, and I opened the door, and I left the door open of the passenger seat and I sat down there. —» Oj And why did you leave the door open? A: Because I was still very hesitant about talking to him. Example 3

DISCOURSE STUDIES

190

A: He went around t o - n o , first, he said, Okay, I'd like to pull the van into the shade. It was a hot day, and there was cars that were parked under the shade for an shade of a tree, I believe, and he got out, and he went and he stepped inside, and he said, Come on up and look. So I stepped up inside, took about two steps in, I didn't, like, walk around in it. And then he went to the door, closed it, and locked it. (some intervening turns) Oj Had you expected him to lock the door? A: Not at all. I left the door completely wide open when I walked in there for a reason. —» Qj And what was that reason? A: Because Ifelt that this was a situation that I shouldn't be in, that I- with anybody to be alone in a trailer with any guy with the door closed. Example 4 Oj Did you talk about other things while you were sitting in front of the van? A: Yes. He asked - we talked a little bit more on a personal level. Oj What do you mean by that? A: I believe 1 told him that I was living on my own. Well, not totally on my own. There was about three other people living there and that I had a boyfriend. • —• Oj Why would you tell him those things? A: Because Ifelt that he should know because I just -1 felt -1 felt that he might feel a little more threatened if I had said that. Example 5 Oj What happened after you talked about your personalities? A: We were still mentioning a lot of personal things. l i k e I was still mentioning that I had a boyfriend. I believe I said his name. Oj What was his name? A: His name was Allan. (some intervening turns) —¥ Q, Why were you mentioning your boyfriend Allan? A: Because, like I said, I felt like ifhe ever - if- it mightprevent himfromgoing beyond any more touching. In example (1), when asked why she suggested going inside the mall, the com plainant explains that it was a public place. (Presumably, sexual advances are less likely to occur in public places.) In examples (2) and (3), when asked why she left the doors open to the front of the van and the trailer, respectively, the complainant explains that she was 'hesitant' about talking to the accused alone 'with the door closed.' In examples (4) and (5), w h e n asked w h y she was mentioning her boyfriend, Allan, the complainant explains that she wanted the accused to feel 'threatened' and that she wanted to 'prevent him from going beyond any more touching.' Clearly, what the Crown attorney succeeds in eliciting in asking these particular Why-questions is a sense that the complainant is not passive, but rather is actively attempting to create circumstances that will -T

'I -

J>

1

I

EHRLICH


191

Examples (6)-(9) are somewhat different from (l)-(5), as the Crown attorney does not ask questions about actions intended to discourage and/or prevent the accused's sexual advances, but rather asks about actions that could be construed as preambles to consensual sex. Example 6 Oj Did he say anything when he locked the door? A: He didn't say anything about the door being locked, but he asked me to sit down. And he sat down cross-legged. Oj What did you sit on? A: Just the floor of the trailer. —> Oj Now, why did you sit down when he asked you to sit down? A: Because IfiguredI was in this trailer, the door was locked, he was not much more than this stand is away from me here, probably only a couple of feet awayfromme. I felt that I was in a situation now where I just better do what I was told. Example 7 Qj And what happened then? A: He told me that he felt very tense and that he would like to have a massage, and he then leaned up against me with his back towards me and told me to rub his shoulders and I did that - Q: And up to the time he told you he was tense and wanted a massage, had the two of you talked about you giving him a massage? A: I believe all he had said right before that is that he liked to have them, and he was tense feeling and that was all. Qj Had you ever offered to give him a massage? A: No. Q: Did you want to give him a massage? ' A: No. (some intervening turns) —» Q: If you didn't want to give him a massage at that point in time, why did you touch his shoulders? A: / was afraid that if Iput up any more of a struggle that it would only egg him on even more, and his touching would be more forced. 1

Example 8 Q: And what happened then? A: Then he asked me to turn around the other way to face him, and he said he would like to touch my feet or he would like to massage my feet, so I did. And he was just touching my feet. Q: Did you want him to massage your feet? A: No. —> Q: Why did you turn around? A: Because I guess I was afraid. I wasfrozen.Ijust did what he told me to do. Q: Did he ever ask you if you would like him to massage his feet - your feet? A: No, he just said, Turn around I'm going to. Q: What happened after you turned around? A: He was massaging my feet, but he didn't stay there. He was moving up my

DISCOURSE STUDIES

192

again, and then he'd move back up again, and I just sat there, and I didn't -1 didn't do anything. I didn't say anything. I knew something was going to happen, and Ididn't want tofight.Ididn't want to struggle. Ididn't want to scream, because Ifelt that that wouldjust egg him on more. Example 9 Oj And what happened when he reached to hug you? A: He just did, and I, at this time, I was trying really hard not to cry. I had been wiping my eyes when he was on top of me when he couldn't see me, and I j u s t . . . . I just responded by just lightly putting my arm on him when he hugged me because I was afraid that he would think I was really scared, and that 1 would leave there telling people. —> Oj And why were you worried about him thinking that? A: Because I didn't think that he would stop there, that it would get worse, and it would be more brutal. Put somewhat differentiy, in examples (6)-(9), the C r o w n attorney asks w h y the complainant complies with the accused's requests: in (6), w h y she sits d o w n w h e n asked; in (7), w h y she begins to massage the accused w h e n asked; in (8), w h y she turns a r o u n d to face the accused w h e n h e asks to massage h e r feet; and, in (9), w h y h e r fear of the accused leads h e r to reciprocate his hug. I n re sponse to these questions, the complainant says a variety of things: that she was afraid; that she felt she should d o what she was told; that she feared if she did not comply with the accused's requests or if she put u p a struggle that she would 'egg h i m o n even m o r e , ' 'his touching would b e m o r e forced,' a n d 'it would b e m o r e brutal.' Indeed, such responses reflect strategies that m a n y victims of sexual violence e m p l o y to prevent m o r e prolonged a n d extreme instances of violence. A s researchers o n violence against w o m e n h a v e asserted, submitting to coerced sex or physical abuse can b e ' a strategic m o d e of action undertaken in preser vation of self (Lempert, 1996: 281). T h a t is, if physical resistance o n the part of victims can escalate a n d intensify violence, as some research shows (e.g. D o b a s h a n d Dobash, 1992) a n d m a n y w o m e n (are instructed to) believe, then submission to coerced sex is undoubtedly the best strategy for survival. Significant about the C r o w n attorney's questioning in (6)-(9) is the fact that her questions allow the complainant's actions to b e revealed as strategies of resistance, rather t h a n as precursors to consensual sex. I n fact, all of the C r o w n ' s why-questions high lighted in examples (l)-(9) function to elicit responses that emphasize the com plainant's active deployment of strategies m e a n t to discourage a n d resist the accused's escalating sexual violence. Put another way, the complainant's actions w e r e contextualized within a framework, or discourse, that acknowledged the potential structural inequalities that characterize male-female sexual relations a n d the effects of such inequalities o n w o m e n ' s strategies of resistance. We will see in the n e x t section that, b y contrast, the trial judge a n d the appeal j u d g e contextualized the complainant's actions within an alternative framework or discourse - o n e that erased the unequal p o w e r dynamics of male-female sexual relations and. concomitantly, construed coerced sex as consensual sex. 3

EHRTJCH

193


Judges' Discourse As stated in a previous section, the trial judge's decision to acquit the accused in this case was appealed to the Alberta Court of Appeal, where the acquittal was upheld, a n d then to the Supreme C o u r t of C a n a d a , w h e r e the acquittal was over turned. I n this section, I discuss b o t h the decision of the trial j u d g e a n d the Alberta Court of A p p e a l judge, including the defense b o t h judges invoked in acquitting the accused - 'the defense of implied consent.' T h e first noteworthy aspect of b o t h of these decisions, given that the C r o w n attorney a n d the complainant depicted the complainant as attempting to discour age a n d resist the accused, was the consistent representation of sexual relations b e t w e e n the accused a n d the complainant as consensual sex. Consider the following excerpt from the trial judge's decision, at a point w h e n the j u d g e is describing the 'facts' of the case: Example 70 1 B [the complainant] told A [the accused] that she was an open, 2 friendly and affectionate person; and that she often liked to touch 3 people. A told B that he was an open, friendly and affectionate 4 person; and that he often liked to touch people. A and B talked. 5 They touched each other. They hugged. They were sitting on the 6 floor of the trailer and they were lying on the floor of the trailer. 7 A told B that he would like a body massage, and B gave A a body 8 massage. For the body massage, A sat in front of B so that B could 9 massage A's back. They later exchanged places so that A could give 10 B a body massage. B later lay on her back, and A gave B a foot 11 massage. After the foot massage, A massaged B's bare legs and he 12 massaged her bare inner thighs. During this period of two and one 13 half hours, A did three things which B did not like. When A was 14 giving B a body massage, his hands got close to B's breasts. B said 15 "No", and A immediately stopped. When B and A were lying on 16 the floor, A rubbed his pelvic area against B's pelvic area. B said 17 "No", and A immediately stopped. Later on A took his soft penis 18 out of his shorts and placed it on the outside of B's clothes in her 19 pelvic area. B said "No", and A immediately stopped. 20 During all of the two and one half hours that A and B were 21 together, she never told A that she wanted to leave. When B 22 finally told A that she wanted to leave, she and A simply walked 23 out of the trailer. (from Reasons for Judgment (Moore, J., C.Q.B.A.), November 10, 1995) These same 'facts' are confirmed by the Alberta C o u r t of Appeal j u d g e in his support of the trial judge's 'doubts about consent.' Example 11 Yet, if review of the evidence

that s u n n n r t s

tho tH_i

inr*.™'-

j„,.u»- _ i — *

•

194

DISCOURSE STUDIES

said to be criminally assaultive were preceded by an exchange of consensual body massages, partially on the floor of the trailer, hugs and assurances of trust and restraint.... Beyond that (and somewhat inconsistent with an appellate profile of Ewanchuk as a relendess sexual predator) every advance he made to her stopped when she spoke against it. (from Reasons for Judgment of the Honourable Mr. Justice McClung, February 12, 1998) I n general, these descriptions of what transpired between the accused a n d the complainant - with the exception of the descriptions of the three times that the complainant said ' n o ' - emphasize the consensual, mutual a n d reciprocal nature of their sexual relations. I n d e e d , the appeal judge states quite explicitly that the hugs a n d the b o d y massages w e r e reciprocal (i.e. 'an exchange') a n d 'consensual' (in italics above). Particularly striking is the fact that b o t h judges represent as consensual a n d reciprocal events that the C r o w n attorney a n d the complainant depicted as coerced sex. T h a t is, o n m a n y occasions - a n d this is exemplified in examples (6)-(9) - the complainant said she complied with the accused's wishes out of fear that his violence would otherwise escalate. Yet, the judges represent these events as ones that the complainant engaged in with out coercion. For example, in line 5 of e x a m p l e (10), the trial j u d g e states that the c o m p l a i n a n t a n d t h e accused h u g g e d e a c h other in a reciprocal w a y (i.e. ' T h e y hugged') whereas in example (9) the complainant says that she re s p o n d e d to the accused's h u g out of fear, that is 'because [she] was afraid that h e would think [she] was really scared' which in turn would lead to even greater brutality. Likewise, in lines 7 - 8 of (10), the judge states simply that the com plainant gave the accused a b o d y massage in response to his request (i.e. 'A told B that h e would like a b o d y massage, and B gave A a b o d y massage'); yet in example (7) the complainant says that she agreed to massage the accused only because she 'was afraid that if [she] p u t u p a n y m o r e of a struggle that it would only egg h i m o n even m o r e a n d his touching would b e m o r e forced.' Finally, we see that in lines 11-12 of (10), the j u d g e states the accused massaged the c o m p l a i n a n t ' s b a r e legs a n d i n n e r thighs, as if this w e r e a consensual act (i.e. 'After the foot massage, A massaged B's b a r e legs a n d h e massaged h e r bare inner thighs'); b u t in example (8) the complainant says that she complied with the accused's massages because she did n o t w a n t to 'egg h i m o n m o r e . ' I n sum, despite the fact that the complainant conveyed in h e r direct testimony that she h a d little choice b u t to comply with the accused's sexual advances, the judges' decisions failed to qualify or modify the sexual advances in such a way. Indeed, there is a sense in b o t h decisions that the complainant's consent was freely given. By contrast, it is noteworthy that in the decision of the Supreme Court of Canada, there are m a n y descriptions of the 'facts' that acknowledge the coerced nature of the sexual contact between the accused a n d the complainant, for e x a m p l e 'At some point the accused said that h e was feeling tense a n d asked the complainant to give h i m a massage. T h e complainant complied, massaging the accused's shoulders for a few minutes.' ' T h e complainant did n o t want the accused to touch her in this way, b u t said nothing as she was afraid that a n y resistance

EHRLICH


195

I n previous w o r k o n the language of C a n a d i a n sexual assault trial judgments, Coates, Bavelas, a n d Gibson (1994) noted that judges recognized resistance o n the part of complainants only w h e n it took the form of persistent physical struggle. Coates etal. (1994:195) elaborate: 'The language of appropriate resistance seemed to us to b e drawn from m a l e - m a l e combat between equals, w h e r e continued fighting is appropriate, rather than from asymmetrical situations... where physical resistance would lead to lirae chance of success a n d a high probability of further h a r m . ' While the excerpts from the judges' decisions do n o t seem to d e e m physical struggle as the only appropriate form of resistance, they do seem to re quire that resistance at least take the form of verbal refusal. For e x a m p l e , b a s e d on lines 14-19 of example (10), it seems that verbal refusals are the only indicators or signals that the trial j u d g e recognizes as resistance o n the p a r t of the com plainant. T h a t is, although the C r o w n attorney is successful in eliciting testimony (for example, (l)-(9)) that depicts the complainant as attempting to discourage a n d resist the accused in a variety of ways - including submitting to coerced sex, the trial j u d g e appears to only perceive her resistance o n the three occasions that she said ' n o ' to the accused. The Defense of Implied Consent U n d e r C a n a d i a n criminal law, a n accused will b e found guilty of sexual assault if the C r o w n attorney proves b e y o n d a reasonable doubt that the complainant did not (or could not) consent to the acts. As stated earlier, the trial j u d g e did n o t find the accused guilty of sexual assault, but rather acquitted the accused, relying o n the defense of 'implied c o n s e n t ' T h e Alberta Court of Appeal upheld this acquittal a n d the defense. For b o t h of these judges, then, the complainant was considered to h a v e implied consent; moreover, the Alberta Court of Appeal judge defined implied consent as 'consent b y conduct.' O n e of the questions that arises from these judgements concerns the kind of'conduct' that these judges d e e m e d as signaling consent. Both the trial j u d g e a n d the Court of Appeal judge found that the complainant was a 'credible witness' (from the trial judge, Moore) and that she was genuinely afraid of the accused: 'Certainly the complainant was afraid of E w a n c h u k as the trial j u d g e found' (from the Court of Appeal judge, McClung). H o w e v e r , both judges also c o m m e n t e d in their decisions that she did n o t communicate her fear to the accused. Consider the follow excerpt from the trial judge's decision: Example 12 All of B's [the complainant's] thoughts, emotions and speculations were very real for her. However, she successfully kept all her thoughts, emotions, and speculations deep within herself. She did not communicate most of her thoughts, emotions and speculations Like a good actor, she projected an outer image that did not reflect her inner self. B did not communicate to A by words, gestures, or facial expres sions that she was'frozen'bya fear of force. B did not communicate that she wasfrozento the spot, and that fear prevented her from getting up off the floor and walkine out of

196

DISCOURSE STUDIES

T h e picture that emerges from this description of the complainant is that she was 'frozen to the spot' a n d 'that she was "frozen" b y a fear of force' (italicized above), yet she kept h e r emotions a n d feelings h i d d e n from the accused. A n d , o n the basis of this type of 'conduct,' the trial judge acquitted t h e accused, sug gesting that the complainant's conduct implied consent. Given that the accused would h a v e to draw inferences i n order to understand the complainant as 'im plying' consent, o n e wonders w h a t kind of cultural background knowledge these inferences w o u l d rely u p o n . I n other words, what kind of cultural assumptions might give rise to t h e inference that concealing emotions a n d being frozen to the spot conveys consent? W h a t b e c o m e s clear from questions such as these is the ideological nature of the cultural assumptions that the accused is understood to rely u p o n i n drawing such a n inference. T h a t is, b y ruling that a w o m a n w h o is emotionless a n d frozen to the spot is implying consent, the judges are invoking assumptions about women's lack of agency a n d passivity in the course of 'normal' heterosexual sex. As C a m e r o n and Kulick (2003:36) r e m a r k about such societal assumptions, 'the denial of sexual agency to w o m e n m e a n s that saying "yes" to sex (or initiating it) is disapproved of. Nice girls should d e m u r coyly i n order to demonstrate that they are n o t sluts or n y m p h o m a n i a c s , b u t this is a ritual, formulaic gesture a n d m e n should n o t b e deterred.' Clearly, the accused in H e r Majesty t h e Q u e e n v. E w a n c h u k w a s n o t d e t e r r e d b y t h e c o m p l a i n a n t ' s emotionless a n d frozen demeanor (indeed, not even b y h e r three verbal refusals); and, such behavior was authorized b y the trial judge a n d the Alberta Court of A p p e a l j u d g e presumably because of the weight of cultural beliefs that equate a w o m a n ' s sexual passivity with consent.

Conclusion Recent approaches to language a n d gender, while emphasizing the agency and creativity of speakers in constructing gendered identities, h a v e paid less attention to the normative discourses (i.e. Butler's rigid regulatory frame) that police and regulate the intelligibility of such identities. I n this paper, I h a v e suggested, fol lowing Goffman (1981), that a full understanding of identity construction requires looking b e y o n d the face-to-face aspects of interaction to what Goffman calls the 'gathering.' Put somewhat differendy, I have used a sexual assault trial to demon strate h o w participants who are not direcdy a n d actively involved in an interaction can nonetheless influence the meanings a n d understandings that are assigned to that interaction. M o r e specifically, I h a v e attempted to show h o w the identity imposed u p o n the complainant in this sexual assault case - as a participant in consensual sex - departed quite dramatically from h e r identity - as a victim of coerced sex - as it was co-constructed in h e r direct testimony. C a p p s a n d O c h s (1995: 21) argue that adjudicators i n legal cases never determine the truth of a case; rather, ' o n the basis of divergent versions of events, j u r y m e m b e r s [and adjudicators] construct a narrative that is plausible and •

'

•

T

, .

L

™ A - „

t

; „ n , l „ „ n „

0

„

n

f t n

EHRLICH


197

are undoubtedly m a n y factors contributing to the determination of an 'official story,' I a m suggesting h e r e that a crucial determining factor, especially in the context of a sexual assault trial, is the intelligibility - or lack thereof - of partici pants' 'performances' of gender. Indeed, in the judicial decisions analyzed so far, the version of events endorsed b y the trial j u d g e a n d the Court of A p p e a l judge relied u p o n a very particular understanding of gender and sexuality - one that viewed w o m e n ' s verbal refusals as necessary to resistance and equated w o m e n ' s lack of physical responsiveness with consent. As demonstrated before, the C r o w n attorney a n d complainant invoked an alternative, feminist sensemaking framework in this trial - one that viewed submission to coerced sex as a w a y for the complainant to resist m o r e extreme a n d prolonged instances of violence. Yet, this alternative way of assigning m e a n i n g to the events in question did not seem to resonate with the sense-making frameworks of the trial j u d g e and the Alberta Court of Appeal judge. McConnell-Ginet (1988,1989) has argued that counter-hegemonic viewpoints or discourses m a y encounter difficulty func tioning as background knowledge in linguistic exchanges that d o n o t take place among familiars (McConnell-Ginet, 1988:92). I n other words, because the Crown attorney's discursive strategy involved a feminist understanding of sexual assault, it h a d difficulty surviving as a contextualizing framework within a public institution, such as a trial. Rather, this trial - or at least crucial a n d defining aspects of it - relied o n dominant, androcentric background assumptions (i.e. Butier's rigid regulatory frame) to inform a n d constrain its interpretation of events.

Acknowledgements I thank Judith Baxter for valuable comments on a previous version of this paper. The research for this article was funded, in part, by a Regular Research Grant from the Social Sciences and Humanities Research Council of Canada, grant #410-2000-1330.

Notes 1. Within the Canadian criminal justice system, Crown attorneys represent the state and complainants assume the role of witnesses for the state. That is, complainants are not directly represented by Crown attorneys. 2. The arrows within the transcripts point to narrow Wh-questions asked by the Crown attorney. 3. In previous work on the language of sexual assault trials (Ehrlich, 2001), I have argued that defense lawyers strategically invoke the 'utmost resistance standard' in trial dis course as a way of undermining the credibility of complainants. Until the 1950s and 1960s in the United States, the statutory requirement of utmost resistance was a necessary criterion for the crime of rape (Estrich, 1987); that is, if a woman did not resist a man's sexual advances to the utmost then the rape was said not to have occur red. While the 'utmost resistance standard' is no longer encoded in legal statutes in the United States or Canada, the a d j u d i c a t i o n n f «.vi,*>i

DISCOURSE STUDIES

198

it is possible that the Crown attorney's questioning of the complainant in the Ewanchuk case anticipated, and was designed to counteract, such a strategy on the part of Ewanchuk's lawyer.

References Atkinson, J. M. and Drew, P. (1979) Order in Court. London: Macmillan Press. Barry, A. (1991) 'Narrative style and witness testimony', Journal of Narrative and Life History, 1: 281-293. Butler,J. (1990) Gender Trouble: Feminism and the Subversion of Identity. London: Routledge. Cameron, D. (1997) 'Theoretical debates in feminist linguistics: Questions of sex and gender', in R. Wodak (ed.), Gender and Discourse, pp. 21-36. London: Sage. Cameron, D. and Kulick, D. (2003) Language and Sexuality. Cambridge: Cambridge Uni versity Press. Capps, L. and Ochs, E. (1995) Constructing Panic: The Discourse of Agoraphobia. Cambridge, Massachusetts: Harvard University Press. Coates, L., Bavelas, J., and Gibson J. (1994) 'Anomalous language in sexual assault trial judgements', Discourse & Society, 5: 189-206. Conley, J. and O'Barr, W. (1998) Just Words: Law, Language and Power. Chicago: University of Chicago Press. Cotterill, J. (2003) Language and Power in Court: A Linguistic Analysis ofthe OJ. Simpson Trial. London: Palgrave Macmillan. Dobash, R. E. and Dobash, R. P. (1992) Women, Violence and Social Change. London: Routledge. Drew, P. (1985) 'Analyzing the use of language in courtroom interaction', in T. A. van Dijk (ed.), Handbook of Discourse Analysis (Volume 3), pp. 133-147. New York: Academic Press. Drew, P. and Heritage, J. (1992) 'Analyzing talk at work: An Introduction', in P. Drew and J. Heritage (eds), Talk at Work: Interaction in Institutional Settings, pp. 3-65. Cambridge: Cambridge University Press. Ehrlich, S. (2001) Representing Rape: Language and Sexual Consent. London: Routledge. Estrich, S. (1987) Real Rape. Cambridge, Massachusetts: Harvard University Press. Goffman, E. (1981) Forms of Talk. Philadelphia: University of Pennsylvania Press. Hale, S. and Gibbons, J. (1999) 'Varying realities: Patterned changes in the interpreter's representation of courtroom and external realities', Applied Linguistics, 20:203-220. Harris, S. (2001) 'Fragmented narratives and multiple tellers: Witness and defendant accounts in trials', Discourse Studies, 3: 53-74. Hutchby, I. and Wooffitt, R. (1998) Conversation Analysis. Oxford: Polity Press. Lempert, L. (1996) 'Women's strategies for survival: Developing agency in abusive relationships',yoaraai of Family Violence, 11: 269-289. Levinson, S. C. (1988) 'Putting linguistics on a proper footing: Explorations in Goffman's concepts of participation', in P. Drew and A. Wooton (eds), Erving Goffman: Explorations in the Interactional Order, pp. 161-228. Oxford: Polity Press. Maley, Y. (1994) 'The language of the law', in J. Gibbons (ed.), Language and the Law, pp. 11-50. London: Longman. McConnell-Ginet, S. (1988) 'Language and gender', in F. Newmeyer (ed.), Linguistics: The Cambridge Survey, Volume IV, pp. 75-99. Cambridge: Cambridge University Press.

EHRLICH


199

McConnell-Ginet, S. (1989) 'The sexual (re)production of meaning: A discoursebased theory', in F. Frank and P. A. Treichler (eds), Language, Gender, and Professional Writing, pp. 35-50. New York: Modern Language Association of America. Reasons for Judgment (Moore, J., C.Q.B.A.), November 10,1995, Queen v. Ewanchuk. Reasons forJudgment of the Honourable Mr. Justice McCIung, February 12,1998, Queen v. Ewanchuk. Walker, A. G. (1987) 'Linguistic manipulation, power and the legal setting', in L. Kedar (ed.), Power through Discourse, pp. 57-80. Norwood, New Jersey: Ablex. West, C. and Zimmerman, D. (1987) 'Doing gender', Gender and Society, 1: 25-51. Woodbury, H. (1984) 'The strategic use of questions in court', Semiotica, 48: 197-228. Wooffitt, R. (2001) 'Conversation analysis', in M. Wetherell, S. Taylor and S. Yates (eds), Discourse as Data, pp. 49-92. London: Sage Publications.

24 The Occasioning and Structure of Conversational Stories Susan M. Ervin-Tripp andAylin Kuntaf

1. O r i e n t a t i o n

W

hat do we m e a n b y a story? Recent usage has turned almost every thing, even the non-verbal, into a narrative. For example, Ochs and Capps (1996) h a v e a r e m a r k a b l y inclusive definition of narratives of personal experience as "verbalized, visualized, a n d / o r e m b o d i e d framings of a sequence of actual and possible life events." T h e absence of a c o m m o n guiding framework h a s led to the blossoming of criteria used b y different narrative researchers for identifying their respective units of study. W h e n o n e looks for personal experience stories in natural conversations, the most striking fact is that they are n o t always clearly recognizable b y traditional narrative-internal criteria such as the presence of a protagonist a n d events creating conflict, refer ence to events in the past, presence of a climactic complicating action, or closure of the storyline with a resolution. Further, the onset of conversational stories does not always clearly demarcate the narrative segment from the preceding talk. S o m e stories are explicitiy introduced into the ongoing conversation with a preface like " d o y o u r e m e m b e r w h e n ..." or "did I tell y o u about ...," or are elicited b y instructions to tell a story. But in children's talk or talk between adults, there are m a n y marginal cases. O u r focus is o n identifying a conversational story a n d h o w it is occasioned. Given that the story is part of discourse a n d the storyteller a participant in an interaction, some of the structural organization of the story is directed to its function in a particular conversation. "Narrators linguistically shape their tellings to a c c o m m o d a t e circumstances such as the setting as well as the knowledge, stance, a n d status of those in their midst" (Ochs a n d C a p p s 1996). We intend to ovntninp h n w t h e e m b e d d i n g contexts of telling alter the internal structure and 2

ERVIN-TRIPP AND KUNTAY

CONVERSATIONAL STORIES

201

features of conversational narratives, attending to some issues raised by Jefferson in her classic study of the conversational embedding of stories (1978).

2. Narrative G e n r e v s . Narrative Production N o w that w e have the video and audio technology to look at conversational sequences, w e have the possibility of two different levels of analysis in the study of talk. O n e level is the live interaction, negotiated o n the spot in dialogue or through listener response. Another level is that of talk about talk using genre and speech act vocabulary, that is, classification by speakers and listeners or even by coders, and reference during talk to something said before or about to be said: "she told a story a b o u t " h e told m e to "he asked m e to ..." This is what Hymes referred to in his development of an ethnography of communication (1972) or what is sometimes called metapragmatics. This level includes our cate gory and prototype system for talk. There is a similar distinction in the study of speech acts like requests. What happens in real-time circumstances can differ from what is retrospectively recognized, remembered, reported, or judged, and therefore from what enters into speech act theories — the recollection of speech acts in tranquillity (Ervin-Tripp, Strage, Lampert and Bell 1987). There are thus two forms w e could call a story. One is the social genre that has prototypes and ideal forms and provides a model for what we expect when we ask "Tell m e that story about when ..." The genre is talked about and has a cultural existence in member instruction and evaluation. The second is whatever happens when people talk about the past, the future, or recurrent events or practices in ordinary conversation. W h e n w e look at conversational events in a transcript, w e do not know how much notions of prototypic genres affected per formance, or even whether the speaker considered what was said to b e a story. Very little work has addressed the relation between talk about talk and the inter actional events such talk refers to. O n e exception is the work of Bauman (1993), who examined a speaker's "metanarrational comments" and "disclaimers of performance o n the grounds of insufficient knowledge" (p. 188) as an indicator of the speaker's construction of genre conventions. However, although Bauman emphasized the relevance of such metapragmatic comments to the analysis of the features of narratives that embed them, h e did not attempt to relate speakers' notions of prototypic narratives to what gets said in ongoing talk. It seems obvious that interactional events as remembered have some bearing on talk about talk. But do the prototypes and genre classifications have any effect on ongoing talk? When w e cannot interview the tellers w e have to make inferences from the nar rative features, evaluations, and changes under conditions of evaluative monitoring. 3

Some theoretical approaches view narrative genre as a "dynamic expres sive resource, in which the conventional expectations and associations that attach to generically marked stylistic features are available for further combination and rKOmhinah'nn i n f t i o n m A ^ u

£

•

202

DISCOURSE STUDIES

genre or norm and narrative as a spontaneous, often ambiguous conversational event - may converge or diverge from one another, depending on the occasion ing conditions which influence narrative features.

3 . Standard Narrative

The standard characterization of personal narratives in social science research grew out of the Labov and Waletzky analysis (1967). Labov'sfirstelicited stories carne from his "danger of death" sociolinguistic survey question, which was designed to produce style shifts toward the vernacular variety of the speakers, in the heat of emotional recall. Eventually more than 600 American east coast adolescents and adults were posed the standard question "were you ever in a situation where you thought you were in serious danger of getting killed?" As a result, Labov elicited stories of near-death experiences that got rendered in fully formed narrative structures, upon which Labov and Waletzky later built their theory. They proposed that it is only spontaneous accounts of past personal experience, "not the products of expert storytellers that have been re-told many times" (p. 12), which could provide a window on the most fundamental forms of narrative structure. However, the methodology they used in order to obtain narratives is not appropriate for either spontaneous or first-time narratives. On the contrary, danger of death stories are likely to be retold, to be seen as attentiongrabbing and entertaining to listeners, and to be thought appropriate for prototypic performance occasions. A human protagonist is at the center, and usually there is some kind of conflict and suspense even in brief telling. Thus, what Labov and Waletzky pursued in personal experience narratives, that is, first telling of thematically dramatic experiences, involves a contradiction in terms; the inherent ever-present conversational relevance of such stories makes them amenable to repetition and stabilizes their structures. In spite of the specificity of the so-called "invariant structural units that are represented by a variety of superficial forms" (p. 1) to this very limited genre of danger of death stories, the internal features that Labov and Waletzky (1967) found have been taken as fundamental in most narrative research since. Labov (1972) defined narrative as "one method of recapitulating past experience by matching a verbal sequence of clauses to a sequence of events which (it is inferred) actually occurred.... We can define a minimal narrative as a sequence of two clauses which are temporally ordered.... There is temporal juncture between the two clauses, and a minimal narrative is defined as one containing a single temporal juncture" (pp. 360-361). The narrative clauses, he pointed out, had to be coordinate, not subordinate or conditional, as are clauses used in representations of practices or recurrent events, or in backgrounding of information. The Labov and Waletzky narrative can be taken as typical of a certain culturally defined type of narrative that optimally occurs under elicitation. Peterson and McCabe, beginning their study of narrative development in

ERVIN-TRIPP A N DKUNTAY

C O N V E R S A T I O N A L STORIES

203

in McCabe, In press). Their strategy for collecting narratives is to have a familiar interviewer engage the child subject in a conversation. During those informal conversations, the experimenters use what they call a conversational map, a series of short, deliberately unevaluated narratives about things that supposedly hap pened to themselves and could possibly have happened to the child (e.g., getting a bee sting). Such a methodology carries considerable ecological validity lince it is the naturally occurring conversation, at least from the perspective of the child participants in the study, that occasions the topics of narration. However, adults could push their agenda of obtaining data quite strongly in such elicitation contexts, sometimes not giving children enough chance to select or expand on their o w n favorite topics. As a result, it becomes difficult to strip off the effects of heavy narrative scaffolding from the outcome structure. The elicitation method, which is employed most commonly to obtain com parable narratives from different subjects in developmental studies, folklore, and sociolinguistic interviews, demonstrates that stories are cultural constructs. If we are asked for stories, w e know what is meant, and w e are able to identifygood stories and good story-tellers. Most research on narratives has relied on elicited stories or prefaced stories, those most likely to be identified b y both speakers and audiences as fitting the cultural prototype. 4

4. A B o u n d a r y C a s e As opposed to elicited stories, stories occasioned in the conversational situation often do not exhibit many of the prototypical narrative genre cues that are thought to be embedded in the structure of a story. The example below illustrates a marginal case of classification that occurs in conversation when stories are spontaneous, not elicited. After the Loma Prieta earthquake, a student taped a conversation that included two brothers and other student friends. H e had re ceived an assignment to tape and analyze about an hour of naturally occurring interaction. 5

(1) Earthquake story Albert and Ned are two brothers. Olga is Ned's friend, Cynthia her roommate. All are college students. '' 53 Al: you know that54 that "nice *glass *china *display case in our *dining room? 55 Ned: =in the *dining room= 56 Cyn: =o-o-oh= 57 Al: **trashed/ 58 Cyn: =forgetid= 59 Ned: =*absolutely= trashed/ 60 Al: whole thing a=bsolutely..yeah = 61 Ned: =*every *single bit= of 'glass and 62 * pottery in th11

'•104

DISCOURSE STUDIES

65 Al: crystal 66 Ned: *everything ..""trashed/ 67 Cyn: =o-o-oh my go-o-o-d= 68 Al: =oh a er *antiques *genuine= *antiques 69 Ned: =and the *amount of *money= we have lost 70 is going to be **astronomical/ (UCDisclab:OJJAKE) Features. This example does n o t m e e t the usual criteria of narrative. T h e r e is (a) n o animate protagonist, (b) n o sequence of events, (c) n o temporal juncture, (d) n o temporal connectives, (e) n o conflict. W h y does this seem to people w h o hear it to b e a story? T h e r e is the introduction of the "protagonist," a piece of furniture, b y a " c o m m o n g r o u n d " preface (53). T h e r e has already b e e n a n event in time, a n earthquake, so t h e story is about an implicit temporal sequence, an earthquake followed b y the outcomes of the earthquake. Most important, the r h y t h m of the telling, the alternation between ellipsis a n d expansion, the lexical a n d syntactic repetitions, the evaluation b y the listeners (56, 67) a n d b y the tellers (68-70) a n d t h e p r o b i n g b y participants (63) is a p p r o p r i a t e to t h e excitement of a dramatic story, so it seems to listeners to b e a story. Context of telling. E x a m p l e (1) occurred i n a spontaneous conversation, i n a series of rounds about different facets of the earthquake experience. To have restated t h e obvious, to h a v e given a temporal, causal sequence such as "we were in the classroom a n d suddenly it shook" would h a v e m a d e n o sense be cause everyone present already h a d experienced the beginning. T h e highly salient experience of the earthquake established the c o m m o n g r o u n d for the r o u n d participants to reveal their individual observations in t h e immediate aftermath of the shared disaster. Stories of disasters recentiy experienced b y everyone in the r o o m probably h a v e these c o m m o n properties - they often occur i n rounds, a n d they are both semantically a n d syntactically elliptical, building o n c o m m o n knowledge and conversational sequences leading u p to the story. As told later w h e n the c o m m o n knowledge is forgotten, or to audiences w h o d o n ' t share the same history, the stories m u s t b e reshaped to provide b o t h orientation a n d temporal sequences to fill in what is n o t shared. I n fact, w h e n Luebs (1992) interviewed 14 people two m o n t h s after the L o m a Prieta quake, the speakers provided extensive orientation a n d full-fledged narrative structures. Moreover, since the narratives were not part of naturalistic conversation, b u t w e r e elicited b y the researcher, most of t h e m included codas, that is, verbal or nonverbal ways of showing that a narrative is over. L u e b s ' data confirm that elicited narratives m o v e closer to t h e prototype. 6

4.1'. Databases for Study Adult data. T h e U C Disclab adult data consist of 180 transcripts collected in a variety of contexts, b u t primarily from informal natural groups t a p e d i n natural t _i__j—1„7 ; „ r < a i 4 W n i a with 500 fines of transcription archived using

ERVIN-TRIFP A N D KUNTAY


205

American school data. We created a non-task waiting r o o m context for best friends of seven a n d ten to tape natural sociable conversation in a California public school, where the children were varied ethnically and some were bilingual. These ten dyads a n d triads are in the U C D i s c l a b archive. Family data. T h e Ervin-Tripp family data consist of 38 transcripts from 8 multichild California families with at least one child 3 years old. S o m e families w e r e videotaped over a p e r i o d of 18 m o n t h s in their h o m e s . We observed meals or child free play with siblings a n d a visiting peer, with the goal of differentiating the social m a r k i n g features in children's speech with peers, siblings, a n d adults. Turkish preschool data. We obtained child-adult conversations at preschool sites in Istanbul, Turkey. Both of the preschools h a d 3-to-5-year-olds, providing around 40 children from w h o m w e t a p e d different k i n d s of elicited a n d spontan eous e x t e n d e d discourse. T h e informal preschool system provided multiple settings such as classroom environments, various organized a n d spontaneous play groups, a n d casual chats to search out the early d e p l o y m e n t of narrative skills. For the examples included in this paper, w e analyzed a r o u n d 60 h o u r s of audiotaped talk.

4.2. Locating Narratives in Transcripts To identify candidate instances of narrative segments, w e employed several methods: computerized search for specific linguistic markers such as temporal connectives, reading over the datasets for larger-level indicators of narrative such as reference to irrealis events or past events, prefaces b y narrators, prompt ing b y audiences, or audience evaluations. No single m e t h o d is adequate b y itself for identifying all the segments of talk that are of interest. This p r o b l e m brings us back to the definitional p r o b l e m we discussed in the beginning of the paper. If we have to determine what constitutes stories, we are probably better off in considering the narrative genre as a continuous cline, consisting of m a n y subgenres, each of which m a y n e e d differential research treatment. Since there currendy exists n o single m o d e l that includes criteria encompassing all the range of oral narrative forms, w e n e e d to a p p r o a c h the p r o b l e m employing different tools. O n c e we identified narratives, the n e x t step was to look back at the antecedent context to find h o w the narratives w e r e situated, a n d h o w they w e r e occasioned by what was said before. We were surprised at h o w often this context changed the interpretation of the narrative.

5. O b s e r v e d Features a n d C o n t e x t s of C o n v e r s a t i o n a l Narratives We found that using judges to identify proto-narratives in the family transcripts produced a wide varipfv nf oli«->«-i— "

8

. 206

DISCOURSE STUDIES

temporal junctures in t h e family a n d the adult data, w e found projections of future events, fantasies, descriptions of regularities, reports of conversations, and reports of events that h a v e occurred - including dreams. It was important in the case of y o u n g children to use a linguistic criterion for choosing text materials, so as n o t to impose categories that are n o t in the system of the child. If w e used temporal connectives as a search criterion - such as when, while, after, before, a n d and then - w e found that they first a p p e a r e d in child speech in c o m m a n d s a n d in p l a n n i n g for future c o o r d i n a t e d action, r a t h e r t h a n in experience or fantasy narratives (Ervin-Tripp a n d Bocaz 1989). I n other words, planning displays s o m e narrative qualities in child usage. T h e t e m p o r a l clause in such sequences defines t h e relation b e t w e e n two activities of a single actor or between two actors' activities. "I'm going to make a garbage can when I'm all through with the train lid." (3.0)° Temporal clauses occurred b o t h in reference to simultaneous events such as "you listen while I r e a d " (3.3), a n d to successive events, as in "can I h a v e y o u r w o r m w h e n y o u get finished" (3.1). Young children describe sequences of actions in elaborate future plans, in cluding directorial planning of complex dramatic play with coordinated scenarios that are temporally detailed. T h e r e is especially rich past a n d future reference whether the child plays the director planning the actions, or an actor undertaking the script. Outside of dramatic play, y o u n g children talk less often about the past t h a n adults d o . A m o n g t h e narrative-like sequences w e found w e r e generic descriptions of scripted events, describing w h a t h a p p e n s o n a regular basis, or if y o u do something, what results. T h e following dramatic play enactment illustrates a n occasioning of talk about the past. T h e actor playing t h e patient is preparing a c o m m a n d to operate o n h e r leg, b u t describes a personal experience in t h e past to account for the injury. W e see from other child texts (see 3 below) that injuries are usually asked about a n d are explained b y a story of w h a t h a p p e n e d . T h e child inserts this c o m m o n type of topic sequence into dramatic play. (2) Doctor-patient play of 4-5 year olds 1 Kit: pretend there's something wrong with my leg. my leg2 let's pretend that I tell you that my leg's-um3 let's pretend I tell you- first, you operate on it. 4 um, but before you operate on it, 5 let me tell you something, okay nurse? 6 Jill: um, 'kay. 7 Kit: um, when I was walking down the street, 8 I saw this piece of glass and I picked it up, 9 then I didn't see too well, then it goes way up to here. 10 see now. it's-now it's over there. 11 can you-can you operate on it, nurse? 12 Andy: lean.

ERVIN-TBIPP A N D KUNTAY CONVERSATIONAL STORIES

207

This segment begins in the directorial voice (1-3) which switches without a n y m a r k i n g into enactment in the patient's voice (5). Kit starts with a request for the floor (4,5) whichJill acknowledges (6), a n d Kit t h e n begins t h e story with a subordinate backgrounding clause (7) establishing the place of the event a n d the protagonist's activity at the time of an action (8) b y t h e protagonist, reported b y a telic verb. Gestures showed the injury (9,10). T h e r e is a sequence of narrative events (9). T h e story provides the explanation of a n injury n e e d i n g repair in the dramatic medical scenario p l a n n e d at the beginning (1-3), so it is an inserted elaboration before the medical scenario unfolds (11). We m a d e similar form-based searches in our extensive database of adult materials, looking for markers of t e m p o r a l j u n c t u r e as a w a y to locate story sequences. We found relatively few extended stories. T h o s e referring to t h e past often were brief, as in the examples to follow. T h e y were stimulated b y t h e sight of objects, pictures, p e o p l e or television. W h i l e the children relied often o n the setting as a source for talk, i n the adults, w e found m o r e stories that continued prior talk, either i n story r o u n d s , or in stories to support or challenge conversa tional claims. T h e r e w e r e scripts, that is stories about customary events, a n d stories that r e p o r t e d dialogue only. T h e focus of m u c h of the students' social talk was o n character, on the implications of events rather t h a n o n the events themselves, a n d o n daily p r o b l e m s that they could n o t solve. I n the last type of interaction, t h e discussions of p r o b l e m s , w e can expect s o m e of the kinds of sequencing that Jefferson (1988) has r e p o r t e d for troubles talk. Unlike L a b o v a n d Waletzky's narrators, our tellers of spontaneous stories did n o t take a long time to build to a high point. It was rare that stories involved d r a m a or suspense. T h e character anecdotes usually presented a brief event, a n d t h e p r o b l e m situations b e g a n with a single event to initiate a topic, with interchanges a b o u t alternative outcomes. T h e s e stories i n y o u n g adults are analogous to the stories Ochs, Smith a n d Taylor (1989) h a v e found i n A m e r i c a n families, which p r o p o s e p r o b l e m s a n d alternative outcomes. T h e s e speakers did n o t seize t h e floor for a long performance of a story with suspense. 10

5.1. Elicited/Prefaced

Narratives

It is elicited or prefaced stories that best demonstrate that the structure of stories is strongly related to the circumstances of their telling. T h e y are likely to fit s o m e paradigm of a g o o d story, since t h e prototype is evoked for tellers. T h e elicitation signals that the r e s p o n d e n t h a s t o u n d e r t a k e a performance. T h e preface signals that the teller is willing to identify what is to b e read as a story. T h e y are, in some sense, accountable to the public standards for a story. W h i l e adults m a y preface or volunteer stories to entertain, in m a n y of the ex amples w e found there was questioning from t h e audience to p r o m p t the teller, and to give the teller the floor. But a question can focus a reply o n a n a r r o w issue rather t h a n o n a full story, thus altering t h e temnriral m^.-;— -•- ^ „„I_J..-—_.

DISCOURSE STUDIES

208

(3) Broken foot elicitation Melody is 4;7, Lisa is 2, Dave and Georgette are researchers in the home of the children making a videotape. There is an overlapping hosting conversation by the father which interweaves the two sets of speakers. 4 Dav: OK. in the course of the next months, 5 we hope to be out here a few more times. 6 Geo: what happened to your foot, Lisa? Fa: =would= you like any7 Mel: =um= [to G] she, she =cut ib= thing to eat or drink? 8 in the wheel. D: =not me= [to Fa] 9 she cut. Fa: =Georgette do you?= 10 =when.when= =when the= G: =no thank you= 11 friend was next door, then, then 12: then, then, then, then she was pushing her on the bike, 13 and it got caught in the spoke or the wheel. 14 G: oh boy 15 M: and she broke a bone. 16 G: oh boy. 17 L: a bone. (Melody) I broke this 18 : =(it go slip) = [to G] Fa: =what kind of= sandwich would 19 : (the bone slipped) you like to eat, Melody? 20 M: Uh, um what kind do you have? (Ervin-Tripp Family Transcripts: Fleursll) T h e focus of Melody's reply was specifically selected b y a n a r r o w "what h a p p e n e d to your foot?" question, which prompts Melody first to a s u m m a r y (7) a n d then backwards in time to the circumstances that brought this result (10-13). As a story, it gets audience reaction (14,16). T h e disruption of the real time tem poral order in the story results from the specificity of the eliciting question. Most children observed in the Turkish nursery schools b y Aylin Ktintay did n o t p r o d u c e stories for other children, w h o gave t h e m n o prompts. It was adults w h o often supported children's elicited stories with p r o m p t s to get started and to continue. I n the following e x a m p l e , E m r e has b e e n talking to the adult researcher (Ad) about a children's entertainment center that h e visited over the weekend: (4) Alligator game prompting [Translated from Turkish] (Age: 4; 11) Emre: there is a Scoreboard, shows our score 1. is that so? how did you - did you make a lot of points? 2. >Ad: Emre: (I) did 3. but once 1 won a lot of things 4. that alligator-shooting game did not give us 5. because some part of it was broken 6. is that so? what happened? 7. > A d : Emre: got broken 8. we had won a lot of shillings 9. at that time it got broken 10. t

m\r

rrncVi 1

ERVIN-TKIPP A N D KUNTAY


209

14. those people who control Piramit - they fixed that 15. and then we got all that shilling. (Aylin Kiintay: Eryavuz preschool) This story has a classic construction with a high point a n d resolution, b u t the temporal m a r k i n g appears to b e affected b y the adult prompting. T h e adult question (2) takes the child from a description to a specific event. In answering the question, the child begins b y a short reply to the question, yet the but signals a newsworthy issue or violation of expectation (4), even as the story is beginning, m a r k e d with once. T h e story t h e n begins. T h e resolution does n o t immediately follow the complicating action, b u t the adult question interrupts the flow (7). At this point the child repeats a n d moves back i n time to recapitulate the sequence of winning before the e q u i p m e n t broke, even using a pluperfect affix to m a r k anteriority before the time evoked b y the question, and then points out the prob lem resolution with a but then marker. While eliciting seems to bring out proto typic stories, prompting can alter the temporal sequence b y its focusing effect. 5.2. Rounds of Stories In m a n y conversational settings, stories implicitly invite related stories from other participants. Goffman (1974) states that "an illustrative story b y one par ticipant provides a ticket another participant can use to allow the matching of that experience with a story from his repertoire" (p. 510). Adults, in their inter actions characterized b y a series of stories, pick out some features from previous stories a n d work t h e m into their ongoing story without bothering to frame each story a n e w (Ryave 1978), m u c h as speakers in sequential self-introductions use the same frame repeatedly. Umiker-Sebeok (1979), analyzing preschool children's narratives p r o d u c e d spontaneously within natural conversation with other children, found that the most c o m m o n response to a narrative was similar nar rative. At age three, response narratives in Umiker-Sebeok's data did n o t s e e m to advance the conversation "as a conjoint elaboration of a conversational topic" (p. 107), b u t a p p e a r e d as "dangling narratives" that generally concerned the same general topic. In the following Turkish preschool example, C a n initiates the series of nar rative contributions b y establishing the t h e m e to b e visits to the doctor's office. Single statement initiations that d o not go b e y o n d minimal event representations are very c o m m o n in this younger age. T h e children await the teacher's interested reaction before building u p o n their original one-line narratives. I n this particular example, the teacher demonstrates interest b y saying "is that so, dear?" (2), set ting u p the topic of visits to the doctor as interesting for all the children. (5) Visits to the doctor Beril (4;0) is one of the two girls in a seven student class. Osman is 4;0. Can is 3;7. It is breakfast time, and all of the children are sitting around a table, with their Teacher (Teach) and the researcher (Res) present.

DISCOURSE STUDIES

210

my mother took me to the doctor, too 3 Beril: shall I say what [he/she) said 4 Can: (he/she) said let him eat waffles a bit later 5 Can: 6 Teach: said let him not eat too much waffles, right? said let him bite in teeny-weeny bites with his teeth 7 Can: and then also pickles.. 8 9 Teach: did you get a shot Can? no: , 10 (1) don't really like 11 Can: 12 Teach: let him eat little litde by little 13 let him not eat much 14 xxx wouid you like biscuits? {to Res} 15 no, thanks 16 Res: Teacher, the doctor told me don't eat anything 17 Beril: 18 Teach: the doctor? 19 but if we don't eat we can't grow no (he/she) said have breakfast but 20 Beril: don't eat those that your mother brings 21 22 Teach: is that so? 23 Osman: my mother did not take me to the doctor's 24 Teach: because you aren't sick, right? don't get sick, ideally 25 (Aylin Kiintay: Ubaruz preschool) Beril, after replicating Can's one-liner in line (3), follows o n C a n ' s subtopic a b o u t his doctor's nutritional advice. Aside from d e m o n s t r a t i n g thematic continuity, Beril's story features formal similarities to C a n ' s , such as em ploying quoted speech of a doctor. O s m a n then contributes (23) b y reporting non-occurrence of the topical event in his life. I n Turkish children's rounds like t h e a b o v e , s e c o n d n a r r a t o r s frequently claimed to h a v e exactly the same experience as first narrators, d o w n to the same details. Since the overarching motivation for rounds of Turkish children in classroom settings seems to b e to gain the teacher's interest, they usually try to top other children's stories with their own versions. If an initiation story attracts the attention of the teacher, other children select out of the thematic possibilities of the original story to construct their own. Usually, some of the thematic content gets replicated in the following stories. As Cortazzi (1993) suggests, usually "later narratives are highly pre-specified, showing m a r k e d parallels of topic, t h e m e , character of events with preceding narratives" (p. 3 2 - 3 3 ) . R o u n d s allow for the ellipsis of p r e s u p p o s e d information or allusion to information from earlier stories. T h e highly elliptical Earthquake narrative (1) occurred in a series of r o u n d s with successively changed protagonists. Students began b y telling w h e r e they were w h e n the earthquake h a p p e n e d , and what they and p e o p l e a r o u n d t h e m did. T h e y a d d e d second-hand stories about what other 1 1

1

1

-

1

. flia^T* n o k a n r l t h e n

ERVTN-TRIPP A N D KUNTAY


211

hierarchy. After a discussion about the behavior of their dogs, Cynthia said that nothing of hers fell down, turning the topic to inanimate objects, including the immediate cues in the r o o m in which the taping occurred. (6) Earthquake round 1 Cyn: = =yeah our room our room looks like nothing happened at all, 2 Dom: = = I was 1 was worried about this [pointing at empty beer 3 bottles on shelves] i-i-it's like coming home to find a big 4 pile of lads, [lads refers to empty bottles] 5 Ned: this is the first this is the first indication 6 that this was something 7 because the last one that happened was like a six 8 [referring to Richter scale] in Los Gatos and 1 was sitting 16 Ned: and then Al and I come back and we are just all *oh my 17 =*god there it is tilted nearly fallen off just hanging= 18 Al: =yeah.. instant hand up supporting them because they are just-= T h e three w h o were present here co-produce a description of the state of the beer botdes. At this point D o m continues a r o u n d of reports o n each person's room, including non-effects. 34 35 36 37 38

Al: nothing happened in my room. Cyn: ==things were moved. Ned: [to A] yeah funny thing Al=your room is a m-= Geo: =you couldn't tell= if something happened to your room.

There is a n acceleration (39) as N e d emphasizes the extent of damage in their house, summarized b y Al (43). At this point the two brothers overlap in two dramatic descriptions, Al with exaggerated generalization (43), N e d with em phatic prosody in (44). 39 40 41 42 43 44 45 46 47 48 49 50 51 52

Ned: oh my mum and dad cannot get to Pete's room to see what it's like.. = they ca*n'fc= get to George's = yeah = Al: Ned .all the book-= Al =*natural *disasters=just do *not* stop *happening at *our house. Ned ==*all the *bookcases *everything *fucking cour *house is **chaos. = =yeah. Al Ned it's *chaos.. = the bookcases= Olga: =the bookcases = came off = the wall?= [toO] =01ga = At: Ned: = = t h e bookcases all fell down= *books *everywhere *fumiture= Al: =you know = Ned: *moved=

DISCOURSE STUDIES

212

supports the generalization b y describing the bookcases. But it is Albert w h o gets to the worst case of all, the shelving containing the antique crystal, a n d the brothers p r o d u c e a duet narrative. 53 Al: you know that54 that.*nice *glass *china *display case in our *dining room? 55 Ned: =in the *dining room = 56 Cyn: =o-o-oh= = 57 A l : **trashed. 58 Cyn: =forget it.= 59 Ned: = *absolutely= trashed. 60 Al: whole thing a =bsolutely..yeah = 61 Ned: = *every * single bit?= of *glass and 62 *pottery in th63 Olga: and *crystal? 64 Ned: * all the crystal..*trashed. 65 A l : crystal 66 Ned: *everything ..*trashed. 67 Cyn: =o-o-oh my go-o-o-d= 68 A l : =oh a er * antiques *genuine= *antiques 69 Ned: =and the * amount of * moneys we have lost 70 is going to be **astronomical. (UCDisclab:QUAKE) T h e r o u n d thus began with what h a p p e n e d to items o n shelves in each of the participant's c a m p u s r o o m s , m o v i n g from a vivid local e x a m p l e in the im mediate environment, t h r o u g h another male's r o o m to the most extreme case personally known, the parental h o m e of Albert a n d Ned. At this point what triggers the dining r o o m narrative is the two strong generalizations i n (43,45). To illustrate the generalization, N e d a n d Albert m o v e emphatically through a coordinated b u t elliptical description, with considerable rhythmic emphasis. T h e line length is shorter, with stronger stress and m o r e focus o n predicated de scriptors, compared to the beer botde description where m o r e action is mentioned and there are h u m a n participants. T h e final evaluation b y Ned, in a longer, syntac tically complex sentence, changes the r h y t h m to summarize the major outcome. R o u n d s can also lead to conventionalized story-telling, since they can provide occasions to b e given the floor a n d to produce a tale that can use prior models as scaffolding and the obligation of mutual appreciation as support." I n the following example, w o m e n at a Senior Center were discussing what they thought of changes in clothing fashions, a n d m o v e d into the topic of m a k i n g graduation dresses. T h e following is an excellent example of the frequent feature in all-female groups of w o m e n telling real-life narratives of their o w n embarrassing experi ences to create amusement. (7) Making grade school graduation dresses Setting: Senior Center

ERVIN-TRIPP A N D KUNTAY


213

A n n told a story about h o w h a r d it was to finish in time, which e n d e d with a n evaluation. 14 Ann: I remember I was so traumatized with that dress [laughs) After general laughter, Bev started h e r own story o n a similar t h e m e . 16 Bev:

what did *I pick but an Indian *linen..

Bev's story quickly m o v e s into a p r o b l e m . H e r laughter (31) suggests that she already h a d the intention of m a k i n g a h u m o r o u s report o n h e r solution. D e b ' s laugh (33) collaborates in the key. 29 Bev: and I uh if you did not leave enough yeah? 30 for your seams you know and then 31 uh, I had [laugh] all but one sleeve for puffed sleeves? 32 and one was not gonna be as full as the other one so-o 33 Deb: [laugh] 34 Bev: my grandmother used to bring a lot of things from overse35 you know, from uh, Europe, she used to travel a lot, my 36 dad's mother? and she had some beautiful ribbon 37 so I used it for the sash I remember it was an orchid. 38 sash, and uh, that- that dress did up, but uh, 39 what was I to do about the sleeves. 40 so I walked crooked up the stage. 41 All: [laugh] (UCDisclab: WCON1) H e r e , Bev reports a situation which for a y o u n g girl would b e a painful dilemma, the lack of fabric for making the dress sleeves symmetrical for a major public event. H e r story about the sash (34-40) sustains the suspense; "but what was I to do about the sleeves? " and takes the listeners explicitly to h e r p r o b l e m of a crooked dress. T h e n she states the p u n c h fine, the embarrassing solution of walking crookedly (40). This story is typical of w o m e n ' s h u m o r o u s narratives about themselves (Ervin-Tripp a n d L a m p e r t 1992). We can surmise that Bev knew at the start that she would present a problem narrative with a witty outcome. This is a prototypic narrative. Adult r o u n d participants use prior m o d e l s as scaffolding, with the possible aim of developing mutual appreciation as similar events. Kirschenblatt-Gimblett (1979) studies "story-dominated conversations" of Eastern Europeanjewish adults finds "in these story-dominated events, there is a preoccupation with narratives as things in themselves. For this reason, and because the narratives are preformulated a n d relatively self-contained (they can b e understood without reference to any preceding conversation or narration), there is a tendency for story-dominated events to b e organized like beads o n a string" (p. 291). I n sum. r o u n d s arp a t v n o n f o CT>OO»I. >•—^ - - • IT

DISCOURSE STUDIES

214

b e cycles of jokes, cycles of h u m o r o u s personal narratives, cycles of stories of one's r o o m in an earthquake. If r o u n d s provide a mutually cooperative audience that allows floor to the speaker, there is the possibility of elaborating a per formance in terms of the cultural n o r m s . R o u n d s can thus involve continuity of genre, continuities of latent topic, continuities of key, emulation of form, a n d presupposition or ellipsis. 5.3. Environmentally Cued Narratives M o r e c o m m o n than r o u n d s in our conversational data are story triggers in the local environment. H e r e , unlike rounds, there are n o models provided in the preceding conversation. T h e r e is considerable variability of structure. (7) Bird burials 1 Ellen: don't look...there's a dead bird. (20 turns) 21 Ellen: a-a-aw let's bury it. 22 Dina: =let's= not [laugh] 23 Bill: =1 -1= don't think so! some cat will probly =get it = 24 Ellen: =my dad= always 25 did it for us 26 =[laugh] =we'd put it in a shoebox and bury it. 27 Dina: =[laugh]= how sweet. 28 Bill: a-aw 29 Ellen: then we'd make crosses out of popsicle sticks. 33 Dina: that's cute though...that's really nice. 34 Ellen: we'd have a funeral for it and everything. we'd say a prayer [laugh] and bury the thing. (UCDisclab: LCON2) This is an example of the reminder report, which describes scripts rather than presenting a conflict with a resolution. T h e r e is a serial connective then, b u t the use of would marks this as a description of customary practice. T h e r e is, however, audience evaluation (27, 33) in this case, as if this w e r e a narrative. Children's conversations often include reference to what is in their immediate environment, since their activities tend to build on available resources for play. T h e following is an example of environment-cued fantasy of ten-year-olds in California. T h e imaginary characters and the disruption of order are characteristic of b o y s ' stories as early as four (Nicolopoulou, Scales a n d Weintraub 1994). (8) Microphone story Two ten-year-old boys alone in a testing room eating lunch notice the microphones strung from the overhead light. 1 Sam: what's this for? 2 Jer: it's a microphone 3 [S & J start to sing into microphones]

^



215

Singing or other stage performances are frequent reactions to m i c r o p h o n e s at this age. At this point, Sam begins to build u p a scene of disorder. 4 Sam: [laughing] I was screaming! 5 [laughing] (hey..l'll be acting like this) 6 hey Jeremy Jeremy Jeremy I'm gonna be laughing in this 7 and the light falls down [laugh] 8 what would you do if the light fell? T h e past tense (4) marks irrealis a n d a possible story beginning. I n (5-6) the tense is m o v e d to the future and in (8) to the conditional, setting the stage for m o r e fantasy. I n (8) w e see the imaginary disruption of order, which creates a problem to b e solved b y the fantasy narrative. 9 Jer: I'll be under the table 10 Sam: [laugh] no., and you did it! 11 Jer: I'd be outta this class [laugh] (5) 12 Sam: it'll hit you on your head! [laugh] 13 Jer: you know what I'd do, 14 I'd be Superman and catch it. 15 be Superman and get squashed like a pancake. 16 Sam: I'll be Flash and get the fuck out. (UCDisclab: GFIVE3) In this sequence, Sam changes his position from challenger (10,12) to capping with a character shift (16). Besides being fantastic responses to an environmental stimulus, these stories show h o w posing alternative outcomes can b e part of col laboration i n competitive bouts of story-telling. 5.4. Narratives Presenting Problems As opposed to stories that resolve the complicating situations they set u p during being told, some narratives invite audience participation in a possible problemsolving process b y formulating problematic situations or protagonist behavior. Ochs, Smith, a n d Taylor (1989), w h o studied w h a t they call "detective stories" during dinnertime talk, noticed that such problem-solving narratives can b e characterized b y interactive negotiation of the incomplete resolution. T h e two most frequent contexts for temporal series in y o u n g adults' talk in the data are personal characterizations and p r o b l e m situations. T h e p r o b l e m situations are presented with potential endings, which participants co-construct. These differ somewhat from "troubles talk" (Jefferson 1988) w h e n they refer n o t to chronic problems b u t one-time events. A Taiwanese engineering student is conversing with two Asian-American w o m e n about thp *TH-.J«-«f * •'

DISCOURSE STUDIES

216

(9) Asian women 1 Hel:

you know...I was talking with a friend about

2 that asian Caucasian dating thing... 3 and my complaint to {[p] to my friend) 4 is that that is that asian guys just don't ask 5 =y' know?. 6 Min: = yeah = they're not as (xxx) as (xx) 7 Hel:

do you agree?

8 Yuan: totally agree 9 Hel: so then if that's true then how come when you guys 10

get like mad at at *us

11 12 13 14 15 Yuan:

if we go out with Caucasian people then you guys don't ask.. you guys don't get off your butts (I)*answer me it's becau:se..

16 Hel:

(I) you expect us to wait

17 Yuan: =*ri:ght 18 Hel: ="yeah = 19 Yuan: *ri::ght 20 Hel:

no, seriously

21 Yuan: (4) hm I don't know I don't what to say 22 Hel:

= =yeah..i don't know either

This segment poses a p r o b l e m of tension between Asian m e n a n d w o m e n , in which H e l e n presents a challenge to Y u a n in the you guys form (12,13) and pressures h i m (14) to answer the challenge, then herself providing the answer (16). I n the conversation, Asian males are accused of causing the p r o b l e m about which they are angry (10) b o t h b y their behavior (12,13) a n d b y their unrealistic expectations (16). T h e topic of m a l e anger triggers a description from M i n in which the t h e m e is Asian m a l e anger. T h e story is basically in the form of a long quotation. 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

Min: = =did you guys see in the Daily Cal about stuff..about the vandalism in Dwinelle ' I actually like saw it.but um..there's this.. there's this like group of people who call themselves the (1) Asian Male something or other they like wrote like all over spray painted or marked =all over= .' Yuan: =o:h AMU?= Min: = =*yea AMU..they like wrote kill whitey and then like in the girls' bathroom downstairs they're like like y'know basically just slamming on asian woman for dating Caucasian men like stay within your own kind..stay within your own group.. know what I mean? and it was just Min: scrawled *all over the stalls and the walls was writing Yuan: hm-mm



217

Helen's evaluation (39) explicitly refocuses the topic again o n anger, a n d triggers a story from Y u a n which begins with the response m a r k e r well (40), suggesting h e intends the listeners to find a continuity from the topic to his story. 40 Yuan: well, I just met a Caucasian guy yesterday 41 and we were just talking right? 42 and he he he speaks a little bit of Chinese 43 and he got a job uh got a job with a company like 44 uh doing tradings and stuff 45 and he was just telling me 46 yeah...you know...some Asian women...man 47 some Chinese women you just take them and ...[clarification of bam — bang] 48 bang them...you know 49 like (2) have sex with them 50 yeah bang them...and then yeah so y'know 51 I'm just going to work here a couple of years 52 and then go back to Thailand and get a wife 53 you know what am I supposed to response54 to response to that? 55 Hel: you can bop him 56 Yuan: all I can say is you know 57 MM: tell him you're pissed 58 Yuan: all I can say is being in this world 59 there's somebody who like to hit other people 60 and there's somebody who like to get hit...you know 61 so (1) what can you say? 62 Hel: did you hit him? 63 Yuan: no 64 Hel: why not? 65 Yuan: cause whoever the girl that wants to go out with 66 him, that's her problem...that's not my problem. (UCDisclab: ASDAT) This story picks u p the racial anger theme, identifying the speaker with the anger of the graffiti writer implicidy b y an example of what could p r o d u c e that anger, white m e n ' s disrespect for Asian w o m e n . T h e story h e r e has a structure of a prototypic narrative involving a conflict, b u t the conflict is unresolved. T h e narrator evidendy did not show his anger to the Caucasian antagonist, a n d was left feeling dissatisfied. T h e businessman's statement (46-49) is presented without comment. T h e r e is a possible narrative transition marker then (50) but n o overt reply is reported, a n d the action shifts to interior reflection. Y u a n describes information about his o w n plans (51-52) implying a conflict with the stance of the businessman. H e leaves the story with a question about what h e should have d o n e (61). I n the context of the graffiti topic, the two w o m e n h a v e n o hesitation in assuming Y u a n ' s anger, and provide a n answer to h i s m i « r f « " 12

i.M

Vft

Ti- i -

' "

*

DISCOURSE STUDIES

218

prompted (64-66). The story has two unusual features: the failure to report the ending, which had to be prompted, and the sanitation b y the teller of alternative outcomes. These mark the story as a problem situation. 5.5. Narratives As Performances Every narrative, insofar as it attempts to get some point(s) across, has an evaluative aspect. However, only some narratives receive a very animated and dramatic enactment so that a performance aspect is attributed to them (e.g., Bauman 1986; Hymes 1972). In such cases, the teller foregrounds the evaluative com ponent of the story, with less emphasis placed on the events comprising the story. Cortazzi (1993), in contrasting performance narratives with interview narratives, asserts that performance becomes highlighted if the topic is appro priate to the audience, and if participants have reciprocal relationships and shared norms for evaluation. Schiffrin (1981) talks about one feature, that is, usage of historical present tense, that "makes the past more vivid by bringing past events into the moment of speaking..." (p. 58). Among some of the other common features that are employed as tools for highlighting the evaluated point of the story are constructed speech and animated intonation (Tannen 1989). The following segment coming from two sisters sitting in a cafe demonstrates how constructed speech can be used to constitute a story. Sara is a topic change; she is involved in Mimi's job hunt. Bringing her up brings up also talk about her personality, first a descriptor from Mimi "standoffish" and then a carefully staged mocking mimicry, which Lina launched (9) before Mimi's comment about weirdness (10) suggesting it is an independent but perhaps corroborating expres sion. The mimicry is a "story", in the form of a dialogue. But it is only a fragment, with no event sequence. (10) Sisters in a cafe talking about a job search 5 Lina: xxso you decided to call Sara? 6 Mimi: yeah, it's weird cause she's kinda..I don't know 7 don't you think that she's kinda standoff8 she's kinda9 Lina: she left this message i 10 Mimi: her and Jill are both kinda weird I think 11 Lina: =she always goes she goes= 12 Mimi: =they're = , , 13 Lina: [slow whiny] "I'm calling for Mi-i-mi-i" 14 Mimi: she's- she =calls me Mi-i-mi-i = 15 Lina: =did you hear her message?. 16 Mimi: yeah. 17 Lina: she goes [slow whiny] "I'm calling for Mi-i-mi-i. 18 um if you want you can work at the video store. 19 um tell her to cali me, bye Mi-i-mi-i" [laughs] (UCDisclab: SISTRS)

ERV1N-TRIPP A N D KUNTAY


219

that Lina gets the floor for a full quotation mimicking Sara's voice, a production she appears to b e planning b y line 11. As a conversational tactic this e x a m p l e corroborates a n d illustrates the points m a d e b y M i m i , b u t the urgency of h e r production, a n d her overlaps while accelerating till she gets the floor suggest another feature, a goal of achieving a g o o d performance, in this case of a witty mimicry. Also, as obvious from the content of the conversation, n o t m u c h of what the story conveys is n e w information. T h e only conversational p u r p o s e of the story appears to b e to display a shared assessment of a third person in a vivid way. I n the Turkish child database, totally "spontaneous" narratives b y a child were n o t v e r y c o m m o n . Teachers, at times explicitly, disallowed children to tell stories outside of the time allotted for story-telling. W h e n interviewed, a teacher reported that she really did n o t like it w h e n H a s a n , w h o told story (11) b r o k e into his "quite fantastical a n d irrelevant" stories in the m i d d l e of "serious" g r o u p activities. But even without encouragement, children occasionally told s o m e stories that w e r e n o t precipitated b y a n y r e m a r k , question, or topic in the preceding conversation. W h a t seemed to trigger such narratives was p r o b a b l y the tellability attributed to the events b y the storytellers. T h e content tends to be n o t banal, b u t out-of-the-ordinary events l e n d i n g themselves to highly theatrical performances that attempt to build a n d u p h o l d some suspense in the audience. L a b o v (1972) proposes that eventfulness is a n intrinsic aspect of events and is expressed b y evaluative devices: "evaluative devices say to us: this was terrifying, dangerous, weird, wild, crazy; or, amusing, hilarious, wonderful; m o r e generally, that it was strange, u n c o m m o n , or unusual - that is, worth repeating. It was n o t ordinary, plain, h u m d r u m , everyday or run-of-the-mill" (p. 371). Children also tended to repeat such narratives, a n d consolidated highly elaborate structures through repetition. As Goffman (1974) a p d y writes "effective per formance requires first hearings, not first tellings" (p. 508). T h e following narrative is from a 4-year-old b o y replaying a particularly dramatic reported near-death experience of a (nonexistent) sibling. 13

(11) Brain-washing During a gymnastic session at the preschool, a four-year old boy spontaneously launched into a dramatic story about his non-existent younger brother. When Hasan was asked to retell the story the second telling was very like the first. 1 Res: tell (it) again. 2 Hasan: my sibling opened medicine/medicine box — took (it)? [self-correction] 3 was able abl to open (it)? broke that lid? 4 ate th< them up 5 6 ate all all all {of them) up? 7 Res: a-ah! [=expressing [= surprise] 8 eee? [= [=so then?] 9

Hacon.

o2). (6) Cataphoric persistence of subjects and objects of transitive clauses in Sahaptin, Panare, Bella-Coola, Korean and Spanish 9

number of occurrences in the following 10 clauses 0-2 language Sahaptin subj obj Panare subj obj

N

>2 %

N

total %

N

%

9 21

19.6% 61.8%

37 13

80.4% 38.4%

46 34

100.0% 100.0%

9 19

31.0% 65.5%

20 10

69.0% 34.5%

29 29

100.0% 100.0% (Continued)

CTVON

COHERENCE IN TEXT VS. COHERENCE IN MIND

265

(Continued) number of occurrences in the following 10 clauses 0-2 language Bella Coola subj obj Korean subj obj Spanish subj obj

>2

total

N

%

N

%

27 82

21.4% 65.1%

99 44

78.6% 34.9%

126 126

100.0% 100.0%

53 106

35.3% 72.0%

97 44

64.7% 28.0%

150 150

100.0% 100.0%

19 70

19.0% 70.0%

81 30

81.0% 30.0%

100 100

100.0% 100.0%

N

%

O t h e r grammatical devices m a y also b e used for the cataphoric grounding of topical indefinite referents. A c o m m o n o n e is restrictive relative clauses in combination with the indefinite article: (7) a. b. c. d.

A man with no shoes on came into the office and... A woman who spoke no English then stepped forward and... Someone who's very anxious to meet you is coming over tonight... A guy I haven't seen in years has just called me and...

I n addition to tagging the newly-introduced N P as a n important topic in the subsequent discourse, the information in the REL-clause makes the referent salient, g r o u n d e d to the current location in the discourse. So that the referent can n o w b e attached at a r e l e v a n t location in the mental representation of the incoming text. 2.3.3 Anaphoric Grounding 2.3.3.1. Preamble: Definite Reference Speakers code a referent as definite w h e n they assume that it is identifiable or accessible to the hearer. By 'accessible' o n e m e a n s that it is represented in and can b e retrieved from - some pre-existing mental structure in the hearer's mind. W h e n speakers re-introduces a referent in such a context, they ground it by various grammatical devices. Such grounding serves to establish a m e n t a l c o n n e c t i o n between the referent's occurrence in the current text-location a n d its previous anaphoric trace in some extant mental structure. Grammar-guided anaphoric devices most c o m m o n l y g r o u n d re-introduced referents into three types of mental structures: (8) Mental structures for anaphoric grounding:

266

DISCOURSE STUDIES

We will take u p these three types of anaphoric grounding in order. 2.3.3.2. G r o u n d i n g to the Speech Situation G r o u n d i n g referents - or other coherence elements - to the current speechsituation is achieved b y indicating their spatial or temporal relation to the two participants in the discourse - speaker a n d hearer. This involves well-known p r o x i m i t y and o r i e n t a t i o n devices such as: 10

(9) a. The interlocutors: T , 'you', 'we', 'y'all' b. Other referents: 'this one', that one', 'that one over there' c. Location: 'here', 'there', 'way over there' d. Time: 'now', 'then', 'long ago, 'in the future' 'today', 'yesterday', 'tomorrow' 'this week', 'last week', 'next week' 2.3.3.3. G r o u n d i n g to Generic-Lexical Knowledge Culturally shared, generic-lexical knowledge is represented mentally in the per m a n e n t semantic m e m o r y . Referents m a y b e g r o u n d e d to this m e n t a l structure in two distinct ways. First, some referents are globally accessible because they are uniquely identifiable to all m e m b e r s of the relevant speech c o m m u n i t y ('culture', 'sub-culture', 'village', 'family') at all times. S o m e examples of such referents are: (10) Globally-accessible generic definite* referent

relevant social unit

a. The sun came out. b. The president has resigned. c. They went to the cemetery. d. The river is frozen over. e. Call the sheriff! f. The Gods are angry. g. Daddy is home!

all humans a nation-state a community a community a county a religion a family

Generic access to definite referents is frequendy intermixed with episodic text-based access, yielding a hybrid system of double grounding. A doublegrounded referent is accessible partiy through a n anaphoric connection to its episodic trace in the episodic representation of the current text, a n d partly t h r o u g h connection(s) to generic-lexical k n o w l e d g e . T h i s h y b r i d t y p e of g r o u n d i n g is often referred to as f r a m e d - b a s e d or script-based reference (Anderson, Garrod & Sanford, 1983; Yekovich a n d Walker, 1986; Walker and Yekovich, 1987). Typical examples are:

GrVON

C O H E R E N C E I N T E X T VS. C O H E R E N C E I N M I N D

267

(11) Double-grounded frame-based reference: a. My boy missed school today, he was late for the bus. b. He showed us this gorgeous house, but the living room was too small. c. She went into a restaurant and asked the waiter for the menu. The definite referent 'the bus' in (11a) receives its anaphoric grounding from two separate sources - the antecedent referent 'school' in the preceding text, plus generic-lexical knowledge of the frame 'school' and its sub-component 'bus'. Similarly in ( l i b ) , the definite referent 'the living room' receives its anaphoric grounding in part from the antecedent referent 'this gorgeous house' in the preceding text, and in part from generic-lexical knowledge of the frame 'house' and its sub-component 'living room'. Likewise in (11c), both definite referents 'the waiter' and 'the menu' receive their grounding in part from the antecedent referent 'a restaurant' in the preceding text, and in part from generic-lexical knowledge of the frame 'restaurant' and its sub-components 'waiter' and 'menu'. Frame-based referential access is often accomplished through conventional knowledge of whole-part, possessor-possessed relations: (12) Whole-part, possessor-possessed frame-based access: a. She grabbed the fish and chopped off its head. b. John just got a job working for his father. c. The house was a mess, the roof leaked. d. She's upset. Her kids keep flunking highschool. e. The table is missing one of its legs. f. My wife called and said... g. Your house is on fire. In (12f, g), the anaphoric antecedent is not accessible from the current text itself, but rather from the speech situation, i.e. the identity of the speaker and hearer. But the use of frame-based knowledge to affect full grounding of the. definite referent is of the same type. 2.3.3.4. Grounding into the Current Text By far, the bulk of the grammar of anaphoric grounding involves access to anaphoric traces in the episodic mental model of the current text. Out of this vast array, one can separate several clusters of devices that seem to specialize in cuing access to rough locations - or configurations - in the mental text structure. One can illustrate the use of such devices b y citing their characteristic anaphoric gap - or referential distance (RD). This purely heuristic measure records the gap - in number of clauses backward - between the referent's current text-location and its last previous occurrence.

DISCOURSE STUDIES

268

(13) Comparison of mean referential distance (RD) values and degree of categorial distribution for common anaphoric devices (from Givon, ed. 1983a; 1983b; 1984b; Sun and Givon, 1985)

construction

mean RD (# of clauses)

degree of clustering around the mean

a. zero anaphora

1.0

100% at mean

b. unstressed PRO

1.0

95% at mean

c. stressed PRO

2.5

90% between 2-3

d. Y-movement

2.5

90% between 2-3

e. DEF noun

7.0

25% at 1.0 35% scattered 5.0-19 40% at 20+

f. DEF-Noun with modifierts)

10.0

55% scattered 5.0-19.0 45%at>20

g. L-dislocated DEF-N

15.0

60%at>20 (25% at 4-9) (13% at 10-19)

O n e must emphasize that referential distance is primarily a heuristic measure of convenience. Distance b y itself is not necessarily of great cognitive significance. Rather, distance tends to coincide with s o m e mental entities - b o t h structures a n d operations. T h e m a i n division in (13) is p r o b a b l y between devices that signal m a x i m a l continuity (13a, b) a n d those that signal discontinuity (13c-g): m a x i m a l continuity

== = == >

discontinuity

===== >

zero anaphora, unstressed pronoun stressed p r o n o u n , full lexical n o u n

As a n example of h o w this major distinction is deployed in text, consider the following passage of fiction. Of the referents that play any significant role in the narrative, continuous ones are bold-faced, and discontinuous ones are italicized: 11

12

(14) "...He circled it wearily as a wolf, [ 0 ] studying it from all angles, and when finally he stopped within a dozen feet of the dead man, he knew much of what had happened at this place. The dead man had ridden a freshly shod horse into the playa from the north, and when [ 0 ] shot he had tumbled from the saddle and the horse had galloped away. Several riders on unshod ponies had then approached the body and out had dismounted to [ 0 ] collect the weapons..." Cognitively, the maximal-continuity anaphoric devices signal the default choice of c o n t i n u e d a c t i v a t i o n of the current topical referent. F r o m the perspective

GIVON


269

u n d e r the same thematic chain-node. T h e topical referent is the n o d e label of the thematic chain. Discontinuous anaphoric devices, o n the other h a n d , signal the t e r m i n a t e d a c t i v a t i o n of the current topical referent, a n d activation of another topic for which there exists a n anaphoric mental trace. A m o n g such devices, the shortdistance devices (13c, d) tend to signal the activation of another referent without terminating the current chain. A n example of such a use of a stressed p r o n o u n can b e seen at the e n d of (14) and its continuation: (15) "... Several riderson unshod ponies had then approached the body and onehud dismounted to [ 0 ] collect the weapons. The clothinghaA not been stripped off, nor was the body mutilated..." T h e full indefinite N P 'several riders' in (15) terminates the activation of the preceding chain awaits topical referent ('the dead m a n ' ) , a n d opens a n e w chain, becoming its topical referent. T h e stressed p r o n o u n ' o n e ' then singles out o n e m e m b e r of this g r o u p briefly, b u t this is d o n e without terminating the thematic chain. T h e chain is then terminated, returning implicidy to the earlier topic 'the dead m a n ' . T h i s is d o n e indirectly b y d o u b l e - g r o u n d i n g the frame-based reference, via 'the clothing' a n d 'the body'. As a n example of the use of the other short-distance discontinuous referent device, Y-movement ('contrastive topicalization' (13d)), consider: (16) "...After buying a copy of The Racing Times at a Manhattan newsstand, Ed Piesman, a dentist and an avid horse-racing fan, said he preferred the new paper to the venerable Daily Racing Form because "the columnists are much better" and some of the statistics "are much much better". But the news dealer who sold him the paper said people like Dr. Piesman are few. "The Racing Time, I sell two a day, maybe five when there is a big race", said Ashok Patel, whose stand at 72nd Street and Broadway is adjacent to an off-track betting parlor. "The Daily Racing Form, I sell 40 or 50 a day"..." 13

Two topical referents persist throughout the paragraph, ' T h e Racing T i m e ' a n d 'Daily Racing Form'. W h e n they are switched for contrast, the object-fronting Y-movement construction is used - without terminating the p a r a g r a p h . Several long-distance anaphoric devices are characterized b y the following conflation of functions: 14

(a)

A •

Like other full-NP definites, they are topic-switching devices, deactivating the current t o p i c / n o d e . (b) Their anaphoric topic is re-activated after a long gap of absence (i.e. of cognitive de-activation). (c) T h e mental trace of the re-activated topic in the episodic m e m o r y is found in another chain, across at least o n e chain b o u n d a r y , but n f i w across a p a r a g r a p h bomr!*'"

270

DISCOURSE STUDIES

As i l l u s t r a t i o n of a l o n g - d i s t a n c e a n a p h o r i c d e v i c e , c o n s i d e r t h e use of L-dislocation (13g) in informal spoken A m e r i c a n English: 15

(17) H:

T: H: T: H: T: H: T: H:

...Well my dad was born in Sherman, that's close to where [...] is. He was born in Sherman in 1881, and he died in '75. Yeah. And ah, so, ah of course, my great grandfather, they came in there, I think, y'know, part of them from Tennessee and part of them from Illinois. And I don't really know much about that far back, Tom. But my grand-dad, he was a hardshelled Baptist preacher, and he just, y'know, farmed and ranched. In Texas? Yeah, yeah. So he was already in Texas? They must've come there when he was small, y'know, 'cause he spent... Your great grandfather moved and your grandfather was really raised in Texas. Yeah, yeah. In other words, about three generations of us... were in Texas... In Texas... And of course we eh, my dad, all he ever did was farm and ranch...

S o m e g r a m m a t i c a l devices are better described as t h e m e switching or p a r a g r a p h - i n i t i a l devices. R a t h e r than coding the re-activated topic, they signal the beginning of a n e w thematic paragraph. Consider, for example, the use of fronted adverbials i n : 16

(18) "...Their trail when they left Wells' body lay in the direction he himself was taking, and that meant the waterhole was off-limit for Shalako unless he wished to fight them for it, and no man in his right mind started a fight with Apaches. When the time forfightingcame, the man Shalako fought with, a cold fury that had an utter impersonal quality about it. He fought to win, fought with deadly efficiency, with no nonsense about him, yet he did not fight needlessly. Despite his weariness and that of the horse, he began backtracking. Peter Wells was not likely to be alone, so his presence indicated a camp nearby, and a camp meant water. Yet Shalako puzzled over his presence there at such a time..." Both n e w paragraphs in (18) are introduced b y pre-posed ADV-clauses. I n both, thematic discontinuity overrides referential continuity, so that in spite of the con tinuing m a i n participant ('Shalako', 'he'), a n e w thematic paragraph is initiated. 17

2.4. The Extreme Bounds of Coherence It is possible, at least for those w h o interpret coherence as a n objective p r o p e r t y of the external text, to view coherence as akin to 'grarnmaticality' in the Chomskian sense - a text either has it or doesn't h a v e it. Within the framework of network like mental text structure I a m proposing, it is m o r e useful to view text coherence as a matter of degree. O n e could then set u p the theoretical extreme u p p e r and lower b o u n d s of coherence, t h e n see h o w various kinds of text stack nr> vis-a-vis

GIVON


271

A useful description of the extreme b o u n d s of coherence can b e extrapolated from Wittgenstein's (1918) discussion of tautology a n d contradiction in logic. T h e u p p e r b o u n d s of coherence is logical t a u t o l o g y , as in: (19) John came home, John came home... I n 'text' (19), all n o d e s in the second clause - subject, v e r b , locative - cohere maximally with corresponding n o d e s in the preceding clause. T h e second clause is t h u s m a x i m a l l y g r o u n d e d , a n d t h e text m a x i m a l l y c o h e r e n t - b u t also maximally r e d u n d a n t . O n e type of lower b o u n d s of coherence is the logical sense of c o n t r a d i c t i o n : (20) John came home, John didn't come home... I n 'text' (20), all nodes in the two clauses seem to c o h e r e maximally - except for the crucial m o d a l n o d e of truth value, which thus m a k e s the two clauses logically incompatible as m e m b e r s of t h e same text. T h e m o r e c o m m o n cognitive, pragmatic sense of incoherence involves the absence of recurrent element(s) across the text, as in: 18

(21) John went to Italy, the cow jumped over the fence... Most coherent - interpretable - texts fall s o m e w h e r e in the m i d d l e b e t w e e n the two extremes of total r e d u n d a n c y a n d utter incoherence. I n m o v i n g across adjacent clauses, o n e e n c o u n t e r s s o m e r e c u r r i n g a n d s o m e n o n - r e c u r r i n g elements. But neither the a m o u n t of n e w (=disjointed) information n o r the amount of old (= connecting) information in a single clause is totally uncon strained. T h e constraint o n the a m o u n t of n e w information i n t h e clause m a y b e given as (Givon, 1975; Pawley a n d Syder, 1983; Chafe, 1986, 1987a): 19

(22) constraint on new information: "Clauses in natural text tend to have only one chunk (usually a word) of new information per clause". If one accepts the clause as the incremental unit of processing n e w textual inform ation (Givon, 1984a, ch. 7; Chafe, 1986,1987a, 1987b), then (22) c a n b e viewed as a constraint o n the a m o u n t of n e w information that c a n b e a d d e d to mentallyrepresented text during a single processing increment. T h e second constraint, at the other e n d of the scale, governs the a m o u n t of old information in the clause, a n d m a y b e given as (Givon, 1984a, ch.7; Givon, 1990a, ch. 2 0 ; Chafe, 1987a): (23) constraint on old information: "Clauses in natural text tend to have at least one chunk (usually a word) of old i«fn«-"*«« *

272

DISCOURSE STUDIES

T h e most c o m m o n element of old (=connecting) information in the clause is its m a i n topic, most often the subject; the vast majority of clauses in natural text have subjects; a n d the vast majority of subjects are definite - i.e. connecting information. T h e r e are good reasons to suspect that principle (23) represents the m i n i m a l g r o u n d i n g r e q u i r e m e n t for coherent interpretation of the clause, rather than an optimal level. T h e n u m b e r of grounding connection in a n average clause is p r o b a b l y larger, a n d referential coherence is only one grounding connection in the clause. If multiple grounding of a clause is indeed the n o r m , then this n o r m can b e cast as a principle of grounding, coherence a n d mental access: 20

21

(24) Multiple grounding, coherence and mental access: "The more grounding connections the clause has, the more mentally accessible it is, and thus the more coherent it is relative to the text in which it is embedded".

3. Knowledge-driven vs. Grammar-cued Coherence 3.1. Preamble I n a recent review, Walter Kintsch (in this volume) has suggested that two parallel processing channels are active simultaneously during text comprehension. O n e , the "strong m e t h o d " , is driven b y domain-specific knowledge contained in the l e x i c a l information (nouns, verbs, adjectives, adverbs) i n the clause. T h e other, the "weak m e t h o d " , is driven b y the g r a m m a t i c a l information in the clause. Kintsch contrasts the two channels as follows: "...We h a v e finally found a use for syntax in a psychological processing m o d e l . It provides the c o m p r e h e n d e r with a "weak" b u t general m e t h o d for comprehension, to b e c o m p l e m e n t e d b y the "strong" knowledgebased a n d domain-specific m e t h o d s . A s in p r o b l e m solving, weak and strong m e t h o d s h a v e their respective advantages a n d uses, a n d the complete c o m p r e h e n d e r would n o t forego either..." (Kintsch, i n this volume, p . 158) Perhaps a m o r e apt characterization of the contrast between grammar-driven and knowledge-driven text comprehension is that of "rough-grained" vs. "fine grained". O n e m a y illustrate the relation between the two processing modes with an analogy from transportation. Suppose you are heading toward a particular house on a particular street in a particular n e i g h b o r h o o d of a distant city. You travel first on the freeway - fast, efficiendy, with relatively little conscious attention - until you reach the city, your rough-grained destination. N o w you must exit

GIVON

COHERENCE I N TEXT VS. COHERENCE IN MIND

273

attending carefully to traffic signals a n d street n a m e s , until - slowly, laboriously - y o u reach your fine-grained destination. 3.2. Grammar as an Automated Discourse-processing Mode O n e must n o t e that text comprehension - the construction of a coherent mental representation of the current text - n e e d not rely o n g r a m m a r cues at all. It is well established that there exist two m o d e s of discourse processing in h u m a n language - the pre-grammatical and the grammatical. T h e pre-grammatical m o d e is b o t h ontogenetically a n d phylogenetically prior (Givon, 1979a, ch. 5; 1989, ch. 7; 1990b; Blumstein and Milberg, 1983; Lieberman, 1984; Schnitzer, 1989; Bickerton, 1990). I n Table (25), the two processing m o d e s are c o m p a r e d in terms of their structural, functional a n d cognitive properties. (25) Pre-grammatical vs. grammatical discourse processing properties STRUCTURAL: a. grammatical morphology b. syntactic constructions c. use of wordorder: d. Pauses: FUNCTIONAL: e. processing speed: f. Mental effort: g. Error rate: h. context dependence: COGNITIVE: i. Processing mode: j. acquisition: k. evolution:

grammatical mode

pre-grammatical mode

abundant

absent

complex/ embedded grammatical (subj/obj) fluent

simple/ conjoined pragmatic (topic/comment) halting

fast

slow

effortless lower lower

laborious higher higher

automated

attended

late late

early early

T h e slow, analytic pre-grammatical m o d e of discourse processing is heavily vocabulary driven. This tallies with the fact that vocabulary is acquired before grammar, in b o t h first a n d second language acquisition. T h u s , pre-grammatical children, adult p i d g i n speaker a n d agrammatical aphasics c o m p r e h e n d and produce coherent connected discourse, albeit at slower speeds a n d higher error rates than those characteristic of grammatical l a n o n t o m . A coherent ™" "~ a

274

DISCOURSE STUDIES

(26) [anticipating a trip] In atnga. Sit dawn. tan ki. in airplane sit down turn key '(We'll go) in the airplane, sit down, turn the key, Vruum vruum! Tan tu da rayt. Atnga! vr. vr. turn to the right airplane And vroom vroom! (We'll) turn to the right. Airplane (flies)!' As a n e x a m p l e of coherent adult second-language pidgin, consider: (27) "...oh me?...oh me over there... nineteen-twenty over there say come... store me stop begin open... me sixty year... little more sixty year... now me ninety... nah ehm... little more... this man ninety-two... yeah, this month over... me Hawaii come-desu... i nineteen seven come... me number first here... me-wa tell... you sabe gurumeru?... you no sabe gurumeru?... yeah this place come... this place been two-four-five year... stop, ey... then me go home... Japan... by-n-by... littie boy... come... by-n-by he been come here... ey... by-n-by come... by-n-by me before Hui-Hui stop... Hui-Hui this... eh... he... this a... Manuel... you sabe-Aa..."

23

1

'•

"

As a n example of coherent narrative produced b y a n agranuriatic aphasia patient, consider (Menn, 1990, p . 165): (28) "...I had stroke... blood pressure... low pressure... period... Ah... pass out... Uh... Rosa and I, and... ' friends... of mine... uh... uh... shore... uh drink, talk, pass out..." i. : "...Hahnemann Hospital... uh, uh I... uh uh wife, Rosa... uh... take... uh... love... ladies... uh Ocean uh Hospital . and transfer Hahnemann Hospital ambulance... uh... half'n hour... uh... uh it's... uh... motion, motion... uh... bad... patient... I uh... flat on the back... um... it's... uh... shaved, shaved... nurse, shaved me... ' uh... shaved me, nurse... [sigh]... wheel chair... uh... i Hahnemann Hospital... a week, a week... uh... then uh... strength... uh... mood... uh... up... uh... legs and arms, left side uh... weak... and... Moss Hospital... two week... no, two months..." 1

1

GIVON


275

lexical words in the clause. T h e s e lexical clues d o n o t disappear in the presence of grammar, in fluent adult speakers. Rather, as Kintsch (in this volume) suggests, vocabulary-guided inferences remains a parallel processing c h a n n e l alongside grammar-cued inferences. -»•

4. Local v s . G l o b a l C o h e r e n c e M u c h of the experimental psycholinguistic work o n text coherence has in fact centered o n vocabulary-driven global processes, i.e. Kintsch's domain-specific general knowledge inferences. This is especially conspicuous in the work o n causal inference (Trabasso a n d van d e n Broek, 1985; Trabasso a n d Sperry, 1985; Fletcher a n d B l o o m , 1988;), as well as i n earlier discourse-oriented w o r k (Rummelhart, 1975; Johnson-Laird, 1980; v a n Dijk and Kintsch, 1983; Morrow, Greenspan and Bowers, 1987; Glenberg et at, 1987). T h e same focus on knowledgedriven global coherence is apparent in the "story g r a m m a r " paradigm (Mandler, 1978; 1982; M a n d l e r a n d J o h n s o n , 1977; J o h n s o n a n d Mandler, 1980; Stein, 1982; Stein a n d Glen, 1979), or its n e a r equivalent "story schema" (Schank a n d Abelson, 1977; Wilensky, 1980, 1982; D e Beaugrande, 1982). In a recent review, M c K o o n a n d Ratcliff (1992) refer to vocabulary-driven global inferences as "elaborative inferences", "instrumental inferences" or "inferences from situation models". T h e y characterize this global approach to text coherence as follows: "...It is widely believed that readers automatically construct inferences to build a relatively complete mental m o d e l of the situation described b y the t e x t . . . " (1992, p p . 17-18) In outlining their niinimalist, localistic alternative, M c K o o n and Radcliff suggest that v o c a b u l a r y - d r i v e n global inferences are s o m e h o w "not automatically encoded" during text comprehension, a n d do n o t d e p e n d o n the immediate recall buffer. I n this, global inferences presumably contrast with local inferences, which are "automatically encoded" during text comprehension and d e p e n d o n the immediate recall buffer. Whatever the ultimate status of this distinction, it is worth noting that the sense of "automatic encoding" used b y M c K o o n and Ratcliff (1992) could n o t b e the same one we used earlier in referring to g r a m m a r as an "automated discourse-processing m o d e " . To begin with, the experimental textframes used b y M c K o o n a n d Radcliff have little to do with grammar-driven inferences. Rather, they contrast local vs. global vocabulary-driven inferences. Further, m a n y processes subsumed u n d e r lexical access, whether u n d e r spread ing activation (Collins a n d Quillian, 1969, 1972; Rips, S h o b e n a n d Smith, 1973; Smith et al., 1974; Collins a n d Loftus, 1975; Glass a n d Holyoak, 1975; Smith, 1978; inter alia) or u n d e r m u l t i p l e activation (Swinney, 1979) are said tn involve C O n s i r l e r a M ^ .»-- - L - i i

t

i :

;

j._ .

in /-i

i

i

„u

:„

GOODWIN

ACTION AND EMBODIMENT WITHIN SITUATED HUMAN INTERACTION

313

a posture that could allow h e r to pursue h e r turn at j u m p i n g further. O n an other level the gestural m o v e m e n t s e n h a n c e a n d amplify the indignant force of the action. I n brief, talk a n d gesture mutually elaborate each other within (1) a larger sequence of action and (2) an e m b o d i e d participation framework constituted through mutual orientation between speaker a n d addressee. It would seem that something like this set of c o n c u r r e n c y relevant semiotic fields is what is being pointed to b y the p h r a s e 'face-to-face interaction'. However, this is b y n o m e a n s a fixed array of fields. T h u s o n m a n y occasions, such as p h o n e calls, or w h e n participants are dispersed in a large visually inaccessible environment (e.g., a h u n t i n g party, or a w o r k g r o u p interacting t h r o u g h computers), visible coorientation m a y not b e present. I'll call some particular subset of possible fields that is being oriented to at a particular m o m e n t as relevant to the organization of a particular action a contextual configuration.

3 . C h a n g i n g C o n t e x t u a l Configurations W h a t h a p p e n s next provides the opportunity to investigate in m o r e detail h o w the shape of the current contextual configuration has consequences for the organization of action. As Carla says 'Quinto' in line 5, Diana looks down, moving h e r gaze away from Carla's face a n d gesturing h a n d , and toward the grid.

314

DISCOURSE STUDIES

T h e participation framework which p r o v i d e d a n essential ground for Carta's use of her gesturing h a n d is n o longer operative. W h e n D i a n a looks away, Carla finds herself in t h e position of looking a n d gesturing toward s o m e o n e w h o is n o w publicly dis-attending her. Such p h e n o m e n a demonstrate h o w a n y partici pation framework is an ongoing contingent accomplishment, something n o t u n d e r the control of a single party (who can at best m a k e proposals about t h e structure of participation that should b e operative at the m o m e n t ) , b u t rather something that has to b e continuously achieved through public displays of orient ation within ongoing processes of interaction. N o t only the gesture b u t also the action Carla is performing, the challenge to Diana, is called into question b y virtue of t h e w a y in which D i a n a is n o longer visibly acting as a recipient to it. Let m e note i n passing that here, unlike in some approaches to 'speech acts', action is being analyzed here as a multi-party inter active p h e n o m e n o n . Does Carla in fact analyze these events in this way? Does she treat what Diana has done as undermining h e r current action, a n d if so what can she do about this? As can b e seen in the transcript o n p . 401 without the slightest break in her fluent, dynamic production of speech Carla restates the argument she h a s just m a d e in a different w a y with a different kind of gesture. As Diana's h e a d moves d o w n w a r d Carla drops h e r gesturing hand. However, she n o w uses h e r foot to d o a deictic stomp at a place constituted b y the intersection of three different, mutually relevant, semiotic fields: • First, the place w h e r e D i a n a is n o w looking, the target of h e r gaze, a n d thus the place that she is visibly displaying to b e the current focus of h e r orientation a n d attention. • Second, o n e of t h e squares i n the hopscotch grid that is the focus of the current dispute, indeed the square w h e r e D i a n a threw h e r b e a n b a g • Third, a square that is explicitly b e i n g talked about within Carla's current speech. T h e structure of Carla's talk also changes i n ways that adapt it to this n e w configuration of orientation a n d gesture. I n lines 4 a n d 5 Carla used numeric expressions functioning as n a m e s to specify the entities being disputed: el Cuatro a n d elQuinto. Such language talked about these p h e n o m e n a , b u t did n o t in a n y way presuppose that the participants w e r e actually looking at the grid squares being talked about. T h o u g h available in the local scene, t h e grid was n o t being p u t into play as something that h a d to b e actively attended to a n d scrutinized i n order to properly constitute the actions in progress at the m o m e n t O n e could look elsewhere, a n d indeed this is precisely w h a t Carla a n d D i a n a b o t h did b y gazing toward each other, a n d it was this structure of mutual orientation that Carla exploited b y placing h e r numeric h a n d s h a p e s directiy i n Diana's line of sight. By way of contrast, after D i a n a looks down, Carla uses the deictic expressions ^ •

'->

1 — -«J—j-> /li

CL i \ j.

:C i.u

— i ; „ . . l — „ „ , , „ , „ „ „t-

GOODWIN


315

Such deictic expressions presuppose that their addressee is positioned to see what is being pointed at (which is being further specified b y the concurrent foot point), and indeed the entities being pointed to are located precisely at the target of Diana's gaze. Orientation to the grid is n o w an explicit, crucial c o m p o n e n t of the operations that have to b e performed to properly constitute the action currenuy in progress. T h e grid as something to b e actively scrutinized is n o w in play as a relevant semiotic field implicated in the organization of the actions of the m o m e n t in ways that it wasn't a m o m e n t earlier. Carla

f

i

y eser— es el quartro. Diana: | 2 | Diana: NN:

I— No- (uhmm) Pero este es el cuattro? N [— o. "Estas en el cinco

And you go in the

FOUR.

You don't go m the

FIFTH.

And that i— is the four L- No-

(uhmm)

But this is the four? Nr— 0 "This is the

DISCOURSE STUDIES

316

I n brief, what one finds within this single turn at talk is a switch from one contextual configuration to another. T h e second contextual configuration con tains a n e w semiotic field, the grid as something to b e looked at, that wasn't necessary for the first. Despite the addition of this field, most of the semiotic fields in play during lines 4 - 5 r e m a i n relevant. T h e w a y in which contextual configur ations are constituted through specific, somewhat contingent mixes of particular semiotic fields provides for the possibility of underlying continuity, even while relevant change is occurring (e.g., sets of fields can overlap from one configuration to another). Rather than replacing one perceptual world with an entirely different one, there is relevant change in a continuing contextual gestalt as configurations are reconfigured. Despite this continuity, the shifts that do occur are b o t h signifi cant a n d consequential for h o w participants build appropriate action. T h u s , the shift in focus to the grid that occurs h e r e also involves changes in the kinds of sign systems, in b o t h talk a n d gesture, used to refer to the entities being talked about. T h o u g h Carla is still pursuing h e r challenge, there has b e e n a change in context or m o r e precisely, the particular contextual configuration of relevant semiotic resources that are providing organization of the action of the m o m e n t . T h r o u g h the shift, Carla can pursue her argument in a different arena a n d this might b e quite relevant. Suppose that Carla a n d D i a n a are in fact counting a n d labeling the squares o n the grid in different ways (this in fact seems quite likely). If D i a n a is n u m b e r i n g squares from her left to right at the top of grid, while Carla is numbering t h e m from the b o t t o m of the grid, they are each labeling the squares where the b e a n b a g was thrown and where D i a n a is n o w standing differently (e.g. for D i a n a the b e a n b a g is in square four, while for Carla it l a n d e d in square five; see the diagrams o n the first data display). Simply telling D i a n a n o t to go in the fifth can b e quite ineffective if D i a n a believes that she didn't d o this, a n d in fact threw her b e a n b a g to square four. By physically stomping o n re levant squares while she n a m e s them, as h a p p e n s in lines 6 - 7 , Carla can visibly show exactly h o w she is counting and labeling the squares in dispute. T h o u g h the shift to the grid m a y have b e e n triggered b y Diana's dropping her gaze there, this provides Carla with a new way of making, and publicly displaying, the grounds for h e r initial argument. T h e most crucial property relevant to the organization of action displayed through w h a t h a p p e n s h e r e is reflexive awareness. Central to Carla's construction of action is ongoing analysis of h o w her recipient is positioned to co-participate in the interactive frameworks necessary for the constitution of that action. W h e n D i a n a looks away, Carla takes into account w h a t D i a n a is doing and reorganizes h e r action in terms of it (see also C. Goodwin, 1981). This reflexive awareness is n o t simply an 'interior' element of the mental processes necessary for defining the action (as it could b e analyzed for example within traditional speech act analysis), b u t a public, visible c o m p o n e n t of the ongoing practices used to build the action, something that leads to systematic, relevant changes in the shape of the action. Moreover, within this process the addressee, as a n e m b o d i e d actor in h e r own right, is as crucial a player as the speaker. 10

GOODWIN


317

is appropriate to specific contextual configurations. H e r e , Carla is able to refer to a n d identify the same entities - specific positions in the hopscotch grid - with a n u m b e r of different sign systems, each of which has quite distinctive properties. T h e s e include n u m e r i c linguistic expressions functioning as n a m e s (which do n o t require looking at the entity b e i n g referred to - lines 4 - 5 , though this can b e built into their structure through syntactic affiliation with a deictic expression lines 6-7), iconic h a n d gestures (which presuppose orientation toward the h a n d rather than the entity being described through the hand), deictic linguistic expres sions a n d deictic or indexical gestures (both of which m a k e relevant gaze toward the entity being pointed to). N o t all of these resources are relevant a n d in play at any particular moment. However, the ability to rapidly call u p o n alternative struc tures from a larger, r e a d y at h a n d tool kit of diverse semiotic resources, is crucial to the ability of h u m a n beings to demonstrate in the ongoing organization of their action reflexive awareness of each other a n d the contextual configurations that constitute the situation of the m o m e n t . Looking at these same p h e n o m e n a from another perspective, w e find that the analyst cannot simply take an inventory of all semiotic resources in a setting that could potentially b e brought into play, a n d use this inventory as a frame to describe a relevant context As these data demonstrate, not all possible and relevant resources are in play at any particular m o m e n t . I n d e e d what h a p p e n s here de pends crucially o n the way in which the grid replaces the h a n d displaying numbers a n d focus o n each other's face, as what is being oriented to at the m o m e n t . To describe the context we h a v e to track in detail the temporal unfolding of the interaction, while attending to what the participants themselves are constituting for each other as the p h e n o m e n a to b e taken into account for the organization of the action of the m o m e n t (see also Schegloff, 1993). We are thus faced with the task of describing b o t h the larger set of possibilities from which choices are being m a d e , a n d the w a y in which alternative choices from that set structure the events of the m o m e n t in consequentially different ways.

4. S e m i o t i c Structure i n t h e E n v i r o n m e n t A n o t h e r crucial c o m p o n e n t of this process is the hopscotch grid being talked about a n d pointed at. T h e grid differs radically from b o t h talk and gesture i n m a n y important respects. Unlike the fleeting, evanescent decay of speech, which disappears as material substance as soon as it is spoken (unless captured in an other m e d i u m such as writing or tape recording), the hopscotch grid has b o t h an extended temporal duration - it is there in exacdy the same form throughout the g a m e , a n d in the present case of a painted grid o n a playground, day after day for n e w games - a n d is built of concrete material so durable that it can sup port the weight of multiple actors j u m p i n g through it. Rather than constituting a mental representation, it is as corporeal, solid, a n d enduring as the ground the players are walking u p o n . H o w e v e r , it is simultaneously a thoroughly semiotic of-,.„^,—

t—j—J

=i.

-J

• « -

-

-

-

-

318

DISCOURSE STUDIES

that m a k e u p the g a m e are impossible i n a hypothetical 'natural environment' unstructured b y h u m a n practice, e.g. a field without the visible structure provided b y the gridlines. Simultaneously, the g a m e is just as impossible without embodi m e n t of the semiotic structure provided b y the grid in a m e d i u m that can b e actually j u m p e d on. T h e notion that the p r i m a r y focus for the analysis of h u m a n action should b e the isolated mental states of individual actors h e r e b e c o m e s impossible. As demonstrated quite powerfully in the w o r k of Hutchins (1995), h u m a n cognitive activity is situated within historically shaped social systems that encompass b o t h actors a n d crucial semiotic artifacts such as the m a p s n e e d e d to navigate ships. Like a m a p , the hopscotch grid does n o t p r o d u c e single actions (a particular 'out' for example), but instead provides a public framework for the constitution of diverse, game-relevant classes of action (outs, successful j u m p s , etc.). Moreover, the m e d i u m in which it exists is crucial for the specific kinds of action that actors perform, such as j u m p i n g through it a n d visually inspecting the feet of other actors to see if a line has b e e n stepped o n etc. H o w e v e r , t h o u g h constructed in a m e d i u m with properties quite unlike those of speech, the grid nonetheless parses its structure into relevant units that are comparable to those being picked out with the language structures used to refer it. T h u s the deictic terms used to talk about particular squares presuppose b o u n d e d entities (e.g. 'this' as visibly contrastive with 'that'), a n d some of the n u m e r i c expressions presuppose elements in a larger series of equivalent units, precisely the structure displayed visibly b y the building of the grid as a n ordered series of smaller identical squares. M o r e generally, once the grid as a relevant semiotic system is taken into account, our framework for the analysis of the organization of action encompasses not only sequences of talk a n d the bodies of actors, b u t also the material structure in the surround. Participants visibly attend to such graphic fields as crucial to the organ ization of the events a n d action that m a k e u p activity reflexively situated within a setting, and which contribute structure to that action. 11

5. E m b o d i m e n t i n Institutional a n d Scientific Practice Viewing action as something accomplished t h r o u g h the juxtaposition of diverse semiotic materials provides resources for specifying in detail precisely those semiotic materials that provide for the uniqueness of culturally situated activities. However, with this framework it is also possible to demonstrate h o w activities that might initially a p p e a r to b e quite u n i q u e a n d esoteric, such as the details of scientific work, are in fact built through use of far m o r e pervasive, indeed generic practices for the accomplishment of action within situated h u m a n interaction. To demonstrate this w e will n o w look at s o m e archaeologists using a Munsell color chart to code the color of the dirt they are excavating. 12

1 Pam: Okay that should be, wet enough. (\ z>\

0

GOODWIN

4 5 6 7 8 9 10 11 12


We're lookin at that right there? (0.3) Pam: Mmm, (0.4) Jeff: Much darker than tha:t. — Pam: Yeah, r I'm not-= Jeff: There Pam: =rmjusttryintaputitiBthe:re.= =eh hih an(h)ywhere. °hih heh huh

319

Jeff:

1

Munsell Book

Most of the same semiotic fields that w e saw in the hopscotch example are relevant here. Two participants using their bodies to sustain a participation frame work for c o m m o n , joint activity are clearly using b o t h talk a n d gesture to carry out some larger sequence of action. However, despite these sirnilarities, this scene has a n opaqueness that the hopscotch dispute hadn't. T h o u g h we can understand (1) the talk of the participants, which consists of quite simple, ordinary English phrases a n d n o t a technical vocabulary; (2) the frameworks of orientation b e i n g displayed through their bodies; a n d (3) crucial aspects of the gestures being used - for e x a m p l e that they are pointing gestures - w e don't quite u n d e r s t a n d w h a t the participants are doing. 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

Jeff: Pam: Pam: Jeff: Pam: Pam: Pam:

Jeff:

I'll take it. ((takes trowel)) (2.0) Down. (1.2) En this one. ((Points)) — (0.4) ((Moves Trowel)) nuhhh? (1.8) Or urat one? (whoops) ((Points)) (0.8) Fou:r. (0.8) Is it that? Na:That's- nottWhat was the Arowness of that? mmfih,

This sense of basic, recognizable interactive organization n r n n i n g smack into an o p a q u e wall, a d o m a i n of p h e n o m e n a which seems absolutely crucial to what the participants are doing, b u t which I d o n ' t understand simply b y speaking the same language or living in the same country, is w h a t has struck m e almost every time I've d o n e fieldwork in a n e w professional or scientific workplace. These difficulties, the gnawing sense of n o t being able to adequately understand what's happening, arise n o t from a general unfamiliarity with the community or setting. Instead, m y inability to understand what is h a p p e n i n g is m a d e apparent b v thp WAV

i n -urliirJ-*

''

-'

320

DISCOURSE STUDIES

access to. H e r e these two archaeologists are staring intently at a weird litde b o o k with holes, pointing to it a n d arguing as they m o v e a trowel with dirt u n d e r it. I n essence, the visible actions of the participants show that they are orienting to a new, semiotic field which is crucial to the local constitution of action, b u t which cannot b e understood without m o r e detailed knowledge of the setting and activities in progress. This same argument could b e m a d e about the documents, television a n d computer screens, and other tools that participants in other settings, such as an Airlines Operations R o o m , an oceanographic ship, a chemistry lab, etc. attend to in performing the distinctive work that constitutes those settings. Q u i t e clearly, ethnography is required (Cicourel, 1992). However, the parameters of that ethnography, what has to b e known, emerge from the visible organization of the activity in progress. T h e issue is n o t what life is like in general for archae ologists, b u t rather w h a t precisely is the structure of the specific semiotic fields a n d activity systems that are providing organization for the actions they are per forming in order to do the w o r k that constitutes their lifeworld. W h a t is going o n with that little b o o k and w h y is it so important? Note that though lodged within professional settings, these charts and documents have structural similarity to the hopscotch grid. I n all cases, the participants are using a semiotically struc tured built environment as a constitutive c o m p o n e n t of the actions in progress. I n brief, the opaqueness of this setting, the way in which its status as something lodged within a specific, technical profession emerges as consequential, is visible through the w a y in which the participants are attending to a special semiotic field as criterial for what they are doing, while in other respects continuing to m a k e use of m o r e general interactive resources for the organization of talk a n d action within h u m a n interaction. Their simultaneous orientation to, a n d use of, this semiotic field as a constitutive feature of the actions that give their work its professional character, m a k e s it relevant for researchers to include that field in their analysis if they want to c o m e to terms with its institutional properties, or even to get a h a n d l e on just what it is that these people are doing. Were this field to b e omitted, it would b e like describing the actions of Carla a n d D i a n a with out taking into account the g a m e they are playing a n d the resources that m a k e b o t h that game, a n d their arguments about it, possible (e.g. the ability to adapt to changes in an interactive participation framework b y using the grid to insistendy pursue a n action, while changing the structure of that action so that it remains visible a n d appropriate to the contextual configuration of the m o m e n t ) . Before proceeding to a brief discussion of what the archaeologists are point ing at, two additional points will b e briefly noted. First, it is possible to adequately c o m e to terms with m u c h , p e r h a p s most, of what is h a p p e n i n g in m a n y inter actions while leaving some fields o p a q u e a n d unanalyzed. A n audio tape of the hopscotch g a m e would preserve a great deal of the structure of talk-in-interaction. Indeed, this acceptance of partial opaqueness is not only a possibility for analysts, b u t one that is systematically exploited b y participants themselves in professional settings characterized b y asymmetries in knowledge a n d access to the resources that m a k e u p the setting. Patients n o t only can't see m a n y of the documents, instruments a n d representations that doctors focus o n during an encounter, but

GOODWIN


321

can see t h e m (e.g., a n electrocardiogram). Such asymmetries are central to work settings as m u n d a n e as checking out in a supermarket, where the clerk is focused on machines for ringing u p the transaction, while the customer stands b y with b o r e d resignation. H o w e v e r , in the archaeological sequence, all relevant par ticipants are using a specific field as a central resource for the actions that give their work b o t h its professional character a n d its local organization as a sequence of interaction. Second, what we h a v e to investigate emerges from the w a y in which the participants themselves display a particular field to b e consequential a n d relevant through the orientation of their bodies a n d the organization of their action. Rather than wandering onto the fieldsite as disinterested observers, attempting the im possible task of trying to catalog everything in the setting, we can use the visible orientation of the participants as a spodight to show us just those features of context that we h a v e to come to terms with if w e are to adequately describe the organization of their action. This has methodological as well as theoretical im plications. For example the participants' visible orientation provides a guide for what should b e included within the frame of the video image, and what materials should b e collected from the setting (e.g. the b o o k they are looking at) to facilitate subsequent analysis. '

6. T h e M u n s e l l Chart as a n Historically S h a p e d F i e l d for t h e P r o d u c t i o n o f A c t i o n T h e activity the archaeologists are engaged in is classifying the color of the dirt they are excavating. T h e y are doing this for a n u m b e r of different reasons. M a n y p h e n o m e n a of interest to archaeologists, what they call features, are visible only as color changes in the dirt they are excavating. For e x a m p l e the cinders p r o d u c e d b y an ancient hearth will leave a black stain a n d the decaying w o o d of a post a n d rubble used to hold it u p will p r o d u c e a tube of dirt with color systematically different from the soil around the post. T h e very activity of excavating features systematically destroys them. As dirt is r e m o v e d to dig d e e p e r the patterns of visible color difference are destroyed. In part because of this, careful records h a v e to b e kept of each stage in the ex cavation. T h e d o c u m e n t a r y proof that a feature existed is to b e found in n o t only photographs, but also records describing in detail h o w the color of the dirt surrounding a feature differed from that within i t These two student archae ologists are coding the color of the dirt in order to fill out one of the forms that tracks their excavation. To code the color of dirt, archaeologists use as a standard reference the subsection of the Munsell color chart (a tool used b y m a n y professions concerned with the accurate description of color) that covers the range of colors that will be found in soil. This is carried to the field in a small loose-leaf notebook, and this is what these archaeologists are looking at here. 13

DISCOURSE STUDIES

322

Feature

Coding Form e fan- ftaiienu

BOIL DESCRIPTION: A ZONE — UW*r Plan WW.

ijMMatt tan S4na«f

SobMil

(

{• clatfc few*

bm n

cn cn O

a: So ts

21

O H

3 w

I

O 2

348

DISCOURSE STUDIES

of the inference and its relation to the explicit text. These classes do not exhaust all of the potential inferences during comprehension, but they d o provide a n initial foundation for discussing the constructionist theory. T h e order in which the inference classes are listed in Table 1 is not altogether arbitrary. Inference classes 1, 2, and 3 are n e e d e d to establish local coherence, whereas inference classes 3 a n d 4 are critical for establishing explanations. Classes 4, 5, and 6 are important for establishing global coherence. Classes 7 through 11 are elaborative inferences that are not n e e d e d for establishing coher ent explanatory m e a n i n g representations. Classes 12 a n d 13 address the prag matic communicative exchange between reader and author. T h e order of listing the inference classes is weakly correlated with the a m o u n t of attention that the inference classes have received in the psychological literature. Referential infer ences (class 1) have received the most attention, whereas inferences about author intent and attitude (class 13) have received the least attention. T h e local inferences that are elicited b y one or two sentences h a v e received m o r e attention t h a n the global inferences. Inferences that are elicited from rich textual cues have received m o r e attention than inferences that are remotely derived from the text b y virtue of background knowledge structures. T h e researchers' preferences in the selection of inferences to investigate can b e attributed b o t h to methodology a n d to theory. T h e most familiar a n d defens ible research m e t h o d s i n experimental psychology carefully manipulate stimu lus texts, so there is a preference for investigating local text-bound inferences. Early theories in psycholinguistics, text linguistics, and reading focused o n infer ences that were derived symbolically from explicit linguistic elements (Brady & Berwick, 1983; Brown & Yule, 1983; Katz & Fodor, 1963; Kempson, 1977). These inferences were derived b y rules, m e a n i n g postulates, a n d compositional analy ses that stripped away most of world knowledge. It is very tedious a n d difficult to specify the n u m e r o u s knowledge structures associated with text, so theorists a n d empirical researchers h a v e normally avoided a systematic analysis of world knowledge. Nevertheless, a mature theory of inference generation would n e e d to analyze world knowledge in detail. Consequently, some researchers have pursued rather ambitious projects that m a p out the world knowledge structures that participate in text comprehension (Alterman, 1985; Dahlgren, 1988; Graesser & Clark, 1985; M a n n e s & Kintsch, 1991; Schank & Abelson, 1977). Clearly, it is important n o t to confuse the likelihood that a class of inferences is generated on-line a n d the a m o u n t of attention that the class of inferences has received in the scientific literature. T h e r e is a n important distinction between text-connectinginfeiences and extratextual inferences (Graesser & Bower, 1990; Singer & Ferreira, 1983; Trabasso & Suh, 1993). I n the case of a text-connecting inference, the current clause being c o m p r e h e n d e d is related to a previous explicit statement i n the text; the previous statement is reinstated (i.e., activated or reactivated) a n d is inferentially linked to the current clause. T h e referential inferences (class 1 in Table 1) are always text-connecting inferences. I n the case of extratextual inferences, the inference is coDied or derived from freneric a n d sDecific knowledge structures that are

GRAESSER, SINGER A N D TRABASSO

INFERENCE GENERATION

349

T h e t a x o n o m y in Table 1 does n o t include some classes of inferences that normally are difficult to generate a n d are therefore off-line. First, there are logicbased inferences, which are derived from systems of domain-independent for m a l reasoning, such as prepositional calculus, predicate calculus, a n d theorem proving (Newell & Simon, 1972; Rips, 1990). Second, there are quantitative inferences a n d statistical inferences that are products of complex formulae a n d mathematical procedures (Kahneman, Slovic, & Tversky, 1982). These inferences are usually a struggle to generate (Bruner, 1986; Graesser & Clark, 1985; Kintsch, 1993; Schank, 1986), as everyone w h o has attempted to solve analytical brain teasers a n d algebra w o r d p r o b l e m s knows. Of course, this does not m e a n that these inferences are never generated during the comprehension of narrative text (Lea, O'Brien, Fisch, Noveck, & Braine, 1990). For the convenience of communication, we consider whether each class of inferences is in one of two discrete states: on-line versus off-line. However, there undoubtedly is a probabilistic continuum between on-line a n d off-line. T h e con tinuum can b e attributed to fluctuations in reader abilities, reader goals, text materials, samples of inferences, experimental tasks, a n d so on. T h e continuum can also b e explained b y the theoretical possibility that inferences are e n c o d e d to some degree rather t h a n all-or-none (Gernsbacher, 1990; Kintsch, 1988; M c K o o n & Ratcliff, 1992; Sharkey & Sharkey, 1992). T h e degree to which a n inference is e n c o d e d might b e strengthened or attenuated as m o r e information is received. Technically speaking, w h e n we claim that a class of inferences is gen erated on-line, our intention is to convey that it has a substantially higher strength of encoding or higher likelihood of being generated than the contrast classes of inferences that are off-line. A full-blown theory w o u l d account for likelihood of generation, the strength of encoding, the time-course of generation, a n d the exact locus of generation (within the text) for each class of inference.

A Constructionist T h e o r y of Inference G e n e r a t i o n As discussed earlier, the proposed constructionist theory emphasizes Bardett's (1932) principle of search (or effort) after meaning. This principle embraces the reader goal assumption, the coherence assumption, and the explanation as sumption. We elaborate o n these distinctive characteristics of the constructionist theory in this section. I n addition to these distinctive characteristics, the theory adopts some components, assumptions, and predictions that are widely accepted by psychological researchers in discourse processing (Gernsbacher, 1990;J u s t & Carpenter, 1992; Kintsch, 1988; Sanford & Garrod, 1981; Singer, 1990; v a n Dijk & Kintsch, 1983). T h e s e uncontroversial assumptions are succincdy enumerated in this section before turning to the theory's distinctive characteristics. Uncontroversial Components and Assumptions

350

DISCOURSE STUDIES

relevant background knowledge structures (both specific a n d generic), a n d t h e pragmatic context o{ the message (i.e., the author, reader, setting, a n d p u r p o s e of the exchange). 2. Levels of cognitive representation. Following Kintsch a n d his colleagues (Kintsch, 1988, 1992; v a n Dijk & Kintsch, 1983), three levels of code are con structed as a result of comprehension: the surface code (i.e., the exact wording a n d syntax), the textbase (explicit text propositions plus inferences n e e d e d for text cohesion), a n d the situation model. 3. Memory stores. T h e r e are three m e m o r y stores: short-term memory (STM; which holds the most recent clause), working memory (WM; which holds approxi mately the last two sentences, plus information that is actively recycled i n W M ) , a n d long-term memory (LTM). It should b e n o t e d that some models do n o t dis tinguish b e t w e e n S T M a n d W M . 4. Discourse focus. Attention m a y b e focused o n a n y of the three levels of cognitive representation (see assumption 2). F r o m the standpoint of the situation m o d e l , the discourse focus is analogous to a mental c a m e r a that scans across the scenario a n d zooms in o n particular characters, actions, events, spatial regions, a n d h o t spots (Bower, 1989). 5. Convergence and constraint satisfaction. Both the explicit information a n d inferences receive m o r e strength of encoding to the extent that they are activated b y several information sources (i.e., there is convergence), a n d they satisfy the conceptual constraints i m p o s e d b y the various information sources (Golden & Rumelhart, 1991; Graesser & Clark, 1985; Kintsch, 1988; M a n n e s & Kintsch, 1991; Sharkey & Sharkey, 1992; St. J o h n , 1991). 6. Repetition and automaticity. Repetition increases the speed of accessing know ledge structures a n d the elements within a knowledge structure. I n t h e case of an automatized package of knowledge (e.g., a familiar generic knowledge struc ture), the content is wholistically accessed a n d activated at little cost to t h e pro cessing resources i n W M . Distinctive Components and Assumptions of the Theory T h e p r o p o s e d constructionist theory embraces the search-after-meaning prin ciple (Bardett, 1932; Berlyne, 1949; Spiro, 1980; Stein & Trabasso, 1985). A t a general level, this search-after-meaning principle asserts that c o m p r e h e n d e r s attempt to construct meaning out of text, social interactions, and perceptual input. T h e principle has three assumptions that empower the constructionist theory with distinctive predictions: (7) satisfaction of reader goals, (8) achievement of b o t h local a n d global coherence, a n d (9) explanation of explicit information. It is important to emphasize that t h e search-after-meaning principle is an effort, not necessarily an achievement. I n most reading contexts, the fruits of these efforts are realized because writers-authors construct messages to b e understood. However, there are conditions that prevent the reader from constructing a mean ing representation that provides coherence, that explains the explicit text, a n d fViat c a t i c f i A C n S a r i a a W a r ' c e m n l c

Tf fn.A r o a r t a r f a i l c frn r i m / n f c * a n v l a f f n r t

f - n A n tn.*ac*»



351

insufficient time for comprehension. A globally coherent message is n o t con structed if the text is choppy, incoherent, a n d pointless. A n explanation m a y not b e generated w h e n there is a r a n d o m sequence of events or a puzzling sequence of propositions that satisfy a researcher's counterbalancing constraints. Inferences are not constructed without the prerequisite b a c k g r o u n d knowledge structures. 7. Satisfaction ofreader goals. W h e n readers c o m p r e h e n d a text, they are motiv ated b y one or m o r e goals. I n s o m e contexts, the goals are ill defined a n d gen eral. This is the case w h e n an adult reads the m o r n i n g n e w s p a p e r to b e c o m e informed about current events a n d w h e n a person reads a novel to b e entertained. T h e r e a d e r m u s t u n d e r s t a n d the m e a n i n g of the text u n d e r these conditions. I n other contexts, the reader's goals are specific. This is the case w h e n a r e a d e r tries to determine w h e t h e r it is a g o o d time to invest s o m e m o n e y in the stock m a r k e t (while reading a newspaper) a n d w h e n a reader tries to form a m e n t a l picture of w h a t a character looks like (while reading a novel). According to the r e a d e r goal assumption, readers are persistent in their attempts to satisfy their goals a n d there fore will construct inferences that address these goals. We distinguish a m o n g three levels of goal specificity: default, genre-based, a n d idiosyncratic. A t the first, m o s t undifferentiated, level, the reader's goal is to construct a meaningful situation m o d e l that is compatible with the t e x t T h i s level is the least constrained a n d is regarded as the default level. At t h e next level, there are goals associated with the genre of t h e text. Readers of fictive narrative h a v e the goal of b e i n g entertained in s o m e fashion (e.g., ex cited, enthralled, a m u s e d , or frightened), whereas readers of expository text h a v e the goal of being informed a b o u t events and facts in the real world (Brewer, 1980; Kintsch, 1980; Pearson & Fielding, 1991; Zwaan, 1993). Brewer (1980) has contrasted four "discourse forms" that reflect different goals of reading text: to entertain, to persuade, to inform, a n d to h a v e a n aesthetic-literary i m p a c t Zwaan's (1993) experiments o n reading a n d m e m o r y h a v e demonstrated that substantially different information is extracted from a text w h e n it is called a n e w s p a p e r article versus a literary story. T h e final level of goal specificity involves idiosyncratic r e a d e r goals. T h e s e goals m a y lead the reader to construct virtually any type of code (i.e., surface or textbase vs. situation model) a n d virtually a n y dimension of t h e situation m o d e l (e.g., spatiality, causality, plans, or traits of characters). A n adequate account of idiosyncratic goals would n e e d to incorporate a general theory of h u m a n motiv ation, which is quite outside the boundaries of our constructionist theory. T h e constructionist theory offers a priori predictions a b o u t inference generation in the context of default goals a n d genre-specific goals, whereas its predictions are ad h o c in the context of idiosyncratic goals. T h e a d h o c prediction is that infer ences are generated if they are directly relevant to the idiosyncratic goals. R e a d e r goals m u s t b e carefully analyzed in t h e experiments conducted b y researchers in discourse processing. T h e task d e m a n d s constrain the goals that readers adopt a n d therefore the inferences that they construct. Readers m a y n o t generate deep inferences, for example, if the tests, materials, and tn«k = „u~n— — • -- -

352

DISCOURSE STUDIES

For example, suppose that an experiment is conducted in which the reader's task is to recognize whether character A or B is activated b y a p r o n o u n she. Suppose further that the words are presented very quickly (250 m s p e r word), a n d the r e a d e r is asked to recognize which of two words h a d b e e n presented, as quickly as possible (namely, character A vs. B after presenting she). Shallow processing is sufficient to complete this task because (a) the task merely involves the recog nition of explicit text, (b) the material is quickly presented, (c) shallow processing is satisfactory for performance, and (d) the recognition test requires the discrim ination of only two characters. T h e reader might a d o p t a n artificial reading strat egy that is similar to monitoring a list of unrelated items. T h e reader might b e frustrated from having t h e n o r m a l comprehension goal of constructing a m e a n ingful situation model, in contrast, a meaningful situation m o d e l would b e con structed in a task that allows readers to read at their n o r m a l p a c e a n d that tests t h e m o n d e e p inferences relevant to the situation model. 8. Local and global coherence. T h e c o m p r e h e n d e r attempts to build a m e a n i n g representation that establishes local a n d global coherence a m o n g the events, actions, and states in the text. O n c e again, this does not m e a n that a coherent representation is necessarily achieved at all levels, however. If the reader believes that the text is n o t considerate a n d lacks global c o h e r e n c e , t h e n the com p r e h e n d e r setties for local coherence. If the reader believes that the text lacks local coherence, then the reader regards the text as incoherent. Stated differently, the reader attempts to construct the most global m e a n i n g representation that can b e m a n a g e d o n the basis of the text and the reader's background knowledge structures. A globally coherent cognitive representation is successfully achieved w h e n the following conditions are met: (a) the textual features support global coherence, (b) the reader has the prerequisite background knowledge, and (c) the r e a d e r does not have a specific goal that prevents understanding of the material. W h e n there is a b r e a k d o w n in one or m o r e of these conditions, the reader settles for local coherence or gives u p trying to achieve any coherence at all. C o h e r e n c e is achieved to the extent that elements and constituents in a text are conceptually connected b y virtue of b a c k g r o u n d knowledge structures, the constructed situation m o d e l , linguistic features of the text, or all three. T h e r e is a n imprecise b u t important distinction b e t w e e n local a n d global coherence. Local coherence is achieved w h e n conceptual connections relate the content of adjacent text constituents (i.e., a phrase, proposition, or clause) or short sequences of constituents. For example, in the story " H o w Leisure C a m e , ' ' there is a causal connection between the "man's haste a n d abstraction" a n d the event "he stuck a pickle fork into his right eye." Global coherence is achieved to the extent that most or all of the constituents can b e linked together b y one or m o r e overarching themes. For example, the t h e m e "haste makes waste" ties together most of the content of " H o w Leisure C a m e . " Features of the explicit text h a v e a p r o m i n e n t role in the construction of local coherence (de Beaugrande, 1980; Halliday & H a s a n , 1976; Kintsch & van Dijk, 1978; M a n n & T h o m p s o n , 1986). I n fact, the term cokesionis sometimes reserved for local connections that are b a s e d entirelv or Drimarilv on linenistir.



353

knowledge. Referential cohesion is established, for example, w h e n an explicit n o u n in a sentence is connected referentially to a previous n o u n phrase or p r o position in the text. A n o t h e r e x a m p l e of cohesion is w h e n a connective (e.g., because, so, and, therefore) explicitly links adjacent clauses with a particular type of conceptual relation. Nevertheless, features of language and the explicit text d o n o t go the distance in establishing local coherence. Sometimes readers n e e d to infer the relations between constituents. T h e classes of inferences that are important for establishing local coherence are referential inferences (class 1 in Table 1), case structure role assignments (class 2), and those causal antecedent inferences (class 3) that connect adjacent constituents. T h e r e is substantial evidence that the process of inferring these relations increases comprehension time (Bloom, Fletcher, van d e n Broek, Reitz, & Shapiro, 1990; Haberlandt & Bingham, 1978; Keenan, Baillet, & Brown, 1984; Myers, 1990; Myers, Shinjo, & Duffy, 1987; Singer, 1990). T h e establishment of global coherence involves the organization of local chunks of information into higher order chunks. For example, a moral, m a i n point, or t h e m e of a text (class 5 in Table 1) organizes m a n y of the events a n d episodes in narrative. A higher order chunk has its tentacles attached to con stituents that span large stretches of text. As a consequence, readers sometimes n e e d to link an incoming constituent to an excerpt m u c h earlier in the text; the earlier excerpt is in L T M but n o longer in W M . I n these situations, reinstate m e n t searches are executed to fetch the earlier text a n d to place it in W M . These reinstatement searches take additional processing time (Bloom et al., 1990; Dopkins, Klin, & Myers, 1993; Fletcher & Bloom, 1988; Kintsch & van Dijk, 1978; Singer, 1990, 1993; Suh & Trabasso, 1993; van d e n Broek & Lorch, 1993). Detailed analyses of global coherence are available in several fields: text lin guistics (Grimes, 1975; Halliday & Hasan, 1976), artificial intelligence (Dyer, 1983; H o b b s , 1979; Lehnert, 1981), education (Meyer, 1985), and cognitive psych ology (J. M . Mandler, 1984; v a n Dijk & Kintsch, 1983). Some global structures are contextually specific a n d detailed, such as a script for eating at a restaurant (Schank & Abelson, 1977). Others are abstract frames. I n one type of R E V E N G E structure, for example, characters A a n d B h a v e a positive b o n d , character C harms B for unjustifiable reasons, a n d subsequently character A h a r m s char acter C. Readers infer these global structures and thereby construct thematic inferences (class 5 in Table 1). Unfortunately, there h a v e b e e n very few empirical tests of w h e t h e r these inferences are g e n e r a t e d on-line (Seifert, M c K o o n , Abelson, & Ratcliff, 1986). Although it is b e y o n d the scope of this article to provide a detailed account of the representation a n d processing assumptions of the constructionist theory, it is worthwhile to point out briefly h o w coherence could b e achieved com putationally. C o h e r e n t representations have traditionally b e e n expressed in a symbolic form a n d h a v e b e e n constructed b y p r o c e d u r e s that m a n i p u l a t e symbols (Dyer, 1983; Fletcher, 1986; G o l d e n & Rumelhart, 1991; Graesser & Clark, 1985; Kintsch & van Dijk, 1978; Lehnert, 1981: Trabasso ™ A— u — ° CI.

inon.

«•••ord.=Now watch. I c'n put more up there for her them. (2.0) Bea: ,What, Barb: An in magic marker "so there. (0.6) Bea: Oh yeah, oh yeah.=5Ae was. she- w's she was in Rochele: house you know, and she said that um, that-1 heard her say um, (0.4) um um uh uh "Julia, said y'all been talking behind my back."=I said I'm a- I'm a say "H:oney, I'm glfl:d. that you know I'm talkm behind your back. Because /- because I meant for you to know anyway." An she said, I- said "I don't have to talk behind your back.=I can talk in front of your face too." II And she saidBarb: That's all I write. I didn't write that, /wrote that. (1.2) Bea: Over here, /write this-I cleared it off. Because Landa wrote and I- *h, rand / made it bigger. Barb: Mmm, (0.2) Bea: So she said, rThat firstBarb: And the other I did with my finger on the cars r and all that. Bea: An-so-/said, an- an so we were playin school you know at Rochele's house? And boy we tore her all- we said, I got uh y'know jl was doin some signs? Barb: I better not go around an catch Kerry. Bea: And Rochele called her baldheaded right - in-fronta-her face. She said "You Sa/rfheaded thing." Because she was messin with Rochele.=I said, and so she said, you know we were playin around with her? And she said "You ftflWheaded l

r DISCOURSE STUDIES

42

94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116

Barb: Bea:

Barb:

Bea: Barb:

Bea:

117 Bea: 118 Barb:

119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141

thing."=She said, "Rochele YOU DON'T LIKE IT?" I said I said rthat's whyYeah she gonna base in some little kid's rface. Yeah. And she said,// I said AND I SAID= I said I said "What~are~ya doin to her." I better not see Kerry today. I'm a say "Kerry / heard you was talkin bout me." Ias.ayTnen she gonna say "I ain't- What / say about you." I say "Ain't none yer business what you said.^You come say it in front a myface since what= you been tell everybody e k e . " (0.4) ((falsetto)) OO:, And I can put more and I'm a put some- some "bad words in today. (0.5) She said, and she was saying. .she saidWozo: n I got somp'm to write r about. /said, I better not catch you t'day.==I'm a

i '4 !rt5 166

TACTICAL USES OF STORIES

43

like that, 'h And she was leanin against-1-1 said, I s'd I s'd Is'd I said, "Hey girl don't lean against that , thing cuz it's weak enough." *h And she said and she said 'h she- she did like that.=She say, "Teh!" ((rolling eyes)) II like that I s'd-1 said "You c'd roll your eyes all you want to. Barb: Yeah if somebody do that to herAnd if .you know what? Bea: Cuz I'm tellm you. (0.5) Tellm- I'm not aslAn you." (0.4) An I ain't say no plea:se «ther. Barb: mm hmm. Bea: ((chewsfingers)) Barb: Don't do that. (1.5) W'll I'm tellin ya / better not catch Kerry today. Cuz if I catch her I'm gonna give her a wor:d from my mouth. (0.6) An if she jump in my / a c e I'm a punch her in her fa:ce. (1.5) And she can talk behind my ba:ck she better say somp'm in front of my face. (1.5) l

((Boy walks down the street))

lo7 Barb: OO: there go the Tack, 'h 'hh 'hh Eh ! n» That's your na(h)me. i*P (1.5) (70 ((Barbara starts down the street)) .171 Barb: °h See y'all. 172 Bea: See you.

Notes This chapter originally appeared in Discourse Processes 13(1990):1.35-71. 1. For a more complete description of this fieldwork see M. H. Goodwin (1990). 2. \ more extensive analysis of the formation of teams, as well as speech activities, occurring in this encounter appears in C. Goodwin and M. H. Goodwin (1990). 3. The complete cycle of storiesfromwhich these data were selected appears in Appendix C. Subsequent line numbers related to this example refer to line numbers in the expanded version of this story in Appendix C. 1 The series of stories from which these data were taken appears in Appendix D. Line numbers in this fragment correspond to line numbers in the more expanded sequence which this story is a part, which appears in Appendix D.

• i

44

DISCOURSE STUDIES

References

1

> i

Douvan, Elizabeth & Adelson, Joseph (1966). The adolescent experience. New York: Joh Wiley & Sons. Eder, Donna & Hallinan, Maureen T. (1978). Sex differences in children's friendships. American Sociological Review 43:237-250. Feshbach, Norman & Sones, Gittelle (1971). Sex differences in adolescent reactions toward newcomers. Developmental Psychology 4:381-386. Gilligan, Carol (1982). In a different voice: Psychological theory and women's developm Cambridge, MA: Harvard University Press. Gluckman, Max (1963). Gossip and scandal. Current Anthropology 4:307-315. Goffman, Erving (1974). Frame analysis: An essay on the organization of experience. New York: Harper & Row. Goodwin, Charles (1984). Notes on story structure and the organization of participation. In Max Atkinson &John Heritage (Eds), Structures of social action (pp. 225-246). Cambridge: Cambridge University Press. Goodwin, Charles (1986). Audience diversity, participation and interpretation. Text 6(3):283-316. Goodwin, Charles & Goodwin, Marjorie Harness (1990). Interstitial argument. In Allen Grimshaw (Ed.), Conflict talk (pp. 85-117). Cambridge: Cambridge University Press. Goodwin, Marjorie Harness (1980a). Directive/response speech sequences in girls' and boys' task activities. In Sally McConnell-Ginet, Ruth Borker & Nelly Furman (Eds). Women and language in literature and society (pp. 157-173). New York: Praeger. Goodwin, Marjorie Harness (1980b). Processes of mutual monitoring implicated in the production of description sequences. Sociological Inquiry 50:303-317. Goodwin, Marjorie Harness (1988a). Cooperation and competition across girls' play activities. In Sue Fisher & Alexandra Dundas Todd (Eds), Gender and discourse: Ih power of talk (pp. 55-94). Norwood, NJ: Ablex. Goodwin, Marjorie Harness (1990). He-said-she-said: Talk as social organization among bl children. Bloomington, IN: Indiana University Press. Goodwin, Marjorie Harness (1990-91). Retellings, pretellings and hypothetical stories. Research on Language and Social Interaction 24:263-276. Jefferson, Gail (1979). A technique for inviting laughter and its subsequent acceptance' declination. In George Psathas (Ed.), Everyday language: Studies in ethnomethodolo (pp. 79-96). New York: Irvington. Labov, William (1972). The transformation of experience in narrative syntax. In William Labov, Language in the inner city: Studies in the black English vernacular (pp. 354-39 Philadelphia: University of Pennsylvania Press. Lever, Janet (1976). Sex differences in the games children play. Social Problems 23: 478-487. Miller, Patrice, Danaher, Dorothy & Forbes, David (1986). Sex-related strategies for coping with interpersonal conflict in children aged five and seven. Developmental Psycholog 22:543-548. Sacks, Harvey, Schegloff, Emanuel A. &Jefferson, Gail (1974). A simplest systematic? for the organization of turn-taking for conversation. Language 50:696-735. Savin-Williams, Richard (1980). Social interactions of adolescent females in natural groups. In H. C. Foot, A.J. Chapman, &J. R. Smith (Eds), Friendship and social rela tions in children (pp. 343-320). New York: John Wiley and Sons. Sutton-Smith, Brian (1979). The play of girls. In Claire B. Kopp & Martha Kirkpatrick (Eds), Becoming female (pp. 229-257). New York: Plenum.

HARNESS GOODWIN TACTICAL USES OF STORIES

45

Tumci. Victor W. (1986). Dewey, Dilthey, and drama: An essay in the anthropology of i'Xj>erience. In Victor W. Turner & Edward M. Bruner (Eds), The anthropology of experience (pp. 33-44). Urbana, IL: University of Illinois Press. Vblosinov. Valentin Nikolaevic (1971). Reported speech. In Ladislav Matejka & Krystyna Pi imorska (Eds), Readings in Russian poetics: Formalist and structuralist views (pp. 149-175). Cambridge: M.I.T. Press.

i i

i

I

II

..

3

I .

Jl~

n

ITTTT.'

2

The Terms of Agreement: Indexing Epistemic Authority and Subordination in Talk-in-Interaction John Heritage and Geoffrey Raymond

S

ocial psychologists h a v e long n o t e d that social actions involve persons in c o m m i t m e n t s that others should recognize a n d validate. Perhaps the preeminent theorist of this perspective was Erving Goffman, w h o observed in " O n Face-Work," [W]hen a person volunteers a statement or message, h o w e v e r trivial or commonplace, h e commits himself and those h e addresses, a n d in a sense places everyone present in jeopardy. By saying something, the speaker opens himself u p to the possibility that the intended recipients will affront h i m b y not listening or will think h i m forward, foolish, or offensive in w h a t h e has said (1967:37).

A n d in Relations in Public, [L] et a participant w h o m others would rather see silent m a k e a statement, a n d h e will h a v e expressed the belief that h e has a full right to talk and is worth listening to, thereby obliging his listeners to give a sign, however begrudging a n d however m e a n , that h e is qualified to speak (1971:95). Goffman conceptualized these obligations in the concept of "face," which h e conceived as central to the organization of social interaction. This work was e x t e n d e d b y Brown a n d Levinson (1987); drawing o n D u r k h e i m ' s (1915) dis tinction between negative a n d positive rites, they distinguished between negative

HERITAGE AND RAYMOND

EPISTEMIC AUTHORITY AND SUBORDINATION

47

face the desire to b e unimpeded) a n d positive face (the desire for approval, appreciadon, or ratification). Brown a n d Levinson operationalized face-work inio a set of specifically linguistic strategies that e m b o d y connections b e t w e e n language use. social distance, power, a n d related variables. I n this way, con nections are m a d e between face processes e m b o d i e d in interaction a n d facets of sociaJ and cultural identity, b o t h Goffman's theory and Brown a n d Levinson's extension are at pains to distinguish these notions. Identity is specific a n d local to persons, groups, a n d cultures, whereas the desire to b e u n i m p e d e d a n d to b e re garded positively - the two central c o m p o n e n t s of face - e m b o d y putatively universal elements of h u m a n conduct and are conceived as basic to its organization. 1

Goffman's theoretical conception of face as situated within "the flow of events in the encounter" (1967:7) invited a n empirical focus o n sequences of talk-ininteraction; these, as Schegloff (1992) observes, are a primordial site of h u m a n socialiry. O n e domain of such research, generally recognized as converging with Goflman's concerns, is the m a n a g e m e n t of conversational actions, such as agree ment and disagreement, that are clearly consequential for social solidarity. Con versation analysts h a v e used the term preference organization to refer to the set of practices through which persons m a n a g e courses of action that either p r o m o t e or undermine social solidarity (Holtgraves 1992; Lerner 1996; Pomerantz 1978; 1984: Sacks 1987; Schegloff 1988, f o r t h c o m i n g ; Schegloff, Jefferson, a n d Sacks 1977). T h e most p r o m i n e n t organizational consequence of these prac tices is to maximize the likelihood of affiliative, socially solidary actions, a n d to minimize the consequences of disaffiliative, socially divisive ones (Heritage 1984a:265-80). In addition to the organization of preference, however, participants' concerns with face can b e found in the m a n a g e m e n t of rights a n d responsibilities related to knowledge and information. For example, conversationalists treat one another as possessing privileged access to their o w n experiences a n d as having specific rights to narrate t h e m (Pomerantz 1980; Sacks 1984); journalists distinguish between firsthand a n d derivative access to breaking news as relevant for the rights to describe it ( R a y m o n d 2 0 0 0 ; R o t h 2002); callers to 911 emergency ser vices report matters in quite distinctive terms if they are bystanders to an incident rather than victims (Whalen a n d Z i m m e r m a n 1990); a n d patients offer medical diagnoses to physicians only u n d e r relatively particular circumstances (Gill 1998; GUI and Maynard forthcoming; Heritage and Robinson forthcoming). I n each of these cases, the distribution of rights a n d responsibilities regarding what partici pants can accountably know, h o w they k n o w it, whether they have rights to de scribe it, and in what terms is direcdy implicated in organized practices of speaking.

2

In this paper w e consider sequences in which participants offer evaluative assessments of states of affairs. We focus o n h o w relative rights to perform these evaluations are indexed within the talk. Although these sequences are occupied mainly with agreement a n d are fundamentally affiliative, w e show that they can involve complex face considerations relating to the m a n a g e m e n t of knowledge and information. We distinguish between assessments that initiate a n assessment sequence as "first position assessments," a n d assessments that are designed to b e • M r u m t i u o t n t t i A Q P as "second oosition assessments." 3

48

DISCOURSE STUDIE^

I n sequences of interaction, first position assessments establish a represen rational field in which second assessments will b e found to position themselves b some fashion: through agreement, disagreement, or adjustment (Heritage 2002a Pomerantz 1984). I n this sense, first position assessments offer a terrain witbii: which agreement will b e sought. We propose that these assessments also carry an implied claim that the speaker has primary rights to evaluate the matter assessed For example, as we demonstrate, persons offering first assessments m a y work to defeat any implication that they are claiming p r i m a r y rights to evaluate the mai ter at h a n d . Conversely, persons w h o find themselves producing a responshv assessment m a y wish to defeat the implication that their rights in the matter art secondary to those of a first speaker. Because assessments are always produced in real time a n d are unavoidably p r o d u c e d as first and second positioned actions,, they bring unavoidable relevance to issues concerning relative epistemic rights to evaluate states of affairs. O u r primary objective is to explicate the m a n a g e m e n t of these rights, and the m e a n s b y which this m a n a g e m e n t is achieved. To do so, w e focus on the intersection b e t w e e n differential rights to m a k e assessments, turn design, and sequential positioning. As we demonstrate here, persons in the midst of joindy evaluating states of affairs are concerned n o t only with agreement, but also with w h o is agreeing with w h o m (Schegloff 1996a: 177) - or, as we shall have it, "the terms of agreement."

Assessments a n d T h e i r Epistemics T h e assessment of states of affairs generally requires some form of access to the state of affairs being assessed (Pomerantz 1984). I n (1), for example, a second speaker struggles to find a basis for affiliating with a first assessment whose very construction ("you sh'd sgjs that h o u s e E(h)mma") denies the access necessary for building agreement: (1) [NB:IV:10:2] 1 2 3

Lot:

[h h] J_eziz Chris' you sh'd s_e that house E (h) mma yih'av 4-no id_a.h[hmhh Emm: [I bet it's a drea:m.

E m m a ' s response in line 3 ("I bgt it's a drea:m.") projects an agreemem with Lottie's assessment of the house, while simultaneously thematizing her lack of firsthand experience. Lacking that experience, she m a n a g e s raw affiliation with Lottie's evaluation b y a n utterance that expresses, at best, a simulacrum of agreement. W h i l e (1) illustrates the limits of verbalized a g r e e m e n t without access, most assessment sequences incorporate the p r e s u m p t i o n of concurrent or serial joint access to a referent state of affairs. Such access can b e first-order and in> mfarliQto

oo

in

tO\*

HERITAGE A N D

RAYMOND

EPISTEMIC AUTHORITY A N D

SUBORDINATION

(2) [JS:il:28|

'

1 J: - > T's tsuh beautiful day out isn't it? 2 L: •-> Yeh it's jus' gorgeous . . .

'

• •

49

i -

Alternatively, it can be second-order and mediated, as in (3). Here J o n and Lyn are a couple who have seen the film Midnight Cowboy, while Eve has not: (3) |JS:I1:«1| 1 Jon; Kve: 3 L\ n: 4 F.ve: 5 6 I've: 7 Jr.n: 8 Lyn: 2

We saw Midnight Cowboy yesterday- or [suh- Friday. [Oh? Didju s- you saw that, [it's really good. [No I haven't seen it - > Jo saw it 'n she said she f- depressed her - > ter [ribly -> [Oh it's [terribly depressing. -> [Oh it's depressing.

As these two examples indicate, access to assessable objects is ranked. Firstorder a C C 6 S S entails rights to assess, which can outweigh the rights of secondorder access. In (2) the initial assessment of the weather is marked by its presenttense declarative form as drawn from direct and current experience; in this context of face to-face interaction, such experience is equally available to the recipient. The co interactant responds with an agreement ("Yen") and an upgraded second assessment with comparable features. In (3), b y contrast, Eve's assessment of Midnight Cowboy at lines 5 - 6 is marked as based in the account of a third party. Jon and Lyn's responsive assessments, however, renew the claim to direct, firstorder access to the movie (first asserted in line 1), and are marked by the "oh"prefaces as asserting stronger rights to assess it (Heritage 2002a). First and Second Position Assessments and Rights to Assess Participants orient to first and second position as involved in claims about rights to make assessments. This is most apparent in cases where there is some incom patibility between the epistemic rights that a speaker wishes to claim and the position in which the assessment is offered. In (4), for example, Norma offers a downgraded first assessment of a third party: (t |SBL 2-1-8:51

1 2 3 4 5

Bea: Nor: - > Bea: "for:

hh hhh We:ll h I wz gla:d she c'd come igo las'ni:ght= = Sh[e. seems such a n]i£e little [1 a dy ] [(since you keh) ] [dAwfl] ly nice l*i'l p*ers'n.thhhhhhhW£:ll,I[;j's ] [I thinjk fiyjyone enjoyed jus . . . t

Ju«* before this sequence, it emerged that Norma had met the person assessed in line I - a longtime acauaintance of Bea - for the first time (see Example (14)

50

DISCOURSE STUDIES

i

d o w n g r a d e d assessment "She. seems such a nice little lady" a n d Bea's declarelively asserted a g r e e m e n t "dAwf'lly nice l*i'l p*ers'n." By downgrading her claimed access with the evidential v e r b "seems" (Chafe a n d Nichols 1986),' N o r m a manages h e r initial assessment so as to defeat any epistemic priority thai might h a v e b e e n inferred sheerly from its first positioning. T h e reverse contingency is evident in (5) below. H e r e A b e a n d Ben are in the midst of a discussion p r o m p t e d b y Abe's a n n o u n c e m e n t that h e has acquired a Burmese c a t As emerged earlier in the conversation, Ben's knowledge of Burmese cats is second-order: it is derived from a neighbor's ownership of the breed: (5) [TCIIA:1] 1 2 3 4 5 6 7 8 9 10 11 12 13

Abe: Ben: Abe: Ben: Abe: Ben:

T

Abe: Ben: Abe:

It's the- only cat I ever saw that chased do:gs. (0.2) [ehhhhu [hhhuhh ] [(Hadda) [go out'n r] gscue a dog thet wz sight times bigger'n he wz th's [morning. [e- .hhhhhh Hurra: :y fer the B]irme:se. e-huh-huh- [heh- [heh-he:h ] [F : : : [ight'n fo]o:ls. (0.2) Pard'n, (0.3) .hhh They're fight'n fools those Burmese, Oh I know'e is.

-> ->

T h e target sequence of assessments (lines 12-13) emerges from Ben's effort (lines 6 a n d 8) to affiliate with A b e ' s illustration of his cat's fighting prowess. His clearly responsive "Hurra: :y fer the Burme:se." is appreciated with laughter, while his a p p e n d e d characterization of Burmese cats as "F:::ight'n foo:ls." is treated as problematic with "Pard'n,". A s a result, Ben's original phrasal assess m e n t , which initially was p r o d u c e d in second position to Abe's first positioned assessment (and therefore fitted to Ben's rights to assess Burmese cats), comes to b e p r o d u c e d as a fully sentential declarative assessment ".hhh T h e y ' r e fight'n fools those Burmese," (line 12). I n this context, A b e comes to treat Ben as asserting rights to assess Burmese cats that are equivalent to his own, a n d it is this putative claim of equivalence that A b e resists. T h r e e features of Abe's turn embody this resistance. First, h e shifts the terms of assessment from Ben's assessment of the breed in general ("those Burmese"), to the cat which h e owns a n d has primary rights to assess. Second, the "oh"-prefaced design of this assessment indexes his claim to p r i m a r y rights in this matter (Heritage 2002a). Finally his use of, and stress on, the w o r d know underscores his claims in this regard. As the talk in these sequences demonstrates, participants w o r k to manage the relationship between rights to assess and sequential position b y manipulating the design of the turns out of which their assessments are built. I n the following sections we examine speakers' resources for m a n a g i n g this intersection of rights 5


EPISTEMIC AUTHORITY A N D SUBORDINATION

51

Assessments in First Position: Resources for Indexing Epistemic Rights Speaker* can design first positioned assessments to create three basic forms that modulaie the extent of their claimed epistemic rights. Using these resources, produc :.TS of first positioned assessments may manage face claims by asserting the sociocpistemic rights associated with particular social identities, or b y deferring to the rights associated with the identities of others. Unmarkedfirst assessments. In unmarked assessments, speakers deploy simple declarative evaluations that claim unmediated access to the assessable. These utterances contain no language that either strengthens or weakens the declara tive claim that is made. For example, in excerpts (6) and (7), speakers assess their immediate experience. (6) |VIYMC 1:4] J ]: 2 R: - >

Let's feel the water. Oh, i t . . . It's wonderful. It's just right. It's like bathtub water.

(7) |NB VII:2] 1 Emm: 2 ->

=We're painting like ma:d in th' kitchen= =a:nd oh evrything's workin out SQ pretty here

In (8) the first speaker deploys similar resources to evaluate past experiences which were shared with the coparticipant: (8) |SBL 2-2-3:51 1 Chi: - > 2 Cla: ! Chi:

We:ll it was[fu:n Clai [re, ((smile voice)) [ hhh [Yea: : [:h,] [°M]m°

A similar pattern is evident in (9), where the assessment concerns a joint acquaintance: (9'i |NB:IV.7:-44] 1 A: - > 2 P:

Adeline's such a swell [gal [Oh God, whadda gal. You know it!

Across these sequences, the first speakers' declarative utterances (arrowed) flatly assert evaluations of states of affairs, and clearly do so on the basis of dir ect access to them. Downgraded first assessments. Contrasting with these are assessments designed to exhibit downgraded epistemic access to a state of affairs. Two main resources can be used to accomplish this. First are evidentials (also see (4) above), through which speakers mark mediated access to a referent, and thus downgrade the claims made by the accompanying assessment. In (10), for example, Hyla has been describing a play that she and Nancy have tickets to see:

52

DISCOURSE STUDffiS

(10) [HG 11:6] 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Nan:

->

Kinyih tell me what it's abouit?= [27 lines of description deleted]

Hyl:

Nan: Hyl:

=.hh En she's fixed up, (0.4) en she meets this gu:y, .hh a:n' yihknow en he's (.) rjlly gorgeous'n eez rilly nice en evrythi [:ng bud li]ke= [Uhhu:h,] =.hh He's ah .hh Hollywood (0.3) s:sta:r's son yihknow who wz a mista:[ke en they [put im in'n [Academy, ] [O o this [ s o u n d s [so goo:: ] _ : : [d? [school, .hh buh wai:t.='n then, .hhm (0.2) .tch en the: (w)- the mother's .hh sister is a real bigot.

i

Nan: - > Hyl:

It is clear from line 1 that H y l a is the better informed about the play. This asymmetry is preserved in Nancy's assessment ("Oo this sounds so goo::::d?'' at line 11; the evidential formulation of this underscores that the basis for her evaluation is premised o n the information provided in Hyla's description. A second means to downgrade an assessment involves the use of tag questions. By introducing an invitation to agree with the assessment as a feature of its surface syntax, such questions index a putatively secondary access to a referent relative to the coparticipant. I n (11), for example, the assessment refers to Vera's grand children, whose recent visit included a stopover at J e n n y ' s house.J e n n y ' s declaratively formulated assessment is m o d u l a t e d with a tag question that defers to Vera's rights to assess h e r own family m e m b e r s : 6

(11) [Rah 14:2] 1 2 3 4 5 6 7

Jen: Ver: Jen: - > Ver: Jen: Ver: Jen:

Mm[I: bet they proud o:f the fam'ly.= [Yg.:s. =They're [a luvly family now ar'n't [they. [°Mm:.° [They arei ye [s. [eeYe[s::, [Yes, Mm: AH they need now is a little girl tih complete i:t.

T h e tag question is positioned so as to invite response as the first matter to b e addressed b y the coparticipant. I n this way, J e n n y formulates her turn as (in the first instance) a question to b e answered rather than as a n assertion to bt agreed with; thus she cedes epistemic authority in the matter to h e r coparticipani T h e introduction of the invitation to agreement in the surface design of the utterance indexes J e n n y ' s position that Vera has primary rights to assess her o w n grandchildren. I n a similar case, N o r m a n and Ilene are dog breeders discussing the breeding potential of o n e of N o r m a n ' s younger dogs. At line 9, Ilene invokes a comparison with Trixie, another of N o r m a n ' s dogs. ^ . t

HERITAGE A N D R A Y M O N D

EPISTEMIC AUTHORITY A N D

53

SUBORDINATION

(12} [Heritage 1:11:4] 1 2 3 4 5 6 " 8 9 10 1)

He:

No well she's i

Nor:

le: Aor: le:

=uh[: [She wz a year: la:st wee:k. Ah yes. Oh well wi any time no:w [then. ] [Uh:::]:[m [Ye:s.= =But =Butshe[:'s( she [:'s ( __U [Cuz Jrixie started] so early [didn't sh [e, [ " O h : : [ye:s. = =°Ye:h°=

Jur:

He: Nor: [le: - > ôr:

->

lie:

litre's assessment asserts that the dog started breeding "early" and the fol lowing tag acknowledges Norman's primary rights to assess his own dog. Uj radedfirst assessments.Just as assessments can be epistemically downgraded telativ to a recipient's attributedly superior rights, so can they be upgraded. The primary resource for this task is the negative interrogative, as in (13): 7

113; |SBL:2-l-8:5] 1 Bea: 2 3 Nor: 4 Bea: 5

>ior:

6

Bea:

Wz las'night th'firs'time you met Missiz Kelly? (1.0) Me:t whoxm? Missiz Missi Kelly? Ye:s. hh[Yihkno]:w what

even Jean said she couldn't do eh uh she said she's ahv'z glad when they gs:. Yeh .h well of eourse you see Bill is so good wih th'm ez well is [11 [n'tl h11 [e:. |C..

f.kl [That's ri:ght yes. Similar to the repeats deployed in the prior excerpts, Vera's "That's right" response treats "confirmation" as the primary business of the response, before going on to agreement with "Yes." I n addition, the formulation "That's right" m o r e overtly takes a n epistemically authoritative stance in relation to J e n n y t h a n do the earlier partial repeats. I n these examples, speakers achieve epistemic alignment b y downgrading rights to assess in first position assessments a n d upgrading t h e m in second pos ition. T h e [confirmation + agreement] format is most commonly used in response to interrogatively formed assessments, particularly those deploying tag questions. This distribution can b e understood as a product of the specific set of resources that tag questions m a k e available and relevant. A n assessment with a tag question a p p e n d e d offers the recipient a n opportunity to disentangle confirmation and agreement as distinct activities in a responding turn. Speakers can simply agree (e.g., "Yes" or "Yes, they are"). Alternatively, b y inverting the order of a confir mation a n d an agreement token, speakers can treat answering a n d agreement as separable activities a n d can exploit their separation to assert their epistemic supremacy. Upgrading with "oh"-prefaced second assessments: A s e c o n d p r a c t i c e for epistemically upgrading second assessments is "oh"-prefacing. I n these cases, the change-of-state sense of "oh" (Heritage 1984b, 1998) is used to index epistemic independence a n d priority, relative to a first assessment (Heritage 2002a). In contrast to the practice of confirming in turn-initial position described above, which normally exploits an earlier tag question, "oh"-prefaced second assess ments are m u c h less constrained in the contexts of their occurrence. I n (20), as n o t e d earlier, N o r m a n is the owner of the dog "Trixie," w h o m Ilene evaluates (line 9) as having "started so early." This assessment is followed b y a tag question which downgrades h e r epistemic access to this information relative to N o r m a n ' s . (20) [Heritage 1:11:4] 1 2 3 4 5 6 7 8 9 10

He: No well she's si Nor: He: =uh[: Nor: [She wz aa;year: la:st wee:k. lie: we any time no:w [then. ] Ah yes. Oh well ' Nor: [Uh:::]: [m He: [Ye:s.= Nor: ==Butshe[:'s( Butshe[:'s( )] He: - > [Cuz Trixie started] so early [didn't sh[e, Nor: - > [°Oh : :[y :s.°=

11

Tin-

1

£

0

0

= YP-1I =

HERITAGE A N D RAYMOND


59

Simultaneous with the tag question, N o r m a n ' s " O h : : ye:s." asserts epistemic priority on the issue. "Oh"-prefacing h e r e conveys a "change of state of orien tation" in response to Ilene's assessment. This is a systematic w a y of claiming that a speaker has i n d e p e n d e n t access to, and already holds a position regard ing, the referent. "Oh"-prefacing is thus a resource through which a second speaker can convey that the assessment which follows is i n d e p e n d e n t of the "here and n o w " of current experience a n d the prior speaker's evaluation. It achieves this outcome through a "change of state semantics," which conveys that the first assessment has occasioned a review, recollection, a n d renewal of the speaker's previous experience a n d j u d g m e n t , a n d that this forms the basis for the second assessment (Heritage 2002a). As with [confirmation + agreement], "oh"-prefacing functions to convey superior knowledge of, a n d / o r rights to assess, the matter u n d e r discussion. T h e following case (21) vividly illustrates this usage. G a y is giving J e r e m y a German telephone number. After she has recited 11 digits, thus exceeding the norm (during the 1980s) for a (British) intracountry call,J e r e m y comments (line 13) on the length of the number, prefacing his c o m m e n t with "Gosh." This expression indicates that, for h i m , this is something new, notable, or surprising. H e r e G a y could have r e s p o n d e d with a simple agreement, which, as in (15) a n d (16) above, would h a v e conveyed that her agreement was g r o u n d e d in the "here a n d n o w " c o m m o n experience of an interminable telephone number. Instead G a y begins her response with " O h it doe:s": (21) [Heritage:0I:7:3] 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Gay: Jer: Jer: Gay: Jer: Gay: Jer: Gay: Jer: -> Gay: - >

So the Tnumber is (0.2) ohj, one oi^lQ_h one oh:, (1.0) Yeup, TFpur niine, (0.5) Right? §ev'n three, u-six o.aie?hh (0.6) Sev'n three: six o:ne? (0.3) Eiight niine, °Gosh°it goe:s (.) goes on'n on Qh it doej.s (Jermany dse.:s.

Two aspects of Gay's " O h it doe_:s" treat J e r e m y ' s r e m a r k as reviving a n earlier observation of the same type that she h a d m a d e independently of this occasion. T h e r e b y they convey that Gay, in contrast to J e r e m y , finds the length of the n u m b e r unsurprising. First, as n o t e d earlier, the "oh"-preface indexes prior a n d i n d e p e n d e n t access to this p h e n o m e n o n . Second, the partial repeat ("it does"), b y confirming rather than simply agreeing with J e r e m y ' s remark, underscores this stance (Stivers forthcoming). By these m e a n s G a y also m a n a g e s

DISCOURSE STUDIES

60

to indicate that she is a n "old h a n d " at p h o n i n g abroad. I n addition, G a y con tinues with a turn c o m p o n e n t that appears to b e designed to further suggest h e r expertise about foreign telephone calls. H e r postpositioned "adjusting" com p o n e n t ("Germany doe:s.") recalibrates the referent of h e r response from this particular telephone n u m b e r to G e r m a n telephone n u m b e r s in general, a n d also conveys a degree of prior knowledge o n the topic. Moreover, with its hint of a further contrast with telephone n u m b e r s in other foreign countries, it implies a still b r o a d e r expertise in placing telephone calls, abroad. S h o r d y afterwards, G a y underscores her expertise, i n f o r m i n g j e r e m y that the "ringing" sound o n a G e r m a n p h o n e sounds like a "busy" signal o n a British p h o n e (data n o t shown). A third case points to the use of "oh"-prefacing as a m e a n s of countering a recipient's u p g r a d e d claim of access to a referent that began as the speaker's informational preserve. By inquiring into Lottie's trip (to Palm Springs), E m m a casts h e r as having direct a n d i m m e d i a t e experiential access to its events that E m m a herself lacks. E m m a , however, meets Lottie's enthusiastic initial assess m e n t of the trip ("Oh:: Go:d wonderful E m m a , " ) with a n "oh"-prefaced, negative interrogative " O h idn'it beautiful do_:wn theire," which embodies a competitive claim to primacy in assessing the attractions of the location. (22) [NB IV.10.R-.1] 1 2 3 4 5

Emm: Lot: Emm: - > Lot: -> Emm:

.h THow wz yer irfcp. Oh" Go:d wojnderful Emm[a, [Oh idn'it beautiful do:wn theire, Oh::Jeeziz ih wz go:rgeous::. Wh't ani;ce Twut time'djih giti:n.Jst ali'i whal ago?

Lottie herself then responds to this assessment with a further a n d equally competitive "oh"-prefaced a g r e e m e n t that u n d e r s c o r e s h e r o w n claims to primacy in the matter. This competitiveness is also evident in the referent shifting that is part of the sequence. T h e sequence begins with Lottie's evaluation of h e r recent trip; E m m a ' s assessment at line 3 shifts the referent to its general locale, a knowledge d o m a i n to which E m m a has some claim. I n turn, Lottie's response at line 4 uses the past tense ("wz") to shift the referent of her assessment back to h e r own arena of expertise - the recentiy completed trip. I n sum, "oh"-prefacing can b e deployed in a wide variety of contexts b y a second or (as in (22)) subsequent assessor to assert epistemic i n d e p e n d e n c e a n d supremacy relative to a first. I n addition to competing with sheer "firstness" (as in (20) a n d (21)), it can also b e deployed (as in (22)) to m e e t a n d compete with epistemically upgraded prior assessments. Upgradingby usurping the firstness"ofa previous assessment Using two resources tag questions a n d negative interrogatives - speakers w h o are responding to an assessment in second position can assert their assessment as a "first positioned" evaluation. I n b o t h cases, interrogative syntax is deployed as a m e a n s of usurping a previous speaker's first positioned assessment with a n e w one that n o w com m a n d s the terms of agreement. With these resources, speakers effectively usurp the socioepistemic claims of others. ,


EPISTEMIC AUTHORITY A N D SUBORDINATION

61

Upgrading with tag questions: While tag-questions downgrade first position as sessments, their function is reversed in second position: there they upgrade second position assessments. I n second position, the [assessment + tag] format invites agreement to the position that is taken b y the second speaker, thus p r e e m p t i n g "first position" in the sequence. I n this w a y it upgrades the second speaker's claimed rights over the first with respect to the matter at h a n d . This usage is il lustrated clearly in (23). I n this sequence, J e n n y a n d Vera are discussing Vera's two grandsons after a recent visit to Vera, during which the children also visited J e n n y ' s h o m e . After some discussion of the children, J e n n y a n d Vera face a p o tential discrepancy regarding what they h a v e just agreed o n (in lines 4-6) w h e n each n a m e s a different child to completeJ e n n y ' s observation "he's a bright little boy." Almost as soon as this discrepancy b e c o m e s apparent, (that is, w h e n Vera, the grandmother, produces "Paul" in overlap with J e n n y ' s "little J a m e s , " lines 5 and 6), J e n n y immediately accepts "Paul" b y repeating his n a m e (line 5) a n d then offering several agreement tokens (in lines 5/7). Perhaps to counter the p o tential inference that she was simply "going along" with Vera in accepting "Paul" as the "bright" boy, J e n n y offers a negative assessment of J a m e s : "Yeh J a m e s ' s a little devil" (line 11). By initiating h e r turn with "Yeh", J e n n y builds it as a con tinuation of the prior sequence, effectively offering the assessment as confirmation that she "meant" Paul in the first place because she already viewed J a m e s as "a little devil." Ironically, in solving this p r o b l e m , J e n n y inadvertently creates trouble of a different kind: in conveying that she m e a n t to refer to Paul, J e n n y comes to p r o d u c e a declaratively formed, first position, negative assessment of Vera's other grandson, J a m e s . 9

(23) [Rahman: 14:1-2] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Ver Jen: Ver: Jen: Ver: Jen: Ver: Ver: Jen: -> Ver: Jen: Ver: - > Jen: Jen:

ehr: they readjer comics:'n eviythink yjhkn[o:w [Yeh: w' 1 I think he's a briight little boy: u [h:m [I_do= =1 [ittleja]: [ rnies, ] uh [Pau.:l.y.es.] [ PauJ, ] [mm- m] mm [Pau : _ 1, ] Mm:. [Yes. [Yes. (0.3) [Yes ( )] [YehJames's a little] divjl ihhh Theh heh [That[.huh .hh [h He:[James is a little bugger[isn'e. [YehYeah [(into) ev'rythi]ng.

Although she steps in to agree with J e n n y ' s assessment, Vera as the grand mother resists the putatively superior access entailed in such a first position assessment.

62

DISCOURSE STUDIES

This resistance is e m b o d i e d most forcefully in Vera's use of a tag question in hum-final position. By deploying an interrogative form as part of a second position assessment, Vera marks h e r turn as a "new" first pair part. T h e "fastness" of h e r assessment is also manifest in her re-use of ' J a m e s " (a locally initial reference form deployed in a locally subsequent position (Schegloff 1996c). Moreover, the r e m a i n d e r of h e r turn is designed as a full-form declarative that effectively disregards J e n r y ' s immediately prior reference to h i m , and, b y extension, her utterance. T h e status of Vera's utterance as a "first position" assessment is sub sequently accepted b yJ e n n y : she cuts off h e r initial response to Vera's c o m m e n t (line 15), a n d then redoes her responsive agreement at line 16 with a typeconforming "Yeah" a n d an elaboration that gives a potential (and decidedly pallid) specification of the "little bugger" as "into everything." I n this case, then, the tag question, b y inviting a response, positions Vera's evaluation as a first action to b e agreed or disagreed with. I n this way, it attenuates its responsiveness to J e n n y ' s initial evaluation, thus asserting Vera's rights in the matter. Upgrading with negative interrogatives: J u s t as negative interrogatives upgrade the epistemic claims e m b o d i e d in first assessments, they also can achieve this outcome for second assessments. Used in second position, negative interrogatives (like second positioned, tags) provide a putatively "new" first pair part for the previous speaker to r e s p o n d to, a n d thus attenuate their status as "second pos ition" assessments. I n the following case (discussed above), Margie's second as sessment downgrades the virtues of the assessed party ("Pat") while deploying a negative interrogative to assert u p g r a d e d rights to h e r opinion. (24) [NB VII: 1-2] 1 2 3 4 5 6 7 8

Emm: Mar: Mar: Emm: Mar: Emm: Mar: ->

=Oh honey that was a lovely luncheon I shoulda c.a:lled you s:soo [:ner but I: ] 1: []o:ved itJh wzjust delhghtfu [: 1. ] = [((f)) Dinu] [°( ) [Well] = =1 wz gla [d y o u ] (came). ] ['nd yer f: ] friends ] 'r so dairlimg, = =Qh:::[: it wz: ] [e-that P]a:t isn'she a do: [dl? ] [iY e]h isn't she pretty,

As we noted earlier, E m m a praises Margie's friend ("Pat") using a negative interrogative (line 7) to u p g r a d e b o t h h e r rights to assess the friend and, with it, the compliment that h e r assessment embodies. Perhaps to m a n a g e the receipt of this compliment in an appropriately d o w n g r a d e d fashion (Pomerantz 1978), Margie responds (line 8) with an initial type-conforming agreement a n d a significantiy weaker token of praise ("pretty" versus E m m a ' s "a doll") (Pomerantz 1984). At the same time, she asserts h e r o w n primacy in assessing her friend b y packag ing her response as a negative interrogative. A somewhat m o r e perverse case can b e found in a return to excerpt (22): (22) [NB IV.10.R:1] 1 Emm: 2 Lot:

.h THow wz yer tri:pOh:: Go:d wonderful Emm [a,

1

•


EEISTEMIC AUTHORITY A N D SUBORDINATION

63

As we noted above, Lottie is the projectedly authoritative evaluator of h e r trip. H e r initial evaluation, however, is intersected with an "oh"-prefaced negative interrogatively formed assessment By shifting the referent to the location, this assessment brings the topical focus of the sequence towards E m m a ' s generalized experience of Palm Springs rather than Lottie's m o r e immediate experiences there. I n these ways, an assessment initiated in response to Lottie's evaluation is designed to exert control over the o n w a r d trajectory of the sequence. Summary Above we described four practices with which speakers can upgrade the epistemic claims of second assessments relative to the claims e m b o d i e d in first positioned assessments. T h e s e four practices are of two types. T h e first two - [repeat/confirmation + agreement] and ["oh"-prefacing] - assert primary rights to assess b y embodying the claim that the position asserted was held already a n d independendy b y the second speaker. T h e second two - [statement + tag] a n d [negative interrogatives] - assert primary rights b y manipulating the sequence to "reclaim" the first position assessment slot a n d thereby the epistemic rights which accrue to that position. T h e s e two latter practices reflect an important w a y in which "first" a n d "second" position, though at first appearance lodged in the t e m p o r a l order, are n o t confined there. Rather, these positions can b e negotiated through practices of speaking that reflexively claim "first position" and "second position" (see note 10) as distinct from the r a w appearances of temporal ordering in the flow of conversation. Reviewing these practices, the first - [confirmation+agreement] - is relatively specialized to environments in which first assessments are d o w n g r a d e d b y using tag questions. For these reasons, this practice is normally used in circumstances where the speakers achieve alignment concerning their relative rights to assess, and therefore is relatively "mild." M u c h m o r e combative are the "all-purpose" resources afforded b y the use of "oh"-prefacing, tag questions, a n d negative interrogatives. W h e t h e r a first assessment is unmarked, upgraded, or downgraded, the change-of-state semantics of "oh"-prefacing asserts an i n d e p e n d e n t stance, and h e n c e epistemic authority, in relation to a n assessable. It can b e used in combination with other resources, a n d can b e deployed to address first assess ments regardless of their specific formulation. Tag questions in second position function b y asserting interactional p r i m a c y i n assessment sequences: their deployment of interrogative syntax formulates the second speaker's assessment as the axial one calling for a g r e e m e n t Finally, negative interrogatives are perhaps the strongest of the four practices. Like tag questions, they deploy interrogative syntax to assert interactional primacy in an assessment sequence, but they are significandy m o r e assertive. T h u s they c o m b i n e b o t h the i m p o r t of a declarative and the sequential implicativeness of a n interrogative as resources to wrest the initiative from the first speaker. Multiplex Deployments of These Resources and the Management ofFace Having developed an analysis of s o m e of the resources through which speakers

64

I

DISCOURSE STUDIES

n o w turn to three cases in which the terms of agreement are m o r e complicated. I n the preceding analysis w e relied o n examples involving a straightforward re lationship between the participants' p r i m a facie socioepistemic rights a n d the practices they deploy to m a n a g e t h e m as a way of establishing a relationship between the two. T h e r e is n o guarantee, however, that speakers will assert the rights to which they m a y b e entitled, n o r that recipients will align with or support the rights asserted b y them. Speakers m a y assert rights that are (or could be) contested; in some cases they m a y defer to a recipient with putatively subordinate rights. H e r e we consider three cases in w h i c h participants assert rights which are potentially problematic, or which b e c o m e a source of conflict or struggle. By explicating the resources through which such circumstances are prosecuted, we can discern the reflexive character of these practices for asserting relative rights to assess in sequences of turns, and thus the m a n a g e m e n t of face a n d identity issues within these interactions. T h e first of our cases involves one party's assertion of the primary right to assess a person that previously h a d b e e n treated as h e r recipient's primary right to assess. I n (25), Vera a n d J e n n y are discussing Vera's grandchildren after a recent visit. As we noted above, in the course of managing a potential discrepancy regarding whatJ e n n y a n d Vera h a v e just agreed o n (in lines 4-6), J e n n y comes to offer a negative assessment of one grandson, J a m e s (line 11), most likely to counter the inference that she was simply deferring to Vera's n a m i n g of Paul as the "bright little boy." Although Vera produces a similarly negative assessment of J a m e s (line 14), she t h e n complicates matters b y r e v e r s i n g herself a n d evaluatingjames's behavior positively (line 17-18). It is this reversal, a n d J e n n y ' s response to it, that we n o w address. This environment m a y b e especially prob lematic forJ e n n y : if she is to agree with Vera, she must produce a positive, second position assessment of J a m e s even though she has just offered a highly negative assessment. As in the sequence regarding Paul, J e n n y has a clear interest in indicating that her reversal is not "merely responsive" to Vera's, especially since she has lesser rights to evaluate the child. (25) [Rahman:14:l-2] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Ver: Jen: Ver: Jen: Ver: Jen: Ver: Ver: Jen: Ver: Jen: Ver: Jen:

ehr: they readjer comics :'n gvry think yjhkn [o:w [Yeh: w'l I think he's a bright little boy: u [h:m [J_do= =l[ittleja]: [imes, ] uh [Pay.: Lyes.] [ Pauxl,] [mm- m ] mm [Pau: : 1, ] Mm:. [Yes. [Yes. (0.3) [Yes( )] [YehJames's a little] divil ihhh Theh heh [That[.huh .hh [h He:IJames is a litfle bugger [isn'e. [Yeh-


16 17 18 19 20 21

Jen: Ver: - > Jen: - > Ver: Jen:


65

Yeah [(into) ev'rythi ]ng. [Mmdju. 'eez good] Jenny, 'e wz mischeevious but w-'e wz good. Oo 'e. wz beautiful here [wuztn't'ee. = [iYes. ='E wz very, well beha:ved.

Despite her putatively subordinate rights in the m a t t e r j e n n y (Une 19) strongly asserts rights to assess J a m e s b y (1) producing a n "oh"-prefaced assessment of James's behavior (2) focused o n his behavior at h e r house. She follows this assessment with (3) a tag question. Each c o m p o n e n t of J e n n y ' s assessment works against the potential inference that h e r positive assessment reflected anything other than the actual position she h e l d i n d e p e n d e n d y of Vera's immediately prior positive assessment. First, the "oh"-preface indexesJ e n n y ' s previous, inde pendent access to J a m e s . Second, J e n n y supplies the basis for this claim b y invokingJ a m e s ' s behavior " h e r e " (at h e r o w n house), in contrast with Vera's as sessment of his behavior in general. Finally, h e r use of the tag question casts h e r turn as a "new" first assessment, even though its production immediately follows Vera's positive evaluation of J a m e s . I n light ofJ e n n y ' s systematic deference, u p to this point, to Vera's primary rights to assess her o w n grandchildren, it appears that J e n n y ' s assertion of primary rights in this sequence is designed to defeat any suspicion that her current, positive assessment could b e motivated simply b y a desire to agree with Vera. Specifically, although h e r turn is p r o d u c e d in response to Vera's, she deploys a n array of practices to establish that the position she takes in that turn is held independently of the circumstances of its production. I n this case, then,J e n n y uses such practices for asserting her p r i m a r y rights to assess J a m e s (despite h e r previous deference to Vera) as a m e t h o d for m a n a g i n g the local, interactional circumstances created b y Vera's reversal. O u r second case offers a m o r e ambiguous deployment of these resources. This excerpt is taken from a conversation between Lesley a n d her elderly mother, in which Lesley offers a series of reports regarding (potential) acquaintances who are currently in the hospital or w h o h a v e just died. I n this case, although Lesley produces an extended description of Mr. Millbeck, M u m fails to recognize him. As Lesley proceeds with the report "anyway," she notes that h e was "still working" w h e n h e died at age 79. Yet despite Lesley's status as teller a n d h e r clearly intimate access to the deceased, she packages h e r first position assessment (in line 25) using a tag question that eschews any claim of epistemic primacy: (26) [HoltX(c)-l-l-l] 1 Les: 2 3 4 Mum: 5 Les: 6

And um (0.4) I Tdon't know T 'f you remember Missiz MOibeck th't use to go to ichu:rch. (0.8) °(Mi[ssiz)° [Uh: uh-he, wz the vicar's ward'n T anyway the die:d suddenly this week .hhh and he wz Tstill wo:r4-king.

66

DISCOURSE STUDIES

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

(0.3) Mum: Les: Mum: Les: Mum: Les:

Mum: Les:

(

[

.- ' )

[He wz Tseventy ni4-:ne, (0.3) My: wo:rd. (0.2) TY g: s [he: wz um [(You've got s'm Trea:l) workers down theire, heh He wz a p- uh: Yg:s. Indee:d .hh He wz a (0.2) .p a Tbuyer for the hoh- i-the tonly harse hair fact'ry left in England. Good (gracious). (0.3) And he wz their buyer, (•)

Mum: Les: Mum: Les: -> Mum: - > Mum: Les:

THm::: .t TtHrri:. So The had a good innimgs Tdid[n't -the. [I should say Iso: Tye: s. (0.2) tMaivelous. (0.2) .tk.hhhh f Anyway we had a very good evening To:n Saturdaiy

It is n o t clear whether Lesley's tag question indexes M u m ' s rights to j u d g e a person of similar age, o r simply acknowledges the self-evident character of her assessment (i.e., that a person w h o worked until the day h e died at age 79 h a d a "good innings''). M u m , however, matches the position taken b y Lesley b y using the [confirmation] + [agreement token] format to upgrade the rights that otherwise w o u l d accrue to h e r second position assessment. I n this case, then, Lesley m a n a g e s the potentially problematic circumstance of p r o d u c i n g a first position assessment regarding a state of affairs that b o t h she a n d M u m c a n claim rights to assess b y simply avoiding a n y claim of epistemic priority, despite h e r status as putatively the better-informed party. Finally, i n our third case, competitive positions regarding epistemic rights to evaluate Margie's friend are coopted into a deeply ambiguous process of agree ment. This example, as noted previously, is a n extended compliment sequence. After a generalized evaluation of Margie's friends as "so dairlimg" (line 5), E m m a ' s "e-that Pa:t isn'she a do::ll?" singles out o n e of the guests for particular praise. As observed earlier, the primacy embodied in this first position assessment is further upgraded b y E m m a ' s deployment of a negative interrogative format to package it. O n the one h a n d , this format (as n o t e d earlier) invites a n agreement p r o d u c e d as a response to a question. O n the other, however, the assessment and its format create two p r o b l e m s for Margie. First, as a question, t h e assessment invites confirmation, b u t as a compliment Margie's agreement is constrained } l v r n n v p n H n n c t r m m m i n i r o o l f . n i ' Q i p a ^ P n m o r o r i f r a 10*7Q\

. ~ U "17~~ — . ~



67

characterizes the persons present as "your friends" (line 5) - thus acknowledging Margie's primary rights to evaluate them - her assessment is packaged with a format which (as we have seen) asserts epistemic primacy and virtually com mands agreement. (27) [NB VII: 1-2] 1 2 3 4 5

Emm: = O h honey that was a lovely luncheon I shoulda ca:lled you s:soo [:ner but I;] 1: [Jg.:ved it.Ih wzjust deli:ghtfu [: 1. ] = Mar: [((f)) Ohiu ] [°( ) [Well] = Mar: =1 wz gla [d y o u ] (came).] Emm: ['nd yer f: ] friends ] 'r so da:rliing,=

6

Mar:

7 8

Emm: Mar:

9

= Q&.::[: it wz:

]

[e-that P ]a:t isn'she a do: [ill? ] [iY e]h isn't she pretty, (•)

10 Emm: Oh: she's a beautiful girl.= 11 Mar: =Yeh I think she's a pretty gir [1. 12 Emm: [En that Reinam'n:: Margie's response in this context embodies two elements: (1) a minimal, pro forma, type-conforming agreement "iYeh", and (2) a negative interrogativeformed assessment, built as a n e w first assessment that invites agreement but which is downgraded significandy ("pretty" versus Emma's "a doll"). In the face of this divergent assessment, Emma's response at line 10 reasserts a version of her earlier assessment ("beautiful girl"). It does so within an "oh"prefacing frame that, in this context, reasserts an overriding epistemic claim. Here it is plausible to see Emma's insistence on this compliment as an effort to overcome Margie's purely "social" resistance to it. Margie's final response again expresses a pro forma agreement; however, her repetition of her earlier evalu ation ("pretty") underscores her position as overfly unmovable (as adumbrated in her turn at line 8), while the introduction of "I think" into the utterance ex plicitly acknowledges its status as a disputed assessment. Here, as in Example (25), one participant's efforts to build a strong positive evaluation of a part of another person's world creates a context in which agreement is contaminated by a com petitive epistemic struggle.

Discussion a n d C o n c l u s i o n In his essay "Territories of the Self," Goffman describes a n array of territorial "preserves" a n d ways in which their boundaries "are ordinarily patrolled and defended b y the claimant" (Goffman 1971:52). Goffman did n o t include a dis cussion of knowledge and expertise in his description of an "information pre serve" (1971:63-64), though control over rights to information is evidendy the object of hnguistic a n d interactional m a n a g e m e n t (Kamio 1997; M a y n a r d a n d Zimmerman 1984) and systematic social competition (Abbott 1988). I n this paper

68

. .. i

(.

DISCOURSE STUDIES

we have suggested that rights to evaluate states of affairs are indeed "ordinarily patrolled a n d defended" b y individuals in routine conversational practices through which these rights are r a n k e d b y speakers relative to o n e another. O u r evidence comprises two sorts of observations. First, w e h a v e argued that assessing a referent state of affairs in first position implies a claim of primary epistemic a n d / o r m o r a l rights to assess that state. W e will abbreviate these as " K + rights," relative to those of a second speaker, w h o has lesser ("K-") rights. T h e evidence for this claim is fourfold: To begin with, first position assessments are rarely u p g r a d e d i n the several h u n d r e d ordinary conversations w e h a v e e x a m i n e d ; however they are quite c o m m o n l y downgraded. Similarly, second position assessments are rarely down graded b u t they are quite c o m m o n l y upgraded. These distributions suggest a recurrent social need to compensate for the p r i m a r y (K+) claims of first position and the secondary (K-) claims of second position. This general distributional observation is supported b y three others: D o w n g r a d e d first position assessments a r e generally p r o d u c e d b y per sons w h o , at least at first appearance, h a v e lesser socioepistemic (K-) rights to evaluate them. I n addition, u p g r a d e d second position assessments are generally p r o d u c e d b y persons w h o , at least at first appearance, have greater socioepistemic (K+) rights to evaluate them. Finally, under conditions where b o t h speakers h a v e putatively equal access to a referent state of affairs, first speakers m a y downgrade initial assessments using a tag question format, while second speakers respond with declaratives. These two practices cooperate to cancel the epistemic implications of the first a n d second positioned status of their contributions (see n o t e 7). This conversational patrol a n d defense of information preserves is m a n d a t e d b y the fundamental association between the positioning of a n assessment a n d the epistemic claims implied b y that positioning. Because social interaction is organized sequentially a n d because s o m e o n e must necessarily b e first to assess a referent, the m a n a g e m e n t of information preserves is inexorably relevant in social interaction. Second, w e h a v e identified a variety of practices that are d e p l o y e d i n m a n aging these epistemic claims. D o w n g r a d e d claims in first position a r e imple m e n t e d p r o m i n e n d y through tag questions a n d evidentials. U p g r a d e d claims in second position are i m p l e m e n t e d through a n array of practices that either invoke a "settled" and preexisting point of view, or deploy interrogative syntax to compete for first position b y usurping a previous evaluation with a n e w "first positioned" evaluation to which the other should respond. T h e practices w e have identified h e r e are somewhat varied in their privileges of occurrence. T h e [confirmation + agreement] response format is enabled b y a n d virtually spe cialized to address, and complement, first assessments that deploy the [statement + tag] format. Although "oh"-prefacing is n o t tied to a specific first position format, and thus is usable in less complementary a n d m o r e competitive ways, it is deployed almost exclusively in second position. Negative interrogatives are most aggressive and can b e d e D l o v e d in either first n r s p m n H n n t W n n «• L:

T's- tsuh beautiful day out isn't it? Yeh it's jus' gorgeous . . .

fVIYMC:l:2] ((J and R are in a rowboat on a lake)) J: R:

->

It's really a clear lake, isn't it? It's wonderful.

[Ravioli Dinner: 6 (Mark and Kim are eating dinner involving many samples of free food] Mar: Not bad for free huh? (0.3) Kim: Hm mm.

8.

9.

10.

11.

In these three cases, it is evident that neither party can claim epistemic primacy concerning the referent states of affairs. Given the concurrent, direct, and there fore equal access enjoyed by both participants, the first speaker's use of a tag question can be understood to downgrade the putatively primary rights to assess the refer ent that might attach to having gone first. This downgrading is accomplished by inviting agreement in the surface design of the turn through the [declarative + tag question] format. The data used in this paper comprise several hundred items of ordinary conversation drawn from Britain and the United States. Most, but not all, of our cases involving symmetry between [assessment + tag question] and [partial repeat + agreement] forms are drawn from British data. What Schegloff (personal communication) calls "the British tag" as a means to downgrade epistemic claims may be much more prominent among speakers of British English than among their U.S. counterparts. For another case of British/U.S. divergence in basic interactional usage, see Jefferson (2002). The fact that tag questions function differendy in first and second positions in a sequence is a clear example of what Schegloff (1996b) calls "positionally sensitive grammar." As noted elsewhere, this analysis is based on an examination of several hundred re corded conversations. Robusdy founded quantitative analysis is hampered, however, by what Schegloff (1993) calls the "denominator problem." Briefly put, without clear evidence of the parties' relative rights to knowledge independent of the talk, we cannot evaluate the extent to which parties assert these rights in the talk. Thus we are ob liged to focus on those cases in which the assertion of these rights emerges as a mat ter that the parties are addressing by talking. These methodological issues, of course, bracket the question of whether, or how, these relative rights exist independent of their assertion in the situation itself. These do not exhaust speakers' methods for downgrading the rights claimed by first position assessments. In addition to evidentials and tag questions (which focus on the authoritativeness of the assessment, or on the access claimed by it), speakers also can modulate the rights claimed by a first position assessment by downgrading its "firstness." For example, speakers can preface their turns with "so" or other ex pressions which indicate that the turn is being offered as an upshot or other product of prior talk, and by that action can acknowledge their recipients' primary rights. For example, in the following excerpt, Shirley offers a "so"-prefaced upshot (in line 43) to display her understanding of Gerri's telling about Dana. By virtue of the — — . . . . . . . . j„ ,.,i,j„v, A i o oamianr-o io initiator) Ip «1~~. « — n _ J — \

37 The Uses and Representations of Local Languages in Tourist Destinations: A View from British TV Holiday Programmes AdamJaworski, Crispin Thurlow, Sarah Lawson and Virpi Ylanne-McEwen

1

Introduction r

I 1 ravel p r o g r a m m e s are v e r y popular in the U K . According to their own I publicity, for e x a m p l e , b o t h t h e B B C ' s Holiday < w w w . b b c . c o . u k / JL holiday> a n d I T V ' s Wish You Were Here? <www.wishyouwerehere.com> h a v e b e e n r u n n i n g for over 3 0 years a n d boast weekly ratings of u p to eight million viewers. While for s o m e package tourists, these p r o g r a m m e s have been found to form an important part of their pre-holiday preparations, providing information a n d tips about holiday destinations, they also aim to provide enter tainment on p r i m e time T V (Hanefors & Mossberg, 2000). T h e basic format of these p r o g r a m m e s follows a n introduction b y a lead presenter a n d then three or four 'episodes' w h e r e other presenters or guest celebrities try out different holidays in a variety of locations. Although most locals/hosts featured in the holiday p r o g r a m m e s use English to communicate with the travelhng journalists regardless of the official status of English in their territories, it is n o t uncommon for the presenters to initiate interaction with the hosts in their native language, or to quote 'foreign' language phrases in their commentaries/narratives. Oc casionally, the presenters also elicit words or phrases in the host language from the hosts. T h e data w e h a v e used in this study c o m e from 18 p r o g r a m m e s of the BBC's Holiday a n d 10 p r o g r a m m e s of I T V ' s Wish You Were Here? broadcast between N o v e m b e r 2 0 0 0 a n d April 2001. T h e p r o g r a m m e s comprise 106 episodes 1

JAWORSKI ET AL. USES AND REPRESENTATIONS OF LOCAL LANGUAGES

ICS

featuring 33 destinations (countries or regions), broadly reflecting the most popu lar destinations of U K holidaymakers (National Statistics, 2002). Perhaps no' surprisingly, the most popular destination o n the p r o g r a m m e s is the U K , w h e n the question of a language barrier in communicating with hosts is also far les: likely to arise. Notably, in episodes featuring the U K , most interviews are wit! other tourists, rather than hosts, although British tourists, tour operators anc expatriates are also occasionally interviewed in 'foreign' episodes. We h a v e n o n e theless chosen not to include any of these interactions in our data analysis. Man] other p o p u l a r destinations, such as Ireland, the U S A , parts of the Caribbean o of Canada, are, of course, largely 'English-speaking' and whatever interactions am undertaken with hosts in these locations, they are invariably conducted in English In other locations, however, English is spoken b y the majority of the local popu lation as a second or foreign language. It is predominantly these settings tha our examples c o m e from. 2

Across all the 106 episodes of the p r o g r a m m e s analysed here, we identifiei 246 instances of interaction between tourists/programme presenters a n d hosts For our purposes, we define 'interaction' as any form of mutual engagement be tween a tourist a n d a host. This m a y b e verbal (often non-reciprocal, with o n l one participant assuming a speaking role), or non-verbal, for example, whei the tourist a n d host m a k e eye-contact, or w h e n a tourist/presenter gets a messag from a host. We exclude from the category 'interaction' the ubiquitous, one-side act of tourist gazing - in a m o r e literal sense t h a n the way it has c o m e to b e u s e in tourism studies, following U r r y (2002). O u r sample includes 6 3 instances c the use a n d / o r representation of languages other than English, which constitut approximately 2 6 % of all interactions with locals in the total sample. T h e data that we deal with in this p a p e r operate at a metalinguistic level i the sense that local languages in the p r o g r a m m e s are represented a n d performe for the benefit of the viewers (see below). This implied reflexivity of languag use invites a n analysis following the traditions of Critical Language Awarenes (CLA) studies (e.g. Clark & Ivanic, 1999; Fairclough, 1992a) as a developmei of Language Awareness (LA) (e.g. Hawkins, 1987) a n d folk linguistics (Prestoi 1996, 2004), a n d building u p o n those of Critical Linguistics (e.g. Fowler . Kress, 1979) a n d Critical Discourse Analysis (CDA) (e.g. Fairclough & Wodal 1997). With a central concern for metalinguistic a n d metapragmatic n o r m s an competencies, C L A has dealt mostiy with educational a n d workplace domain in which desirable or required versions of communication are spelled out ft pupils, trainees a n d employees. However, the framework offered b y C L A als encourages an examination of h o w everyday communication n o r m s a n d coi ventions of 'appropriacy' (Fairclough, 1992b) act as manifestations of underlyir power differentials, competing ideologies a n d 'gatekeeping' strategies. This a] proach is consistent with our general view of metalanguage as a b r o a d soci linguistic category a n d a site of m u c h ideological 'work' that social actors c through their representations of a n d manifested attitudes to language (s< Jaworski et al., 2004). Following C a m e r o n (1990/1999: 62), o n e of the central tasks of critical a

37 The Uses and Representations of Local Languages in Tourist Destinations: A View from British TV Holiday Programmes AdamJaworski, Crispin Thurlow, Sarah, Lawson and Virpi Ylanne-McEwen '

Introduction

T

ravel p r o g r a m m e s are very popular in the U K . According to their own publicity, for e x a m p l e , b o t h t h e B B C ' s Holiday < w w w . b b c . c o . u k / h o l i d a y > a n d I T V ' s Wish You Were Here? <www.wishyouwerehere.com> h a v e b e e n r u n n i n g for over 30 years a n d boast weekly ratings of u p to eight million viewers. While for some package tourists, these p r o g r a m m e s h a v e been found to form a n important part of their pre-holiday preparations, providing information a n d tips about holiday destinations, they also aim to provide enter tainment o n p r i m e time T V (Hanefors & Mossberg, 2000). T h e basic format of these p r o g r a m m e s follows an introduction b y a lead presenter a n d t h e n three or four 'episodes' w h e r e other presenters or guest celebrities try out different holidays in a variety of locations. Although most locals/hosts featured in the holiday p r o g r a m m e s use English to communicate with the travelling journalists regardless of the official status of English in their territories, it is n o t uncommon for the presenters to initiate interaction with the hosts in their native language, or to quote 'foreign' language phrases in their commentaries/narratives. Oc casionally, the presenters also elicit words or phrases in the host language from the hosts. T h e data w e h a v e used in this study c o m e from 18 p r o g r a m m e s of the BBC's Holiday a n d 10 p r o g r a m m e s of I T V ' s Wish You Were Here? broadcast between N o v e m b e r 2 0 0 0 a n d April 2001. T h e p r o g r a m m e s comprise 106 episodes 1

Source: Language Awareness vol. 12, no. 1, 2001, pp. 1-25.

JJWUKSKl tl A L .

USES A N D REPRESENTATIONS O F L O C A L LANGUAGES

li

featuring 3 3 destinations (countries or regions), broadly reflecting the most popi lar destinations of U K holidaymakers (National Statistics, 2002). Perhaps m surprisingly, the most popular destination o n the p r o g r a m m e s is the U K , whei the question of a language barrier in communicating with hosts is also far le likely to arise. Notably, in episodes featuring the U K , most interviews are wil other tourists, rather than hosts, although British tourists, tour operators an expatriates are also occasionally interviewed in 'foreign' episodes. We h a v e n o n theless chosen not to include any of these interactions in our data analysis. Man other popular destinations, such as Ireland, the U S A , parts of the C a r i b b e a n c of Canada, are, of course, largely 'Engnsh-speaking' and whatever interactions ai undertaken with hosts in these locations, they are invariably conducted in Englisl In other locations, however, English is spoken b y the majority of the local p o p i lation as a second or foreign language. It is p r e d o m i n a n d y these settings th; our examples c o m e from. Across all the 106 episodes of the p r o g r a m m e s analysed h e r e , we identifie 246 instances of interaction between tourists/programme presenters a n d host: For our purposes, w e define 'interaction' as any form of mutual engagement b« tween a tourist a n d a host. This m a y b e verbal (often non-reciprocal, with onl one participant assuming a speaking role), or non-verbal, for example, whe; the tourist a n d host m a k e eye-contact, or w h e n a tourist/presenter gets a messag from a host. We exclude from the category 'interaction' the ubiquitous, one-sidei act of tourist gazing — in a m o r e literal sense than the way it has come to b e usei in tourism studies, following U r r y (2002). O u r sample includes 6 3 instances o the use a n d / o r representation of languages other than English, which constitub approximately 2 6 % of all interactions with locals in the total sample. T h e data that we deal with in this p a p e r operate at a metalinguistic level h the sense that local languages in the p r o g r a m m e s are represented a n d performec for the benefit of the viewers (see below). This implied reflexivity of languagi use invites a n analysis following the traditions of Critical Language Awarenes; (CLA) studies (e.g. Clark & Ivanic, 1999; Fairclough, 1992a) as a d e v e l o p m e n of Language Awareness (LA) (e.g. Hawkins, 1987) a n d folk linguistics (Preston 1996, 2004), a n d building u p o n those of Critical Linguistics (e.g. Fowler & Kress, 1979) and Critical Discourse Analysis (CDA) (e.g. Fairclough & Wodak 1997). With a central concern for metalinguistic a n d metapragmatic n o r m s and competencies, C L A has dealt m o s d y with educational a n d workplace d o m a i n s in which desirable or required versions of communication are spelled out foi pupils, trainees a n d employees. However, the framework offered b y C L A also encourages an examination of h o w everyday communication n o r m s a n d con ventions of'appropriacy' (Fairclough, 1992b) act as manifestations of underlying power differentials, competing ideologies a n d 'gatekeeping' strategies. This ap proach is consistent with our general view of metalanguage as a b r o a d socio linguistic category a n d a site of m u c h ideological 'work' that social actors do through their representations of a n d manifested attitudes to language (see Jaworski et al, 2004). Following C a m e r o n (1990/1999: 62), one of the central tasks of critical an alyses of language practice is to reveal h o w institutions like the broadcast m e d i a 2

I

170

.

.

.

DISCOURSE STUDIES

p r o d u c e and r e p r o d u c e Knguistic n o r m s , and, in turn, to consider ' h o w these n o r m s a r e a p p r e h e n d e d , a c c e p t e d , resisted a n d s u b v e r t e d b y individual actors a n d w h a t their relation is to the construction of identity'. O n this basis, we view our data as ' c o m m o n sense understandings' or representations b y the T V producers of h o w communication b e t w e e n hosts a n d presenters shapes up in face-to-face encounters. Specifically, it is our aim in this p a p e r to examine the uses of host languages (other than English) in the T V holiday p r o g r a m m e s , and to establish h o w these language choices position hosts, presenters a n d T V audi ences vis-a-vis one another. We address the latter aim in our discussion b y defming the uses of languages other than English in our data as instances of crossing, which 'is concerned with switching into languages that are n o t generally thought to belong to y o u ' (Rampton, 1995: 280). T h e s e heavily mediated encounters are thus ideologically situated practices which help to better understand the subject-positionings and relations of p o w e r which u n d e r p i n tourism as both an international industry a n d a n intercultural playing field. It is precisely this kind of strictiy linguistic or socioKnguistic approach which, we believe, can m a k e a useful contribution to the already interdisciplinary tourism literature. Indeed, part of the impetus for this m o v e comes from within Tourism Studies, w h e r e writers like Franklin and C r a n g (2001) advocate the n e e d for a m o r e 'multi-sense' understanding of tourism, a n d a concomitant re appraisal of its u n d u e emphasis o n the visual - typified b y the sway of Urry's (2002) otherwise trenchant notion of the 'tourist gaze'. To some extent, any lack of scholarly attention to language is explained b y nature of the tourist experience itself; in Cronin's (2000: 82) words, 'sightseeing is the world with the sound turned off. As D u n n (2002: 3) further explains, tourists usually e n d u p gazing simply because they cannot u n d e r s t a n d the languages spoken b y the objects of their gaze. Either way, analyses of language and social interaction in tourism h a v e b e e n noticeable b y their absence. With this said, our concern here is less about m a k i n g a contribution to the tourism literature, as it is about demonstrating the application a n d relevance of C L A to a very significant d o m a i n of everyday, interethnic interaction - in fact, one of the most substantial cultural industries today. This is d o n e also in the sense of Language Awareness"'s long-standing atten tion to the evaluation a n d prioritisation of different languages a n d their speaker:;, especially within the context of 'foreign language' learning (see, for example. Hawkins, 1999). It is all of these concerns which stimulate our particular interest here in the ideological a n d discursive functions of languages other than English as they are represented in holiday travel p r o g r a m m e s . Before we discuss their functions, however, we want also to comment on thts languages as a form of performance, a n d consider their particular functioning in terms of what Goffman calls the 'theatrical frame' (1974: C h a p . 5), i.e. strips of behaviour which are produced a n d interpreted as, or b y analogy to, a rather narrowly conceived stage performance. I n this sense, performance involves one or m o r e performers engaged in a make-believe interaction (or monologue in the case ofjust one performer), and an audience, where performers and an audience are likely to b e separated spatially (the former, traditionally, performing or. ± j K„ n n r l i p n r p from a n auditorium). Of particular importance 3

m

JAWORSKI Er AL. USES AND REPRESENTATIONS OF LOCAL LANGUAGES

171

our data here, however, Goffman recognises h o w a n audience operates in a dual role. O n the one h a n d , every m e m b e r of t h e audience engages in t h e act of buying a ticket, taking a seat in the auditorium, spending real time in the theatre, taking part in the non-performance activities such as having a drink during intermission, a n d so on, i.e. every m e m b e r of the audience is a 'theatregoer'. O n the other h a n d , every m e m b e r of the audience also collaborates as a n onlooker in the 'unreality onstage' (Goffman, 1974:130) - or what, in dramaturgical terms, is known as the suspension of disbelief. I n other words, the audience is an essential constituent or co-creator of the performance since it is only through the audience that the performance is itself ratified. Goffman (1974) also notes h o w participants' activities m a y fall out of the on going frame. This m a y take place in m o m e n t s of disattention to specific actions or participants a n d usually results in changing patterns of alignment, or footing, among participants (Goffman, 1981). I n 'real-life' conversation (i.e. that which is perhaps less overtly, self-consciously staged), this m a y b e achieved linguist ically, for example, b y adopting a vocative form of address to turn one's attention to a new participant, although this is usually achieved through non-verbal m e a n s •mch as facial expressions, b o d y orientation, gaze (cf. Levinson, 1988). I n perform ance (at least in its m o s t canonical form discussed with all t h e caveats above), or the theatrical frame, the audience, despite its status as accepted onlookers, is such a disattended participant b y the actor-in-character, although it is also true that the specific status of the audience as ratified eavesdroppers is 'maximally facili tated' and, to continue, 'theatrical audiences h a v e only restricted rights to reply to the show they watch a n d are allowed only a restricted role, but unlike the inlookers at excavation sites, they d o h a v e some expectations in this regard' Goffman, 1974: 226). Examples of n e w positions that m a y b e taken u p b y audiences are those brought about b y the special role of chorus, 'orator,' or other mediating, editorialising func tions that can easily b e built into the performance. T h e mediator - a spe cialised viewer w h o also participates as a staged character - can c o m m e n t on whole aspects of the production, treating as a n object of direct attention what the projected characters h a v e to treat as something in which they are immersed. H e is a footnote that talks. (Goffman, 1974: 226-227) Other e x a m p l e s of such m e d i a t i o n s b e t w e e n t h e p e r f o r m a n c e a n d t h e idience m e n t i o n e d b y Goffman are prologues and epilogues, subtitles in silent :ims. and other slightly out-of-frame stage voices such as soliloquy i n which the itior allows the audience to 'hear' his or h e r thoughts, a n d different forms of direct address, which is u s e d to c o m m e n t o n or explain a twist in the plot, elab orate on a moral issue or summarise what h a s h a p p e n e d so far. I n unstaged inter action, an out-of-frame activity m a y involve asides a n d collusion, as 'individuals can turn from their c o m p a n i o n s a n d give fleeting v e n t to their "real" feelings, irough gesture and sotto zw*commenf (Goffman, 1974:233). Goffman's example fcof such 'electronic collusion' in the case of T V s h o w s is f

DISCOURSE STUDIED

172

voiceover in a quiz show 'telling the folks in T V land what the answer was while the contestant manfully strained at pretending to strain at recalling it" (1974: 233). Following Goffman's work on the theatrical frame and out-of-frame activities, we treat the T V holiday p r o g r a m m e s as instances of scripted performances, with the T V audiences being the accepted onlookers, whose right to participate in and respond to the staged activities is necessarily minimal, b u t whose role in ratifying die p r o g r a m m e s is hugely important. I n simple economic terms known as the 'ratings g a m e ' , without the viewers' mass participation, the programmes would b e scrapped b y the T V stations p r o d u c i n g them. A n d as we hope to demonstrate in this paper, the use a n d representation of host languages serves as an important aspect of the p r o g r a m m e m a k e r s ' realignment with the viewers.

Functions of L a n g u a g e s O t h e r T h a n English i n T V Holiday Programmes A n initial analysis of the primary functions of languages other than English in T V holiday p r o g r a m m e s identified four m a i n categories of function: • expert talk (e.g. guided tours, explanations, instructions); • service encounters (e.g. purchasing foodstuffs in shops or markets); • phatic c o m m u n i o n (e.g. exchanging greetings, thanking - often as part of a service encounter); • n a m i n g a n d translating (e.g. providing labels for local concepts, artefacts or dishes, translating local place names). These functions are performed b y the presenters a n d / o r hosts either when the two parties are shown interacting with one another, or b y either party when no face-to-face interaction is present, for example, in n u m e r o u s voiceovers by the presenters or w h e n racing c o m m e n t a r y is h e a r d during a visit to a race course. Of course, m o s t communicative utterances perform multiple functions, and we found such overlaps in our data. As such, it was not always possible to isolate com municative intent. I n looking to code our data, therefore, interactions were cai egorised in terms of their primary functional orientation, while allowing coding in terms of m o r e than o n e function to retain something of the multifunctionalit*. of exchanges w h e r e relevant. Expert Talk We define talk as 'expert' w h e n a host is portrayed in the role of a n authority or guide a n d shown explaining a process, system or local custom to the presenter, w h o is portrayed in the role of learner, listening a n d perhaps asking questions E x p e r t talk appears to b e the most c o m m o n function of language used by hosb in the p r o g r a m m e s . T h e majority of these instances take place in English (in oir sample, almost as m u c h as 9 0 % of all instances of expert talk). After all, this

[SBL:3:3:R:5] Milly : Oikay that's all ah wan'duh know r I thought it w I z mui c h i Keith : Y a h w' 1 we we don' know how much is Tur.icorne h*gre though, J

1

Somehow 'Okay that's all I wanted to know' has a strong sense of finality about it. But, no, one can perfectiy well go on with more. The last fragment in this little set is another one of those prototypical adjacency-pair first pair-parts, 'How are you?' In this case the one asking the question goes on to produce a candidate answer. 1.5

-> -»

[Her:0II:2:7:5:R] Doreen : Yes well pop in on th'way back'n pick it up Katie : "Thhank you ve 'y much" eh ha-how are you 4-all. r Yer a L itde ti:red °nah° Doreen : Q_hwir all fi:ne, Yes I'm jus: sohrta clearing up

We can note the reasonableness of such further talk for such an item as 'How are you?' That is, it is in the first place merely a pro-forma question which recurrendy gets just the sort of pro-forma answer it gets here, 'Fine.' The additional material here provides that there was a bit more to the question than that. I'll just quickly go through two more positions in the transition space, again showing a case of clean transition, and then a few in which there is overlap, both minimal and more florid. These next two positions entail a slight incursion into the utterance in progress. Terminal Onset 1.6

[Fr:USI:99:R:2] ((Vic has just accused Mike of having made a statement that 'came from an attitude.' This idiom rather defies translation; suffice it to say it's not a good thing.)) Mike TDin' come Fm'n 4attittoo;d? (.) "tDon'tell me it cameT f m 'n Tatti-kood. (•) Vic °14-came fum-l en a:ttitr 0 )° Mike [ Ahha:h. Carol °Ehka:v.° 0

1

DISCOURSE STUDIES

196

I show this particular case because it has two people starting up 'there'. And somehow I find it useful for, in the first place proposing that it is a 'there', to have such consensus. 1.7 -» -> 1.8

-> -> 1.9

-» -*

[Her:I:ll:10] Norma : Oh: yes cz Min:x now is (.) three:: (0.3)-three pohinds eleve r n. ou:nc r e.s, 11 Doreen : Yg:s, Ye:s, 1

[NB:IV:1:R:1] Emma : W't time yih haftih g_ e:e]_ven? (1.2) Lottie : iNoj.: abah a quarduh twe:l, ve twgnny, minnis t'twelve. Emma : Oh:: 1

J

[Rah:B:2:(14):10:R] ((re. why Vera's visiting grandchildren didn't want any lunch.)) Vera : •_ they muucked intuh biscuits. They had (.) quite a lotta biscuit r s n che , e _ i :: i s e, Jenny : Oh :: J well thahfs it th ^ en T

1.10 IJG:I:8:3] Marge : W_ll so anyway I c_rtainly hfipe thatchu w_uld go with -> hi r m:beec- hh-hh -> Colin : Ygah well tuh tell tell'm r ah'mMarge : BECAUSE HE ha:s already arranged uhr a h-room up there so I know thetchu'd have a place tih st_:y. 1

'Last-Item' Onset 1.11 [Her:01:2:2] -> Jean : So well they won't be here Boxing r Day? -> Doreen : 0_t i well that doesn'mattuh 1.12 tSCC:DCD:23] Sokol : Ah'U __1 you'ow she does Ut? -hhh That's oil sewn -» tihgethuh by r hay:nd. -> Bryant : I thflught this wz a very expensive (business) I, I S£S r this. Sokol : Yeh.° lo

1.13 [GTS:I:1:68:R:4] -> Roger : u-But heck I do the same damn r t h i_n g , buh-> Louise : TOhev'ry Jb'ddydoes s'sa°l:me th r ing." Roger : Bud vers wz pretty obvhhi(h)ou:s heh

JEFFERSON

NOTES O N 'LATENCY' IN OVERLAP ONSET

197

Note that the overlap in this case is resolved b y the current speaker d r o p ping out; i.e., it is n o t in the first place 'minimal' in the sense that some of the others are. I've b e e n trying to show very simple cases. I'll just show o n e - another case of overlapped 'last-item' onset - in which the p h e n o m e n o n is not so obvious. As it h a p p e n s , a recipient is 'interrupting' a n 'if-then' formatted statement which eventually, after several rounds, turns out to b e ' . . . if it quiets o n e thing m a y b e it q u i e t s . . . your w h o l e . . . g a s t r i c . . . kind of business.' W h a t m a y b e h a p p e n i n g here is that the recipient is utilizing a simpler version, 'If it quiets one thing it quiets everything,' for which, then, she is starting u p at the projected 'last-item'. 1.14 [SBL:2:1:8:R:5] Bea: VERY often u-*u:: anything thet's given to yah to:: qhuiet things do:wn quiets mp.:re then you wah:nt. • ((ca. 8 lines omitted)) Nora Bea Nora Nora Bea Nora Bea

Well I spoze it dp.es if it quiets one thing maybe it quiet r s yer , uSURE J ly. r E V RJE t h u . EVRY thi:ng. yer w h o : 1 e (0.4) who.::le gas-s-tric (0.6) E r v- eh- , kind'v business. EvTry.thi:ng. Uh haji L

'

So there is this p h e n o m e n o n : Transition-space onset, speaker continuation, resulting in overlap, it is massively present; obviously so in m a n y cases, m o r e obscurely in others. This is n o t to say that all cases of overlap are accountable in that way. People do, for various reasons, start u p before a current utterance is anywhere n e a r completed or transition-ready. I'll just show a few cases of these. I tried to find a neutral t e r m for t h e m a n d came u p with 'interjacent' onset.

2. Interjacent O n s e t 2.1

-> -»

2.2

[GTS:I:l:44:R:2-3] Louise : That's FI:VE TEiH? en tha:t's six foojt two;? (0.2) Louise : 'N how t •• all u h r . you iA: •, 1, Roger : How tall'r-' 4-you Al. ' (0.7) Louise : ih ,Thhuh, thu, Al : l I'm fi * :ve ^seven, [MDE:60-l:3:l-2] Sheila : What time did'e set on the nlarn»

DISCOURSE STUDIES

198

Tom

-> —>

2.3

: Uh::: (0.2) I: don't know exactly I think ih wz arou:nd thrge uh'cloxk er so.mething a'that sort. (0.2)

Sheila Tom

—>

[SBL:2: Zoe Amy Zoe Amy

—>

Zoe

; r

o t sm i He took it-* et fou:r. Gerda gays.

an' he. sorta iscares me,h Have you sgen im? •hhh Wj-11 I(m) I've Umet eem,° •hhhhhW'l'uh actually: when s h e's •, ' En the way th ^ e;y_ lpla:y. Oh.-: (•) Serious huh? •h TY-ah, r

Amy Zoe 2.4

—> —>

[NB:II: Nancy Emma Nancy Emma Nancy

L

TWQNderiPul. TSo: we w'r s*itting in , L Y E R LIFE-* is CHANG ing TTEEYE::AiH r

r

1

A m o n g the cases I h a d stuck together as 'interjacent onset', I h a d a subset that I was calling 'post continuation onset'. I n these, a recipient would start up just after the current speaker h a d p r o d u c e d a clear indication of going on, fol lowing a possible completion. H e r e are a few of these. 2.5 -> ->

2.6 -> -> 2.7 -> -» 2.8 -> ->

[SBL:3:1:R:6] Amy : all thetu stuff Maryoui? requires a lo:Ttta spai*:ce. Ah me a:n , i h . -hhh Marylou : ^ Sp£' cially if yer gonna have it open fun the public, r

[Fr:GS:4] Mary : hn (h)he shiz wejl, (0.5) u-ghesshu get ha:lf en j. I sid I GET M_Q , RE THEN H(h)ALF Sue : Wuh yg.ar wuzit? [SCC:DCD:26] Bryant : that branding that wouldn't of bean touched with'n i:yn. The rn . o:st we'd of done , Sokol : 'Oh: n o a: w ' right, steam [NB:IV:4:R:4] Emma : cz you se_e she. wz: depending on: hhim takin'er in tuh the L.A. dgeple s:- depot Sundee so •• 'e. siz , Lottie : "-Ah^ll feke'er in:

1

JEFFERSON NOTES ON'LATENCY IN OVERLAP ONSET

199

And that's a reasonable e n o u g h place to 'interrupt'. For example, if w h a t has been - adequately and syntactically possibly completely - said so far is something to which a recipient wants to respond, and n o w it looks like the speaker is at least continuing and p e r h a p s m o v i n g on to other matters, t h e n one might want to get in now, while the initial matter is still relevant, even if it m e a n s interrupting. Or, for example, as in the following fragment, the utterance in question might itself be treated as interruptive of something prior. This is taken from an interview of a group of y o u n g girls. O n e of them, Cindy, produces an utterance which could be just a remark, b u t could also be a story preface. Another, Kathy, then produces a similar sort of utterance. After a first possible sentence com pletion, Kathy starts a n e x t clause, and simultaneously b o t h the interviewer and the initial speaker, Cindy, m a k e a m o v e ; the interviewer asking a question which will provide for a story b e i n g told, the initial speaker simply continuing her story. 2.9

[Labov:Quadro:I:1720:5:Simplified] Cindy : Th:gy_ got mad on Friday though I dint get home t'leven thirty, Debbi : -hhhh,, Oh:::::: Kathy : •'Yehinymo 1 ther wzkindamad I din get Louella : t Where were you. Cindy : En I had the ca:r, I don'know w£ were just- we went t'the basketball game . . . f

—>

r

—> -»

So it looks like this is certainly some kind of 'there', a place for another or others to start up. Now, there was one really puzzling bunch of these things: some one would start at this 'post-continuation' place, a n d then immediately d r o p out. For example: 2.10 [Fr:USI:43:R:2:Simplified] Mike : Least'e c'd'v] dane wz c'm dah::n en letchu know w't -» happ'n hey r look y i , h gkuss broke, -» James : >• Thaj.t2.11 [Henlll: 1:4:2-3] Desk : Just ane moment I'll uh he's in uh-actually in suhngery -» et the moment ah'll , sge'f I c'n get hold of him, -> Heath : '•eeeeYg:s, r

2.12 [SCC:DCD:9] Phipps : Didju hev it ma:de (0.3) ju- soon ahftih you bowt th' -» m'terial? ohr ohr ilay::tuh. Was i t in •, Febru'ry: -» Sokol : ' l W'H r

J

2.13 [NB:II:1:R:12-13] Emma : °God ih° jist like a night max re, Lottie : '•Ihd e - s e e m s r

a

200

DISCOURSE STUDIES

-> —>

fatay storeh I i couldn'believe it. Emma: Ir

L -

2.14 [GTS:I:l:50:R:l-2] Dan : it is part a'the function a'°th'J-group to,begin d-«ha;re —» in° Isome a'these things so : the others c'n understand . . . ° -> Roger : W'l

0

r

L

The question was, what o n earth is this? They start up after it's obvious that someone is continuing, and then do this 'oops sorry' and drop out. I kept col lecting the things, with this sort of question in mind, and after awhile it occurred to m e that I'd got it wrong. Maybe they're not starting up 'post continuation'. Maybe the way to characterize where they're starting up - and this would handle, account for, these immediate dropouts - is that they're starting up by reference to the possible completion which precedes the continuation. Further, there is a massively recurrent, systematic position for next speaker startings, which these things might be treated as instances of. It's another transition-space position. Where, then, just as the other 'transition space' onsets can end up in overlap as the current speaker continues talking, so can this one. I'm calling this position 'unmarked next' position.

3 . ' U n m a r k e d N e x t ' Position I call it 'unmarked' simply because in m y transcripts I don't use a symbol to show this particular relationship. A recipient/next speaker produces his talk in such a way that it occurs with neither haste nor delay. It is not pushed up into or latched immediately onto the prior utterance, but permits just a bit of space between the end of the prior and the start of the next. It is 'simply next'. And I have the impression that this is the most common, the usual, the standard relationship of one utterance to another. Here is an instance. 3.1

[NB:I:1:R:1:R] Curly : Helloi? Bud : Curly? Curly : Y :h. Bud : Bud Hoffmeier. Curly : Hi Bud how you doin. Bud : Fine. £

To get a sense of 'unmarked next position' as a locus of overlap I'll sneak up on it gradually, starting with innocuous cases. For example, there are cases where the overlapped material, while it might well indicate that a current speaker is 'going ahead', does not constitute 'talk' as such. Rather, the speaker is taking a breath, perhaps in preparation for further

JEFFERSON

NOTES ON 'LATENCY' IN OVERLAP ONSET

201

talk which does not seem to be of the sort which would motivate 'interruption'; such things as brief'receipts', 'acknowledgement tokens', etc. The resulting data is innocuous to the point of vacuousness. 3.2

->

[SBL:2:1:6:R:3] Tess : becuz I wz going dih haftih get out (.) early in the morning I have (.) a p - (0.3) transaction et the ba::nk.,-h. , hh Bea : (—' I iY :h. Tess : A.:nd uh -p but it wouldn't a'made any diffrence cuz ah wz gunnuh dress t'guh down'n do it anyiWA:Y..-h. hh Bea : V_J 4-Y :h.

3.3

[Her:I:ll:12]

->

Doreen

: en Jonathan's gone tuh Kingston fer a meeting,

~>

Norma

: L.^

3.4 -»

[MDE:60-1:5:1] Sheila : they were already off someplace so Tony called..-hh. ,hh Erma : ' _ > l-Oh:.

-> -> ->

a

f

->

L

a

fohi:.

But these little receipts and acknowledgements also occur 'in the midst' of another's talk; specifically, after it is clear that a speaker is going ahead after a possible completion. 3.5 -> -» 3.6 -> -» 3.7 -»

[Rah:B:l:(13):3] Jenny : I think she's got a- uh-m: one a'those: uh permanent caravahns. ,up> .. in: Gro ,zeida:rle. Ann : *—' ' • M m : - ' '•A:oh. [NB:IV:3:R:5] Lottie : becuz they would really be the Spri:ng. ,Let i ' r s see tha t's twunny fl:'dgHars . . . Emma : — ' Y e a h . 1

L

1

[SBL:3:5:R:4] Milly : Well that's it,an»' r it's diffrent then I've e:ver tlived Ginny : —' °Yah° l

3.8 -» -» 3.9 ->

1

[Owen:8B15(A):36] Bette : It really keeps you going actually hoi- ah mean both looking forward an'looking ba"-ck. ,on th>' r holiday . . . Andrea : * ' '•Yg.ih [Friedell:30] Hank : Well we better not spend three hundred fifty a month . we. better be able tuh hold it, (0.6) three hundred, r

202

DISCOURSE STUDIES

While in fragments 3.2-3.4 a current speaker m a y b e not only pausing for b r e a t h b u t m a k i n g a space for various orientation signals from a recipient, in fragments 3.5-3.9 n o such opportunity is provided. T h a t is, the recipient is pro ducing such a display just after a speaker has shown that n o such thing is called for. This p h e n o m e n o n could b e accounted for if the recipient/next speaker was designing his talk b y reference, not to w h a t occurred immediately prior to the start of his o w n talk, b u t b y reference to w h a t occurred a bit earlier; i.e., the pos sible completion of the other's utterance. T h e design of the recipient/next speaker's talk could b e characterized as pro viding a bit of space between the two utterances. T h e bit of silence, then, being part of the recipient/next-speaker's w o r k in the first place. It would not then b e a matter of a n utterance reaching possible completion, a silence occurring, a n d the recipient/next speaker understanding thereby that the prior turn is completed, w h e r e u p o n h e starts to talk, as is sometimes argued. Rather, as is nicely evidenced b y 'last item' onset and 'terminal' onset, upcoming completion is projected as the talk in progress is unfolding. I n the cases at hand, it m a y b e that the recipient/next speaker undertakes to p r o d u c e a particular, standard sequential configuration, with a small inter-turn space as a projected part thereof. W h a t the recipient/next speaker is doing, then, might best b e described as producing a n action which consists of: permitting a bit of space and then starting to talk. Now, such an action might n o t have the recipient/next speaker in such an orientation as would h a v e h i m listening for or attending to the occurrence of that small particle of sound inhabiting the little silence h e is now, as part of his activity, providing for. H e is, as it were, in a 'blind spot'. H e is n o longer in re cipient orientation, but in a state of speakership, although h e is n o t yet produ cing sounds. Such an account handles these 'post-continuation' acknowledgements and receipts, a n d also the set of 'post-continuation' onsets followed b y immediate dropouts (see fragment 2.10-2.14). T h e y are not being produced 'post-continuation', b u t 'post possible completion plus a bit of silence'. T h a t is, they are b e i n g pro duced to occur in ' u n m a r k e d next position'. As with other transition-space onsets, they h a p p e n to collide with a current speaker's producing further talk. T h e possibility of a systematic, designed 'latency' in a recipient/next speaker's talk provides for an orderliness to some other sorts of overlap. I'll mention two. O n e of t h e m I h a v e n ' t d o n e any work on, a n d the data are often very messy. But they might at least partially b e accounted for if this 'latency' is p r o d u c e d n o t only b y recipients, but b y speakers as well. T h a t is, in a sort of correlate to the way in which a recipient might shift into speakership orienta tion a n d n o t listen for or h e a r a particle of further talk, a speaker having reached a point of possible completion might not shift into recipient orientation b u t main tain speakership orientation, and thus n o t listen for or h e a r that a recipient has started to talk. T h e speaker, then, could b e characterized as providing a bit of space between two adjacent activities, a n d / o r two adjacent syntactic components. I haven't systematically collected these, b u t h e r e are a few cases that will do for the time h p i n o r

JEFFERSON NOTES O N 'LATENCY' IN OVERLAP ONSET

203

3.10 [SBL:2:2:3:R:38] Zoe : well with Bud? en the way he says I'm not bid uhhhw(h)'l (h)if you s(h)ay two(h)o sump'n then I'm not gunna bi:d, -> • i hh\ r hh r I -I Tdpn'ti think Elva si - Elva would = -» Amy : [ o h t h t e y . don't -hh J

Zoe

J

: = take that kindly

3.11 [SBL:2:1:8:R:12] Nora : En the] little edge ed curled up'n wz showing rg:d.h -» • /hhh. r h r En ah thought gee wh -. iz what's thi:s. -> Bea : ^ ' £>h | f e r g'dness°say° J

3.12 [NB:IV:10:R:50] Lottie we bought s'm hats et Wah uh Wakd-u-er CJa.:rk's they had uh: those uh (0.2) fisheen ha: ts y i h , know, Emma ^ Mm: J hmj., -» Emma = /doze lid% rdle-i (badges) -i Yeh -i -> Lottie ' *a- doi:llar s o l .Jbaughtone fer hei:r, r

—

O n e that I h a v e collected systematically occurs after a completion followed by a silence. O n e participant starts u p a bit after another. Again, the possibility is that each n o w being in speaker orientation is n o t listening for or hearing that the other h a s started. I n such cases the 'latency' can b e o n t h e part of prior speaker or recipient, or as in fragment 3.20 below, can occur as between two recipients of a prior utterance. I ' m showing cases across a range of lengths of silence, from a 'micropause' of just over l/10th second, to a substantial silence of one a n d four-tenths second. 3.13 [Her:I:6:ll] Helen : Alright ifine. Doreen : Okay, (•) Doreen : /Lo> r ve t'you ia^ 11 Helen : U B y e : , (•) Helen : Bye:, Doreen : tBye:.

C

[

J

3.14 [JG:II(a):6:l] Maggie : But'e said a lotta times it happens 'e said I could've even been allergic tih something thet I js:t yihknow (.) ATE (0.2) Maggie : ,em pit j's -.t effected me like tha rt. Sorrell : —> °Yeah, °(Mm)°

C

[

3.15 [Erhardt:8:l]

L

oJ

1

DISCOURSE STUDIES

204

(.) to work o:n um tomorrow night (0.4) Vicky : / Thm\ r s d a n y? r -i Pam : bec'z Yes. 1

;

l

J

U

3.16 [Owen:8B15(A):25] Bette Qh::. , Andrea o n a G ood 4-Fri Bette (0.5) /So^ r we, V e g o Andrea — Where'v you Bette = something Andrea f l

J

r day. Friday. tto take th'MQj tbrbike or = got to ge_t to this

3.17 [Labov:Bat:II:4] Parky : They've got f&'ovuh there, (0.7) /You s> r ee two (.) trainees en then you see the two= Parky Loma Parky T

3.18 [GTS:I:2:94:R:l-2] Louise : ennit's discussion igroups:. on a subject with one te_ach*er. -> (0.8) Louise : , L •• t ' s u •, m Tvery. few classes, as classes. Roger : U °Hm,°

C

L

J

3.19 [JG:I:24:8-9] Laura : But I know thet Joe did say he had a letter from im. (1.2) Êh dij ^ d he tell you- -hh Marge That's all he said. Laura Well did he tell you that when you phoned im Marge 3.20 [G:84:AD:52] Cal : You know why niggers are always goin like dis? (1.4) Bart : /Ban % i d e o d orant? -i . I I tru™. J„„>U 1what they wanna do:. hey don'know

C

So there is a great deal of data in which, in varying degrees, s o m e o n e starts up 'in the course o f another's talk, which can b e characterized in terms of'transitionspace' onset or, as in fragments 3.13-3.20 in terms of for all practical purposes 'simultaneous' starts. T h e account for the 'interjacent' starting has to do with a systematic 'blind spot', each participant being in speakership orientation and not listening for or hearing a bit of talk b y another. T h a t is, s o m e o n e can b e sys tematically described as not h a v i n g heard. Now. this is not to sav that participants cannot b e hearing a n d dealing with

JEFFERSON


205

possible completion, or with what is being said, in particular, in that bit of talk. People can and do react to small particles.

4. Consensus or Reaction? In the following two fragments, someone does a checkout on a prior utterance, and the prior speaker produces a confirmation. In each case, after j'ust a bit of the confirmation object has b e e n produced, w e get an acknowledgement of that confirmation. 4.1

-> -» 4.2

-> ->

[JG:II(a):3:2] Sorrell : she's a'ready gone cz Claire's there. (0.3) Maggie Oh: (.) Claire came in? Sorrell : /Vex r ah. en she sizi no prob'm she'd work tihda:y Maggie : *—' "O h; okayj J

QG:I:16:6] Pete : No ah'll call again, Grace : You will. Pete : /Ye» r ah. Grace : * — ' Ah-hah -hh Ah'll tell'im then. 1

And across a range of circumstances and activities w e can see recipients at tending to small particles. The cases I'm showing might be seen as a systematic subset of 'interjacent' onset; i.e., 'recognitional' onset. 4.3

-» -»

[Her:I:6:9] Doreen : If by any chance, (0.8) theh isn't anybody heuh I've gQt tih go out jus fer awhi-'le, j -hh j What is your- newHelen Ye: Doreen What is your .telee^ r phone, nummuh?, Helen * ' Wgll we're n^ ot on the phone y.et. Doreen : Ah I see:. J

1

4.4

-» -»

4.5

[Rah:II:ll:R] Jenny : But- ee-ee "two pihleece cahrs'd" stopped outsi:de. = Ida

:

= eeYe:ir s?

Jenny

:

Ida Jenny Ida

:

^ -h An'that whether he tho:wt thet I ed'n / acv r cidn't r r ohr someth ing I don't kno_:w. = ^_' I_j_: : kneo:w L

1

[

l

niNo.:::

[Fr:TC:1:1:2-3] Shirl : Who w'yih Ttadking to.

J

DISCOURSE STUDIES

206

->

Geri Shirl

1

: Jis nouv? : -hhhh £2g_ I called be- like between ,ele.

',,

:I

r v g n en

*—' '•L, wasn't talkeen tuh amybuddy. (b) Bo-oth Maria 'n I slept'ntil about noo_:n, 4.6

[MDE:60-l:5:2-3] Erma : did you hear w't'e: she- tol:d Lisa? -hhh thet they w'r c'nsidered the cutest couple in r the claiss? r ( ) Sheila : -hhhhhh Y e h - W u l this: this was official, apparendy this goes into the .yga.. . rbook. Erma : * — ' Qh: yer kidding. How adorable. 1

-* -» 4.7

-» -»

l

•

[NB:IV:13:R:12] Emma : What city is] it ii:n. (0.8) Lottie : Wul't'sin .Cos. .taMe , isa , Emma : ^ ' Costa Mgs a

Some of these 'recognitionals' can get quite delicate. In the following fragment it appears that an initial consonant is adequate for acknowledgement to occur. Here, a repair is being made from 'fifty cents' to 'fifty dollars'. 4.8

[Friedell:32] Sandy : I wz figuring- six twunny five is what I get Pm Merit So rotherr t h e n-i b e ing fiifty cents. Hank : A'right Okay, r

1

J

l

J

(•)

-» ->

Sandy Hank

: u-mean fifty , d . . ollars.,(.) ou:t, we'd be . . . : ' UiYe:h, J

And in the following fragment a possible alternation between 'shoot' and 'shit' is reacted to upon the occurrence of the crucial vowel. I'll just note that we have a great deal of talk between these two women, and their friends, neighbors, relatives, and this is the only occurrence of the word 'shit'. They routinely use 'shoot', 'doggone' and others of a mild grade, 'damn' and 'hell' being the boundarying expletives. 4.9

-> ->

[NB:II:5:R:3] Emma : 'e wantih pull a tooth'n make me a new go:ld uh: -hhhh (0.2) bridge fer (.) EI-.GHT HUNDER'DOLLARS. Lottie : °Oh:: sh:: ,i. .:t.° , Emma : - Shi *t. (0.2) is Xri:ght.

In fragments 4.1-4.10 it seems quite clear that a recipient is exhibiting that his response is directed to the particle which directly precedes that response. I've found a few cases where such monitoring and reaction might be occurrinff. at those ooints I've proposed to be 'blind spots'. But at least as far as

JEFFERSON

NOTES O N 'LATENCY" IN OVERLAP ONSET

207

I can see, they're ambiguously i n d e p e n d e n t actions or reactions. T h a t is, I can't decide or argue that s o m e o n e is hearing a n d responding to s o m e particle, or starting a n i n d e p e n d e n t bit of talk with that 'latency' I've b e e n proposing. I'll start with an innocuous case. I n the following fragment, a n object which is alternatively used to solicit response or link into a n e x t part of a n utterance in progress, 'See,' (cf. 'You know,' 'Right?', etc.), occurs after a possible completion. It is immediately followed b y a n acknowledgement token b y the recipient. 4.10 [SBL:3:5:R:6] W'l this is what I ?mea:n. ,See 4- it's ggyn be so = -> Milly I HiMmhrJ —» Ginny Milly = I w h u - Toi4-n:derf:uU I just don't h a v e the analytic resources to argue one way or the other; the acknowledgement token could b e responding to 'See,' as a response-solicit, or could b e occurring i n d e p e n d e n d y of it, in u n m a r k e d next position to the prior part of the utterance. T h e r e are others, rather m o r e interesting a n d delicate, in which that sort of ambiguity is present. T h e following fragment is taken from a telephone con versation b e t w e e n two m e n o n duty at different locations during the 1964 Anchorage Alaska earthquake. T h e y refer to each other b y their locations: 'City' is the Anchorage fire d e p a r t m e n t a n d ' E l m o n d o r P is an oudying a r m y base. They've b e e n connected b y a short circuit in t h e telephone wires, a n d h a v e taken the opportunity for a chat. T h e fragment starts after they've b e e n talking for awhile. 4.11 [FD:Finger: 2-3] E'dorf : D'you know w't-w't kinda ngws'ere broadcastin'down'th' States et (.) pres'nt? City I: heard r d'finst Squawk (xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx xx) City The firs' one thet dey uh, (0.7) broadcas'wz sjxty rtuh thrE'dorf Yer loud'n clear Muldoon Tower, (2.0) Pardjn? E'dorf City I heard d'rfirs'broadcas'Stateside, , Squawk '•xxxxxxxxxxxxxxxxxxxx-'xr xxxxx) E'dorf '• Jjista minnit. (1.4) E'dorf Gfl'head. (1.9) E'dorf Go'head, (0.2) -> E'dorf ( ) L* ' -> City — ' Y e - u - ah heard d'firs'broadcas'state det deh wz bout sixty t'three. hunner'dea:d'n (0.4) city of Anc'rage is on dih grou:n' t

C i

v

7

DISCOURSE STUDIES

208

J u s t as City starts to answer Elmondorf's question, a squawkbox o n the Elmondorf side starts u p , with a r e p o r t in from M u l d o o n air field. A n d w e can watch City's work b y reference to the Squawk. H e initially d r o p s out, a n d then, p e r h a p s because h e gets n o indication from his coparticipant that h e should maintain his silence, m a y take it that someone else o n duty is h a n d l i n g it, so he starts u p again. But it turns out that his coparticipant is handling the call, and interrupts h i m to r e s p o n d to it. Thereafter, City remains silent until he's invited back b y Elmondorf, with 'Pardon?' A n d p e r h a p s h e hears it as directed to h i m a n d n o t to the fellow on the Squawk because 'Pardon?' is a 'conversational' object, in strong contrast to the 'instrumental' 'You're loud a n d clear'. But again, just as h e gets going, the Squawk goes again, h e produces a first clause, a n d Elmondorf, again with a 'conversational' object, 'Just a m i n u t e ' , gives the Squawk priority. N o w comes the problematic stuff. Elmondorf uses ' G o ahead', which is am biguously conversational or instrumental. This m a y cause a p r o b l e m for City; w h o is being asked to ' G o ahead', himself or M u l d o o n Tower? A n d it is possible that Elmondorf comes to see w h a t the p r o b l e m is; after two such invitations, he shifts to a non-ambiguous item, 'City'. It's possible that City hears, in that small particle, ' C i . . . ' that it is h e who is being invited to talk, w h e r e u p o n h e starts. It is also possible that it's a matter of consensus; that whatever it is that brings E l m o n d o r f to see that his recipient. City, is having a p r o b l e m , brings City to see that h e is i n d e e d Elmondorf's re cipient. T h a t o n e of t h e m starts minutely later than the other is just another in stance of post-silence latency, as in fragments 3.13-3.20. Again, I don't have the resources to choose b e t w e e n these possibilities. T h e following fragment might b e quite exquisite. I ' m using the real names of two of the participants,Jeff a n d j o a n n , to show just h o w delicate it might be. This is from a g r o u p therapy session for teenagers. At this session, there are observers in a r o o m b e h i n d a one-way mirror. 4.12 [GTS:I:2:19:R:51 ((Jeff is reporting a success with his parents, who have stopped interrogating him about his comings and goings.)) Jeff : Nob'ddy se]z inning yih jis keep °whha:lkin.° hh i "yihknow (0.2) It'sjTbghuggin'lmhhe(h)now r h m hm •, Jeff TDpn'ta lk tih ithem talk Joann t'm*:s:. (1.3) No (.) th- (0.4) drapes er closed now I c'n see through Jeff that liddle crack et th'window over there (2.0) Yer very T°conscious'v th'm being in the:re.. Je. ff. Dan - > He Joann keeps:: rT£alk r in'ithere.-i i h ° It doesn' rilly J-bother me, Jeff L

J

r

l

u

1

J

L

I

JEFFERSON


209

The overlap in question could be occurring at a very touchy moment. Joann having raised the issue of the observers in the first place, it's possible that Dan's remark is addressed to her. Indeed, the appending of the boy's name may be directed to clearing such a possible ambiguity, similar to Elmondorfs work in the prior fragment, with his shift from 'Go ahead' to 'City.' And similarly to City's problem in the prior fragment, Joann could be monitoring for which of the two who have shown themselves to be 'conscious of them being in there' is being addressed. If that is so, then the 'recognitional' work in this case is a bit finer than in the prior, the alternative names in this case starting with the same consonant. What would be involved here, then, would be another instance of that very fine sort of monitoring and response seen in fragment 4.9 vis-a-vis the 'shoot'/'shit' alternatives; i.e., recognition upon the oc currence of the crucial vowel. One problem with this fragment is that it is face-to-face interaction and we don't have the video. D a n might be looking straight at Jeff, in which case we have a trivial instance of 'unmarked next position' onset happening to overlap with an appended and unnecessary particle. O n the other hand, it turns out that frequendy just such actions as D a n is producing; e.g., challenges, confrontations, etc., are done 'blind' - a term introduced by a graduate student of Harvey Sacks, Blaine Roberts. That is, they are done specifically not looking at the re cipient. And there is at least one other reasonable target for Dan's eyes - the controversial observation window. So, we have a possible case of a very delicate monitoring for and reaction to an otherwise 'blind spot' particle. Or just another case of overlap in the tran sition space. I'll show one more case of possible intermediate reaction to a small particle, or possible independent actions which exhibit consensus. It's taken from a conversation between the two ladies of fragment 4.9. In this case, in a range of ways, we can see Lottie working to have the conversation end, without herself having been the one to end it. 4.13 (NB:IH:4:R:3] Emma : •hhhh Well ah won't KEEP yih honey, Lottie : Oh Tthat's ok_:y, /uh, , (I'm just) Emma : »—' ^ Have Tfun fishinT ah hope yih catch a big one, Lottie : Y_ah. (0.2) -* Emma : .AJ. .right,, -> Lottie : »—' '•OJkay-' honey, (•) Emma : r •. Bye bye, Lottie : *• ^ Bye bye For example, 'Oh that's okay' does not disagree with the proposal that she's being 'kept' on the phone. In other such cases we see strong disagreement items, and/or devices for showing that one is happy to keep talking, e.g., introducing a new topic, referring back to a prior one, asking about something in recioient's

DISCOURSE STUDIES

210

T h e n , the 'Yeah'. It occurs at a point w h e r e she could initiate Closings with, e.g., 'Okay.' So she's clearly n o t doing Closings. But in terms of being in con versation, she's p u t the talk into a state which J o h n Lee a n d I call 'attrition/ hold'. Roughly, it hasn't b e e n closed off, a n d thus is still alive, b u t such an ob ject as 'Yeah' provides n o materials for the coparticipant to take off from or work with. If the talk is to keep going, it's u p to the coparticipant to find somediing to say. T h u s , the sense of 'attrition'. Focal for us is h e r ' Q k a y ' which is alternatively 'latent' or 'recognitional'. It might b e occurring just a n d only when E m m a has recognizably initiated Closings. T h a t is, withholding an overt entry into Closings until E m m a does the initiating, a n d t h e n p o u n c i n g o n it the m o m e n t she has, reifying it with what might stand as a 'simultaneous second', similar to the terminal pair (the 'goodbyes') (see work o n closings of G. Button, forthcoming). For the terminal pair there is indeed perfect i n d e p e n d e n t consensus; they occur utterly simultaneously. But for the prior pair there is, ambiguously, con sensus or fast reaction, o n e party waiting until the other has recognizably made a m o v e , a n d t h e n responding to it. To get a sense of the issues involved here, w e can look at another case of Closings; o n e which is in some ways very similar, b u t in crucial ways quite contrastive. I n this case there is n o possibility of seeing 'consensus'. This is a man talking to his elderly mother. 4.14 [Her:I:5:6:R] Heath : Well yyou leave that w'n With me. (0.3) -» Janet : Yes ,alri. * :ght,, -> Heath : ^ — ' AJ J right, (0.5) Heath : Bye:, Janet : Bye darling, 1

I n the first place, J a n e t ' s 'alright' is n o t necessarily a Close-initiation, but an agreement to a n arrangement. By coming in on it in terminal overlap, with another 'Alright', H e a t h retroactively provides that it is the first of a Closeinitiating pair. A n d in contrast to fragment 4.13, there is n o subsequent consensus. Janet remains silent at a point where she might herself volunteer the first of the ter minal pair. I n this case we can see H e a t h placing his 'Alright' in such a w a y that, that J a n e t has d o n e a first is unequivocal; i.e., e n o u g h of it has b e e n p r o d u c e d before h e starts u p . I n fragment 4.13, Lottie places h e r ' O k a y ' in such a w a y that it is n o t clear whether she is responding to E m m a ' s , or whether they have both i n d e p e n d e n d y a n d for all practical purposes simultaneously, m o v e d to close the call. I n fragment 4.14 we h a v e a clear case o f ' y o u started it'; in 4.13 a possible case of 'we agree'. (And see fragment 2.1 as another ambiguous case involving a possible 'saying the same thing at the same time' consensus).

JEFFERSON NOTES O N 'LATENCY' IN OVERLAP ONSET

211

Now, the fact that people d o h a v e the capacity to react to a very small particle - w h e n they are in various ways primed for its occurrence - m a y provide them with the capacity to achieve a display of consensus in contrast to response, specifically b y r e s p o n d i n g quickly e n o u g h to m a k e it at best ambiguous as to whether they're acting i n d e p e n d e n t i y or responsively. A n d in fragments 4.11-4.13 we m a y b e seeing such achieved i n d e p e n d e n t consensus occurring in what is otherwise a 'blind spot' in conversation. T h a t is, the consensus is achieved in part b y p r o d u c i n g what is systematically characterizable as transition-space overlap; a m e r e coincidence of two i n d e p e n d e n t activities. I n a sense, I've g o n e to a lot of trouble a n d e n d e d u p within a m e m b r a n e ' s thickness of w h e r e I started. T h a t is, I started out with a batch of cases which I initially took to b e s o m e o n e starting u p after hearing that another was still talking after a point of possible completion. I argued that they might not b e that at all, but yet another sort of transition-space onset, providing for a 'blind spot' in which just those particles which m a k e the new talk 'interruptive', are not heard. And now I ' m trying to argue that perhaps some of those cases are indeed perfectly well heard. But I think that m e m b r a n e - t h i n difference matters. It can matter in terms of managing to exhibit i n d e p e n d e n t - m i n d e d consensus w h e n o n e has, in fact, re sponded to what another is doing. It m a y also matter in terms of something like 'rights to a turnspace'. Roughly, if one can m a n a g e one's talk in such a w a y as to achieve a lawful p l a c e m e n t of it, i.e., in the transition space, t h e n o n e m a y h a v e every bit as m u c h right to b e talking as did the prior speaker in his turnspace. A n d I would argue that the vast majority of apparent transition-space onsets - Lastitem, Terminal, Possible Completion, and U n m a r k e d Next-Positioned - are lawful. That is, that they are p r o d u c e d b y reference to seeing that a n utterance is just ending or just ended. But there is this very recurrent business of a recipient starting u p in the tran sition space for a n utterance, finding that what h e took to b e ' a n utterance' was not yet a n utterance, a n d that the current speaker is still talking.

5. W h a t is ' C o m p l e t i o n ' Perhaps the materials and considerations I've g o n e through could b e the basis for a complaint against the w a y a lot of recipients d o a lot of their talk. T h a t is, if they would only wait a bit longer, they'd see that the speaker isn't finished just because he's arrived at a possible completion point. T h a t is, we could propose that something like Utterance + Pause should b e the w a y talk proceeds. Perhaps, indeed, the way it is designed to proceed, where, then, we're seeing a 'breakdown' of conversational rules. But that doesn't take into account the fact that even a very tiny silence can be treated b y a speaker as problematic; where, then, they will engage in activities to deal with that problem. Anita Pomerantz (1984) considers this issue: she works with rather substantial silences, of approximately one second. I'll just show a few fragments in which even a tiny delay can cause problems.

DISCOURSE STUDIES

212

Here are two fragments from a same corpus; calls between a family and some of their friends and neighbors. There are two w o m e n who call routinely to talk to the mother. When the phone is answered b y the youngest boy, h e tells his mother that it's 'Auntie Lorna' or 'Auntie Myra'. W h e n the older boys answer, they just tell her 'It's for you.' And that sets up the problem. In one such call we get the following. 5.1

-» -» -»

[Rah:B:l:(12):l] Jessie : "Hellof (0.5) Myra : Jessie? (0.3) Myra : It'smei, Jessie : Oh hellg. My:ra.

I don't know what the first silence of half a second is about, but I doubt whether it involves Myra waiting to see ifJessie is finished talking or not. What I want to note here is the shorter (0.3) silence, in the course of which I take it that Myra decides thatJessie hasn't recognized her voice, whereupon she pro vides more voice (with 'It's me'), whereupon Jessie, in unmarked next position, displays recognition - and with the 'Oh' shows that there was indeed a problem (see John Heritage, 1984, on 'Oh' as a 'change of state' token). In a subsequent call we get a tighter version of this phenomenon, with a silence of under two-tenths of a second; i.e., a 'micropause'. 5.2. [Rah:C:l:(l6):l] Jessie : Helloi? -> Myra : Hello Jessie. 0

1

(•)

->

Ida : Jessie :

r r

It's m £_:. •. Oh hello the' ah.

And in this case it so happens that just as Myra sees that there is a problem and moves to resolve it, her recipient has herself resolved it. This very short 'tolerance' b y a speaker for inter-turn silence can be seen ac ross a range of activities. I'll just show a few cases, in which, as in fragment 5.2, a speaker's problem-resolutional activity collides with a recipient's (slighdy de layed) response. [Rah:A:2:(9):l-2:R] Jessie : I've only jus goht in en ah thowt w'l p'hahps ih wz a bit late tih come round r d feoff n e e : . r i Alice : N o T;: ^ If you've ti.me, L

(•)

Alice : r r Have yih? -j Jessie : It's jist * thet evrybuddy seems fbe in t&wntihday.

JEFFERSON 5.4 -» -»

5.5 -> -»

213

[GTS-.1:1:44:4] Roger : I guess I jig didn'wanna be like evrybuddy Tg:lse. (•) Roger : j- rY'knp. i w? i Louise : You: a rn't. [G:93:AD:7:7] Bart : Well yp_u r'member when McKyen did tha:t? (0.3) Bart : r r L o:ng itime ago . . . Cal

5.6 -» -» -»


:

L 1

4-(NQ: )

J

)

[SBL:2:2:3:R:27] Zoe : Ellen had fiftfiejn. (0.3) Zoe : r [-hundred. Claire : Oh:, 1l

Here's another, that I think is particularly lovely. I t occurs i n t h e course of a conversation which is part of a series taking place during a hurricane a n d flood in New Orleans. J o s h is the c o m m a n d e r of the N e w Orleans Civil Defense. O p a l is a w o m a n w h o h a s b e e n trying to reach h e r b r o t h e r b y telephone. She finally has reached h i m , a n d h a s told J o s h s o m e of t h e eyewitness materials she got from h i m . T h e p o i n t at which t h e fragment starts involves a nice sort of p r o b l e m . J o s h is attempting to elicit further information, with a device Anita Pomerantz (1980) has described. Roughly, o n e does n o t ask, b u t provides for its relevance b y de livering t h e partial knowledge o n e h a s oneself. I n this case, however, that J o s h is C o m m a n d e r of Civil Defense m a y b e relevant to Opal's hearing of his ut terance. S h e (eventually) produces n o m o r e than a n acknowledgement token, and a token of the sort I h a v e elsewhere proposed to display particularly 'passive' recipiency. Parenthetically, the fact that she doesn't d o e v e n that for half a second m a y constitute a real case of silence telling a recipient that some sort of speaker tran sition should occur. 5.7

-» -» -»

[CDHQjII:252:R:3-4] Josh : Ah heard uh: first thet mere wz rilly s'm water in: Mohg'n City up uh: waiz deep uhr something. (0.5) Opal : Mm hm, (0-2) Opal : j- r Well i now he l Josh : U h how fo :r inlin.d uh did he Tsay 'nything

abfiut s'im: uhl::*: s'm Qthuh (0.6) -k *uh: cities like Lahfeeyet...

DISCOURSE STUDIES

214

What I want to be focussing on is that in the (0.2) silence following the ac knowledgement token whereby Opal shows herself to be still aligned as a recipi ent, both parties come to see that there is a problem, and what the problem is. Simultaneously they produce the same class of remedial activity. Opal starts to report something that a 'he' told her, and Josh asks what a 'he' said, 'he' being the temporarily lost brother with his eyewitness materials. It seems to m e that fragments 5.2-5.7 are particularly nice data for the issue of completion not being a matter of an utterance + pause. Even such tiny silences can be problematic and generate remedial activities. And, as in these cases, re sult in overlap. But it might be argued that I've got it wrong again; that what I'm treating as completion + problematic silence + remedial activity is in fact a speaker making the briefest of intra-utterance hesitations; that what they say next was in the first place designed to be part of a single utterance. That is, that neither the recipients nor I have waited long enough, and once again what we're seeing is some form of interjacent onset, or 'interruption'. Well, how long shall we wait? Over half a second? 5.8

-> ->

[NB:IV:13:R:25] Emma : yih dan't go tuh work til three.:,h (0.4) Lottie : Fou:r. (0.6) Lottie : i- rT'da:y 's Sa: turday. Emma : >• ^D_h: f * ou:r. (•) J

Emma

: TAoh that's i°r*ight.°

How about one second? 5.9 i-» U -> ->

Vic Wut's left aftih you take th'chez up [Fr:USI:8-9:R] (•) Mike Nothi ng. ° ( )° '•There's ^ ba:re fla2:r. Fred (1.0) Mike r Just f 1 -I ao:r. i Vic : Thez no ho: Ik? NO::-!.::. Mike f

r

L L

J

I'll stop at one second because some work I've done recendy indicates that one second is some sort of'standard maximum' silence (Jefferson, forthcoming), so whatever one might mean by 'waiting long enough', waiting beyond one second is waiting too long. Every now and then we can find materials in which a best characterization might in fact be that a speaker takes it he's finished and a recipient isn't sure, so he waits a bit. There was that possibility in fragment 5.7. And here is the one other case I have that feeling about.

JEFFERSON


215

5.10 [SBL:3:5:R:10] Milly : En I said I'd never heard Hunter so that w'd be most interesting. = *-> Ginny : = It certainly could . the. . et le . ast tha:t, one . thet = Milly : ( — H Y a h J shetsJ' J m e J Ginny : = I heard was (0.7) -> Milly : She Jnl'me he wz wonder, ful, Ginny : ^Yah Roughly, there m a y b e a particularly problematic combination of structure and intonation which makes it unclear whether 'At least that one that I h e a r d was' is a completed pro-formed re-use of the prior 'that would b e most mteresting', or is as yet incomplete, its speaker in a wordsearch for a description, e.g., 'At least that o n e that I h e a r d was ((pause)) fascinating.' But as far as I can see, for the vast majority of cases 'utterance + pause' does not capture the routine ways that recipients monitor talk in progress. W h a t it does recurrently catch is a particular sort of p r o b l e m posed for speakers. W h e n I talk about h o w recipients monitor a n utterance in progress, I talk in terms of possibly complete actions, a n d / o r syntactic possible completion. N o w and again people ask if I've considered intonation, and if so, does it matter? I say I have, a n d it doesn't seem to. At the start of this project o n overlapping talk I did a mass of transcribing, and was careful to catch a range of intonational features, assuming that would have a lot to d o with the p h e n o m e n o n . But while I was working with the materials, it didn't s e e m to matter. Speakers would, for example, p r o d u c e what I h e a r d as 'full stop' intonation a n d t h e n k e e p going (see, e.g., fragments 1.2, 1.3, 1.5, 1.7, 1.9,2.5,2.7,2.12,3.5,3.6,3.8,3.11,3.18,4.1,4.5,4.10 a n d 4.12). O r , for example, speakers would p r o d u c e 'continuing' or indeterminate intonation, a recipient would start u p , a n d there would b e clean transition (see, e.g., fragments 1.1, 1.5 and 2.3). W h a t held across cases of clean transition a n d recipient onset resulting in overlap alike was observable completed actions or syntactic possible completion. But recently, J o h n Local a n dJ o h n Kelly at the University of York have b e e n doing some interesting work with intonation in actual sequences of talk, a n d getting some possible results in terms of turn transition, etc. (see the following paper in this issue: Eds.); and their w o r k b r o u g h t m e to w o n d e r if I h a d b e e n looking at the p h e n o m e n o n properly. Perhaps intonation does matter, a n d somehow I'd missed it. I u n d e r t o o k a data run, to see h o w often 'clean tran sition' a n d 'overlap' were associated with 'completion' intonation and 'noncompletion' intonation. I immediately r a n into two problems. O n e was that I have only the grossest idea of what 'completion intonation' is, or could b e defended to b e . But there were some contours I supposed a n y b o d y would agree to b e completion or noncompletion intonation, so I focussed o n those. T h e second p r o b l e m has b e g u n to b e familiar. Roughly, if something 'works' it's there and if it doesn't 'work' it isn't there. So there w o u l d b e a n intonation contour that I'd identify as 'comple tion intonation' Tf it O r n i r r o r l in a n c a nf rAann f ^ n — n i f l ^ — '* H tr> v>oT*«n~ ±- « r

DISCOURSE STUDIES

218

I take it that the issue is far from settled. And I don't have the analytic re sources to handle it. Fortunately, there are sequentially-oriented phonologists like Local and Kelly who are beginning to make some inroads o n these sorts of problems.

Author's Note Some of the results are reported in Tilburg Papers in Language and Literature, No. 28, 'Two Explorations of the Organisation of Overlapping Talk in Conversation: Notes on Some Orderliness of Overlap Onset', and 'On a Failed Hypothesis: 'Conjuntionals' as Overlap Vulnerable'. For a version of the former see Jefferson (1983).

References Button, G. (forthcoming). Moving out of closing. In G. Button andj. Lee (Eds.), Talk and social organisation. Avon: Multilingual Matters. Heritage, J. (1984). A change-of-state token and aspects of its sequential placement. In J.M. Atkinson andj. Heritage (Eds.), Structures of social action: Studies in conversation analysis. Cambridge: Cambridge University Press. Jefferson, G. (1983). Notes on some orderlinesses of overlap onset. In V. D'Urso and P. Leonardi (Eds.), Discourse analysis and natural rhetoric Padua: Cleup Editors. Jefferson, G. (Forthcoming). Notes on a possible metric which provides for a 'standard maximum' silence of approximately one second in conversation. In Interdisciplinary approaches to interpersonal communication. Avon: Multilingual Matters. Pomerantz, A. (1980). Telling my side: 'limited access' as a 'fishing' device. Sociological Inquiry 50: 186-198. Pomerantz, A. (1984). Pursuing a response. In J.M. Atkinson a n d j . Heritage (Eds.), Structures of social action: Studies in conversation analysis. Cambridge: Cambridge University Press.

39 ... Sociolinguistic Resources, Individual Identities, and Public Speech Styles of Texas Women BarbaraJohnstone

n

" " t o g e t h e r with two colleagues, I a m studying the speech and writing of _ w o m e n from Texas to see h o w they draw o n linguistic resources provided J L b y history, culture, a n d society in constructing public voices that display personal a n d regional identity. (See J o h n s t o n e et al. 1993 for a n overview of the project. See McLeod-Porter 1992; Bean 1993; a n d j o h n s t o n e 1993 for preliminary reports of some findings.) O u r study explores the relationship between individual identity a n d shared sociolinguistic resources in the description of linguistic variation. It takes a per spective on sociolinguistic variation that supplements the results of studies of language use in relatively tight, focused, h o m o g e n e o u s communities. Using the particularistic techniques of e t h n o g r a p h y and discourse analysis, it asks a n d answers questions about h o w speakers draw o n regional a n d social models of phonology, morphology, syntax, a n d pragmatics as they construct a n d express individual identities in a mobile, multiethnic, linguistically heterogeneous con text. O u r approach to sociolinguistic variation expands o n w o r k b y LePage and Tabouret-Keller (1985) and others on linguistic "acts of identity." LePage a n d Tabouret-Keller point out that people in a given area can b e expected to speak alike only if they interact regularly a n d exclusively with the same people a n d are trying to express an identity similar to the identities projected b y the people they interact with. Their examples of the linguistic heterogeneity possible in a relatively nonfocused speech c o m m u n i t y c o m e from the Caribbean, w h e r e societies are racially a n d ethnically diverse a n d w h e r e various forms of English, French, a n d Spanish are available as resources. T h e heterogeneity that t h e y

Source: Journal of'Linguistic Anthroboloey vol. 5. no. 2.1995. oo. 183-202.

220

find leads t h e m to suggest that each individual creates his or h e r own ways of talking, ways that overlap m o r e or less with the ways others talk; they call into question the idea that languages and varieties are widely shared a n d that people necessarily form speech communities with joint orientation toward group norms. Sociolinguistic research in the U n i t e d States a n d Great Britain has until recently typically taken place in relatively tightly knit communities in which everyone is m o r e or less focused o n the same models for talk a n d m a k e s use of the same sets of linguistic variants. Beginning with the Linguistic Atlas projects, research o n language in the A m e r i c a n South has likewise sought to describe the speech of relatively closed, stable populations such as small rural towns and u r b a n ghettos, with the assumption that sociolinguistic models that accounted for the relatively stable patterns of variation found there would also account for patterns of variation in other settings. Southern states m a y indeed at one time have consisted of h o m o g e n e o u s speech communities. For most of the region's history, people could b e said to talk the way they did because of where they came from, since people stayed w h e r e they w e r e b o r n a n d interacted with the same people in their community, a n d few from outside it, all their lives. But does it still make sense, in view of current demographic and economic facts about the South, to suppose that Southerners, or even lower Southerners, or inhabitants of the Appalachians, or Texans, form speech communities analogous to those of H a r l e m or Detroit or inner-city Belfast? Outsiders a n d Texans alike tend to think that there is a single linguistic m o d e l toward which all Texans either aim or decide not to aim. But clearly all Texans do n o t speak alike or share the same n o r m s for speech. It is our view that c o n t e m p o r a r y Texas is m o r e like the C a r i b b e a n than like a n e i g h b o r h o o d in Belfast. T h e r e are m a n y very different resources available h e r e for the creation a n d display of idiosyncratic ways of being a n d sounding, a n d the m a n y Texans w h o interact with people outside their immediate com munities, either privately or in m o r e public media, h a v e access to varying sets of these resources. Texas is a particularly suitable locale for a study of the linguistic effects of multiple cultural models. It is the quintessentially southwestern state, culturally (and geographically) b o t h southern a n d western, a n d it is ethnically diverse, with large Hispanic, Anglo, a n d African A m e r i c a n populations that influence o n e another. To varying degrees, people in Texas speak English like Southerners, like Californians, like Midwesterners; they speak Spanish, TexM e x , Vietnamese, G e r m a n , Czech; they use features of Black English Vernacular and traditions of African A m e r i c a n oratory; they project colorful, direct western personalities or genteel, indirect southern ones (both are stereotypes that they can describe); they try to sound like the people o n Dallas, the people on The Cosby Show, or the p e o p l e on Beverly Hills 90210. Texans shape languages to use as they shape individual identities in the "multidimensional social space" (LePage a n d Tabouret-Keller 1985:180-186 et passim) suggested b y these and other possibilities. Not all linguistic possibilities are o p e n to all Texans: not all speak Spanish equally well, for example, a n d some are only able to m a k e small symbolic stabs as sounding southern. Some Texans do continue to live in tightly focused communities

B

JOHNSTONE

SPEECH STYLES OF TEXAS WOMEN

221

in which linguistic resources are limited and h o m o g e n e i t y is accordingly high. We are n o t trying to u n d o the results of dialectology b a s e d in such communities, but rather to extend the study of linguistic variation so that it provides a m o d e l of the language of p e o p l e in other, m o r e public contexts as well. O u r research population consists of w o m e n w h o live in Texas. T h e m e t h o d ology combines two techniques. (1) Interviews are conducted with w o m e n of a variety of demographic backgrounds to elicit attitudes about place, personal identity, a n d language use. A n understanding of generally shared sociolinguistic models o n which Texas w o m e n draw is thus allowed to emerge from their o w n senses of w h o they are and h o w they show it. (2) Detailed case studies of the lan guage often w o m e n w h o h a v e occasion to speak or write in public settings h a v e also b e e n conducted. O u r case-study subjects are w o m e n w h o use language in public in a variety of ways: they are teachers, politicians, journalists, storytellers, singers, a n d orators. These case studies are based o n two things. First, w e conduct with each w o m a n an interview that is designed to elicit speech in a variety of styles, as well as information a b o u t the subjects' histories as speakers and writers and their attitudes about language a n d speech. As part of the interview, we ask each w o m a n to tell us about what she sees as the m a i n influences o n the ways she talks, h o w she learned to u s e language as she does, a n d w h o a n d w h a t her models are. Second, w e e x a m i n e other examples of the subjects' writing and speaking in a variety of genres. Subjects of case studies include politician a n d professor Barbara J o r d a n , writer a n d entertainer S u n n y Nash, editorialist Molly Ivins, a n d county s p o k e s w o m a n Eliza Bishop, as well as other writers a n d pol iticians, teachers, a n attorney, a labor leader, a n d a country-western singer a n d lyricist. Each case study shows h o w o n e particular w o m a n draws on shared models for Texas speech, as well as o n particular models of speech from family and acquaintance as she creates a n d expresses a personal identity. T h e overall project is b o t h sociolinguistic a n d ethnographic in nature, as w e are interested both in h o w Texans talk a n d w h y they talk as t h e y d o and in Texans' o w n understandings of their ways of speaking. 1

In this article I explore h o w s o m e of these successful professional w o m e n conceive of themselves as language users a n d w h y ; I also illustrate s o m e ways in which their choices of h o w to say things serve as a metalinguistic display of the things that they say about w h o they are. S o m e of the cultural a n d linguistic resources that our subjects draw o n are easy for t h e m to talk about: they are glad, for example, to discuss h o w they feel about b e i n g Texans a n d h o w features associated with Texan speech fit into their linguistic repertoires, a n d they readily acknowledge the influence of speech styles associated with ethnicity and with vocation. But w h e n w e ask about b e i n g a n d sounding like women, they hesitate. These accomplished, visible w o m e n - m a n y from the generation that grew u p struggling for equal rights a n d influenced b y feminism - do n o t attribute very much about the w a y they act a n d speak to female role m o d e l s or m o d e l s of women's speech. This article examines w h y this should b e and looks at h o w these w o m e n define themselves as language-users. W h a t will emerge is that the sources of identity expressed in talk a n d other action are idiosyncratic a n d particular. Although all of the w o m e n w h o m w e h a v e stndiori î— •

222

DISCOURSE STUDIES

expectations a n d stereotypes as t h e y construct public voices, each does so differently, all draw equally or m o r e o n other linguistic a n d cultural resources, and all attribute the largest part of their identity to very particular facts about their o w n lives a n d families. Ideologies about region, gender, a n d ethnicity do n o t a p p e a r to serve, for these influential w o m e n , primarily as constraints on what they can d o ; rather, m o d e l s of speech a n d action that express linguistic ideologies are sources of creativity, selected from a n d combined in the expression of unique selves. I look in this article at what four w o m e n h a d to say to us about the sources of their linguistic abilities a n d styles. All four are middle-aged, between 40 a n d 60; all are natives of Texas, a n d all are successful and visible in professions that re quire t h e m to speak in public, to audiences of people w h o m they d o not and p r o b a b l y n e v e r will k n o w a n d w h o d o n o t k n o w t h e m , as representatives of their v o c a t i o n s . L i n d a C h a v e z - T h o m p s o n is E x e c u t i v e D i r e c t o r of the San Antonio Local of the A m e r i c a n Federation of State, County, a n d Municipal Employees. She has b e e n involved in u n i o n organizing a n d administration for most of her adult life. She grew u p in a Spanish-speaking family in various small towns n e a r Lubbock, Texas. Barbara J o r d a n , from Houston, is a former U.S. Representative, the first African A m e r i c a n w o m a n from the South to b e elected to the H o u s e . She is especially well k n o w n for speeches to the H o u s e Judiciary C o m m i t t e e concerning the N i x o n i m p e a c h m e n t a n d for two keynote addresses to Democratic conventions. She is n o w a professor at the University of Texas. Sunny Nash, also African American, is a freelance writer of newspaper columns, screenplays, short stories, and novels, as well as a musician, photographer, and art teacher. She is from Bryan, a medium-sized city in central Texas. Bettye Springer is an attorney in Fort Worth w h o presents workshops about issues in labor law such as sexual harrassment. She is a West Texan (from Lamesa) of European origin. I examine h o w these four w o m e n talked, in interviews with us, about lin guistic models and resources associated with gender, region, ethnicity, and vocation. M y m e t h o d o l o g y for this aspect of the study is simply to listen carefully and critically to w h a t our research subjects told us about w h o they are a n d h o w they talk. M y analysis is of course informed b y m y understanding that people are not accurate reporters of the finer details about their speech a n d of the fact that texts, b e they literary works or the conversations that I a m working with here, d o n o t completely speak for themselves. T h e s e w o m e n m a y well b e unaware of some of the real sources of their styles, or they m a y h a v e reason to conceal them. W h a t I a m interested in here is h o w they understand what motivates their choices a n d h o w their o w n understandings of their linguistic images are reflected in their speech.

1

Resisting G e n d e r Identification

T h e r e are vivid traditional m o d e l s for female s p e e c h b e h a v i o r in Texas. (See Carrington 1975, Crawford a n d Ragsdale 1992, Exley 1985, Fernea and

JOHNSTONE


223

2

D u n c a n 1977, F u r m a n 1980, a n d Marshall 1983, for example.) A s is Texas i n general, Texas w o m e n tend to b e treated in the state's popular m e d i a with a fair a m o u n t of positive hyperbole, as tougher a n d stronger, m o r e glamorous, m o r e neighborly, a n d "friendlier'' than other people are. For white Texans, m a n y of w h o m are of southern U.S. ancestry, one powerful image is that of the courtly but indirectly powerful "southern belle,'' with her "iron fist in a velvet glove," as it is often put. T h e r e are also popular images (propagated in magazines like the Texas Monthly, o n T V shows like The Eyes of Texas, a n d in local a n d national ad vertising campaigns) of rugged pioneer w o m e n working ranches in the southern and western parts of the state. Rural African Americans preserve m e m o r i e s of fiercely i n d e p e n d e n t former slaves farming a n d raising families b y themselves while their menfolk w e n t to the cities to find work. (Some of S u n n y Nash's w o r k has in fact involved documenting this history in central Texas.) C o n t e m p o r a r y church life provides African A m e r i c a n w o m e n with other m o d e l s that involve public speech and action. Traditional models for Hispanic w o m e n , p e r h a p s the most restrictive with respect to public action a n d speech, allow for consider able p o w e r within the family. Texas w o m e n can also draw o n models of w o m a n l y speech a n d behavior from elsewhere. T h u s a w o m a n wishing or needing to express a particularly female identity has m a n y resources to draw on, a n d except perhaps for Hispanic w o m e n , such a n identity n e e d not imply powerlessness or the lack of a public voice. But all four research subjects resisted or rejected explanations of their linguistic behavior based o n gender. T h e y insisted that they d o not, a n d h a v e never, m a d e or oriented to g e n d e r distinctions, a n d all except labor leader Chavez-Thompson denied that g e n d e r discrimination h a d figured m u c h in their lives. W h e n we asked t h e m whether they h a d b e e n aware as children of different linguistic or social expectations for girls a n d boys, they said that such distinctions had not applied to t h e m . H e r e , for example, Chavez-Thompson talks about noticing different expectations for interaction a m o n g m e n a n d w o m e n w h e n she was a child: 3

Chavez-Thompson.-'Wb.eri [my parents] moved to town, the biggest shock for my mother is that three or four couples could sit and visit in the living room together. The standard for them was that they would visit, eat; the men would go to one place and the women would go to the kitchen and, and, and do the gossip in the kitchen. We approached it differendy. In fact, when I was very young, I wondered, "Why did they do that?" you know, because I wanted in and listening in to what the men were saying, because I found that very interesting. And, and they said, "Get away from here," you know, you're sort of like, you're bothering us." Well, I, I was litde, so maybe I was bothering them. [Laughs.] Johnstone: [They were] attributing it to being a child, yeah. Chavez-Thompson: Yeah, because I was a child at the time. But I never, I, I really never separated that which women are supposed to do and that which men are supposed to do. My

first marriage didn't set any barriers; uh, certainly my second marriage did not. And, and so, I've really perhaps not, been there for, obstacles many many other women have had

placed before them. Although some obstacles are, you know, have always been there.

224

DISCOURSE STUDIES

Chavez-Thompson describes h o w females in h e r family were expected not to participate in the m e n ' s conversation, b u t w h e n the m e n would ask her to leave, she attributes it to h e r having b e e n a nuisance, not to h e r having b e e n a girl. T h e same reluctance to identify herself with other w o m e n characterizes her descriptions of interactions as an adult. She brings u p the topic of "obstacles m a n y other w o m e n h a v e h a d placed before t h e m " b u t denies having faced such obstacles in her marriages. Chavez-Thompson did describe h o w being a w o m a n m a d e h e r j o b difficult at times. H e r e she tells about first coming to work for the government employees' union: I came to work for the organization in June of 1971. But it was basically just, you know, having to battle the fact that I am a w o m a n . . . and that I am at the same level as business agents of plumbers and pipe fitters and, and whatever organization; it was difficult. Even, even my own - even here within this organization; it was difficult, because a HISPANIC woman, in the union business, is a bigger, a, a, a more rare t h i n g . . . . I signed up the first 13 people in this union. And I had botties thrown at me, when, when I was out there. . . . It was a long time before, uh, people of that mentality treated me differentiy than, than a secretary. W h e n she was asked, however, whether being in this position required her to learn to talk or act in m o r e traditionally male ways, Chavez-Thompson hesitated, first appearing to say yes a n d then firmly saying n o : Johnstone: Do you think that, when women, including yourself, get into, get into pos itions like yours, they have to learn to talk like men or to act more like men or would you just, would you describe it some other way, how you managed to do that? Chavez-Thompson: There are times when you have to, to be just like a man, I guess, if that's, that's, talking is, is Johnstone: - Yeah, that's part of it. Chavez-Thompson: - is part of that. But I've, I've had a few of, quite a few of those moments, but I've never acted like anything but what you'd describe as a lady. Or a woman, you know, I just, I do things, as I think they are. The biggest respect that I getfrom someone is that they look at me as their union representative, not as a woman. Not as a woman acting like a man, but just because of the talent that I may have, to resolve their problems. Chavez-Thompson does n o t want others to identify h e r with her gender; she wants the people w h o m she works with to "look at m e as their u n i o n represen tative, n o t as a w o m a n . " N o r does she want to attribute to g e n d e r a n y influence h e r o w n speech or behavior ("not as a w o m a n acting like a m a n " ) . Yet she described situations in which it is clear that others d o identify h e r with her gender a n d in which it would b e easy to describe Chavez-Thompson's behavior as "acting like a m a n . " T h e following excerpt illustrates b o t h (she was describing a recent conversation with a group of sheriffs deputies): Chavez-Thompson: But, but, in fact, in fact, in fact, they st-they call me "Mom." Johnstone: Oh yeah?

JOHNSTONE

SPEECH STYLES O F TEXAS W O M E N

225

Chavez-Thompson: Yeah. Now they call me "Mom," but every once in a sein a, in a while, / ask them, "What kind of mother are you calling me?" Uh, because [laughing] honest! They're bad sometimes! But the question here for them - and, and for me is that occasionally a city manager or occasionally a department head, ah, that because I'm a woman they're going to be able to walk all over me, or because I'm a woman I don't know how to take them on. And every once in a while, ah, I have to show them, that they're getting to deal with me at the same level as they would a man. And, uh, I don't often have to use the strong language that I, I, I am sometimes prone to do. Ah, I know some words, but I, and, and, and I am prone to do that. Both in turning " M o m " into a j o k e about a familiar profanity ("What kind of mother are you calling me?") and in describing her "proneness" to use "strong lan guage," Chavez-Thompson is, by standards that she herself is aware of, "acting like a m a n " ; a n d in calling h e r " M o m , " others are, at least superficially, treating her like a w o m a n (though the playfully aggressive tenor of the whole interaction could b e described as m o r e typically m a l e t h a n female; Tannen 1993a). Yet Chavez-Thompson continues throughout to b e very reluctant to acknowledge either. She does not want to describe herself in these ways and does so only w h e n pressed, interestingly often becoming less articulate - hesitating and stuttering when she does. Barbara J o r d a n also denied having noticed gender distinctions in expected linguistic behavior as a child: Johnstone: Were there other contexts in which you noticed, uh, that girls were supposed to speak differently than boys, or less, or more, or [more carefully? Jordan:[I really can't, I really have no recollection ofthat being a distinction. J o r d a n rejected a n explicit question about g e n d e r differences in public oratory and denied speaking differently to w o m e n than to m e n : Porter:\ have a couple of questions for you about, about what I'd call women's style of, of speech making. Do you, have any sense of whether or not your style is different from say some of the men counterparts that you've uh encountered in your career? Jordan: I, uh, I just, I s-1 have, I have no sense ofthat because I don't know how a man does his speeches or gets them and, and, um, presentation. I, I have no idea. Porter: Well, whenever you're preparing a speech and you know that you're speaking to an audience predominandy of men as opposed to an audience predominantly of women, do you make adjustments to meet the needs of that particular, those particular groups? Jordan: [Laugh] I don't. I really don't. I, uh, and I have, I will of course think of the group, and, uh, for instance it was late December when I went to Dallas to speak to a group of, uh, the women's foundation or whatever, and when in preparing that speech I thought, "This is a, a women - this is a woman's group and they will want something specifically about women and so after I do the rest of the stuff that I want to say here I better throw in something like that," and I'll try to find a good quote that may be appropriate for the occasion, that sort of thing.

U n d e r Porter's repeated questioning, J o r d a n did eventually acknowledge that a n audience of w o m e n might "want something specifically about w o m e n " and that as a result she "throws in" something like "a g o o d quote" "after I do the rest of the stuff that I w a n t to say." This acknowledgment of a n effect of audience g e n d e r o n J o r d a n ' s speech is about as m i n i m a l as possible, a n d its phrasing reflects its intent, as J o r d a n departs from h e r n o r m a l l y very formal style to use expressions like "stuff" a n d "throw in." W h e n w e pressed h e r a b o u t w h a t a w o m e n ' s g r o u p m i g h t want to h e a r about, J o r d a n a p p e a r e d very deliberately to frustrate our expectations: Porter: Now would you move towards something that was more story-oriented, anecdotal for women, as opposed to Jordan: No, more theoretical, philosophical in general. Porter:¥oT, for? Jordan: For women, right. 1

Porter was finally forced to b a c k off the issue, a n d J o r d a n h a d the last w o r d : itorfer.Tnteresting. Um, so then you're not conscious of any kinds of adaptations that speakers make when they're addressing different [audiences.] Jordan: [I'm not]lam not. Attorney Bettye Springer also resisted suggestions that h e r linguistic style might b e related to h e r g e n d e r or that of h e r audiences. H e r e , asked w h e t h e r h e r "par ticipatory" style in the w o r k s h o p that w e observed could b e a style typical of w o m e n , she b e g a n to agree, b u t immediately backtracked, ending b y suggesting that a person's style is a matter of "who y o u go a r o u n d with." Johnstone: Some, uh, some people would want to claim that that kind of more partici patory style was more more likely to be a woman's style, and the more kind of lecture style was more of a man's style. How do you, how do you feel about that? Big generalization, of course, but Springer: Let me think. That very well could be; now the fellow that's in there for my lawfirmrightnow that's participating in the mock trial, uh, M F , I, I've worked with him a lot on programs and I've seen him take questions and do participatory, but maybe not quite to the extent I do. But I have seen him do that some, uh, but and I will say that the woman who works with me, uh, who but you didn't meet, but her name is S G - she's a fifth-year lawyer. I guess you'd say I've trained her, and I, we still work together in doing training, and I see that she's adopted the style that I have, because she's seen me the most. So maybe it's who you go around with. Springer b r o u g h t u p the topic of role m o d e l s (she called t h e m "proteges") in the interview, partly in order, it seemed, to m a k e the point that hers h a d b e e n men: Springer: You didn't - One thing that you didn't ask me about, that I thought you might, was did I have a protege. Bean: Do tell. Springer: And, and I really did have a protege that, that might be of interest to you.

JOHNSTONE


227

Chavez-Thompson agreed that she h a d not h a d m a n y female role models: Bean: So you've had role models, Hispanic role models, but not too many women. Chavez-Thompson: No, not too many women. Just very few and far between. Writer Sunny Nash also repeatedly denied having h a d role models of either gender: Johnstone: When you, when you do your writing now, are you influenced by some authors more than others, or? Nash: No Generally, generally, the things just kind of pour out of my own experience. More specifically, she said that she h a d not learned h o w to b e a public speaker from either of the w o m e n in h e r family: Johnstone: How did you learn to be a public speaker, for example?= Nash:=Oh. Johnstone: From your parents, from your grandmother? Nash: No, no, my mother is probably the shiest person in the world [laughing]: she doesn't even talk to me. [Laughs] She's very, very shy. . . . And my grandmother talked to herself a lot and, uh, and prayed out loud s- but I think, see I had all kinds of uh projects of my own when I was a kid. Although a closer linguistic analysis would indubitably show that the Unguistic interaction of the interviews, and the speech of our subjects, was in some ways clearly affected b y our a n d their gender, the subjects a p p e a r e d to b e m a k i n g little strategic use of "sex-class-linked" linguistic features (Tannen, in press) in their talk. Like that of Chavez-Thompson, Springer, a n d J o r d a n , Nash's speech in the interview reflects her rejection of traditionally female m o d e l s for Texas speech a n d of stereotypical m o d e l s of w o m e n ' s speech in general. N o n e , for example, displays the indirectness or the positive politeness associated with southern female speech style; n o n e makes a display of linguistic hesitancy; all speak with fairly low-pitched voices. 4

M a n i p u l a t i n g R e g i o n a l Identity The w o m e n ' s reluctance to talk about their speech styles in terms of g e n d e r contrasts strikingly with their willingness to talk about themselves in terms of region. All but N a s h claim to orient their style to m o d e l s of Texanhood, and all including Nash talk about using their Texanness strategically, especially with outsiders. Models for speech a n d other action identified as particularly Texan are ubiquitous: Texans h a v e for several generations b e e n encouraged to think of their state as unique and superior, a n d Texas's size a n d population n o w makes it a target for specialized advertising that plays o n traditional images such as the state flag a n d the shape of the state, as well as o n ways of dressing, acting, a n d talking which are seen as especially Texan. (See Duffer a n d Sewell n.d.) People can "talk Texan" in varying degrees. Some use speech a n d discourse

228

DISCOURSE STUDIES

only occasionally, mainly with non-Texans, for strategic purposes such as getting attention or seeming exotic. Many Texas speakers symbolize their regional iden tification on a more regular basis, making more consistent use of a few features that have become linguistic badges of Texas identity. These include the use of y'all, at least in the nominative and oblique cases and often also in the possessive (y 'all's), the monophthongization of / a y / preceding voiceless obstruents (in words like night and like), and discourse-level displays of extreme "positive politeness" (Brown and Levinson 1987) such as extended, elaborate greetings. There are, of course, still many small-town or rural Texans, many older, who "talk Texan" because, living in small, homogeneous communities and interacting mainly in a dense, multiplex social network, they have limited linguistic resources from which to choose. These are the speakers favored as informants in traditional dialectology and modeled best b y much current sociolinguistic theory, but none of our research subjects are in this category. Since the traditional stereotypical Texan - the relaxed, neighborly, southmidland-accented cowboy in roper boots and hat - is a white man, state imagery is more easily available to white w o m e n than to others. Bettye Springer talked at length about the role of Texanness in her presentation of self to outsiders: 5

Bean: We do seem to have some sort of attitude about the state. Do you have any particular feeling about the state? Springer: [Emphatically] Oh I, I do, um and I can tell you, I can - with an anecdote. When I lived in Chicago, particularly in Chicago and everyone was from somewhere, and they were either Irish or they were Polish or they were this or they were that and my best friend in law school was Hungarian, but everybody was something. Johnstone: That's real important in Chicago. Springer: And, and people would always ask me, "Well, what are you?" and I'd always tell them, "I'm a Texan." [Bean laughing] Because I can remember growing up, I never was a German, which is what my Daddy was, or anything else; you were a Texan. . . . So I always had that identity and I'm still - there's things that that we have in the state that I'm not particularly proud of, like guns and stuff, but u h . . . but I'm proud of where I'm from. Springer talked at length about what she felt characterizes the interactive style of Texans - a combination of independence and friendliness - which she illustrated with the example of Texas drivers who sometimes make their own rules on the road but still wave at one another. W h e n we asked Barbara Jordan if she felt that Southernness was part of what shaped her style, she hesitated, saying that she hadn't always been proud of the region. But her identification with Texas was clearer. Well, I, uh, uh, I'm proud of the South too. I wasn't always but, uh, I grew to become, uhh, fond of the South and of course it's the race question that, uh, made the South a negative for me for so long, but when I was a member of Congress, uh, the Southerners would y- sort of get together: there was an affinity among the Southern contingent in the United States House of Representatives which you could not count on in any other region of the country.... When when I was there, uh, Texas was was sort ofa a

JOHNSTONE


229

thing apart, and we at least we considered ourselves a thing apart [laughing] andjust a little bit superior, eh, to tell the truth of the matter. And that's a junction of Texas braggadocio, that's what that is. L i n d a Chavez-Thompson told us an anecdote of a genre that one hears repeatedly in Texas: a story about surprising outsiders with what is called Texas "friendliness." ("Friendliness," in this context, usually m e a n s active courtesy to strangers. For example, "driving friendly," as motorists are encouraged to do with road signs, m e a n s pulling onto the shoulder to let others pass. T h e most com m o n l y a d d u c e d contrast with this m o d e of behavior is the behavior, experienced or imagined, of p e o p l e from "the East" or "the North.") Johnstone: You think of yourself as a Texan, then, in addition toChavez-Thompson: - Yeah. Yeah. Very much a Texan.... We had a convention here in 1988. The people from California came. I do not know what they expected. And I got more compliments from, "Ohh, ya'll are just WONDERFUL; the people HERE they're so friendly!" I wondered what they had ever heard about Texas for them to think that we wouldn't be. And, I think that this image of Texans, you know, as, uh, guns, you know, the the saloon, the the twang. I, I have no idea what other people think of us Johnstone: The kind of Wild West, [do you think? That's what they were expecting? Chavez-Thompson: [I think so, sort of like the Wild West, yeah. And, and so they were absolutely floored by how friendly - "We even asked somebody on the street for directions, and they were so nice! They gave us directions." [Laughter] Well, aren't you supposed to? Springer told a similar story, about surprising an airport bus full of dentists attending a convention b y being a "friendly Texan." Both Chavez-Thompson and N a s h talked about using their Texanness strategically. Chavez-Thompson spoke of "the showmanship issue": I'm very guilty, ever since [former governor] Ann Richards came into the picture in Texas, of using her phrases, her Texas twang. I find myself doing that when I am somewhere else. I twang it, you know, I actually d o ! . . . I can talk like Ann Richards, and I do it deliberately. Again, the showmanship issue. Nash also employs her Texanness for professional reasons ("I capitalize o n the Texas aspect"), although she denied feeling t h e n e e d for a real regional identity. For example, she used being a Texan (an exotic identity in the context, especially for an African American) to get bookings in the East Coast jazz-club scene that she tried to break into for a time. All four w o m e n , then, claimed that the ways they speak are sometimes influenced b y their being from Texas. S o m e illustrated this in their speech to us d u r i n g the interviews, t o o . B o t h S p r i n g e r a n d C h a v e z - T h o m p s o n use monophthongal / a y / , Chavez-Thompson variably a n d Springer fairly invariably, before voiceless obstruents, and b o t h used y'all occasionally, even in the fairly formal context of an interview with two or three nrofpssm-s

230

DISCOURSE STUDIES

Orientations t o Ethnicity Except in occasional, often commercial displays such as G e r m a n Oktoberfests a n d Czech kolache-baking contests, the Anglo Texans w h o m w e h a v e studied m a k e reference to their national origins less often t h a n do m a n y other E u r o p e a n Americans. (This is attested, for example, in Springer's comments quoted above about h o w m u c h m o r e important b e i n g "from s o m e w h e r e " was in Chicago than at h o m e . ) White ethnicity in Texas is to a considerable extent conflated with regional identity. But if regional identity can b e m o s t easily d r a w n o n and exploited b y white Texans, ethnic identity is most easily available as a m o d e l for action a n d talk to Hispanic a n d African A m e r i c a n Texans. J o r d a n , Nash, and Chavez-Thompson all talked about ethnic identity. E a c h orients to ethnicity in two ways: she draws on ethnic m o d e l s in some respects, a n d she actively resists a n d rejects ethnic models in others. Nash, for example, talked about rejecting h e r community's ways of talking, at the price of b e i n g seen as different: 6

Nash: My mother didn't speak that often, but when she did, she read to me, and she always used correct language. She always did; she never told me I had to, but she always did, so that's what I heard. Bean: This is the way it's done, kind of Nash: -Un-hnn. And, and now down the street, at the Candy Hill Market, now that's not the way people talked. And they would look at me real funny. I n h e r writing, Nash orients herself to h e r perception that whites expect blacks to write in an emotional way a n d attempts n o t to write that way, choos ing instead a reportorial style. Some of Nash's w o r k is also p a r d y aimed at undercutting what she sees as ethnic expectations: she has, for example, collected a n d restored antique photographs of a n d written newspaper columns about local rural African Americans w h o , contrary to stereotype, w e r e sometimes quite wealthy. For BarbaraJ o r d a n , African American speech style is an inalienable element of h e r own. Although she does n o t think of herself as a "preacher," she describes the influence of black preaching style o n that of a n y black orator: The, uh, what black or-, oratory is an important ingredient of the black experience because preachers have been so paramount in leadership roles in the black community throughout the country. You are going to find that there are any number of people who are public speakers who try to reflect what they have heard from their minister, and the black minister has a definable and totally, uh, in my opinion different style of oratory than anybody else in the country. And it is because of that role of the black minister, a role which I don't believe is paramount in any other race or group of people but it is that, it is there, it is there for us, and it is because of that preaching that, uh, influences, I think, the way that a person who is black would deliver an, an address or a speech. W h e n asked if she was "part of that tradition," J o r d a n exclaimed, "Of course I

JOHNSTONE


231

L i n d a Chavez-Thompson speaks of rejecting ethnic stereotyping as a child b y refusing to notice it: Chavez-Thompson: In school, uh, I know that I was less than ten years old and I remember feeling different because, uh, a little boy and a little girl called me a Mexican. And, Is-you know, I said, "Well, I know THAT" you know, it was like, Johnstone: [So what? Chavez-Thompson: [so what's wrong with that? And then, my own, my own, when I say my own, my, my friends, who were Hispanics, Mexican Americans treating me differendy, because I used to hang out with some of the Anglo girls in school, because they did, they, they, and I got, I never really saw and never treated myselfas any different. But she also described insisting to the m a n a g e r o n p r o p e r service in a restaurant where h e r family was treated sloppily, attributing the sloppiness to their being "the only Hispanic family in the place." Being Hispanic is connected with ChavezThompson's vocational identity in a very direct way: most of the m e m b e r s of the union local that she manages are Hispanic, a n d her j o b involves using Spanish fairly often. H e r speech (in English) to us was n o t markedly Hispanic-sounding. (Features of Hispanic English include, for example, the pronunciation of -ing&s [in], the devoicing of / z / , the deletion of the regular past-tense m a r k e r w h e n the root ends in a n alveolar consonant, and the use of intonation patterns that sound hesitant to non-Hispanics; see LeVine a n d Franco 1982 a n d Penalosa 1981.) She did, however, display her ethnic identity during the interview in nonlinguistic ways, describing at length h o w her grandfather taught her to m a k e formal speeches (in correct M e x i c a n Spanish) as a child a n d h o w m o v e d she felt w h e n she a n d her daughter visited Mexico City for the first time. 7

J o r d a n displayed her willingness to identify herself with h e r ethnic g r o u p in several ways. H e r speech style in general is characterized b y displays of verbal virtuosity: elevated diction, self-conscious word choice (and c o m m e n t a r y about word choice), extremely careful articulation, grammatical complexity, and a lack of editing. T h e s e are features that h a v e b e e n identified with African A m e r i c a n style (see Balester 1993) a n d traced historically to the valuation in black culture of the " m a n of w o r d s " (Abrahams 1976). J o r d a n also switched into markedly African American phonology a n d intonation at one point in the interview, during a narrative about talking to a n elementary-school class. Nash's linguistic style was not noticeably (to us) influenced in the interview b y African American models, though p e r h a p s the m a i n t h e m e of the interview was her assertiveness and uniqueness, characteristics that H e c h t et al. (1993) h a v e described as "core symbols" for African Americans.

E x p r e s s i n g Individuality To summarize, the w o m e n whose talk I h a v e e x a m i n e d h e r e are uncomfortable defining themselves, or having others define them, in terms of gender. Even when they can cite cases in which they are treated differendy because they are

232

DISCOURSE STUDIES

t h e m well, they d e n y talking or acting like w o m e n a n d claim to h a v e b e e n un influenced b y gender discrimination. O n the other h a n d , they readily acknow ledge acting a n d talking like Texans, either all the time or for strategic purposes; a n d they are not unwilling to cite ways in which their ethnic identities influence their behavior a n d speech. W h y should this be? O n e explanation is psychological. W h i t e h e a d a n d Smith (1986:166) cite a n u m b e r of studies showing that Americans attribute their own success to their ability, effort, a n d other internal factors, while they attribute failure to external factors such as the difficulty of the task or b a d luck. T h e w o m e n w h o m I have described h e r e are all highly successful, s o m e despite considerable odds. It m a y n o t b e surprising, t h e n , that t h e y e x p l a i n their a c h i e v e m e n t s i n t e r m s of psychological predispositions: they have n o n e e d to present themselves as victims of the process of identity ascription a n d n o real reason to present themselves as its beneficiaries. But this does not explain w h y they are, in fact, willing to talk about e m p o w e r m e n t via region a n d ethnicity. Another part of the explanation m a y have to do with the nature of the handicap imposed b y female gender. It could b e that, b y these people, talking or acting "like a w o m a n " is still seen as a liability in a way talking like a Texan or like an African A m e r i c a n or like a Hispanic is not. It could b e , in other words, that for these w o m e n ascribed gender identity is still a greater social and linguistic burden than is regional identity or ethnic identity. This is despite the fact that a Texas "twang" is a decidedly m i x e d blessing in terms of linguistic prestige for people in public life (and Texans are well aware of the stigma attached to Southernsounding speech) and despite the continuing stigmatization of African Americans and Hispanics. To understand completely why Chavez-Thompson,J o r d a n , Nash, a n d Springer talk about talk they w a y they do, we n e e d to e x a m i n e their o w n m o s t c o m m o n and most spontaneous explanations for the provenance of their acting and speak ing styles. For these w o m e n , ethnicity, gender, a n d region are n o t determinants of h o w people talk. Instead, they are resources that are selectively a n d sometimes consciously used in the presentation of self. T h e y reject the notion that they talk and act as they do because they are w o m e n or because they are African Americans, or Hispanics, or Anglos, or Texans. W h a t they present to the world w h e n they write, orate, negotiate, or lecture are, they insist, their u n i q u e selves, shaped by particular circumstances of birth, family, inclination, a n d luck. Each claimed repeatedly that she talked a n d acted the way she did because she always had, because she was b o r n that way, because of influences of particular friends and relatives early in life. Nash, for example, "always k n e w " that she was talented with words. She refers to h e r style as something that "started h a p p e n i n g " when she was in elementary school. Most explicitty, N a s h points out that the linguistic choices she makes are in service of the expression of h e r individual uniqueness, "who I am": Nash: Well, it's, it's actually in in the handling of the language. I want to make sure that I am not digressing from who I am when I speak. When I, when I speak, uh,

JOHNSTONE


233

verbally or when I speak, uh, in the written [mode], "This is me." I don't want to be somebody else when I'm writing a story about [my father-in-law], I want to be the same person who wrote the story about [a farmer]. . . . "That's the same person," you can read the story and say "I bet that's the same person [who wrote that]," even though the material is completely different. Springer attributes h e r success in court to "being herself" in response to a question that was m e a n t to get h e r to talk about b e i n g a woman: Johnstone: Do people ever talk about differences in style between male and female attorneys in the in the courtroom? Springer: Yes, they do, a lot. Uh, in the courtroom, uh, difference in style and people have told me that they think that I'm successful in the courtroom because I can identify with the jury, that the juries like me. And I haven't everfiguredout why, except that I try to smile, and I try to just be myself, and I don't put on any airs. W h e n Springer does attribute aspects of h e r style to the influence of a w o m a n (her mother), she does so, interestingly, b y m e a n s of a biological m e t a p h o r that immediately particularizes the source of influence: she claims to h a v e inherited her mother's "teacher genes." Linda Chavez-Thompson also speaks of a relative, the grandfather w h o taught her to m a k e speeches in Spanish, as a n important source of influence. But she attributes her success as a negotiator and administrator mainly to her o w n inborn talent, a "sixth sense" that she has always h a d , as recognized b y a m a n for w h o m she worked as a secretary: When I met this gentleman, he, maybe recognized the potential. Uh, whether he rec ognized a talent that I might have had. I have almost this, this sixth sense of people sitting down and beginning to tell me their problem, and I automatically can figure out what the solution might be and how do we get there. And I've always had this; I, I, I'm sort of like a middle child in the family. There's two older, and there's, there's, uh, five younger, but I'm the one that, that everybody looks to to make a decision. For these w o m e n , the primary reasons for Hnguistic choice are particular reasons. T h e y talk as they d o because they are the particular individuals they are, shaped by the particular circumstances of their birth, family, and community. If there are speakers w h o are victims of their gender, ethnicity, or place of origin, speakers who "are spoken b y " language rather than creating language as they speak, the successful public speakers w h o m we are studying d o n o t feel that they are in that group. T h e s e w o m e n are resourceful in two ways: as m e m b e r s of a mobile, heterogenous society, they h a v e m a n y resources from which to choose as they construct ways of expressing their uniqueness, and they are good at doing so. As Linda Chavez-Thompson puts it: A n d I think that the advantage - a n d I use it, believe m e , I use it - the advantage is, uh, I A M female. I A M Hispanic. I am, uh, labor. Of course, some people would take that as three strikes against m e , but, b u t that's

DISCOURSE STUDIES

234

N o t all speakers are as resourceful. For some, limited education or limited contacts outside a small, tight c o m m u n i t y m e a n that fewer choices are available; for others, the aptitide or motivation to choose effectively is lacking. I d o not m e a n to suggest that ideologies about gender and other social attributes do not exert e n o r m o u s influence over h o w p e o p l e talk. I d o want to point out, though, that the process is a n indirect one, always m e d i a t e d b y the individual.

Discussion I n a 1993 article, D o n n a Eder reports on a study of working-class adolescent girls' teasing about r o m a n c e a n d sex. E d e r found that the girls referred to traditional concepts of h o w females were supposed to act in teasing each other, b u t that the teasing functioned m o r e to strengthen b o n d s a m o n g the girls t h a n to socialize t h e m as females. Stereotypical romantic a n d sexual behavior was a resource to b e played with rather than a source of inescapable influence. This finding leads Eder to call for m o r e study of "the spontaneous a n d innovative side of an individual's behavior" (1993:29) and to wonder whether adults as well as children might not selectively use d o m i n a n t social ideologies as resources. O u r study of Texas w o m e n suggests that the answer is yes. To see ideologies about gender, ethnicity, and region primarily as constraints o n what a person can d o is to miss the fact that individuals use ideologies creatively. As speakers decide, consciously or unconsciously, h o w to sound, they pick a n d choose from a m o n g the available ways of sounding. O u r interviews with Texas w o m e n show that this is in fact the way they themselves u n d e r s t a n d their language. People, at least people like these, those w h o shape public discourse, are motivated at least as m u c h b y the n e e d for self-expression - the expression of a unique, differ entiated self - as they are b y the necessity to replicate social ideologies. T h e w o m e n w h o m w e are studying take various approaches to being Texans and to being w o m e n , as well as different approaches to their ethnic backgrounds a n d to their vocations. T h e y are m e m b e r s of a c o m m o n speech community in that they all participate, to o n e extent or another, in a large public network of influence, but they cannot b e said to orient to the same set of n o r m s for speech or other behavior. Instead, their sociolinguistic identities are constructed in discourse. These Texans shape languages to use as they shape individual identities in the social space defined b y the axes of region, gender, vocation, ethnicity, a n d ideologies a b o u t talk, as well as b y m o r e p a r t i c u l a r a x e s of family, community, psychology, and the n e e d for individuation.

Acknowledgements An earlier version of this article was presented at the American Anthropological Association Annual Meeting in November 1993, in a session on gender ideology in discourse organ ized by Deborah Tannen and Susan U. Phillips. I am grateful to them and to the session's discussants and audiences for comments and suggestions. I am also grateful to Judith Mattson Bean, one of my coworkers on the Texas women's speech project, for ideas arising

JOHNSTONE


235

from many discussions of the interview transcripts that I draw on here, and to Bean and Delma McLeod-Porter for their help with interviewing and transcribing. Anonymous reviewers for the Journal of Linguistic Anthropology also provided useful suggestions.

Notes 1. Both in its concern with speech and writing for public audiences and in its use of sociolinguistic data collected in relatively formal interviews, the project thus focuses on relatively non- "vernacular" speech, in contrast to many other sociolinguistic studies. This is because we are interested in the public sources of influence that increasingly supplement private neighborhood and family networks as agents of linguistic change. 2. The material in this paragraph is gready condensed, in the interest of space. It is drawn from the language-attitude interviews described above and from the enormous liter ature in history, political science, literary studies, and anthropology on Texas women. 3. Transcription conventions: I have indicated overlapping talk with brackets and latched talk with equal signs; the transcriptions otherwise employ normal English orthography and punctuation. I am more interested in this article in what the subjects said than in how they said it, and this system makes the transcripts easiest to read. Capitalization indicates unusual emphasis. Material that I have left out is indicated with ellipses. 4. There is not much scholarship to date on discourse-level features of southern speech; see Johnstone 1992 for an overview of what has been done. There is, of course, an enormous body of scholarship on language and gender in general, but this is not the place to review it. See Tannen 1993b, for example, for work on gender and inter actional style; Eckert 1989 on sociolinguistic effects of gender; Graddol and Swann 1989:12-40 and Sachs et al. 1973 on biological versus learned aspects of vocal pitch. 5. Monophthongized or glide-reduced /ay/ before voiced sounds is common throughout the South; it is specifically monophthongization before voiceless obstruents that appears to be associated with Texas identity. Bailey 1991 shows that opinion-poll respondents who rate Texas as a "good" or "excellent" place to live are more likely than others to use monophthongal /ay/ before voiceless obstruents. 6. The issue of religious identity and ethnicity connected with it is more complicated. I have not dealt with it here. 7. The variety spoken in this sort of context is not actually called, by its speakers, "Spanish." It is a variety that involves frequent, extremely fluent switches from English vocabulary and syntax to Spanish and back, with the conversation as a whole some times seeming more Spanish and sometimes more English. The case can be and has been made that this is a distinct "language" or "dialect" (linguists do not have an appropriate term) rather than the result of switching between two separate languages. Linda Chavez-Thompson, like many Texas Hispanics, calls the variety "Tex-Mex" (another of the case-study subjects uses the expression "speaking mixed") and em phatically denies that people are speaking Spanish when they speak this way; further west, it is sometimes referred to as Pocho Spanish. See Anzaldua 1987, OrnsteinGalicia 1981, and Penalosa 1981.

References Cited Abrahams, Roger D. 1976 Talking Black Rowley, MA: Newbury. Anzaldua, Gloria 1987 Borderlands/La Frontera: Thf. Nan MecK™ <s°" * ? , — =

A — *

236

DISCOURSE STUDIES

Bailey, Guy 1991 Directions of Change in Texas English. Journal of American Culture 14:125-134. Balester, Valerie 1993 Cultural Divide: A Study ofAfrican American College-level Writers. Portsmouth, NH: Boynton/Cook-Heinemann. Bean, Judith Mattson 1993 "True Grit and All the Rest": The Expression of Regional and Individual Identities in Molly Ivins' Discourse. Southwestern American Literature 19:35-46. Brown, Penelope, and Stephen C. Levinson 1987 Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Carrington, Evelyn M. 1975 Women in Early Texas. Austin: Pemberton Press. Crawford, Ann Fears, and Crystal Sasse Ragsdale 1992 Women in Texas: Their Lives; Their Experiences; Their Accomplishments. 2nd edition. Austin: State House Press. Duffer, Dyanne, and Amy Sewell n.d. Marketing towards Texans. Manuscript, Texas A&M University. Eckert, Penelope 1989 The Whole Woman: Sex and Gender Differences in Variation. Language Variation and Change 1:245-267. Eder, Donna 1993 "Go Get Ya a French!": Romantic and Sexual Teasing among Adolescent Girls. In Gender and Conversational Interaction. Deborah Tannen, ed. Pp. 17-31. Oxford: Oxford University Press. Exley, Jo Ella Powell, ed. 1985 Texas Tears and Texas Sunshine: Voices ofFrontier Women College Station: Texas A&M University Press. Fernea, Elizabeth W, and Marilyn P. Duncan, eds. 1977 Texas Women in Politics. Austin: Foundation for Women's Resources. Furman, Necah Stewart 1980 Texas Women vs. the Texas Myth. In The Texas Heritage. Ben Procter and Archie P. McDonald, eds. Pp. 167-183. Arlington Heights, IL: Forum Press. Graddol, David, and Joan Swann 1989 Gender Voices. Oxford: Basil Blackwell. Hecht, Michael L., Mary Jane Collier, and Sidney A. Ribeau 1993 African American Communication: Ethnic Identity and Cultural Interpretation, Newbury Park, CA: Sage Publications. Johnstone, Barbara 1992 Violence and Civility in Discourse: Uses of Mitigation by Rural Southern White Men. SECOL Review 16:1-19. 1993 Discourse, Dialect, and Identity among Texas Women. Paper presented al Southeastern Conference on Linguistics, Auburn, AL, April. Johnstone, Barbarajudith Mattson Bean, and Delma McLeod-Porter 1993 Ethnography and Discourse Analysis in Dialectology: Methods of a Study of Language, Region, and Personal Identity in Texas. Paper presented at Methods VII International Conference on Dialectology, Victoria, BC, August. LePage, R. B., and Andree Tabouret-Keller 1985 Acts of Identity: Creole-based Approaches to Language and Ethnicity. Cambridge: Cambridge University Press. LeVine, Elaine S., and Juan N. Franco 1982 New Dimensions in Cross-Cultural Communi cations: Some Anglo/Hispanic Comparisons. Palo Alto, CA: R&E Research Associates. Marshall, Carol 1983 The Fairy Tale and the Frontier: Images of Women in Texas Fiction. In The Texas Literary Tradition: Fiction, Folklore, History. Don Graham, James W. Lee. and William T. Pilkington, eds. Pp. 195-208. Austin: University of Texas Press. McLeod-Porter, Delma 1992 Guardian of Linguistic Tradition: A Case Study of a Southern Lady. Paper presented at ADS session, South Adantic Modem Language Association meeting, Memphis, TN, Oct. Ornstein-Galicia, Jacob 1981 Varieties of Southwest Spanish: Some Neglected Bask Considerations. In Latino Language and Communicative Behavior. Richard P. Duran.

JOHNSTONE


237

Penalosa, Fernando 1981 Some Issues in Chicano Sociolinguistics. In Latino Language and Communicative Behavior. Richard P. Duran, ed. Pp. 3-18. Norwood, NJ: Ablex. Sachs,Jacqueline, Philip Lieberman, and Donna Erickson 1973 Anatomical and Cultural Determinants of Male and Female Speech. In Language Attitudes: Current Trends and Prospects. Roger W. Shuy and Ralph Fasold, eds. Pp. 74-84. Washington, DC: Georgetown University Press. Tannen, Deborah 1993a The Relativity of Linguistic Strategies: Rethinking Power and Solidarity in Gender and Dominance. In Gender and Conversational Interaction. Deborah Tannen, ed. Pp. 165-183. Oxford: Oxford University Press. In press The Sex-Class-Linked Framing of Talk at Work. In Communication in, through, and across Cultures. Proceedings of the Third Berkeley Women and Language Conference. Mary Bucholtz, Anita Liang, and Laurel Sutton, eds. Berkeley, CA: Berkeley Women and Language Group. Tannen, Deborah, ed. 1993b Gender and Conversational Interaction. Oxford: Oxford University Press. Whitehead, George I., Ill, and Stephanie H. Smith 1986 Competence and Excuse-making as Self-presentational Strategies. In Public Self and Private Self Roy F. Baumeister, ed. Pp. 161-177. New York: Springer-Verlag.

i

i

40 The Role of Knowledge in Discourse Comprehension: A ConstructionIntegration Model Walter Kintsch

D

iscourse comprehension, from the viewpoint of a computational theory, involves constructing a representation of a discourse u p o n which various computations can b e performed, the outcomes of which are commonly taken as evidence for comprehension. T h u s , after c o m p r e h e n d i n g a text, one might reasonably expect to b e able to answer questions about it, recall or sum marize it, verify statements about it, paraphrase it, a n d so on. To achieve these goals, current theories use representations with several m u t u a l l y constraining layers. T h u s , t h e r e is typically a linguistic level of representation, conceptual levels to represent both the local a n d global meaning a n d structure of a text (e.g., the micro- a n d macrostructure, constituting the text base in v a n Dijk & Kintsch, 1983), a n d a level at which the text itself has lost its individuality a n d its information content has b e c o m e integrated into some larger structure (e.g., v a n Dijk & Kintsch's situation model). M a n y different processes are involved in constructing these representations. To mention just a few, there is w o r d identification, where, say, a written word like bank must s o m e h o w provide access to w h a t we k n o w about b a n k s , money, a n d overdrafts. T h e r e is a parser that turns phrases like the old men and women into propositions such as A N D [ O L D [ M E N ] , O L D [ W O M E N ] ] . T h e r e is a n inference m e c h a n i s m that concludes from the phrase The hikers saw the bear that they were scared. T h e r e are macro-operators that extract the gist of a passage. There are processes that generate spatial imagery from a verbal description of a place. It is o n e thing for a theorist to provide some formal description (e.g., a simula tion model) for h o w such processes can occur a n d for what the computational

Source: Psychological Review vol. 95, no. 2,1988, pp. 163-182.

KINTSCH

ROLE OF KNOWLEDGE IN DISCOURSE COMPREHENSION

239

steps were that led to a particular w o r d identification, inference, or situation model. It is quite another to control construction processes in such a way that at each point in the process exactiy the right step is taken. Part of the p r o b l e m has to do with the characteristic ambiguity of language: H o w d o w e m a k e sure that we access the financial m e a n i n g of bank, and n o t the m e a n i n g of riverbank? W h y did we parse the old men and women as w e did - m a y b e the w o m e n were n o t old at all. W h y did w e infer that the hikers were scared rather than that they h a d their eyes open, or a myriad of other irrelevancies? Of all the m a n y ways m a c r o operators could b e applied, h o w did we get just the right sequence to r e a c h a plausible gist without making the w r o n g generalizations? T h e n u m b e r of possible alternative steps is distressingly large in constructing discourse representations, and without firm guidance, a computational m o d e l could n o t function properly for long. T h a t is w h e r e knowledge comes in. General knowledge about words, syntax, the world, spatial relations - in short, general knowledge about anything - constrains the construction of discourse representations at all levels. Indeed, this is w h a t makes it possible to construct these representations. T h e r e is a striking unanimity a m o n g current theories about how this is d o n e . O u r conceptions a b o u t k n o w l e d g e use in discourse c o m p r e h e n s i o n are dominated b y the notions of t o p - d o w n effects a n d expectation-driven processing. Knowledge provides part of the context within which a discourse is interpreted. T h e context is thought of as a kind of filter through which p e o p l e perceive the world. At the level of w o r d recognition a n d parsing, it lets through only the ap propriate m e a n i n g of a n ambiguous w o r d or p h r a s e a n d suppresses the in appropriate one. T h r o u g h semantic priming, the feature counter of the logogen for bank as a financial institution will b e incremented a n d will reach its threshold before that of riverbankm the right context (Morton, 1969). Parsing a sentence is often thought of as predicting each successive constituent from those already analyzed o n the basis of syntactic rules (Winograd, 1983). Scripts, frames, a n d schemata constrain the inferences a n understander makes (as in Schank & Abelson, 1977), t h e r e b y p r e v e n t i n g the process from b e i n g s w a m p e d in a flood of irrelevancies a n d redundancies. Arithmetic strategies generate just the right hypothesis in solving a w o r d p r o b l e m a n d preclude the w r o n g ones (Kintsch & Greeno, 1985). I n a word, knowledge m a k e s understanding processes smart: It keeps t h e m o n the right track a n d avoids exploring blind alleys. People under stand correctly because they sort of k n o w what is going to c o m e . This p r o g r a m of research is well expressed b y the following quotation from Schank (1978, p. 94), which served as a motto for Sharkey's (1986) m o d e l of text comprehension: We would claim that in natural language understanding, a simple rule is followed. Analysis proceeds in a top-down predictive m a n n e r . U n d e r standing is expectation based. It is only w h e n the expectations are useless or w r o n g that b o t t o m - u p processing begins. Empirically, this position is questionable: Even fluent readers densely sample the words of a text, as indicated b y their eye fixations (Tust & Oarnpntor io»n\

240

DISCOURSE STUDIES

making the b o t t o m - u p m o d e appear the rule rather than the exception. C o m p u tationally, it is n o t a n easy idea to m a k e work. It is difficult to m a k e a system smart e n o u g h so that it will m a k e the right decisions, yet keep it flexible enough so that it will perform well in a b r o a d range of situations. O n t h e o n e h a n d , o n e needs to m a k e sure that exactly the right thing (word meaning, proposition, in ference) will b e constructed; for that purpose o n e n e e d s powerful, smart rules that react sensitively to subtle cues. O n the other h a n d , h u m a n s c o m p r e h e n d well in ever-changing contexts a n d adapt easily to n e w a n d unforeseen situations; for that p u r p o s e o n e n e e d s robust a n d general construction rules. Scripts a n d frames, as they were first conceived, are simply not workable: If they are powerful enough, they are too inflexible, a n d if they are general enough, they fail in their constraining function. This d i l e m m a has long b e e n recognized (e.g., Schank, 1982; v a n Dijk & Kintsch, 1983), a n d efforts have b e e n undertaken to m a k e expectation-driven processes sufficiently flexible (e.g., Schank's m e m o r y organ ization packets, or M O P s ) . I n this article, a n alternative solution to this problem will b e explored.

Construction of D i s c o u r s e R e p r e s e n t a t i o n s T h e traditional a p p r o a c h to modeling knowledge use in comprehension has b e e n to design powerful rules to ensure that the right elements are generated in the right context. T h e p r o b l e m is that it is v e r y difficult to design a production system powerful e n o u g h to yield the right results b u t flexible e n o u g h to work in a n environment characterized b y almost infinite variability. T h e approach taken h e r e is to design a m u c h weaker production system that generates a whole set of elements. These rules n e e d to b e just powerful enough so that the right element is likely to b e a m o n g those generated, e v e n though others will also be generated that are irrelevant or outright inappropriate. A n integration process will then b e used to strengthen the contextually appropriate elements a n d inhibit unrelated a n d inappropriate ones. Weak productions can operate in m a n y dif ferent contexts because they d o n o t h a v e to yield precise outputs; o n the other h a n d , a context-sensitive integration process is then required to select among the outputs generated. T h e integration phase is the price the m o d e l pays for the necessary flexibility in the construction process. T h e m o d e l p r o p o s e d h e r e has b e e n t e r m e d a construction-integration model to emphasize its most salient feature. I t combines a construction process in which a text base is constructed from the linguistic input as well as from the compreh e n d e r ' s knowledge base, with a n integration phase, in which this text base is integrated into a coherent whole. T h e knowledge base is conceptualized as an associative network. T h e construction process is m o d e l e d as a production system. I n d e e d , it is a generalization of the production system used in earlier work, such as the simulation-of-comprehension processes developed b y Fletcher (1985) and Dellarosa (1986) after the m o d e l of Kintsch a n d G r e e n o (1985). T h e m a i n dif ference is that instead of precise inference rules, sloppy ones are used, resulting in a n incoherent, potentially contradictory output. However, this output structure

KINTSCH

ROLE O F K N O W L E D G E I N D I S C O U R S E C O M P R E H E N S I O N

241

is itself in t h e form of a n associative net, which can b e shaped into a coherent text base via relaxation procedures in the connectionist manner (e.g., Rumelhart & McClelland, 1986). Thus, the m o d e l represents a symbiosis of production systems a n d connectionist approaches. Certain limitations of the present article are worth noting at this point, for it does n o t offer a solution to all the p r o b l e m s i n discourse understanding. T h u s , it is n o t primarily concerned with t h e specific strategies (or rules) for t h e construc tion of text propositions or inferencing. Instead, it relies in this respect o n what is available in t h e literature as well as o n whatever future researchers will b e able to come u p with. T h e only point it makes is that whatever these strategies or rules are, they will b e easier to formulate within the present framework, which allows t h e m to b e b o t h weaker a n d m o r e general. Thus, o n e n e e d n o t worry about constructing just t h e right inference, b u t can b e content with a m u c h slop pier rule. Sometimes, of course, even the latter type of rule m a y b e h a r d to c o m e by, whereas i n other cases (e.g., i n t h e w o r d p r o b l e m s discussed later) p r o miscuous hypothesis generation is straightforward (while selecting just t h e right one can b e tricky). 1

Knowledge Representation T h e process of constructing a discourse representation relies heavily o n know ledge. To u n d e r s t a n d h o w it operates, o n e must first h a v e a n idea of h o w the tobe-used knowledge is organized. Typically, theorists have tried to create knowledge structures t o s u p p o r t smart processes: s e m a n t i c nets, frames, scripts, a n d schemata. As has b e e n argued elsewhere (Kmtsch, in press), such fixed structures are too inflexible a n d cannot adapt readily e n o u g h to t h e d e m a n d s i m p o s e d b y the ever-changing context of t h e environment. Instead, a minimally organized knowledge system is assumed h e r e in which structure is n o t prestored, b u t gen erated in the context of t h e task for which it is n e e d e d . A n associative n e t with positive as well as negative interconnections serves this purpose. Knowledge is represented as a n associative net, t h e n o d e s of which are concepts or propositions. T h e nodes in this net are interconnected. Connections among n o d e s h a v e a strength value, which m a y b e positive, zero, or negative, ranging from 1 to - 1 . Nodes consist of a head plus a n u m b e r of slots for arguments. Thus, t h e nodes of the knowledge net are formally equivalent to the propositions used to represent texts (e.g., Kintsch, 1974). T h e slot specifies the nature of t h e relation b e t w e e n t h e h e a d a n d t h e argument. Slots m a y represent attributes, parts, cases of verbs, or arguments of functions. T h e y n e e d n o t b e n a m e d , b u t m a y b e n a m e d if the relation is a c o m m o n o n e (such as t h e cases of verb frames). The arguments of a proposition a r e concepts or other propositions. T h e n u m b e r of arguments i n a proposition m a y v a r y from o n e t o s o m e small n u m b e r . E x a m p l e s of c o m m o n types of n o d e s in t h e k n o w l e d g e n e t a r e (a) M A R Y , 2

3

(b) CAKE, (c) SWEET[CAKE], (d) BAKE[agent:MARY,object:CAKE], (e)

CONSEQUENCE

[condition:NOT[wATCH[agent:MARY,object:CAKE],effect:BURN[object:CAKE]].

Examples A a n d B are lexical nodes that h a v e associated with t h e m perceptual procedures that identify certain patterns i n the environment - pithpr f V i « n W o / . ^

242

DISCOURSE STUDIES

themselves or the written or spoken words, such as MARY and CAKE, respectively. I n the following I shall not deal with these perceptual procedures exphcidy. T h e semantic and associative relations into which MARY and CAKE enter, which constitute a part of the general knowledge net, are the focus of interest here. MARY and CAKE appear as arguments in E x a m p l e s C through E in various roles (the agent and object slots, etc.). There are two ways of looking at the list of propositions in Examples A through E. O n the one hand, it could be considered simply as a portion of a general knowledge network, whereas on the other hand, it could be considered the propositional base of a (brief) discourse, in which a particular M a r y bakes and burns a particular cake. Thus, the elements of which knowledge nets and text bases are constructed are the same. Indeed, as will be detailed later, text bases are formed by selecting, modifying, and rearranging propositional elements from the knowledge net. However, text bases are not part of the knowledge net, but separate structures with their own properties. Concepts are not defined in a knowledge net, but their meaning can be con structed from their position in the net. T h e immediate associates and semantic neighbors of a node constitute its core meaning. Its complete and full meaning, however, can be obtained only by exploring its relations to all the other nodes in the net. Meaning must be created. As a first step one could add all propositions in the net directly related to a node to obtain what Mudersbach (1982) termed thefirstlevel of meaning; then all propositions direcdy related to the propositions at the first level can be added to form a second level, and so on, until the whole knowledge net is involved. Note, however, that such a construction is a theoretical exercise without direct psychological correspondence. It is not possible to deal with the whole, huge knowledge net at once. Instead, at any moment only a tiny fraction of the net can be activated, and only those propositions of the net that are actually activated can affect the meaning of a given concept. Thus, the meaning of a concept is always situation specific and context dependent. It is necessarily incomplete and unstable: Additional nodes could always be added to the activated subnet constituting the momentary meaning of a concept, but at the cost of losing some of the already activated nodes. T h e notion of an associative net is not unfamiliar, but it is usually thought of as relating concepts only, not propositional nodes. Two extremely simple ex amples will illustrate the nature of such an associative net. First, consider the representation of the homonym B A N K in an associative net. Positive connections are indicated by arrows, negative ones by circles. Asterisks indicate further, unspecified nodes. Of course, each of the concepts and propositions shown in Figure 1 participate in the general knowledge net beyond the single connection shown here. A s a second example, consider the proposition BAKEJagent: P E R S O N , object:CAKE] (see Figure 2). O n c e again, only a fragment of the complete network is shown, just to illustrate certain types of connections. Representing knowledge in a propositional network has several advantages. Primarily, it provides a common format for the knowledge base and for the mental representation of discourse. Furthermore, we have by now considerable 4

5

KINTSCH


FIRST-NTL-BANK

MONEY

BANK1 »-

ISA

[BANK2

-0BANK2

ISA[banki,finan,inst]

bank

243

.RIVERBANK]

RIVER

OVERFLOW [RIVER.BANK2 ]

Figure 1: A fragment of the associative net for BANK (Positive connections are indicated by arrows, negative ones by circles; asterisks indicate further, unnamed nodes.) LiKE[Ag:PERS,obj: [ e a t [ p e r s , c a k e ] ] ] BAKE[Ins:SUN, obj:PART-OF-COUNTRY] 0

Q ^

PUT[Ag:PERS,obj:CAKE,LOC:iN-OVEN]

BAKE[Ag:PERSON,Obj:CAKE] •* *-* result[bake[...],hot[cake]]

BAKE[Ag:PERS,obj:SQUASH

PREPARE [Ag:PERS,Obj : d i n n e r ]

BAKE[Ag:PERS,Obj:BRICKS]

Figure 2: A fragment of the associative net for BAKE experience working with propositional structures, whereas other forms of repre sentation are less well understood (e.g., the spatial-imagery and linear structures of Anderson, 1983; the mental models of Johnson-Laird, 1983; or whatever the appropriate representation in the affective system might be, as in Zajonc, 1980). However, the decision to use a propositional representation does not imply that all other forms of knowledge are to be considered unimportant or nonexistent. It would b e desirable to expand the model to include nonpropositional repre sentations, but one would first have to learn h o w to operate with such forms of representation. Construction Processes The steps in constructing a text base according to the construction-integradon model involve: (a) forming the concepts and propositions directly corresponding to the linguistic input; (b) elaborating each of these elements by selecting a small number of its most closelv associated neiu-hhnrR from t l w ironoral Vrr-v..,io»-in-n

244

DISCOURSE STUDIES

net; (c) inferring certain additional propositions; a n d (d) assigning connection strengths to all pairs of elements that have b e e n created. T h e result is a n initial, enriched, but incoherent a n d possibly contradictory text base, which is then subjected to an integration process to form a coherent structure. I n Step A of this process, a propositional representation of the text is con structed from a parsed linguistic input, such as the words of a text with suitable syntactic annotations, and from a knowledge system as envisioned earlier. Note that the parser itself is n o t a part of the present model. T h e basic process of proposition building has been described in van Dijk and Kintsch (1983, chapter 4) a n d Kintsch (1985). I will illustrate it h e r e with some simple examples. Consider the sentence Mary bakes a cake. T h e parser output n e e d e d is Mary (agent of BAKE) bakes (predicate) a cake (object of BAKE). Mary, bake, and cake activate their corres p o n d i n g lexical nodes, a n d MARY and CAKE are assigned the roles of agent and object in the BAKE proposition. A s was suggested in Figure 2, BAKE requires a P E R S O N as agent, h e n c e a test is m a d e whether MARY is a person. This m a y either involve a search through the knowledge net for the proposition ISA[MARY,PERSON] or, should that search p r o v e unsuccessful, a n attempt to infer this proposition (e.g., the net m a y contain only propositions to the effect that MARY is a n a m e and that persons have names; exactiy h o w such problem-solving activity occurs w t h i n an associative net will n o t b e considered here). T h e present m o d e l , however, differs in a significant w a y from m y earlier conceptions: It does not require that the right, a n d only the right, proposition always b e formed. Instead, the construction rules for building propositions can b e weakened, allowing for the formation of incomplete or "wrong" propositions. Proposition building is on-line, and frequently, all the relevant information for building just the right one is not available on-line, leading to false starts or in completed attempts. I n the aforementioned example, this has n o interesting consequences; for example, if in response to the phrase Mary bakes... the pro position BAKE[MARY,$] - the dollar sign indicates an unfilled slot - is formed, it will simply b e replaced b y the complete proposition when the rest of the sentence is processed. However, consider a n e x a m p l e discussed b y Frazier a n d Rayner (1982): The linguists knew the solution of the problem would not be easy. H e r e , the on-line construction of propositions is n o t so simple. First, the proposition K N O W [ L I N G U I S T S , $ ] is formed. T h e n , b y the strategy of m i n i m a l attachment, the subsequent n o u n p h r a s e is interpreted as the object of K N O W , yielding K N O W [ L I N G U I S T S , S O L U T I O N ] . T h e final v e r b phrase, however, requires a subject, so [NOT[EASY[SOLUTION]]] is constructed. As Frazier and R a y n e r pointed out, this does n o t involve a reinterpretation of the sentence. Subjects d o n o t go back, noting in some way that solution of the problem h a d b e e n attached to the wrong proposition, a n d repair this error. Instead, the incorrecfly formed K N O W pro position s o m e h o w just disappears; the description of the integration process that follows shows how. A third example of proposition building, involving p r o n o u n identification, will b e discussed here. T h e r e exists g o o d psychological evidence that pronouns m a y activate m o r e than o n e possible referent (e.g., Frederiksen, 1981). T h u s , in

KINTSCH


245

U

The lawyer discussed the case with the judge. He said I shall send the defendant to prison" the following propositions would b e formed: D I S C U S S [ L A W Y E R , J U D G E , CASE]; SAY[LAWYER,[SEND[LAWYER,DEFENDANT,PRISON]]]; and

SAY[JUDGE,

Eventually, of course, the right interpretation comes to dominate the w r o n g one, as will b e shown shordy. I n Step B of the construction process, each concept or proposition that has been formed in Step A serves as a cue for the retrieval of associated n o d e s in the knowledge net. T h e retrieval process itself is m o d e l e d after well-known theories that h a v e b e e n developed a n d tested in the m e m o r y literature (Raaijmakers & Shiffrin, 1 9 8 1 ) . Suppose that n o d e i in the knowledge n e t is positively associated with n other nodes in the n e t Let s{i\j) be the associative strength between nodes i and j . T h e n the probability that the retrieval cue j will retrieve n o d e j is [SEND[JUDGE,DEFENDANT,PRISON]]].

Zs(i,h)

W

Note that each concept or proposition in the text base serves as a n independ ent retrieval cue, h e n c e the particularly simple form of the retrieval process. (An intersection search would b e required if the items in the text base acted as a c o m p o u n d cue.) O n each retrieval attempt, a n item a m o n g the associates of i is selected according to Equation 1. A sampling-with-replacement process is assumed so that d o m i n a n t associates m a y b e retrieved m o r e t h a n once. T h e n u m b e r of retrieval attempts with item i as the cue is assumed to b e fixed and is a parameter of the m o d e l , k. I n the examples that follow, k was chosen to b e 2 or 3, m o s d y to r e d u c e the complexity of these examples. However, o n e m a y speculate that the most realistic value of k would n o t b e m u c h higher, p e r h a p s between 5 a n d 7. Consider some simple examples. 1. Suppose the w o r d bank is presented as part of a text. It will activate the lexical n o d e s B A N K I (financial institution) as well as BANK2 (riverbank), plus some of their associates; for e x a m p l e , the construction process m i g h t pick from Figure 1: B A N K I , M O N E Y , FIRST-NATIONAL-BANK, B A N K 2 , R I V E R , O V E R F L O W [RIVER,BANK2].

2. Suppose the sentence Lucy persuaded Mary to bake a cake is presented as part of a text. T h e parser should provide a phrase structure tree as output, from which the proposition PERSUADE[LUCY,MARY,BAKE[MARY,CAKE]] is constructed. Each text proposition activates propositions closely related to it in the general knowledge net, regardless of the discourse context. For instance, in the case of BAKE[MARY,CAKE] we m i g h t thus obtain LIKE[MARY,EAT[MARY,CAKE]], P U T [ M A R Y , CAKE,IN-OVEN], RESULT[BAKE[MARY,CAKE], HOT[CAKE]],

PREPARE[MARY,DINNER].

These propositions are all closely associated with baking a cake (Figure 2). Note, however, that elaborating the text base in this w a y is not just a question of retrieving associated propositions from the knowledge n e t T h e arguments of these retrieved propositions m u s t b e treated as variables that are to b e b o u n d to the values specified b y the retrieval cue. T h u s , because MARY is the agent of the text proposition, MARY is m a d e the agent in the knowledge propositions it brines

246

DISCOURSE STUDIES

into the text representation, instead of P E R S O N in Figure 2. Similarly, although the informality of the present notation hides this, CAKE n o w is the particular one M A R Y bakes, n o t the generic one in Figure 2. These knowledge propositions func tion as potential inferences. O u t of context there is n o w a y of determining which of t h e m are relevant: M a y b e M a r y really likes to eat cake, b u t p e r h a p s she is in the process of cooking dinner, i n w h i c h case P R E P A R E [ M A R Y , D I N N E R ] might b e c o m e a macroproposition (what v a n Dijk, 1980, calls a construction). But it is also possible that n e x t she will b u r n h e r fingers w h e n she takes the cake out of the oven, m a k i n g HOT, which plays n o role at all in the other contexts, the rele vant inference. A t this point, the construction process lacks guidance a n d intel ligence; it simply produces potential inferences, in the h o p e that some of them might turn out to b e useful. 3. I n the third example, if the proposition S E N D [ L A W Y E R , D E F E N D A N T , P R I S O N ] has b e e n formed, the knowledge n e t contributes nothing, because o n e pre sumably does n o t k n o w anything about lawyers sending defendants t o prison. (Of course, LAWYER, D E F E N D A N T , a n d P R I S O N would each b e associatively elabor ated separately.) If, however, J U D G E rather than LAWYER w e r e the agent of S E N D . the elaboration process w o u l d contribute the information that this implies that the j u d g e is sentencing the defendant a n d so forth. Step C in the construction process, the generation of additional inferences, is necessary because n o t all inferences that are required for comprehension will, in general, b e obtained b y the r a n d o m elaboration m e c h a n i s m described earlier. I n some cases m o r e focused problem-solving activity is necessary to generate the desired inferences. E x a c d y h o w this is to b e d o n e is, however, b e y o n d the scope of this article. I merely wish to point out h e r e that in addition to the undirected elaboration which results from Step B of the construction process, there is still a n e e d for controlled, specific inferences. Two types of such inferences are of particular importance i n comprehension. Bridging inferences (Haviland & Clark, 1974; Kintsch, 1974) are necessary whenever the text base being constructed is incoherent (i.e., w h e n e v e r either t h e original text base itself or the elaborated text base remains incoherent b y the criteria discussed in v a n Dijk a n d Kintsch, 1983, chapter 5). Second, macropropositions h a v e to be inferred (as discussed in general terms in chapter 6 of v a n Dijk & Kintsch, 1983, a n d operationalized as a production system b y Turner, McCutchen, & Kintsch, 1986). Macropropositions are also elaborated associatively, as described in Step B for micropropositions. W h a t h a s b e e n constructed so far is a set of propositions containing the (micro)propositions directly derived from the text, a r a n d o m l y selected set of associates for each of these, the macropropositions generated from the text, and their associates. T h e final Step D of the construction process involves the speci fication of the interconnections between all of these elements. T h e r e are two ways in which elements are interconnected, (a) T h e propositions directly derived from the text (hence referred to as "text propositions") are positively inter connected with strength values proportional to their proximity in the text base. Specific realizations of this principle are described i n the discussion of Figure 4. fr\\ Tf

niTinnoitinnB

«

rrnA ia-rex pnnna^fjarl i n f-ho manoral V n

orlrro net

witH flip

KINTSCH

ROLE O F KNOWLEDGE IN DISCOURSE COMPREHENSION

247

strength value s[i,j), - 1 < s(i,j) < 1, and if iand^'become m e m b e r s of a text base, the strength of their connection in the text base is s(i,j). I n other words, proposi tions in the text base inherit their interconnections from the general knowledge net. Strength values are additive, u p to a m a x i m u m of 1, in those cases in which an inherited strength value combines with a text-base-determined connection. Consider, for instance, the portion of a network that is generated w h e n the word bank activates b o t h B A N K I a n d B A N K 2 , as well as the associations M O N E Y and RIVER. A possible pattern of connections is shown in Figure 3, where for simplicity, connection strengths h a v e b e e n limited to +.5 or 1. Alternatively, the graph shown in Figure 3 can b e expressed in matrix form as shown in Table 1. BANKI is associated with M O N E Y , B A N K 2 with R I V E R , b u t inhibitory connections exist b e t w e e n M O N E Y a n d BANK2 a n d between R I V E R a n d B A N K I .

.5 MONEY-"

u

1 BANKI O - ^ — o BANK2

u

5

— * ~ RIVER

-.5

Figure 3: Connections between

BANKI

and BANK2 and their associates

W E E D [LUCY, G A R D E N ]

L

U

C

Y

GARDEN

VEGETABLE[GARDEN]

Figure 4: The text base for Lucy weeded the vegetable garden An example of text propositions that are interconnected via their positions in the text base is shown in Figure 4. L U C Y is connected most strongly to WEED[LUCY, GARDEN], a n d least strongly to VEGETABLE[GARDEN]. Although there are m a n y possible ways to assign numerical connection strengths to express this pattern of connectivity, the o n e chosen h e r e results in the matrix shown in Table 2. Table 1: Connectivity matrix for the graph shown in Figure 3 Proposition 1. M O N E Y

1 -

2

3 0.5

-0.5

0.0

-

-1.0

-0.5

2. BANKI

0.5

3. BANK2

-0.5

-1.0

-

0.0

-0.5

0.5

4. RIVER

4

0.5

DISCOURSE STUDIES

248

Table 2: Connectivity matrix for the graph shown in Figure 4 Proposition

1

2

3

4

0.7 0.9

0.4 0.7 0.9

1.

LUCY

—

0.9

2.

WEED

0.9 0.7 0.4

0.9 0.7

3.

GARDEN

4.

VEGETABLE

-

-

0.9

-

Inferences inherit positive a n d negative interconnections from the general knowledge net, as seen in Figure 5. T h e result of the construction process is, therefore, a network expressable as a connectivity matrix, consisting of all the lexical n o d e s accessed, all the propositions that h a v e b e e n formed, plus all the inferences a n d elaborations that were m a d e at b o t h the local a n d global level and their interconnections. Integration T h e network that has b e e n constructed so far is n o t yet a suitable text representa tion. It was carelessly constructed a n d is therefore incoherent a n d inconsistent. At all levels of the representation, c o m p o n e n t s associated with the text elements were included without regard to the discourse context, a n d m a n y of t h e m are inappropriate. A n integration process in the connectionist m a n n e r can b e used to exclude these u n w a n t e d elements from the text representation (e.g., see Rumelhart & McClelland, 1986, a n d Waltz & Pollack, 1985, for discourse). Text comprehension is assumed to b e organized in cycles, roughly corres ponding to short sentences or phrases (for further detail, see Kintsch & van Dijk, 1978; Miller & Kintsch, 1980). I n each cycle a n e w net is constructed, including whatever is carried over in the short-term buffer from the previous cycle. Once the net is constructed, the integration process takes over: Activation is spread around until the system stabilizes. M o r e specific, an activation vector representing the initial activation values of all n o d e s in the net is postmultiplied repeatedly with the connectivity matrix. After each multiplication the activation values are renormalized: Negative values are set to zero, a n d each of the positive activation values is divided b y the s u m of all activation values, so that the total activation o n each cycles remains at a value of o n e (e.g., R u m e l h a r t & McClelland, 1986). 6

WEED [LUCY.GARDEN]

PULL-OUT [LUCY.PLANTS]

IN-GARDEN [PLANTS j NOT-IN-GARDEN [PLANTS]

O

Figure 5: Inferences generated from W E E D [ L U C Y , G A R D E N ] and their interconnections

KINTSCH


249

Usually, the system finds a stable state fairly rapidly; if the integration process fails, however; new constructions are added to the net, a n d integration is attempted again. Thus, there is a basic, automatic construction-plus-integration process that normally is sufficient for comprehension. This process is m o r e like perception than p r o b l e m solving, b u t w h e n it fails, rather extensive problem-solving activity might b e required to bring it back o n track. These processes will not b e considered further here. T h e result of the integration process is a n e w activation vector, indicating high activation values for some of the nodes in the net and low or zero values for many others. T h e highly activated n o d e s constitute the discourse representation formed o n each processing cycle. I n principle, it includes information at m a n y levels: lexical n o d e s , text propositions, knowledge-based elaborations (i.e., various types of inferences), as well as macropropositions. A few simple examples will illustrate what is at issue h e r e . Consider Lucy persuaded Mary to bake a cake, which was discussed earlier. T h e P E R S U A D E p r o position will pull in related knowledge items, just as was shown for BAKE. However, out of context the integration process will n o t yield any striking results. I n the context of Lucy made tomato soup and sauteed some porkchops with herbs. She set the table and persuaded Mary to bake a cake, the integration process has very different results: P R E P A R E [ L U C Y , D I N N E R ] emerges as the d o m i n a n t proposition (macroproposition) because most of the other propositions in the text base contribute to its activation value. T h a t the cake was hot, or that she put it into the oven, dis appears from the representation with activation values a r o u n d zero. Next, consider the example just discussed, w h e r e a perfectiy g o o d propositional strategy led to a wrong result. For The linguists knew the solution ofthe problem would not be easy, the text base that was constructed is shown in Figure 6. It cor responds to the connectivity matrix exhibited in Table 3 if connection strengths are assigned as in Table 2. (KNOW[SOLUTION] a n d NOT[EASY] are connected posi tively via KNOW[$] b u t negatively via EASY, which adds u p to 0.) T h e activation vector (.25, .25, .25, .25) corresponding to the assumption that all text p r o positions are equally activated initially is repeatedly multiplied with this matrix, r e n o r m a l i z i n g t h e o b t a i n e d activation values after e a c h multiplication as described earlier. To decide w h e n the activation vector has stabilized, the fol lowing criterion was established: A stable state is reached w h e n the average

T H E LINGUISTS KNEW T H E SOLUTION OF T H E PROBLEM W O U L D N O T B E EASY

KNOW[LING,$]

-NOT

KNOW[LING,SOL] EASY[SOLUTION]

Figure 6: The strategic construction of a text base: SOLUTION-OF-THE-PROBLEM is assigned to KNOW, then to EASY (The Dollar sign is a placeholder.)

250

DISCOURSE STUDIES

Table 3: Connectivity matrix for the graph shown in Figure 6 Proposition

1

2

—

0.9

1.

KNOW[$]

2.

KNOW[SOL]

0.9

3.

EASY

0.7

-1.0

0.9

0.0

4 . NOT

-

3

4

0.7

0.9

-1.0

0.0

-

0.9

0.91

-

change in the activation values after a multiplication is less t h a n .001. Although this is a n arbitrary criterion, even large changes (by one order of magnitude in either direction) m a k e only m i n o r differences in the final activation values obtained in this and m a n y other cases. I n the present case, this criterion is reached after 10 operations, yielding the final activation vector (.325, .000, .325, .350) that is, the w r o n g K N O W [ L I N G U I S T S , S O L U T I O N ] , which does n o t fit into the text base, has b e e n deactivated. T h e integration process similarly resolves the problem of multiple p r o n o u n referents. For The lawyer discussed the case with the judge. He said "I shall send the defendant to prison" propositions w e r e constructed for both lawyer a n d j u d g e as referents of he. H o w e v e r , the process of associative elabor ation generated some additional information for S E N D [ J U D G E , D E F E N D A N T , P R I S O N ] , but n o t for S E N D [ L A W Y E R , D E F E N D A N T , P R I S O N ] . T h e resulting text base is shown in Figure 7. To obtain the corresponding connectivity matrix (see Table 4), connection strengths a m o n g text base propositions were assigned as in Table 2, DISCUSS [LAWYER J U D G E , C A S E ]

SAY

SAY

[LAWYER, [ . ] ]

[ JUDGE, [.]]

SEND

SEND

[LAWYER,DEFENDANT,PRISON]

[JUDGE,DEFENDANT,PRISON]

^MPLY[[.],[.]]

SENTENCE IJUDGE.DEFENDANT]

Figure 7: The strategic construction of a text base: The pronoun he is identified with two potential, mutually exclusive referents (Instead of writing out whole propositions, the abbreviation [.] is used for the arguments of a proposition when they can be readily inferred.)

KINTSCH


251

Table 4: Connectivity matrix for the graph shown in Figure 7 Proposition 1. DISC 2. SAY[LAWYER] 3. SAY[JUDGE] 4. SEND[LAWYER] 5. SEND[JUDGE] 6. IMPLY 7. SENT

1 —

0.9 0.9 0.7 0.7 0.0 0.0

2 0.9

-

-1.0 0.9 0.0 0.0 0.0

3 0.9 -1.0

-

0.0 0.9 0.0 0.0

4

5

6

7

0.7 0.9 0.0

0.7 0.0 0.9 -1.0

0.0 0.0 0.0 0.0 0.5

0.0 0.0 0.0 0.0 0.5 0.5

-

-1.0 0.0 0.0

0.5 0.5

—

0.5

-

and among associates as in Table 3 (other assignments result in different numerical values for the final activation vector, b u t its pattern remains the same as long as the essential features of the matrix are preserved - for example, which con nections are positive, negative, a n d zero). Assume a n initial activation vector of (.25, .25, .25, .25, .25, 0, 0), reflecting t h e fact that only the text propositions themselves are activated initially. After 19 multiplications with t h e connectivity matrix, the two propositions i n which Aehad b e e n identified as the /aoiyerhave activation values of 0, w h e r e a s t h e c o r r e s p o n d i n g judge propositions h a v e activation values of .2 61 a n d .283, respectively. J u s t a little knowledge was enough to choose the correct referent. After this general description of the construction-plus-activation m o d e l , two specific applications will b e discussed i n m o r e detail: h o w words are identified in a discourse context, a n d h o w a propositional text base a n d situation m o d e l are constructed w h e n comprehension depends heavily o n activating a rich know ledge set. For that p u r p o s e , arithmetic w o r d p r o b l e m s w e r e chosen as the example, because the knowledge that needs to b e activated is particularly well defined in that domain, a n d u n a m b i g u o u s criteria of understanding exist - a solution is either right or wrong. T h e purpose of these examples is twofold: to show h o w t h e general framework p r o p o s e d can b e elaborated into specific models in these experimental situations, a n d to c o m p a r e the performance of these models with empirical observations a n d experimental results as a first test of the psychological adequacy of these models.

Word Identification i n D i s c o u r s e The first p r o b l e m to b e considered i n detail is h o w knowledge is used i n under standing the m e a n i n g of words i n a discourse. T h e previously sketched m o d e l implies that w o r d meanings h a v e to b e created a n e w in each context, that this is initially strictly a b o t t o m - u p process with context having its effects in the inte gration phase, a n d that this construction-plus-integration process takes time, with different factors influencing successive phases of the process. Context effects i n w o r d recognition are ubiquitous in t h e experimental literature, a n d the explanation of these context effects h a s b e e n a primary goal

252

DISCOURSE STUDIES

of theories of word recognition. Typically, it is taken for granted in these theories that because context influences w o r d recognition, contextual factors interact with the perceptual processes. Context effects are said to b e t o p - d o w n and expect ation driven a n d are said to facilitate (or sometimes interfere with) the perceptual analysis. Similar ideas w e r e once current in related fields, such as the "New Look" in p e r c e p d o n (Bruner & Postman, 1949) a n d the filter theory of attention (Broadbent, 1958). People perceive what they expect or want, attention filters out the irrelevant. Some words are recognized because the context favors them; others are n o t because the context inhibits them. H o w these t o p - d o w n effects of context are realized differs in detail a m o n g theories, b u t all the most influential current theories postulate interactive processes in which contextually expected words are favored. I n the logogen m o d e l (Morton, 1969, 1979), context primes semantic features that enter into the feature counter of a logogen and therefore strengthen that logogen. I n Forster's search m o d e l (Forster, 1976), perceptual analysis defines a candidate set that is then searched b y semantic relations or b y word frequency. I n Becker's verification m o d e l (Becker, 1976), b o t h a set of sensory candidates and a set of semantic candidates are created, with the latter being verified first. I n the cohort m o d e l of Marslen-Wilson a n d Welsh (1978), context is used to exclude m e m b e r s of the cohort from the very beginning. Norris (1986) has recendy reviewed these models and pointed out that they all involve s o m e sort of p r i m i n g m e c h a n i s m t h r o u g h which context effects are mediated. T h e m o d e l of h o w knowledge is used in discourse suggests a fundamentally different approach. Following earlier w o r k b y Kintsch and Mross (1985) and Norris (1986), the present m o d e l is neither interactive, n o r does it involve priming. As these authors h a v e argued, w o r d identification is not simply a matter of lexical access. Rather, it is a complex process that responds to different in fluences at different stages. These stages, however, are merely convenient verbal labels. In fact, processing is continuous, a n d there is significant temporal overlap between the different subprocesses defining these stages. I n the first stage (which was t e r m e d sense activation b y Kintsch & Mross, 1985), the n u m b e r of word candidates consistent with the perceptual input is progressively reduced through perceptual feature analysis. A s in Forster or Becker, a set of sensory candidates is created through perceptual analysis, b u t its size decreases as the analysis progresses. This process rapidly reduces the n u m b e r of w o r d candidates to some manageable n u m b e r , b u t n o t necessarily to one. At this point (probably at about 50 m s , see Fischler & G o o d m a n , 1978), the semantic context comes into play. Some small n u m b e r of lexical nodes has n o w b e e n selected, each one activating a few of its strongest semantic or associative neighbors in the knowledge network. If there is a n o d e whose associates fit into whatever context is present, it will be taken as the m e a n i n g of the to-be-identified word. W h a t fits is determined by the integration process sketched earlier. This is the sense-selection stage of Kintsch a n d Mross. Note that if the perceptual analysis h a d b e e n allowed to continue for a sufficient period of time, for most words it would have yielded a result eventually b v itself, and D r o b a b l v the s a m e nnp. Tt is hist that dip aKsnriaHnn r h p r t IIPII-IPH

KINTSCH


253

to shortcut this process. With h o m o n y m s , however, the association check plays a crucial role: Perceptual analysis alone cannot decide which m e a n i n g of bank to select in any given context. Sense selection b y m e a n s of an association check is the very first of a possibly very long series of contextual plausibility checks (Norris's term). It comes first because the associative/semantic context of a lexical n o d e can b e c o m p u t e d rapidly. As m o r e information about the context b e c o m e s available, the sentence and discourse m e a n i n g begin to emerge, and m o r e and deeper plausibility checks can b e performed as long as there still is time. This is the sense-elaboration phase, in which the meaning of a word is contextually explored and elaborated. However, once a response has b e e n m a d e in a recognition experiment, or once the process moves o n in a discourse, elaboration is terminated. T h u s , w o r d meanings are usually identified long before complex inferences are m a d e in c o m p r e h e n d i n g a discourse. At this point, a "meaning" has b e e n constructed for the word in this particular context. It consists of the lexical n o d e that has b e e n activated (the contextually inappropriate n o d e s that h a d b e e n activated h a v e b y n o w b e e n deactivated through the various context checks), the associative a n d semantic neighbors of that n o d e , the sentence a n d discourse context in which the word participated, and some inferences a n d elaborations that w e r e p r o d u c e d in the course of the various plausibility checks that explored the role of that word in the given context. W h a t do we n e e d to m a k e such a m o d e l of word identification work? We shall disregard the perceptual analysis a n d take for granted that a certain n u m b e r of appropriate lexical nodes has b e e n activated (e.g., multiple semantic nodes for a h o m o n y m ) . We then n e e d to c o m p u t e the sentences a n d phrases in which the word in question participates, or m o r e accurately, the propositions in which the corresponding concept token (for which the lexical n o d e serves as the type) plays a role. Finally, we need to construct inferences and elaborations w h e n necessary. A m o d e l of w o r d recognition that thus far is identical with the one favored here has r e c e n d y b e e n developed b y Norris (1986). Norris called it the "checking model" a n d compares and contrasts it with the other extant models of word recognition in the literature. I n Norris's model, the plausibility of w o r d candidates in any given context is evaluated. T h e recognition criterion for contextually plaus ible words is lowered a n d that for implausible words is increased. By manipulating criterion bias in this way, Norris accounted for a wide range of observations from threshold a n d other types of recognition experiments. Instead of equating plausibility with criterion bias, a different m e c h a n i s m integration - is used h e r e . This m e c h a n i s m has the great advantage of being applicable n o t only at the word-recognition level (which is w h a t Norris was concerned with), b u t it is equally suited to m o d e l i n g knowledge integration at higher levels. W h e n a word is perceived, one or m o r e lexical nodes are accessed, and some of their neighboring n o d e s that are closely related associatively or semantically are also activated. Similarly, w h e n a proposition is constructed, a n u m b e r of associatively a n d semantically related propositions are also constructed. Both rolof-orl

r n n r o n t o

-raid fori n r n n A e i r i A n c c o n m

f/-i r\oformina

f V i a nlonciVwliHr r v f

254

DISCOURSE STUDIES

the core words and propositions. A richly interconnected structure is thus formed, through which activation can spread, so that positively interconnected items strengthen each other, while unrelated items drop out a n d inconsistent items b e c o m e inhibited. O r , said differendy, implausible items will b e suppressed, whereas plausible ones support each other - at the level of word recognition as well as of textual integration. Time Course of Activation of Words in a Discourse Context T h e m o d e l of w o r d recognition just outiined is consistent with a great deal of experimental data. Norris (1986) has reviewed the w o r d recognition literature in great detail and shown that his checking m o d e l accounts for the rich empirical findings in that area better than any of its competitors. T h e construction-integration m o d e l is closely related to Norris's m o d e l . O n the o n e h a n d , it is m o r e specific in that it p r o p o s e s c o m p u t a t i o n a l p r o c e d u r e s b y m e a n s of which Norris's "plausibility check" could actually b e achieved, whereas o n the other hand ii replaces Norris's shift in criterion bias with the computationally m o r e feasible integration mechanism. It appears likely that the present m o d e l can handle all the data the checking m o d e l accounts for, in just the same way a n d for just the same reasons as the checking model. T h e r e is, however, another part of the literature o n word recognition that is n o t discussed in Norris (1986): the work o n w o r d identification in discourse. T h e empirical findings in this area are also in good agreement with the construction-integration model. I n a lexical decision task, the subject sees a string of letters and must decide as quickly as possible whether it forms an English word. If a target word is p r e c e d e d b y a closely related word, the response to the target w o r d is speeded u p (on the order of 2 0 to 4 0 ms) in comparison with unrelated control words. This priming effect has b e e n well d o c u m e n t e d for some time a n d is obtained in list contexts (e.g., Meyer & Schvaneveldt, 1971) as well as in discourse contexts (e.g., Swinney, 1979). However, the discourse context is actually irrelevant to the p r i m i n g effect. W h a t matters is merely the associative relation between the prime w o r d and the target word. As has b e e n shown repeatedly (Kintsch & Mross, 1985; Swinney, 1979; Till, Mross, & Kintsch, in press; also equivalent results ob tained with a n a m i n g task b y Seidenberg, Tanenhaus, Leiman, & Bienkowsky, 1982), h o m o n y m s will prime strong associates of both their meanings, irrespective of the discourse context and in spite of the fact that the context-inappropriate m e a n i n g of the h o m o n y m never enters consciousness. Furthermore, contex; appropriate inferences that are n o t associatively related to a priming word are n o t r e s p o n d e d to any faster than unrelated control words. However, all of this depends o n the a m o u n t of time allowed for the processing of the priming word. If the target w o r d closely follows the p r i m i n g word, so that the processing of the p r i m e is still in its initial stages, the results are as already described. However, if there is e n o u g h time for m o r e complete processing of the priming word in its discourse context, quite different results are observed. I n this case, context-^ appropriate associates are still primed, b u t inappropriate associates no longer!

KINTSCH


255

are, whereas context-appropriate inferences n o w b e c o m e strongly p r i m e d . This dme course of knowledge activation can b e described in m o r e detail b y some illustrative experimental results. In the study b y Till et al. (in press), subjects read sentences like The townspeople were amazed to find that all the buildings had collapsed except the mint. After the priming word mint they w e r e given a lexical decision task, with the target w o r d being either money, candy, or earthquake. T h a t is, the target was a context-appropriate associate of the p r i m e (money), a context-inappropriate associate (candy), or a topical inference word (earthquake), respectively. I n addition, the interval between the presentation of the p r i m e a n d the target w o r d (stimulus-onset asynchrony, or SOA) was varied from 200 m s to 1500 m s . I n the first case, the p r i m e could only b e incompletely processed; with an S O A of 500 m s , a somewhat d e e p e r processing of the p r i m e was possible before a response h a d to b e given to the target word; a n d with 1,000 m s , extensive processing of b o t h the p r i m e w o r d and its discourse context was possible. T h e data are shown in Figure 8. To keep this presentation simple, Figure 8 shows the average priming effects observed in the three experiments of Till et al. for S O A s of 2 0 0 , 3 0 0 , 4 0 0 , 5 0 0 , 1 , 0 0 0 , and 1.500 ms. T h e value shown for associates at 200 m s , for instance, is the differ ence between the m e a n reaction time for context-inappropriate a n d contextappropriate associates at that p r i m e - t a r g e t asynchrony. It is the average of two such values obtained in two different experiments - showing the data separately for each experiment merely complicates the picture without changing its essential features. T h e value for inferences, similarly, is based o n the difference between topic words a n d unrelated control words. T h e purpose of Figure 8 is, therefore,

60

31 40w 20

6 o

o

-20

1

1

1

1—

200 300 400 500 1000 PRIME-TARGET ASYNCHRONY (IN MSEC)

mt

lgu

• Associates

1500

Inferences

:re 8: Context effects as indexed by the reaction time difference to context ipropriate and appropriate associates or inferences as a function of processing time, Till, Mross, and Kintsch (1988)

256

DISCOURSE STUDIES

m e r e l y to give a n impression of the over-all shape of the results of this study (for m o r e detailed analyses, the original article must b e consulted). Targets that are contextually appropriate associates of the priming w o r d are primed at all four SOAs. Contextually inappropriate targets, however, are primed only w h e n the p r i m i n g w o r d is still in its initial processing stages; b y 400 ms inappropriate associates are n o longer activated. Topical inferences are primed only if there is ample time, m o r e t h a n 500 m s , for the processing of the prime a n d its discourse context. This observation implies that the topic was not inferred immediately as soon as the relevant information b e c a m e available, b u t was left for the sentence wrap-up period. Till et al.'s sentences w e r e written in such a way that the topic could h a v e b e e n inferred before the last w o r d in the sentence. This, however, is not what h a p p e n e d : Topics w e r e inferred only after the whole sentence was read, requiring m o r e than 500 m s of processing time. T h u s , the full contextual m e a n i n g of the p r i m e required about 1 s to emerge. D a t a like these suggest that the initial activation of lexical knowledge is i n d e p e n d e n t of the discourse context. W h a t matters is only the (relatively fixed a n d stable) associative/semantic context of each word b y itself. This stage of sense activation, however, is quickly followed b y a process of sense selection in which the discourse context becomes effective: By 500 m s , context-inappropriate associates are deactivated (see also Seidenberg et al., 1982, a n d Swinney, 1979). If given m o r e time, context effects g r o w even stronger: By 1,000 ms, contextu ally appropriate inference words are strongly a n d reliably p r i m e d even in the absence of associative connections (similarly for recognition, see M c K o o n & Ratcliff, 1986). Clearly, this pattern of results is in excellent agreement qualitatively with the m o d e l of knowledge use in discourse presented earlier. Right after a word is perceived, it activates its whole associative neighborhood in a context-independent way, with the consequence that strong associates of a w o r d are likely to b e re presented in working m e m o r y a n d h e n c e will b e p r i m e d in a lexical decision task, whether they are context appropriate or not. T h e knowledge-integration process then results in the deactivation of material that does not fit into the overall discourse context (such as context-inappropriate associates). N o t e that in order to disambiguate words on-line, the integration phase cannot b e delayed until the e n d of a processing cycle; w o r d senses are disambiguated before that. I n the m o d e l , therefore, as soon as a text proposition is constructed a n d its asso ciates h a v e b e e n generated, they will b e integrated into whatever context exists at that time in working m e m o r y . Thus, each processing cycle involves many integrations, a n d the single integration operation performed at the e n d of each cycle in m a n y of the examples discussed h e r e is merely a simplification, adopted whenever one is not concerned with the on-line generation of w o r d meanings. Finally, contextual inferences should require the most time to b e c o m e activated on the average because although they sometimes result from the initial knowledge sampling, in other cases repeated sampling or, further, strategic elaboration might b e required. Earlier, a n example was given of o n e of the texts used in the Till et al. (in press) study. T h e predictions of the m o d e l will b e illustrated b y m e a n s of this

KINTSCH


257

example. T h e aforementioned text (The townspeople were amazed to find that all the buildings had collapsed except the mini) h a s t h e following p r o p o s i t i o n a l representation: 1. 2. 3. 4. 5. 6.

TOWNSPEOPLE AMAZED[TOWNSPEOPLE,P3] COLLAPSE[P4] ALL-BUT[BUILDING,MINT] BUILDING MINT

Connection strengths of .9, .7, .4, a n d 0 were assigned to text propositions one, two, three, or m o r e steps apart in the text base (e.g., P I is two steps away from P 3 , connected via P2). Next, each text proposition was allowed to access at r a n d o m two of its neighbors in the long-term associative net. This process was simulated by having an informant provide free associations to phrases based on each of these six propositions. For instance, the phrase all buildlings but the mint elicited the associations many buildings a n d mint is a building. Of course, M O N E Y a n d C A N D Y were chosen as the associates of M I N T . Each text proposition was connected b y a value of .5 to its associates, yielding a n 18 x 18 connectivity matrix. Activation was t h e n allowed to s p r e a d from t h e text p r o p o s i t i o n s to t h e k n o w l e d g e elaborations. Specifically, an initial activation vector with 1/6's corresponding to the text propositions a n d zeros otherwise was multiplied with the connectivity matrix until the pattern of activation stabilized. As a result, text propositions achieved activation values b e t w e e n .0987 a n d .1612, d e p e n d i n g o n h o w closely they were tied into the text base, and the knowledge elaborations h a d m u c h lower activation values, b e t w e e n .0142 and .0239, with b o t h M O N E Y a n d C A N D Y having a value of .0186. Thus, at this stage of processing, M O N E Y a n d C A N D Y are equally activated. Activation continues to spread, however, a n d differences begin to emerge among the activation values for the various knowledge elaborations that h a v e been added to the text base. T h e reason for this is that the knowledge elaborations are connected n o t only to the text propositions that h a d pulled t h e m into the net b u t also to other text propositions as well as to each other. To a p p r o x i m a t e these interrelations, a connection value of .5 was assigned to any two propositions sharing a c o m m o n argument. Because the h o m o p h o n e mint contributed associ ations to the subnet that refers to b o t h of its senses, an inhibiting connection of -.5 was assigned to M I N T / C A N D Y a n d B U I L D I N G , whereas C A N D Y a n d M O N E Y them selves w e r e connected b y a - 1 . Continued multiplication of the activation vector with this connectivity matrix yielded a stable pattern (average change W Z

o S

u

o o Final

2

1

< D

CA w

EARTHQ

a

1

o

Immediate

Figure 9: The changing meaning of MINT (The activation values of all propositions direcdy connected to MINT at the beginning and at the end of the process. The [.] notation if used as an abbreviation for the arguments of a proposition.)

260

i

DISCOURSE STUDIES

Although m u c h m o r e research is needed to produce a m o r e ade quate picture of h o w the contextual m e a n i n g of words is constructed during discourse comprehension, h e r e is a technique that at least m a y help us to d o so. EARTHQUAKE.

1

A r i t h m e t i c Word P r o b l e m s

H o w children understand a n d solve simple word arithmetic p r o b l e m s provides an excellent d o m a i n to try out the construction-plus-integration model. Unlike with m a n y other types of discourse, there are clear-cut criteria for when a problem is solved correctly, and the formal knowledge of arithmetic that is necessary for its solution is easily defined. However, w o r d problems, like all other texts, share the ambiguity a n d fuzziness of all natural language. Not only formal, arithmetic knowledge is involved in understanding these problems, b u t all kinds of linguistic and situational knowledge. W h a t makes word problems h a r d - and interesting are often not their formal properties, b u t the w a y a p r o b l e m is expressed lin guistically a n d the w a y formal arithmetic relations m a p into the situations being described. T h u s , word p r o b l e m s are ideal from the standpoint of knowledge integration because it is precisely the integration of formal arithmetic knowledge a n d linguistic a n d situational understanding that is at issue here. Another reason for choosing the d o m a i n of w o r d p r o b l e m s is that there already exist alternative formal models of h o w children solve simple w o r d arith metic problems (Briars & Larkin, 1984; Kintsch & G r e e n o , 1985). Specifically, the w o r k of Kintsch a n d G r e e n o will b e taken as a starting point here. Their m o d e l represents a union of the work o n p r o b l e m solving in arithmetic b y Riley, Greeno, and Heller (1983) o n the one h a n d , and that o n discourse understanding b y v a n Dijk a n d Kintsch (1983) o n the other. Kintsch a n d G r e e n o (1985) added to the discourse-comprehension strategies of the v a n Dijk a n d Kintsch model some special purpose strategies for solving w o r d arithmetic problems, which they n a m e d the arithmetic strategies. For instance, if the m o d e l encounters a quantity proposition, such as "six marbles," it forms a set and tries to fill in the various slots of the set schema: what the objects are, the cardinality of the set, a specification of the objects (e.g., that the marbles are o w n e d b y Fred), and the relation between the present set a n d other sets in the p r o b l e m (the six marbles were given to Fred b y Tom, which might identify t h e m as a "transfer set"). Thus, the Kintsch a n d G r e e n o m o d e l for w o r d p r o b l e m s builds a text base in quite the same w a y as in the van Dijk a n d Kintsch general theory of text comprehension, but it then forms a very specialized situation or p r o b l e m m o d e l in terms of sets of objects a n d their interrelations. It solves a p r o b l e m b y recognizing a particular pattern of relations a m o n g sets (such as transfer-in or superset) and then using a stored-solution procedure appropriate to that case. T h u s , in terms of the foregoing discussion about knowledge use in discourse, the Kintsch and G r e e n o m o d e l is a "smart" m o d e l : Production rules are formulated in such a way that in each situation exactly the right arithmetic strategy is fired. T h e Kintsch a n d G r e e n o m o d e l of solving arithmetic word problems is useful in several ways. T h e m o d e l identifies different classes of errors, such as errors 7

KINTSCH


261

caused b y a lack of arithmetic knowledge, errors caused b y linguistic misunder standings, and errors that do not reflect a lack of knowledge at all but result from resource limitations. Certain formulations of word problems overload the resources of the c o m p r e h e n d e r , especially short-term m e m o r y , leading to a b r e a k d o w n in processing. A s Kintsch a n d G r e e n o h a v e shown, within each arithmetic p r o b lem type there exists a strong correlation between the frequency of errors m a d e in solving the p r o b l e m and the m e m o r y load imposed b y it, even though there are n o differences within p r o b l e m types in either the arithmetic or linguistic knowledge required for solution. T h e m o d e l distinguishes b e t w e e n linguistic and arithmetic errors a n d helps us to investigate to what extent errors m a d e b y second- a n d third-grade pupils are caused b y a failure to understand properly the text of the word p r o b l e m , rather t h a n b y a faulty knowledge of arithmetic (e.g., Dellarosa, 1986; Dellarosa, Kintsch, Reusser, & Weimer, in press; Kintsch, 1987). If certain linguistic mis understandings about the meanings of such key words as have more than, have altogether, or someaie built into the knowledge base of the model, the model produces a pattern of wrong answers and misrecall of the problem statements that strikingly parallels some of the m a i n types of errors that experimental subjects m a k e . This is a good example of h o w m u c h can b e achieved even with the use of knowledgepoor representations in studies of discourse processing. T h e Kintsch and G r e e n o model knows about arithmetic (its arithmetic strategies), a n d it knows about the meaning of words (its lexicon; a semantic net'm Dellarosa, 1986). However, it has no general world knowledge that would allow it to understand the situation described in a w o r d p r o b l e m . It merely picks out the crucial arithmetic informa tion from the discourse and builds a propositional text base for it. This is good enough for some purposes (e.g., the investigation of resource limitations or linguistic factors in understanding as mentioned earlier, or to predict recall, summarization, or readability as in Kintsch & van Dijk, 1978, a n d related work), b u t it is n o t good e n o u g h for other purposes. The limits of this a p p r o a c h are illustrated b y a well-known observation: If a word p r o b l e m is e m b e d d e d into a concrete, familiar situation or action context, it is much easier to solve than w h e n the same p r o b l e m is expressed abstracdy (e.g., Dellarosa et al., in press; H u d s o n , 1983). Thus, Five birds saw three worms on the ground, and each bird tried to get a worm. How many birds didn't get a worm? is easy for first graders, but There are five red marbles and three green marbles. How many more red marbles are there than green marbles? is very hard, even though the two problems are equivalent in form. The Kintsch a n d G r e e n o m o d e l does not account for this difference. W h a t is needed is a m o d e l in which all knowledge relevant to the understanding of a word p r o b l e m b e c o m e s integrated into a representation that is sensitive to arith metic as well as to situational information. I n the m o d e l to b e described shortiy, this is achieved b y forming m a n y different hypotheses about the arithmetic relations in the problem, instead of only a single one, a n d then b y looking for information in the text in support of each hypothesis. T h u s , situational a n d arithmetic information can c o m b i n e in formino- the ™-~vo~_

2fi2

DISCOURSE STUDIES

Arithmetic Strategies Arithmetic knowledge forms a special subset of a person's general knowledge network. Sets of objects can b e represented b y a propositional schema with the slots object, specification, quantity, and role (i.e., their relation to other sets) equivalent to the set schema of Kintsch and Greeno (1985). Superordinate schemata can b e similarly defined. T h u s , a T R A N S F E R - I N s c h e m a can b e set u p with slots for a START, T R A N S F E R - I N , a n d RESULT SET. With each such superordinate schema, various arithmetic procedures (such as the counting strategies of Kintsch & Greeno, 1985) can b e associated. Arithmetic knowledge is used in the same w a y as general world knowledge. T h a t is, propositions that represent various hypotheses about the arithmetic structure of a w o r d p r o b l e m are constructed as the text of the w o r d p r o b l e m is r e a d a n d b e c o m e part of the subnet. Several possible arithmetic hypotheses are constructed at each point, a n d whatever evidence in the text favors o n e or the other of these hypotheses is connected positively to it. T h e strategies required for solving arithmetic w o r d p r o b l e m s h a v e b e e n described in Kintsch a n d G r e e n o (1985) and Kintsch (1984) and h a v e been incorporated into the computer simulation of Dellarosa (1986). H o w e v e r , they are used h e r e in a different way. I n the aforementioned works, the intent was to circumscribe the conditions u n d e r which each strategy is applied so accurately that only t h e correct o n e is fired in each w o r d problem. H e r e , strategies fire pro miscuously w h e n e v e r they are supported, however weakly, b y the text, and it is left for the integration process to w e e d out w h a t is n o t wanted, just as all sorts of general knowledge propositions are activated that later turn out to b e useless. A problem is solved w h e n the right superordinate schema is m o r e strongly activated than its alternatives, which then triggers the desired arithmetic procedures. T h r e e forms of arithmetic strategies n e e d to b e considered. T h e r e are strat egies that form arithmetic hypotheses about sets, strategies that determine the nature of the connections between various text propositions and these hypotheses, a n d strategies that form superordinate arithmetic schemata o n the basis of which arithmetic calculations can b e performed. 1. H y p o t h e s e s a b o u t sets a r e p r o p o s i t i o n s of t h e form SET-[object:X, specification:Y,quantity:Z,role:W], w h e r e X refers to a n object, such as T U L I P ; Y is one or m o r e other text propositions, specifying X further - for example, PAST[LOCATION[TULIP,IN-GARDEN]] ; Z is a quantity proposition with X as argument - for example, F O U R T E E N [ T U L I F ] ; a n d W indicates the role of the set in some superschema, such as W H O L E or PART. 2. W h e n e v e r a quantity proposition is encountered in the text base, possible arithmetic hypotheses derivable from it are constructed (e.g., two otherwise identical propositions with the roles W H O L E a n d PART). Propositions in the text base that provide evidence for any of these alternatives are t h e n connected positively to it. Key words can b e used for this purpose, as in Kintsch a n d Greeno (1985): Collection terms such as altogether indicate W H O L E sets; give/take, ofthese, a n d have more/less than indicate PART sets. I n addition, general world knowledge about situations a n d actions is used to determine w h a t is a W H O L E a n d what are

KINTSCH


its PARTS. T h e strategies involved have b e e n described in Kintsch (1984): restricted subsets, conjunction, a n d time-ordered possession/location. Restricted subsets. If the specification of one set is m o r e general than that of another, the former is assigned t h e role of W H O L E a n d the latter that of PART. Ex amples are L A R G E - W I N D O W , S M A L L - W I N D O W versus W I N D O W , or O N - U P P E R - S H E L F , ON-LOWER-SHELF versus O N - S H E L F . Conjunction. If the object or specification of o n e set consists of the conjunction of the objects or specification of two other sets, the former is the W H O L E a n d the others the PARTS. This conjunction m a y b e explicit as in YESTERDAY, T O D A Y , a n d YESTERDAY&TODAY, or implicit, as in TEDDYBEAR, D O L L , a n d TOY. Time-ordered possession/location. If the specification slots of three sets contain either HAVE[agent,object] or LOCATiON[object,place], or the negations of these propositions, as well as information to establish a t e m p o r a l order, W H O L E a n d PART roles can b e assigned to the three sets according to the resulting patterns. For instance, if the specifications of three sets are T I M E I [ H A V E , J O E , P [MARBLES]]; TIME2 [GIVE, J O E , T O M , Q, [MARBLES]] , which implies TIME2 [NOT[HAVE,JOE,Q[MARBLES]] ; and TIME3[HAVE[JOE,Z[MARBLES]], S E T I is indicated as the W h o l e set. 3. T h e P A R T - W H O L E s c h e m a is t h e o n l y arithmetic s u p e r - s c h e m a to b e considered in the examples that follow, though various TRANSFER a n d C O M P A R I S O N schemata could h a v e b e e n treated in the same way, as in Kintsch a n d G r e e n o (1985). T h r e e hypotheses can b e formed about the P A R T - W H O L E schema, depend ing o n whether the first, second, or third of the sets formed is to b e considered the W H O L E set. (Note that the order in which the sets were formed in the w o r d problem, not their true temporal order, is at issue here.) T h u s , a proposition with the h e a d PPW, which is simply a m n e m o n i c for P A R T - P A R T - W H O L E , expresses the hypothesis that the p r o b l e m is a P A R T - W H O L E p r o b l e m with the third set as the WHOLE:ppw[role[sETl,PART],role[sET2,PART],role[sET3,WHOLE]]. Associated with this schema is the equation Ql + 02= 03, w h e r e 01 is the quantity of the r-th set, as well as procedures to solve this equation, d e p e n d i n g o n which of the quantities h a p p e n s to b e u n k n o w n .

E x a m p l e s a n d Issues Three examples will b e analyzed h e r e to show h o w the m o d e l understands, or fails to understand, as the case m a y b e , arithmetic w o r d p r o b l e m s . To see h o w these examples work, it is necessary to present at least the first o n e in sufficient detail. This p r o b l e m is intended simply as an illustration of the basic mechanisms of the m o d e l - n o t h i n g m u c h of interest h a p p e n s with respect to the arithmetic, and textually, the only thing of significance is that a simple inference is formed, which, however, is crucial for the understanding of the p r o b l e m . Two m o r e examples, which will n o t b e presented in as m u c h detail, will serve as illustrations of how the m o d e l can account for s o m e well-known facts about word-problem solving that alternative models (Briars & Larkin, 1984; Kintsch & G r e e n o , 1985) do not h a n d l e readily.

DISCOURSE STUDIES

264

Inferences

> -

Manolita tried to weed h e r father's garden. "You sure w e e d e d it," said Mr. M u n d o z a . "There were fourteen tulips in the g a r d e n a n d n o w there are only six." H o w m a n y tulips did she pull out b y mistake? This problem, modified from the "Thinking Stories" of Willoughby, Bereiter, Hilton, a n d Rubinstein (1981), requires for its solution the application of one of the L O C A T I O N strategies: T h e r e were so m a n y tulips in t h e garden, t h e n some were pulled out, a n d n o w so m a n y are left. A simple, knowledge-based infer ence becomes necessary: that the tulips that were pulled out are n o longer in the garden. T h e knowledge-activation m e c h a n i s m of the present m o d e l readily supplies this inference, a n d the p r o b l e m will b e solved successfully. T h e m o d e l processes this p r o b l e m in three cycles, which includes the first sentence, the statement b y Mr. M u n d o z a , a n d the question sentence. T h e first sentence simply sets u p a context a n d is n o t directly relevant to the arithmetic. I n Figure 10, the w a y the m o d e l understands this sentence is indicated, albeit in abbreviated form. T h e propositions constructed from the sentence itself are ( P I ) MANOLITA, ( P 2 ) G A R D E N , ( P 3 ) TRY[MANOLTTA,P4], ( P 4 ) WEED[MANOLITA,GARDEN], ( P 5 ) F A T H E R ' S [ G A R D E N ] . O n l y the first terms of these propositions are shown in Figure 10. Also shown in Figure 10 are the propositions that w e r e added to the text base through the process of associative-knowledge elaboration (they are m a r k e d with an asterisk and, o n c e again, abbreviated: * N A M E stands for ISA[MANOLITA,NAME], etc.). Because n o simulation of a general knowledge network is available, or even conceivable, the process of knowledge elaboration must be approximated empirically. A n informal procedure was adopted for this purpose: T h r e e persons were asked to provide free associations to phrases corresponding to P I through P 5 (as well as to corresponding phrases from the remaining two sentences of this word problem), a n d the responses generated b y at least two

*PULL-OTJT / *IN-GARDEN

/

*WORK *HELP FATHER'S /

WEED *DIFFICULT

/

^ *ENJOY

-J T R Y *CONFUSE

'

/

MANOLITA *FLOWER

*COLORFUL

•UNSUCCESSFUL *NAME

\

*VEGETABLE

*FEMALE

Figure 10: The elaborated text base for the first sentence of the Manolita problem (Each proposition is indicated by a single word. Text propositions are connected by solid lines, their associates bv broken lines. Associates are marked with an asterisk.)

j

KINTSCH


265

persons w e r e considered as the top associates of each proposition in the general knowledge n e t (up to a m a x i m u m of three associations p e r proposition). T h e text base s h o w n in Figure 10 serves as a basis for deriving a connectivity matrix, using the principles illustrated earlier in Tables 1 a n d 2: Text propositions are connected d e p e n d i n g o n their proximity in the text base, each text propos ition is connected to its associates b y a value of .5, a n d knowledge derived p r o p ositions are interconnected b y the same value if they share an argument, or b y -.5 if different word senses are involved (this does not occur in the present example). A n initial activation vector consisting of .2 s for the five propositions directly derived from the text, followed b y 13 zeros for t h e propositions generated from the knowledge net, was t h e n repeatedly u p d a t e d b y multiplying it with the con nectivity matrix until the activation values stabilized, as in the examples discussed previously. I n the present case, activation levels stabilize after 10 iterations. T h e resulting pattern of activation is shown in the first panel of Figure 11. W E E D [MANOLTTAJGARDEN], whose centrality in the text base is apparent in the graphical representation, has the highest activation value, a n d the other text-derived p r o p ositions also h a v e fairly high activation values. Knowledge-derived propositions are considerably less activated. T h e four most strongly activated propositions (PI through P4) are retained in the short-term buffer a n d enter the second pro cessing cycle. T h e second processing cycle is shown in Figure 12. T h e four propositions held over in the short-term m e m o r y buffer from Cycle 1 are joined b y 9 n e w text propositions a n d 11 associated propositions from t h e knowledge base. (Because of lack of space, the latter are indicated only b y asterisks.) T h e quantity propositions FOURTEEN[TULIP] a n d SIX[TULIP] generate four arithmetic hypotheses: that the 14 tulips that were in the g a r d e n in the past are, respectively, a PART or WHOLE set, a n d that the 6 tulips n o w in the garden are a PART or W H O L E set. W h a t the reader knows about weeding gardens provides the crucial information that discriminates a m o n g these hypotheses: T h e tulips before the garden was w e e d e d are the W H O L E set, a n d only a PART is left after the weeding. This knowledge is expressed in the connectivity matrix b y connecting PAST with WHOLE[14], a n d PRESENT with PART[6]. T h e last t h r e e propositions that enter t h e subnet are the superordinate arithmetic hypotheses P P W , P W P , a n d W P P . T h e y receive support from their corresponding first-order arithmetic hypotheses. Thus, whatever strength each arithmetic hypothesis gathers from the text is fed into the superordinate arithmetic schemata consistent with it. These schemata are mutually exclusive a n d inhibit each other with connection values of - l . Note that only at this final level is in hibition a m o n g arithmetic hypotheses used: T h e hypotheses that a particular êt of objects plays t h e role of W H O L E or PART set are also mutually exclusive, b u t they are not allowed to inhibit each other; they merely collect m o r e or less posi tive evidence, which they then transmit to the superordinate stage where a selection among alternatives is m a d e . The resulting connectivity matrix then becomes the multiplier of the activationstate vector for the 2 8 propositions participating in this second processing cycle, initially, these activation values are D o s i t i v e fnr thp t-»v+.«i«ir*.™»«i ™ —

DISCOURSE STUDIES

266

.13

A

WEED

O-..

p p

.12 GARDEN Q T R Y 0~.\

p

.11 M A N O L I T A O.

\

..O .10 FATHER'S O

O" / / /

.09 A

SAY O ' / ' V

.08 A

\ P A S T O (3 P R E S E N T OO CONTAIN

.07 A O MUNDOZA

.06

' * NOT-IN-GARDEN • P(?)

"'"••OOOTDUP /:'

.05

O PULLOUT

.04 O HOW MANY O BY MISTAKE

.03 A

.02

*m .01

w(14) p(6) p(14) w(6) i WPP

PW PW P P a-"— DCYCLE I

CYCLE I I

| •w(?) * -D - a-n :

CYCLE I I I

Figure 11: The result of the integration process for the three sentences in the Manoliu problem (Propositions are indicated by single words; inferences are marked by an asterisk; their arrangement in thefigureis approximate. The ordinate shows the activation values of each proposition after the process has stabilized. Propositions carried c from one processing cycle to the next are connected by arrows.)

T

KINTSCH

ROLE O F KNOWLEDGE I N DISCOURSE C O M P R E H E N S I O N

BUFFER

267

TEXTBASE

Figure 12: The elaborated text base for the second sentence of the Manolita problem Four propositions were carried over from the previous cycle in the short-term memory buffer. Solid lines connect text propositions, broken lines inferences; nonarithmetic inferences are indicated by asterisks only.) and zero otherwise, except for the propositions carried over in the buffer, which retain the activation values they reached in the last cycle. I n this case, the activ ation vector stabilizes already after seven operations. T h e results are shown in ihe second panel of Figure 11. (If the activation process is extended to twice the number of cycles, the activation values for the arithmetic hypotheses, measured to four decimal places, d o not change at all.) All text-derived propositions r e m a i n strongly activated, while n o n e of the textual inferences (e.g., M U N D O Z A I S a N A M E of a MALE, TULIPS are FLOWERS, R E D , a n d G R O W - I N - H O L L A N D ) reach a high level of activation. This is intuitively quite plausible. As far as the arithmetic is concerned, the problem is at this point understood correctiy a n d practically solved: WHOLE[14] is more strongly activated than its alternative, PART[14]. Similarly, PART[6] is stronger than WHOLE[6]. T h e correct hypothesis, W P P , is the most strongly activated of the three alternative superschemata. Note that the text propositions a n d inferences are, in general, m u c h m o r e strongly activated than the arithmetic hypotheses. Therefore, the activation values ot ihe latter must b e considered separately, relative to each other, rather than in relation to the text propositions w h e n it comes to selecting propositions to b e


268

maintained in the short-term m e m o r y buffer. This imbalance is required for the m o d e l to work. If the arithmetic hypotheses are weighted m o r e heavily, they draw the activation away from the text itself, a n d the system cannot stabilize: It will flip-flop between alternative, mutually contradictory arithmetic schemata. T h e arithmetic hypotheses h a v e to b e anchored in a stable text representation. For the third a n d final sentence, the short-term m e m o r y buffer needs to carry over b o t h text propositions to establish textual coherence a n d arithmetic hypotheses to take advantage of the understanding of the p r o b l e m that has b e e n achieved so far. It has b e e n assumed h e r e that the four strongest text propositions as well as the four strongest arithmetic hypotheses are carried over in the buffer, as shown in Figure 13. (There are, of course, other plausible alter natives.) T h e three text propositions generated o n the basis of this sentence bring with t h e m into the net six knowledge propositions, o n e of which is NOT [ C O N T A I N [ G A R D E N , TULIP]], which turns out to b e crucial for the solution of the problem. In addition, n e w hypotheses about the question set are formed, and the s c h e m a t a P P W a n d P W P , which w e r e lost after the s e c o n d cycle, are reconstructed. Because the child knows about weeding gardens, the tulips that were pulled out are identified as a part of those that w e r e in the g a r d e n in the

BY-MISTAKE

PRESENT

PAST

PULL-OUT NOT-INGARDEN

HOW-MANY

WHOLE (14)

PART

WPP

WHOLE ( ? )

PART(6)

P A R T ( ? ) '"'

(6)

O-

pp

-O w O -

--O

p W P

-o L BUFFER

TEXTBASE

Fienre 13: The elaborated textbase for the third sentence of the Manolita problem

1

KINTSCH


269

beginning. H e n c e , a connection that favors the PART hypothesis over the W H O L E hypothesis is formed b e t w e e n the inference N O T [ C O N T A I N [ G A R D E N , T U L I P ] ] a n d PART[?]. It completes the pattern that is the condition for the use of a LOCATION strategy: some tulips at o n e place in the past, t h e n s o m e n o t there, n o w some are left. T h e n e w net requires 4 3 operations to stabilize. T h e knowledge-based infer ence N O T [ C O N T A I N [ G A R D E N , T U L I P ] ] achieves an activation level above the range of the text propositions (Figure 11, third panel). T h e picture is completely clear as far as the arithmetic is concerned: All the correct hypotheses are strongly activated, a n d all incorrect alternatives h a v e low or zero activation values. T h e final steps in the solution of the p r o b l e m are procedural. F r o m infor mation associated with the W P P pattern the equation 14 = 6 + ? is generated, which is then used to obtain the correct answer. A lot of mountains h a d to b e moved to achieve a very simple result! T h e Manolita p r o b l e m was solved without p r o b l e m solving. T h e basic com prehension operations were sufficient; that is, it p r o d u c e d the inference that the pulled-out tulips are n o t in the garden, which was required for the application of the L O C A T I O N strategy. However, this is n o t always the case. I n m a n y , n o t necessarily difficult, problems, m o r e focused problem-solving operations are required because the random-inference generation process described earlier fails to generate the required inference. Consider the following "thinking problem": Mrs. Nosho was telling M a r k about the two huge aquariums she kept w h e n she was a little girl. "There were 30 fish in one a n d 40 fish in the other, so you can tell h o w m a n y fish I h a d . " H o w m a n y fish did Mrs. N o s h o have? In a simulation r u n of this p r o b l e m t h e m o d e l failed because it did not c o m e u p with the transitive inference H A V E [ X , Y ] & C O N T A I N [ Y , Z ] implies H A V E [ X , Z ] . At this point, the process needs to go into a problem-solving m o d e in which the infor mation in the text is elaborated in a m o r e focused m a n n e r than is possible with the automatic-comprehension mechanisms discussed here. Context Effects Problems e m b e d d e d into a familiar situational context are m u c h easier to solve than problems that must b e solved without this situational support (e.g., H u d s o n , 1983). T h u s , birds catching w o r m s present a concrete, understandable situation that makes it clear what is the whole a n d what are the parts, whereas abstract, ill-constrained p r o b l e m s do not. All depends o n whether the right arithmetic strategy is used; the situation is of n o help. In the worm-and-bird problem, the text provides a situational constraint for the interpretation of the p r o b l e m that has v e r y little to do with arithmetic per se. It is the knowledge about birds eating w o r m s that matters. T h e birds trying to catch the w o r m are understood as the W H O L E set, with the birds catching w o r m s as one PART, a n d the birds unable to get a w o r m as the other PART. This under standing was achieved n o t because a certain key phrase, like how many more,

270

DISCOURSE STUDIES

was parsed correctly b u t o n the basis of general world knowledge. If there are birds, some of w h o m catch a n d some of w h o m do n o t catch a w o r m , w h a t is the W H O L E set a n d what are the PARTS is given b y general world knowledge that is n o t specific to arithmetic. T h e arithmetic can hardly go w r o n g h e r e because the well-known situation guarantees the right interpretation of the p r o b l e m . It is this aspect that the present m o d e l deals with m o s t effectively. Context, however, does n o t always facilitate p r o b l e m solution, it m a y also interfere with it. Consider this typical school problem, with its highly impoverished context: Fred has four Chevies a n d three Fords, (a) H o w m a n y cars does h e have altogether? (b) H o w m a n y m o r e Chevies does h e h a v e than Fords? Context is n o help with this problem; it must b e solved o n the basis of specialized arithmetic strategies, o n the basis of the key words have altogether'for Question A a n d have more than for Question B. O f course, children are m u c h m o r e familiar with the former (e.g., Riley et al., 1983), b u t if the right strategies are available, b o t h problems will b e solved. In the model, too, the altogether in Question A will b e connected with the H O W - M A N Y / W H O L E hypothesis, a n d the have more than will b e connected with the H O W - M A N Y / P A R T hypothesis in Question B, and both questions will b e answered equally well. After the first sentence, PART a n d WHOLE hypotheses are established for b o t h the Chevies a n d the Fords, b u t there is noi m u c h to distinguish t h e m ; the superordinate schemata P P W , P W P , and WPP are only weakly activated a n d hardly differentiated. Question A, o n the other h a n d , correctly activates the P P W hypothesis, a n d Question B yields the WPP result. Thus, if the arithmetic knowledge is available, it makes very little difference which question follows the p r o b l e m statement. I n contrast, if the p r o b l e m is only slightly contextualized, the m o d e l can be biased in favor of one of the questions, a n d actually fails w h e n it gets the wrong one. Suppose, the foregoing p r o b l e m is changed to r e a d Fred has a nice collection of antique cars. Four of his cars are Chevies, a n d three are Fords. Collection, like some, is constructed as a quantity proposition, and hence PART a n d W H O L E hypotheses for a set of cars with unspecified quantity are established in the first processing cycle. T h e y are b o t h activated equally, however, at this point. This changes dramatically with the second sentence: T h e four Chevia a n d three Fords are b o t h identified as PART sets because of the phrase of his. In consequence, the m o d e l begins to favor the W P P hypotheses. W h e n it receives Question A , the W P P hypothesis is decisively strengthened, and the problem is solved correctly. O n the other hand, if it is given Question B, the model become* confused b e t w e e n the W P P a n d P W P hypotheses, which are both equally activated, a n d fails to solve the p r o b l e m . T h u s , w e h a v e h e r e a n example where the p r o b l e m context interferes with the solution of a problem. It biases the p r o b l e m in favor of o n e particular inter•*

*

' * --

«—

-'-

—™»J,-«.J

fV.o

TATII/II/J

nrnrim

KINTSCH

ROLE O F KNOWLEDGE I N DISCOURSE COMPREHENSION

271

fails. It is important, however, to analyze exactly w h y the m o d e l failed to answer Question B correcdy: After processing the second sentence, it was so strongly convinced that the four Chevies a n d three Fords were b o t h PART sets that it did n o t carry over the corresponding W H O L E set hypotheses a n d therefore h a d n o w a y of using the information in the have-more-than question in support of the C H E V I E S / WHOLE hypothesis. T h u s , rather special circumstances p r e v e n t e d the m o d e l from answering Question B. I n slighdy different circumstances, it could h a v e done so: (a) if the buffer were large enough, the C H E V Y / W H O L E hypothesis would not have b e e n lost, or (b) if the m o d e l h a d b e e n allowed to reread the p r o b lem statement. Question Specificity The final example illustrates some different aspects of word-problem solving; namely the complex role that redundant specifications of sets m a y have. O n the one hand, overspecifying a set can b e helpful because it provides m o r e than one way to refer to it. O n the other h a n d , r e d u n d a n t specifications increase the length of the text and thus the likelihood that some important piece of information is no longer in active m e m o r y w h e n it is required. I n the following problem, three versions of the question are possible: . J o e h a d a collection of n i n e marbles. H e started his collection with some beautiful red marbles. T h e n Lucy a d d e d six pink marbles to his collection as a present, (a) H o w m a n y beautiful r e d marbles did h e start his collection with? (b) H o w m a n y marbles did h e start his collection with? (c) H o w many beautiful red marbles did h e have? The first processing cycle results in undifferentiated hypotheses a b o u t the nine marbles. T h e set constructed in the second cycle, on the other h a n d , is clearly a FART set, as is the o n e constructed in the third cycle. Indeed, at the e n d of the third cycle, the m o d e l understands the p r o b l e m essentially correcdy, with the WPP schema g r e a d y exceeding alternative hypotheses in activation value. To understand what h a p p e n s next, it is necessary to k n o w which text propositions were maintained in the buffer at the e n d of the third cycle: O n l y propositions from the third sentence are carried over, while the propositions from the second sentence are n o longer held in active m e m o r y at this point. This has non-trivial consequences w h e n the question is asked. I n Versions A a n d B eveiything is all right, because the question itself identifies the question set as a PART set - starting a collection serves this function, just as it did in Sentence 2. Version C of the question, on the other hand, does n o t yield a correct solution. T h e question itself does n o t indicate the role of the question set, a n d there is n o information from the second sentence still available in active m e m o r y that would help to identify its role either: because there are already several strong PART hypotheses around, the model tends toward the hypothesis that the question set has the role of a W H O L E ; the PWP schema thus b e c o m e s m o r e activated than the correct W P P schema. However, this is far from an unequivocal prediction of failure for Version C

DISCOURSE STUDIES

272

intervening {pink marbles, as a present), the critical information from the second sentence could have been maintained in the buffer and used to solve the problem. O r even m o r e obviously, the problem solver could reread the problem or perform a reinstatement search (Kintsch & v a n Dijk, 1978; Miller & Kintsch, 1980) to activate the required information from long-term memory. Rather the prediction is that children, like the m o d e l , would h a v e m o r e trouble with Question C, a n d fail m o r e frequendy, t h a n with either A or B. Thus, the more specific the question the better. But h o w irrelevant or redundant material will affect the difficulty of a w o r d p r o b l e m is a m o r e c o m p l e x story. It m a y b e quite harmless, or m a y even facilitate p r o b l e m solving, if the question exploits a r e d u n d a n c y in the specification of a set. But it m a y b e a source of dif ficulty a n d even a cause of failure w h e n the question is asked in an unhelpful way. T h e present m o d e l has the flexibility to h a n d l e these c o m p l e x effects of context: M a n y small effects are allowed to add u p a n d pull the m o d e l one way or another. T h e "smart" m o d e l s of Kintsch a n d G r e e n o (1985) and Briars and Larkin (1984) h a v e n o r e a d y way to cope with these subde contextual d e m a n d s : Either the right strategy is used or n o t Discussion H o w people recall relevant knowledge w h e n they r e a d a text is reminiscent of another experimental p a r a d i g m that has b e e n studied extensively in psycho logical laboratories: h o w p e o p l e recall lists of words. A widely used explanation for the recall of word lists is based o n the generation-recognition principle. Some words are recalled directiy, perhaps from a short-term m e m o r y buffer, a n d these words are then used to generate other semantically or contextually related, plausible recall candidates. Words that have actually appeared in the to-be-learned list will b e recognized a m o n g these candidates and recalled, whereas intrusions will tend to b e rejected. Generation-recognition theories h a v e h a d their detractors, a n d in their most primitive form they are certainly inadequate to account for the m o r e complex p h e n o m e n a of list recall. H o w e v e r , sophisticated versions of this theory are widely accepted now. Almost every current m o d e l of list recall includes a generation/retrieval as well as a recognition/editing stage. T h e m o d e l of knowledge use in discourse comprehension p r o p o s e d here has two analogous stages: First, a propositional network must b e constructed, and then it must b e edited or integrated. T h e way the construction process is thought of h e r e is a straight extension of previous w o r k o n discourse processing. T h e words and phrases that m a k e u p a discourse are the r a w material from which a mental representation of the m e a n i n g of that discourse is constructed. This mental representation takes the form of a propositional text base. Text bases c o m b i n e two sources of information: the text itself a n d knowledge knowledge about language as well as knowledge about the world. To construct even a single proposition, an appropriate frame must b e retrieved from one's store of knowledge, and its slots must b e filled in the w a y indicated b y the t e x t T h e novel aspect of the present m o d e l is that the role of knowledge is greatiy e x p a n d e d in this process. Previously, o n e could think of the text base - to put it

KINTSCH


273

crudely - as a translation into "propositionaleese" of the sentences in the text. Now, the text base b e c o m e s a m u c h richer structure t h a n before. Not only does it contain the propositions direcdy derivable from the text, b u t also each of these propositions brings with it a n u m b e r of other propositions that are closely con nected to it in the general knowledge net. T h u s , propositions are constructed just as before (e.g., v a n Dijk & Kintsch, 1983); yet w h e r e previously a single proposition was formed, a whole cluster is generated now. Crucial in the present m o d e l is h o w this cluster of propositions is obtained: b y a context-free process of activation of the closest neighbors of the original text-derived proposition in the general knowledge net. O f course, such a process will inevitably activate a lot of material that is irrelevant for any given context and, indeed, inconsistent with it. However, the price that has to b e paid for p r o m iscuity is n o t v e r y high: T h e resulting text base is a connectionist net in which further spreading activation processes rapidly take care of inconsistencies a n d irrelevancies. W h a t is gained b y this d u m b a n d seemingly wasteful process of r a n d o m knowledge activation is flexibility a n d context sensitivity. T h e research o n knowledge activation in psychology, as well as the experience with artificial intelligence systems, suggests that it is very difficult to activate knowledge intelligendy. Prediction or expectation-based systems that use frames or scripts d o not adapt easily to new contexts; prestructured knowledge hardly ever is exacdy in the form that is needed. T h e construction-integration scheme p r o p o s e d h e r e m a y turn out to b e m o r e successful in this respect. T h e general framework sketched earlier could b e extended a n d elaborated in various ways as m o r e experience with it is obtained. It might prove necessary, for instance, to resort to greater formalization in the propositional notation used here. However, until it b e c o m e s quite clear what the gains of greater formaliza tion would be, a robust, easy-to-use system is to b e preferred, even at the cost of some imprecision. Perhaps m o r e important might b e elaborations of the knowledge-sampling mechanism. A s presented here, each text-derived proposition activates its own strongest associates. It might b e worthwhile to explore schemes w h e r e b y pairs or clusters of propositions activate their strongest joint associates. Similarly, other criteria for stabilizing a network might b e explored. For instance, networks might b e m a d e to maximize some statistic like h a r m o n y , as in Smolensky (1986). This m i g h t have considerable advantages. For instance, it is n o t always possible n o w to c o m p a r e different networks in terms of h o w fast they reach equilibrium, because the n u m b e r of cycles required depends strongly on the n u m b e r of n o d e s in the network. I n addition, at present there is n o really satisfactory w a y to tell h o w good a n equilibrium a process achieves. I n the w o r d arithmetic problems, all one can tell is whether the right hypothesis is more strongly activated than its competitors, but comparisons of the size of that difference across problems are problematic. Constructive processes other than the ones explored h e r e will n e e d to b e considered. For w o r d arithmetic problems, the most important constructions involved the arithmetic hypotheses. T h e construction of macropropositions could be neglected, mostly because the w o r d p r o b l e m s were short ones a n d their

DISCOURSE STUDIES

274

8

macrostructure played no role in the problem-solving process. For many other types of text, construction rules to form successive layers of abstractions and generalizations, as described by Turner et al. (1986), would be of primary interest The macro-structure of a text could thus be made an integral part of a text base rather than a separate component, as it is presentiy treated. Thus, there are a great many rules necessary to make the constructionintegration model work for proposition building, assigning references and prefer ences, bridging inferences, forming macrostructures, elaborating knowledge, and so on. Some of these construction rules are reasonably well worked out at this point, others are available within restricted domains, but many problems remain as yet unsolved. Thus, some of the same problems are encountered here as in conventional expectation-driven, top-down models of comprehension but with one difference: Weaker, more general rules can be used here because these rules need not be fine-tuned to an ever-changing context. Whatever rules are still needed ought to be easier to work out within the construction-integration framework. In van Dijk and Kintsch (1983), an imporant distinction was made between text bases and situation models. The former correspond to the propositional representation of a text, both at the level of the micro- and macrostructure. The latter correspond to a representation of the text that is integrated with other knowledge. Thus, in terms of the present model, the integrated text base - after irrelevant and inconsistent information has been deactivated and important knowledge elements have been absorbed - is a kind of situation model. The qualifying phrase "a kind of" is needed because text bases, integrated or not, are always propositional, whereas van Dijk and Kintsch specifically left open the possibility that situation models may be nonpropositional (e.g., Perrig & Kintsch, 1985). Situation models, under certain circumstances, may thus be like Johnson-Laird's (1983) mental models. 9

The theory of knowledge use in discourse comprehension has been presented here at two levels: first, it is presented in terms of a general computational mech anism, at the level of what Pylyshyn (1985) called the "cognitive virtual machine"; and second, as a particular model that specifies h o w this mechanism is used in word identification in discourse and in understanding and solving word problems. The fiinction of the model is primarily explanatory. Certain phenomena can n o w be interpreted within the framework of the model; for example, why a particular formulation of a word problem is especially hard or easy. Unlike less complex theories, however, there is n o direct link between explanation and prediction in the present case. Unqualified experimental predictions are hard to come by in a model as complex as the present one. At best, one might predict that a particular problem should be a difficult one, but that might mean several different things at the empirical level: that the solution fails, that a particular error occurs, that extra memory resources are required, that a reinstatement search will occur, that the problem must be read twice, and so forth. Even if we knew precisely what the "knowledge-use virtual machine" was like, our ability to make precise experimental predictions that are testable in conventional ways would still b e severely limited. That, however, is not to say that such theories

KINTSCH


275

are without empirical consequences. Although w e cannot predict particular events, predictions concerning classes of events are quite feasible (e.g., the differ ent ways people might have trouble with word problems). Furthermore, our new found understanding of why and h o w certain things happen can have important consequences for h o w certain texts are created in the first place or for instructional practices designed to help people with particular comprehension tasks.

Acknowledgements This research was supported by Grant MH 15872 from the National Institute of Mental Health. The work on word arithmetic problems was supported by Grant BNS 8741 from the National Science Foundation.

Notes 1. Conceivably, a purer connectionist model might be constructed. In the present model, an associative knowledge net is used to build a text-base net, which is then integrated. McClelland (1985) has put forth the idea of a connection information distributor, which is a subnetwork in which the units are not dedicated and connections are not hardwired. Instead, this subnetwork is programmable by inputs from the central network where the knowledge that controls processing in the subnetwork is stored. One could say that the production rules in the present model have the function of programming such a subnetwork. 2. Formally, concepts and propositions can be treated alike (e.g., Anderson, 1980). 3. This use of the term proposition differs from the standard one in logic. Furthermore, not all authors who use comparable semantic units in their analyses use the same term. For instance, Dik (1980) talked about predicates and terms combining to form predications. Wilensky (1986) used relationand aspectuals. In spite of this terminological disarray and the possibility of confusion with the meaning of proposition in logic, proposition appears to be the most widely accepted term and will be retained here. 4. The extreme informality of this notation is chosen for ease of exposition. Frequently, of course, a more precise formalism is needed. It is fairly straightforward to elaborate the present informal notation whenever that is the case. For example, in the computer simulation of word problem solving by Dellarosa (1986), the LOOPS language pro vides a ready-made type-token distinction. There seems to be no reason, however, to burden a general discussion like the present one with a cumbersome, formal notation when it is not needed. 5. As with proposition, this is a nonstandard use of the term's meaning. Meaning is used here as shorthand for the momentary, subject- and situation-specific activated semantic and experiential context of a concept. Clearly, this is not what many people have in mind when they speak about the meaning of a word - though it is a conception of meaning quite appropriate for a psychological processing model. 6. That integration occurs at the end of each processing cycle is proposed here merely as a simplifying assumption. Although there is clearly something going on at the end of sentences (e.g., Aaronson & Scarborough, 1977), integration does not need to wait for a sentence boundary (see the evidence for the "immediacy assumption"; Just & Carpenter, 1980; Sanford & Garrod, 1981). It would be quite possible to aoolv the

DISCOURSE STUDIES

276

relaxation procedure oudined here repeatedly in each cycle, as propositions are being constructed. This would allow for the disambiguation of word senses before the end of a cycle. Because inferences and macro-propositions are usually not available before the end of a processing cycle, end-of-cycle integration plays an especially important role. 7. Computer simulations of this model have been developed by Fletcher (1985) and Dellarosa (1986) and are available from the author. 8. Longer problems in which the macrostructure does play a role have been investigated by Dellarosa et al. (1988). Depending on whether a word problem establishes a theme of competition or cooperation between two protagonists, compare or combine problems will be solved most easily. 9. Unlike the representation of the text itself - the text base, which is always propositional - situation models may have a different representation format, although this possibility was not considered in the present article. Both text bases and situation models are mental models of one kind or another in the sense of Gentner and Stevens (1983), though not necessarily in the more restrictive sense ofJohnson-Laird (1983).

References Aaronson, D., & Scarborough, H. S. (1977). Performance theories for sentence coding: Some quantitative models./oarna/ of Verbal Learning and Verbal Behavior, 16,277-303. Anderson, J. R. (1980). Concepts, propositions, and schemata: What are the cognitive units? Nebraska Symposium on Motivation, 28, 121-162. Anderson.J. R. (1983). The architecture of cognition. Cambridge: Harvard University Press. Becker, C. A. (1976). Semantic context and word frequency effects in visual word recogni tion. JoumfiZ of Experimental Psychology: Human Perception and Performance, 2,556-566. Briars, D., & Larkin, J. H. (1984). An integrated model of skill in solving elementary word problems. Cognition and Instruction, 1,245-296. Broadbent, D. A. (1958). Perception and communication. New York: Pergamon Press. Bruner,J. S., & Postman, L. (1949). Perception, cognition, and behavior. Journal ofPer sonality, 18, 15-31. Dellarosa, O. (1986). A computer simulation of children's arithmetic word problem solving. Behavior Research Methods, Instruments, & Computers, 18,147-154. Dellarosa, D., Kintsch, W., Reusser, K, & Weimer, R. (in press). The role of understanding in solving word problems. Cognitive Psychology. Dik, S. C. (1980). Studies in Junctional grammar. London: Academic Press. Fischler, I., & Goodman, G. O. (1978). Latency of associative activation in memory.ToarnaZ of Experimental Psychology: Human Perception and Performance, 4,455-470. Fletcher, C. R. (1985). Understanding and solving arithmetic word problems: A computer simulation. Behavior Research Methods, Instruments, and Computers, 17,565-571. Forster, K. I. (1976). Accessing the mental lexicon. In R. J. Wales & E. C. T. Walker (Eds.), New approaches to language mechanisms (pp. 257-287). Amsterdam: NorthHolland. Frazier, L., & Rayner, K. (1982). Making and correcting errors during sentence compre hension: Eye movements in the analysis of structurally ambiguous sentences. Cognitive Psychology, 14, 178-210. Frederiksen, J. R. (1981). Understanding anaphora: Rules used by readers in assigning pronominal referents. Discourse Processes, 4,323-347.

KINTSCH


277

Gentner, D., & Stevens, A. L. (1983). Mental models. Hillsdale, NJ: Erlbaum. Haviland, S. E., & Clark, H. H. (1974). What's new? Acquiring new information as a process in comprehension. Journal of Verbal Learning and Verbal Behavior, 13,512-521. Hudson, T. (1983). Correspondences and numerical differences between disjoint sets. Child Development, 54, 84-90. Johnson-Laird, P. N. (1983). Mental models. Cambridge: Harvard University Press. Just, M. A., & Carpenter, P. A. (1980). A theory of reading: From eye fixation to compre hension. Psychological Review, 87, 329-354. Kintsch, W. (1974). The representation of meaning in memory. Hillsdale, NJ: Erlbaum. Kintsch, W. (1984). Strategiesfor solving word arithmetic problems: Extensions ofthe Kintsch & Greeno Model (Technical Report No. 133). Boulder: Institute of Cognitive Science, University of Colorado. Kintsch, W. (1985). Text processing: A psychological model. In T. A. van Dijk (Ed.), Handbook of discourse analysis (Vol. 2, pp. 231-243). London: Academic Press. Kintsch, W. (1987). Understanding word problems: Linguistic factors in problem solving. In M. Nagao (Ed.), Language and artificial intelligence (pp. 197-208). Amsterdam: North Holland. Kintsch, W. (in press). The representation of knowledge and the use of knowledge in discourse comprehension. In C. Graumann & R. Dietrich (Ed.), Language in the social context. Amsterdam: Elsevier. Kintsch, W., & van Dijk, T. A. (1978). Towards a model of text comprehension and production. Psychological Review, 85, 363-394. Kintsch, W., & GreenoJ. G. (1985). Understanding and solving word arithmetic problems. Psychological Review, 92,109-129. Kintsch, W., & Mross, E. F. (1985). Context effects in word identification. Journal of Memory and Language, 24, 336-349. Marslen-Wilson, W. D., & Welsh, A. (1978). Processing interactions and lexical access during word recognition in continuous speech. Cognition, 10, 29-63. McClelland, J. L. (1985). Putting knowledge in its place: A scheme for programming parallel processing structures on the fly. Cognitive Science, 9,113-146. McKoon, G., & Ratcliff, R. (1986). Inferences about predictable events. Journal of Experimental Psychology: Learning, Memory, and Cognition, 12, 82-91. Meyer, D. E., & Schvaneveldt, R. W. (1971). Facilitation in recognizing pairs of words: Evidence of a dependence between retrieval operations. Journal of Experimental Psychology, 90, 227-234. Miller, J. R., & Kintsch, W. (1980). Readability and recall of short prose passages: A theoretical analysis. Journal of Experimental Psychology: Human Learning and Memory, 6, 335-354. Morton, J. (1969). Interaction of information in word recognition. Psychological Review, 76,165-178. MortonJ. (1979). Word recognition. InJ. Moron &J. C. Marshall (Eds.), Psycholinguistics, Series 2: Structures and processes (pp. 107-156). London: Paul Elek. Mudersbach, K. (1982). Hol-Atomismus als Vereinheitlichung von Atomismus und Holismus. In Proceedings of the 7th International Wittgenstein Symposium (pp. 347-349). Wien: Hoelder-Pichler-Tempsky. Norris, D. (1986). Word recognition: Context effects without priming. Cognition, 22, 93-136. Perrig, W., & Kintsch, W. (1985). Prepositional and situational representations of text. Journal of Memory and Language, 24, 503-518. Pylyshyn, Z. W. (1985). Computation and cognition. Cambridge: MIT Press-

DISCOURSE STUDIES

278

Raaijmakers, J. G., & Shiffrin, R. M. (1981). A theory of probabilistic search of associative memory. Psychological Review, 88, 93-134. Riley, M. S., Greeno, J. G., & Heller, J. I. (1983). Development of children's problem solving ability in arithmetic. In H. P. Ginsburg (Ed.), The Development of mathematical thinking (pp. 153-196). New York: Academic Press. Rumelhart, D. E., & McClelland, J. L. (1986). Parallel distributed processing. Explorations in the microstructure of cognition. Vol. I: Foundations. Cambridge: MIT Press.

Sanford, A. J., & Garrod, S. C. (1981). Understanding written language. New York: Wiley. Schank, R. C. (1978). Predictive understanding. In R. N. Campbell & P. T. Smith (Eds.), Recent advances in the psychology of language - Formal and experimental approaches (pp.

91-101). New York: Plenum Press. Schank, R. C. (1982). Dynamic memory. New York: Cambridge University Press. Schank, R. C , & Abelson, R. P. (1977). Scripts, plans, goals, and understanding. Hillsdale, NJ: Erlbaum. Seidenberg, M. S., Tanenhaus, M. K , Leiman,J. M., & Bienkowsky, M. (1982). Automatic access of the meaning of ambiguous words in context: Some limitations of knowledgebased processing. Cognitive Psychology, 14, 489-537. Sharkey, N. E. (1986). A model of knowledge-based expectations in text comprehension. InJ. A. Galambos, R. P. Abelson, &J. B. Black (Eds.), Knowledge structures (pp. 49-70). Hillsdale, NJ: Erlbaum. Smolensky, P. (1986). Information processing in dynamical systems: Foundations of Harmony Theory. In D. E. Rumelhart &J. L. McClelland, Parallel distributed processing. Explorations

in the microstructure of cognition. Vol. 1: Foundations (pp. 194-281).

Cambridge: MIT Press. Swinney, D. A. (1979). Lexical access during sentence comprehension: (Re)Consideration of context effects. Journal of Verbal Learning and Verbal Behavior, 18, 523-534.

Till, R., Mross, E. F., & Kintsch, W. (in press). Time course of priming for associate and inference words in a discourse context. Memory & Cognition, 16, 283-298. Turner, A. A., McCutchen, D., & Kintsch, W. (1986). The generation of macropropositions during comprehension. Paper presented at the meeting of the Psychonomic Society, New Orleans. van Dijk, T. A. (1980). Macrostructures. Hillsdale, NJ: Erlbaum. van Dijk, T. A., & Kintsch, W. (1983). Strategies of discourse comprehension. New York: Academic Press. Waltz, D. L., & Pollack, J. B. (1985). Massively parallel parsing: A strongly interactive model of natural language interpretation. Cognitive Science, 9, 51-74. Wilensky, R. (1986). Some problems and proposals for knowledge representation (Tech. Rep.

No. UCB/CSD 86/294). Berkeley: Computer Science Division, University of California. Willoughby, S. S., Bereiter, C , Hilton, P., &RubinsteinJ. H. (1981). Healmath: Thinking stories book. Level 2. La Salle, IL: Open Court. Winograd, T. (1983). Language asa cognitiveprocess. Volume I: Syntax. Reading, MA: Addison-

Wesley. Zajonc, R. B. (1980). Feeling and thinking: Preferences need no inferences. American Psychologist, 35, 151-175.

41 Heteronormativity in Action: Reproducing the Heterosexual Nuclear Family in After-hours Medical Calls Celia Kitzinger

n e of the major achievements of the lesbian, gay, bisexual a n d transgender (LGBT) m o v e m e n t s of the last 30 years has b e e n to transform V . - ^ at least in m a n y quarters - "the p r o b l e m of homosexuality" into "the problem of heterosexism." So, for example, the focus of social scientific research has shifted from a focus o n L G B T people per se (e.g., assessing their m e n t a l health of their parenting capacities) to the multiple oppressions to which they are subjected - ranging from state-sanctioned execution, torture, a n d enforced psychiatric treatment (Amnesty International 2001), to institutional discrimination and h a t e crimes (Herek a n d Berrill 1991), to the m u n d a n e oppressions of everyday life, such as anti-gay jokes and the social gaffes m a d e b y well-meaning heterosexuals in talking to, or about, L G B T p e o p l e (Conley et al. 2002). J u s t as "homosexuality" is a social construction invented to diagnose, circum scribe, and control certain kinds of behavior (same-sex sexual relationships) treated as social problems (Kitzinger 1987), so too is "heterosexism" such a social con struction. T h e term represents a second attempt b y the L G B T m o v e m e n t (the first was " h o m o p h o b i a " ; see Kitzinger 1987) to label certain kinds of behavior (e.g., discrimination, prejudice, a n d violence against L G B T people) as social problems. Insofar as the social construction of "heterosexism" is successful, it is the behavior of heterosexual people in discriminating against L G B T people that is the legitimate target of social reform. As a social constructionist, one approach I might take to "heterosexism" as a topic (not the o n e pursued here) would b e to investigate h o w (i.e., through what claims-making actions, with what degree of success, and against w h a t contestations) the L G B T m o v e m e n t s h a v e struggled

Source: Social Problems vol. 52, no. 4, 2005, pp. 477-498.

280

DISCOURSE STUDIES

to p r o m o t e heterosexism as a social p r o b l e m . A n alternative approach - the o n e exemplified h e r e - is. to explore h o w everyday heterosexist reality is con structed: that is, h o w does it come about that people in ordinary interactions (not motivated b y heterosexist prejudice or discriminatory intent) commonly invoke a n d p r o d u c e a normative heterosexual world as a taken-for-granted reality? As a lesbian-feminist activist and scholar, I a m concerned about the full spec trum of heterosexist oppression - from the deportation of h o m o s e x u a l asylum seekers to countries in which they are tortured a n d imprisoned, to the denial of equal marriage rights to same-sex couples, to the minor indignities and exclusions of everyday life. T h e analysis presented h e r e aims to gain purchase o n the latter p h e n o m e n o n . It is motivated b y m y strong sense that while L G B T activists are campaigning against blatant oppression a n d overt discrimination, at the same time all a r o u n d us a heteronormative social fabric is unobtrusively rewoven, thread b y thread, persistency, without fuss or fanfare, without oppressive intent or conscious design. M y current research p r o g r a m uses conversation analysis to explore h o w that is d o n e a n d aims to understand what strong threads constitute and b i n d together a normative heterosexist culture that m a k e it so impervious to challenge a n d so slow to change. I n this study, m y aim is to m a k e visible (and thereby to enable us all to challenge) s o m e of the m u n d a n e quotidian actions that result in the routine achievement of a taken-for-granted world that socially excludes or marginalizes non-heterosexuals. T h e term "heteronormativity" is widely used in contemporary political, social, and critical theory to describe socio-legal (e.g., Phelan 2001), cultural (e.g., Lancaster 2003), organizational (e.g., Grace 1999), a n d interpersonal (e.g., Blasius 2000) practices that derive from a n d reinforce a set of taken-for-granted presump tions relating to sex and gender. T h e s e include the presumptions that there are only two sexes; that it is "normal" or "natural" for people of different sexes to be attracted to one another; that these attractions m a y b e publicly displayed a n d celebrated; that social institutions such as marriage a n d the family are appro priately organized around different-sex pairings; that same-sex couples are (if not "deviant") a "variation on" or a n "alternative to" the heterosexual couple. Hetero normativity refers, in s u m , to the m y r i a d ways in w h i c h heterosexuality is produced as a natural, unproblematic, taken-for-granted, ordinary p h e n o m e n o n . I n contrast with whatJ o h n Heritage (1984) calls the "normative determinism" (p. 16) of theorists such as Durkheim, Freud, and Parsons, the conversation analytic approach pursued h e r e (like ethnomethodology m o r e generally; Garfinkel 1967) does n o t d e p e n d u p o n an understanding of actors as the bearers of internalized, culturally transmitted norms that guide or determine their conduct. Heteronorm ativity is embodied in what people do rather than in their beliefs, values, ideologies, or faiths. Complicity with heteronormativity does not necessarily imply prejudiced attitudes or beliefs (e.g., as these are usually conceptualized b y h o m o p h o b i a scales; see Kitzinger 1987) or any deliberate intent to discriminate against LGBT people (Kitzinger forthcoming). Rather, heteronormativity - like other social n o r m s - is e m b o d i e d a n d displayed endogenously, in the details of conduct, a n d m a y b e studied empirically as such.

KTTZINGER

HETERONORMATIVITY IN ACTION

281

M y empirical investigation of heteronormativity builds on H a r v e y Sacks's (1995a) analysis of n o r m s , most especially in relation to m e m b e r s h i p categoriza tion as outlined in his classic lecture "The Baby Cried. T h e M o m m y Picked It Up." N o r m s , according to Sacks (1995a), provide rules of relevance for selecting categories ("viewers use n o r m s to provide the relevant m e m b e r s h i p categories in terms of which they formulate identifications of the doers of those activities for which the n o r m s are appropriate"; p . 260), and people "use n o r m s to provide some orderliness, a n d p r o p e r orderliness of the activities they observe" (p. 260). According to Sacks, n o r m s are used to "see a family" w h e n observers see a m a n beating u p a w o m a n a n d do nothing because "we thought you were married and it wasn't any of our business" (p. 91). Also according to Sacks, p e o p l e can "pass" as m e m b e r s of categories to which they do not in fact belong simply b y virtue of the n o r m that leads t h e m to b e seen that way, as w h e n "a w o m a n walks away from a supermarket with the b a b y carriage filled with a b a b y that's not h e r s " (p. 254). N o r m s might also occlude families as w h e n older parents observed caring for their child are taken to b e the child's grandparents (p. 226), or as in the following account from the m o t h e r of two-year old Nathaniel, describing what happened w h e n they w e r e referred to a n e w doctor: "After I h a d introduced her to Nathaniel a n d explained his symptoms she r e s p o n d e d b y looking at m e a n d asking, 'And w h o are y o u ? ' Since parents d o not n e e d to legitimate their right to discuss their o w n children's illnesses with a doctor, I inferred from this that she did not think I could b e his m o t h e r " (Liladhar 1999:240).Janine Liladhar (1999) understands the doctor's displayed inference that she is n o t the m o t h e r of h e r child with reference to race: she is white a n d h e r son is black. Observers apparendy did not see a family w h e n a white m o t h e r playing race-you-to-the-car with h e r 20-year-old black son was taken to b e pursuing a thief (Rosenblum and Travis 2000:169) or w h e n two lesbian parents were asked, "which o n e of y o u is the m o t h e r " (Dalton and Bielby 2000). I n such instances, other cultural n o r m s (about "age," "race," a n d the p r o p e r gender constitution of family members) overturn inferences that could or should b e m a d e normatively o n the basis of category-bound activities. T h r o u g h the displayed assumptions of the interactants in such situations, analysts can begin to understand the constitution of the norma tive concept of the family as, for example, "private," mono-racial, and bi-gendered. 1

T h e research reported h e r e develops the Sacksian notion of n o r m s with specific reference to the heterosexual nuclear family. It puts conduct in interaction at the center of theorizing about the social world, a n d shows h o w n o r m s a n d normativity can b e g r o u n d e d and elaborated in detailed empirical analysis of their deployment in action. As such, it is part of the ethnomethodological trad ition that aims to "treat the obvious as a p h e n o m e n o n " ( Z i m m e r m a n a n d Pollner 1970:80) and to explicate the seen-but-unnoticed tacit presuppositions (or "norms") of everyday life a n d contributes thereby to the "sociology of the un marked" (Brekhus 1998) a n understanding of the practices that produce ordinary mundane heterosexism. Insofar as conversation analysis has focused o n social problems - a n d certainly in relation to the social p r o b l e m s highlighted in the 1988 special issue of Social Problems o n "Language, Interaction, a n d Social P r o b l e m s " - these h a v e b e e n

282

DISCOURSE STUDIES

treated in the form of "trouble" as oriented to b y participants in interaction (Maynard 1988:325). However, from the point of view of m a n y social activists, and others concerned with social problems - indeed, including Sacks, the founder of conversation analysis himself, in his early lectures (Sacks 1995b: 175-87) - social p r o b l e m s can also b e p r o d u c e d , a n d reproduced, b y social actors w h o are not oriented to a n y trouble in their interactions. A social p r o b l e m exists only for us, as analysts eavesdropping o n their talk, w h o see in it the untroubled reproduction of a heterosexist (or racist or classist or otherwise oppressive) world. O n e important theoretical goal of this article, then, is to highlight the value of conversation analysis in analyzing interactions in which, b y and large (and with orderly exceptions), the taken-for-granted, ordinary world is r e p r o d u c e d a n d reconstituted with no trouble at all - because it is precisely the assumptions with which this taken-forgranted w o r l d is i m b u e d that cause " t r o u b l e " or " p r o b l e m s " (at best), and oppression a n d discrimination (at worst), for those w h o d o not conform to its cultural n o r m s .

Family as a C a t e g o r i z a t i o n D e v i c e This article focuses on person reference terms d r a w n from the category of the "family" as these are deployed b y interactants in after-hours calls to a doctor. Previous analyses h a v e shown that social participants readily p r o d u c e family terminology that displays their heterosexuality without b e i n g h e a r d thereby as doing anything out of the ordinary (Eatzinger 2005), whereas w h e n people make it apparent in the course of ordinary activities (such as m a k i n g a dentist appoint m e n t or taking out car insurance) that they h a v e same-sex partners there are r e c u r r e n t interactional p r o b l e m s (Land a n d Kitzinger 2005). T h e analysis presented h e r e is particularly g e r m a n e , then, to the sociological tradition of work o n the family - long-considered a social problem, especially in its "deviant" forms: for example, black families (e.g., Stack 1974), or working-class families (e.g., R u b i n 1976; c o m p a r e to Smith 1993). L G B T families, in particular, are at the forefront of c o n t e m p o r a r y debates about marriage a n d the family (e.g., Bernstein a n d R e i m a n n 2 0 0 1 ; Weston 1998). Family researchers h a v e b e e n preoccupied b y issues of definition. As James A. Holstein a n d j a b e r F. G u b r i u m (1999) argue, however, "family is not objectively m e a n i n g f u l . . . it is constantiy under construction, obtaining its defining character istics somewhere, somehow, in real time and place, through interpretive practice" (p. 4). This, they suggest, mandates an empirical focus o n "the myriad social pro cesses through which persons in the course of everyday life produce and organize 'family' as a meaningful designation for social relations" (p. 4). I n common with their innovative empirical research program analyzing discourse about the family across a range of organizational contexts (e.g., G u b r i u m and Holstein 19P; Holstein 1993), the work to b e reported h e r e explores h o w "family" is constructed in interaction. T h e production of the heteronormative family is understood as a n ongoing, locally m a n a g e d accomplishment, a n d the analysis focuses on one of the key practices through which family is constituted in these calls: the use ai person references a n d the inferences that attach to these. I

KTTZINGER

HETERONORMAT1VITY IN ACTION

283

Person references drawn from the category of "family" - such as "wife," "husband," "daughter," "brother-in-law," and so on - are common nonrecognitional person references (Sacks 1963, 1972a; Sacks and Schegloff 1979; Schegloff 1996). That is, they are terms routinely deployed as references to third parties in talk-in-interaction where the speaker is treating their coconversationalist as someone who does not know the person being referred to and/or would not recognize them by name. (Other examples of typical nonrecognitional person references include categorical or role references such as "a doctor," "an African-American," "a diabetic," "a woman.") Family reference terms are among the categories whose deployment Sacks began to explore most famously in "the baby cried, the mommy picked it up" discussion (Sacks 1972b, 1995a), but also at various points scattered throughout his work, includ ing an interrogation of the use of inferences attached to "wife," "sister," and "child" in a counseling call (Sacks 1995a: 116) and the (hypothetical) deployment of "heterosexual couple" inferences to disguise otherwise stigmatized identities (Sacks 1995a:593). The present research, then, builds on the early work by Sacks both on norms and on membership categories, and contributes to con versation analysis a more systematic and data-driven approach to the selection and deployment of non-recognitional person reference forms. It explores - in a single corpus of calls - the difference the selected person reference term makes to the sequence that subsequendy unfolds, due to the inferences that, thereby, can be seen to attach to the term selected. I begin with a description of the dataset and then consider the way in which the talk produces a heteronormative version of family.

Data

The data corpus comprises recordings of 59 after-hours calls made to a medical practice in a large town in the English Midlands. In 50 of the 59 calls, the caller is someone other than the patient; these calls are the focus of analysis here. According to the terms of reference of the British National Health Service, gen eral practitioners - physicians with no particular clinical specialization - have 24-hour responsibility for the medical care of their patients. This includes the condition that they should visit patients in their own homes if necessary, either because the patient cannot travel to the surgery or because the surgery is closed. The normative purpose of these after-hours calls is to request such a home visit Drew forthcoming). This data corpus cannot be treated as representative (as Drew forthcoming points out), either in terms of the kinds of after-hours calls made generally or in terms of the manner in which doctors manage such calls. Moreover, it is likely that definitions of "the family" vary - for the participants in these calls and for pi-ople generally - across their occasions of use. Analysis of these calls begins to chart the deployment of family reference terms in one such specific context. In all of these calls, the doctor is a "locum" (a stand-in for the doctor with whom the patient is registered), such that he knows neither the call 2

DISCOURSE STUDI]

284

patient, n o r does h e know anything about their family circumstances. Interactions between t h e patients a n d their usual physicians would b e conducted very dif ferently because most references to nuclear family m e m b e r s could b e d o n e by using recognitional forms of reference (i.e., their names). H o w e v e r , t h e range of recurrent occasions in m o d e r n societies in which recognitionals cannot b e used is vast, a n d the use of family reference terms is a potent resource in all such contexts. T h e data were transcribed according to t h e system devised b y Gail Jefferson (see Atkinson a n d Heritage 1984) a n d subsequentiy developed b y other convers ation analysts, designed to preserve those small but significant features of talk (cut-offs, silences, etc.) that interactants use in systematic a n d orderly ways, many of which turn out to b e essential to t h e analyses presented here. Non-recognitional person reference forms - in particular, person reference forms drawn from the category set of the family - were very c o m m o n across these calls, a n d I collected a n d analyzed t h e m in their sequential context. I examined, i n particular, initial person references - a n d t h e inferences associated with t h e m - for what they show about normative understandings of t h e family a n d t h e interactional uses these understandings are p u t to.

R e p r o d u c i n g t h e H e t e r o n o r m a t i v e Family Of the 5 0 calls i n the corpus where t h e patient is n o t t h e person calling, 33 are calls o n behalf of p e o p l e referred to as family m e m b e r s (e.g., spouses, children a n d grandchildren, siblings, a n d in-laws), a n d 17 are o n behalf of non-family m e m b e r s (e.g., friends, neighbors, a n d people for w h o m the callers have an in stitutional responsibility. Extracts 1 a n d 2 are examples of openings of calls on behalf of family m e m b e r s : (1) [DEC 1-2-12 BREATHLESS]

01 02 03 04 05

Doc: Clr: Doc: Clr:

Heklp, Hel.io, is tha' du- doctore sorry< how long's thatCld: .hh[h] Doc: [b]een going on for. Cld: Aw, two days. But it's just gettin' wp:rse, Doc: Has- Have you: seen: (.) your own doctor at alU

7

Similarly, in Extract 21 a m o t h e r calling about a child's sickness is asked, "who'i your doctor," t h e r e b y displaying the assumption that mother and child share the s a m e doctor: (21) [DEC 2-1-7 SICK III] 01 Clr: ... my daughter who's uh:: just coming fuh nineteen 02 months old, um: .thh yesterday she was feelin' very 03 sick. // ((74 lines omitted, of problem presentation, diagnostic questioning, and diagnosis)) 78 Doc: Who's your doctor, By contrast, non-family m e m b e r s a r e a s s u m e d n o t to share a doctor, with the question formulated as "Who'[s his doctor?" [DEC 2-1-16 ] or "Who p his doctor actually" [ D E C 1-2-17]. A caller o n behalf of an adult son is asked " W h o - who's her doc- your: t! his d o c t o r i " [ D E C 1-2-19]. Likewise, the caDo w h o identified herself as "only 'is g r a n d m a " (i.e., kin, but not close kin) is asked " 0 : k a y , an' who's your doctor normally, or his doctor." [DEC 2-1-9]: in d* original formulation the doctor displays an orientation to the caller as fwmh". in his repair, h e treats her as distant kin, unlikely to share with a grandson l family doctor. Callers - especially spouses - also m a k e available in their talk the infereocr that nuclear family m e m b e r s h a v e the same doctor and that others do not A h u s b a n d of a wife w h o has just fainted explains that "She's been to the doctor*! a n d our doctor at ((deleted)) gave h e r some painkillers" [DEC 1-1-13]; a husband of a couple n o t registered with the practice asks, "Did you want to know what our doctor was?" [ D E C 1-1-6]; and awife of a h u s b a n d with possible food poon* ing reports unproblematically the advice of a previous doctor to "contact am o w n doctor" [ D E C 2-1-15]. W h a t is interesting about this latter case is tha* turns out that h u s b a n d and wife d o not, in fact, h a v e the same doctor. The «change in which this statement is m a d e is r e p r o d u c e d in Extract 22 (an extensoi of Extract 5), described earlier as unique in the corpus for the deploy men! of i non-familial person reference (line 5) - typical of institutional calls - followed in Lesbian Kinship." Cultural Anthropology 10:41-63. Herek, Gregory M. and Kevin T. Berrill. 1991. Hate Crimes: Fighting Violence against Lanm and Gay Men. Thousand Oaks, CA: Sage. Heritage, John. 1984. Garflnkel and Ethnomethodology. Cambridge: Polity Press. Heritage, John and Jeffrey Robinson. Forthcoming. "Accounting for the Visit: Gtvoy Reasons for Seeking Medical Care." In Communication in Medical Care: lnttmam between Physicians and Patients, edited by John Heritage and Douglas W. Maynwi Cambridge: Cambridge University Press. Holstein, James A. 1993. Court-ordered Insanity: Interpretive Practice and favtlwun Commitment. Hawthorne, NY: Aldine de Gruyter. Holstein, James A. andJaber F. Gubrium. 1999. "What is Family? Further Thought! or a Social Constructionist Approach." Marriage and Family Review 28(3/4): 3-20. Ingoldsby, Bron B. 1995. "Marital Structure." Pp. 117-38 in Families in Mydtiadlm. Perspective, edited by Bron B. Ingoldsby and Suzanna Smith. New York: GuiMarc Press. Jacobson, David, Joan H. Liem, and Robert S. Weiss. 2001. "Parenting from Septnfa Households: A Cultural Perspective." Pp. 229-45 in New Directions inAnlhnptitffa. Kinship, edited by Linda Stone. Lanham, MA: Rowman & Utdefield Jefferson, Gail. 1984. "Laughter Organization in Talk about Troubles." Pp. 346-69 ir Structures of SocialAction: Studies in Conversation Analysis, edited byJ. Max tv ell Aikinra and John Heritage. Cambridge: Cambridge University Press. Kitzinger, Celia. 1987. The Social Construction ofLesbianism. London: Sage . 2000. "Doing Feminist Conversation Analysis." Feminism & Psychology 10: 163-SJ . 2005. "Speaking as a Heterosexual: (How) Does Sexuality Matter for Talkie interaction?" Research on Language and Social Interaction 38:221-65. . Forthcoming. "After Post-Cognitivism." Discourse Studies. Kitzinger, Celia and Sue Wilkinson. 2004a. "The Re-Branding of Marriage." Ftmam & Psychology 14:127-50. . 2004b. "Social Advocacy for Equal Marriage: The Politics of 'Rights' and dx Psychology of 'Mental Health.'" Analyses of Social Issues and Public PoUn •: 173-9* Lancaster, Roger N. 2003. The Trouble with Nature: Sex and Science in Popular CihtBerkeley: University of California Press. Land, Victoria and Celia Kitzinger. 2005. "Speaking as a Lesbian: Correcting L Heterosexist Presumption." Research on Language and Social Interaction AWA): 371-41 Lempert, Lora B. and Marjorie L. DeVault. 2000. "Guest Editors' Introduction: SpecIssue on Emergent and Reconfigured Forms of Family Life." Gender &Sutittj 14 1 Liladhar, Janine. 1999. "Racial Categories: Appearance and Othering." Femtttsm Psychology 9:239-42. Maynard, Douglas W. 1988. "Language, Interaction, and Social Problems." Social 35:311-34. 1

KITZINGER

HETERONORMATIVTTY INACTION

305

Miller. Leslie. 2003. "Claims-making from the Underside: Marginalization and Social Problems Analysis." Pp. 92-119 in Challenges and Choices: Constructionist.Perspectives e» SoaalProblems, edited by James A, Holstein and Gale Miller. New York: Aldine de Grower. Itthailavich. Kaga. Sarah Martin, and Stephen Lawton. 2001. "Lesbian and Gay Parents: Their Experiences of Children's Health Care in Australia.'' InternationalJournal of Stxwlity and Gender Studies 6:181-91. PW, Elizabeth. 2001. "Mundane Heterosexism: Understanding Incidents of the Everyday" Women's Studies International Forum 24:541-54. Relin. Shane. 2001. Sexual Strangers: Gays, Lesbians, and Dilemmas of Citizenship. Philadelphia, PA: Temple University Press, tardan. Daniel C. 2004. "Interaction Strategies of Lesbian, Gay and Bisexual Healthcare Practitioners in the Clinical Examination of Patients: Qualitative Study." British JHrd«/>unw/328:1227-29. Raenblum, Karen E. and Toni-Michelle Travis. 2000. The Meaning of Difference: American Ctnstrucnons of Race, Sex and Gender, Social Class, and Sexual Orientation. Boston: McGraw-Hill. Rabin, Lilian 8. 1976. Worlds ofPain: Life in the Working-class Family. New York: Basic Books. Sadb, Harvey. 1963. "Sociological Description." Berkeley Journal ofSociology 8:1-16. — . 1972a. "An Initial Investigation of the Usability of Conversational Data for Doing Sociology.' Pp. 31-74 in Studies in Social Interaction, edited by David N. Sudnow. N o York: Free Press. —. 1972b. "On the Analyzability of Stories by Children." Pp. 329-45 in Directions in Seaolingutstics: The Ethnography of Communication, edited by John J. Gumperz and Del Hymes. New York: Holt, Rinehart, and Winston. — 1984. "On Doing 'Being Ordinary.'" Pp. 413-29 in Structures ofSocial Action: Studies n Convmation Analysis, edited byj. Maxwell Atkinson andjohn Heritage. Cambridge: Cambridge University Press. — 1995a. Lectures in Conversation Part I. Oxford: Blackwell. —. 1995b. Lectures in Conversation Part II. Oxford: Blackwell. Sacks, Harvey and Emanuel A. Schegloff. 1979. "Two Preferences in the Organization of Reference to Persons in Conversation and Their Interaction." Pp. 15-21 in Everyday Lmguage: Studies inEthnomethodology, edited by George Psathas. New York: Irvington. -

42 ~ .1

Oral Genres of Humor: On the Dialectic of Genre Knowledge and Creative Authoring Helga Kotthoff

Introduction u m o r poses a challenge to genre theories w h o s e concept of genre imposes an excessively high standard of rigor and is too narrowly based J L - L o n a n interest in classifying ideals of pre-patterned discourse. T h e high degree of creativity, emergent construction a n d artistry typical of h u m o r call for a concept of genre which makes sense of modifications a n d transgressions in communicative processes, as shown b y L u c k m a n n (1986, 2002), Giinthner/ Knoblauch (1995) a n d Muntigl/Gruber (2005). S o m e forms of conversational h u m o r are generic, for example, standardized jokes. O t h e r forms exploit our knowledge of serious genres and activity types (thereby relying o n it). Despite, or better because of this, generic boundaries are often transgressed a n d disregarded in joking; n e w sub-types arise, such as absurd meta-jokes which violate the wellknown expectation of a p u n c h line or other features of the genre. There are genres such as teasing which m u c h m o r e than jokes reflect their emergent construction. We can define prototypes of teasing (and other h u m o r genres), b u t the genre is nevertheless realized in a great variety of ways. l i n g u i s t s h a v e also identified some basic generic features of stories, b u t there are as well m a n y sub-types of h u m o r o u s stories d e m a n d i n g a certain style of performance, a special framing arid significant deviation from their serious counterparts. T h e r e are also genres for which w e lack a folk t a x o n o m y b u t which are nevertheless quite generic in their dialogical structure a n d e m e r g e n t p e r f o r m a n c e (i.e. j o i n t fantasizing). I will take a close look at the above-mentioned genres a n d discuss h o w genre, contextualization a n d keying relate to each other. Conversational h u m o r often works with contextualization procedures such as code switching, social stylistics,

Source: Pragmatics (forthcoming, 2007).

KOTTHOFF

ORAL GENRES OF HUMOR

307

features of oral art, repetition, m a r k e d wording, prosody, interjections, laughter, mimicry, etc. that create a h u m o r o u s keying. T h e s e cues i n d e x the continually changing contextual presuppositions necessary for situated interpretations in oral discourse. M u c h of spontaneous humorous discourse involves conversational inferencing in the sense of G u m p e r z (1982). Finally, I will discuss what it m e a n s for a genre such as gossip to b e performed as a play with gossip. Is "playing gossip" still gossip? Intertextuality is another important concept that I draw u p o n .

1. G e n r e i n t h e S o c i o l o g y of K n o w l e d g e C o m m u n i c a t i v e processes following m o r e or less fixed patterns are called "genres." L u c k m a n n (2002: 163) describes typification a n d routinization pro cesses as going o n naturally in h u m a n action. It comes as n o surprise that in the course of history interlocutors consolidate certain structural expectations about how an activity might begin, develop a n d c o m e to an end, what role relations it allows, w h e r e it typically takes place, what its reputation is, a n d what functions it is able to realize. Like poetics, theology and literary criticism, classical rhetoric works with a genre concept. Because of the excessive d e m a n d s for rigor a n d clarity expected in scientific definitions of genre, "thus far in the illustrious history of the discipline, n o t so m u c h as o n e genre h a s b e e n completely defined" (Dundes, cited following Swales 1990, 34). Genre theorists h a v e mainly b e e n preoccupied with written texts, whereas the work of Bakhtin (1986/1994) and Voloshinov (1929/1975) p r o m p t e d a 'communicative turn' in genre theory, as discussed b y G u n t h n e r / K n o b l a u c h (1995). T h e two authors opposed a static concept of genre such as the ones c o m m o n in folklore studies a n d literary critic ism. In linguistic anthropology (Hanks 1995, Foley 1997), a n d likewise in the ethnography of communication, there is a tendency to n o longer consider genres as static, monological products, b u t rather to adopt a performance-centered ap proach a n d to study genres in the process of their interactive production within a conversational a n d socio-cultural context. This includes showing h o w and w h y speakers violate conventions a n d i n d e x originality. Conversation Analysis has also h a d an important influence o n genre research. Sacks (1974, 1978), for ex ample, analyzed joke-telling in natural settings. H e showed h o w joke-telling suspends the n o r m a l turn-taking procedure. For the length of the j o k e , the teller reserves the right to speak. This is w h y jokes are usually introduced before they are told. T h e a n n o u n c e d intention to tell a j o k e must first b e ratified b y the prospective listeners. Joke-telling is temporally a n d sequentially organized. T h e story unfolds in a simple series of events located in time (Sacks 1974,1978). T h e sequential structure of the j o k e relies on a series of implausibilities. I n order for a j o k e to b e received as such, it is necessary from the very beginning to secure an appropriate reception by the listeners. T h e conflation of temporal a n d sequential order allows the sequence of events to appear c o h e r e n t

308

DISCOURSE STUDIES

Earlier studies h a v e suggested that disbelief is suspended for the length of the j o k e . Sacks, however, emphasizes that the implausabilities of jokes must b e h a n d l e d systematically. J o k e s are not invalidated b y implausibility, b u t rather implausibilities help to secure the appropriate reception through the canonical order of time (1978). H e writes that in receiving a story, listeners should believe the events b e i n g told; if necessary, they are expected to suspend 'disbelief. I n telling a j o k e , the teller concentrates o n constructing the p u n c h line so that listeners can figure it out (get it) as easily a n d quickly as possible. Recipients s h o u l d u n d e r s t a n d a j o k e directly, w i t h o u t r e c e i v i n g h i n t s or a d d i t i o n a l information, a n d laughter is the preferred a n d desired reaction. But Sacks also showed that joke-telling can b e used for context-specific purposes. Speakers can, as is the case in Sacks' (1978) example, use jokes to show their knowledge of sexual behavior. A n interlocking of functions a n d goals arises. Of course, one function of j o k e s is to amuse people. Individual a n d context-specific functions can also b e added. Genres m a y b e reframed strategically in various ways. As G u n t h n e r / K n o b l a u c h (1995: 7) explain, reframing can only succeed if there are pre-fixed communicative patterns. G u n t h n e r / K n o b l a u c h h o l d that pre-patterning is located o n three different structural levels: the level of internal structure, the situative level a n d the level of external structure. M a n y features of the internal structure of jokes can b e identified. At the situative level, it remains clear that joke-telling is typical of informal settings. It m a y b e used to deformalize a c o n t e x t T h e external struc ture is quite loose because w e seldom find situations that m a k e joking obligatory. O n e such occasion is the G e r m a n "Buttenrede," a speech delivered during the Carnival season. A n d there are situations w h e n joking is forbidden, for example at funerals. T h e r e is an ideology underlying standardized joking as there is for every genre. I n Germany, a n d p e r h a p s throughout Western E u r o p e , m e n were in the past m o r e likely to tell jokes than w o m e n . T h e r e was a critical metadiscourse about jokes. M a n y jokes were, for example, r e g a r d e d b y the w o m e n ' s m o v e m e n t as carrying sexist messages. Most jokes took place in a m a l e world; w o m e n w e r e often the b u t t of j o k e s (Legman 1970, Kotthoff 2006a). T h e n the w o m e n ' s m o v e m e n t b e g a n to p r o d u c e n u m e r o u s jokes aimed at m e n . In this way, the genre gained a n e w status in society's communicative household. B r i g g s / B a u m a n (1992: 147), applying Bakhtin's c o n c e p t of intertextuality, describe an "intertextual relationship" as a linkage of texts that are "ordered, unified, a n d b o u n d e d , o n the o n e h a n d , a n d fragmented, heterogeneous, and open-ended, o n the other." G u n t h n e r / K n o b l a u c h (1995:21) sum u p approaches that underline the interrelationship b e t w e e n generic speech practices a n d social structures, values a n d ideologies. W i t h L u c k m a n n they see genres as part of a cultural system, as an important link between language and culture. Speakers are always o p e n to modifying typified forms of communicative behavior. I see this flexibility as the advantage of discussing genres within the context of the sociology of knowledge. Knowledge of typified genre realization frees speakers to inscribe n e w social meanings into a genre. Identical realignments h o l d true for everyday communication, as h a s b e e n discussed i n connection with the evolution of the creative arts:

KOTTHOFF


309

H o w a c o m p e t e n t reader approaches a work of literature, his attitude a n d expectations, d e p e n d importantly u p o n the genre h e sees it as exemphfying. A w o r k that rebels against genre-conventions equally relies o n the reader's recognition of the conventions being rejected. Aesthetically relevant features of a work m a y stand out only if its reader has a back g r o u n d awareness of the historical d e v e l o p m e n t of the genre, or of the style, that the work is transforming in its distinctive w a y a n d p e r h a p s without direct allusions within the text itself. T h e work d e m a n d s to be seen against the foil of the whole tradition from which it stems, a n d which it modifies b y its v e r y existence. ( H e p b u r n , 1983: 496, cited following Swales 1990: 37). It is also the case that h u m o r is p r o d u c e d a n d perceived in relation to t h e n o r m constituted b y codification. T h e fact that communicative activities violate t h e norms of their genres does not m e a n that those genres necessarily disappear. I will look at jokes, teasing activities, h u m o r o u s stories, joint fantasizing, h u m o r o u s gossip a n d h u m o r o u s counseling in order to discover the creative potentials that d e p e n d o n genre knowledge.

V

2. B e y o n d t h e Standards of Standardized J o k e s

The genre of the "joke" is familiar to everyone in our culture, and this can b e relied o n . I h a v e already s u m m a r i z e d Sacks' j o k e analysis. H e writes that the joke, constructed as a test of comprehension, always makes special inter pretative d e m a n d s o n reception. T h e d e m a n d s for plausibility a n d coherence are different from those in serious discourse. Freud referred to the high 'density' of jokes already in 1905. Sacks (1978) stresses that there are n o divergences from t h e central focus. 'Embellishment' is typical of stories, b u t n o t of jokes. Everything that does n o t direct attention to the p u n c h line should b e eliminated in jokes. But there are i n d e e d aesthetic strategies that i m p r o v e a j o k e . T h e n e x t j o k e was told in the U S state of M i n n e s o t a during a dinner shared b y several friends. David, W e n d y and Vivian are Americans, R o l a n d is German. Datum l

l

(David (D), Vivian (V), Wendy (W), Roland (R)) 1 2 3 4 5 6 7 8 9 10

D: W: D: V: W: V: D:

...which reminds me of aJOke i wanna tell you. oh yeah, hehe uhm dAve it's time to GO. oh NO:::: hehe hehehe uh this uh THIS guy came over from EUrope in about the Middle of eighteenslxty you know, around that PEriod, middle eighteenhundreds Am™,,, crmre-.i . "

DISCOURSE STUDIES

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

W: W: D:

W: D:

W: D:

and the REAson he came Over of course is to make • his fOrtune in this new LAND. he heard, the further wEst you WENT, the bEtter chance you had at making a FORtune. remEmber there was a GUY [that said, [go WEST young mAn. go WEST young mAn. go WEST. yeah. (? ?) i think his nAme was. Anyway he comes over and he he lAnds in new YORK. well All the money that he had OBviously was spent for passage. sEE' yeah, ((glass is tipped over)) so, Anyway he uh he goes and he hears that there=s a WAggon train that's being set Up (-) in new YORK, and he wants to gEt on this WAggon train. but (-) he has no MOney. so he goes to the WAggon master and he says, uh sOmething about, (-) uh i would like to go as far WEST as you people are gOing. and he said, wEll, we're going ALL the way over to the Oregon territory. wEll he says, i'd like to go WITH, but i don't have any MOney. he says, but i'd be glAd to do ANY kind of wOrk that yOU would like me to DO' wEll, he said uh, how are you with a Rifle. (-) and he says i'm uh really vEry GOOD with a with a gun. see. so they said wEll, we're gonna set you on the last WAggon of the waggon train. you're gonna sit on the BACK and watch out for Indians. okay? so they take Off from new YORK. well they go through you know ohlo, (-) and pennsylvAnia, and Illinois, and the whole wOrks, and they go to minnesOta and they get into the dakOtas before they ever see an INdian. hehehehe so and they're in the dakOtas and they are trAveling, (-) you know, this guy has got his rifle, he hollers to the driver, and he says, HEY DRIVER, he says, (-) I SEE AN INDIAN. (-) and the driver says, HOW BIG IS HE: he says, well, HE'S ABOUT THAT BIG. ((indicates a very small distance with his lingers)) well, he says, he's TOO far away to shOOt. he says. jUst WATCH=m. so they go for miles and that and he says uh DRIVER, THAT INDIAN'S STILL FOLLOWING US: HOW BIG IS HE NOW. OH HE'S THAT BIG. ((indicates a greater distance)) he says, - TOO small to shOOt. they're TOO far awAy. so: Anyway they GO and that and he says uh HEY that Indian's still - following us and he's getting a little CLOser.

KOTTHOFF

59 60 61 W: 62 D: 63 64 65 66 67 68 69 70 71 72 a: 73 W: 74 V: 75 D: 76 m:


311

how big's he NOW? wEU about THAT big. i , ((indicates a greater distance)) ah, he says, he's still TOO far awAy. hehe so they keep on gOing and that and finally, he says HEY that Indian's getting closer. and the driver says, HOW CLOSE IS HE NOW? WELL, he says, HE'S ABOUT THAT BIG. ((indicates a greater distance)) he says, SHOO::T=m. (-) he sis, i CAN't shOOfc=m. so why nOt. he sis he's a frIEnd of mine. (- -) hehe, he says, a frIEnd of yours? how the heck do you figure he's a frIEnd of yours. he sis, hey, I've known him since he was THAT big. ((indicates a small distance)) hahahahahahahahahahahahaha thAt's good, hehehehehehehehehehe hehehehehe okay HARD to find a good, clean jOke. hehehehehehe ((Baby cries))

Generic features go b e y o n d those described b y Sacks. T h e j o k e prefers implicit person characterization for which direct quotations are very important. A n essential question for j o k e performance always remains whether the typification process is staged in such a way that it could b e shared b y listeners. 2.7. Comical Effects of Social Typifications and Stylizations With T a n n e n (1989), C o u p e r - K u h l e n (1999) and Giinthner (1999), I regard reported dialogue as a play with double voicing in the sense of Michail Bakhtin. The persons whose speech is delivered are stylized a n d evoked (Gunthner 1999, Kotthoff 1998). These implicit typifications of the dramatis personae are easily identifiable b y the listeners because they are based o n shared knowledge about typical speech styles, which is confirmed in this m a n n e r . I n a well presented j o k e , scenic narration occurs. O n o m a t o p o e t i c callingsout, expressive evaluations, gesticulations, mimicry, interjections, typical oral and structural discourse markers a n d so o n are employed as stylistic procedures, just as with funny stories. 63 64 65 66

HEY that Indian's getting closer. and the driver says, HOW CLOSE IS HE NOW? WELL, he says, HE'S ABOUT THAT BIG. ((indicates a greater distance)) he says, SHOO: :T=m.

Direct quotes are i m p e r a t i v e . E l e m e n t s of s p o k e n discourse, interjections such as "hey" and "well," call prosody a n d shortenings ( S H O O : : T = m ) can b e reproduced most effectively in a quote. T h e r e are good a n d b a d realizations of

312

DISCOURSE STUDIES

a genre. I n a b a d realization (which would n o t contain the m e n t i o n e d strategies) the genre remains valid. A good j o k e performer not only stylizes direct quotes, b u t also integrates other special effects, for example, s/he manipulates the t e m p o of the j o k e , such as h e r e the iconization of great distance with long lists of the territories traversed. 40 41 42 43

well they go through you know ohlo, (-) and pennsylvAnia, and Illinois, and the whole wOrks, and they go to minnesOta and they get into the dakOtas before they ever see an INdian.

Even the reproduction of a standardized j o k e is m u c h m o r e than simple reproduc tion. T h e specific speaker-listener constellation influences the performance of the joke. Beyond the goal to amuse the public, jokes can h a v e m o r e specific goals, such as to introduce A m e r i c a n folklore to a G e r m a n guest as is the case in the example. 2.2. Meta-Jokes For a theory of genre, it is important that interlocutors play with the features of the genre. T h e r e are question-and-answer jokes, such as elephant jokes: • H o w d o y o u fit four elephants into a V W ? • Two in the front, two in the back. T h e p u n c h line is based o n a demonstration of normality. Most cars seat two persons in the front and two in the back. T h a t elephants are too large to fit into a V W is suppressed, a n d thus the question is n o t really answered b u t it does not present an o d d solution either. W h a t Attardo (1994: 285) discusses as an example of a joke, which fails to deliver the expected p u n c h line a n d b e c o m e s funny precisely because of the failure to d o so, is in fact a meta-joke b a s e d o n the failure to fulfill the normal genre expectation: "Have you heard the latest?" "No? Well, neither have I." T h e r e are typical s u m m o n s which o p e n u p a j o k e telling sequence, such as "have you h e a r d the latest?" or " D o you k n o w the o n e about X ? " o n e can play with.

3 . Stability a n d Variability i n Teasing Teasing is a genre with e v e n greater variation in situative performance. We start from the c o m m o n d e n o m i n a t o r that teasing is a personally addressed jocular remark with a bite, often performed in front of a public. T h e h u m o r o u s quality is m a r k e d , for example, b v the inadequate wording of attributions.

KOTTHOFF


313

In the next episode, a playful framing is initiated in the first line b y the inap propriate combination of opulent and social life. This steers the reception towards irony since the utterance is stylistically marked. Datum 2

2

(David (D), Ernst (E), Inge (I), Johannes (J), Katharina (K), Maria (M), Rudolph (R), several persons at once (m)) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

M: du hasch grad son opulEntes [soziALleben. R: [(? ?) D: totAL. totAL was los grad, well ich namlich initiatlv geworden bin Qetzt. M: [hahahahahaha K: [habichscho(h)nerZA(h)HLT.haha[hahaha S: [haha= s: hahahahahaha [hahahahaha E: [WAS sagt er, er freut sich schon auf wEIhnachten und silvEster. a: hahahahahahahahaha [hahahahahaha E: [mUnkelt man. mUnkelt man. D: ich hab angeregt entWEder. Oder, hab ich angeregt. s: hehehehehehehehehe K: wenn nichts 10s sei, weihnachten und silvEster, dann wiirde er (-) verREIsen. hat er gesagt. [dann FLIEGT er D: [mhm dann FLIEG ich. E: in die karlbik. karlbik. hehehehe [hehehe s: [ha [hahahahahahaha M: [hehehehehehehe

Translation: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

M: you are leading such an opulent [social life of late. R: [(? ?) D: a lot. a lot has been going on lately, because i have taken the initiative [now. M: [hahahahahaha K: [i have just told about that, haha [haha [haha= s: hahahahahaha [hahahaha E: [What's he saying, he is already looking forward to Christmas and new years. a: hahahahahahahahahahahahaha [hahahahahaha E: [it is rumored, it is rumored. D: i have suggested either, or. i have suggested. s: heheheheheheheheheh K: if nothing were happening, christmas and new year's eve, then he would (-) take a trip, he said, [then he would fly D: [uhm then i fly. E: to the Caribbean. Caribbean, hehehehe [hehehe s: [ha [hahahahaha M: [hehehehehe

DISCOURSE STUDIES

T h e dinner takes place at Katharina and David's h o m e . Maria focuses on David's social life. She employs a n elevated a n d inappropriate formulation (opulent social life), thereby creating a playful-ironic modality. Everybody knows that David prefers a quiet lifestyle. Recendy, however, h e has taken part in two social events: dinners at their h o m e at Christmas a n d o n N e w Year's Eve. David likewise responds ironically to Maria's remark. A b o v e all the formu lation taken the initiative is quoted from M a r i a a n d his wife Katharina, w h o immediately reacts affirmatively to this a n d laughs. M a n y p e o p l e present know that the view that David normally does not take the initiativeis not his own. David's self-irony thus draws its potential from M a r i a a n d Katharina, w h o sometimes use such psychological jargon. T h e others also u n d e r s t a n d the conflict-laden point of David's reclusive social life a n d h o w it is discussed. I n lines 6, 7, and 8 several persons laugh. David not only processes Maria's irony, but h e immediately counters it. We must assume that shared knowledge of personal habits a n d speaking styles and a high degree of familiarity m a k e it possible to r e s p o n d very quickly and creatively to irony. I n lines 9 and 10, Ernst alludes to n e x t Christmas a n d New Years Eve, which further amplifies the topic a n d the teasing of David. T h e background is that Katharina h a d invited n u m e r o u s guests this year n o t only for Christmas, b u t also for N e w Year's Eve, a m o n g others those present, and that this was definitely too m u c h for David. H e has resigned himself good-naturedly to his fate. Everyone laughs at the teasing jest that h e is looking forward to the n e x t Christmas a n d N e w Year's Eve, which also implies that there will again be n u m e r o u s invitations to social events. Ernst expands o n the irony b y himself starting to tease David. A contrasting perspectivation is still at stake. H e speaks about David in the third person, which is typical of teasing (Straehle 1993, Giinthner 1996). Maria's ironic implication that David considers his present social life to b e "opulent" is n o w teasingly in tensified. David is portrayed as wanting nothing m o r e t h a n to h a v e m o r e parties. I n line 12 Ernst refers to a rumor h e pretends to h a v e h e a r d , thereby suggesting the fictitious quality of his remarks. I n line 13 David seriously reports w h a t his stated preference is (he would like to h a v e a party o n either Christmas or N e w Year's Eve in the c o m i n g year). Everyone laughs again. D r e w (1987) has shown that teased persons initially react seriously to a teasing attack. David seriously explains h o w h e m a d e it clear to his wife Katharina that in the future h e would prefer n o t to h o l d big parties at their h o m e o n Christmas a n d N e w Year's Eve. This reveals his real m o o d . But the teasing continues. Starting at line 13, Katharina links David's distaste for a n opulent social life with his disinclination to travel. D a v i d really does not enjoy trips a b r o a d a n d seldom takes them, a n d the threat to take o n e would be the last thing we would expect from him. Everyone present shares this knowledge. David starts to take part in the teasing himself (15). H e confirms the views attributed to him. This again is a reaction to the literal meaning. Ernst augments this once m o r e b y referring to the Caribbean. D a v i d himself h a d recently teased h i m because of his flight to the Caribbean. For environmental reasons, David was critical of traveling long distances o n vacation trips. Again, the participants

KOTTHOFF


315

laugh. David's "leg is pulled", b u t h e shows the ability to laugh at himself. Teasing can w o r k with irony, as is the case h e r e , a n d it always works additionally with exaggeration. "Tangential address" (Gunthner 1996) is typical; it underlines t h e performance character of the teasing. Ironic activities are being carried out h e r e teasingly, which everyone present seemingly finds amusing. People communicate knowledge of o n e another in this way a n d t h e r e b y affirm their identity as part of the in-group. Friendly irony allows the in-group to deal playfully with social differences, which thereby receive acceptance. T h e participants leave the d o m a i n of official face politics and play fully create a high level of intimacy (Kotthoff 1996). Teasing in this case indirectly communicates a social difference to David. H i s friends playfully convey to h i m a construction of h o w h e might see himself in regard to the topic - a n d also m a k e it clear that they k n o w his real perspective. Friendly irony combines social disson ance u n d consonance (in the sense of Radcliffe-Brown 1940/1965), individuation and solidarity. David's self-irony shows that h e does not feel insulted b y the teasing. I n d a t u m 2 t h e social function of t h e teasing episode can b e described as negotiating social n o r m s a n d accepting differences in r e g a r d to t h e m . T h e r e are other forms of teasing, a n d the knowledge of teasing can b e ex ploited. T h e listener can define a simple critique as teasing a n d t h e r e b y invite t h e emergent construction of a teasing episode. Datum 3 (from Drew 1987) (Gerald has a new Mustang Sport wagon) Gerald: Martha: Gerald: Lee: Gerald:

Hi, how are you. Well, you're late as usual. eheh eheh eheh eheh What's the matter, couldn't you get your car started? hehe That's right. I had to get it pushed, eheh eheh

Gerald's laughter in line 3 defines t h e criticism as teasing. We get a n impression here of the recipient's p o w e r to negotiate the m e a n i n g of a speech activity. Lee ratifies Gerald's definition of the situation through teasing. D r e w does n o t discuss that this is e m e r g e n t teasing invited b y the activities of the recipient. I see d a t u m 3 as a n e x a m p l e of a recipient's refraining of a critique. M a r t h a ' s remark is quite serious. But Gerald refuses to offer a serious reception of h e r complaint (late as usual). T h e occasion of the teasing is a critical incident here. As D r e w claims, there is an evident contrast h e r e b e t w e e n the n e w car a n d Lee's comment that h e could not get it started. T h e critical c o m p o n e n t of datum 3 is more evident t h a n in d a t u m 2. S o m e teasing e p i s o d e s a r e totally fictional. Schmitt (1992) describes a group that meets everyday at a newsstand to chat a n d h a v e a drink. This g r o u p likes to tease a n older customer (Mtiller) for c o m i n g just to see Iris, a 22-year old student. This teasing is a provocation without a n y underlying real event or critical attitude. T h e g r o u p likes to see Miiller's reactions. I n a playful way, the group invents special situative identities for t h e steady customers. T h e owner of the newsstand, G e r h a r d , for e x a m p l e , p r e t e n d s to u n m a s k Miiller as a serrpf

DISCOURSE STUDIES

316

admirer of Iris. Midler is spoken about in the third person, which is typical of teasing activities. Miiller often gets into the act, a n d Iris likewise reacts with laughter (1992: 92ff.). T h e teasing episodes take place in front of a public. Often they thematize aspects of social relations. Customer Miiller counters the newsstand's owner's teasing b y pointing out that Gerhard's shoes could use polishing. Although the teasing is carried out in a playful m o d e , relevant cultural values are at stake, such as age differences in love affairs or standards of cleanliness. Already, w e h a v e seen three m o m e n t s of variation in teasing: • Playful provocation based o n behavioral differences within a social group • Teasing as a redefinition of criticism • Fictional teasing S o m e anthropological hnguists h a v e analyzed the teasing of children, which shows further variants of that genre. Eisenberg (1986) discusses h o w two Mexican families in California mani pulate their children b y teasing them. It is important that the children learn not to believe what is being said. Very often a n adult says something that is highly threatening to the child, like: "We are going to throw Marissa into the garbage can!" This is said with a lot of laughter a n d a marked sing song intonation. Smiling also contextualizes a humorous keying. Very often emotive threats form the kernel of the provocation. For example, a m o t h e r might say that everyone is going to visit grandfather, b u t N a n c y will h a v e to stay h o m e . After N a n c y gets excited, it is m a d e clear that Nancy will of course accompany the others. I n teasing, children are first threatened, b u t then the threat is taken back completely, and the adults thus create a possibility to communicate closeness, security a n d love. T h e teasing in this setting is used as the first part of a ritual with two parts. T h e second is the celebration of love, togetherness, a n d solidarity within the family. Again, it becomes evident that the genre is used in various contexts for various purposes. Sexual and romantic teasing a m o n g youngsters would reveal additional aspects (Eder 1993, Lampert/Ervin-Tripp 2006). A m o n g themselves, adults do not close a teasing sequence b y demonstrating their principal social conjunc tion. T h e y often develop a n e w teasing topic from a previous one (as we saw in d a t u m 3). A teasing topic can b e transformed into a r u n n i n g gag as is the case with Miiller a n d Iris at the newsstand.

4. H u m o r o u s Stories a b o u t P r o b l e m s T h e dialectic of genre knowledge a n d creative authoring can also b e seen in narratives, for example, w h e n c o m p a r i n g talk about problems with humorous talk about problems. H u m o r o u s stories about p r o b l e m s deviate significandy in production and reception from serious stories about problems. Jefferson (1984) has dealt with trouble telling in conversations and shown that in this context laughter produced b y the speaker does not necessarily demand

KOTTHOFF


317

that listeners also laugh. She discusses episodes in which speakers laugh while talking about difficult problems, e.g.: •

i

Datum 4 (fromJefferson 1984) (l)[Frankel:TC:l:4:SO] G: You don't want to go through all the hassle? S: 'hhh I don't know Geri, (•) S: I've stopped crying uhheh-heh-heh-heh-heh. G: Why were you crying? Person S laughs after saying that she has stopped crying. Person G (Geri) does not join in. While laughter o n the side of recipients is the n o r m a l response to the speaker's initial laughter a n d the n o r m a l case in daily discourse, the listeners here display "trouble-receptiveness" (1984:348). Precisely b y n o t laughing they indicate that they take the p r o b l e m seriously. Jefferson writes that initial laughter in the context of p r o b l e m presentations shows that the narrator displays resistance to the p r o b l e m ; she wants to take the p r o b l e m lightly. But the hearer m a y n o t necessarily share this attitude. T h e hearer indicates "trouble sensitivity" if s/he tends to react to the problem content by posing questions about it or m a k i n g serious c o m m e n t s on it. I n connection with p r o b l e m presentations, there is a n e e d for especially strong contextualizations of h u m o r intended to evoke laughter from the hearers. Jefferson's examples suggest that the trouble teller herself laughs relatively late in the p r o b l e m presentation; in h e r data the first laughter particles usually occur in the closing p h a s e of the topic. T h e positioning of laugh particles thus contributes significantly to the social m e a n i n g of the utterances. It makes a differ ence in w h a t phase of presenting a p r o b l e m the tellers laugh. If the problematic aspects are already introduced with laughter, the h u m o r o u s potential of the topic takes the u p p e r h a n d . Recipients expect something funny to follow. In the following we focus o n strategies of contextualizing harmlessness a n d humor in regard to the presented p r o b l e m . In the next datum, A n n i jokes about losing her student status at the university. From the v e r y start, the contextualization of h u m o r prevents a possible "trouble sensitive" reception. Datum 5 (Anni (A), Bernada (B), David (D),Johannes (J), Maria (M), Katharina (K), Ulf (U)) 1 A: aber STELLT euch vor, ich musste mich jetzt im 2 NEUNunddrei (h) Bigsten semester exmatrikulIEren. 3 da hamse Extra ne STUdienberatung eingerichtet. 4 M: nEI:: (heheh) n 5 U: es war FOLgende mEldung in der presse, in berLIN 6 hattense jetzt mAssenhaft ihre [lang= 7 A: [genAU.ja 8 U: [eh die bummela (h) Anten 9

A-

f b e in his fifty-ninth semester. A n n i immediately competes with him -.htkaswrn than i (12). Ulf laughs. I n line 14 she laughs as she announces her own arnbiboo

OFF ORAL GENRES OF HUMOR

319

study goals. Ulf laughs with her. Katharina asks seriously, a n d A n n i explains that a professor h a d hypocritically tried to find out w h a t was w r o n g with these dmaettn. Anni adopts prpforma t h e professor's perspective. But since h e h a d already been presented as hypocritical, it is m a d e clear that the professor was not interested in finding o u t t h e reasons, b u t only in ridding t h e university of dawdlers. Anni ironically plays with thought patterns that are indirectly attributed to the professor (Kotthoff 1998). Her announcement that she h a s b e e n forced to e n d h e r university studies in her thirtv ninth semester seems anything b u t contrite. N o r does she present hiving been enrolled for so long as a personal inadequacy, b u t rather as a sort of game-, the longer t h e better. Katharina's question is n o t understood as a ques tion of how the overly long enrollment could h a v e h a p p e n e d , b u t rather of h o w the termination could have h a p p e n e d . I n h e r presentation, A n n i turns t h e tables on the usual societal evaluations. T h e professor is a negative sort of person: h e has spoilt her fun. She is i n o n e sense a loser - b u t i n a g a m e whose n o r m s she rejects anvway. She invites h e r hearers to laugh with h e r over t h e incongruity of the norms. Here a representative of t h e institution h a s indeed won, b u t at least she has had her fun. In order to accept the casualness of Anni's representation of h e r withdrawal Ixmatrikulation), however, background knowledge is helpful. A n n i h a s b e e n professionally active as a sinologist for some time a n d was enrolled as a student only secondarily, in order to s u p p l e m e n t h e r knowledge of Chinese art history. If she were actually unable to cope with her studies, the presentation and reception would probably have b e e n different. The listeners' reception h e r e again shows that they share Anni's distanced tad amused perspective o n losing one's student status. Ulf n a m e s a negative mribution [Bummelanten/slowpokes) as a quotation, which A n n i emphatically coofirms '91. If Anni h a d presented h e r termination as awkward, it would h a v e Been tasteless to stress t h e awkwardness b y negative attribution. But Ulf can feel Hre that Anni will recognize t h e quotation character, a n d thus h e indicates that he shares her amused perspective o n w h a t h a s h a p p e n e d . Narratives at t h e xOer's own expense can encourage a sharing of perspectives. It is important to bear i n m i n d that s o m e stories might b e told from different perspectives, tailoring t h e m as m u c h as necessary to fit the current context Norrick 20uO). In a different setting, A n n i might very well describe t h e same fvent as really creating a p r o b l e m for her.

5. J o i n t Fantasizing iaracterisiic of this genre is t h e emergent production of a shared fantasy, &ea *»ith several conversational participants m a k i n g short contributions which rttie coherent scenes through t h e incremental structuring a n d augmentation ' unreality. T h e genre shows h o w interlocutors p u t each other o n inferential i d t s and how these tracks can b e processed, drawing o n t h e relevant con.-unial knowledge, so that t h e h u m o r c a n b e immediatelv " h m t - J " ' * —

320

DISCOURSE STUDIES

ethnographical term). It shows h o w several persons closely oriented to each other select formulations which produce a coherent fiction, until the created scene is conversationally p h a s e d out again. I n this case a scene is an imagined situation, in the sense of Karl Biihler's phantasm, which lies outside the ongoing conversational situation (Bange 1986, E h m e r 2004). T h e conversation takes place a m o n g thirty-year-old Viennese in a b a r in Vienna. Most h a v e academic degrees. Before the transcript starts, the group has already b e e n discussing a certain H e r m e s Phettberg. At the time w h e n the recordings were m a d e (1995), Phettberg was a very popular television m o d e r a t o r in Austria a n d Germany. H e wrote a column in a well-known cultural magazine in V i e n n a (Falter), in which h e aired his views on life as such, a n d h e h a d a late-night T V talk show called "Nette Leit Show", o n which h e interviewed celebrities. T h e tide includes a p u n - Leit could suggest either Leute (people) or Lite (light). T h u s it could either m e a n "nice people show" or "nice light show." His professional n a m e is also a pun: it literally m e a n s "mountain of fat" (German: "Fettberg"). His popularity was partly b a s e d o n his unusually corpulent figure, especially for a m e d i a personality; h e is also a confessing h o m o s e x u a l masochist. With his o p e n way of talking about intimate subjects a n d his critical attitude toward the Catholic Church, which is quite powerful in Austria, h e appealed to an intellectual public. T h e group j o i n d y imagines h o w Phettberg, the anti-type, could b e presented as a typical celebrity b y the yellow press. T h e group n o t only cooperates in creating the content of the fantasy b u t also in the style of speaking. Datum 6 (Conversation 19 (Viennese Group I) Episode 9) Conrad (C), Hugo (H), Lilo (L), several (m), Peter (P), Renate (R) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 IS 16 17 18 19 20

C: wieviel kilo dEr hat, waaB A kana, wieviel kilo der WIRKlich hat. R: hundertvierundsIEb [zig? [jEnseits der zwahundert SIcher. P: H: [(? ?) C: [i bin kA BRAvo- leser mit steckbriefsammlu (h) [Ung. L: [he [hehehehe [hehehe m: P: was? H: woher WEIBT du das iiber [haupt. [na, aber zwAhundert, des konnt P: wlrklich sein, ja. C: Amal hob is glesen. P: TdEs war was. THERmes phettberg lEbensgroB. tSTA:Rschnitt in der brA:vo, [na? [hahahahahahahaha m: [kommst a jAhr lang aua. L: H: na, na. im PLAYgirl. nO viel besser. P: ajOa. m: hahahaha 1

KOTTHOFF

21 22 23 24 25 26 27 28 29 30 31 32 33


321

L: als EIne ausgabe fUnfundzwanzig [meter hoch. hehehehehe i P: [dErfalltEHnet i untersjugendverbot= L: hehe P: =weil genitAlien sichst bei dem EH kane durch den bauch, also C: na DER kann nackert ruhig sEIn, [des des H: [DER kann ruhig nackert sEIn. C: mAlen nach zahlen. [HERmes phettberg zum sElber malen. P: [das PHETTberg puzzle. L: hahahapfuiTEUFL. C: da hast a LEbensaufgabe.

Translation: 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

C: how many kilos he really weighs, nobody knows that, how many kilos he really weighs. R: hundred forty-sev [en? P: [over two hundred surely. H: [(? ?) C: [i am not a bravo-reader with a collection of fan [c(h)ards L: [hehehehehehehe m: [hehe P: what? H: how do you know that [anyway. P:

[no, but two hundred could

be possible, yeah C: once i read that. P: Tthat would be something. Thermes phettberg life-sized. Tcelebrity cutouts in bravo, [huh? m: [hahahahahaha L: [lasts for a whole year. H: [no, no in playgirl, much better. P: a whole year. m: hahahahaha L: as a single picture twenty-five [meters high, hehehehehe P: [he does not come under the youth age limits. L: hehe P: =because with him you don't see genitals anyway because of his belly, well C: no for him to be nude is all right [the the H: [for him it's all right to be nude. C: painting by numbers, [hermes phettberg to paint yourself.

31

P:

32

L:

[the phettberg puzzle. hahaha yucky.

33 C: that's a lifetime task. Before t h e transcript begins, s o m e o n e h a d c o m m e n t e d that H e r m e s Phettberg would n o t reach the aee of s e v e n t v s i n r p ht> «rac m n A f«f r> 1 •—

322

DISCOURSE STUDIES

line 1/2 a n d notes twice that n o one knows h o w m u c h Phettberg weighs. R e n a t e offers a concrete guess with a slight question intonation. Peter in line 4 offers an even higher estimate. So far the talk is serious. I n line 6 C o n r a d makes a statement about himself, which d e m a n d s extra processing effort to fit it into the topical context. T h e last w o r d contains a laughter particle. Lilo a n d some others respond with laughter (lines 7,8), which suggests that they understood Conrad's statement to b e funny. C o h e r e n c e seems to b e formed. C o n r a d refers to a youth magazine, Bravo, which publishes so-called Steckbriefe (celebrity trading cards) containing all sorts of information about film, p o p a n d rock stars. H e r e we h a v e a n unusual combination of elements from the life world. Phettberg does n o t at all fit into t h e t r e n d y y o u t h m a g a z i n e Bravo, w h i c h d o e s n o t c o v e r intellectual a n d unconventional personalities. Peter has not u n d e r s t o o d something (maybe h e could n o t see the comical dimension). H u g o ' s question in line 10 is directed at Peter, w h o in line 4 m a d e a claim about Phettberg's weight. Peter in the following softens his assertion a bit. In line 13, C o n r a d agrees with h i m . I n the l e x e m e Steckbriejsammlung (celebrity trading cards), we find the first laugh particles of this episode, which elicit responsive laughter; thereby ratifying a h u m o r o u s perspective o n the conversational topic Phettbergia connection with Bravo. T h e utterance in line 6 does not immediately m a k e sense a n d violates the m a x i m s of quality and quantity, because nothing is ever written about Phettberg in Bravo. According to Sperber a n d Wilson (1985), the listeners h a v e to seek a context for the information which requires as little processing effort as possible, thus preferably the o n e already opened, to which the n e w information can relate. T h e widely differing topics of Phettberg a n d Bravo can relevantly b e connected if one takes Bravo as a context in which celebrity trading cards about Hermes Phettberg could b e created. This presupposes that Phettberg could b e presented as a teenage idol. I n reality, this is so far from the truth that it is amusing. It works as a n invitation to enter the realm of fiction. T h e normative world of stars a n d starlets is connected with the anti-normative world of H e r m e s Phettberg. Line 6 bi-sociates two contrastive frames, as A r t h u r Koestler described in his b o o k o n "acts of creation" (1964). T h e inferencing does not explode maximally in m a n y directions, b u t in a coordinated m a n n e r elaborates the context that was o p e n e d u p b y C o n r a d . I n line 14, Peter continues to fantasize about the topical area of Bravo. The phrase that would be something introduces something that is m a r k e d as unusual. T h e line is syntactically a n d prosodically subdivided into three phrase units, which all h a v e the same r h y t h m and intonation. T h e accent is on the first syllable in each phrase; the intonation falls at the e n d of each. R h y t h m a n d intonation can contextualize comicality, if semantics support this. I n this way a h u m o r o u s key ing is reinforced. T h e youth magazine is k n o w n for its celebrity cutouts; one can gradually fit together small pieces to create life-sized celebrity portraits. The pieces can b e clipped out of Bravo o n e after the other like puzzle pieces. This sort of world knowledge h a s to b e activated. D r a w n out laughter in line 16 indi cates that something funny is being processed. Bravo is introduced in headline style. Formulation (fines 14ff.) certainly is a m o n g the procedures which key h u m o r . Metonymically, elements of Bravo 3

323

(a j o u r n a l has headlines) are used in order to n o t only denote this magazine, b u t simultaneously to stylistically evoke it. T h e syntactic and prosodic forms recreate the semantic content iconically. This "likeness o n several levels" (Jakobson 1960: 369) characterizes aestheticized speech. Conversational h u m o r h e r e enters into the realm of verbal art (Knoblauch/Kotthoff 2002). Furthermore, the continuation of a very elliptical speech style is striking. Ellipses usually contain the theme, the n e w information. I n line 18, for example, the proposal to present Phettberg in Play girl, as well as the evaluation no viel besser/ much better axe like small spotlights o n a n already set stage; lines 19 a n d 21 also cohere in form a n d content to lines 14f. T h e presented fictionalization is absurd, since Phettberg absolutely does not come into question as a teenage idol, suitable to b e presented in Bravo. I n line 16 the women present laugh. Lilo comments on the fantasy that in the case of Phettberg it would take an especially long time to collect all the pieces ( a j a h r / o n e year); she t h e r e b y alludes to his e n o r m o u s girth. Allusions further aestheticize the discourse. I n conversational h u m o r the recipients often actively top the ongoing humor (Norrick 1993). T h e humorous fantasy is displayed as a n ongoing achieve ment. H u g o in line 18 intensifies the fiction about Phettberg in Bravo; h e suggests Playgirl, a magazine k n o w n for erotic photos of n u d e or scantly clad males. Previously created images are then detailed. Peter would like to see him presented a year long in this magazine as well. Several persons laugh (20). In line 21 Lilo stretches the life-sized figure to o n e twenty-five meters high a n d likewise laughs. Starting in line 2 2 , Peter takes u p another aspect of Phettberg's girth a n d erotic self-display. His belly hangs down, serving as though it w e r e a sort of fig leaf. C o n r a d a n d H u g o affirm this impression. Phettberg a n d youth age limit form a n e w combination, which is also developed b y other interlocutors. C o n r a d in line 30 alludes to a g a m e for children. T h e g a m e is called Malen nach Zahlen/ painting by numbers. Peter then continues with a further fictionalization from children's games (the Phettberg Puzzle). Lines 30 a n d 31 again use a headline style. The games are presented like an ad. Lilo laughs a n d inserts a n interjection of dismay. Conrad, b y pointing out that one t h e r e b y has a Lebensaufgabe/lifetime task, again alludes to Phettberg's e n o r m o u s girth. All the fictionalizations draw o n cultural knowledge of entertainment m e d i a and thereby m a k e coherence easy. T h e topic development goes from Phettberg in Bravo, Phettberg in Playgirl, to Phettberg in children's puzzle games. T h e co ordinated imaginings h a v e a meta-message: H e r m e s Phettberg, w h o markets himself as nonconformist, is mercilessly m a r k e t e d in the fantasies of the y o u n g Viennese in a conforming way. T h e y take his b o d y as a starting point for var ious h u m o r o u s quips. T h e interlocutors d o something that Phettberg himself very often does but they do it so-to-speak in a diametrically opposite manner. In numerous interviews, Phettberg himself has referred repeatedly to his unusual body, b o d y feeling a n d sexuality. H e acquired his popularity to a considerable degree due to the fact that h e staged himself as a n appealing anti-type. H e con tradicts several n o r m s of the boulevard press. H e notoriously presented himself in interviews as fat, u n k e m p t , homosexual a n d masochistic - thereby trying to shock a n d simultaneously win over the public, which amused at least part of the

DISCOURSE STUDIES

324

intellectual public in Austria. Consequently, it is amusing to see h i m being inte grated into the yellow press world as though h e were a quite typical T V celebrity. T h e r e b y the y o u n g V i e n n e s e also implicitly c o m m u n i c a t e t h a t t h e y find Phettberg's self-presentation contradictory. T h u s , distance can b e simultaneously displayed toward b o t h Phettberg a n d the yellow press. T h e participants show their knowledge of m e d i a contents, and as well their critical attitude to them, without explicidy evaluating them. T h e evaluation is n o t m a d e explicit, but is jointiy performed. I n their imaginings, the n o r m s of the magazine world are violated, and at the same time the social n o r m s of self-presentation are negotiated, using Phettberg as an example. With a high level of personal participation, thirteen different turns come about which sketch out a fiction a n d amplify the absurdity of its elements (lines 6, 14, 15, 17, 18, 19, 21, 2 3 , 2 5 , 27, 30, 31, 33). I n this genre the most important thing appears to consist in s o m e h o w continuing to spin out the topic within theater frame. This m e a n s that a performative special frame was created. A dif ferent example of this type is discussed in Kotthoff (2005). I n joint fantasizing, condensed information is quickly a d d e d o n to produce the most absurd fictive scene; the short turns iconize the t e m p o of building u p the structure. T h e particular artistry of participation in the formation of such fantasies consists in doing this rapidly (Ehmer 2004). T h e taking over of construc tions is functional in the sense of "on-line" syntactic p h e n o m e n a (Auer 2005).

6. Playing w i t h G e n r e N o r m s 6. 1. Humorous Nutritional Counseling Below, I present another transcript from a dinner shared b y friends (30 to 45 years old) in a G e r m a n academic milieu. Vegetables are being passed around. A n n i responds to t h e m in a n unusual way; she claims that she needs to eat a few carrots just for the sake of vitamins, and this time with butter. T h e n she advises the others with exaggerated emphasis that raw fruits a n d vegetables ought to be eaten with butter. H e r explanations b e c o m e still m o r e amusing w h e n she reveals the source of her nutritional expertise: from the Backerblume ('Baker's Flower', a free magazine available in m a n y G e r m a n bakeries, offering a m o n g other things nutritional advice). A n amusing episode follows. I will explain the sequence conversation analytically a n d pragmatically. Datum 6 (Everyone (a), Anni (A), Bernada (B), David (D), Johannes (J), Katharina (K), Maria (M), several (m), Ulf (U)) 1 B: noch jemand TrUebli::? ((bietet diese an)) 2 A: ICH muss noch welche Essen. 3 ich hab zu wenig vitamin A: und bE:. 4 weil, du musst se ja mit BUTter essen. 5 sonst is das vitamin A hehe ni (h) cht w (h) Irksa (h) m.

KOTTHOFF

7 8 9 10 11 12 13 14 15 16 17 18 18 19 20 21 22 23 24 25 26 27

A: U: A: D: B: A: M: A:

D:

D: a: M: A: m: K: B:

28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

A: B: K: A:

B: U:

U: M:


325

((Zwischensequenz, A nimmt sich Butter)) jetzt tu main bisschen BUTter drAU (h) f. hehe des heifit, man muss eigendich e:h [die dann auch [na ja, ohne BUTTER UNbedingt. rohkost wlrkt nich ohne. (? ?) wUsst ich gar nich. ja::::hehe ich AUch net. Ich hab [rOhkost IMmer SO gegessen. [genau des NUTZT nix. brauchst tBUTter mit dabEI. TBUTTER. [hehehe [deshalb war das schOn ernahrungs [poli (h) tisch WICHtig. hehe was wi (h) r vorhln gesagt ham. [hehehehehehe [hehehehe [hehehehehehe hehehehehehehe °wohEr WEISCHT du des? sowas WEIB ich. weil Ich die tBACKerblume lese. hahahahahahahahahahahaha [hahahahaha [ICH wEIB es auch daher. aber bei Unserm backer TGIBTS jetzt < [keine backerblume me:::hr. ((kindlich)) > [ja genAU. des wird jetzt Alles EINgespart. das musst Ich jetzt AUCH schon mal feststellen. sonst wiird ich sie mir AUCH noch hOlen. [(? ?) [ich hab [die friiher AUCH immer gelEsen. [die BACKerblume und die METZgerzeitung. ich war ganz verzwEI (h) felt als [ich merkte, hehe die BACKerblume kOmmt nich mehr. [kann sein dass es die bei Uns schon LANge nich mehr gab. da hab ich ja noch gAr keinen konTAKT aufgenommen. [in meiner journalistischen lAUfbahn. [hehehehehehehehe wo ich eigentlich [versUch JEdes blatt irgendwie EInzubeziehen [zur bAckerblume? 0

Transla 1 2 3 4 5 6 7 8 9 10

B. A:

anyone else rUebli:::? (Swiss German for carrot) I must eat some more. I have too little vitamin A: and bE:. because, you must eat them with butter. otherwise the vitamin A will no (h) t be (h) effe (h) ct (h) ive. B: well ((incomprehensible side sequence, A helps herself to the butter)) A: now put a little butter on t (h) em. hehe U: that means, one should practically e.h [then also r ii —i * A:

DISCOURSE STUDIES

326

11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

D: B: A: M: A:

D: D: a: M: A: m: K: B:

28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43

A: B: K: A:

B: U: m U M

Absolutely, raw fruits and vegetables have no effect without. I really didn't know that. we::::ll hehe I didn't either. I have always eaten [raw fruits and vegetables plAIn. [that's exacdy what doesn't hElp. you need TbUtter with them. BUTTER, therefore it was already hehe nutritional [poli (h) tically important, hehe what w (h) e said before, [hehehehehehehehehehehehe [hehehehehehehehehe [hehehehehehehehehe hehehehehehehehehehehe "where did you find that out?" I knOw that sort of thing, because I read the TBaker's Flower, hahahahahahahahahaha [hahahahahaha [I also found out about it there, but at our bakery there T is now > ((childish)) [no more Baker's Flower. ((childish)) < [yes exactly, everything is being sAved on now. I Also could not help but notice that, otherwise I would still pick it up. [(? ?) [I [also always used to read it. [the TBAker's Flower and the BUtcher's Journal. I was really desp (he) erate when [1 noticed that BAker's Flower doesn't come anymore. [could be that it hasn't been available here anymore for a lOng time already. I haven't contacted them at all. [in my journalistic career, [hehehehehehehehehehehe whereas I normally try [to include every publication somehow [even Baker's Flower?

W h a t is going on here? Bernada, w h o is from Berlin, asks the group whether anyone else would like Riiebli [carrots). This term, spoken in Swiss G e r m a n dialect (Standard G e r m a n would b e 'Karotten' or 'Mohren') in a diminutive form, re presents a code switch a n d thereby draws attention to the expression as such. The group is meeting in a Swiss town on the G e r m a n border. Talking like the Swiss (German dialects are a popular source of G e r m a n humor) affects the creation of a play layer of action in Clark's sense (1996: 357f.): T h e Berliner even prolongs the i of the Swiss diminutive li exaggeratedly, thereby reinforcing the playful effect. M a r k e d registers a n d "borrowed" varieties are effective performative strategies (Coupland 2001). M a r k e d formulations draw attention to themselves a n d highlight performance. A n n i pretends i n line 2ff. that for the sake of vitamins she still has to eat some i o„rrV.vor t - ^ r t t r A p s in \me 5mdicate\ivatt\ieseieasoTisaietiottobetakeTV 4

KOTTHOFF


327

overly seriously. T h e scene switches back a n d forth b e t w e e n j o k i n g a n d serious ness. But n o p u n c h line h u m o r m o v e s the text from the realm of bona-fide into that of non-bona-fide (as it is sometimes discussed in h u m o r theory, see Attardo 1994). A n n i actually does eat some carrots (Ruebli) with butter a n d offers a nutritional theory that the others r e s p o n d to seriously ( 9 , 1 2 , 1 4 , 2 3 ) . H e r accept ance of the carrots is n o t the usual w a y of accepting offered food, but is easy to understand. Transcripts from conversations often show that utterances contain laughter particles in places w h e r e nothing really funny is b e i n g said. A b o v e all Jefferson (1984, 1985) corrected the d o m i n a n t view in h u m o r research that laughter follows from a h u m o r o u s stimulus. Laughter itself often functions as a stimulus, as a contextualization cue, as a n extra element that lends a n utterance additional m e a n i n g in the sense of: "Take it easy," or " W h a t I a m saying is a bit funny." I n line 5 the laughter contextualizes comicality (see also Glenn 2003). I n line 8 A n n i tells the others laughingly that they should also put butter on their carrots. W h a t inferences could b e m a d e from her laughing speech a n d nutritional counseling? M a y b e she can thereby avoid the danger of being con sidered a glutton. A n n i stylizes herself in a transparent w a y as s o m e o n e w h o al ways acts sensibly. This perspective remains totally implicit a n d vague. As we all know, this is not unusual in everyday talk. We often invite listeners to m a k e a little extra effort to construct additional meaning. All contextualization cues create, as G u m p e r z (1982) and Auer (1986) have pointed out, information o n h o w to interpret what is said. Since these cues are analogous, they m a k e sense only in combination with what is said. D a v i d seriously confirms Anni's theory in line 11. H e must h a v e stayed in the realm of what Attardo (1994) calls the bona-fide. We see that it is not problem atic to react bona-fide to a non-bona-fide utterance. This does n o t m e a n that the discourse is shifted back to the bona-fide. It shows instead that two levels are activated simultaneously. Bernada admits that she does n o t k n o w h o w one ought to eat r a w fruits a n d vegetables. This sequence is also spoken seriously. A n n i then reacts with a drawnout ja (well), in which laughter particles are integrated (line 13): This well, spoken with a gradually falling contour, has a playful sense. We could translate it as, 'There, n o w you see h o w m u c h / k n o w ' . A n n i presents herself as an expert o n nutrition a n d simultaneously comicalizes this role with the laugh particle a n d other strategies. M a r i a reacts seriously to the information (14). T h o u g h they switch back a n d forth between both keys, this does not seem to present a problem. T h e playful keying does not necessarily suspend any of serious meanings that are created. W h a t A n n i tells h e r friends is true, and her suggestions are m e a n t in this sense. But there is a n additional layer of meaning. She could of course h a v e pointed out seriously that vitamins in r a w fruits a n d vegetables are m o r e easily digested if they are eaten with some sort of fat, b u t serious nutritional counseling is not very entertaining. A n n i communicates h e r advice in a double framing: she shows h e r knowledge a n d at the same time invites comical perspectives. I n tines 16 a n d 17, A n n i again gives advice in a very exaggerated manner. Above all the word Butter, repeated twice with a high onset a n d spoken loudly, has a comic effect. D a v i d responds with laughter.

328

DISCOURSE STUDIES

But only with the term "nutritional politically" (ernahrungspolitisch) d o most of the hearers begin to laugh. This academic-sounding t e r m is too elevated for h e r m o d e s t suggestions. A n n i alludes to something they have said before (vorhin gesagt ham). Prior to the episode recounted in the transcript, there was another in which the participants developed the absurd theory that if something tastes good then it must also b e g o o d in a nutritional sense. T h e extended laughter in lines 19 a n d 2 0 also has to d o with r e m e m b e r i n g this absurd theory. T h e partici pants are also invited to infer that A n n i is supporting h e r current culinary tastes with theories whose value d e p e n d s arbitrarily o n whether they fit h e r current tastes or not. This would b e o n e possible w a y to construct relevance for h e r talk. But w h y d o people present arbitrary theories a n d simultaneously pull the rug out from u n d e r t h e m ? S u d d e n shifts into a quasi-scientific register, which are contextually inappropriate, c o m e u p repeatedly with these intellectuals. T h e y play with academic terminology, theatricalize it as though quoting scientific texts, a n d thereby frame it as something unusual. O u r everyday conversations are full of citation-like speech that is n o t intro duced as such (Kotthoff 1998,2002). We can use this m o d e of speech in order to blend in other persons' ways of speaking. Chiefly shared knowledge is what guarantees that such utterances are n o t understood o n a direct level, b u t rather as playing with typical utterances in typical genres in a sort of theater frame b e it academic or parental or whatever - that people can laugh about. P h e n o m e n a like prosody, gestures and mimicry, seen b y Bateson (1953,1954) as meta-messages indicating, "this is play," perform a basic function in the creation of humor. Goffman (1981) calls these "footing," and Clark (1996,2004) calls t h e m "pretense" and "layering." T h e speaker downplays her responsibility for the utterance. This is not to b e confused with non-bona-fide speech. A n n i really motivates the group to take m o r e butter, b u t does this in a n entertaining m a n n e r , contextualized b y integrated laugh particles, the manipulation of v o l u m e a n d a n exaggerated authoritative intonation, w h e r e b y the utterance is m a d e recognizable as a cit ation. We seem to h a v e n o p r o b l e m s with vague communication, as m a n y pragmaticists since Grice h a v e recognized. Nevertheless, Anni's h u m o r o u s way of speaking can only vaguely b e assigned an intention. H u m o r o u s intentions can seldom b e p i n n e d down exacdy. Apparently, such vagueness can b e handled without problems in communication. Now, u p to line 18 Anni is the only o n e w h o laughs. Bernada, David, Maria a n d Ulf react seriously, which is not a p r o b l e m . T h e r e is n o p u n c h line that must b e ratified b y laughter. O n e can very well respond to the serious level of the advice. Anni's comments require n o specific sort of response. Maria asks quiedy h o w A n n i "knows that" (23). A n n i constructs h e r answer in a suspense-creating way. T h e first part of the answer can b e u n d e r s t o o d as in itself bragging (24); it is syntactically independent. A long, i m m e d i a t e laugh follows the information contained in the second statement, Baker's Flower (spoken with high onset). But what is funny about Baker's Flower? Baker's Flower (Backerblume- n o p u n between 'flower' a n d 'flour' is intended, since the corresponding G e r m a n words are 'Mehl' and 'Blume') is a free customer magazine available in m a n y G e r m a n bakeries. A n n i focuses o n this magazine

KOTTHOFF


329

as t h o u g h it w e r e s o m e t h i n g special. Cultural k n o w l e d g e of t h e m u n d a n e , everyday kind comes into play here, since Baker's Flower is well k n o w n to b e a very m o d e s t publication, b o t h topically a n d intellectually. Anni's almost p r o u d reference to Baker's Flower as the chief source of h e r nutritional expertise is h u m o r o u s to the extent that the magazine clearly does not m e e t our expectations about what a university graduate normally reads. This is o n e of the tacit assump tions that Anni rejects w h e n she violates such milieu-specific cultural expectations of n o r m a l behavior. Cultural knowledge has to b e shared if we are to grasp this sort of h u m o r . Katharina announces that she has also acquired similar knowledge from Baker's Flower (26). Bernada's c o m m e n t that h e r baker n o longer carries this magazine is given a complaining and childishly whining u n d e r t o n e through the extended o in more (German: e in mehr). She thereby joins in the h u m o r o u s play of exaggerating the prestige of this lowbrow magazine. Anni confirms the negative trend (28) in bakeries. Bernada also confesses to reading Baker's Flower (32), a n d A n n i extends the r e p o r t of h e r reading interests to the Butcher's Journal (34). A confessional discourse arises, keyed as h u m o r o u s , a n d several participants join in. T h e y perform their e x t r e m e d i s a p p o i n t m e n t at the d i s a p p e a r a n c e of this valued source of information. T h e staging of h u m o r focused on Baker's Flower is j o i n d y p r o d u c e d . I n the emergent discourse, the scope of the h u m o r gradually b r o a d e n s . T h e introduction in line 3 5 of the sense of feeling 'desperate' as a reaction to the disappearance of the magazine points again to exaggeration as a keying procedure. Baker's Flower is h u m o r o u s l y transformed into a n intellectually signi ficant journal. O n e can take part in this h u m o r b y p r e t e n d i n g to b e quite serious o n the explicit level, as Ulf does in the following (39). Ulf is a journalist with high standards a n d n o w pretends that h e might want to publish in Baker's Flower a n d that that this would help to m a k e u p for a personal journalistic deficiency. Several of the friends laugh. T h r o u g h his participation in exaggerating the status of a trivial popular magazine, Ulf also shows that h e is ratifying the h u m o r of the preceding discourse. His r e m a r k is integrated into the already constituted h u m o r o u s discourse a n d e x p a n d s o n it. Ulf speaks in a serious tone, although the g r o u p knows that h e normally publishes in m u c h m o r e prestigious journals. I n the e m e r g e n t discourse, the h u m o r is developed step b y step. A level of bona-fide is nevertheless preserved. Again a n d again w e find exaggeration used as a h u m o r strategy. T h e h u m o r o u s counseling as well as the h u m o r o u s play with the upgrading of m o d e s t journals allow t h e group, consisting of academics, to distance them selves from the n o r m s of the academic world. Speakers index informality b y maximizing an intertextual gap from serious counterparts of genres such as counseling or talk about journals. 6. 2. Humorous

Gossip

Bergmann (1987/1993) analyzed gossip as a reconstructive, collaborative genre of m o r a l communication. I n gossip, "discreet i n d i c r r ° f ; « - "

DISCOURSE STUDIES

330

gossip has a bad reputation, it is widely indulged in and even fulfills important functions for group formation. The producer of the gossip and the addressees talk about an absent object. The information is delivered as delicate. In the next datum, the interlocutors play with the genre of gossip. Datum 7 (Conversation 14 Episode 10) (David (D), Ernst (E), Inge (I),Johannes (J), Katharina (K), Maria (M), Rudolph (R)) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 22 23 24 25 26 27 28 29 30 31 32 33 34

K: R: K: R: D: R: D: R: M: I: R: E: R: D: K: M: I: R: a: I: K: D: M: R:

I: D: a: R: a: K:

Irgendwie hast du=s AUCH nicht mEhr so mit frau donner. ne? [m:::::::: [frUher war das mal, da gab=s doch mal mEhr kontakt. sie hat en FREUND jetzt wieder, m:::: wie langgOht des schon? '; he? wie lang gOht des schon. ja seit SECHS wOchen ungefahr. und sEIther ist sie wieder UMganglicher. achsO? was ISCH des fur EIner? oh jE. jetzt eh eh RED ich naturlich wieder ausm nAhkastchen. dEs erfahr ich sowieSO. he? [(? ?) [bei UNS ist das gUt [aufgehoben. [hehe [hehehehehe [hehehehehe es ist der VAter ihres sOhnes. NA::::I:::N po::: erzA::h::l. WER das is wollnwaja gar nicht wlssen, rudolph. das dArf doch KEIner wissen. [(? ?) [also wenn ihr jetzt nAchstens ins TREppenhaus geht, und die lllo kommt, sagst du, wir wissen NICHT, dass dun verhAltnis mit dem vAter [deines SOHnes hast. [hehehehehehehehehehehehehehe wir wissens auch NICHT von rUdolph. hahahahahahahahahahahahaha wir hams namlich letzten mOntag GAR nicht erfahren. hahahahahahahahahaha [haha [rudolph, magst du den sEkt....

Translation: 1 K somehow things are not going so well for you and frau. donner. isn't that so? 2 R 3 K [previously there was, there was really more contact, 4 R now she has a boyfriend again, . 5 D how long has that been going on already?

KOTTHOFF

331


7 D: how long has that been going on already? 8 R: well for about six weeks, and since then she is more approachable again, 9 10 M: oh really? 11 I: what sort of person is he? 12 R: oh well. 13 now uh uh i am naturally giving away secrets, 14 E: i will find that out anyway, 15 R: he? 16 D: [(? ?) 17 K: [it is safe with us. 18 M: [hehe [hehehehehe 19 I: [hehehehehehe 20 R: it is her son's father. 21 a: NO:::::::: 22 I: po::: 22 K: te:::ll us 23 D: who that is we definitely do not want to know, rudolph 24 no one is supposed to know. 25 M: [(? ?) 26 R: [so if you step into the stairwell sometime soon, 27 and lilo comes, you will say, we do not know, that you are having a relationship with 28 the father [of your son. 29 I: [hehehehehehehehehehehehehehe 30 D: nor do we know it from rudolph. 31 a: hahahahahahahahahahahahaha 32 R: we definitely did not learn about it last monday. 33 a: hahahahahahahahahaha [haha 34 K: [rudolph, do you like the sparkling wine.. Katharina asks R u d o l p h about his current relationship with Frau Dormer, w h o works as a psychologist in the n e i g h b o r h o o d and is usually addressed informally by everyone as Lilo. T h e formal reference to Frau Dormer instead of Lilo indexes distance from the lady in question. This is the first step to establishing a possible object of gossip. R u d o l p h ' s drawn-out interjection in line 2 is h a r d to interpret, but s o m e h o w signals the delicacy of the subject. H e answers in line 4 that Frau Dormer apparently has a boyfriend again. A wonderful topic of gossip is thus established. D a v i d reacts with the same drawn out interjection m::: (thereby underlining the delicacy) a n d with a question in Alemannic (a south-German dialect), which h e otherwise never speaks (wie langgoht des schon? H o w long has that been going on already?), and which thereby becomes an indicator of comicalization, a stage separator in H a i m a n ' s sense (1990). N o w a gossipy conversation is imitated. I n line 7 David repeats the question in Alemannic in response to Rudolph's questioning signal. T h e stylization of the question in Alemannic dialect gives his words a quotation-like character. R u d o l p h replies normally a n d com ments on Frau Donner's relationship. M a r i a a n d I n g e would like to k n o w m o r e (10, 11). R u d o l p h verbalizes the difficulties h e is having in his role as a gossip informant I ietrt eh oh r*>J —-* " 5

J

332

DISCOURSE STUDIES

u h I a m naturally giving away secrets.) T h e group seems to b e fully aware thai gossip has a b a d reputation. Ernst allays his misgivings. H e also has other sources. Katharina pretends to b e acting therapeutically (17), as t h o u g h the informani would b e relieved if she could share the secret. T h e r e is a h u m o r o u s incongruity in this, because the opposite is obviously the case. Maria a n d Inge laugh. Rudolph presents the key piece of information in formal syntax (es ist... it is...). In line 21 a sort of exclamation goes through the r o u n d which is so exaggerated that it re inforces the theater frame of the gossip. Inge inserts an interjection which play fully underlines the outrageousness of the news (po::.). Katharina insists that R u d o l p h tell t h e m about it. Everyone knows about Frau D o n n e r ' s child and hat already o n various occasions w o n d e r e d about the father's identity. Now David turns the tables: H e evinces explicit disinterest in exactly what everyone is so anxious to k n o w (23,24). Starting at line 2 6 , R u d o l p h stages scenes of meetings with Lilo Bonner. T h e y imagine saying to h e r that they d o n ' t k n o w any of the things they h a v e just b e e n talking about. Inge laughs (29). D a v i d expands on R u d o l p h ' s fantasy of the dialogue (30). E v e r y o n e laughs. R u d o l p h speak' even m o r e concretely in line 32. Everyone laughs again. T h e imagined dialogue with Frau Donneris absurd. T h e joking episode reaches a climax and ends, amonj. other things, with the drinks being refreshed. I n playing with gossip, some gossiping really is going on. I n a humorous frame, people can distance themselves from a speech genre with a b a d reputation (Bergmann 1987/1994) and simultaneously still carry o n the activity. T h e mail, piece of information, that Frau Donner is having a relationship with the fathe; of her child, whose identity they h a v e k e p t secret, is in any case passed on. T h e emergent play is so successful because e v e r y b o d y knows not only thir genre, but also the ideology underlying it. K e y information about Frau Donner is being transmitted in a play frame.

7. Final R e m a r k s : Transcending G e n r e by Relying on Genre Knowledge I n this paper I have considered humorous genres (jokes, teasing, joint fantasizing) a n d determined that we can find a basic pattern for them. Nevertheless, the realizations of these genres are related only b y a sort of family resemblance in Wittgenstein's sense. G e n r e knowledge is, however, e m p l o y e d precisely when the speaker goes outside the genre a n d w h e n the pattern is violated in such a way that further information is located precisely in the violation. T h e n I considered humorous realizations of genres that modify a serious genre (humorous stories about problems, h u m o r o u s counseling, humorous gossiping), H e r e the framing is d o n e from the start in such a w a y that a serious mode of understanding is u n d e r m i n e d . T h e h u m o r o u s realization is co-constructed. The co-construction is of course emergent, but nevertheless (or precisely for thii reason) it relies o n genre knowledge.

KOTTHOFF


333

Along with Clark (1996), we can say that in h u m o r o u s realizations of genres a second meta-communicative layer is m a d e relevant. A level of c o m m e n t a r y on the said arises b y m e a n s of which speakers distance themselves from their messages. I n playing with gossip, the gossip is k e p t as a n intertext. T h e partici pants also communicate knowledge of the b a d reputation of the genre. T h e y take the offensive toward the ambivalence attached to the genre in everyday life (on the o n e side a b a d reputation - o n the other pleasure in the exchange of discrete indiscretions). O t h e r intertexts are also included in the g a m e , e.g. therapeutic discourse. People act as though it is m o r e in the interest of the teller to tell something, than in the interest of satisfying their o w n curiosity. We viewed genres from a performance perspective a n d witnessed h o w an actual co-construction of ongoing discourse indexes social relationships, m o r a l stances and a certain context. With Briggs and B a u m a n (1992) we can see datum "». 6 and 7 as maximizing an intertextual gap. Creative improvisation blurs any sharp distinctions a m o n g genres. Although the intertextual gap is smaller in realizing jokes, teasings or joint fantasies also these genres of h u m o r d e m a n d high performance standards - rather monological in the case of jokes and rather dialogical in the case of teasing and joint fantasies. Transcription conventions (based o n GAT, Selting et al. 1998)

(--) (0.5) (? what?) (??)

one hyphen indicates a short pause two hyphens indicate a longer pause (less than half a second) pause of half a second; long pauses are counted in half seconds indicates uncertain transcription indicates an incomprehensible utterance

indicates overlap or interruption latching of an utterance of one person; no interruption laughter hahaha slight laughter hehehe integrated laughter goo(h)d audible exhalation (h) audible inhalation »'h) slightiy rising intonation rising intonation falling intonation ongoing intonation indicates elongated sound lower amplitude and pitch blabla° COME ON emphatic stress (pitch and volume shift) iOme ON primary and secondary accent syllable within a sentence (only in the original language of the transcript) high onset of pitch pitch goes down low pitch register within the brackets < smiling)> comments ((sits down)) nonverbal actions or comments :,

334

DISCOURSE STUDIES

Author's N o t e This article will also appear in the journal Pragmatics in 2007. I wish to say thanks to Pragmatics.

Notes 1. The data stem from various circles of friends who were at the time of the recordings between 30 and 40 years old; most have an academic training background. The data are characterized in Kotthoff 1998. 2. The irony in this scene is also discussed in Kotthoff 2002. 3. Literature theorists such as Iser 1992 see that as a typical procedure to create an. They discuss only written texts such as novels or poetry. Humor shows all features which Iser calls artistic staging. 4. This diminutive is well-known in the German speaking world and is identified by most speakers as typical for Swiss German. See for German dialects Barbour and Stevensen 1999. 5. I discuss the potentials of shifting and crossing dialects as a humor strategy in Kotthoff 2006b.

References Attardo, Salvatore (1994): Linguistic Theories of Humor. Berlin/New York: Mouton de Gruyter. Auer, Peter J. C. (1986): Kontextualisierung. Studium Linguistik 19: 22-48. Auer, Peter (2005): Syntax als Prozess. InList 41 [http://www.uni-potsdam.de/u/inlist/ issues/41]. Bakhtin, Mikhail, M. (1986/1994): Speech genres. In: Pam Morris (ed.): The Bakhtin Reader. London: Edward Arnold, 80-87. Bange, Pierre (1986): Fiktion im Gesprach. In: Werner Kallmeyer (Hrsg.): Kommunikationstypologie. Diisseldorf: Schwann, 117-153. Barbour, Stephen/Stevenson, Patrick (1999): Variation in German. A critical approach to German sociolinguistics. Cambridge: Cambridge University Press. Bateson, Gregory (1953): The Position of Humor in Human Communication. In: H. von Forster (ed.): Cybernetics Ninth Conference. New York: Josiah Macy. Jr. Foundation, 1-47. Bateson, Gregory (1954): A Theory of Play and Phantasy. In: Steps to an Ecology of Mind. San Francisco: Chandler, 177-193. Bergmann,Jorg R. (1987/1993): Discrete Indiscretions. The Social Organization of Gossip. New York: Mouton de Gruyter. Briggs, Charles/Bauman, Richard (1992): Genre, Intertextuality, and Social Power./ounw/ ofLinguistic Anthropology 2,2: 131-172. Clark, Herbert (1996): Using Language. Cambridge University Press. Clark, Herbert (2004): Pragmatics of Language Performance. In: Laurence R. Horn/ Gregory Ward (eds.): The Handbook of Pragmatics. London: Blackwell, 365-383. Couper-Kuhlen, Elizabeth (1999): Coherent Voicing. On Prosody in Conversational Reported Speech. In: Wolfram Bublitz et al. (eds): Coherence in Spoken and Written -

--'

"

1

v

1

J

•_

T—T

:c

1\

KOTTHOFF


335

Coupland, Nikolas (2001): Dialect stylization in radio talk. Language in Society 30, 3: 345-375. Drew, Paul (1987): Po-Faced Receipts of Teases. Linguistics 25: 219-253. Eder, Donna (1993): "Go get ya a French!": Romantic and Sexual Teasing among Adole scent Girls. In Deborah Tannen (ed.): Gender and Conversational Interaction. New York/Oxford: Oxford University Press, 17-30. Ehmer, Oliver (2004): Spontane Inszenierungen in Alltagsgesprdchen. Diplomarbeit, Universitat Halle-Wittenberg. Eisenberg, Ann R. (1986): Teasing: Verbal Play in Mexican Homes. In: Bambi B. Schieffelin/ Elinor Ochs (eds.): Language Socialization across Cultures. Cambridge: Cambridge University Press, 182-199. Foley, William A. (1997): Anthropological Linguistics. An Introduction. Maiden and Oxford: Blackwell. Glenn, Phillip (2003): Laughter in Interaction. Cambridge: Cambridge University Press. Goffman, Erving (1981): Forms of Talk. Philadelphia: University of Pennsylvania Press. Giinthner, Susanne (1996): Zwischen Scherz und Schmerz. Frotzelaktivitaten im Alltag. In: Helga Kotthoff (Hrsg.): Scherzkommunikation. Beitrage aus der empirischen Gesprdchsforschung. Opladen: Westdeutscher Verlag, 81-109. Gunthner, Susanne (1999): Polyphony and the 'layering of voices' in reported dialogue: An analysis of the use of prosodic devices in everyday reported speech. Journal of Pragmatics 31: 685-708. (also published as InList-paper 3) Gunthner, Susanne/Knoblauch, Hubert (1995): Culturally Patterned Speaking Practices - the Analysis of Communicative Genres. Pragmatics 5, 1: 1-32. Gumperz, John J. (1982): Discourse Strategies. Cambridge: Cambridge University Press. Haiman, John (1990): Sarcasm as Theater. Cognitive Linguistics 1-2: 181-205. Hanks, William (1995): Language and Communicative Practice. Boulder, CO: Westview Press. Iser, Wolfgang (1992): Staging as an Anthropological Category. New Literary History 23, 4: 877-888. Jakobson, Roman (1960): Closing Statement: Linguistics and Poetics. In: Thomas Sebeok (ed.): Style in Language. Cambridge: MIT Press, 350-377. Jefferson, Gail (1984): On the Organization of Laughter in Talk about Troubles. In: Max Atkinson/John Heritage (eds.): Structures of Social Action. Cambridge: Cambridge University Press, 346-269. Jefferson, Gail (1985): An Exercise in the Transcription and Analysis of Laughter. In: Teun A. van Dijk (ed.): Handbook of Discourse Analysis. Vol 3. London: Academic Press, 25-34. Knoblauch, Hubert/Kotthoff, Helga (2002) (eds.): Verbal Art across Cultures. Tubingen: Narr. Koesder, Arthur (1964): The Act ofCreation. London: Hutchinson. (Dt. 1966): Der gStdiche Funke. Bern/Munchen/Wien: Scherz. Kotthoff, Helga (1996): Impoliteness and ConversationalJoking: On Relational Politics. Folia Linguistica XXX, 3-4, 299-327. Kotthoff, Helga (1998): SpaB Verstehen. Zur Pragmatik von konversationellem Humor. Tubingen: Niemeyer. Kotthoff, Helga (2002): Irony, Citation, and Other Forms of Staged Intertextuality. In: Carl F. Graumann/Werner Kallmeyer (eds.): Perspectivity in Discourse. Amsterdam: Benjamins, 201-233. (also published as InList-paper) Kotthoff, Helga (2005): Artistische Dimensionen der Alltagsrede. Performanzorientierte Miindlichkeit bei Kindern und Erwachsenen. Muttersprache, VierteljahresschriH fiir

336

DISCOURSE STUDIES

Kotthoff, Helga (2006a): Humor and gender. The state of the art. Journal of Pragmatics 1 (vol. 38): 4-26. Kotthoff, Helga (2006b): The humorous stylization of "new" women and men and con servative others. In: Peter Auer/Werner Kallmeyer (eds.): Communicating Social Identity. New York/Berlin: de Gruyter (forthcoming) Muntigl, Peter/Gruber, Helmut (2005): Introduction: Approaches to Genre. Folia Linguistics XXXIX, 1-2,1-18. Lampert, Martin D./Susan M. Ervin-Tripp (2006): Risky Laughter: Teasing and SelfDirected Joking among Male and Female Friends. Journal of Pragmatics 1 (vol. 38): 51-73. Legman, Gershon (1970): Der unanstdndige Witz, Hamburg: Hoffmann und Campe. Luckmann, Thomas (1986): Grundformen der gesellschaftlichen Vermitdung des Wissens: Kommunikative Gattungen. Kolner Zeitschrift fur Soziologie und Sozialpsychologit 27: 513-523. Luckmann (2002): Kommunikation. In: Thomas Luckmann: Wissen und Gesellschafl. Ausgewa'hlte Aufsatzf 1981-2002. Konstanz: Universitatsverlag, 157-201. Norrick, Neal R. (1993): Conversationaljoking. Humor in Everyday Talk Bloomington, IN: Indiana University Press. Norrick, Neal R. (2000): Conversational Narrative. Storytelling in Everyday Talk Amsterdam: Benjamins. Radcliffe-Brown, Alfred R. (1940): On Joking Relationships. In: (1965): Structure and Function in Primitive Society. New York/London: Cohen/West, 90-104. Sacks, Harvey (1974): An Analysis of the Course of ajoke's Telling in Conversation. In: Richard Bauman/Joel Sherzer (eds.): Explorations in the Ethnography of Speaking. Cambridge: Cambridge University Press, 337-353. Sacks, Harvey (1978): Some Technical Considerations of a DirtyJoke. In: Jim Schenkein (ed.): Studies in the Organization of Conversational Interaction. New York: Academk Press, 249-275. Schmitt, Reinhold (1992): Die Schwellensteher. Sprachliche Prasenz undsozialer Austausch an einem Kiosk Tubingen: Narr. Sperber, Dan/Wilson, Deidra (1985): Relevance. London: Blackwell. Straehle, Carolyn A. (1993): "Samuel?" "Yes, Dear?": Teasing and Conversational Rapport In: Deborah Tannen (ed.): Framing in Discourse. New York/Oxford: Oxford Universirv Press, 210-229. Swalesjohn M. (1990): Genre Analysis. English in academic and research settings. Cambridge : Cambridge University Press. Voloshinov, Valentin N. (1929/1975): Marxismus undSprachphilosophie. Frankfurt/Berlm Wien: Ullstein. 1

43 Against Arbitrariness: The Social Production of the Sign as a Foundational Issue in Critical Discourse Analysis Gunther Kress

1. Critical R e a d i n g

O

n e aspect of critical linguistics (CL) which has attracted the particular attention of s o m e commentators is the status of critical readings. ' W h a t is it', so the challenge is formulated 'that privileges the readings of the critical linguist over those of the ordinary reader?' This is a challenge of a funda mental kind. W h y should the readings of o n e reader h a v e greater legitimacy or privilege than those p r o d u c e d b y other readers? O f course, this question has particular topicality in the current, postmodern, period, a n d in poststructuralist theories, in which the plurality of readings is taken as a given, a n d the authority of certain readings is challenged, along with a far-reaching challenge to power and its effects. A n d indeed, as critical discourse analysis (CDA) in its very aims seeks to reveal the structures, locations a n d effects of p o w e r - whether in the operation of discourses of race, or of gender, or of ethnicity; or in the operation of power at micro-levels as in relations across the institutional/individual divide (doctor-patient interactions, for instance); or in the interactions of socially pos itioned individuals in everyday relations - this challenge is a fundamental o n e to CDA. There is always the possibility that practitioners of C D A will replicate the operations of p o w e r in the readings which they produce. My suggestion, in response to this challenge, a n d in its refutation, is that C D A needs to produce a clearly articulated theory of the reading of texts as m u c h as a theory of the production of texts, a n d that such a theory needs, crucially, to b e founded on a theory of the social production a n d reading of signs. A s such,

Source: Discourse & Society vol. 4, no. 2,1993, pp. 169-191.

338

DISCOURSE STUDIES

C D A will b e c o m e , a n d this too is an essential step, p a r t of that larger enterprise which deals with the production of m e a n i n g in any semiotic m e d i u m , namely that of social semiotics. T h a t enterprise is o n e which attempts to account for the making a n d remaking of m e a n i n g in all those semiotic m o d e s which cultures employ, differentially, b u t as a matter of course. A s just o n e consequence of such an approach language will n e e d to b e seen as just o n e such semiotic m o d e , and therefore the longstanding theoretical a n d political buttressing of language as the primary, most significant, m e d i u m ('logocentrism') will b e challenged. That has far-reaching effects o n structurings of power, n o t least in w h a t present them selves at the m o m e n t as u n m e n t i o n e d givens about language and its role in social, cultural a n d cognitive structurings. I n this p a p e r I (1) outline a theory of the sign as a motivated a n d conven tionalized semiotic entity; (2) focus on a n account of the reception/reading of signs in cultural histories, which introduces and deals with the notions of opacity, mystification a n d ideology; (3) deal with the question of the boundaries of signs, a n d the interplay of p o w e r a n d the question of boundaries as punctuations of ongoing semiosis; a n d (4) provide two analyses which demonstrate both the need for a n d the critical effect a n d potential of this theorization. T h a t is, m y argu m e n t will b e that this account acts not merely to r e m i n d C D A analysts of other m o d e s through which m e a n i n g is m a d e , but d e m a n d s a significant rethinking of practices within C D A . Some of the most crucial assumptions which u n d e r p i n critiques of critical reading (CR), conducted in C L , or in C D A , are: (a)

unproblematized notions of text, a n d of the text which is to b e read: that is, text is seen as unproblematically established, fixed, bounded; (b) that reading is treated as a mirror-image of text production, that is, the conditions of the production of the text are regarded as identical with those of the reading of the text, in all essential factors; (c) that as a corollary of (b) producers of texts and their purposes are treated as identical to those of readers of texts; (d) that language is a n a u t o n o m o u s , self-contained system, discrete in all r e s p e c t s from o t h e r semiotic a n d social s y s t e m s of practices, whether of representation (as kinds of social action) or of action, so that the meanings of the verbal semiotic seem in n o sense to d e p e n d on the c o n t r i b u t o r y m e a n i n g of any other semiotic; a n d lastly, a n d most significantly, (e) the c o m m o n sense, including the theoretician's c o m m o n sense, sur r o u n d i n g the question of the sign, particularly (i) the assumption of the arbitrariness of the relation of the signifier and the signified in the sign the relation of 'form' a n d 'content' in common-sense parlance; and (ii) the assumption of the stable nature of the sign. I n m y view some or all of these underlie criticisms of the enterprises of C L C D A , C R ; and inevitably underlie a n d buttress the theoretical edifices of most mainstream versions of linguistics, semantics and, surprisingly, even pragmatics. A n y activity which goes b e y o n d assumptions of the naturalness a n d transparency

KRESS

SOCIAL PRODUCTION OF THE SIGN

339

of language brings these to the surface. In this sense, then, mainstream linguistic theories and descriptions are founded on these assumptions. For instance, (b) and (c) above together imply either a theory in which power is treated as so signifi cant that readings are enforceable and enforced, or else an asocial theory which treats a communicating group as composed of atomistic individuals of exactly the same (biological, cognitive, cultural, social, etc.) make-up. Such implications are naturalized in most mainstream linguistic theories, and in much semantic and pragmatic work also. However, even in critical theories of text production and reception these questions are not usually brought into focus or discussed: though it is the case that most practitioners in C D A would regard most of these as relevant points, or questions. The exception to this in C D A is the question of the sign; that theoretical category, either in the traditional Saussurian sense, or in any sense of critique, is absent from discussion. At the same time there is an increasing amount of work which is beginning to engage with the question of the sign, particularly in terms of a critique of the traditional conception of the Saussurian sign (Halliday, 1978, 1985; Hodge and Kress, 1988,1993; Kress, 1978; Lentriccia, 1980). Nevertheless it is the case that even where the discussion has been broadened, its implications have not been taken up. This means, in any case, that through all of this century linguistics, with some exceptions, has remained as the science of the signifier (to borrow a phrase from Tobin [1990], who himself, however, perpetuates the idea of the sign as an arbitrary construct). To the extent that C D A relies on existing linguistic theories, it is always in danger of importing that discipline's unexamined baggage of common sense. A social theory of the production of the sign is the essential underpinning not only of critical theories of text production and reception, but also of any plausible theory of language and text. In this paper I articulate aspects of a social theory of the sign in relation to four points: the characteristics of signs; the production/ reception of signs; rules determining the boundaries of signs; and the multimodality of all texts as signs. As a part of this argument I attempt to provide refu tations of the assumptions outlined earlier.

2. Characteristics of Signs This drawing was produced by a 3V2-year-old child. The child first drew the four circles in the centre, then added two circles on the left, and finally the one circle in the top right corner. Having drawn it, he said 'This is a car'. What is crucial here is the relation of the signified - a 3V2-year-old child's idea of 'car' at this point in time - to the signifier - the seven circles. It illustrates a number of fundamental characteristics of all signs. This relation, and the resultant sign, encodes, on the one hand, characteristics of the producer of the sign, and, on the other, characteristics of the object to be represented. For the producer of the sign, wheels seems to be the most prominent, the criterial aspect of the car, they become, for him, at that age, the signified. Given this producer's physical size it is plausible to assume that when he looked

T h i s is a car'

at his parents' car the wheels were the most obvious features, especially as at that time 'his' car used to b e parked on a driveway which sloped upward from the front door of the house w h e r e h e a n d his family were staying, m a k i n g the wheels seem e v e n m o r e prominent. At the same time, the sign is also motivated b y fea tures of the object to b e represented, the object which is t h e stimulus of the sig nified for this child sign-producer. T h a t is, the car does h a v e wheels. Similarly, circles, a n d the arrangements of the circles, are b o t h an apt signifier for the signified 'wheel'. This sign therefore is n o t the product of a n arbitrary association of a signifier a n d a signified, either from the point of view of the producer, or from a consideration of characteristics of the object. F r o m the point of view of the producer it represents his particular 'interest' in the object, an interest which is itself a reflection of his place in the world, physically, cognitively, socially, culturally, conceptually. This 'interest' is n o t fixed but is the expression of a t e m p o r a r y configuration of socially a n d culturally p r o d u c e d internal rep resentations - a drawing d o n e the n e x t day might have produced a different sign, fixing on, selecting different aspects of the car as criterial. Certainly, drawings of cars d o n e b y h i m now, around the age of five years, reflect different interests, and, above all, reflect the increasing awareness of culturally existent conventions, itself an aspect of 'being in the world' - the w a y cars are represented in books for children, for instance. It is important to insist that the sign is always motivated from the point of view of the object; that is, that the sign always has referential dimensions, in that it is particular aspects of the object which are affected. These t h e n affect the characteristics both of what m a y b e used b y the p r o d u c e r to con struct signifieds, a n d of what can function as apt signifiers. To give another example, from the verbal semiotic, one which complicates the picture somewhat. T h e same child, at this age, while climbing a very steep hill, said: 'This is a heavy hill.' As before, aspects of his 'interest' - great expenditure of effort for a small child - a n d aspects of features of the object to b e signified produce a particular signified - p e r h a p s something like: 'This activity takes con siderable physical effort' - and this expresses aspects of the referent that pro duced, in part, this signified - the difficulty of climbing the steep hill. T h e r e are two differences c o m p a r e d to the previous example. T h e object/ referent is n o t a relatively integral object, such as the car, but a highly complex behavioural event a n d a resultant conceptual structure. A n d whereas in the case of the car the signifier could b e produced newly b y the child, h e r e h e is operating

KRESS


341

in a semiotic system (spoken language) in which signs already exist. He therefore selects an existing sign which in his view most nearly expresses aspects of the signified, namely 'heavy'. That is, an existing sign in one semiotic system becomes the signifier for a newly produced signified. This view of the production of signs has close affinities, formally, with the theoretical work of Hjelmslev (1943/1969) and, deriving from it, that used in the writings of Barthes (1982) and Eco (1976). The significant and crucial difference lies in my attempt to locate this process firmly in the social and cultural histories of the producer of the sign. In other words, the production of signs, in semiotic systems with already available sign/signifiers, is subject to constraints which the child (or adult pro ducer of a sign) is not subject to in a semiotic medium which does not already provide ready-made signs. Even here, however, the sign is a motivated construct, even though some limitation or circumscription is placed on the production of the sign by the presence of already existent signs. In other words, as a sign-producer moves into greater facility with existing semiotic systems, the production of signs takes place in a situation of ever increasing tension between the meanings of existing signs, which provide the signifiers (Hjelmslev's 'expression substance'), and the producer's need or wish to produce new signs. Generalizing, I wish to say that signs are always motivated in this manner by the producer's 'interest', and by characteristics of the object. It is 'interest' which determines the characteristics that are to be selected and to be represented. The relation of signifier to signified, in all human semiotic systems, is always motiv ated, and is never arbitrary. This may seem a tenable hypothesis in relation to some signs only - certain visual icons, onomatopoeia in verbal language perhaps; it will seem entirely untenable in relation to the vast majority of signs in all kinds of semiotic systems, and in particular in relation to signs in verbal language words, for instance, or sounds. It is as well to recall here Volosinov's statement: 'The form of the sign is conditioned above all by the social organization of the participants involved and also by the immediate conditions of their interaction' (1930/1973: 21). I will give some more examples and then extend the hypothesis even fur ther, to include all those instances of signs which appear to be quite outside this hypothesis, namely all those instances which appear to be simply re-uses of 'existing' signs. If the notion of 'interest' is extended to the making of signs by the fully acculturated, fully socialized individual, we will have to take into consideration the individual's social and cultural histories, and her or his present social pos itioning in the whole complex of social structures which make up an individual's social life. 'Interest' is the articulation and realization of an individual's relation ship to an object or event, acting out of that social complex at a particular moment, in the context of an interaction with other constitutive factors of the situation which are considered as relevant by the individual. These might, for instance, include other individuals and their social positionings, present purposes and in tentions, structures of power which obtain in the situation, etc. Signs, then, reflect, in their construction, that relation of signified to signifier which expresses this interest That is, in relation to a particular object or event, 'interest' leads the

342

DISCOURSE STUDIES

p r o d u c e r of the sign to focus o n a particular characteristic of a n object o r event (whether an object or an event in the physical or in the social/cultural, semiotic world) to m a k e that the criterial characteristic of the object or event, that is, m a k e it the basis of the production of a signified. It m a y b e w o r t h drawing attention to the fact that the process which I a m de scribing here is the process underlying the construction of a metaphor. A n aspect of the object or event to b e represented b y being d r a w n into the metaphoric relation is selected b y the speaker or writer as having, at that point a n d in relation to the speaker/writer's purposes, criterial characteristics in respect to the object to b e represented; that is, it b e c o m e s the signified. This criterial aspect simul taneously specifies the characteristics which will m a k e an apt signifier. So in re lation to the earlier e x a m p l e of the car, the child decides that wheels are criterial in representing the signified 'car'. T h a t decision t h e n determines w h a t m a y b e an apt signifier, in this case circles. W h e e l s are selected as the criterial aspect of the object to b e represented (the signified), a n d circles are apt signifiers in the sense that t h e y adequately signify, represent or express the characteristics of the signified. All signs are formed in this metaphoric process. All signs are metaphors. Two consequences follow for C D A . O n the o n e h a n d all signs are metaphors, h e n c e code ideological positions in that they realize the social, cultural and therefore political position of their producer. O n the other h a n d all signs are thus equally subject to critical reading. This last point h a s t h e important theoretical a n d methodological consequence that n o sign is innocent. T h e r e are therefore n o texts which can escape critical reading. All texts equally c o d e the ideological positions of their producers. T h e everyday, innocent and innocuous, the mundane text is as ideologically saturated as a text which wears its ideological constitu tion overfly. Consider, for instance, the signified 'social distance'. A n y sign which already expresses distance - temporal, spatial or other kinds of distance - can become an appropriate signifier of the signified 'social distance'. H e n c e , as an example, the so-called past tense can serve as a m a r k e r of social distance - the sign of dis tance in time serving as a n appropriate signifier for social distance: the so-called polite formulations, in English: 'I wanted to ask you, could y o u . . . ' (compared to T want to ask you, can y o u . . . ' ) . With this conception of the sign the first question which arises is about the interest of the user of this sign. W h a t is it that he or she is taking as criterial, and in what process o r situation? O t h e r signifiers of distance are available, of course: spatial distance (proximal versus distal) - this rather than that, conceptual distance (hypothesis versus factuality) - If it were... rather than It is . . .; ontological distance (possibility versus certainty) - possibly rather than certainly, etc. Take as a real example the following contrasting pair of textual extracts. Each is the abstract preceding an article in a scientific journal, in one case the presri gious disciplinary journal Science, in the other case the popularizing journal Scientific American. T h e articles b o t h report the s a m e set of findings (from Myers, 1990. (1) Experiments show that Heliconius butterflies are less likely to oviposit oa host plants that possess eggs o r egg-like structures. T h e egg niimics are

I

KRESS


343

a n unambiguous example of a plant trait evolved in response to a hostrestricted group of insect herbivores. (From Science) (2) Heliconius butterflies lay their eggs only on Passiflora vines. I n defense the vines seem to have evolved fake eggs that m a k e it look to the butterflies as if eggs h a v e already b e e n laid o n them. (From Scientific American) Given the hypothesis that signs are constructed out of the interest of a specific individual we are forced to ask about that interest i n each case. W h a t is it that changes the interest of the writer i n the two instances, so that in one case we have o n e characteristic set of verbs, a n d of tense/modality forms (show, are, are, has) a n d in the other case another (lay, seem to h a v e evolved, m a k e it look, as if... laid). I n a traditional account the difference would b e noted, and assigned, either in a correlational fashion, or simply as a matter of description, to a (formal and less formal) 'scientific style'. I n C L and C D A the m o d a l effect would b e noted, though the present account is n e e d e d to provide the link between social location of the p r o d u c e r of the form, a n d the form itself. Signs can b e of any size: a word; a syntactic unit; a clause; a text consisting of many sentences. So, for instance, the choice of o n e specific textual genre rather than another - an informal note rather than an official m e m o r a n d u m ; an informal chat versus a n official on-the-record exchange - constitutes the production of a sign, just as does the example above. I n this instance too, the producer's interest, as a reflection of a present socal positioning, leads h i m or her to focus on particular aspects of the object to b e represented (the n e e d for formality), which becomes the signified, and to select a specific textual form as a n appropriate signifier.

3 . T h e P r o d u c t i o n a n d R e c e p t i o n of Signs The choice of a generic form of text m a y b e thought to b e n o t an instance of the production of a sign, b u t rather a n instance merely of the reproduction of a sign, because (in most instances) the use of a generic text type seems to b e simply that: the use of a pre-existing text type, involving n o n e w production. This is cer tainly the common-sense, naturalized view. However, until I as speaker or as writer have actually m a d e the decision to use this text type, n o specific sign exists. There is nothing that fully predicts m y production of a sign. Nor is the textual instantiation of any o n e generic type ever like that of a n y previous instantiation, even in conditions of great external constraint. Even in a situation of great con straint and awareness of convention, I can act unpredictably, assessing in a par ticular instance the consequences of m y action differendy to what would normally be predicted. I can speak formally to m y partner over the dinner table w h e n rhere seems n o clear, contextually plausible reason for doing so; a n d I can speak mformally to m y superior in a formal situation, p r e p a r e d to accept the conse quences in b o t h cases. But even m y decision to act in conformity with a n understood convention is the result of an act of choice, a n d as such is m y n e w production of the m e a n i n g of conformity. That, after all, is the source of the im mense social, ideological and psychic p o w e r of that a c t Consequentiy the notion

344

DISCOURSE STUDIES

of the re-using of the sign, which is the essential underpinning of the idea of the stable sign, is not tenable even in those instances which seem most exemplary. H o w e v e r , the m o r e significant point is to realize that in all instances the char acteristics of the signified are n e v e r entirely knowable or predictable, given the producer's interest - as a p r o d u c t b o t h of m o m e n t a r y a n d of m u c h m o r e per sistent factors. Let m e give another example. Crawling along in a traffic j a m o n the motor way, I notice next to us a car towing a speedboat o n a trailer. T h e b o a t is called 'Weekend Warrior'. T h e two signifiers Weekend and Warrior axe not new, of course; a n d quite probably, the conjoined signifier Weekend Warrior has b e e n used before, even in British contexts (ignoring its colloquial U S usage), so that it m a y well be an existing sign. I could therefore treat the use of this n a m e for the speedboat as the re-use of a sign, a cliche; a n d from that I could m a k e m y predictable judge ments about our temporary neighbour o n the motorway. H o w e v e r , I cannot k n o w the signified that h e h a d constructed - what mixtures of age a n d gender values, what notions of self, p e r h a p s even degrees of irony. I n other words, it is m u c h m o r e plausible to see this as the production of a n e w sign, a newly m a d e conjunction of a n existing sign/signifier ('Weekend Warrior') with a n e w signi fied - our m o t o r w a y neighbour's conception of himself. A n d I would certainly h a v e to a b a n d o n a n y notion of this merely as cliche if I discovered that his wife, w h o on this occasion was a passenger in the car, was driving the speedboat for all of the weekend. N o r does this begin to touch o n my remaking of this sign in m y reading of it; that is, a treatment of signs in a communicative framework. This view contrasts sharply, o n the o n e h a n d , with a view that sees an indi vidual's m e a n i n g pushed, squeezed, forced into the 'straightjacket' or the 'mould' of existing word-signs; or, o n the other h a n d , with a view which sees the indi vidual's m e a n i n g as already a c c o m m o d a t e d to existing signs, either in the trad itional a n d persistent view of the speaker's re-use m e r e l y of existing signs, or in an alternative view, language structuring thought, in effect the Whorfian view of language. T h e notion of sign which I a m putting forward m a k e s it possible to connect the specificities of semiotic forms, in any m e d i u m , with the specificities of social organizations and social histories, via the actions of social individuals in the production of signs. It permits the m o v e from the actions of individuals as social agents, with their interests, to the actions of the group of which individuals are m e m b e r s , with that group's interest, in the fullest sense. This accounts at the same time for the p a r a d o x of group interest a n d values, a n d divergences and differences within group interests a n d values. T h e history of any semiotic system, for instance the history of a language, is thus accounted for b y the histories of social groups through the actions of their m e m b e r s in constandy making and remaking the semiotic system. It also explains the p a r a d o x of the appearance of the systematicity of the semiotic system, a characteristic which always recedes a n d dissolves w h e n it is subjected to analytic probing. A 'language' can b e seen to b e the working fiction constructed by, and in the interest of, a larger social group, itself consisting of multiplicities of groups: a fiction insofar as its i m p u t e d systematicity dissolves wherever searching analysis is applied; yet a working fiction, in the sense that

KRESS


345

large 'language communities' d o w o r k with such fictions, a n d h a v e to d o so for the achievement of particular social ends, a n d in the sense that such fictions d o work to a certain extent, in very m a n y practical domains. This view of signs also permits the m o v e from t h e micro-histories of semiotic events to the macro-history of semiotic systems. T h e history of a language, for instance, m u s t b e accounted for, can only b e accounted for, t h r o u g h the m y r i a d of small-scale histories which are the records of linguistic events. It is through the two related notions of interest in the production of the sign a n d the motivated relation of signifier a n d signified that (critical) readings are m a d e possible. T h e former b o t h allows a n d obliges (critical) readers to seek to hypothesize about the producer's actions, situation, intentions; the latter m a k e s it formally possible to do so. Without either notion critical reading would b e impossible, and reading itself would b e seen as an activity of the same kind as the looking u p of the code h a n d b o o k to decipher the message in front of the reader. Nevertheless there is a n element absent from m y account of the production and particularly the r e p r o d u c t i o n / r e a d i n g of the sign. If signs are motivated conjuncts of signifiers a n d signifieds, h o w d o they c o m e to b e opaque? For after all, if t h e relation is motivated, w h y is it n o t transparent? M y answer is founded on the notion of interest, as before. I n the case of the child's drawing it was the ac cident of the parent's presence at the n a m i n g of the drawing which enabled the reconstruction of the child's interest a n d so m a d e the sign relatively transparent: the reader, at least hypothetically, was p r o v i d e d with sufficient information necessary for that. Without that accidental presence, neither interest n o r motiv ation would b e easily recoverable: a n y reading, critical or ordinary, would b e more difficult, or impossible in those circumstances. A 3V2-year-old child's pos ition in the world is, simply, very different from that of a n adult. In quite that w a y d o the interests of fully social individuals differ from each other. A t times, of course, that difference m a y n o t b e very m a r k e d . A n d o n very many m o r e occasions, our possibilities of attending to potential discrepancies in interest are severely curtailed, b y the multitudinous pressures which surround every act of c o m m u n i c a t i o n and which curtail attempts at a critical reading. At tention to the interest of the p r o d u c e r of a sign d e m a n d s the hypothetical re construction b y the reader of the conditions of production of the sign; given the complexity of a n y message, that task is impossible other t h a n in extremely un usual circumstances. This factor alone, that is, the complete usuality of difference of interest, and the impossibility of attempting to recover the conditions of pro duction of the sign in each instance, is a factor that pushes 'ordinary readers' into common-sense readings - w h e r e the category of 'ordinary r e a d e r ' is not defined b y a n individual's qualification as a r e a d e r - t h o u g h that is a significant factor - b u t b y the fact that the reading is d o n e in ordinary circumstances. Not only are the interests of social individuals different, b u t also those of so cial groups, a difference ranging from the m e r e l y other to the directly oppos itional, antagonistic a n d conflictual. A sign p r o d u c e d b y an individual m e m b e r of one g r o u p within the interest of that g r o u p is thus most likely to b e o p a q u e in its constitution to a m e m b e r of another group. Differences in interest are likely to be lesser a m o n g m e m b e r s of a particular social g r o u p t h a n across different groups, especially in relation to those matters which are constitutive of that group.

34fi

DISCOURSE STUDIES

I n these conditions too the motivations for the production of the sign are likely to b e difficult or impossible to recover easily, so that the most likely reading is one which simultaneously treats the sign as stable and known, and, the condi tions of its production b e i n g u n k n o w n , as an arbitrary construct. O f course, where interests are oppositional a n d conflictual, there is m o r e likely to b e a contestation about the characteristics of the sign. It is at this point that it becomes essential to m a k e overt what I h a v e so far kept implicit, n a m e l y that any understanding of the production a n d reception of signs, that is, a theory of semiosis, has to b e set in the context of a social the ory of communication in which p o w e r is an inevitable component. (It m a y seem tautologous to use the t e r m 'social' with 'theory of communication', and I do so as a rhetorical m e a n s to draw attention to the fact that so m a n y theories of com munication are cognitive/psychological i n their constitution, treating the com municating individuals as asocial atoms.) W h i l e it is quite likely that the child p r o d u c e r of the car/sign p r o d u c e d that sign without any notion of a n audience in mind, either specific or general, m o r e m a t u r e makers of signs d o h a v e an idea of an audience, not, of course, necessarily consciously. A n d most events of com munication take place in relatively well-established social structures, which bring together n o t only individuals in events of communication, b u t also social groups in structures of interaction which h a v e relative stability a n d persistence. I n other words, in any society most communication takes place in pre-existing networks which b r i n g together b o t h the individual participants a n d the groups of which they are members in relatively stable configurations. This ensures a relatively welldeveloped understanding of w h o the audience is, a n d it also ensures a relatively well-developed awareness of interests at issue in that relation. T h e presence of p o w e r introduces a factor into the production of signs not unlike, in its effects, that of the childish producer. While the child/producer does not attend to his or h e r audience because there appears n o possible area of difficulty in that the child simply a n d naively expresses his or h e r interest, the powerful p r o d u c e r of the sign can act in a similar fashion motivated n o t by childish naivety, b u t b y the wilfulness of power. This can lead to its o w n naivety because ease of communication is n o t his or h e r p r i m a r y consideration, or even a consideration at all. I n other words, with an increase in the power of one par ticipant in relation to that of his or h e r interlocutor's comes the possibility of forcing t h e m to accept m o r e a n d m o r e esoterically constructed signs. Whereas the d e m a n d s of'good', efficient communication suggest that the p r o d u c e r of the sign m a k e s the interest which is c o d e d in the sign relation as transparent as possible for the addressee, the presence of p o w e r allows h e r or h i m to override that consideration. This is n o t a question of bias - there is n o m e a n s of producing an unbiased sign, there are only m e a n s of p r o d u c i n g signs in which the char acteristics of the object to b e represented are coded as transparentiy as possible for the envisaged audience. It is nevertheless a question of p o w e r producing 'distortion', in the sense of an increase in the opacity of the sign, with a consequent increase in the degree of p o w e r b r o u g h t to bear, a n d expressed. T h e actual inhprpnt n n w e r of the individuals involved in the event of communication is

KRESS


347

not a telling issue - powerful individuals d o n o t n e e d to use or display their power - rather it is the degree of power actually b r o u g h t to b e a r o n a particular occasion which produces opacity. W h e r e group interests are at issue, the same factors obtain; w h e r e o n e group is actively seeking to control the other, this b e c o m e s a strong part of the p o tentially controlling group's interest. T h e signs constructed b y the powerful group m a y take at least two forms: as transparent as possible, o n t h e grounds of a wish to maintain the appearance of relations of solidarity with another group within an established structure of interaction; or opaque, o n the grounds that the degree of p o w e r possessed b y the makers of signs allows t h e m to ignore requirements for ease of understanding a n d to enable t h e m to force the interlocutor into attempts either to recover the motivation that p r o d u c e d the sign, or to accept it at its p r o posed value. Of course, the m a n n e r in which I a m setting out these possibilities is a n oversimplification. First, the situation is one of a scale, along which there is a vast range of possibilities. Secondly, interlocutors are perfectiy capable of resist ant, oppositional a n d subversive reading, tugging a forelock while ridiculing the boss. H e n c e the question is n o t 'is this sign o p a q u e , or transparent?' or 'To what degree is it either o p a q u e or transparent?' All signs are always transparent to the m a k e r s of the sign; a n d all signs are always o p a q u e to s o m e degree for the readers of signs. T h e question is: ' H o w o p a q u e is this sign for this specific reader, and w h y ? ' All signs are motivated in their relation of signifier to signified, and all signs are always transparent to their makers, and all signs are m o r e or less opaque to readers. Motivatedness of signs, a n d opacity and transparency, are distinct con siderations. T h e former relates to the conditions of production of the sign; the latter relates to the uses of signs in structures of communication affected b y dif ferential distributions of power: to the conditions of reception. Ideology can n o w b e seen to h a v e at least two polar forms, i n terms of the production of the sign out of a producer's interest. At the one extreme, a pro ducer's position in the world will m a k e h i m or h e r p r o d u c e signs which arise out of that position a n d reflect his or h e r interest in relation to particular events and objects. This will p r o d u c e a set of signs reflecting one k i n d of interest. O f course, a producer's position i n the social a n d cultural world is always a position which includes multiple locations, m a n y of which will b e at o d d s with others, either simply different, or different in the sense of contrast or opposition. O n e part of b e c o m i n g a social subject is to p r o d u c e meta-signs for the self which pro vide reconcihations, temporarily or m o r e permanently, of such differences, which provide integration for the individual. These sets of signs, including the metasigns, p r o d u c e a particular account of the world, a n individual's ideology. At the other extreme, a sign-maker's position in the world vis-a-vis specific others, individuals or groups, m a y lead h e r or h i m to p r o d u c e sets of signs which are reflections of their interest in the relation to the other individual or group. That is, the sign is p r o d u c e d to provide a particular sense of the relationship of individuals or groups - for instance, a set of signs which provides a n account of relations of domination. If the individual or group producing the sien or set of

DISCOURSE STUDIES

348

signs is successful in having that sense accepted b y the other individual or group, then a n ideological position in the traditional Marxist sense has b e e n produced. I n most instances of sign production b o t h kinds of interest are at issue, so that signs or sets of signs are nearly always ideological in b o t h senses, with particular degrees of each. T h a t is, signs are fundamentally contradictory, in the sense that they code the producer's interest per se, and their interest vis-a-vis an interlocutor in a specific situation of communication. I n Social Semiotics (Hodge a n d Kress, 1988) we dealt with this issue b y produc ing the term 'ideological complex'. T h e ideological complex is a structure which contains, at the same time, the representation of a state of affairs as it appears from the point of view of the producer's interest, a n d another representation of a state of affairs as it appears from the point of view of the interest of the p r o d u c e r vis-a-vis another individual or g r o u p . Text 1 S c h o o l b a n i s h e s b o y t o o p o o r to p a y for uniform b y I a n M a c G r e g o r (Education Correspondent) 1. A S C H O O L B O Y was taught in a class o n his o w n because his parents could n o t afford a uniform, it was revealed yesterday. 2. T h e 13-year-old was separated from classmates for several weeks, a c o m m u n i t y advice organisation said. 3. His misery only e n d e d w h e n friends found e n o u g h j u m b l e in a n Oxfam shop for a make-shift uniform. 4. 'It was the worst case of this kind w e h a v e seen', said Nichola Simpson, h e a d of policy at the National Association of Citizens' Advice Bureaux. 5. 'It was v e r y distressing for a b o y so young. 6. A n d h e missed out edu cationally. 7. Children should not suffer like that, especially because they cannot afford a uniform. 8. It was a very hard-hearted thing to d o ' . 9. T h e incident h a p p e n e d in the East Midlands earlier this year w h e n the b o y ' s family m o v e d into the area. 10. His parents were o n income support, b u t education chiefs told t h e m they did n o t qualify for a grant because their son m o v e d schools mid-term. 11. T h e case was revealed b y the association in a report o n uniforms. 12. It said m a n y p o o r children w e r e suffering at school because local bur eaucrats are n o t helping t h e m p a y for uniforms. 13. M a n y parents o n benefits or low incomes w e r e n o t receiving grants, which are discretionary. Discipline 14. T h e news comes as m a n y schools return to uniforms in a bid to make pupils m o r e presentable a n d disciplined. 15. But the N A C A B found local authority provision for uniform and clothing grants was tending to 'wither away'. 16. Assistant director Peter Hildrew a d d e d that grants were n o w a low prior ity for m a n y education authorities.

KRESS


349

17. 'Parents are penalised according to w h e r e they live', h e said. 18. T h e system is often arbitrary a n d seems close to collapse. {Daily Express, Friday, 3 M a y 1991) Bearing i n m i n d several of the points m a d e just above, I shall briefly analyse this text. M y interest h e r e lies initially i n t h e construction of the concept of poverty in t h e popular m e d i a (see Meinhoff a n d Richardson, forthcoming); a n d m y specific focus is t h e clause 'his parents could n o t afford a uniform' i n sentence 1. Given t h e theory of text briefly sketched earlier, m y view is that the text itself is a large, complex sign, in which t h e interests of a particular p r o d u c e r are real ized. This m e a n s that these interests will b e realized in other parts of this text. Particular points to focus on are t h e headline, sentences 3 , 7, 8, 10, 12, 13, 15, 16, 17 a n d 18. T h e cited clause i n sentence 1 presents a syntactic ambiguity. W h a t syntactic analysis/description d o w e give to afford? O v e r d y it looks to b e a transitive v e r b , with a uniform as direct object, a n d hisparents as subject. A syntactically analogous form to that reading of this clause is 'His parents (could not) b u y a uniform', which is clearly transitive. H o w e v e r , 'affording a uniform' is n o t clearly transitive; its subject n o u n is n o t clearly agentive: Hisparentsis n o t a n unambivalentiy agentive subject. Afford is a state of being, n o t a process u n d e r the control of an agent. A n d clauses which are not clearly transitive d o n o t passivize easily: 'a uniform was (not) afforded b y his parents' (and similarly with further tests, such as prenominalizing of the adjective, ' T h e afforded u n i f o r m . . . ' ) . Yet m a n y readers of the Daily Express m a y r e a d across this clause in reading t h e text, a n d read it as fully semantically transitive, which I shall call, following H o d g e a n d Kress (1993), a 'transactive'. I n support of that reading, these readers might say: ' W e scrimped a n d saved, a n d we afforded a uniform for our children, so w h y can't they?!' T h e r e are thus at least two syntactic readings for this clause, very likely cor responding to t h e readings of different audiences - a hypothesis which could b e tested. O t h e r forms which line u p with the non-transitive, non-transactive reading are ' t o o poor to p a y . . . ' , (3) 'friends found . . . j u m b l e ' , (7) ' C h i l d r e n should n o t suffer...'; a n d b y a further extension, all those clauses which are n o t transactive, for instance (3) 'His misery . . . e n d e d w h e n . . . ' ; (6) ' h e missed o u t education ally'; (9) ' T h e incident h a p p e n e d in t h e East M i d l a n d s . . . w h e n the boy's family m o v e d into t h e area'; etc. Forms which align with t h e transitive, transactive reading are 'too p o o r to pay (money) f o r . . . ' , (3) 'friends f o u n d . . . j u m b l e ' , (2) 'the 13-year-old was sep arated (by someone) from classmates', (10) 'their son m o v e d schools', etc. I n other words, m a n y of the forms which can b e r e a d transactively can also have a non-transactive reading. This serves, or can serve t h e producer's interest. A sign which can b e r e a d in both ways - a syntactic form which can b e read as assigning agency to subjects, or which can b e r e a d as simply 'involving' sub jects - m a y b e what best reflects the producer's interest: o n t h e o n e h a n d , it but tresses a politically reactionary view which wishes to h o l d t h e p o o r responsible for their o w n poverty; o n t h e other h a n d it serves a n interest n o t t o alienate those readers w h o m a y b e poor, or w h o h a v e a divergent account of the causes of poverty.

350

DISCOURSE STUDIES

It is important to see that there is a whole pattern of these ambiguous forms (and of clauses with a relational syntax, which tend towards a stative reading of the linguistic-social world). T h e pattern sets u p , facilitates m o d e s of read ing. Most of the clauses in this net-like pattern involve the social participants w h o are implied a n d implicated - entailed - in m y 'focal' clause - 'his parents could n o t afford a uniform': the b o y , the parents, the uniform, the school, the school authorities, the local education authorities, bureaucrats. All of these have a complex place in the t h e n a n d still current educational a n d social politics of the British Conservative government. This has, broadly, entailed - apart from far m o r e widely ramified politics - a m o v e to destroy the place of local education authorities (which are seen as obstacles to t h e imposition of centralizing control) in order to gain control of schools direcdy, a n d centrally. Part of the strategy of b o t h g o v e r n m e n t a n d the right-wing press has b e e n a n attempt to undermine the credibility n o t only of local governmental structures, b u t also of all forms of social organization which m a y provide resistance to the 'modernizing' drive of the Conservative state; such as, for instance, 'bureaucracy', seen itself as a leftover from a 'socialist' version of a centralized state (the ' N a n n y State'), and an obstacle to Conservative m o v e s aimed at the renovation of the state. Linguistically, semiotically, this leads to the presence of the clearly, unam biguously transactive clauses 'School banishes b o y ' (headline), (1) 'A schoolboy was t a u g h t . . .', (2) ' T h e 13-year-old was separated . . .' (both having the im personal, 'faceless' agent of the passive deleted), (8) 'It was a v e r y hard-hearted thing to d o ' (where the e m b e d d e d clause has a deleted agent 'Someone did a hard-hearted thing') ( 1 2 ) ' . . . local bureaucrats are not helping t h e m . . . ' (when the real culprits are identified), etc. I n this analysis the agents of the transactive clauses are the m e m b e r s of the demonology of the Conservative right: local education authorities, local councils, schools, teachers, bureaucrats. It is they w h o act oppressively towards the hapless, helpless, powerless individual, b o t h parents a n d child. T h e syntax thus clearly represents one facet of Conservative discourse. T h e r e are, of course, other facets, a n d these are b y n o m e a n s complementary. So the transactive reading of 'his parents could n o t afford' is a poHtically Conservative reading, which assigns re sponsibility for 'their' poverty to the poor. A n d in other texts, schools are rep resented as sites of resistance to (socialist) local council tyranny, through tht agency of parent-power. Of course, this kind of ' c o m m o n sense' in the Gramscian sense is exactiy what we should expect. At the s a m e time this text provides, as I h a v e mentioned, another reading the non-transactive, which allows a space for those readers w h o have an op positional view of poverty. T h e text is thus the site of at least two discourses on poverty: the Conservative, which assigns responsibility to the poor and to a malign set of social agents; a n d the oppositional (in this context), which represents po verty as a n event in which participants are caught u p . M y account provides a description a n d explanation of this seemingly con fused or contradictory syntactic pattern in terms of the production both of the sign of the clause, the sign of the sentence, b u t m o r e , of the complex sign of the overall i text, in terms of the interests of the producer of the text, including his interest in J

KRESS


351

the perceptions of the text in terms of the receiver's interest. T h e complexities of that possible reception, which includes divergent audiences with divergent interests, are coded in the complex sign of the text, a n d in the smaller level signs of clause a n d sentence. T h e micro-history of the production of this text/sign is set in the context of the macro-history of British society at this time, a n d the latter provides a n account of the former. However, the text as sign also encodes a n d produces as a n object a sign which embodies that social state, a n d in doing so b o t h naturalizes a n d deproblematizes that state of affairs, ana produces a sign which in itself projects and produces n e w signs which r e p r o d u c e its meanings. This text also illustrates the operation of p o w e r in producing opacity of signs, at all levels. To m a k e that point succinctly, imagine the instructions one might give to a student writing this report, encouraging h e r or h i m to b e clearer, or to write m o r e clearly. M o r e can b e said about this, in particular in relation to m y subsequent dis cussion of the boundaries of the sign. This text/sign occurs o n the right half of a page, which o n its left has a largely pictorial 'story', u n d e r the rubric 'photonews' of a m e m b e r of the royal family cuddling, caring for, a child from a fosterfamily which has dealt with child victims of child abuse. It is h a r d to think h o w the reading of the o n e text can n o t affect the other. In relation also to a discussion on the stability of signs, it is relevant to consider the effect of this text o n the semiotic system of language, and, in particular, the system of transitivity or valency of verbs. T h e production of signs which are con structed to enable differentiated readings b y two distinct groups o n the one h a n d Introduces a systematic instability into the system a n d to terms in the system in as far as w e wish to maintain that theoretical construct - and, o n the other hand, facilitates a n d entrenches particular readings of verb valencies for particular readers; that is, it shifts the system for particular readers in specific directions. 1

From the producer's o w n interest, the 'poverty text' is to b e r e a d with the adjacent visual text o n the care for abused children. F r o m within the producers' interest vis-a-vis a group of potentially sceptical or oppositional readers of the •poverty' text the visual 'caring' text is to b e taken as merely co-present, a n d the features of the layout, which bring t h e m together in this m a n n e r , are to b e read as an accidental factor. I n m y experience these are the readings given to the text b y the readers whose readings I h a v e tested. In any interaction involving power, signs will therefore b e ideological or produce ideology, in the two senses, either as 'world view', or as 'mystification', encapsulated in M a r x ' s m e t a p h o r of the c a m e r a obscura. C o m m o n to b o t h is the effect p r o d u c e d b y the wilfulness of the powerful, whose p o w e r can b e real ized, among other things, in their indifference to the n e e d for the best obtainable conditions of communication in a n y particular situation. This is a factor which makes their communication as o p a q u e as that of children, for reasons which are in part the same - in the case of the child, 'indifference' is an accidental com ponent of the child's interest - a n d in p a r t entirely different, n a m e l y the causes of the ^difference in the case of the powerful, that is, the wilful obliteration of the needs of addressees.

DISCOURSE STUDIES

352

4. T h e B o u n d a r i e s o f Signs Text 2 Consider the text from t h e Sunday Telegraph (16 J u n e 1991). W h a t determines the boundaries of texts? D o w e h a v e h e r e two texts, or o n e . O n the face of it, the answer seems clear: there are two texts, o n e which I shall call the 'Bullying' text, a n d another which I call the 'Correctness' text. T h e relation between t h e m is, it seems, determined b y the (accidental) contingencies and exigencies of layoutit just h a p p e n e d that the sub-editor responsible for laying out that p a g e n e e d e d a small bit of text to fill a gap, a n d so stuck the 'Correctness' text where it fitted, into the 'Bullying' text.

Bullying is beyond a joke for victims SARAH, a pretty and intelli gent M-year-old, was twice forcibly stripped to the waist by a gang of 10 boys and girls at her school. After a failed suicide attempt, she is now thriving at a new school. Sarah (not her real name) was photographed and punched and had her hair pulled during the attack in which she was forced to beg on her bands and knees for her clothes back. The head at the Edinburgh school where her ordeal took place dis missed it as "horseplay". Her story is told in a book on bullying and disruptive behaviour in schools, which sets out a 10-point plan that teachers can use to eliminate a menace which has long been a feature of school life. . The book attacks the Brit ish education system for fail ing to train teachers to cope with disruptive behaviour. It says they arc "hopelessly illequipped" to deal with the growing problem of bullying. its author, Michelle Elliott, is director of the children's safety campaign Kidscape. In Bullying

to coping

— a practical

(or schools

guide

— pub

lished tomorrow by Longman at £10-95 — Elliott and other experts offer what is probably the first comprehensive com pendium of advice. They also tell some horrific tales of recent cases including those of: Q "Mark", 9,-whose arm was

on her desk or a sanitary towel stuck to her blazer. In the school dinner queue broken by a gang who set on girls screamed obscene him as he walked home from rhymes and names at her — school in Essex. such as "rat-faced lesbian" — D "Tony", 7, who was set on and Sandra, whose father has fire by a child of the same age. a history of violence against D "Sandra" , 14, who was her mother, recalls: "I think driven to truancy and eventu the teachers on dinner duty ally contacted Esther Rant- thought it was just a joke," zen's ChildLine for help. The book advises teachers The girl had suffered weeks to set up "courts" in which of physical attacks and cruel bullies should be arraigned by okes, such as a pack of the whole class. Dut Mrs ictchup-soakcd tampons left Elliott warns; "Dully courts may not be the right answer for every school. One head• ALMOST all the members teacher saw the courts as just of a group set up by the another form of bullying." Government to advise on the She also encourages teach leaching of English «?pose ers to take a tougher line with the Idea of "correctness'" In the classic excuse: "It was the language. only a joke. Miss" — SO why Twenty-one of the group's wasn't everyone laughing? 25 members belong to a pres She Is concerned about the sure group representing a inability of British teachers to highly ideological minority of cope with bullies. "When I teachers. was trained 25 years ago in One expert outside the the US," she says, "wc had a roup nid: "A lot of them course in dealing with the elieve teaching children disruptive child. irammar and Standard Eng. "Teachers' in this rnuntrv ish is oppression of the work, Icdfft virtually notinng. They inn tlassî ney aim io reshape get ov' or training and arc fcflgiisn on multicultural, suddenly confronted with a anti-racist, anti-sexist lines." group of 10-year-old thugs Tne dtniytufn ~tn tlie and don't know what to do." National Curriculum project, SheCTiti>i«-—

DISCOURSE STUDIES

382

I n the case of Narrative 1, we find m o r e than one evaluation section. Narra five 1 begins with a long orientation section of 10 clauses. (1) a 0

30

ib c

w

2

3

28

d7

But he had a young wife

e f g h

and those days I dressed well. And seemingly, she was trying to make me. I never noticed it. Fact is, I didn't like her very well, because she had she was a nice looking girl until you saw her feet. She had big feet. Jesus, God, she had big feet! Then she left a note one day she was going to commit suicide because he was always raising hell about me. He came to my hotel. Nice big blue 44 too. I talked him out of it, and says, "Well, we'll go look for her, and if we can't find her, well, you can - go ahead, pull the trigger if you want to." I was maneuvering. So he took me up on it. And we went to where they found her handkerchief the edge of a creek And we followed down a little more, And we couldn't find anything. And got back it was a tent show she was laying on a cot with an ice bag on her head. She hadn't committed suicide. But - however - that settled it for the day. But that night the manager, Floyd Adams, said, "You better pack up and get out, because that son of a bitch never forgives anything once he gets it in his head." And I did. I packed up and got out. That was two.

2

4

5

6

7

26

2s

24

23

i ji ko

8

22

2

9

10

1 m in 2p, 0

0

0

3

2

3

0

P q r

0

0

0

0

oSo to oU] O

2

iV

9

]W, ix iy Zo 2

7

0

0

aao

O

bb, cc ddo ee

0

(

0

0

0

0

Well, in the business I was associated at that time, the Doc was an old man . . . He had killed one man, or - had done time.

9

T h e first narrative unit is k ("Then she left a note o n e day . . . " ) , followed by 1 ("He c a m e to m y hotel") a n d m ("I talked h i m out of it"). We then have two clauses coordinated with m - clauses n ("And says") a n d o ("I was maneuvering."). These multicoordinate clauses suspend the action at a critical moment - when the danger of death is greatest, a n d they contain a n explicit statement of the attitude of the narrator. His coolness in a m o m e n t of crisis emphasizes the danger

anrl

rap]arte

urall nn nimcolf

1AB0V AND WALETZKY

NARRATIVE ANALYSIS

383

Five narrative clauses follow this suspension, resolving the crisis introduced by 1 and m. A second evaluation section occurs at a subsidiary point w h e n the situation is further resolved - the fate of the lady in question is determined, a n d simultaneously the immediate threat to the narrator. T h e action is suspended at this point by the use of a free clause that might h a v e occurred in the orientation section, v ("it was a tent show"), a n d a direct comment, x, that might h a v e b e e n inferred from w. T h e resolution is stated with some finality in y ("that settled it"). Finally, there is an a d d e d explicit evaluation of a third party that confirms the im plications of the previous evaluation section, followed b y a conclusion. T h e overall diagram shows h o w evaluation sections outline the structure of the narrative. It should b e apparent h e r e that the evaluation sections are responsible for those deviations from the order of the p r i m a r y sequence of the narrative that complicate the a-then-b relation of narrative. T h e functions of the evaluation sec tion must b e added to the p r i m a r y narrative function in order to understand how the primary sequence is transformed into the m o r e complex structure that we see here. All of the evaluation sections shown h e r e are related to the originating function of the narrative. F r o m a structural point of view, the first section is the major break in the complicating action. Not all evaluation sections have the structural feature of suspending the com plicating action, as shown in the Figure 4. I n m a n y cases, the evaluation m a y b e present as lexical or phrasal modification of a narrative clause, or it m a y b e itself a narrative clause or coincide with the last narrative clause. For this reason, the fundamental definition of evaluation must b e semantic, although its implications are structural. 10

Oause.

3.

a

0

18

b

1

17

2

16

d

3

15

s

4

14

f

5

13

g h i

6

0

0 0 1

0 1 0

j k

0

1

1

1

0

m

0

3

n

13

5

0

6

3

P

7

2

q r

0

0

0

0

s

0

0

YJ 0

0

W = Complication © = Evaluation r~~\ = Resolution

T h e evaluation of a narrative is defined b y us as that part of the narrative thai reveals the attitude of the narrator towards the narrative b y emphasizing the relative importance of some narrative units as c o m p a r e d to others. This may be d o n e b y a variety of m e a n s : Semantically defined evaluation: 1. direct statement: "I said to myself: this is it." 2. lexical intensifiers: " H e was beat u p real, real b a d . " "I w h u p p e d that d u d e half to death." Formally defined: 3. suspension of the action: a. t h r o u g h coordinate clauses a n d restricted clauses: See Figure 1 b . repetition (subtype of the above): See Narrative 2, at the moment of crisis w h e n the d o g is gone for the 3rd time: "And h e didn't come back. A n d h e didn't c o m e back." Culturally defined: 4. symbolic action: " T h e y p u t an egg o n his door." "I crossed myself." "You could h e a r the rosaries clicking." 5. j u d g m e n t of a third person: h e r e the entire narrative is reported to a per son n o t present at the narrative. Narrative 12 is a heavily evaluated narrative that shows three of these char acteristic forms of evaluation. It is typical of m a n y fight narratives in its two-pan Structure. T h e first subcycle deals with the events leading u p to the fight, and its conclusion is the beginning of the second subcycle, the fight itself. In this case, the evaluation of the first section is a statement of the narrator: do I told him that - it's impossible for him to find downtown, 'cause all those people were walking by, and just his father is the only one that find it?

0

Although the very length of this closely reasoned argument serves to suspend the action, the structural criteria we h a v e b e e n using show it as a single narrative clause. We identify this clause as an evaluation o n semantic grounds: It is an ex plicit statement b y the narrator of his attitude towards the situation. T h e conclusion of Narrative 12 is also a n evaluative statement that coincides, with the last narrative unit: T h e statement of a third person after the entire se quence of events is reported to h i m . ojo Then he start crying ok, and run home to his father. 1 And his father told him, he ain't find no glove 0

0

I n addition, w e have the evaluation of the act of clause i

LABOV A N D WALETZKY

„h i„ 0

0

NARRATIVE ANALYSIS

So he say he give And I kept on hitting him.

385

•

i

It is normal n o t to hit s o m e o n e after h e says "I give." This incident evaluates the narrative b y indicating that the anger of the narrator was so great - d u e to ex cessive a n d unreasonable provocation - that h e was carried away to the extent of violating this n o r m . T h e other b o y h a d placed himself outside of normal sanctions by his behavior. All of these forms of evaluation serve the function of self-aggrandizement, showing the narrator in a favorable position as c o m p a r e d to the other boy. It is evident that there are a great variety of evaluation types, m o r e or less deeply embedded in the narrative. But this variety should not obscure the fact that unevaluated narratives are exceptional as representations of personal experience, and unevaluated narratives lack structural definition. An important characteristic of narratives is the degree of e m b e d d i n g of the evaluation in the narrative framework. T h e r e is a wide range, from the most highly internalized type - a symbolic action or the evaluation of a third person to the most external - a direct statement of the narrator to the listener about his feelings at the time. I n the examples given previously, w e find internalized evalu ation in Narrative 1, in the dramatic statements of narrator a n d m a n a g e r ; a n d in Narrative 3, in the statement of the doctor, ("just about this m u c h m o r e , " h e says, "and y o u ' d a b e e n dead."). T h e last narrative, Narrative 14, has a dramatic statement of the narrator ("I say, 'Calvin, I ' m bust your h e a d for dat.'"). Sometimes the evaluation occurs in a statement of the narrator to himself, less well integrated into the narrative, as in Narrative 7: "So I says to myself, 'Well, there's g o n n a b e times . . . ' " . T h e other e n d of the scale is shown b y a c o m m e n t at the end of the narrative directed towards the listener, as in Narrative 13: "Just over two dollars that h e was sent for peaches with." Still m o r e direct is Narrative 2: "I'll tell y o u if I h a d ever walloped that d o g I ' d h a v e felt some bad." We might construct a scale of degrees of e m b e d d i n g of evaluation, following examples of the following sort: Internal

1. And when we got down there, her brother turned to me and whispered, "I think she's dead, John!" 2. And when we got down there, I said to myself, "My God, she's dead!" 3. And when we got down there, I thought, "She's dead." 4. And when we got down there, I thought she was dead. 5. Later, the doctors told us she was close to death. 6. I think she must have been close to death. External 7. You know, in cases like this, it's clear that she was likely as not dead. Resolution With this definition of evaluation, we can n o w return to the p r o b l e m of defining the result of a narrative. T h e p r o b l e m is n o w quite simple. We can establish the break between the complicating and resolving action hv I n c o H - " 1

DISCOURSE STUDIES

386

of the evaluation. Thus, the resolution of the narrative is that portion of the narra tive sequence that follows the evaluation. If the evaluation is the last element, then the resolution section coincides with the evaluation. I n the examples given pre viously, the complicating clauses are symbolized u a n d the resolving clauses n . Coda M a n y narratives end with a resolution section, b u t others have an additional element that we m a y call the coda. T h e actual sequence of events described in the narrative does not, as a rule, extend u p to the present. T h e coda is a functional device for returning the verbal perspective to the present m o m e n t . This is accomplished b y a variety of means., so that the codas cannot b e identified b y such simple tag lines as "And they lived happily ever after." a. O n e device used in a coda is deixis. This is the linguistic category that points to a referent instead of n a m i n g it explicitiy: I n this case, it has the effect of stand ing at the present m o m e n t of time a n d pointing to the e n d of the narrative, iden tifying it as a T e m o t e point i n the past. (1) oaaj) I packed up obb and got out. o o That was two. 0

cc

(7) obb„

That was one of the most important

(8) oggi ihho

And that - that was it, you know. That was it.

(11) And that was that. This use of the obviate deictic category - that, there, those- contrasts sharply with the use of the proximate in the b o d y of the narrative - this, here, these. For example, w e h a v e the following proximate evaluation in Narrative 8: (8) bb and they was on me and cc and I said "like this is it, man." oddo I pulled a knife 0

0

o

0

b . A n o t h e r device used in codas is an incident in which one of the actors can b e followed u p to the present m o m e n t in actions that m a y not b e totally relevant to the narrative sequence: And you know that man who picked me out of the water? he's a detective in Union City,


NARRATIVE ANALYSIS

387

c. T h e effect of the narrative o n the narrator m a y b e e x t e n d e d t o the present moment: I was given the rest of the day off, and ever since then I haven't seen the guy, 'cause I quit. I quit, you know. No more problems. It is interesting to n o t e that all codas are separated from the resolution b y temporal juncture. At the same time, it seems that some semantic criterion is necessary to identify codas: T h e fact that they are frequendy n o t descriptions of events, or of events necessary to answer the question: "What h a p p e n e d ? " The overall structure of the narratives that we h a v e examined is n o t uniform; there are considerable differences in the degree of complexity, in the n u m b e r of structural elements present, a n d h o w various functions are carried out. However, a composite view of narrative performance leads us to posit a normal form for irai versions of personal experience; the degree to which any one narrative ap proximates this normal form is a significant fact about that narrative - p e r h a p s more significant than a n y other in terms of fulfilling the originating function of th» narrative. The normal form is quite distinct from the primary sequence of the narrative. L • noted above, the n e e d for an evaluation section motivates the transformation he primary sequence into the m o r e characteristic n o r m a l form that appears . he linear sequence presented b y the narrator. One can represent the n o r m a l form of narrative using the diagram in Figure 5. H re the originating function of the narrative is applied at the base of the mond; we proceed u p a n d to the left with the orientation section, t h e n u p to • apex with the complication. Frequently, but n o t always, the evaluation sus; ids the action at this apex, as represented b y the circle. T h e resolution proceeds wnward and to the right, a n d the coda is represented b y the line that returns : lie situation (point in time) at which the narrative was first elicited. T h e sim plest possible narrative would consist of the single line of the complication,

DISCOURSE STUDIE5

388

without a clear resolution; frequently w e find minimal narratives that have both complication and resolution ("He hit m e hard and I hit him back"). As we proceed to more complex narratives, told by speakers with greater overall verbal ability, w e find a higher percentage of narratives that duplicate the exact form of this diagram. Perhaps the most frequent variant is the case in which the evaluation ends the resolution: jokes, ghost stories, and surprise endings take this form, as the story is reshaped b y many retellings.

Conclusion This view of narrative structure helps us to answer the two questions raised af the beginning of this discussion. First, w e have related the sequence of narrative elements to the inferred sequence of events in the experience that is being recap itulated, through the definitions of narrative units, restricted clause, free clause, and narrative clause. Secondly, w e have outiined the principle elements of simple narratives that perform both referential and evaluative functions. We have shown that the evaluative function requires the transformation of the primary sequence, based on the a-then-b relation, into the more complex normal form of the nar rative as presented by the narrator. With this framework, w e are beginning to analyze relative effectiveness and completeness of narrative structure among various subgroups of our population, and, furthermore, to analyze the more complex types of narration developed by skilled storytellers and preserved by oral tradition. It is clear that these conclu sions are restricted to the speech communities that w e have examined. This view of narrative structure will achieve greater significance when materials from radically different cultures are studied in the same way.

Acknowledgment The work described in this article was supported in part by the U.S. Office of Education in connection with Cooperative Research Branch Project No. 3288, "A Study of Non Standard English of Negro and Puerto Rican Speakers in New York City," under the overall program of Project Literacy.

Notes This article was originally published in Essays on the Verbal and Visual Arts: Proceeding? -:/ the 1996Annual Spring Meeting of the American Ethnological Society (pp. 12-44), edited tv June Helm, Seattle: University of Washington Press. Copyright 1967 by University "f Washington Press. Reprinted with permission. 1. The materials include: 70 interviews with speakers from various occupations, ethn!-: membership, and ages on Martha's Vineyard, Massachusetts; 230 interviews wuh speakers representing a stratified random sample of the Lower East Side of New V i i. City; 250 interviews of children and adults from our current research in Central " - - * • » — n » w o frnm p v n l n r a t o r v work in Cleveland, Boston, Philadelph ia. ;

1

J

c

n


NARRATIVE ANALYSIS

389

Chicago, Phoenix, and Beaufort County, South Carolina. The basic interview techniques are described in Labov (1964,1966) and Labov, Cohen, and Robins (n.d.). 2. In our current research in Central Harlem, we are concerned with the functional conflicts between standard English and the nonstandard English of Black and Puerto Rican children. Many of these children show great verbal ability in many areas, in cluding the construction of narratives, but cannot read at all. One purpose of this work on narrative analysis is to show how children use language to carry out the functions that are important in their system of values. 3. As in, "We were running, walking, and then creeping down the road." We might better say that an apparent appositive turns out to be a coordinated clause. Coordinate verbs are always analyzed separately if they are independent, and in most cases in which they are subordinated to verbs of saying and telling. See Narrative 2, which follows: "I said, 'You git back there/and get that duck.'" If the narrator had cited himself as saying, "You get that duck and get back there," he would have been reversing the inferred sequence of events - in this case, two utterances. The same argu ment holds for the example in Narrative 1 (cited previously), clauses n and o. On the other hand, if someone says: "You try and get it," we cannot understand these as two independent verbs, but rather the use of "and" is equivalent to an infinitive em bedding, the same as "You try to get it." 4. As noted previously, the subordination of "get back there" and "get that duck" to "I said" is not the type of subordination that removes clauses from temporal sequence. We can consider this coordination: "I said, 'You git back there,' and I said, 'You get that duck.'" 5. If "was laying" is accepted as a narrative clause, it cannot have the basic grammatical meaning of "simultaneous" as stated by Diver (1963). It would rather differ from the simple past layby the feature "extended." The meaning of "simultaneous" can be sup ported by arguing that these clauses are equivalent to "When we got back, she was laying..." In other cases, Diver pointed out, the use of the past progressive may force a metaphorical interpretation "the action was so swift that it was as if it was simul taneous with the preceding," as in "I was on the masthead; the ship gave a lurch; I was falling through the air; I hit the water." These and other interpretations can be subjected to an increasing number of empirical tests through the analysis of narra tives such as the ones given here. 6. Diver (1963) showed this form in his narrative axis with the meaning of "present, before," and gives a constructed example: "All day the sun has warmed the Spanish steps " One can find such examples in literary works that use historical present sequences freely, perhaps, but they have not occurred in the material we have examined to date. 7. Here the usual adjustments in anaphoric reference have been made. It may be noted that this series of past perfect clauses is one answer to a difficult problem produced by a narrative of this type. The result would lose its surprising effect if these clauses were placed in narrative sequence with regular preterit verbs. By placing the three clauses well out of temporal sequence, it is more difficult for the listener to follow the explanation and surprise is achieved at the risk of a certain awkwardness and con fusion. Again, we find that even partial success signals the fact that the narrator of Narrative 2 is a practiced storyteller and has probably told this story many times. We do not take narratives of this type as primary data. 8. As noted previously, Narrative 2 has many formal features that set it aside from the others and identify it as the product of a practiced storyteller. One can point to the embedding of an essentially anonymous "other" in the complicating action, frequent if traditional metaphor, the trinle s u h r v r - l o »fj—'— " 1 r

DISCOURSE STUDIES

390

repetition, and also the determination shown by the narrator in introducing the story. The preliminary material illustrates how a narrator of this sort will get the topic of his favorite stories into the conversation despite the fact that the original stimulus was only marginally relevant The transition of the interview theme was accomplished solely by the subject, and the actual stimulus for the narrative was his own. Despite the fact that features appear in this narrative that are distinct from the simpler examples, a formal analysis of Narrative 2 is possible only after consideration of the simpler narratives or, at least, a formal analysis based on such functional con siderations as we have introduced. 9. The phrase "Nice big blue 44 too" might as well be considered a narrative clause, derived from "He had a nice big blue 44 too." However, the status of hadas the head of a narrative clause is still at issue, and it would be tendentious to use a deleted form as evidence. We have therefore been treating this phrase as subordinated to "He came to my hotel," equivalent to "with a nice big blue 44 too." 10. The three evaluation sections of Narrative 1 raise the possibility that we can analyze this narrative as consisting of three distinct subcycles: that it is a complex narrative consisting of three structural units. This article is limited to the consideration of sim ple narratives, and this possibility must be postponed to a later study of subcycles and complex narratives.

References Colby, B. (1966). Cultural patterns in narrative. Science, 151, 793-798. Diver, W. (1963). The chronological system of the English verb. Word, 19, 141-148. Labov, W. (1964). Phonological correlates of social stratification. American Anthropologist. 66{6, Pt. 2), 164-176. Labov, W. (1966). The social stratification of English in New York city. Washington. DC: Center for Applied Linguistics. Labov, W, Cohen, P., & Robins, C. (n.d.). Final report on Project 3091. Washington, D l U.S. Office of Education, Cooperative Research Branch. Labov, W., & Waletzky, J. (1967). Narrative analysis: Oral versions of personal experience. InJ. Helm (Ed.), Essays on the verbal and visual arts: Proceedings of the 1996 Annual Spring Meeting of the American Ethnological Society (pp. 12-44). Seattle: University of Washington Press. Propp, V. (1958). Morphology of the folk tale (Publication 10). Bloomington, IN: Research Center in Anthropology, Folklore and Linguistics. Schatzman, L., & Strauss, A. (1955). Social class and modes of communication. American Journal of Sociology, 60,329-338.

45 Gender, Discourse and Semiotics: The Politics of Parenthood Representations Michelle M. Lazar

Introduction 1

h e aim of this article is to examine the politics at work in representations of p a r e n t h o o d in a n a t i o n a l , g o v e r n m e n t advertising c a m p a i g n in Singapore. Although it is n o t u n c o m m o n these days to find egalitarian or symmetrical models of parenting in the media, this is not to say that the m o r e traditional, asymmetrical division of labour between w o m e n a n d m e n in this do main has disappeared. T h e asymmetry m a y not b e as blatant as it was decades ago, but as this study will show, it continues to thrive in sometimes less obvious and seemingly innocuous forms. T h e particular concern of this article is to exam ine the c o e x i s t e n c e of b o t h symmetrical a n d asymmetrical forms of represen tation of p a r e n t h o o d in the same ad or set of ads. T h e dynamics of the two will be analysed in terms of w h a t I call the Discourse of Egalitarian G e n d e r Relations and the Discourse of Conservative G e n d e r Relations, respectively.

i

Representational practices are deeply political i n at least three key ways. The first involves the question of relative representation and non-representation who is represented (or not) in what ways i n relation to the other? Secondly, there is the issue of m o d e s of representation - h o w is representation achieved textually? What linguistic a n d non-linguistic resources are deployed, a n d in what ways? rhe two points intersect with the third, which is a n interrogation of the p o w e r dynamics at work - what does the representation/non-representation tell us about the (changing or unchanging) contemporary balance of p o w e r between, in this case, w o m e n a n d m e n in the d o m a i n of p a r e n t h o o d ? W h o s e interests are served b y it, a n d whose suppressed? Part of the interrogation of interests, in this study, inevitably includes the stake the state has in these representational prac tices, and the support (or otherwise) for particular orderings of gender relations. Source: Discourse & Society vol. 11. no. 3 . smnn

m

w*-Ai\r\

DISCOURSE STUDIES

392

I n what follows, I will begin b y locating the present study in terms of the socio political context of Singapore, a n d also in relation to previous work that I have undertaken in this area. This is followed b y a discussion o n the view of 'discourse' adopted in the study, a n d a case for the inclusion of multisemiotic analysis in the study of discourse. I n the succeeding section, I will present the analysis of the two apparently contending discourses of g e n d e r relations (the Egalitarian Discourse and the Conservative Discourse) in the advertising campaign in question, which I refer to as the 'Family Life' ads. I n the campaign, approximately 14 of the ads deal with representations of p a r e n t h o o d ; a brief description of these is outiined in the A p p e n d i x . I will conclude b y summarizing the findings of the analysis, and suggesting that far from being straightforward, the Conservative/EgaHtarian distinction itself is in n e e d of problematization. 2

3

T h e Study i n C o n t e x t T h e present study is part of a larger project o n a critical discourse analysis of gender (and sexuality) in Singapore's Family Life advertising campaign that I h a v e u n d e r t a k e n over several years. This campaign arose from the changed, pronatalist fertility policy actively pursued b y the government of Singapore since the 1980s. At that time the g o v e r n m e n t h a d b e c o m e concerned with the rapid decline in the country's birth rate, b o t h in terms of the 'quantity' of births (i.e. nationally), as well as with the 'quality' of births (i.e. the better-educated segment of the population, in particular, were reported as m a r k e d l y under-reproducing). Amongst the various measures u n d e r t a k e n in order to rectify the situation was the inception of the national, multi-media 'Family Life' advertising campaign. It needs to b e said that the launching of such a campaign in itself is quite unre markable in the Singapore context, which m a y b e d u b b e d a 'campaign country'. Historically, since the present leadership came to p o w e r 41 years ago, an assort m e n t of national campaigns h a v e b e e n utilized as regulative instruments in the service of social engineering. T h e present campaign u n d e r study is n o exception. T h e 'Family Life' campaign is a two-pronged endeavour: one half of the pro g r a m m e is targeted at single w o m e n a n d m e n , encouraging t h e m to get married, and another, concurrent, half is addressed to those already married, urging them to h a v e babies soon. I n an earlier p a p e r to Discourse & Society (1993), I had dealt, in part, with the 'first' half of the campaign. I n the present p a p e r , m y focus is on the 'second' part, in which I a m specifically interested to ask w h a t it means in these ads for w o m e n a n d for m e n to b e c o m e parents, a n d what are the power dynamics at work in a n d through these representations. I n order to appreciate the workings of the Egalitarian a n d Conservative Dis courses of G e n d e r Relations that are co-present in the 'Family Life' ads (and. a* we shall see, the complexity of their inter-relationship), it is necessary first to con sider w h y it is that childbearing h a d b e c o m e generally unpopular in Singapore, thus precipitating the inception of this national campaign. A p a r t from the effectiveness of the previous (now defunct) Family Planning or Limitation cam paign, which h a d extolled the benefits of a small family, the practice of having

LAZAR

POLITICS OF REPRESENTATIONS

393

fewer children h a d b e c o m e symptomatic of a coping strategy employed b y m a n y Singaporean w o m e n w h o were in the paid labour force. This is p e r h a p s best understood in terms of what Bernard (1972) has described as the transition from the ' O n e Role Ideology' to the 'Two Role Ideology' (cited in Q u a h , 1994: 189). According to the ' O n e Role Ideology', w o m e n ' s roles were exclusively those of childbearing, childrearing a n d housekeeping. I n the case of the 'Two Role Ideo logy', w o m e n ' s roles w e r e e x t e n d e d to include p a i d work outside the h o m e in addition to their original domestic responsibilities. I n the 1960s a n d 1970s, w h e n the government h a d urged large n u m b e r s of w o m e n to enter a n d stay in the paid workforce, m a n y Singaporean w o m e n found themselves at grips with the 'Two Role Ideology'. T h e change occurring for w o m e n in regard to the public sphere, however, was n o t m a t c h e d b y concomitant changes in gender roles a n d expect ations in the (distincdy separated) private sphere w h e r e housework and childcare were concerned. Optionally, some m e n h a v e helped out in the h o m e , b u t it has b e e n just that - 'helping out' or 'lending a hand', which in itself is d e e m e d praiseworthy. T h e r e is nothing remarkable, however, w h e n w o m e n perform daily domestic chores a n d care for children - in addition to working outside the h o m e - for that is considered 'women's work' anyway. With little restructuring of the g e n d e r order, this m e a n s that w o m e n in paid e m p l o y m e n t h a v e come to hold two equally d e m a n d i n g jobs. C o m m e n t i n g to The Straits Times (Singapore's main daily), one (female) academic summed u p the impact this has h a d o n women: A n y o n e w h o has attempted to h o l d two j o b s will understand w h y the key terms recurring again a n d again in w o m e n ' s letters to the press are 'exhaustion', 'guilt', 'conflict', ' b u r d e n ' , a n d 'tired[ness]'. T h e y will also see w h y so m a n y w o m e n try to curtail one j o b or the other - b y d r o p p i n g out of the work force or having only o n e child. (Heng, 1983) In fact, some w o m e n have opted for childless marriages. Figures have shown that since 1980, the proportion of all ever-married w o m e n with n o children h a d risen b y a third, to 12 percent, in 1990 (Liak, 1994:56). T h e majority, w h o would like to have children, have tended to postpone childbearing, or space out the birth of their children over a wider period (Fawcett a n d K h o o , 1980, reported in Hill and Lian, 1995:150), a n d they generally have small-sized families. T h e n u m b e r of births per woman, moreover, correlated direcdy with her educational attainment, which caused m u c h consternation to Singapore's eugenic-minded leaders: they were worried that the better-educated class of w o m e n , w h o m they saw as hav ing superior or m o r e intelligent genes, w e r e n o t reproducing sufficiently. T h e r e has also b e e n a n o t e d discrepancy between the n u m b e r s of children this class of women are generally willing to h a v e c o m p a r e d to m e n of a similar educational background. I n a relatively recent g o v e r n m e n t survey, it was found that whilst most female tertiary graduates only w a n t e d two children, most m a l e graduates wanted three. T h e difference was explained b y the fact that ' w o m e n were aware of the heavy responsibilities in the dual roles of being mothers and career w o m e n ' [The Straits Times, 21 April 1997). T h e double b i n d that has confronted w o m e n is n o t only asymmetrically ap portioned at the level of private interaction between w o m e n and m m h u t ic

DISCOURSE STUDIES

394

underscored b y various public policies a n d practices of the state (see Lazar, in press). W o m e n are granted (almost) the same access to educational a n d employ m e n t opportunities as m e n , and are much needed for their economic productivity. H o w e v e r , at the same time, a range of state policies h o l d sacrosanct traditional, Confucian-Asian values that support m e n ' s position as heads of households, a n d unequivocally emphasize w o m e n ' s reproductive a n d nurturing roles as mo thers. Whilst w o m e n are u r g e d to r e m a i n economically active, their maternal roles are prioritized for them, as the following rhetorical question b y the former Prime Minister, M r L e e K u a n Yew, m a k e s a m p l y clear: ' D o our w o m e n need to value their careers m o r e than, a n d at the expense of, their families?' (The Straits Times, 19 February 1984). 4

Discourse and Semiotics Discourse^) 'Discourse' is here understood in Foucauldian terms to m e a n a set of related state ments that p r o d u c e a n d structure a particular order of reality, a n d which within that reality m a k e s available specific subject positions. I n this vein, I find it use ful to think of discourse as a socio-historically contingent 'meaning potential' to b o r r o w a Hallidayan terminology - that b o t h enables a n d constrains possible ways of knowing about the world, a sense of w h o we m a y (and m a y not be) within that world order, a n d h o w w e m a y (and m a y not) relate to o n e another. I n the d o m a i n of parenthood, this m e a n s that there is nothing inherendy fixed about the identities of ' m o t h e r ' a n d 'father', or in the w a y gender relations be tween t h e m h a v e t e n d e d to b e structured. Rather, these are socially formed and fixed in a n d through discourse. We perceive these relations a n d identities in certain ways (say, m o t h e r h o o d as 'natural' a n d fatherhood as 'social') as a result of conventional ways in which the societies we live in h a v e c o m e to express and think about them. W h a t is fixed, therefore, can b e analytically un-fixed or dis mantled. T h a t is to say, discourses (and the realities a n d subjectivities that they m a k e available) can b e taken apart in such a way as to reveal that they are noi i m m a n e n t truths, b u t rather are constructed that w a y from particular positions that serve particular interests, whilst subordinating others. In this study, I analyse the interplay of two discourses of gender relations that operate in the domain of p a r e n t h o o d : a d o m i n a n t Discourse of Conservative G e n d e r Relations, a n d a c o u n t e r Discourse of Egalitarian G e n d e r Relations. Both underscore the relational aspect of g e n d e r construction, that is, ways of being a ' w o m a n ' a n d a ' m a n ' (or, as is the case in this study, 'mother' and 'father j are always explicitly or implicitly co-constructed one in relation to the other. T h e two discourses, however, structure this relationship in different, indeed (potentially) contradictory, ways. T h e difference is o n e of parity; specificallv, whether possibilities for ways of being a n d b e c o m i n g are equally available and interchangeable between w o m e n a n d m e n . T h e Conservative Discourse deriving

LAZAR


395

from a Confucian-Asian ethic favours a traditional, asymmetrical arrangement between the genders, whereby w o m e n a n d m e n each h a v e gender-specific roles, responsibilities a n d expectations to fulfil. T h e Egalitarian Discourse, in contrast, which is motivated in part b y feminism, strives for gender parity in all aspects of personal a n d public life. At stake i n the two discourses is the b a l a n c e of p o w e r between w o m e n a n d m e n as they enact their respective parental (and other, careerist) identities. Bridging Discourse and Semiotics A valuable insight worth importing into C D A from Foucault is that discourse is not equivalent to language use alone. I n his view, discursive s t a t e m e n t s are n o t reduced to sentences a n d propositions, as is the usual understanding of the t e r m 'statements' in linguistics. Indeed, discursive statements are not units or structures (in the way, say, that sentences are), b u t a f u n c t i o n (or meaning) that operates as m u c h through language as through other semiotic modalities. I n other words, graphs, m a p s , classificatory tables, statistical calculations and algebraic formulae (Foucault, 1972) are as m u c h forms of discursive statement as is language. Extending from this conception of discourse, 'text' can b e viewed as the multisemiotic manifestation of discoursal meaning(s). Typically, however, m a n y studies have t e n d e d to conflate 'discourse' with 'language' and therefore, b y analysing only linguistic structures a n d meanings in texts, h a v e h a d a restricted scope. A n d if images (for example) are also discussed, they h a v e b e e n considered separate from 'text' - a term which, like 'discourse', is reserved only for linguistic p h e n o m ena. W h a t I propose, therefore, is the uncoupling of the two categories 'discourse' and 'language' in favour of discourse encompassing semiosis of various kinds (including language). A critical analysis of discourse, in other words, involves a commitment to the analysis of various strands of semiosis that configure in the realization of particular discoursal meanings in texts. T h e present study will demonstrate, for instance, h o w the interplay of language a n d visual images (including layout, gestures a n d actions) contributes to the manifestation of the two contending discourses of g e n d e r relations in the 'Family Life' texts. Not to attend concurrently to b o t h m o d e s of semiosis, in this case, would b e to offer a less than complete analysis of discourse (Lazar, 1999). Analysing Discourse Discourse is analysable in terms of a set of representations that express a particular discoursal m e a n i n g (e.g. Conservatism or Egalitarianism). To give an example, part of the Discourse of Egalitarian G e n d e r Relations in the 'Family Life' ads can b e analysed in t e r m s of particular representations of g e n d e r identities: the 'New M a n ' a n d the 'New Woman'. T h e aim of the analysis is to show u p these various representations, singly a n d in constellation with others, which evidence the operation of a particular discourse in the text(s). This being the aim, the start ing point of the analysis will b e to identify types of representation articulated in

396

DISCOURSE STUDIES

the ads, supported b y a n a m a l g a m of textual structures a n d processes that cue these representations. T h e explication of textual resources involves the interplay of several categories of textual structure (be that linguistic, non-linguistic, or lin guistic a n d non-linguistic) that work in t a n d e m to manifest a particular type of representation. I n order to analyse the linguistic a n d visual structures in the texts I use the g r a m m a r s of Halliday (1994), a n d Kress a n d v a n Leeuwen (1996), respectively. T h e various aspects of the g r a m m a r s will b e d r a w n u p o n in different degrees in the study. As m a n y will know, Halliday's functional g r a m m a r is based o n a view of language as simultaneously involved in three kinds of meaning-making: •

Ideational (representing the w o r l d 'out t h e r e ' a n d our inner states analysable in terms of participant types, processes, and circumstances). • Interpersonal (enacting social relationships between participants - analytic categories include types of m o o d a n d modality choices). • Textual (making connections within the text, a n d the text a n d its social context - analytic categories include thematic a n d information structures, a n d cohesion).

T h e grammar is analysable at the ranks of word, group a n d phrase, and clause and clause complex (sentence). T h e analysis of the images in the texts is based largely o n the grammar of visual design developed b y Kress a n d van Leeuwen, following Halliday's theory of triple meaning-making: •

Ideational m e a n i n g is of two types: (a) narrative, which includes dual par ticipant ('transactional') or single participant ('non-transactional') action and reaction (or perceptual) structures, realized b y vectors of various sorts; a n d (b) conceptual, which includes classificational ('type o f ) and analytical ('part-whole') structures. • Interpersonal m e a n i n g includes, amongst others, contact (or mood) - real ized b y eye contact; a n d distance or social affinity - realized b y frame size a n d types of shots. • Textual meaning (or composition) includes such categories as salience - real ized b y relative size and sharpness of focus; and information value - realized b y relative position (centre or margin) within a frame.

M y use of the visual g r a m m a r is supplemented b y insights selectively drawn from Goffman's work on gender in advertisements (1979).

Analysis of P a r e n t h o o d i n Discourses of Egalitarianism a n d C o n s e r v a t i s m T Koorin

W nresentinu the analysis of the Egalitarian Discourse of

LAZAR


397

G e n d e r Relations in terms of the interplay of linguistic a n d visual resources deployed in the ads. T h e two discourses are co-present within single ads, as well as across the intertextual spread of the 14 ads in the campaign. T h e basic distinc tion between the two discourses, as previously indicated, is o n e of gender parity, i.e. w h e t h e r there is an equitable access to possibilities for being. I n m o d e r n in dustrial societies, the issue of g e n d e r parity largely rests o n the possibilities, a n d the extent of negotiation, that are available to w o m e n a n d to m e n between the private a n d the public spheres of life (i.e. between the h o m e / t h e family a n d paid work/career). The Discourse ofEgalitarian Gender Relations The Discourse of Egalitarianism in the ads is identifiable in terms of three types of representation: (1) the representation of parenthood as symmetrical for w o m e n and m e n ; (2) the depiction of m e n in the domestic sphere as devoted, nurturing fathers; a n d (3) the portrayal of w o m e n as mothers as well as successful careerists outside the h o m e . Women and men O n e way that egalitarianism between the genders is con strued in the ads is through an invitation to view p a r e n t h o o d as being identical for w o m e n a n d m e n . I n m a n y instances, w o m e n a n d m e n are collectively referred to as joint participants. For example, in terms of lexical choice the gender-neutral noun parents or parenthood is overwhelmingly favoured (over the gender-specific terms ' m o t h e r ' / ' m o t h e r h o o d ' a n d 'father'/'fatherhood'). A n d in terms of the choice of pronouns, we repeatedly find either the collective out/we or the indefinite pronoun you (depending o n whether the ad is presented from the first or second person point of view). I n either case, gender is again elided, a n d it appears to represent w o m e n and m e n alike, as constituting a single undifferentiated unit. For example: (1) (2) (3) (4)

It's the most precious giftparents can give (Lonely Child). Becomingparents changes our lives completely... (Something Wonderful Happened). As you gaze at your child, you willfeel so proud to be parents (Experience The Joy). Understandably, parenthood is a big decision (Experience The Joy).

As a corollary of being referred to as joint participants, we also find in the above clauses that w o m e n a n d m e n are represented as sharing a c o m m o n experience. They are affected b y p a r e n t h o o d in the same way, a n d appear to respond to it identically. T h e representation of egalitarianism expressed in the language struc tures, moreover, finds visual support in one of the (print) ads. I n Something Wonderful Happened, the symmetry is striking o n two counts. T h e couple are shown in a joint action process of pushing a b a b y stroller together, thereby showing that what they are doing is a joint endeavour. I n addition, the couple are portrayed as distributed symmetrically across the picture space (i.e. they are at equal distance from each other, a n d are roughly equal in size and orientation), which according to Kress and van Leeuwen (1996:88) expresses a 'covert taxonomic

DISCOURSE STUDIES

396

the ads, supported by an amalgam of textual structures and processes that cue these representations. The explication of textual resources involves the interplay of several categories of textual structure (be that linguistic, non-linguistic, or lin guistic a n d non-linguistic) that work in tandem to manifest a particular type of representation. In order to analyse the linguistic and visual structures in the texts I use the grammars of Halliday (1994), and Kress and van Leeuwen (1996), respectively. The various aspects of the grammars will be drawn upon in different degrees in the study. A s many will know, Halliday's functional grammar is based on a view of language as simultaneously involved in three kinds of meaning-making: • Ideational (representing the world 'out there' and our inner states analysable in terms of participant types, processes, and circumstances). • Interpersonal (enacting social relationships between participants - analytic categories include types of m o o d and modality choices). • Textual (making connections within the text, and the text and its social context - analytic categories include thematic and information structures, and cohesion). The grammar is analysable at the ranks of word, group and phrase, and clause and clause complex (sentence). The analysis of the images in the texts is based largely on the grammar of visual design developed b y Kress and van Leeuwen, following Halliday's theory of triple meaning-making: • Ideational meaning is of two types: (a) narrative, which includes dual par ticipant ('transactional') or single participant ('non-transactional') action and reaction (or perceptual) structures, realized by vectors of various sorts; and (b) conceptual, which includes classificational ('type o f ) and analytical ('part-whole') structures. • Interpersonal meaning includes, amongst others, contact (or mood) - real ized b y eye contact; and distance or social affinity - realized b y frame size and types of shots. • Textual meaning (or composition) includes such categories as salience - real ized by relative size and sharpness of focus; and information value - realized by relative position (centre or margin) within a frame. My use of the visual grammar is supplemented b y insights selectively drawn from Goffman's work on gender in advertisements (1979).

A n a l y s i s of P a r e n t h o o d i n D i s c o u r s e s o f Egalitarianism a n d C o n s e r v a t i s m In this section, I begin b y presenting the analysis of the Egalitarian Discourse of Gender Relations, followed b y the analysis of the Conservative Discourse of

LAZAR


397

Gender Relations in terms of the interplay of linguistic and visual resources deployed in the ads. The two discourses are co-present within single ads, as well as across the intertextual spread of the 14 ads in the campaign. The basic distinc tion between the two discourses, as previously indicated, is one of gender parity, i.e. whether there is an equitable access to possibilities for being. In modern in dustrial societies, the issue of gender parity largely rests on the possibilities, and the extent of negotiation, that are available to women and to men between the private and the public spheres of life (i.e. between the home/the family and paid work/career). The Discourse of Egalitarian

Gender

Relations

The Discourse of Egalitarianism in the ads is identifiable in terms of three types of representation: (1) the representation of parenthood as symmetrical for women and men; (2) the depiction of men in the domestic sphere as devoted, nurturing fathers; and (3) the portrayal of women as mothers as well as successful careerists outside the home. Women and men One way that egalitarianism between the genders is con strued in the ads is through an invitation to view parenthood as being identical for women and men. In many instances, women and men are collectively referred to as joint participants. For example, in terms of lexical choice the gender-neutral noun parents or parenthood is overwhelmingly favoured (over the gender-specific terms 'mother'/'motherhood' and 'father'/'fatherhood'). And in terms of the choice of pronouns, we repeatedlyfindeither the collective ourfive or the indefinite pronoun you (depending on whether the ad is presented from the first or second person point of view). In either case, gender is again elided, and it appears to represent women and men alike, as constituting a single undifferentiated unit. For example: (1) (2) (3) (4)

It's the most precious gift parents can give (Lonely Child). Becomingparents changes our lives completely... (Something Wonderful Happened). As you gaze atyour child, you willfeel so proud to be parents (Experience The Joy). Understandably, parenthood is a big decision (Experience The Joy).

As a corollary of being referred to as joint participants, we also find in the above clauses that women and men are represented as sharing a common experience. They are affected by parenthood in the same way, and appear to respond to it identically. The representation of egalitarianism expressed in the language struc tures, moreover, finds visual support in one of the (print) ads. In Something Wonderful Happened, the symmetry is striking on two counts. The couple are shown in a joint action process of pushing a baby stroller together, thereby showing that what they are doing is a joint endeavour. In addition, the couple are portrayed as distributed symmetrically across the picture space (i.e. they are at equal distance from each other, and are roughly equal in size and orientation), which according to Kress and van Leeuwen (1996:88} exnrpscoc < > — " — „1

•£•--••

398

DISCOURSE STUDIES

belong to the s a m e category o f ' p a r e n t s ' , rather t h a n being classified according to the differential a n d value-laden roles of 'father' a n d 'mother'. C o m m e n t i n g o n t h e n o w widespread use of such gender-neutral terms as 'parenting' a n d 'parenthood', some scholars (e.g. Busfield, 1987) have noted that two different b u t related assumptions underlie their usage. Both of these sup port the egalitarian thesis put forward in this paper. T h e first assumption is that gender relations are b e c o m i n g m o r e symmetrical (as analysed above), and the second assumption is that m e n are b e c o m i n g m o r e involved in the day-to-day care of their children - which is the focus of the n e x t sub-section. Men T h e Discourse of Egalitarianism in relation to p a r e n t h o o d is also ar ticulated in terms of a construction that is widely k n o w n in g e n d e r and cultural studies as the 'New M a n ' . This is a type of masculinity that has gained popularity in the (mainly Western) m e d i a since the 1980s, in response to feminism's critique of traditional forms of masculinity. A significant dimension of the m o d e r n New M a n is a caring, sensitive a n d nurturing depiction of fatherhood - an involve m e n t that is a far cry from 'authoritarian' or 'distant breadwinner' (Pleck, 1987, cited in C h a p m a n , 1988) images of fatherhood in the past. According to this new portrayal, m e n are very comfortable with infants a n d very y o u n g children, and are expressive of care a n d emotion, which traditionally w e r e seen to b e the pre rogative of w o m e n a n d m o t h e r h o o d . Indeed, what makes the N e w Father a pari of t h e egalitarian discourse of p a r e n t h o o d is its b r e a k i n g d o w n of gender stereotypes. As Rutherford (1988: 34) describes, the N e w Father 'looks soft and gentle and, what's m o r e , he's n o t afraid to show it'. Somewhat ironically, there fore, the egalitarianism appears to b e based o n the feminization of men. I n the 'Family Life' ads, the N e w Father is especially strongly represented in interactions between m e n and their n e w b o r n babies. T h e portrayal is jointly real ized b y visual representational structures as well as in compositional structures. I n terms of the visual representational structures, three characteristic features of tenderness a n d emotional b o n d i n g are evident: cradling, intently gazing, and sweetly smiling at the baby. Quite obviously, cradling is a transactional action structure w h e r e b y the m a n (Actor) carries the infant (the Goal) gentiy in his arms. At the same time, however, the cradling is also indicative of an analytical structure that represents a c o m p o u n d relationship between the m a n as Carrier a n d the b a b y as an Attribute, which is visually seen to b e extension of him. C o u p l e d with cradling, are the reactional structures of smiling a n d gazing upon the infant in his arms. C o m p o s i t i o n a l structures, further, frame these depictions in ways that heighten their emotional value. Close-up shots are especially deployed for this purpose. So, for example, a close-up in Your Family is Your Future focuses on the facial expression of a y o u n g father, w h o gazes lovingly a n d smiles almost tear fully at his infant. Also, in Something Wonderful Happened, there is a close-up shoi of a large m a l e h a n d holding the very small h a n d of a b a b y to the accompani m e n t of the following reiterated adjective in the lyrics 'tiny fingers, tiny toes'.i Emphasizing the contrast in the size of the two h a n d s , a n d the gentleness ofl the touch, evokes a heart-tugging quality of the father's sensitivity vis-a-vis thel baby's vulnerability. A

LAZAR


399

T h e tender, nurturing role of the N e w Father in these representations approxi mates so m u c h to stereotypical constructions of m o t h e r h o o d that in a family shot in Your Family is Your Future, there is a reversal in the positions occupied b y the father and the mother in relation to their baby. I n contrast to traditional repre sentations of the m o t h e r carrying the b a b y and the father protectively w r a p p i n g his a r m a r o u n d h e r shoulders (Goffman (1979) calls these 'shoulder-holds'), in this particular ad, it is the father w h o is shown cradling the baby, with the mother's arm a r o u n d h i m . T h e Sensitive N e w M a n is constructed not only visually, b u t also linguistically: (1) And it was only then, as I took her [the baby] into my arms for the very first time and looked down into her tiny, perfect face that I realised my whole life had changed. (Your Family is Your Future, TV) (2) They're [My children are] my life and my future. They are my hope, my strength. The reason lean carry on. (Collage Ad, TV) In these extracts w e find an emotional response to the experience of fatherhood. (The first e x a m p l e from Your Family is Your Future is the verbal equivalent of the visual structures earlier analysed). T h e y are b o t h expressed from the first person point of view (I, my) a n d stress (via repetition of my life) that the lives of these men are profoundly affected b y children (note die material processes: had changed and (can) carry on). W h e r e a s the impact is succinctiy encapsulated in the first of these ads via premodification (my whole life), this is elaborated over a series of identifying relational processes in the second ad, in which the father identifies himself wholly with his children. More generally, too, m a n y of the ads show m e n to b e highly visible a n d actively involved in family life. A p a r t from representations of m e n as Actors in the visual structures of these ads, the m e n are also depicted in analytical terms. To take just one example, in the (TV) ad Why Build Your Career Alone?we are presented with various snapshots of a father playing a b o a r d g a m e with his family, sitting along side and watching his son d o school work, a n d sitting with his family gathered around him. I n all of these scenes, the m a n together with the other characters are represented as parts that m a k e u p 'the (whole) family'. Such a representation invites an interpretation of egalitarian g e n d e r relations, for h e r e is a m a n w h o is portrayed as very m u c h involved, and part of the everyday family scene. Women Whilst m e n in the Discourse of Egalitarian Gender Relations are rep resented as involved participants in the domestic sphere, w o m e n in this discourse are shown as participants not only in the private b u t also in the public work space as well. This is a depiction of the m o d e r n 'New Woman', as opposed to traditional images of w o m e n as full-time caregivers at h o m e . I n two of the ads, Make Room For Love (TV) a n d Babies and Careers (Print), it is suggested that w o m e n can h a v e a career a n d b e successful at it, whilst also fulfilling familial roles. I n Make Room for Love, this is conveyed in the depiction of 'Mrs Tan', a senior, newly retired career woman. Two pertinent aspects of her identity are revealed in the following congratulatory utterance b y a younger (former) colleague: You had a wonderful career, Mrs Tan. T h e clause neatiy brines together the w o m a n ' s rlnî —

DISCOURSE STUDIES

400

in the public a n d private spheres: she is represented as Carrier both in terms of a career (and note t h e epithet wonderful to describe it) a n d in terms of marital status, indicated b y t h e tide Mrs. Later in t h e ad, M r s Tan is also called mom a n d granny b y another w o m a n (her daughter) a n d two youngsters (her grandchil dren), respectively, w h o enter t h e scene. 'Wife', 'mother', a n d ' g r a n d m o t h e r ' all belong within a c o m m o n semantic field of familial identities attributed to this career w o m a n . I n other words, she is offered in the a d as an example of a w o m a n w h o h a s h a d it all - a marriage, a family, a n d a successful career. If the portrayal of M r s Tan - s o m e o n e of an earlier generation - appears opti mistic, then t h e following declaration i n Babies and Careers b y a 29-year-old w o m a n (the age group of the target audience) is even m o r e heartening: It's now easier to be successful in both [a career and motherhood]. T h e choice of declarative m o o d together with categorical modality renders the proposition a simple undis p u t e d statement of fact. These, plus t h e time indicators - now a n d easier (& com parative reference presumably being m a d e to an earlier generation) - construe a 'reality' that purportedly reflects current, m o r e enlightened times.

The Discourse of Conservative Gender Relations Although thus far w e h a v e seen a discourse at work that p r o m o t e s a symmetrical and egalitarian gender order, there is a concurrent overwhelming presence of the Discourse of Conservatism that maintains gender asymmetry in the ads. I n what follows, I analyse the g e n d e r asymmetry in t h e construction of p a r e n t h o o d on two levels: (1) t h e systematically divergent roles performed b y w o m e n a n d m e n within the domestic sphere itself, a n d (2) t h e different ways in which w o m e n a n d m e n m a y negotiate their respective parental identities b e t w e e n t h e private and public spheres of life. I shall first present the two-level 'asymmetry analysis' in regard to fathers, followed b y t h e analysis concerning mothers. A discussion on the dynamics between t h e conservative a n d egalitarian discourses of g e n d e r re lations will b e undertaken in the concluding section of t h e paper.

Men (A)Within the Private Sphere: I n terms of the construal of the conservative discourse, let us n o w consider t h e (asymmetrical) representation of m e n ' s involvement in t h e private sphere. First of all, there appears to b e a clear representation of gender role differentiation in the h o m e front. Fathers are mainly depicted in executive (functional) roles, realized i n transactional action structures. T h e role of the executor as Goffman (1979: 32) h a s noted, is typically performed b y m e n and n o t w o m e n in representations w h e r e t h e two are co-present. To take an example, it is the fathers in the 'Family Life' ads w h o are consistendy represented as the Actors operating such equipment as cameras a n d camcorders i n family situations. I n Something Wonderful Happened, a y o u n g father uses a camcorder to 'record' images of his sleeping infant; i n Because That's Your Family, t h e father sets u p t h e c a m e r a a n d r u n s back to j o i n his family in a self-timer shot; a n d in Precious Moments, t h e m o t h e r passes t h e c a m e r a over to t h e father to take a 5

LAZAR


401

p h o t o g r a p h of their children playing together, instead of taking the p h o t o g r a p h herself. Fathers-as-executors extend to m e n ' s interactions with children. Fathers are c o m m o n l y represented with sons, a n d in these interactions, m e n are shown in various ways directing their y o u n g sons in g e n d e r e d play a n d behaviour. For instance, in Because That's Your Family (TV), the father is represented showing his son the mechanics b e h i n d the r u n n i n g of a toy train, and in Kids Make You See (TV), the father teaches his son to ride a bicycle. E v e n w h e r e the m a n is n o t in an active instructor or executor role (i.e. as Actor), we h a v e a powerful represen tation in Fam, Fam, Fam (TV) of a b o y learning gendered behaviour, nonetheless, from his father t h r o u g h observation a n d imitation. T h e father, shown leisurely reading the newspapers with his legs widely crossed, b e c o m e s the P h e n o m e n o n (or object) of his son's avid gaze, a n d his b o d y posture and activity are direcdy replicated in the b o y ' s subsequent actions. T h e b o y ' s emulation (Phenomenon), in turn, does n o t go unnoticed b y the father (Senser), w h o smiles approvingly to himself. I n this way, the father is still shown very m u c h in the executive role, only this time directing indirecdy. Following Pleck (1987) (reported in L u p t o n and Barclay, 1997), who undertook a historical study of fatherhood in America, all these portrayals seem to support the m o r e traditional, 'sex role m o d e l ' type of fatherhood, which pre-dates the N e w Father. A p a r t from the executor representation above, fatherhood (compared to motherhood) is largely construed in terms of fun a n d physical play. While the rep resentation of m e n ' s involvement with children in this w a y is undeniable, it is asymmetrical in so far as gender roles a n d relations are concerned. Fun, play a n d popularity appear to b e the prerogative of fatherhood a n d take centre-stage, whereas, as we shall see in the section o n 'Women', support a n d routine care are largely left to m o t h e r h o o d , which is relegated to the periphery. Taking centrestage m e a n s that there is greater frequency in the portrayal of fathers' activities with children, a n d these appear saliently in relatively large shots (e.g. Fam, Fam, Fam (Print), a n d Kids Make Your World Brand New (Print)). Moreover, father-children interactions typically centre o n leisure activities rather than on intensive day-to day care-giving activities. This is indicative of differential understandings of care that apply to fathers as opposed to m o t h e r s : 'care', where fathers are concerned, is understood purely in terms of fun a n d relaxation. ('Care' and m o t h e r h o o d will b e dealt with in the following section.) T h e popular 'fun dad' representation is realized mainly through transactional action structures, I n m a n y instances, fathers are the Actors a n d their children are the Goal or Beneficiary at or for w h o m the m e n ' s physically affectionate b e h a v i o u r is directed. For e x a m p l e , a father is shown ruffling his son's hair, tick ling a n d playing with him, nuzzling a n d kissing his children, a n d m a k i n g funny faces at t h e m (in Kids Make Your World Brand New (TV), Kids Make You See (TV), a n d Fam, Fam, Fam (TV)). I n some instances, the roles are reversed, that is, fathers are the recipients a n d the children are the Actors. O n e such case is found in Fam, Fam, Fam (Print), w h e r e three y o u n g children (Actors) sitting o n their p a r e n t s ' b e d pillow-fight their dad (Goal), w h o erood-natnrpHlv r p r p i w e tho 'kl™.™,' 8

DISCOURSE STUDIES

402

T h e playfulness of dads is also significant in portraying t h e m as ' o n e of the kids', which accentuates their popular appeal. For example, in the same ad (Fam, Fam, Fam, TV), the father a n d children are united in their role as Actors, w h o con spire to 'steal' and eat cookies from the kitchen counter - cookies freshly b a k e d b y the children's mother. T h e representation of being 'one of the kids' is especially well expressed via a classificational (covert taxonomic) process. Preceding the cookie stealing shot, the father a n d children are identically represented as they place their heads one on top of the other, in a totem-pole fashion, forming a sym metrical vertical composition. A s Kxess a n d v a n L e e u w e n (1996: 81) h a v e observed, 'For participants to b e p u t together in a syntagm which establishes a classification m e a n s that they [are] j u d g e d to b e m e m b e r s of the same class and are to b e r e a d as such.' It is important to b e a r in m i n d that representations of gender role differentia tion - whether in terms of m e n ' s executive role or their 'fun daddy' role - are not simply about functional difference but, rather, are implicated in asymmetrical power relations. This is p e r h a p s most striking in representations of fathers as heads-of-households. I n Why Build Your Career Alone. (Print a n d TV) this is indi cated b y the central position occupied b y the father in relation to m e m b e r s of his family. Such a shot demonstrates a n exhaustive analytical structure, whereby the two adults and three children in this ad are parts (or 'Attributes') that altogether comprise 'the family' (the 'Carrier'). However, although the father is o n e member of the family (and one of t w o adults), h e is represented as the most salient - lit erally the central - figure, flanked b y his wife, w h o stands partially hidden behind him, and his children w h o encircle h i m in the front Furthermore, other portrayals of h i m in relation to his family represent h i m as a Carrier himself, with mem bers of his family r e p r e s e n t e d as his Attributes. T h i s interpretation of the analytical structures evokes a proprietal relationship, expressed b y the m a n plac ing a n a r m a r o u n d the shoulders of his family m e m b e r s . I n one scene, h e has an a r m a r o u n d the w o m a n in the ad, which defines h e r in relation to him as h i s wife, a n d in another scene, the m a n has an a r m a r o u n d a little boy, which defines the child in relation to h i m as h i s son. Shoulder-holds are a c o m m o n enough ges ture also found in other ads where m e n are represented with their family (for example. Experience theJoy, Because That's Your Family, a n d Kids Make You See). ?

7

W h a t I h a v e analysed so far in terms of the Discourse of Conservative Gender Relations are asymmetrical representations of men's roles in the family (i.e. in the private sphere) as (1) executors, (2) popular a n d fun dads, a n d (3) heads of house holds. I n the n e x t sub-section, I shall focus o n h o w m e n are represented as being able to negotiate b e t w e e n their identity as fathers a n d their public work identity. (B) The Public versus the Private: I n contrast to ads that address women, those directed at m e n construct a world in which m e n basically can h a v e it all withoui significant tension. T h r e e b r o a d types of representation are found in the ads. which I will examine in turn: (1) the h a r m o n i o u s co-existence of a man's career with farnily life; (2) absenteeism of fathers on account of their careers as excusable: a n d (3) a family as beneficial to m e n ' s own self-interests.

LAZAR

403


First, in some ads family life a n d paid-work life are represented as co-existing harmoniously for m e n , that is, the identities of father a n d careerist are reconcil able without one compromising the other. T h e compatibility b e t w e e n the identi ties is enacted b y bridging representations of m e n in the public a n d private worlds. T h e portrayal of m e n in the public world of w o r k is signalled b y the clothes they wear (long-sleeved shirts and ties), i.e. m e n here are seen as Carriers whose professional identities are inscribed u p o n their bodies, a n d (optionally) also in terms of the representation of a n office setting. Although the m e n , in these instances, are seen primarily in their professional capacity, their identity as fathers is quite easily co-enacted. This is explicidy the case in Why Build Your Career Alone? and Fam, Fam, Earn: t h e words for rfarfhandwritten on a gift b o x and dad written in icing o n a birthday cake, respectively, are displayed in the fore ground. T h e m e n are shown in turn, via transactional actional or reactional structures, as claiming this identity. I n Why Build Your Career Alone? (TV) the m a n is represented lifting the gift u p from his office desk with a smile. I n Fam, Fam, Fam the m a n , still in his office attire, is represented peering through the kitchen doorway of his h o m e and being pleasantly surprised b y a birthday cake b e i n g prepared for h i m b y his family. T h e direction of his gaze forms a vector linking him (in his professional identity) to the cake with the words Happy Birthday Dad written o n it. 8

Even w h e r e there are admissions that a m a n ' s career commitments m a y compete with his family role, this is r e n d e r e d as expected a n d understandable. The following are two verbal examples from the Collage Ad (TV): (1) even though my work takes me away, when it comes to joy and dreams, my children are the key. (2) even though my work takes me away, my children are my hope andjoy supreme. The concessive (conjunctive) adjunct at the beginning of the clause complexes sets u p a presupposition that m e n would b e away on account of their career, a n d that this in itself is h a r d l y surprising. Further, the choice of the material process attributes my careerwikh an agency of its own, positioning the m a n himself as the unwitting Goal, haplessly 'taken away' from his family. T h e r e m a i n d e r of the clause complexes, moreover, appears to compensate the absenteeism b y represen tations of effusive declarations of the importance of his children to him. T h r o u g h the rhetorical strategy adopted in these clauses, absenteeism of fathers is accepted and forgiven, and n o t seen as a censure on m e n to better balance their twin roles. Further, the representations in the examples are one-sided: w e are told of the effect children h a v e o n the father, b u t n o t the effect of his absence u p o n them. The latter is b o r n e out b y P h o e n i x a n d Woollett's (1991) observation that the existing literature on fatherhood rarely deals with the impact that fathers' frequent absences on account of their careers h a v e o n the d e v e l o p m e n t of their children. Further to the representation of the unproblematic co-existence of family and career for m e n , m e n ' s careers are represented as positively enriched b y family life. It is not only t h e case that, as w e saw, fathers are showered with gifts a n d birthday

DISCOURSE STUDIES

404

surprises from their family members, but that having a family is construed as positively benefiting their personal and career development. In two of the ads Why Build Your Career Alone? and Your Family is Your Future - the family is por trayed in instrumental terms, as bringing about certain personal qualities deemed necessary for men's professional success. In the following clauses, the selection of material processes (in examples 1-4) and causative constructions (in examples 5-7) systematically represents family life as a helping, enabling agent, and men, on their part, as the recipients - the ones w h o stand to gain. (1) Family Life helps (Why Build Your Career Alone? Print) (2) It also provides stability, encouragement and support (Why Build Your Career Alone? Print) (3) It's broadened my horizon (Why Build Your Career Alone? TV) (4) It gives you a direction, a purpose. And most ofall, it gives you afitture.(Your Family is Your Future Print) (5) [...] a happy, well-rounded Family Life makes people wider in their outlook (Why Build Your Career Alone? Print) (6) Family Life has made my life really good (Why Build Your Career Alone? TV) (7) [It has] made me more stable, understanding and less selfish (Why Build Your Career Alone? TV) Men are also represented as active participants, but in clauses with mental pro cesses. These convey men's response to the benefits they have received from having a family. The self-revelatory construals b y means of the mental processes indicate the positive transformation family life has had on these men. (1) I've learnt a ht(Why Build Your Career Alone? TV) (2) And it was only then, as I took her [the baby] into my arms for the very first time and looked down into her tiny perfect face that I realised my whole life had changed (Your Family is Your Future TV) All these personal benefits accrued on account of having a family translates direcdy into men's professional success. Herein lies the instrumentality in men's involvement in family life; it enables the preservation and development of their own career interests. Nowhere is this raised so pointedly as in the ad Why Build Your Career Alone? (Print), where upon oudining the usefulness of family life (see examples of clauses above), the ad closes with the rhetorical question: Isn't that what you need for a successful career? The question presupposes that the answer it> in the affirmative. Another way that family life is construed as having an instrumental effect upon men's careers is that it is shown to imbue the pursuit of a career itself widi greater, tangible meaning. Consider the following excerpt from Your Family is Your Future (TV): Now for the first time, I seem to have a direction and a purpose (cl.7) And I know what I'd been working for all my life (cl.8) Not for money or status (cl.9) I've been workingfor the future (cl.10) '

- •» '

•-

J2.J

/_1

1

t\

LAZAR


405

H e r e , b y m e a n s of the conjunction ' a n d ' (used twice), the instrumental link between family life a n d work life is established. O n the first occasion (clauses 7 and 8), fruits of family life [direction and purpose) are causally linked (by and) to the realization of the significance b e h i n d the pursuit of a career. O n the second occasion (clauses 10 and 11), the n e w realization concerning his career is tangi bly reinforced in terms of the b a b y in his arms. T h e depiction of meaningfulness of one's career in this way, in fact, m a y b e viewed in relation to what scholars such as Burgoyne (1987) have suggested about conventional styles of father hood, namely, that m e n tend to translate their n e w responsibilities as fathers in terms of a n increased c o m m i t m e n t to (paid) work. Women (A) Within the Private Sphere: I n the preceding section, we n o t e d that there w e r e particular executive tasks in the family that almost certainly fell to m e n . Similarly, in this section, we shall see that most routine, practical care-giving tasks are reserved for w o m e n . W h a t is most striking in the conservative discourse of gender relations is the w a y that m o t h e r h o o d is constructed as entailing total devotedness to others, whereas as we saw in the previous section, representat ions of fatherhood, to a large extent, looked to the preservation of the self-interest of men. I n the final part of this section, we shall see h o w other-centeredness is also the operative principle b e h i n d the negotiation of the public a n d private spaces for w o m e n . Firsdy, let us consider the representation of w o m e n within the private sphere as mothers. By 'other-centeredness', I m e a n women's acute consciousness (or con sideration) of their h u s b a n d s a n d their children in the e n a c t m e n t of their motherhood identity. T h e implication to b e drawn from this is that w o m e n derive self-fulfilment indirectiy through the happiness a n d fulfilment of others. Already, in the first instance of expecting a baby, women's experience of preg nancy a n d would-be-motherhood is defined almost entirely in relation to their husbands. Consider the following two scenes from the ad Something Wonderful Happened, where the would-be-mother is represented as responding directiy to h e r husband's evident interest in having the baby. I n the two scenes, the pregnant belly of the w o m a n is the prime object of h e r husband's gaze and actions towards her. In the first scene, the husband's gaze is eye-level with his wife's a b d o m e n as she emerges from a gynaecologist's office. T h e shot cuts to a close-up of the woman's face as she slowly meets h e r h u s b a n d ' s gaze. W h e r e a s his gaze is full of eager expectancy, her look is less certain, pensive, a n d searching. I n other words, her own feelings a n d thoughts about having a child are unclear to us, but the fact that she locks eyes with h i m suggests a willingness to engage with h i m in terms of his evident interest. T h e reactive (as opposed to the active or initiative) position of the w o m a n in relation to h e r h u s b a n d is also manifest in a later scene in the same ad. Triggered by her husband's action of leaning to keenly listen at h e r pregnant belly, the woman responds through a configuration of reciprocal transactional processes, whereby she looks a n d smiles at h i m (reactional processes) and kisses his forehead actional orocessK WTiilcf oil ti^>»

about having the child is left unrepresented. W h a t is represented, however, is h e r j o y as a mother-to-be that is experienced o n account of his paternal interest T h e only time a w o m a n is represented as claiming motherhood as a selfdetermined choice occurs linguistically in Babies and Careers: I want to be a mother. My life would be incomplete without kids of my own. Notwithstanding the reason she gives for wanting children, she is represented here on h e r own terms, without ref erence to h e r husband. Clearly identified in the first person singular (possessive) p r o n o u n , the w o m a n is represented as the Senser, w h o desires motherhood, and explains the consequence that n o t having children would otherwise have on her life. However, o n e clause later, h e r wish gets subsumed within our wish, i.e m a k i n g sure that this is also what h e r h u s b a n d wants: We want someone else to make our lives complete. Ib finally make us a family and bring netn meaning to our lives. T h e shift in p r o n o u n from singular to plural, in other words, signals the similar sort of other-centeredness found in t h e representation of would-be-motherhi >d, as discussed earlier. M y point in earlier discussion is n o t that I find problematic representation.-: of m e n being k e e n o n fatherhood, or that w o m e n would want to consult dieir part ners in parenthood. Such portrayals in themselves are fine; indeed, sharing ind (men's) keen involvement, as w e have seen earlier, m a k e for an egalitarian model of parenting. M y point, rather, is that whereas in these examples the experience of m o t h e r h o o d is directly defined vis-a-vis their husbands, the representation of fatherhood is relatively i n d e p e n d e n t of their wives, since it only focusc-. on their o w n interaction with the children. Further, as I h a v e shown in Something Wonderful Happened, t h e other-centeredness of w o m e n entails a submergence of their o w n dispositions o n becoming mothers within their husbands' expressed desire for fatherhood. W h a t m a y appear as a joint decision b y couples to have children, therefore, m a y belie divergent a n d potentially conflicting views held by each party, and one party m a y b e 'pressured' into having the baby. Burgoyne t i*>87) has found, for example, that m e n in partnerships with w o m e n who are highly career-oriented, or w h o wish to use c o m m i t m e n t to p a r e n t h o o d as a means of cementing an insecure relationship, are m o r e anxious to b e c o m e parents dian their partners, a situation which is b o r n e out in t h e story-line of Something Wonderful Happened I n t h e light of putting to w o r k t h e principle of othercenteredness, I would argue that the 'pressure' placed o n w o m e n m a y also be selfgenerated, thus obliging t h e m to respond 'appropriately'. 3

Other-centeredness applies n o t only to wives' relationships with their hus b a n d s , b u t also to m o t h e r s ' relationships with their children. In the 'Family Life' ads, o n e of the most powerful other-centred gestures mothers can express, is to consider the interest of a n only child b y producing siblings for that child to grow u p with. Consider the following extract from Lonely Child (TV): You may give your child the best things money can buy (cl. 1) But the most precious gift of all is a brother or a sister (cl. 2) The most precious gift you can give your child is a brother or sister (cl. 3)

LAZAR POLITICS OF REPRESENTATIONS

407

What we find in the above is the construction of a 'giving' mother. A giving mother, b y implication, is a G o o d Mother, s o m e o n e w h o has the child's best interest at heart. T h e adversative conjunction in clause 2 suggests to mothers that they may b e somewhat misguided, a n d directs t h e m to choose what is set u p as the ultimate thing that mothers can give to their only child. T h e full impact of the verbal text is best understood in the light of the visual portrayal of the child in the ad. In the visual image, in spite of b e i n g surrounded b y toys of various kinds (cf. clause 1 above), the child is portrayed as aimless and miserable. T h e dis satisfaction of the child is conveyed in two concurrent grammatical structures: although the b o y is in Actor position, holding a toy m o t o r car in one h a n d , h e is at the same time a Reactor in a non-transactional reactional process, w h o looks away from the immediate scene into the mid-distance. Similar to Goffman's licensed withdrawal', the child is physically present in the scene, yet mentally 'absent' from it. His sullen facial expression a n d listless b o d y posture, moreover, suggest that h e is not looking at anything in particular outside the picture frame, but is merely b o r e d a n d u n h a p p y . If the child is u n h a p p y , the implication is that the mother has n o t b e e n other-centred enough, n o t giving enough. I n short, she lias been a Bad Mother. Clauses 2 a n d 3 in the extract (p. 389) are grounded, in this visual context, in which the onus is placed o n m o t h e r s to ensure the happiness of the child by having another baby. From the point of view of the construction of othercenteredness in this discourse, this is a rather interesting expectation: w o m e n are encouraged to h a v e m o r e children - regardless of h o w they themselves might leel (cf. the discussion earlier in relation to men) - out of a sense of maternal duty wards their children. T h e onus on w o m e n is particularly emphatic in the print ersion of this ad: But there's one precious gift, which only you can give - a brother or ister. 'Only' accentuates the obligation w o m e n h a v e as G o o d Mothers to keep [heir child h a p p y b y g i v i n g h i m or h e r a sibling. T h e final segment of the T V ersion of the ad concretely plays this out through the deployment of a transscrional action process, w h e r e b y the mother, the Actor, gives a new-born baby, he Goal, over to the delighted b o y , the Beneficiary. Less dramatically, women are also represented in a range of gender-differentiated asks that emphasize their other-directedness. Visually, this is manifested in trans actional actional structures in which the m o t h e r is the Actor, and h e r family, the < ioal or Beneficiary. For example, mothers are depicted towel-drying children's wet hair, baking cookies for the family, cheering a n d applauding children's efforts at tasks, getting children dressed and preparing t h e m for public events, and taking are of children's safety at outings (Fam, Fam, Fam and Because That's Your Family). The gendered nature of the m u n d a n e care-giving tasks the m o t h e r performs is •..mphasized in contrast to what the father is shown doing at the same time, if copresent. To take o n e example from Fam, Fam, Fam, whilst the m o t h e r is rep resented as watching over the safety of her youngest son at the beach b y holding ,n to his float, the father, although also represented as an Actor, is engaged in m activity entirely different in nature from the basic care-giving function periormed b y the mother. T h e father is shown enacting his 'popular d a d ' role b y entertaining the child (and the child's siblings) b y m a k i n g funny faces at them.

408

DISCOURSE STUDIES

T h e care function performed b y m o t h e r s is not only different from the roles performed b y fathers in the same settings, since m o t h e r s are also sidelined b y the fathers' enactment of their 'fun d a d d y ' role. Unlike the father, the mother is rarely the focus of the children's attention. Instead, she stands at the margin looking in at the b o n d i n g that goes o n between the father a n d the children. The spectator role of the m o t h e r is realized b y transactional reactional structures, w h e r e she is the Senser watching her h u s b a n d , h e r children, a n d the interaction that transpires between those two, without being included in their activities. The implication is that w o m e n derive happiness indirecdy through witnessing the hap piness of others. Further, her o w n care-giving function (e.g. holding o n to the boy's safety float) is subsumed within the father's interaction with t h e children. T h e implication is that the type of care performed b y w o m e n taken for granted as something mothers d o 'naturally', a n d is therefore quite unremarkable. Femi nist scholars (e.g. A n n Oakley, 1974) h a v e long criticized assumptions about the naturalness of motherhood, and the construal of m o t h e r h o o d as being the greatest achievement of w o m e n ' s lives, since these contribute to the perpetuation of w o m e n ' s disadvantaged status in relation to m e n (Lupton and Barclay, 1997). (B) Between Private and Public Spheres: Let us n o w look at the representation of w o m e n i n terms of their negotiation between the public and private spheres of life. O n e of the heartening aspects of the egalitarian discourse, which we saw earlier in the article, was that in encouraging motherhood, the pursuit of a woman's career was n o t overlooked. However, further analysis shows that the negotiation of the public in relation to the private spheres is remarkably dissimilar for women a n d for m e n . W h e r e a s for m e n , the representation was that the two spheres are entirely compatible, for w o m e n , the relationship is fraught with tension. In the following examples from Babies and Careers, the two spheres of life are set up as co-existing contentiously, through a set of presuppositions. (1) Fm really excited about parenthood, but I also love my job. (2) Babies and Careers. Who says you can't have both? (3) How will you divide your time between the kids, housework, and the office? I n Example 1, the use of the adversative conjunction sets u p a presupposition that the two interests are conflictual. I n E x a m p l e 2, such a question can only b e asked based o n the presupposition that according to some discoursal point of view, it is inconceivable or problematic for w o m e n to have babies as well as a career outside the h o m e . A further presupposition that arises from this is that (as in Example 1) babies a n d careers are incompatibly matched. T h e asking of the question in E x a m p l e 3 also conveys presuppositions, n a m e l y that it is an extremely difficult juggling act, and that the n e e d for time m a n a g e m e n t between h e r public and pri vate roles is a concern that is unique to her alone as a w o m a n (since n o indica tion is given of sharing the tasks with a partner). N o t e also the order of the list presented: the kids come first and the office comes last, suggesting an implied order of priority prescribed for w o m e n .

LAZAR


409

T h e fact that these utterances are even m a d e suggests that w o m e n ' s experi ence of having a family a n d a career is markedly different to m e n ' s . T h e assump tions that underlie each of these examples are absent i n the representation of m e n in relation to their career a n d family life. N o w h e r e is it questioned, though, w h y the tension should exist for w o m e n a n d n o t for m e n . Since m e n , as w e h a v e seen, can h a v e it b o t h ways without any problems, the issue clearly is n o t that one's public a n d private identities are i n h e r e n d y irreconcilable. Rather, w h a t is at issue in these representations is the asymmetrical assumptions and expectations in regard to gender. I n constructing a world in which, for w o m e n , the twin pursuits of a career and a family are d e e m e d problematic, the solution offered to w o m e n (and only women) is to self-regulate the two carefully; to strike a balance. N o t e the reiteration of this concern: (1) Oh, I'm sure you'll do very well, Lin. But do balance your career with a family. (Make Room For Love, TV) (2) One ofmy major concerns right now is balancing family and career. But I have friends who have shown me that it can be done. A lot ofSingaporean women are making that choice, too. (Babies and Careers) (3) And along the way, balancing what's best by making the right choices and practical decisions. (Babies and Careers) T h e n e e d for balance, furthermore, is represented in terms of choices and deci sions that w o m e n are obliged to m a k e . This is found in Examples 2 a n d 3 above. Also, it is found in Make Room For Love (TV): There are choices we make today that we'll be living with for the rest of our lives. Ln these three examples, note the reiter ation of the w o r d choice as well as the transitivity structures, in which w o m e n are represented as the Actors w h o (need to) m a k e those choices. Further, note that choice in this conservative discourse does n o t connote freedom or the availability of a range of options. O n the contrary, it suggests that w o m e n are n o t at liberty to pursue a career a n d a family anyway they like, b u t are constrained to select very particular options - which, according to E x a m p l e 3 above, are d e e m e d to be the right choices. W h y is striking a balance a n d m a k i n g the right choice so important w h e n it comes to w o m e n ? W h a t is this 'right' choice? As w e shall see, to balance a career and family does not m e a n giving the two equal weightage. Rather, it is tipped in favour of fulfilling the family role. I n d e e d as the analysis of the ad Too Old will show, the juggling act is construed as necessary in order to ensure that a woman's maternal role is accomplished. I n other words, w o m e n ' s identity as mothers is prioritized for t h e m as their primary identity. T h e cornerstone of m o t h e r h o o d is other-centeredness, and in the Discourse of Conservative Gender Relations, this is something that needs safeguarding. Thus, a career associated with self-fulfilment and self-directedness is regulated in relation to women's prioritized, other centered maternal identity. T h e regulation is achieved via two threats or sanctions o n women: (1) the inability to b e a r (more) children as they get older, and; (2) the implication this has for w o m e n as deficient or Bad Mothers. These sanctions b o t h

DISCOURSE STUDIES

410

draw u p o n a n d reinforce the social disapproval of 'late motherhood'. Let us take each of the sanctions in turn. T h e threat of inability to bear children is explicitiy presented in the ad Too Old. T h e following is an excerpt from the print version of the ad: Because while you're busy building your career (cl.9), or savingfor your dream home (cl. 10), Mother Nature just won't wait (cl.l 1). The older you get (cl. 12), the more difficult it can be to conceive (cl. 13). You mayfindpregnancy a whole lot tougher (cl. 14). And the risk ofsomething going wrong increases (cl. 15). Note that there is a n implied causal relation h e r e that is established linking a w o m a n ' s career building (thematic clause 9) to her getting older (thematic clause 12). T h e ramifications of this are built u p through a combination of lexical choices in clauses 13-15. T h e words more difficult, tougher and risk- accentuated by the comparatives (in the first two instances) and b y the verb increases (that foilows risk) - portray a consistendy bleak, problematic scenario. Such an outcome, more over, is constructed as inevitable, based o n (the personified) representation of Mother Nature as a n Actor following h e r own natural course. T h e threat of a reduced childbearing capability foregrounds the reproductive function of women, a n d uses that as a 'natural' justification for getting w o m e n (and, therefore, not men) to k e e p their careerist motivation in check. Further, the scare tactic em ployed here through a n appeal to a (pseudo-)medical perspective on later mother h o o d ignores a host of other contending, liberatory views o n the subject. These include advances m a d e in obstetric knowledge; the fact that in previous gen erations w o m e n continued to have children until later in life; that in many respects w o m e n (in industrialized countries) are n o w healthier than ever; and the positive view taken b y 'older' w o m e n themselves of b e c o m i n g mothers after 40 (Woollett and Phoenix, 1991: 40, 2 2 0 - 1 ) . It is also the case that the threat of a diminishing childbearing capability is not represented in terms of h o w this would impact on w o m e n themselves, but o n h o w it impacts on others in the family. Unlike representations of men and the family, therefore, the focus is once again fixed o n other-centeredness. In Too Old. the inability to have children is not about the w o m a n b e i n g childless, but about h e r n o t producing m o r e than o n e child. A n d in this respect it is construed as a failure - her failure - to give h e r only son a sibling. Consequentiy, as we see in the following extract from the T V version of the ad, she is ascribed blame and guilt for depriving h i m of a brother or a sister. (The extract involves a boy of about six years old chatting with his friend, a girl.) Boy:

[...] And as for starting a family (14), mom and dad put that off for some more years. (15) And then they couldn't. (16) Girl: They had you, silly. (17) Boy: Eventually. (18) After a lot oftrying. (19) But by that time, mom was getting so old. (20) The doctor said (21) it wasn't advisable to have any more children. (22)

LAZAR POLITICS OF REPRESENTATIONS

Girl: Boy:

That's awfitl! (25) She was too old, you see. (26)

411

,

[•••]

The boy attributes b l a m e for being a n only child squarely o n his m o t h e r : the demonstrative in clause 2 3 - That's why - refers to his m o t h e r being too old to have a safe pregnancy. Interestingly enough, although in clauses 14 a n d 15 h e says that both his m o t h e r a n d father h a d delayed parenthood, w h e n it comes to allocating b l a m e , she alone is singled out. N o t e the shift from the co-ordinate •itructure mom and dad in clause 15 (and the third person plural in clauses 16 and 17) to the singular n o u n mom in clause 20. Subsequently all reference p r o n o u n s are in the third person singular feminine, whilst all traces of the father disappear. I n this way, b l a m e is asymmetrically apportioned. W h a t is m o r e , according to conservative discourses of m o t h e r i n g prevalent in developmental psychology, mother-blaming appears to b e justified. Woollett a n d Phoenix (1991:216) explain that studies in the field m a k e a n expressed link between m o t h e r s a n d children, f his mother's deficiency as a ' G o o d Mother'. Therefore, w h e n she is b l a m e d , this happens because she has 'asked for it'.

Conclusions In this article I h a v e suggested a view of discourse-as-meaning-potential that is loindy realised b y (though n o t hmited to) linguistic a n d visual structures a n d •.trategies. By arguing for the co-presence of two discourses of g e n d e r relations in the domain of p a r e n t h o o d in the ads, therefore, we h a v e two sets of sociohistorically contingent meaning potentials at work at the same time. This is indica tive of the contemporary social order that Singapore finds itself in as it negotiates between competing forces of traditionahsm (hence, the Discourse of Conservative Gender Relations) a n d m o d e r n i s m (hence, the Discourse of Egalitarian G e n d e r Relations). T h e question I n o w want to address in this section is h o w the nego tiation between the two discourses is accomplished in the text(s). Before I d o this, however, let m e recapitulate the analyses undertaken of the two discourses. The Two Discourses in a Nutshell In the Egalitarian Discourse, we saw the representation of identical parenting expressed gender-neutrally in such terms as 'parent' and 'parenthood'); the con struction of the sensitive N e w M a n , a n d the portrayal of the New W o m a n as someone w h o has career c o m m i t m e n t s outside the h o m e . By contrast, we found that in the Conservative Discourse the relationships between the public a n d pri vate spheres, and the involvement within the private sphere itself, were structured in markedly dissimilar ways for m e n a n d for w o m e n . In th*» h r > m o tv™,* ™—

DISCOURSE STUDIES

412

depicted as all fun a n d play. Moreover, there was n o significant tension between their identity as fathers a n d their professional identity at work. T h e two were rep resented as coexisting well, absenteeism on account of their careers was excused, and family life was shown to e n h a n c e their self-interest. A different picture e m e r g e d in the representation of w o m e n as mothers. W o m e n w e r e constantly encouraged to b e other-centred, to the extent of their o w n self-effacement. I n the private sphere, their experience of m o t h e r h o o d was defined in relation to the concerns of their h u s b a n d s a n d children, and they performed gender-differentiated tasks that put the others in the centre, occupying a marginal a n d vicarious role themselves. We also saw that w h e r e a career was concerned, this was represented as causing tension, a n d that w o m e n were advised to strike a balance b e t w e e n a career a n d their families. T h e 'balance', too, was construed unequally, in favour of prioritizing w o m e n ' s family role, with the underlying threat that they might otherwise face personal a n d social censure. T h e principle that u n d e r p i n n e d the analysis of the two discourses was relationality, namely, h o w w o m e n a n d m e n were represented in relation to each other. I n so far as the Egalitarian Discourse was concerned, there was an evening out in terms of representations of gender responsibilities a n d expectations. The N e w M a n a n d the N e w W o m a n representations, for example, worked in tandem to redress g e n d e r stereotypes; fathers h a d a strong presence in the h o m e front which has b e e n traditionally the preserve of w o m e n , whilst w o m e n were por trayed as 'working mothers', disrupting the association of m o t h e r h o o d solely with the private sphere. I n so far as the Conservative Discourse was concerned, the asymmetries in representation c a m e into stark relief only w h e n w e looked through the lens of g e n d e r relationality. If the representations of motherhood a n d fatherhood are considered in isolation, these in and of themselves m a y not b e faulted. However, where relationality becomes the interpretive principle in the analysis, this changes the complexion o n matters. T h e question that gains prom inence is h o w m e n are systematically represented in ways that w o m e n are not. and h o w w o m e n are systematically represented in ways that m e n are not. In other words, the politics of representation is o n e of relative presences as well as of relative absences that are organized along g e n d e r e d lines. This is fundamentally an issue of power, since possibilities for ways of being (or not being) are unequally available to m e n and to w o m e n , a n d the inequity benefits o n e party at the ex p e n s e of the other. 10

The Dynamics between Ae Two Discourses W h a t is the dynamic b e t w e e n the two discourses of g e n d e r relations in the Family Life ads? H o w are the two apparently dissonant discourses co-articulated in the same campaign without jeopardizing a n overall sense of coherence? After all this is a strategic a d campaign launched b y the Singapore government with very specific social a n d d e m o g r a p h i c goals a n d i n t e n d e d outcomes. T h e two discourses, I would argue, are ' m a n a g e d ' in the ads via an implicil strategy of d i s p r o p o r t i o n a t e c o e x i s t e n c e . By this I m e a n that although certain pjralitarian values m a v h e nresent in the ads thev are overwhelmed b v a Drevailine

LAZAR


413

conservative discourse that is simultaneously at work in the same ads, and which, as a result, renders the egalitarian virtues non-threatening. Further, I would argue that the 'brand' of egalitarianism supported in the ads is itself far from subversive. T h a t is, it does n o t seriously u n d e r m i n e or challenge the conservative project, which is w h y the co-existence is even possible. It is a n egalitarianism which can b e a c c o m m o d a t e d a n d / o r displaced b y conservatism. I n the light of this, it is worth revisiting the three representations of egalitar ianism discussed originally. Firstly, the neutering or de-gendering in the use of such terms as 'parent' a n d 'parenthood' requires scrutiny. At first sight, it suggests symmetry a n d the m u t u a l sharing of roles a n d responsibilities. However, as we h a v e seen, in the light of the p r e d o m i n a n t conservative discourse of g e n d e r re lations, the apparent neutrality actually glosses over the unequal, gendered w o r k that continues m u c h as before. Further, the notion of sharedness is a tricky one. It encourages a view of gender role complementarity, w h e r e b y the functions per formed b y fathers and mothers are d e e m e d 'different b u t equal'. Such a view, however, is deeply problematic in that it overlooks the structural arrangements in society that support not just dissimilarity, b u t a rigidly dichotomous a n d hier archical gender order. I n other words, we m a y critically ask what exactiy is m e a n t b y 'sharing', what a n d h o w m u c h is shared, a n d what this sharing entails for w o m e n a n d for m e n . Secondly, it has to b e stated that the N e w M a n is not a subversive construction. Scholars in gender a n d cultural studies h a v e pointed to the disjuncture b e t w e e n the m e d i a rhetoric of the N e w M a n and m e n ' s actual contributions to childcare in reality. This is n o different in the present study. Whilst images of fathers' interest a n d enjoyment of their children in the ads are heart-warming, there is litde evidence to show that there is a significant redistribution of practical childcare responsibilities between fathers a n d mothers. Moreover, in the light of what I have shown in terms of m e n ' s negotiations between the private a n d public spheres, the N e w M a n or N e w Father can easily b e assimilated within the C o n servative Discourse of G e n d e r Relations. T h e N e w M a n is n o t a n antithesis of conservative masculinity, but m a y b e viewed as a hybridized form of masculinity that gets the best of b o t h worlds with little significant cost to m e n . I would call this a politically c o r r e c t (PC) m a s c u l i n i t y , which is geared to appeal to m o d e r n w o m e n without sacrificing the benefits that accrue to conservative masculinity. As we h a v e seen from the analysis, fathers stand to gain personally a n d profes sionally even though their type of involvement in family life is rather limited. An added dimension of P C masculinity is that it is a class indicator. Griswold (1993) (cited in Lupton a n d Barclay, 1997), for example, argues that the N e w M a n is a middle-class p h e n o m e n o n , w h e r e b y this class of m e n are shown u p as m o d e r n and 'enlightened', setting t h e m apart from the crude a n d overfly sexist working class m e n . In the Singapore case, the class argument is construed m o r e b r o a d l y to include educational class. Finally, like the N e w M a n , the N e w W o m a n represented in the ads is insufficientiy threatening to the status quo. A n important characteristic of the N e w Woman is that she wants a m o d e r n career. However, she is not a careerist alone; as we h a v e seen, she is typically represented as s o m e o n e w h o combines a career 11

wit H a f a m t K r T n / - > f V . . — I ~

-1

]

'

1

**"

DISCOURSE STUDIES

414

the N e w W o m a n is g r o u n d e d obligatorily, in part, in m o t h e r h o o d . (It is in the light of the conservative discourse that it becomes clearer that this part of her identity is unequivocally prioritized for h e r as h e r core identity.) Further, the depiction of the N e w W o m a n itself as a careerist-cum-mother is inherendy prob lematic, in that it is based o n rather dubious assumptions. Recall the following example from the egalitarian discourse that read: It's now easier to be successful in both [apaid career and motherhood] (Babies a n d Careers). Eayier'presupposes thai at some point in time it h a d b e e n 'easy' for w o m e n to balance the two, and thai it has n o w b e c o m e even easier. However, one m a y challenge that premise by asking: h a s it ever b e e n easy for w o m e n to m a n a g e the two j o b s successfully? In leaving this unquestioned, however, I suggest that the N e w W o m a n construction blurs with that of a Superwoman. Far from being a Uberating influence, this Superw o m a n assumption loads a further b u r d e n o n to w o m e n . It presents them with an unrealistic ideal of 'having it all' unproblematically, a n d it denies the actual struggles experienced b y countless w o m e n w h o are forced to cope with increas ingly onerous private a n d public responsibilities. A very recent study b y a socio logist in Singapore, in fact, reports that mothers w h o also worked outside the h o m e struggled as a result of the h i g h expectations placed u p o n them, as few duties in the h o m e w e r e shared between couples in dual-career families. (The Sunday Times, 18 September 1999). (Cf. H e n g ' s quote o n page 375; also recall the reasons earlier cited for Singaporean women's reluctance to b e c o m e mothers, or to h a v e fewer children.) 12

I n sum, p a r e n t h o o d evidendy m e a n s different things for fathers and for mothers. T h e analysis has shown that m e n h a v e a lot to gain from becoming fathers. T h e y h a v e the option of being m o r e , or less, involved in childcare; but either way, they are n o t tied definitively to this identity. A n other-centered m o t h r e h o o d , however, is a compulsory identity for w o m e n . It is an identity that has strategic importance for others: for children, for m e n / h u s b a n d s , and for the state. It is little wonder that m o t h e r h o o d thus construed has long b e e n considered b y feminists as oppressive to w o m e n , for the n e e d s of others are m e t at the expense of w o m e n ' s o w n desires, aspirations a n d e m p o w e r m e n t (Koh and Wee. 1987; Lazar, 1993; Soin, 1996). A critical discourse analysis of media repre sentations of p a r e n t h o o d , fatherhood a n d m o t h e r h o o d has shown how the d i s e m p o w e r m e n t of w o m e n is discoursally maintained b y the state in its efforts to boost national fertility rates. A s we h a v e seen, the discoursal mechanics are less t h a n straightforward: although at first glance there appeared to b e two con tending threads of discourse mnrring through the campaign, u p o n closer scrutiny, the two appear to b e b r a i d e d together. I n other words, instead of providing a subversive (emancipatory) current, the discourse of egalitarianism is itself sub verted (and deflected) b y the discourse of conservatism.

Notes 1. An earlier version of this paper entitled 'Gender, Discourse and Semiotics: Asym metrical Constructions of Parenthood' was first presented at a meeting on CDA ai the University of Birmingham (UK), 6-7 April 1999.

LAZAR


415

2. I have named these the 'Family life' ads, based on the fact that the majority were produced by the Family Life Co-ordinating Unit specially setup under the Ministry of Finance. 3. Most of the 14 ads were produced in print and televisual media, which means that the total number of actual texts analysed is around 28. 4. Indeed, the original Women's Charter (1961) description of 'head of household', which had been gender-neutrally worded, was recently amended to ensure that only men could acquire the legal status of being the 'head of the household' and 'principal breadwinner' (Soin, 1996). 5. The depictions support an American study cited in Goffman (1979), which reported that 'the male head of household used the camera most of the time' (Chalton, 1975:94). 6. Kress and van Leeuwen's (1996) grammar does not have 'Beneficiary'. This is a cat egory that I have imported from Halliday's analysis of transitivity structures. 7. The shoulder-hold, although appearing to be an affectionate gesture, is clearly asym metrical (Goflman, 1979). It is also evident between men, where someone senior can put an arm around his junior, but not vice versa. An example of this can be found in the interaction between 'Peter' (the protagonist) and his younger male colleague in Why Build Your Career Alone ?(TV). 8. In Hallidayan linguistic terms, this is an attributive relational process that is intensive, i.e., the attributes describe what the Carrier 'is'. 9. In the 'story' of Something Wonderful Happened, it is the husband who wants the couple to try for a baby, in order to salvage their ailing marriage. They had drifted apart on account of their respective busy careers. 10. The importance of relationality is reflected in the names given to the two discourses: Egalitarian Discourse of Gender Relations, and Conservative Discourse of Gender Relations. 11. See also Chapman (1988), who uses the term 'hybrid masculinity' to refer to an adapted form of masculinity that is better able to retain control in contemporary society. 12. Another reading of the clause may be that it used to be hard for women to balance their twin roles in the past, but that it has now become less difficult to do so. The point, however, remains that the ad glosses over answering another pertinent question, namely how this change has become possible. In keeping silent, the ad leaves unaddressed important social issues regarding the general stasis in gender relations in Singapore society.

References Burgoyne, Jacqueline (1987) 'Change, Gender, and the Life Course' in Gaynor Cohen (ed.) Social Change and the Life Course, pp. 33-66. London: Tavistock. Busfield, Joan (1987) 'Parenting and Parenthood' in Gaynor Cohen (ed.) Social Change and the Life Course, pp. 67-86. London: Tavistock. Chapman, Rowena (1988) 'The Great Pretender: Variations of the New Man Theme' in Rowena Chapman andJonathan Rutherford (eds) Male Order: Unwrapping Masculinity, pp. 225-48. London: Lawrence & Wishart. Foucault, Michel (1972) The Archaeology of Knowledge. London: Tavistock. Goffman. Erving (1979) Gender Advertisements. London: Macmillan. Halliday, M.A.K. (1994) An Introduction to Functional Grammar, 2nd edn. London: Edward

DISCOURSE STUDIES

416

Hill, Michael and Lian, Kwen Fee (1995) The Politics ofNation Building and Citizenship in Singapore. London: Roudedge. Koh, Tai Ann and Wee, Vivienne (1987) 'Editorial: Women's Choices, Women's Lives'. Commentary 7(2/3): 1-4. Kress, Gunther and Van Leeuwen, Theo (1996) Reading Images: The Grammar of Visual Design. London: Roudedge. Lazar, Michelle M. (1993) 'Equalising Gender Relations: A Case of Double-Talk', Discourse & Society 4(4): 443-65. Lazar, Michelle M. (1999) 'Family Life Advertisements and the Narrative of Heterosexual Sociality', in Phyllis G.L. Chew and Anneliese Kramer-Dahl (eds) Reading Culture: Textual Practices in Singapore, pp. 145-62. Singapore: Times Academic Press. Lazar, Michelle M. (in press) 'For the Good of the Nation: "Strategic Egalitarianism" in the Singapore Context', Nations and Nationalism. Dak, Teng Kiat (1994) 'A Declining Birthrate', in Derek da Cunha (ed.) Debating Singapore: Reflective Essays. Singapore: Institute of Southeast Asian Studies. Lupton, Deborah and Barclay, Lesley (1997) Constructing Fatherhood: Discourses and Experiences. London: Sage. Phoenix, Ann and Woollett, Anne (1991) 'Introduction', in Ann Phoenix, Anne Woollett and Eva Lloyd (eds) Motherhood: Meanings, Practices and Ideologies. Sage: London. Quah, Stella R. (1994) Family in Singapore: Sociological Perspectives. Singapore: Times Academic Press. Rich, Adrienne (1980) 'Compulsory Heterosexuality and Lesbian Existence', Signs5(4): 631-60. Rutherford, Jonathan (1988) 'Who's That Man', in Rowena Chapman and Jonathan Rutherford (eds) Male Order: Unwrapping Masculinity, pp. 21-67. London: Lawrence & Wishart. Soin, Kanwaljit (1996) 'National Policies: Their Impact on Women and the Family', in AWARE (collective). The Ties that Bind: In Search of the Modern Singapore Family. Singapore: Armour Publishing. The Straits Times and The Sunday Times. Singapore. Selected issues. Woollett, Anne and Phoenix, Ann (1991) 'Psychological Views of Mothering' and 'Afterword: Issues Related to Motherhood' in Ann Phoenix, Anne Woollett and Eva Lloyd (eds) Motherhood: Meanings, Practices and Ideologies. Sage: London.

Appendix The following is a list of print and televisual ads referred to in this paper, with a brief description on each ad. 1. Make Roomfor Love. Mrs Tan, a senior colleague, advises her younger female colleague on the importance of combining a career with a family. 2. Why Build Your Career Alone? Family life is represented as contributing to a man's career by providing him with invaluable support and stability. 3. Precious Moments. Snapshots are shown of three children (siblings) playing together, while their parents warmly look on. 4. Lonely Child A miserable only child cheers up when his mother brings home a newborn baby - a sibling - for him. 5. Experience The Joy. Married couples are advised on the benefits of starting a family tArliiln

am clill vnnnff

LAZAR


417

6. Fam., Fam, Fam. A family is engaged in a number of leisure activities, with the father prominendy portrayed in his 'fun daddy' role. 7. Because That's Your Family. On the pride and appreciation of two parents as their children surprise them in little ways. 8. Your Family is Your Future. As a young father holds his infant in his arms for the first time, he realizes that from then on his whole life will take on a new meaning. 9. Kids Make Your World Brand New. This presents a collection of photos of many different children, some on their own, others with a parent. 10. Too Old A young boy complains to a friend that he has no siblings because his mother had given priority to developing her career. 11. Kids Make You See. A family represented as enjoying active leisure activities together. 12. Babies and Careers. The personal account of a young career woman, who describes her decision to become a mother. 13. Something Wonderful Happened. The birth of a baby helps to reunite an estranged young married couple. 14. Collage Ad A special compilation of shots on fatherhood, selected from the full series of ads.

46 Reconstructing Topical Sensitivity: Aspects of Face-Work in Talks between Midwives and Expectant Mothers Per LineU andMargareta Bredmar

Sensitivity i n C o n v e r s a t i o n

W

h e n people meet and talk about various topics, whether in institutional encounters or in the private spheres of everyday life, their com munication is n o t a matter simply a n d only of efficient and rational information transfer about anything that might b e present in their minds. Rather, speakers follow a rationality informed b y m o r a l choices: "all verbal behavior is governed b y social n o r m s specifying participant roles, rights a n d duties vis-a-vis each other, permissible topics, appropriate ways of speaking a n d ways of intro ducing information" (Gumperz, 1982, p . 165). A sensitive, or interactionally delicate, topic m a y b e defined as one that cannot b e addressed directly or explicidy b y the speaker without endangering the inter actional h a r m o n y of the encounter b y threatening the listener's face (and there fore also the speaker's own face). Sensitive topics are primarily those with "moral" implications, that is, they often touch u p o n interlocutors' responsibilities for leading their lives in g o o d or bad, acceptable or b l a m e w o r t h y ways. T h e inter actional treatment of morality a n d delicate topics h a v e a strong affinity with issues of mutual face preservation discussed b y Erving Goffman (e.g., 1955,1959, 1983, p . 28). O n e can talk about b o t h a positive face-need (to present oneself a n d to cast others as, e.g., competent a n d socially likeable) a n d a negative faceneed (to want to b e unimpeded). For example, if an expectant mother is invited to have an H I V test, there is an implied threat to h e r negative face b o t h b y the mere fact of the tests a n d b y the impositions that would follow a positive test result.

Source: Research on Language and Social Interaction vol. 29, no. 4, 1996, pp. 347-379.

LINELL A N D BREDMAR

419

RECONSTRUCTING TOPICAL SENSITIVITY

Brown a n d Levinson (1987), arguing largely along Goffmanian lines, pointed to the tension in interaction b e t w e e n the two o p p o s e d poles of positive politeness (being authentic, genuine, involved, immediate, sincere) a n d negative politeness (being considerate, tactful, respectful, discreet). Sensitive, potentially facethreatening topics generally cluster toward the latter, a n d h a v e b e e n found to b e expressed in m o r e indirect a n d implicit ways, with m o r e deference a n d negative politeness, t h a n topics that are m o r e neutral with respect to m o r a l im plications. We use indirectness (of expression) as a cover t e r m for various kinds of "expressive caution" (cf. Silverman, 1994a), defined as any type of deviation from a straightforward ("bald-on-record"), immediate (e.g., nondeferred), expli cit, a n d u n a m b i g u o u s expression of the things a n d issues m e a n t (including their implications). 1

I n health care there are two major kinds of sensitive topics. Both occur in maternal (prenatal) health care ( M H C ) talk a n d are well represented in our data O n e category is lifestyle-implicating topics, that is, those that concern m o d e s of living for which the addressee (normally the patient or client) can b e held re sponsible. I n our case, these are the topics of smoking a n d drinking habits, a n d sexually transmitted diseases. T h e other recurrent category of sensitive topics in health care contexts is those that relate to serious diseases a n d disabilities with strongly intimidating aspects, potentially affecting life a n d death. I n our data, these are primarily issues concerning malformations of the fetus/baby, which, in turn, entail issues of serious m o r a l choice, that is, abortion. Topics such as sex, illness, a n d dying are difficult to address in ordinary con versation, a n d they w e r e found to b e sensitive also for participants in the institu tional contexts of A I D S counseling sessions, as studied b y Perakyla and Silverman (Silverman & Perakyla 1990; Perakyla & Silverman, 1991; Perakyla' 1993; Silverman, 1994a, 1994b). We will see h o w such topics are dealt with in a seem ingly m u c h less threatening context, namely, that of routine visits b y newly pregnant w o m e n to midwives working in primary (preventive) health care units. We argue that the clinical frame m a y b e seen as involving efforts to r e m o v e some of the m o r a l concerns from the delicate topics, a n d yet the participants' actual handling of the topics as "sensitive" reinvokes the m o r a l dimension all over again. 2

Data Our data are drawn from a project o n authentic encounters between midwives, who are qualified nurses working within preventive M H C units, a n d expectant mothers, w h o p a y regular visits to these units. Preventive M H C is a service p r o vided to p r e g n a n t w o m e n in Sweden. T h e purpose of this societal service is primarily to give personal and professional advice a n d assistance to w o m e n , but it naturally also affords health authorities with some opportunities to check and control progression and conduct during pregnancy. It is optional to take advantage of these c o m m u n a l M H C services, b u t a large majority of w o m e n (95% according to national statistics) d o it.

DISCOURSE STUDIES

420

At the time w h e n our data were collected (in the years 1990-1993), a woman could see her midwife as m a n y as 15 times during h e r pregnancy. (Recent cuts in Swedish maternal services have n o w brought the n u m b e r d o w n to about 10.) I n our data, the first visit usually takes place in the 10th or 11th week of pregnancy, and involves a good deal of history-taking and information transfer b y the midwife. This talk is c o m m o n l y called the booking interview (Swedish: inskrivningssamtal). It is followed by a series of visits, the last one taking place a few weeks or a month after delivery. A couple of these appointments also include a n examination by (and talk with) a doctor (either a specialized gynecologist or a general practitioner, the ordinary doctor of the p r i m a r y health care clinic). I n our project, ethnographic fieldwork was conducted at nine M H C centers, and a large n u m b e r of interviews with w o m e n clients a n d with midwives and other personnel were conducted. T h e core data, however, are the tape-recorded (and observed) midwife-pregnant w o m a n encounters. There are two corpora of such data, one of which is used in this article. It consists of 30 booking interviews, that is, the first encounters b e t w e e n the individual w o m e n a n d their midwives. T h e booking interview serves to acquaint the expectant m o t h e r with the midwife a n d with the routines and n o r m s of the M H C . T h e midwife takes in a considerable a m o u n t of information about each w o m a n , a n d this information is entered o n a form with preprinted slots. T h e midwife also communicates infor mation about what is going to h a p p e n at different stages of a "normal" pregnancy, particularly with regard to what the different visits to the M H C center will involve. T h e client receives some printed brochures, a n d times are fixed for some of the future visits. T h e overall phase structure of the b o o k i n g interview is as follows: 3

(a) Greetings (b) M(idwife) asking W(oman) "how she is," a n d the ensuing (usually chit chat type) conversation (c) Information collection: M filling in W ' s case-book, p o s i n g specific questions, and then entering W ' s answers o n the preprinted form (d) Instructing W to call the hospital to fix a date for the upcoming ultrasound examination (to b e d o n e in Week 12) (e) Taking W ' s b l o o d pressure (f) Informing about the A F P test (g) Informing about the routines (opening hours, working schedules) of the M H C center (h) Talking about b l o o d tests to b e taken after the interview (i) Agreeing o n times for later visits to the M H C center (j) Various issues (optionally b r o u g h t u p b y W ) (k) Closing the session 4

T h e order of these subactivities m a y vary slightly. I n particular, the talk on the blood tests (h) m a y sometimes occur earlier, after (d), (e), or (f). T h e topics of the booking interview are largely agenda-bound, that is, they are predetermined, a n d m a n y points correspond to items o n the printed form. The

IJNELL A N D BREDMAR


421

the predefined agenda will therefore provide a frame in which topics, which are sensitive b y m u n d a n e standards, could b e b r o a c h e d in a rather "mechanical" or "bureaucratic" way, along with other nonsensitive topics. We see that, n o n e t h e less, they are treated differendy from these other topics. I n our data, there are three different topics that are almost always b r o u g h t u p in the interviews a n d that m a y b e candidates for sensitivity. T h e s e are (1) a lifestyle issue, the w o m a n ' s smoking and drinking habits; (2) tests for sexually transmitted diseases (syphilis, H I V ) ; a n d (3) a " d r e a d e d " issue implicated b y the A F P test. T h e A F P test is used as a n indicator of serious anomaly or malformation in the fetus, and this topic will then bring u p the morally loaded issue of abortion. These three topics belong to rather different domains, a n d in the booking inter views they belong to different frames and contexts; in terms of the aforementioned phases, they occur in (c), (f), a n d (h), respectively. Accordingly, they exhibit b o t h features that are c o m m o n to all of t h e m a n d other features that distinguish between them. 5

T h e E x p r e s s i o n of Sensitivity Analyses of conversational episodes pertaining to the three topics just mentioned yield a n array of eight features of indirectness a n d mitigation, which we can interpret as signs of sensitivity. T h e s e features range from phonetic a n d prosodic features of delivery over various grammatical, lexical, a n d sequential properties of discourse to aspects of contextual embeddings. We n o w proceed to a systematic account of these observed features of indirectness, a n d illustrate t h e m through the use of selected excerpts from the three domains. T h e s e features are pausing and other perturbations of delivery, reduction of topic p r o m i n e n c e , deferral of topic introduction, limited depth of topic penetration, use of special (mitigating) vocabulary, restrained interactional style, use of neutralizing activity contexts, and use of neutralizing cognitive contexts (frames). Pausing and Other Perturbations of Delivery Pauses, especially if nondeliberate, m a y b e characterized as a perturbation of a s m o o t h delivery of information and a r g u m e n t A s is shown in the literature (Schegloff, 1980; Perakyla, 1993; Weijts, H o u t k o o p , & Mullen, 1993), predelicate sequences often involve such features as hesitations (filled a n d unfilled pauses), a b a n d o n e d utterance attempts a n d r e n e w e d turn beginnings, self-repairs, topicless introductory turns {let's see then), and so on. Such disfluency features also serve to delay the introduction of the delicate issues. I n our data they are characteristic particularly of the syphilis a n d H I V episodes: ^^^k ^^^H ^^^H ^^^P

6

7

Example I (Tema K:BU4: Episode on rubella /syphilis/HIV; M = midwife, W = pregnant woman, in this case expecting her first child) {{M returns to the printedform)) 1 M: ehrm (.) then let's see, (.) German measles, do vou

IM

.

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

W: M: W: M: W: M: W: M:

W: M: W: M: M: W: M: W: M: W: M: W: M:

I,, - i n .

Lii

DISCOURSE STUDIES

no:, I haven't, but on the other hand r l am (.) -i •mm vacc- ^ of course vaccinated "right" we must see to it that the vaccination has taken an too mm mm (.) chrm you could stay away from hugging spotty children "before" *yeah. hh*= =°we know" (.) this answer. (2.0) then we'll take (.) a test, also a blood test (1.0) where one can see (.) that (.) so that you haven't got syphilis (•) I see ((Swedish: jaha)) this, you see, is something we take on everybody, (.) an' I haven't yet found anybody *no:*= =in all these years (1.0) HIY, (0.3) we will test, mm it's optional, d'you want to dp. that? yes, it- "it ca-"= =(that) I take, you know, in the same pricking when I now (.) when I do it on you, right? mm so it won't be anything (.) extra ° no" ((M turns the page, 3.0)) then I talked about ultrasound

In Example 1 the midwife informs the woman about three blood samples. We note that she uses the topic of the rubella test as an introduction to the whole block of tests for which she seeks W's consent. When, in lines 12-14, she moves on to the syphilis test, she does not say, for example, straightforwardly "syphilis we will also test" (let alone, in analogy with lines 1-2, "syphilis, do you know if you have [had] that?"); instead, her approach is slow and stepwise, replete with pauses and with some repetitions. Such perturbations and signs of guardedness are quite typical, and distinguish the syphilis and H I V topics from the adjacent rubella topic. In this case, the word "HIV" is produced in the beginning of a turn (line 22), but it is surrounded by pauses and followed by some verbal material ("we will test"). In the following excerpt, the beginning of a fairly long episode on the AFP test, w e see the midwife producing quite a lot of pauses: 8

Example 2 1 M: "

(TemaK: BUI: Episode on AFP: W expecting her first child) ((previous topic: appointment dates)) mm. (0.7) yes, then I thought I should inform you II J — l l \ c\

T, . , W < u n

/ ^ than

vnn'Il


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


423

(•)

W: M:

W: M:

W: M: W: M: W: M:

W: M: W: M:

W:

yes. and this blood test that will show then if one has heightened values of (.) °alpha-fetoproteins, that is a protein which is found" (.) mm °in the bload". an' that can then, if one has too much of it it CAN mean that there is something (0.5) which is not well with the baby, °right° °mm-hm° an'(.) what above all you see then, that is those s(.) Qpen spinal hernias mm. (.) °mm° one can then also, pjften in combination with this it can be that they don't have (.) that much brain either ODT

.-CO

1 see plus hernias in (2.5) ehrm the abdo.minal wall, that ehrm (1.5) the abdominal wall is not closed so to speak, right 0

O

mm but the bowels °so to speak" mm "come out like this , and-ah (.) that is not that (.) bad actually, you know, but it's only that it's good to know about it. ° yeah ° 0

A

A

Pauses, if consciously planned, m a y express cautiousness a n d guardedness in ap proaching the sensitive topic. Nondeliberate pausing, too, m a y indicate some resistance toward introducing face-threatening topical aspects direcdy. Both of course lead to deferring the direct expression of sensitive aspects. In addition to pauses in introductory sequences, pauses typically occur just before a n d / o r just after the keyterm, that is, the expression most unambiguously naming the sensitive topic (see Table 1). Although it is certainly c o m m o n p l a c e that there is a hitch in speech delivery before keywords (speakers often seem to hesitate a little before finding or deciding on which w o r d to use), pauses a n d hitches are particularly frequent and salient in our data, especially considering the fact that midwives are involved in carrying out routine activities. T h e prekeyterm pauses are p e r h a p s m o r e convincing evidence of the speaker's addressing the keyword as delicate t h a n are postkeyterm pauses, which could b e interactionally driven (solicitation of acknowledgment, reestablishing eye contact, etc.). Yet, a pause after the keyword(s) m a y also b e h e a r d as an attempt to let the intimidating information sink in. O n this point, however, there seems to b e a competing, partly opposed, strategy available, namely, to m o v e o n quickly without dwelling o n the sensitive point (this is addressed further later). Virtually all the midwife's turns in E x a m p l e 2 are replete with pauses. T h e r e are micropauses in the introduction (e.g., E x a m p l e 2, lines 6-9) a n d some longer pauses in the "dreadful" description (Example 2, lines 2 0 - 2 2 , 2 4 , 2 6 ) . A n o t h e r significant feature is M ' s (apparendy)

DISCOURSE STUDIES

424

Table 1: Some properties of midwives' strategies for presenting some topics (incidence in percentages) Topics

Pauses in introductory sequence" Pauses before or after keyterm Prekeyterm utterance segment* Postkeyterm utterance segment" Proxy topic Quick retreat from topic c

1

r

h

AFP (n = 30)

Rubella (n = 21)

27 23 70 50 100 47

24 10 52 57 76« 24

HIV (n = 30) 27" 63 87 60 100 90

Syphilis (n=19) 21 58 63 74 100 95

"A sequence was positively coded if it contained at least two pauses (micropauses or longer). Note, however, that HIV and syphilis topics often do not have an introductory sequence of their own: they are introduced through the "safe" topic of the rubella test (see text). A sequence was positively coded if it contained at least one pause before or after the key expression. Any verbal material except hesitation markers like (Swedish equivalents of) uh, ehrm, and so forth. "Compare Footnote d. This notion is explained in the text under "use of neutralizing cognitive contexts" (p. 369). *In the remaining cases (24%). M asks direcdy about rubella or immunity against rubella. A quick retreat must not contain any additional remark on the topic.

b

C

d

h

Reduction of Prominence O n e strategy for diminishing the face-threat of some topics seems to consist of reducing the perceptual or cognitive salience of their corresponding expressions. For example, lowering the volume to sotto voce, thereby p e r h a p s also switching to another voice quality, m a y b e a w a y of indicating the delicacy of a particular issue. W e find it particularly in the A F P test episodes, w h e n (some) midwives start to lead u p to mentioning particular malformations (as in Example 2, lines 7-8,10). A n o t h e r feature quite frequent in our data is that of syntactic embedding; the use of some verbal material before a n d / o r after the particular delicate or threatening keyterm. To see this point, let us first introduce a n example of an episode that is quite different in that it involves morally relatively neutral topics. Consider the following sequence from the dialogue accompanying the nlling in of the form: Example 3 1 M: 2 3 W: 4 M:

(Tema K: BA5:6. Sequence from information collection by M; W ex pecting her first child) >then I'll go through different diseases an'such.< diabetes, you haven't got that, have you? no. urinary tract or kidney disease?

LINELL AND BREDMAR

5 6 7 8 9 10 11 12 13 14 15 16 17


W: M: W: M: W:

no. epilepsy?

M: W: M: W: M:

heart disease? no lung disease?

o

425

o

no . "high blood pressure?" no (•)

W:

A

A

o

o

•

no any (.) gynecological disease or operation, that you had, hadn't you? yes, there I had an operation ((etc.))

I n contrast to the H I V episodes, the keyterm (usually a n o u n , explicidy iden tifying the topic) is h e r e routinely introduced b y M straight on (and it is, of course, taken as such from the printed form). In most cases, like in E x a m p l e 3, the w o m a n will give short negative answers, which receive n o overt verbal responses from the midwife, w h o , instead, m o v e s o n quickly through the list of items, without m u c h ado. T h e rubella, H I V , and syphilis tests are also three items that m a k e u p a small list (corresponding to a specific subsection on the printed form), and yet there is a m p l e evidence that they are not treated in the same way. Typically, the keyterms H I V a n d syphilis d o not c o m e first in a turn (or even turn-constructional unit); they tend to b e preceded, a n d / o r followed by, other, often rather empty, verbal material. We saw this in Example 1. Example 4 shows something of the same: Example 4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

M.

W M W M W M W M W: M W:

(Tema K: BUS: Episode on rubella/syphilis/HIV; W has children from before) ((Previous topic: tallness and weight. W chats with her little child, M leafs through papers, clears her throat. No talk for about 15 seconds)) ""let's see her-"" "then we'll fill in these blood tests we're gonna take." then I guess they checked up HIY last time you were pregnant? mm. but that's something one repgats. "yeah exactly". you have nothing against that? no. in the same way we check up "syphilis too". yes= =German measles on the other hand, that you were immune against, so that's OK, you know mm so that w e - "we need not" (.) that you can't change anything on that, right? ° no " ((W talks to her child, pause for 7 seconds, then a new topic)) A

A

DISCOURSE STUDIES

426

This episode is a rather exceptional case in terms of its absence of pauses (but note M ' s local use of lowered v o l u m e ; E x a m p l e 4: lines 1-2, 9). Yet, rubella is again treated differently from H I V a n d syphilis. I n this case, however, w e have a w o m a n whose i m m u n i t y to rubella is k n o w n from her previous pregnancy, b u t H I V a n d syphilis are different a n d must b e checked again. Note, though, h o w the rubella test is used as a n exit, a m e a n s to retreat rather rapidly from the topic of testing for the sexually transmitted diseases. T h e r e seem to b e e m b e d d i n g s of sensitive material at several structural levels. At the level of turn-constructional units, we find the use of prekeyterm or postkeyterm utterance segments, which allow the speaker to avoid having the (threatening) keyword (usually a noun) stand out as the perceptually salient first or last w o r d of the utterance. A t a somewhat higher level, w e find the use of a "safe" topic, the rubella test, used as an entrance (as in E x a m p l e 1) a n d / o r an exit (as in E x a m p l e 4) to the topic of venereal diseases. I n the following we encounter other ways of e m b e d d i n g delicate topics in suitably neutral contexts. 9

Deferral of Introduction Some of the aforementioned m e t h o d s serve to defer the introduction of the sensitive topic. A n o t h e r m e t h o d to achieve deferral is the use of presequences (Schegloff, 1980; Heritage, 1984, p p . 265ff.): Example 5 1 M: 2 W: 3 M: 4 W: 5 M: 6 W: 7

(TemaK: BA8: Episode on AFP test; W has children from before) (5.0) AFP, did you take that last time? , no. what are you going to do this time? no, I think I'll stick to this. yes, but then I write abstains here. yeah. (11.0)

I n our data, presequences are characteristic primarily of the A F P topics; for instance, E x a m p l e 2, line 1: "I thought I should inform you," a n d E x a m p l e 5, line 1: "AFP, did y o u take that last time," a n d E x a m p l e 6, line 1, following. T h e syphilis and H I V episodes m a y exhibit similar features (Example 4, line 1: "we'll fill in these blood tests we're g o n n a take"). T h e use of pre-acts, which w e can see as a kind of sequential e m b e d d i n g (cf. the uses of syntactic embedding, discussed in the prior section), m a y serve the purpose of indicating the delicate character of that which is going to b e introduced later o n (Schegloff, 1980, p p . 131 ff.). Perakyla (Silverman & Perakyla, 1990, p . 308; Perakyla, 1993, p . 303) observed this feature in counseling sessions.

L i m i t e d D e p t h of Penetration If the sensitive topic is sometimes approached slowly, indirectly, or gradually,



427

midwife, soon start retreating. This is especially salient in t h e malformation/ abortion issue. A s we noted before, the A F P test is a major m e t h o d in screening pregnant w o m e n for fetal anomalies, a n d it implicates serious m o r a l consider ations. However, if the midwife finds out that the w o m a n knows about the test and knows what stance she will take, the midwife will invariably, as in Example 5 , m o v e on without a n y penetration of the topic at all, thus avoiding the implicated m o r a l issue completely. E x a m p l e 6 is just a little m o r e explicit: Example 6 (TemaK: BU10: Episode on AFP: W has children from before) (4.0) 1 M: AF£, did you take that last time? 2 W: well, that I dan't (know) that was this thing with the c h i r Id having, an open 3 M: L l well u r a l l that 4 W thatJ 5 M; spinal r hernia I 6 W: '•well Wt I took (2.0) I probably took 7 all the tests, ((laughter)) M: (xx) yes, you did. W: mm that was good. 10 M: A yeahA •\ 11 W: 12 M the offer stands of course, mm 13 W: 14 M and one doesn't take it until preg- the sixteenth 15 pregnancy week so there will bethat I want to take ("anyway") 16 W m m 17 M: A yeahA 18 W: 19 M: (.) then you'll get a new appointment when you've 20 been to the doctor's check up 21 W: yes 22 M: so we'll take it in week so it's fairly important that 23 one takes it in week sixteen you know 24 W: mm 25 M so that there won'tA yeahA 26 W 27 M: be any change on that. ((new topic)) I n some cases, the A F P gives rise to a long episode. I n E x a m p l e 2 we saw the start of such an episode; E x a m p l e 7 is almost a direct continuation of the sequence i n E x a m p l e 2: Example 7 (TemaK: BUI; see Example 2) 32 M: hh. ehrm yeah, so that one has then an opportunity to 33 take such a test. 34 W: mm. 35 M: now I should of course also tell you that this is 3fi

„f

„ — * — = t - l

DISCOURSE STUDIES

428

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79

W: M: W: M: W: M: W: M: W: M: W: M: W: M: W: M:

W: M: W: M:

W: M:

W: M: W: M: W: M: W: M: W:

it is, isn't it. yes, it is, so that it is so to speak *no.thing* which (.) no happens often, but it is terribly unusual "yes exactly" y e s . and it is completely optional if you want to take it. (0.3) whereabouts hh d'you take °it°? it's a common blood sample that I (.) take here ((M shows where on her arm the sample is to be taken)) take it here then. mm= =mm. but one takes it in week sixteen. m-hm mm. (0.5) so you can then (.) read through this brochuor ehrm (.) slip at home then, and- ah (0.3) then you can (.) tgll me (.) mm when you come to the doctor then, if you want to take the test "or not". "yes exactly" mm. what I think perhaps is important that is I think that one (.) goes one step further what do I do. if I get a— (.) yes. that one has so to say (.) thought it over a little bit, so that one won't take a test that one actually doesn't hh *want* "mm" (0.3) to know, right mm (1.0) so th- (.) but talk about it at home an'~ mm (2.0) then you'll see. yes m m . that (brochure) you'll have (.) also then.

A

DO

A

A

00

A

((Mgoes on to leaf through papers, then a new topic)) W h e r e a s morally indifferent topics can b e a p p r o a c h e d head-on, the topical episodes of morally sensitive topics are quite different. T h e co-text for the AFP test a n d its implied issues is tests and examinations that the M H C "offers." The footing is clearly that of information delivery, b u t such information could of

LINEIX A N D BREDMAR


429

course b e heard as a (delicate) w a y of providing advice. T h e strategy used b y the midwife usually involves a preparatory action, "have y o u h e a r d about A F P ? " O n l y if this is m e t with a negative answer, or an answer indicating uncertainty, does the midwife go further into the issues. This naturally h a p p e n s m o r e often with w o m e n w h o h a v e h a d babies before. If we generalize across the A F P episodes of the whole corpus, we can discern an approach phase, a central phase, a n d a retreat phase. T h e approach phase is characterized b y indirectness (in the sense of deferral). For example, prepara tory sequences are c o m m o n ; metacommunicative pre-acts (Example 2, line 1: "I thought I should inform you about"; also, "I a m going to ask you about"; " d o you k n o w anything about") are used to signal in advance that a certain topic will b e broached, a certain t y p e of sensitive question will b e posed or sensitive information given, a n d so on. This gives the w o m a n an opportunity to p r e p a r e herself, a n d it gives M (and W) a chance to withdraw promptiy, if possible. I n fact, midwives regularly take t h e chance n o t to g o into the matter if the w o m e n say that they are familiar with the test a n d that they k n o w what they want to d o . Even if the w o m a n is very laconic o n the matter, the midwife accepts this with out p r o b i n g the issue (Example 5). I n most cases, the episode develops into a central phase, defined as the sequence in which parties get maximally close to the (sensitive) topic. I n m a n y cases the central phase is characterized b y limited degrees of concreteness, p r e cision, and depth of penetration. Often there are abstract a n d vague references and anonymizing strategies; the talk is temporarily at a general level, rather t h a n addressed to the individual w o m a n ; c o m p a r e the use of " o n e " (Swedish: man) rather t h a n "you" (Swedish: du): "if o n e has heightened values" (Example 2, lines 6-7), "one can then also" (Example 2, line 17), "one has an opportunity" (Example 7, line 32). Ways of retreating from the sensitive parts are often hinted at, and, if possible, also exploited. I n some cases, the dreadful aspects and m o r a l implications are avoided altogether. Accordingly, we can say that parties, led b y the midwife, prefer to leave the central phase rather quickly a n d m o v e into a retreat phase. T h e retreat can sometimes b e v e r y rapid, with M posing questions in forms that allow for short replies o n the part of W. I n other cases, however, w h e n the intimidating topic has b e e n penetrated, there is a n e e d to neutralize negative a n d threatening aspects, if these h a v e b e e n brought u p . This calls for a longer retreat. I n E x a m p l e 7, M takes s o m e pains to assure W that the malformations are "terribly unusual" (Example 7, lines 36, 40). T h e morally l o a d e d issue d o r m a n t in the A F P episodes is of course that of a possible abortion (medical reasons for abortion becoming relevant if the test indicates increased levels of AFP). Although this issue is explicitly brought u p in those cases in which the midwife goes into describing the possible malfor mations involved, it never h a p p e n s that the midwife asks for a particular stance from the w o m a n (at the stage of the booking interview), n o r do the w o m e n ever overtly speak their m i n d s o n this point.

430

I

DISCOURSE STUDIES

Compared to the A F P episodes, those on smoking and drinking exhibit a rather different form. Let us look at three cases of varying extension, Examples 8-10: Example 8

1 2 3 4 5 6 7 8 9

M: W: M: W: M: W:

Example 9

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

M: W: M: W: M: W: M: W: M: W: M:

W: M: W: M: W: M: W: M:

(Tema K:BU5: Episode on smoking and drinking: W expecting her first child) ((straightfrom previous topic, housing, no pause)) d'you smoke? "no" (3.0) ehrm (.) drink alcohol? °no° (.) ngver?= =°no° (3.0) ((new topic)) (Tema K: BM3: Episode on smoking and drinking: W has children from before) (1.0) well, an'then (.) d'you smpjke by the way? no (0.5) "your alcohol habits?" well, they are *awfully mo_derate* ((laughs)) they are, yes. *yes* once in a while or-? well, now that I am pregnant, then I am naturally completely abstinent, of course you are, right? yes mm. then I put down never then, an'you- >we have talked, haven't we, about your- we talked about it last time< yes >this thing about temporary abstinence and that this< r t h a t is so well worked into-i y e s it would nevit would never occjir to me no, if you don't take (.) headache tablets ((i.e., aspirins)) no then you shouldn't drink anything either no ((M laughs)) no (.) then that is fine.



431

Example 10 (Tema K: BM6: Episode on smoking: W expecting her first child) {{previous topic: W's overweight)) (2.0) 1 M: "yes, that's it° °an'then°, d'you smp_ke? 2 W: yes, unfortunately. 3 M: about how much then? 4 (.) 5 W: "one package per day" 6 M: "one package per day" (0.5) that was pretty much. 7 W: W = 8 M: =mm 9 (0.3) 10 W: I've tried to reduce since I did my test at home then, but I 11 I can't manage you see, I can't dp. it. 12 M: mm. do you feel stressed from (.) feeling that now I must 13 quit, or? 14 W: yes, I do that too. (0.5) yes, it's it's hard, it fge_ls hard 15 'cause you know one (.) one knows, one has heard so much 16 M: mm mm 17 W: rif it isn't-, 18 M: _ an'clearly, =twenty cigarettes is (.) a whole let, you know 19 W: mm 20 (0.3) 21 M: so that (.) if you (.) tried22 (0.3) 23 W: yes, most (.) of all (.) I would of course want to be able 24 to quit once rfor all I 25 M: *• completely straight oyer, r yes •. 26 W: '•so I mean 27 W: I can't keep it (.) somewhere in between for L

J

J

28

rthat I can't ma

29 M: 30 W:

Hhat you can fgel no, no, no, I know that it is like that, "actually" ({discussion goes on for about 30 more turns))

nnage J

I n E x a m p l e 8, M introduces smoking bald-on-record whereas t h e midwives in E x a m p l e 9 a n d E x a m p l e 10 display a few features of indirectness. T h e ques tion in E x a m p l e 9, line 1, is e m b e d d e d between " a n ' then" a n d " b y t h e way." T h e use of " a n d (then)" prefacing in E x a m p l e 9, line 1, a n d E x a m p l e 10, line 1, indicates that the questions have a routine or agenda-based character (Heritage & Sorjonen, 1994), a n d this m a r k i n g m a y somewhat contribute to diminishing m o r a l implicativeness. W e also note t h e lowered v o l u m e used in E x a m p l e 9, line 4, a n d Example 10, line 1; it seems as if the midwife in Example 10 anticipates that t h e w o m a n is a smoker (her voice quality m a y b e a reliable symptom). T h e typical pattern of the smoking a n d drinking episodes, however, is that the midwife poses t h e basic questions rather straightforwardly, without a n y previous pauses

DISCOURSE STUDIES

432

or other perturbations a n d without prekeyterm utterance segments; the keywords are taken direcdy from the form; "do y o u smoke?", "your alcohol habits?" (the latter typically in this elliptical form). If these questions receive clear "nos" from the w o m a n , the topics are typically left without further c o m m e n t s ; E x a m p l e 8 is a case in point. If the w o m a n hesitates just a little before answering " n o , " a few reminders b y the midwife are usually forthcoming. I n E x a m p l e 9, the w o m a n does not hesitate, b u t prefaces h e r response with a weak Swedish ja (here translated as "well" in E x a m p l e 9, line 5) a n d then concedes that she drinks al cohol moderately, before she, after a follow-up question b y the midwife, assures the midwife that she does n o t drink at all during h e r pregnancy (Example 9, lines 9-10). I n such cases, the midwife would still m e n t i o n the desirability of "temporary abstinence" (Example 9, line 17; using an abstract coinage that means, in literal translation, "point abstinence"; Swedish: punktnykterhet) a n d t h e n leave the issue. I n cases like E x a m p l e 10, w h e n the w o m a n admits that she does smoke a n u m b e r of cigarettes p e r day, or that she m a y drink a glass of wine n o w and then, a discussion usually evolves in which the midwife issues the "recommen dations" b y the M H C ("we issue a r e c o m m e n d a t i o n of t e m p o r a r y abstinence"; Swedish: vi gar ut och rekommenderar. . .) a n d expresses some cautious moralizations. T h e general pattern of the episodes is this: first ask for information about W ' s habits, then either leave the topic (if W provides unambiguous, negative answers), or explain what the M H C wants. Midwives are clearly aware that their r e c o m m e n d a t i o n s imply an intrusion into private lifestyle. (On this point, one m a y compare recommendations on other non-lifestyle-related issues, such as the prescription of iron tonics, which, in our data, never involves any face-preserving measures o n the part of midwives.) In some talks, the midwives explicitly mention that they are "aware that they cannot set limits" a n d that they d o n o t want to evoke feelings of guilt with regard to (especially) smoking habits. So, if midwives try to educate a n d control, they do it in a mitigated m a n n e r . T h e desiderata of the M H C with regard to w o m e n ' s conduct are often expressed in terms of mitigated recommendations, rather than straightforward requests. T h u s , the force of prescriptions is d o w n g r a d e d ; a r e c o m m e n d a t i o n of "temporary abstinence" is a fairly limited intrusion into the w o m a n ' s lifestyle (midwives do n o t say "we think y o u should n o t drink at all"). 10

Use of Special (Mitigating) Vocabulary O t h e r strategies for warding off face-threats a n d avoiding embarrassment in clude the use of vague or abstract expressions, circumscriptions and euphemisms, mitigators, attenuations, and various qualified "weak" expressions (cf. Drew, 1992, p . 503). Speakers often prefer such expressions to n a m i n g the things talked about bald-on-record. R a t h e r than b e i n g indiscreet, speakers leave it to their addressees to draw conclusions about things m e a n t a n d implied. For example, Macintyre (1982) m e n t i o n e d gynecologists' use of initiations like "Let's have a look down there" a n d Weijts et al. (1993, p p . 304ff) found the same, as well as an excessive use of D r o n o m i n a l expressions in referring to sexual organs a n d activ-



433

particularly salient in the episode types considered in this article. N o t e however, abstract, a n d somewhat clumsy, coinages like "temporary {point) abstinence." I n asking about drinking habits, midwives regularly use the abstract n o u n "your alcohol habits?" rather than "do you drink anything?" which would h a v e focused o n the personal agent a n d h e r conduct m o r e directly. I n o n e case (not quoted here), in which the midwife first addresses the w o m a n m o r e personally o n the matter, she immediately corrects herself, thus taking immediate measures to re establish the anonymization. 11

Restrained Interactional Style T h e two last paragraphs point to the role of vagueness a n d incompleteness of discourse o n sensitive topics. ' A n o t h e r characteristic of the interactional style is the asymmetrical pattern, with the midwife providing information, without too m u c h penetration, and the expectant m o t h e r responding, often rather minimally. This is actually quite noteworthy, because there are other episodes in the booking interviews w h e r e the two w o m e n interact in rather different ways. Consider a spate of talk like E x a m p l e 11, where the two w o m e n indulge in a m o r e sym metrical exchange: 1 2

13

Example 11 (Tema K:BA5:4-5. Sequence preceding the information collection

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

M: W: M: W: M: W: M: W: M: W:

M: W: M: W: M: W: M: W:

°I see" y e s . (.) so, mentally it has been tough yes, it must be almost even tougher when one (.) keeps trying anyway so to speak yes (0.3) y e a h . that's uh (2.5) no, it's been terrible, I think [(lisht laughter)) A

A

A

A

DISCOURSE STUDIES

434

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

W: so that was the last try, then I thought never anymore. so that this third time I relaxed in some way, I wasM : yes then rit~ '•it must be as it may. this time, that is (0.5) W: I go just 'cause I (.) sh- should do it-i should do •* i t ((laughs)) M: yes yes, but it is kind of nice to have that attitude for r it's a little— yeah, perhaps that's-! why i t W: M : it works, right? W: yeah yeah M: yeah °so that i t - ° W: (0.5) M : 'cause it's a terrible-1 think terribly much lies (0.3) perh- tgrribly much but it can surely be terribly much on the mental plane if W: mm M : how one succeeds W: mm M : and so on W: mm. I think so too. M : so that uh (.) yeah: (0.5) but it's nice W: yeah M : mm. ((M starts asking about housing conditions)) r

1

A

A

A

A

N o t e h e r e features like the u p g r a d i n g in second assessments (Example 11, lines 1,3: "tough," "very tough"; lines 21, 22, 24: "tough," "even tougher," "terrible"), collaborative completion of turns (e.g., lines 3 5 - 3 6 ) , overlapping turns (lines 2 9 - 3 5 ) , s h a d o w i n g (e.g., lines 3 1 - 3 2 ) , a n d s e q u e n c e s of m e r e confirmations (lines 8-10, 3 7 - 3 9 , 49-51). By contrast, in the case of morally delicate topics, the two interlocutors d o not engage in Ughtheartedness and they seldom go into details. Yet, along with paralinguistic strategies for mitigating face-threatening mes sages such as pausing, sotto voce, and perturbations of delivery, o n e might ex pect the use of laughter, which has often b e e n shown to alleviate face-threats (cf. Jefferson, 1985; Adelsward, 1989). Beck and R a g a n (1992) found that patients and nurses often key their talk with laughter and jokes during actual gynecological examinations, and Aronsson a n d R u n d s t r o m (1989) found doctors bracketing sensitive talk b y using ajoking m o d e w h e n they issued lifestyle recommendations to child patients a n d their parents. Neither of these latter p h e n o m e n a is common in our data, however. * 1

Use of Neutralizing Activity Contexts Discourse cannot b e understood without analyzing its contexts. Turning now to contextual resources, w e note the following aspects of indirectness. U n d e r this 15



435

First of all, the institutional encounter itself is of course a particular type of context, partly clinical a n d medical in character, which makes it different from a n everyday-life context. I n part, the institutional context m a y b e regarded as a n e n v i r o n m e n t created partially to m a k e it possible to talk professionally about issues which are not naturally talked about in everyday life. O n the other h a n d , midwives (and other [parajmedical personnel) seem to try to create an informal atmosphere, so as to partially reconstruct a m o r e everyday-life type of conver sational frame. O n e m a y think of the communicative genre as one in which the midwife, together with h e r interlocutor, seek a compromise between a profes sional neutrality a n d a m o r e conversation-like affiliative stance. I n Fairclough's (1992, p . 205) terms, we can talk about the "conversationalization" of a type of institutional discourse. Secondly, o n e m a y exploit particular activity types or phases in the encounters. For example, a m o n g conversational activities one m a y m e n t i o n types of in formation collection being administered b y the professional in going through a form. This generates a kind of highly constrained, routine-like interaction, in which specific questions are often posed in sequences of elliptical formulations that allow for short answers (cf. E x a m p l e 3). With this specific kind of interactional footing, delicate issues might b e smuggled into a "neutral" context (cf. Silverman & Perakyla, 1990, p . 309). I n our case, this is true of smoking and drinking, but n o t of H I V a n d syphilis. A n o t h e r strategy m a y b e to exploit the specific footings of situations in which the medical professional pursues a physical examination of the patient. I n this kind of situation, the patient has already b e e n stripped of some of h e r personal integrity, a n d she is in fact often p a r d y undressed, a n d the doctor or nurse m a y then insert a few pertinent questions and comments, as if it were in the back g r o u n d of the physical manipulations. O u r data for this article do not include any physical examinations, though, except for taking a patient's blood pressure (for other findings from the m o r e comprehensive corpus, see Bredmar, in press). 16

A third, and somewhat related, strategy is simply to sneak in sensitive topics in phases that m a y b e regarded as marginalized in the encounter as a whole. T h e r e is a tendency for the H I V and syphilis test topics to b e relegated to inconspicuous phases, that is, rather late in the encounter, w h e n , one m a y p e r h a p s assume, all important things h a v e already b e e n dealt with. This squares well with the results of Weijts et al. (1993) o n the topic of sex in gynecological consultations. Use ofNeutralizing Cognitive Contexts (Frames) U n d e r this rubric, w e also find a n u m b e r of different p h e n o m e n a . Most of t h e m concern the discursive refiaming ofissues, that is, one tends to bring u p the delicate topics within frames that are not in themselves morally loaded. Thus, we observed h o w H I V a n d syphilis, in our data, are never talked about as diseases, let alone as morally implicative p h e n o m e n a (cf. note 11). Instead, they are always topically framed as targets of blood tests. We could say that the topic of the tests stands p r o x y for the m o r e delicate topic (accordingly, the term proxy topic). T h e issue of anomalies and potential abortion is oblieatorilv introduced thrcmtrh thp nmrhVal

436

DISCOURSE STUDIES

A n o t h e r strategy in the discourse is the d e p l o y m e n t of a n o n y m i z a t i o n ("we," "the M H C ) ; c o m p a r e what D r e w a n d Heritage (1992, p . 30) t e r m e d "the self-referring we [used] to invoke an institutional over a personal identity." A related strategy in our data is the invocation of routines, that is, practices applied to everybody (although the individual w o m a n m a y decline to take the tests involved). In our data, a sensitive issue is sometimes framed as if it were of n o concern for the individual w o m a n , a n d instead something forced u p o n the individuals b y the n e e d to check the state of health in the Swedish population. A characteristic feature of Example 1 is the midwife's immediate comment (line 17) that the test "is something w e take on everybody." T h u s , the testing has nothing to do with any suspicion in the specific case (cf. also E x a m p l e 1, lines 13-14, cast with the negative presumption: "a b l o o d test where one can see that y o u haven't got syphilis").

T h r e e Sensitive D o m a i n s : Similarities a n d Differences T h e three types of sensitive topics are different in fundamental respects. T h e smoking a n d drinking topic is concerned with a lifestyle, which the w o m a n , a n d indirecdy the M H C , can influence, a n d h e n c e it b e c o m e s rational to discuss appropriate conduct during pregnancy. T h e syphilis a n d H I V topics are also potentially lifestyle-related, but little can b e changed here, at least n o t within the scope of the current pregnancy. T h e topical complex of fetal anomalies and possible abortion implied b y the talk o n the A F P test does not involve lifestyle issues; we are faced with something for which the individual cannot b e held responsible. O n the other h a n d , these topics involve a rather frightening future perspective a n d lead to the morally loaded question of what decision the w o m a n would take on the issue of abortion. T h e r e are also other premises a n d conditions that distinguish the discursive treatment of the three topics. T h e context for the smoking a n d drinking issue is a rather early phase in the conversation, in which the midwife goes through the case-book, that is, there is a printed form to b e filled in. T h e syphilis a n d H I V episode, on the other h a n d , sometimes occurs in a late phase of the encounter, as if it was some insignificant matter that the midwife h a d forgotten to deal with before. It is related to the form (the consent b y the w o m a n to take the tests is en tered there) a n d is always introduced in the frame of obligatory tests to b e taken. These differences lead to distinct ways of managing the topics in the dialogues. T h e major patterns can b e summarized as in Table 2. Briefly, we can say that the topics of smoking and drinking are characterized b y an immediate approach, a mitigating vocabulary, a n d a limited depth of penetration. T h e latter aspects are what indicates sensitivity. T h e bald-on-record a p p r o a c h m a y h a v e to do with the fact that the issues involved are routinely raised in m a n y types of health care talk these days. By contrast, H I V a n d syphilis are n o t talked about in most doctor consultations, a n d h e r e t h e y are introduced through a refraining and



437

Table 2: Typical patterns of indirectness and mitigation and their incidence in three kinds of sensitive topics in midwife-pregnant woman talks Smoke/Drink Reduced volume Pausing Perturbations of delivery Syntactic embedding Presequences Limited depth of penetration Mitigating vocabulary Elliptic Q_-A sequence Inconspicuous phase Practical refraining (proxy topic) Anonymization

no no no no no yes yes yes no no yes

HIV/Syph AFP (Malf/Abortion) yes/no yes yes yes yes/no yes no no yes yes yes

yes yes/no yes/no yes/no yes yes" yes/no no no yes yes/no

"Limited depth of penetration occurs whenever possible, that is, when the woman says she knows about the test. delayed approach. T h e same is true of the issue of malformation a n d possible abortion. T h e r e , the presence of presequences is particularly noteworthy. T h e r e are additional c o m m o n features. W i t h respect to morally loaded as pects, there is a general preference for avoidance, a reluctance to detail accounts or to penetrate issues. T h e issues are anonymized; they are talked about as h a n d l e d b y the M H C in general, rather t h a n b y the midwife herself. Downgrad ing also applies to some descriptions ("it is v e r y unusual'' about anomalies, thus reducing the range of applicability of the description). Perakyla (1993) likewise found counselors generalizing a n d anonymizing descriptions a n d prescriptions, a n d mitigating the applicability of dreadful descriptions (upgrading their conditionality: "if it were to h a p p e n . . . " ) . At the same time, such similarities m a y conceal differences. For example, mitigations h a v e a rather different m e a n i n g in prescribing lifestyle as c o m p a r e d to estimating the likelihood of a birth defect. Moral aspects are also typically recontextualized b y the institution. R a t h e r t h a n being couched as m o r a l in nature, they are framed as issues of what is medically appropriate. Sometimes, the medical aspects too are backgrounded, in favor of practical aspects. This pattern is p e r h a p s most clearly shown in the case of the A F P test a n d its background a n d implications. T h e course of progression is nearly always this: 1. First practical aspects: information about h o w a n d w h e n blood tests are taken. 2. T h e n medical aspects: describing the anomalies that m a y b e indicated b y heightened levels of AFP. 3. Last moral aspects: the dilemma of deciding on abortion is explained. This order m a y b e a natural a n d logical one, but it is still a significant fact that the moral issues are backgrounded, deferred, and never brought to a climax.

438

DISCOURSE STUDIES

' "

Conclusion

Sensitivity in conversation has b e e n treated in terms of tact (Leech, 1983), facethreats a n d politeness strategies (Brown & Levinson, 1987), sequential properties of structurally dispreferred responses (especially within Conversation Analysis), a n d so on. H e r e w e relate it to m o r a l implicativeness, a n d we take indirectness to b e its overt expression. Sensitivity is therefore related b o t h to accountability (what the speaker can b e held accountable, or responsible, for; Buttny, 1993) a n d reportability (what topics can b e freely reported on, or not). Indirectness has b e e n shown to h a v e m a n y different shapes. M a n y of them can b e understood in terms of the speaker's distancing himself/herself and his/ h e r interlocutor from the delicate or "dangerous" issues involved. For example, sensitive topics are often gradually approached, only partially penetrated, and quickly retreated from in m a n y episodes. Given that speakers m a y b e said to h a v e several voices or identities available (Bakhtin, 1986; Wertsch, 1991), anonymization m a y b e understood as the speaker's attempt to not speak in his/ her personal identity (rather, the midwife is the representative of the M H C sys tem), a n d to speak as if the topic did n o t direcdy concern the individual other (anonymization, Perakyla's "universalistic approach," 1993, p . 305). Other means to distance oneself from one's own a n d the other's person include the collective construction of depersonalizing situations with less conversation-like footings, for example, those involving the use of technology or the manipulation of the body. A n o t h e r possibility of distancing oneself from sensitive topics lies in the ex ploitation of the participation framework of a multiparty encounter. For obvious reasons this is not available in a dyadic situation like ours. However, Aronsson (Aronsson & Rundstrom, 1989; Aronsson, 1991) has shown, in studies of pediatric consultations featuring a doctor, an allergic child, a n d his/her parent, h o w phys icians m a y address sensitive topics to the child w h e n t h e y are actually targeted at the parent. Perakyla's studies of participation framework in his A I D S counsel ing session data showed that a counseling t e a m could avoid addressing delicate questions to the client direcdy in basically two different ways. Either the cocounselor could request the m a i n counselor to ask the client a sensitive question, thereby relieving the m a i n counselor of the b u r d e n to act as principal and author (in Goffman's, 1981, terminology) of the sensitive question. Another possibility sometimes used was for the counselor to ask the co-client, a close relation ac c o m p a n y i n g the client (a spouse or a girl/boy friend) first to tell her/his views o n topics dealing with the (HIV-positive) client's experience; only then was the latter himself addressed on the same issue (Perakyla & Silverman, 1991). O n e w a y of neutralizing sensitivity is to reconstruct it within n e w frames or contexts (cf. Perakyla, 1993, p . 306). As w e have argued, the establishment of institutional contexts m a y b e taken to function as a m e a n s to provide oppor tunities for talking about sensitive issues in neutral contexts, with neutral pro fessionals having n o personal interests involved. O n e aspect of the treatment of m o r a l issues in such contexts is their recontextualization (and reinterpretation) in the direction of a framing of practical, administrative, or technical consider ations. For example, Adelsward (in press) described this "de-moralization" in interviews with conscientious objectors, and m o d e r n judicial procedures exhibit 17



439

the same p h e n o m e n a (Linell & J o n s s o n , 1991). A similar tendency can b e seen in the present data, for example, in reconstructing sensitive issues in terms of medical appropriateness ("temporary abstinence"), practical business (taking blood samples), and so forth. T h e institutional context in our recorded talks involves several features that seem to warrant a m o r e neutral treatment of sensitive issues. Yet, the actual per formance shows that the issues are still sensitive, with (it seems) their m o r a l im plications lurking below the surface. Indeed, the exploitation of some of the specific contextual features just referred to contribute to m a k i n g the issues stand out as special, that is, as delicate or sensitive. Parties try partly to avoid the mor ally loaded aspects, and in doing just that, that is, exhibiting partial avoidance, they reconstruct the m o r a l loadedness. Bergmann (1992), in his study of "veiled morality" in psychiatric interviews, m a d e a similar point with respect to vagueness; if the psychiatrist describes a n event in an indirect a n d vague way, h e / s h e makes the event "embarrassing, delicate, morally dubious" (1992, p . 154) precisely b y speaking about it in this indirect way. I n a similar vein, Schegloff (1980, p . 143) noted in the discursive treatment of delicate issues that it is "not so m u c h a mat ter of 'masking the delicateness' as of 'doing masking of the delicateness', which is not masking at all." I n our case, the institutional contexts seem to b e arranged in order to recontextualize issues as nonsensitive, and yet parties reconstruct topics as sensitive in a n d through their conversational conduct. Sensitivity b e c o m e s both a precondition for a n d a product of indirectness in discourse. 18

Acknowledgements This article is based on research carried out with Grant Number B91:0058 from The Swedish Research Council for Social Research. The article was read at the Annual Meeting of the American Association for Applied Linguistics in Baltimore, MD, USA, on March 8, 1994. For valuable comments, we thank Viveka Adelsward, Karin Aronsson, John Heritage, Doug Maynard. Deborah Tannen, and Ruth Wodak.

Notes 1. Not all sensitive aspects in discourse are exclusively topical in nature. Some are more endogenous to the discursive interaction as such, in the sense that certain moves or illocutionary actions may be face-threatening in themselves. For example, orders, inquisitive questions, or threats and other imposing moves are often interpersonally delicate, apart from the fact that their topics are often sensitive too. At a general level, Linell and Rommetveit (in press) talked about "discourse-internal ethics" as an intrinsic moral dimension of human discursive interaction. 2. Politeness strategies in health care encounters were treated by Aronsson and Larsson (1987), and Larsson (e.g., Larsson, Johanson, Saljo, & Svardsudd, in press) described in later studies how lifestyle issues, relating particularly to smoking and drinking, tend to be talked about in indirect, vague, or incomplete ways. Aronsson and Rundstrom (1989) pointed to the use of pauses, hedges, diminutions, and other

440

3.

4.

5.

6.

DISCOURSE STUDIES

indirectness types in doctors' presequencing of lifestyle critiques and potentially face-threatening advice. Weijts et al. (1993) described the exercise of "expressive caution" (Silverman, 1994a), represented by delays, avoidances, and depersonal izations, in talk on sexuality during gynecological consultations. Conversational strategies to approach morally implicative issues in health care were also treated in Heritage's studies of health visitors (Heritage & Sefi, 1992; Heritage & Lindstrom, in press). We disregard here the fact that a prior telephone contact occurred in some cases. Also, some of the women had met their midwives during one or several previous pregnancies. The 30 booking interviews are divided among 10 "young" women (ages 17-24), 10 "middle-range" women (ages 25-32), and 10 "older" women (over age 32). Eight of these were first-time mothers, the other 22 had delivered at least one child before (and had also gone to a MHC center before). The other data corpus is longitudinal; 13 women were followed through pregnancy, and in each case six encounters were observed and tape-recorded. These 13 women were all expecting their first child, so the total number of primigravidae in the entire corpus was 21. Six different midwives appeared in the 43 booking interviews. The whole project deals with how the participants construct, or reconstruct, views of the pregnancy, the projected childbirth and the circumstances connected with these events, and how these topics and understandings are differently talked into being in the course of the progressing pregnancy. Issues like solidarity talk between women (in spite of the institutional environment), the role of medical technology as a resource for promoting understanding, and the prevalence of the notion of normality in these talks are treated in other studies. For a full documentation of the entire corpus, see Bredmar (in press). Alpha-fetoprotein (AFP) is produced by the fetal liver and is passed via the amniotic fluid through the placenta into the mother's blood circulation, where it can be measured. The maternal serum AFP screening is not in itself a diagnostic test Other tests (ultrasound, amniocentesis) are used to diagnose suspected fetal anomalies (see Katz Rothman, 1993, p. 236). Other sensitive topics may occur in individual talks, for example, cases of prior abortions, previous problematic deliveries and their circumstances, and issues of malformation and abortion may sometimes surface in other contexts than the AFP test. Likewise, information about problematic marital life or social networks may occasionally make topics like residence and housing conditions (an obligatory topic) morally delicate. However, these types only occur in a few isolated cases. Excerpts are close English translations of Swedish originals. We use standard transcription conventions (such as brackets for overlapping speech, ° ° for sotto voce speech, and (.) for a micropause), but note also the following points: underscore = marks (the stressed syllabic nucleus of) a focally stressed word * * = laughter in the speaker's voice while pronouncing the words enclosed > < = accelerated tempo (relative to surrounding talk) = words pronounced on in-breath — = utterance fades out, usually with a prosody indicating that the speaker leaves it unfinished A

A

7. The Swedish lay term for "rubella" is rbda hund, which is consistendy used in the talks. A literal translation of the term would be "red dog." It is translated here by "German measles." 8. These tests formed part of the screening procedures undertaken on pregnant women when our investigation was carried out. In the last phase of our data collection, syphilis



441

9. The rubella test as an entrance is, not surprisingly, more common with first-time mothers (nth-time mothers are usually known to be immune), but the correlations between discursive strategies and first-/nth-time pregnancy seem otherwise less pronounced in our data than could be expected a priori. 10. In general, our results on this point corroborate those of other researchers, for ex ample, Larsson et al. (1994, in press), who have found that topics like smoking and drinking (and sex-related issues) are usually touched upon only superficially, in spite of their potential importance in primary health care. 11. The discursive treatment of HIV and syphilis (in our data) do not seem to involve euphemisms or mitigated substitutions, but note that both words (HIV, syphilis) are rather clinical in nature. Sex is never mentioned in the episodes involved. Note also that midwives almost invariably talk about "HIV" or "HIV positiveness"; only ex ceptionally do they mention "AIDS," that is, the (dreadful) illness. 12. Vagueness and incompleteness as face-preserving techniques (Brown & Levinson, 1987) have been described for many communicative genres. For example, Markova (1989) analyzed semantic indeterminacy, as well as unfinished sentences and other kinds of incompleteness, in dialogues between patients and therapists as strategies for handling emotionally difficult topics, and Bergmann (1992) showed how euphemistic descriptors are used by psychiatrists to describe events with discretion. 13. It is undoubtedly a significant factor that our data involve only women. Features of indirectness have sometimes been considered as a gender-based indicator of female conversation. Moreover, the two persons of each dyad in our data are, in general, more equal in status than are, for example, a doctor and his/her patient. Because we lack a commensurable corpus of conversations with male doctors (or midwives!) we must, however, abstain from exploring the gender dimension here. 14. The longitudinal corpus includes some instances, as when the topic of sex is raised by the midwife in the postdelivery checkup encounter, for example: TemaK: 1 M: 2 3 4 W: 5 M: 6 W: 7 M: 8 W: 9 10 11 12 M: 13 W: 14 M: 15 W:

B8:6:7 it feels good dow- down thej£?= did you try intercourse? ((The translation "down there" in line 1 is a free one; the Swedish original was i underlivet "in the womb")) =no, we haven't done that. =you haven't done that, no. I forgot how to do it, I say ((laughs)) ((laughs loudly)) *you have to take a course* ((laughs, 2.0)) >he said that, *when we talked about it earlier*, an'then he said,< (0.3) °you can tell her this" we thought of quitting, now that we have got a child. *and* see what she says, ((laughs)) *yes* ((laughs)) yes, right, we (xx) there's no point any longer, you say ((laughs)) *no* ((laughs)) ((shift of topic)) r

L

15. Actually, textual and contextual features are heavily intertwined in discourse, and hence the taxonomy used here is a considerable simplification. We have chosen to be a bit inconsistent in dealing with anonymization, largely a matter of wording, ad dressing it under the heading of "Use of Neutralizing Cognitive Contexts." For a

DISCOURSE STUDIES

442

16. Larsson et al. (1994) mentioned that the issue of self-examination of the breasts, a sensitive issue bordering on other topics with sexual overtones, tends to be broached during the physical examination phase. Interestingly, the related issue of mammography is not as sensitive, judging from Larsson's (admittedly limited) data. Mammography, on the other hand, is associated with technology (see later discussion under "Use of Neutralizing Cognitive Contexts"). 17. Our ethnographic data, however, have documented a case where the midwife ad dresses, in a joking mode, a 6-year-old boy accompanying his (pregnant) mother with the question: "Well, Charlie, how's mother doing on smoking?" 18. Whether this is due to strategic choice or to psychological-emotional factors, or both, is something that can hardly be determined on the basis of our data.

References Adelsward, V. (1989). Laughter and dialogue: The social significance of laughter in institutional discourse. NordicJournal ofLinguistics, 12,107-136. Adelsward, V. (in press). Moral dilemmas, moral rhetoric, and face-work in interviews with conscientious objectors. In J. Bergmann & P. Linell (Eds.), Moral dimensions in dialogue. Aronsson, K. (1991). Facework and control in multi-party talk: A paediatric case study. In I. Markova & K. Foppa (Eds.), Asymmetries in dialogue (pp. 49-74). Hemel Hempstead, England: Harvester Wheatsheaf. Aronsson, K., & Larsson, U. S. (1987). Politeness strategies and doctor-patient communi cation: On the social choreography of collaborative thinking. Journal ofLanguage and Social Psychology, 6,1-27. Aronsson, K., & Rundstrom, B. (1989). Cats, dogs, and sweets in the clinical negotiation of reality: On politeness and coherence in paediatric discourse. Language in Society, 18, 483-504. Bakhtin, M. M. (1986). Speechgenres and other late essays. (V. McGee, Trans., C. Emerson & M. Holquist, Eds.). Austin, TX: University of Texas Press. Beck, C. S., & Ragan, S. L. (1992). Negotiating interpersonal and medical talk: Frame shifts in the gynaecologic exam. Journal ofLanguage and Social Psychology, 11, 47-61. Bergmann, J. (1992). Veiled morality: Notes on indiscretion in psychiatry. In P. Drew & J. Heritage (Eds.), Talk at work (pp. 137-162). Cambridge, England: Cambridge Uni versity Press. Bredmar, M. (in press). Barnmorskesamtalet som social konstruktion [Conversations with midwives as social constructions]. Dissertation in preparation. Department of Com munication Studies, University of Linkoping, Sweden. Brown, P., & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge, England: Cambridge University Press. Buttny, R. (1993). Social accountability in communication. London: Sage. Drew, P. (1992). Contested evidence in courtroom cross-examination: The case of a trial for rape. In P. Drew & J. Heritage (Eds.), Talk at work (pp. 470-520). Cambridge, England: Cambridge University Press. Drew, P., & Heritage, J. (1992). Analyzing talk at work: An introduction. In P. Drew & J. Heritage (Eds.), Talk at work (pp. 3-65). Cambridge, England: Cambridge Univer sity Press. Fairclough, N. (1992). Discourse and social change. Cambridge, England: Polity Press.



443

Goffman, E. (1955). On face-work: An analysis of ritual elements in social interaction. Psychiatry:Journal ofInterpersonal Relations, 18, 213-231. Goffman, E. (1959). The presentation ofselfin everyday life. New York: Doubleday Anchor. Goffman, E. (1981). Forms of talk. Philadelphia: University of Philadelphia Press. Goffman, E. (1983). Felicity's condition. AmericanJournal of Sociology, 89,1-53. Gumperz,J. (1982). Discourse strategies. Cambridge, England: Cambridge University Press. Heritage, J. (1984). Garfinkel and ethnomethodology. Cambridge, England: Polity Press. Heritage, J., & Lindstrom, A. (in press). Motherhood, medicine and morality: Scenes from a medical encounter. In J. Bergmann & P. Linell (Eds.), Moral dimensions in dialogue. Heritage,J., & Sefi, S. (1992). Dilemmas of advice: Aspects of the delivery and reception of advice in interactions between health visitors andfirsttime mothers. In P. Drew & J. Heritage (Eds.), Talk at work (pp. 359-417). Cambridge, England: Cambridge University Press. Heritage, J., & Sorjonen, M.-L. (1994). Constituting and maintaining activities across sequences: ^4n is. and with respect to what we want our theory to do for us, especially where we want to use it to intervene in social processes (as advocated b y critical discourse analysts; see Caldas-Coulthard a n d Coulthard, 1996; Tanks and Ivanic, 1992; Kress. 1996). I take u p a little of this challenge here, drawing on systemic functional linguistics to explore the text in relation to the generalized system of meanings it instantiates. I n addition I consider the roles of text a n d image in a multimodal version of the text (as framed b y two photographs a n d margin notes in Mandela. 1996). The text in question comprises the final six paragraphs of M a n d e l a (1995), which folluw a section break. H e r e I interpret the text generically as a recount, and divide ii into major phases, Orientation Record of Events Reorientation ifor discission of the recount genre see Martin 1985, 1992, 1997; Martin a n d Plum, l!''7l. In addition I h a v e highlighted in small caps the meanings that move the recount forward from one stage to another in Mandela's life (see Appendix for guide to other style features). A

A

1

Example 1 [Orientation] I w i s n. »t born with a hunger to be free. I WAS BORN free - free in every way that I could know. Free to run in the fields near my mother's hut, free to swim in the clear stream

68

DISCOURSE STlJi'*!F>

that ran through my village, free to roast mealies under the stars and ride the b; uJ backs of slow-moving bulls. As long as I obeyed my father and abided by the cusi n< of my tribe, I was not troubled by the laws of man or God. [Record of Events] It was only when I began to learn that my boyhood freedom was an illusion, when I discovered as a young man that my freedom had already been taken from me. thai I began to hunger for it. AT FIRST, AS A STUDENT, I wanted freedom only for my-elf, the transitory freedoms of being able to stay out at night, read what I pleased ami. where I chose, LATER, AS A YOUNG MAN in Johannesburg, I yearned for the basic honourable freedoms of achieving my potential, of earning my keep, of marr and having a family - the freedom not to be obstructed in a lawful life. Birr THEN I slowly saw that not only was I not free, but my brothers and si were not free. I saw that it was not just my freedom that was curtailed, but the f dom of everyone who looked like I did. That is WHEN I JOINED THE AFRICAN NATICONGRESS, and that is when the hunger for my own freedom became the grsjiii hunger for the freedom of my people. It was this desire for the freedom of my people to live their lives with dignity and self-respect that animated my life, thai transformed a frightened young man into a bold one, that drove a law-abiding attorney to become a criminal, that turned a family-loving husband into a man without i home, that forced a life-loving man to live like a monk. I am no more virtuous 01 self-sacrificing than the next man, but I found that I could not even enjoy the pew and limited freedoms I was allowed when I knew my people were not free. Freedom is indivisible; the chains on any one of my people were the chains on all of them, the chains on all of my people were the chains on me. It was DURING THOSE LONG AND LONELY YEARS that my hunger for the freedom of my own people became a hunger for the freedom of all people, white and black, i knew as well as I knew anything that the oppressor must be liberated just as surely as the oppressed. A man who takes away another man's freedom is a prisoner of hatred, he is locked behind the bars of prejudice and narrow-mindedness. I am not truly free if I am taking away someone else's freedom, just as surely as I am noi frw when my freedom is taken from me. The oppressed and the oppressor alike art robbed of their humanity. WHEN i WALKED OUT OF PRISON, that was my mission, to liberate the oppressed and the oppressor both. Some say that has now been achieved. But I know that this is not the case. The truth is that we are not yetfree;we have merely achieved thefreedomto be free, therightnot to be oppressed. We have not taken the final step of our journey, but the first step on a longer and even more difficult road. For to be free is not merely to cast off one's chains, but to live in a way that respects and enhance ! the freedom of others. The true test of our devotion to freedom is just beginning. 1

[Reorientation] i HAVE WALKED THAT LONG ROAD TO FREEDOM. I have tried not to falter; I have roadt missteps along the way. But I have discovered the secret that after climbing a greB hill, one only finds that there are many more hills to climb. I have taken a moment here to rest, to steal a view of the glorious vista that surrounds me, to look back oc the distance I have come. But I can only rest for a moment, for with freedom roiw responsibilities, and I dare not linger, FOR MY LONG WALK IS NOT YET ENDED. (Mandril. 1995: 750-1) Informally speaking, the Orientation phase sets the story in time and place, introducing its m a i n character a n d the habitual behaviours of his early dayi

N THE LOGOGENESIS OF FREEDOM

69

Then the lime line begins to unfold as we transition t h r o u g h the R e c o r d of Events chronicling Mandela's life. Finally, as the events culminate, w e m o v e to a retrospective Reorientation which b o t h summarizes the chronicle and bridges Groin 'present' to 'future' concerns. In calling this text a recount, w e are placing it a m o n g the genres available for the reconstruction and evaluation of experience. By w a y of exploring its social function. Mandela's recount can b e usefully c o m p a r e d to other story genres, such as the following personal recount recorded b y Guenter Plum (1988): Example 2 [Orientation] Well l \ v always wanted one [old English sheepdog; GP]. (Record of Events] And 1 made enquiries for about 12 months, prior. And I was able to get a young bilch. nine weeks old, from a friend. No, it wasn't a friend then; I didn't know her then bui «t turned out to be a good friend, Harriet U. up at Winston Hills. She had this young bitch that she was going to keep herself and she decided that she'd let it go. So >vt- got her. That was our Sophie. And so we started to show her of course. I think herfirstshow was the day she turned 3 months old. She was eligible for it and 1 think she won her class. And, yes, she did win her class that day; it happened to be our championship show, the old English sheepdog championship show. (Reorientation] And wi- just took it from there. Every weekend we brushed h e r . . . off to dog shows, you know, joined the rat race. No, I thoroughly enjoy it. Compared with E x a m p l e 2, M a n d e l a ' s text reads m o r e like history than narrative, since it deals with a m u c h longer time line a n d is especially concerned with its interpretation (the m e a n i n g of freedom). Another useful point of comparison would b e historical recounts of insti tutional experience, in the following example:

J

Example 3 . . . By the mid twenties, big changes took place to meet the needs of the fishing industry and government fisheries managers. The Board began to employ full-time srientifK staff, and Technological Stations were established in Halifax, Nova Scotia, and Prince Rupert, British Columbia. The staff quickly became multi-disciplinary. Al the Biological Stations physics and chemistry were added to zoology in recognition of the importance of understanding fish habitat At the Technological Stations zoolo gists did some initial work on refrigeration offish, but chemists, bacteriologists, and engineers soon took over responsibility for industrial research and development During the depression years the fisheries research and development program was maintained despite constrained budgets. Facilities for volunteer investigators could only be provided to those with independent financing, and the small permanent staff endured a ten-percent reduction of salaries. During the World-War-II years that followed, there was the additional setback of loss of some staff to war service . . . (W.R. Martin, 1991)

C o m p a r e d with E x a m p l e 3, M a n d e l a ' s text is m o r e personal, focussing o n the individual life of the writer; note that there are n o individuals named in E x a m p l e 3, whose m a i n protagonist is the Fisheries Research Board of Canada. We might also compare Mandela's recount with a text from secondary school geography dealing with the destruction of habitats: Example 4 Man has been making animals rare and even extinct for thousands of years, and ow of the main ways man has achieved this is BY THE DESTRUCTION of their habitai. The destruction of a habitat MEANS THAT the vital balance between an animal and lis en vironment is disturbed. In ancient times the destruction of habitat and the extinction of animals was quite small. Since then it has rapidly increased. People began to make more use of machines and industrialisation occurred bringing with it change! which would destroy the face of the earth's environment forever. As the demani grew wood and later coal, supplied the resources needed, this in turn RESULTS B THE DESTRUCTION of forests and habitats. At the same time that industrially lios was taking place humans were settling in new parts of the world. Whenever ihcy setded, nests were cut down and farms established. This destroyed the habiial of many animals. THE EFFECTS OF INDUSTRIALISATION and the need of more land DUE TO THE uwiwra OF POPULATION seriously AFFECTED WILDLIFE and still is today already half the w>rld"i tropical rainforests have already been destroyed or irreversibly damaged. This reekfes ravaging of some of the most amazing habitats on earth MEANS THAT by the year 2000 the destruction will be complete and the world will be without these areas. C o m p a r e d with E x a m p l e 4, Mandela's text is m o r e story than explanauon. since it unfolds largely through setting in time rather than explicit causal conn« tions between o n e event a n d another (note the explicit causality highlight* small caps in the preceding Example). We could continue positioning M a n d e l a ' s recount in this way for v time (see Martin, 1997; Martin a n d Plum, 1997; Rothery and Stenglin, 19^ story genres; a n d Coffin, 1997; Veel a n d Coffin, 1996, o n the genres of hist Cutting short the discussion, w e can place E x a m p l e 1 genetically as auu graphical recount, as opposed to personal recount (Example 2), historical rer (Example 3), historical account (Example 4) and a range of affine but n. closely related narrative and expository genres (see Martin, 1985,1993,19v press, for discussion). Unlike the rest of Mandela's autobiography, howc the text reconstrues a n d interprets a great deal of time in just seven paragraph a n d in this respect it is m o r e like the chronicles of history than the narrai;-.': of everyday life. H o w does M a n d e l a m a n a g e so m u c h time and value in : littie time?

2. A b s t r a c t i n g F r e e d o m M a n d e l a begins his story using relatively concrete language. Participants (people, places and things) are realized as n o u n s ; qualities (descriptive and attitudinal, are realized as adjectives; processes (doings a n d happenings) are realized a verbs; assessments of epistemic and deontic modality are realized as modal

OGENESIS OF FREEDOM

71

verbr, and logical connections are realized as conjunctions. S o m e examples fol low, illustrating the choices M a n d e l a uses to construct his childhood, a n d sym bolizing perhaps in their directness the untroubled nature of his life (labelling from Halliday, 1994: see also Halliday, 1989 [1985], o n spoken a n d written language). 1

participant as Thing (noun) I. fields. hut. stream, village, mealies, stars, bulls, father... quality as Epithet (adjective) free, clear, broad process as Process (verb) was born, to run, to swim, to roast, rise, obeyed . . . assessment as Finite (modal verb) could logical relation as Textual Theme (conjunction) as long as As the recount continues, however, the relation of m e a n i n g to wording b e comes much more indirect (Martin, 1993). Alongside participants realized as n o u n s , we iind processes, qualities a n d m o d a l assessments realized as n o u n s as *>•T t'mB

what each a n d every person in a discussion h a d to say before venturing m y o w n opinion. Oftentimes, m y own opinion will simply represent a consensus of what I h e a r d in the discussion. I always r e m e m b e r the re gent's axiom: a leader, h e said, is like a shepherd. H e stays behind die flock, letting the most n i m b l e go o n ahead, w h e r e u p o n the others follw, not realising that all along they are being directed from behind. (Mandeia, 1995: 25-6)

5. R e c o n t e x t u a l i z a t i o n

So far I have explored M a n d e l a ' s recount with respect to system - especially transitivity, m o o d , theme, grammatical m e t a p h o r , lexical relations, conjunct tion, tense, phase, process type, hierarchy of periodicity, polarity, conunuirj and elaboration. I n this regard I h a v e dealt with intertextuality implicitly, wilt reference to the climate of m e a n i n g i m m a n e n t in our culture (as outlined it Halliday, 1994; Martin, 1992; Matthiessen, 1995). I h a v e also explored ioi« local weather, where it seemed important to deal concretely with intertextualityExamples 2 - 4 to situate the genre; the two quotes from elsewhere in the auto biography a n d Vincent Lingiari's speech to highlight the orality of Mandela ! recount. Along b o t h time scales of analysis, m y approach has been selectiw. reflecting m y reading of the text. As analyses accumulate and results converge, m y reading, I expect, aligns with the reading position naturalized b y the recount H o w might this b e confirmed? O n e tack is to consider closely related texts, a n d for the recount under coa-l sideration here we have the recontextualization published in The IllustraiJla^ Walk to Freedom (Mandela, 1996). T h e last two pages of this edition of the auto-; biography position an elided version of the recount in relation to two iiiiapi a n d margin notes, as outiined in Figure 1. Following Kress and Van Leeuwa (1996), this multimodal presentation consists in b r o a d oudine of two multiniodil triptychs. D o w n the left-hand margin of page 202 (Mandela, 1996) we have thf triptych formed b y an image of a y o u n g b o y a n d the following notes: 1

10

[above photo of y o u n g boy] O n the day of the inauguration I was ov,. r w h e l m e d with a sense of history. I n the first decade of the twentieth century, a few years after the bitter Anglo-Boer war a n d before my ovn birth, the white-skinned peoples of South Africa patched u p their diffr r ences a n d erected a system of racial domination against the dark-skinnd peoples of their o w n land, [below p h o t o of y o u n g boy] T h e structu;: they created formed the basis of one of the harshest, most inhuman--, societies the world has ever known. Now, in the last decade of the rwe;i tieth century, a n d m y o w n eighth decade as a m a n , that system had bee:i overturned forever a n d replaced b y one that recognised the rights and J freedoms of all peoples regardless of the colour of their skin. (Mandela, I 1996:202) I 11

LOCOGENESIS OF FREEDOM

'lex! ' regimes' ,

Ideal

87

[fi lines continuing previous page] te* 'Mandela's recount'

1

tap jlwy; clenched

fist salute'

Given Real text •effects'

Figure I: Outline of the illustrated recount (Mandela, 1996) Across the pages we have the image/verbiage triptych formed b y the abridged recount, with the photo of the y o u n g b o y to the left, a n d a bigger p h o t o (which takes up all of page 2 0 3 a n d spreads over a fifth of p a g e 202) taken from the back of the crowd at Mandela's presidential inauguration, featuring the South African flag. Space precludes a detailed analysis of the two images here. I n brief, the smaller photo, in naturalistic colour, features the h e a d a n d u p p e r torso of a primary school-aged African b o y (medium shot). H e is in the foreground, facing us, at eye-level, but focussing o n something out of frame to our left. His right arm is raised in a clenched fist salute which forms a vector at a 45° angle to the triptych, reinforced b y a parallel stripe o n his sleeve a n d the fact that h e is leaning to our right. T h e b o y is wearing a n orange shirt with white horizontal tfripes on its body, four of which form a vector at 90° to his salute. Behind h i m a a crowd of what appear to b e other school children, out of focus, some of whom are also saluting. The larger photo, in naturalistic colour, is dominated b y a large South African flag (medium shot). It is being held, at a 4 5 ° angle, b y s o m e o n e in a large crowd; the vectors formed b y the flag's design are at 9 0 ° to the flagstaff, towards the lower right-hand corner of the image. T h e flag is at eye-level (since we are sanding higher u p . a few rows back). T h e crowd is overwhelmingly African; we see only .he backs of their heads, out of focus in the immediate foreground, ia focus around the flag, and increasingly out of focus towards the stage. Their gaw f o r m s a unresolved vector towards the centre of the stage, which is outside ifae frame of the photo in the distance to our left. Read in relation to one another, the images involve b o t h similarity a n d con trast As for similarities, the crowds are African; a n d the gatherings are con tracted as political b y the two symbols (the clenched fist salute a n d the flag). I n both images the vectors formed b y gaze are to our left a n d unresolved. T h e vectors formed by the boy's salute a n d striped shirt point in the same directions

88

DISCOURSE STUDIES

as those formed b y the flagstaff a n d the flag's design. Beyond this, the bold black typeface above a n d below the boy's image, along with his orange shirt a n d white stripes corresponds strikingly with the colour a n d design vectors of the South African flag. T h e relationship between the images is further enhanced b y contrasts - the y o u n g b o y vs the m a t u r e crowd; the b o y facing us vs the spectators' backs; the b o y gesturing with his a r m vs the flag waving in a crowd. I take these parallels a n d differences as evidence that we are m e a n t to r e a d one image i n relation to the other, m e d i a t e d b y the recounted text, as the triptych presentation implies. Kress a n d Van L e e u w e n (1996) suggest a reading of polarized images in which the left is glossed as Given a n d the right as N e w a n d this seems to fit the left-right triptych u n d e r discussion h e r e . To the left, as Given, w e have a young individual, gesturing with his a r m ; to the right, as New, we h a v e a symbol of nationhood, aloft in a crowd of m a t u r e spectators. Mediating this relation in the triptych we have M a n d e l a ' s recount, in which a b o y becomes the president of the republic, a n d through which the freedom to play evolves into the freedom to govern (with responsibilities). Such is the nature of the photographic modality that the two images t e n d to linearize as Given a n d N e w the spiral texture of the recount; but, as noted, the colours a n d vectors of the left-hand triptych resonate so strongly with the flag that a spiral m o v e m e n t from youthful solidarity to m a t u r e nationhood is implied. N o t e as well that w e observe the boy, and his peers, a n d are in this sense excluded from the politics constructed b y his salute (which might even b e r e a d as protesting against us, t h o u g h w e are not addressed b y his gaze); with the image of the flag, however, w e are included in the crowd, and aligned with their celebration of Mandela's inauguration. T h u s the triptych reinforces b o t h the enUghtenment a n d engagement motifs of the original monom o d a l recount, with the recount mediating the transition from personal to insti tutional values. 12

T h e t o p - d o w n triptych is also polarized; Kress a n d Van L e e u w e n (1996) suggest the terms Ideal a n d Real for the top-to-bottom opposition, characterizing the Ideal as the generalized essence of the information, the Real as m o r e specific, down-to-earth or practical information. I n this triptych the semantic contrast seems to b e b e t w e e n systems of g o v e r n m e n t a n d their consequences. Above the p h o t o M a n d e l a deals with b o t h the inauguration of a n e w republic and erec tion of the former apartheid regime; below, M a n d e l a notes the effect o n people of first the old regime (harsh a n d inhumane) a n d t h e n the n e w (respect for the rights a n d freedoms of all peoples regardless of the colour of their skin). Linking these messages is the h y b r i d image of the boy, w h o can b e r e a d either as a pro testing victim of old, or a celebrating c h a m p i o n of the n e w ; the ambiguity of the clenched first salute (protest or celebration) thus mediates b o t h facets of the triptych's Ideal and Real, with youthful zeal translating system into practice (see Figure 2). Alongside enhancing the m e a n i n g of Mandela's recount, the interlocking multimodal triptychs also extend it. T h e key m e a n i n g h e r e is the boy's salute, which seems at first blush to contradict Mandela's Orientation: ' I was not bom with a hunger to b e free'. Recall, however, that it was this salute which mediated

MARTIN THE LOGOGENESIS OF FREEDOM

I

now I

then [photo]

'—" then '

now

89

'inauguration' 'system of racial discrimination' ['protest/celebration'] 'harsh, inhumane society' 'rights and freedoms of all peoples'

Figure 2: Mediating ideal and real in the left-hand triptych Ideal with Real, t h r o u g h its ambiguity in the left-hand triptych, a n d which resonated strongly with the flag in the right-to-left o n e . T h e very m e a n i n g which is hard to reconcile with M a n d e l a ' s recount is i n fact foregrounded as the a p e x of the semiotic vortex constructed b y the multimodal text. I suspect that t h e n e w m e a n i n g h e r e has to d o with regeneration. T h e margin notes recontextnalize Mandela's recount with respect to the wider context of South African history (in the first decade of the twentieth century . . . now, in the last decade of the twentieth century). A n o t h e r aspect of this recontextualization is the absence of M a n d e l a himself from the images here, though his presence is implied by the unresolved gaze of the inauguration crowd, b y his 'rebirth' in the p h o t o of the boy, a n d p e r h a p s even b y the unresolved gaze of the boy, since to his right we have the previous page a n d the rest of the book, throughout which Mandela's image figures p r o m i n e n d y . I n each case, the virtual M a n d e l a is realized to the left, as Given, in relation to a range of News (i.e. the flag, the president, the politicized youth, respectively). A r n h e i m ' s reading of the m e a n i n g of diagonal vectors reinforces these points. For A r n h e i m (1982: 107) diagonal vectors such as those formed b y the salute a n d flagstaff are characterized as releasing', while those formed b y the shirt stripes a n d flag are 'holding'. I n these terms, the salute a n d flagstaff point forward to future accomplishments, supported b y the shirt stripes and flag of the youth and nation that will get the work done. 13

As Mandela c o m m e n t s in his margin notes, the apartheid system h a d b e e n 'overturned forever'; in this sense Mandela's journey has ended. T h e interlocking triptychs construct his legacy - one nation, united, u n d e r p i n n e d b y a generation of empowered youth for w h o m the freedom to respect a n d enhance the freedom of others is point of departure. By the time M a n d e l a published the 1996 edition, the long walk has almost ended; the triptychs m a k e way for others to carry on.

6. G r a c e In this article I h a v e tried, from m y own specific reading position, to analyse this instance of discourse in relation to the meanings I have b e e n trained to de code. If allowed a reaction, the term that comes to m i n d is grace, in every meaning of the word. T h e gracefulness with which the recount unfolds, the charm of its rhetoric, the goodwill to all p e o p l e s . . . I cannot h e l p admiring the texture, and the Mandela it construes for m e . I n this kind of reaction I a m n o t alone. Consider

DISCOURSE STUDIES

90

for example the evaluative terms used promotionally o n the covers of the 1995 edition: anger, sorrow, love, joy, grace, elegance, riveting, brilliantly, emotive, compelling, uplifting, exhilarating, epic, hardship, resilience, triumph, clarity, eloquence, burns with the luminosity of faith, invincible, hope, dignity, enthralling; great, indispensable, unique, truly stunning, extraordinary, vivid, unusual, courage, persistence, tolerance, forgiveness, extraordinary, well worth, greatness, epic, struggle, idealism, inspired, cynicism, compulsory. W h a t is the appeal? I suspect we are examining here Mandela's ability to naturalize radical values in terms that disarm rather t h a n confront. Both the recount a n d its multimodal recontextualization p r o m o t e a politics of freedom that involves respecting and enhancing the freedom of others. Put into practice, this involves m o r e than an end to apartheid a n d reconciliation with its perpetrators. Ultimately it involves the reconfiguration of a global economic order which distributes resources so unevenly that it has to b e p r o p p e d u p b y all m a n n e r of unbearable regimes. In a sense then, M a n d e l a is promoting socialism in the n a m e of freedom; h e natural izes a comfortable reading position for those w h o might oppose his aims, and at the same time gives his sympathizers a n inspirational shot in the arm. If discourse analysts are serious about wanting to use their work to enact social change, then they will h a v e to b r o a d e n their coverage to include discourse of this kind - dis course that inspires, encourages, heartens; discourse w e like, that cheers us along. We need, in other words, m o r e positive discourse analysis (PDA?) alongside our critique; a n d this m e a n s dealing with texts we admire, alongside those we dislike a n d try to expose (Wodak, 1996). Let m e close with two m o r e comments. T h e first is that discourse analysis inevitably involves multimodal analysis, even if we begin with language. Trad itionally, the extra-linguistic has b e e n glossed as context a n d (i) ignored; (ii) dis cussed in common-sense terms; or (hi) analysed o n a n interdisciplinary basis with apparatus b o r r o w e d from another discipline (psychology, sociology, phil osophy, cultural studies and the like). T h e alternative explored here is to construe context as a social semiotic, m o d e l it as a system of meanings a n d treat it as add itional texture. Kress a n d Van Leeuwen's (1996) pioneering work in this transdisciplinary project was drawn o n earlier in order to unpack the multi-modal discourse in M a n d e l a (1996). W h e t h e r we adopt an interdisciplinary or transdisciplinary perspective o n discourse analysis is a n important variable. I expect that modelling context as m e a n i n g is the m o r e promising tack for sign-oriented linguists, since it encourages t h e m to explore n e w ground a n d renovate their discipline in ways that are p e r h a p s long overdue. T h e final point I would m a k e is that specific instances of m e a n i n g do matter, alongside our studies of generalized instances, of corpora, a n d of systems as a whole. I n part, this issue is simply a matter of perspective: the m o r e variables we are interested in, the fewer texts we can analyse. So the challenge lies in de veloping frameworks in which system can b e brought into focus without losing contact with text, a n d text can b e explored in detail without losing contact with system. In this article I have leaned towards the instance, in order to focus on 14

15

MARTIN

THE LOGOGENESIS OF FREEDOM

91

the contingency of meanings as they unfold from one step to another in a single text (Halliday, 1987,1993c; Martin, 1995a). This local contingency is critical to understanding the social impact a text m a y h a v e , o n political developments in South Africa, for example, or o n linguistic changes to the recount genre. I n b o t h connections it is important n o t to lose sight of grace in discourse, as outlined here. We h a v e taken a short stroll t h r o u g h m e a n i n g - in order to value a n instance of semiotic weather in relation to a climate of political change. As socially engaged discourse analysts, however, our journeys h a v e just begun. N o doubt this n e w journal will guide us o n our way.

Author's N o t e Other style features in this article: bold type is used for highlighting; italics for citations in running text; small caps for realization of counterexpectation through continuity; underlining for through conjunction; equals sign for elaboration (Halliday, 1994).

Notes i

My thanks to Malcolm Coulthard and his colleagues at the University of Birmingham for the opportunity to present an early version of this article in September 1997, and for iheir insights during and following the presentation; and to Anne Cranny-Francis, Rick ledema, Teun van Dijk and Theo van Leeuwen for their guidance on revisions. 1. Cf. Halliday (1971) who explores the role of grammar symbolizing consciousness in Golding's The Inheritors. 2. Halliday's ideational metafunction has two subcomponents, the experiential and the logical; we are focussing on the experiential here. 3. The Mood adjunct already is part of Mood; it has not been separately labelled here. 4. For Halliday Finite and Predicator conflate in the simple present; so takes appears as both Mood and Residue. 5. I am adopting a minimalist position here, simply taking the last group or phrase of the clause as New, based on an unmarked reading of the clause with the tonic falling on its last salient syllable; parts of what I have called transition could arguably be taken as New in a carefully reasoned clause by clause analysis. '6. Or, undoing the nominalization, I was deluded. 7. Following Halliday (1994), the tense choice is [past in present], the 'what happened still matters now' option. K. Semantically, if not grammatically negative: an illusion = 'not real'. 9. Paul Kelly (1991) comments on his song celebrating Vincent Iingiari's achievements as follows: '"From little Things Big Things Grow" is dedicated to Vincent Lingiarri, the Gurindji stockmen and their families who walked off Lord Vestey's cattle station in 1966 thus initiating a land claim that lasted 8 years. The Whidam government handed back much of the Gurindji country in 1974, Gough Whidam himself pouring dirt into Vincent Lingiarri's cupped hands in a ceremony symbolizing the legal restoration of their lands. From this simple action of walking off in 1966 many consequences flowed'.

92

DISCOURSE STUDIES

10. The elisions are presented in the Appendix; the larger cuts are to paragraphs 3 and 4. Paragraphs 4 and 5 from 1995 (communal and democratic freedom) are collapsed in the 1996 version. 11. This text is taken from pages 747-8 of the 1995 autobiography, a couple of pages before the end of the book. 12. Neither triptych resolves the question of whether the boy's salute is to be read as is a sign of protest or of celebration. 13. The previous pages feature images with a photo of Mandela and his daughter on stage during the inauguration, on top of smaller photos of visiting dignitaries - Fidel Castro, Benazir Bhutto, Yasar Arafat, the Duke of Edinburgh and four unnamed Arabs (in royal Saudi Arabian dress). 14. See also Kress (1997), O'Toole (1994). 15. For discussion see Halliday (1991, 1992a, 1992b, 1992c, 1993a, 1993b), Halliday and Matthiessen (forthcoming), Nesbitt and Plum (1988).

References

Arnheim, R. (1982) The Power of the Center: A Study of Composition in the Visual Arts. Berkeley: University of California Press. Caldas-Coulthard, C. and Coulthard, M., eds (1996) 'Preface', in Text and Practices: Readings in Critical Discourse Analysis, pp. xi-xii. London: Routiedge. Coffin, C. (1997) 'Constructing and giving value to the past: an investigation into second ary school history', in F. Christie andJ.R. Martin (eds) Genre and Institutions: Social Processes in the Workplace and School, pp. 196-230. London: Cassell. Fries, P.H. (1981) 'On the Status of Theme in English: Arguments from Discourse', Forum Linguisticum 6(1): 1-38. [reprinted in J.S. Petofi and E. Sozer (eds) (1983) Micro and Macro Connexity of Texts. Hamburg: Helmut Buske. Gee, J. (1990) Social Linguistics and Literacies: Ideology in Discourses. London: Falmer. Halliday, M.A.K. (1971) 'Linguistic Function and Literary Style: An Inquiry into the Language of William Golding's "The Inheritors'", in S. Chatman (ed.) Literary Style: A Symposium, pp. 362-400. New York: Oxford University Press, [reprinted in M.A.K. Halliday (1973) Explorations in the Functions of Language, pp. 103-40. London: Edward Arnold. Halliday, M.A.K. (1989[1985]) Spoken and Written Language. Oxford: Oxford University Press. Halliday, M.A.K. (1987) 'Poetry as Scientific Discourse: The Nuclear Sections of Tennyson's In Memoriam\ in D. Birch and M. O'Toole (eds) The Functions of Style, pp. 31-44. London: Pinter. Halliday, M.A.K. (1991) 'Towards Probabilistic Interpretations', in E. Ventola (ed.) Func tional and Systemic Linguistics: Approaches and Uses, pp. 39-61. Berlin: Mouton de Gruyter. Halliday, M.A.K. (1992a) 'How do you Mean?', in M. Davies and L. Ravelli (eds) Recent Advances in Systemic Linguistics, pp. 20-35. London: Pinter. Halliday, M.A.K. (1992b) 'Language as System and Language as Instance: The Corpus as a Theoretical Construct', inJ. Svartvik (ed.) Directions in Corpus Linguistics: Proceed ings of Nobel Symposium 82, Stockholm, 4-8 August 1991, pp. 61-77. Berlin: De Gruyter. Halliday, M.A.K. (1992c) 'The Act of Meaning', inJ.E. Alatis (ed.) Georgetown University Round Table on Languages and Linguistics 1992: Language, Communication and Social Meaning. Washington, DC: Georgetown University Press.

MARTIN THE LOGOGENESIS OF FREEDOM

93

Halliday, M.A.K. (1993a) 'Quantitative Studies and Probabilities in Grammar', in M. Hoey (ed.) Data, Description, Discourse: Papers on English Language in Honour of JohnMcH. Sinclair [on his Sixtieth Birthday], pp. 1-25. London: HarperCollins. Halliday, M.A.K. (1993b) Language in a Changing World. Canberra, ACT: Applied Linguis tics Association of Australia. Halliday, M.A.K. (1993c) 'The Construction of Knowledge and Value in the Grammar of Scientific Discourse: Charles Darwin's The Origin of'Species', in M.A.K. Halliday andJ.R. Martin (eds) Writing Science: Literacy and Discursive Power, pp. 86-105. London: Palmer. Halliday, M.A.K. (1994) An Introduction to Functional Grammar. London: Edward Arnold. Halliday, M.A.K. (1998) 'Things and Relations: Regrammaticising Experience as Technical Knowledge', in J.R. Martin and R. Veel (eds) Reading Science: Critical and Functional Perspectives on Discourses of Science, pp. 185-235. London: Routledge. Halliday, M.A.K. and Matthiessen, C.M.I.M. (forthcoming) Construing Experience through Language: A Language-Based Approach to Cognition. London: Cassell. Henderson, W. (1996) Review Article, 'The Struggle is my Life: Nelson Mandela's Auto biography', African Affairs 95(379): 288-93. Hymes, D. (1995) 'Bernstein and Poetics', in P. Atkinson, B. Davies and S. Delamont (eds) Discourse and Reproduction: Essays in Honor of Basil Bernstein, pp. 1-24. Cresskill, NJ: Hampton Press. Janks, H. and Ivanic, R. (1992) 'CLA and Emancipatory Discourse', in N. Fairclough (ed.) (1992) Critical Language Awareness, pp. 305-31. London: Longman. Kelly, P. (1991) PaulKelly & the Messengers Comedy. Sydney: Mushroom Records. Kress, G. (1996) 'Representational Resources and the Production of Subjectivity: Ques tions for the Theoretical Development of Critical Discourse Analysis in a Multicul tural Society', in C. Caldas-Coulthard and M. Coulthard (eds) Text and Practices: Readings in Critical Discourse Analysis, pp. 15-31. London: Routledge. Kress, G. (1997) Before Writing: Rethinking the Paths to Literacy. London: Roudedge. Kress, G. & Van Leeuwen, T. (1996) Reading Images: The Grammar of Visual Design. London: Roudedge. Lingiari, V. (1986) 'Vincent Lingiari's Speech (translated by P. McConvell)', in L.A. Hercus and P. Sutton (eds) in This is What Happened, pp. 312-15. Canberra: Australian Institute of Aboriginal Studies. Mandela, N. (1995) Long Walk to Freedom: The Autobiography ofNelson Mandela. London: Abacus. Mandela, N. (1996) The Illustrated Long Walk to Freedom: The Autobiography of Nelson Mandela. London: Little, Brown and Co. Martin,J.R. (1985) Factual Writing: Exploring and Challenging Social Reality. Geelong, Vic: Deakin University Press, [republished by Oxford University Press, 1989]. Martin, J.R. (1992) English Text: System and Structure. Amsterdam: Benjamins. Martin,J.R. (1993) 'Life as a Noun', in M.A.K. Halliday and J.R. Martin (eds) Writing Science: Literacy and Discursive Power, pp. 221-67. London: Falmer. Martin,J.R. (1995a) 'Interpersonal Meaning, Persuasion and Public Discourse: Packing Semiotic Punch', Australianjoumal ofLinguistics 15(1): 33-67. Martin, J.R. (1995b) 'More than what the Message is About: English Theme', in M. Ghadessy (ed.) Thematic Development in English Texts, pp. 223-58. London: Pinter. Martin, J.R. (1996) 'Evaluating Disruption: Symbolising Theme in Junior Secondary Narrative', in R. Hasan and G. Williams (eds) Literacy in Society, pp. 124-71. London: Longman.

94

DISCOURSE STUDIES

Martin, J.R. (1997) 'Register and Genre: Modelling Social Context in Functional Linguis tics - Narrative Genres', in E. Pedro (ed.) Proceedings of the First Lisbon International Meeting on Discourse Analysis, pp. 305-44. Lisbon: Colibri/APL. Martin, J.R. (in press) 'A Context for Genre: Modelling Social Processes in Functional Linguistics', in R Stainton andj. Devilliers (eds) Communication in Linguistics. Toronto: GREF. Martin, J.R. and Plum, G. (1997) 'Construing Experience: Some Story Genres'./oarna/ ofNarrative and Life History 7(1-4): 299-308. Martin, W.R. (1991) "The Fisheries Research Board of Canada', Bulletin of the Canadian Society of Zoologists 22(3): 36-40. Matthiessen, C.M.I.M. (1995) Lexicogrammatical Cartography: English Systems. Tokyo: International Language Sciences Publishers. Nesbitt, C. and Plum, G. (1988) 'Probabilities in a Systemic-Functional Grammar: The Clause Complex in English', in R.P. Fawcett and D. Young (eds) New Developments in Systemic Linguistics. Vol. 2: Theory and Application, pp. 6-38. London: Pinter. Olson, D. (1994) The World on Paper. Cambridge: Cambridge University Press. Ong, W. (1982) Orality and Literacy: The Technohgizing ofthe Word. London: Methuen. O'Toole, M. (1994) The Language of Displayed Art. London: Leicester University Press. Plum, G. (1988) 'Textual and Contextual Conditioning in Spoken English: A GenreBased Approach', unpublished PhD thesis, Department of Linguistics, University of Sydney. Rothery,J. and Stenglin, M. (1997) 'Entertaining and Instructing: Exploring Experience through Story', in F. Christie andJ.R. Martin (eds) Genre and Institutions: Social Processe in the Workplace and School, pp. 231-63. London: Cassell. Veel, R. and Coffin, C. (1996) 'Learning to Think Like an Historian: The Language of Secondary School History', in R. Hasan and G. Williams (eds) Literacy in Society, pp. 191-231. London: Longman. Whitaker, R. and Sienaert, E., eds (1986) Oral Tradition and Literacy - Changing Visions of the World. Durban: Natal University Oral Documentation and Research Centre. Wodak, R. (1996) Disorders ofDiscourse. London: Longman.

Appendix Wordings from Mandela (1995: 750-1) which are elided from Mandela (1996: 202), by paragraph: [free in every way that I could know.] . . . [near my mother's hut] . . . [and ride the broad backs of slow-moving bulls] . . . [as a student]... [freedom only for myself.]... [as a young man inJohannesburg,] . . . [and honourable]... [- the freedom not to be obstructed in a lawful life]. . . . [not only was I not free, but my brothers and sisters were not free. I saw that] . . . [I am no more virtuous or self-sacrificing than the next man, but I found that I could not even enjoy the poor and limited freedoms I was allowed when I knew my people were not free. Freedom is indivisible; the chains on any one of my people were the chains on all of them, the chains on all of my people were the chains on me.] . . . [as well as I knew anything]... [A man who takes away another man's freedom is a prisoner of hatred, he is locked behind the bars of prejudice and narrow-mindedness.

MARTIN

THE LOGOGENESIS OF FREEDOM

95

I am not truly free if I am taking away someone else's freedom, just as surely as I am not free when my freedom is taken from me. The oppressed and the oppressor alike are robbed of their humanity.]... [Some say that has now been achieved. But I know that this is not the case.] . . . [I have tried not to falter; I have made missteps along the way.]

51 I

Division and Rejection: From the Personification of the Gulf Conflict to the Demonization of Saddam Hussein Luisa Martin Rojo

Le discours, en apparence, a b e a u etre bien p e u d e chose, les interdits qui le frappent revelent tres t6t, tres vite, son lien avec le d e s k et avec le pouvoir. (Foucault, 1971: 12) t

1 i h e analysis I present in this p a p e r is basically of a 'critical' nature, in I Foucault's terms. It aims to discover a n d delimit the m e a n s of exclusion J L a n d appropriation which work within discourse, a n d then reveal the motivation behind t h e m b y m e a n s of analysis. Discourses emerge from society a n d h a v e effects on it as well: they contribute to building u p social networks a n d conceptual frameworks. It is these social and political effects that I want to analyse. This critical view is closely linked with Foucault's conception of discourse. Discourses are practices which follow rules. Discourse production is controlled and structured b y a cluster of procedures that function as a m e a n s to avoid ex ternal threats a n d power. These procedures m a k e discourse opaque. Foucault invites us to 'read' the effects of a variety of discourse forms a n d verify h o w they c o m e to b e a 'representation'. This 'representation' exists because, from the very beginning, every piece of discourse is an interpretation, a n d not a reflection of a state of affairs. Foucault and 'critical linguistics' (Fowler, 1991:231; H o d g e a n d Kress, 1992: 161; for a critical approach in discourse analysis, see also v a n Dijk, 1988 and 1993) are interested in the same type of discourse: a n o n y m o u s discourse, a dis course of a kind which is p r o d u c e d in those situations w h e r e p o w e r relation! 1

2

3

Source: Discourse & Society vol. 6, no. 1,1995, pp. 49-80.

MARTTN ROJO

DEMONIZATION OF SADDAM

97

are most evident - rules within mental hospitals or prisons, for instance, or, i n the case of critical linguistics, the analysis of language in the news. This p a p e r deals with the latter, in particular with the type of discourse that the press pro duced o n the Gulf War, an event that has b e e n considered as o n e of the most outstanding examples of appropriation of discourse. H o w e v e r , I focus o n w h a t might b e called t h e process of 'demonization of S a d d a m Hussein', especially the period just before the war, w h e n this demonization was necessary in order to turn a western ally (against Iran) - or, at least, an ' u n k n o w n quantity' - into an enemy, a n d in this w a y justify the attacks. T h e change in t h e perception a n d presentation of S a d d a m Hussein was especially relevant in Spain, where t h e war was, from the very beginning, unpopular. T h e controversy which surrounded it reminds o n e of the polemic over Spain joining N A T O . O n b o t h occasions, the Spanish position was presented as a chance to join the developed countries and overcome t h e country's traditional isolation. This process implies, first of all, an inclusion p r o c e d u r e , which ensures that the Spanish position does n o t simply respond to U S d e m a n d s , b u t is an ideological option in which all E u r o p e is involved a n d , of course, some A r a b countries. Nevertheless, w h a t m a k e s t h e inclusion process possible is an exclusion procedure which establishes an ideo logical us, opposed to S a d d a m Hussein. 4

In this p a p e r I study b o t h processes - inclusion a n d exclusion - a n d try to delimit the role played b y the latter in the achievement of t h e former. Both are realized through discourse a n d they are carried out b y means of different linguistic and argumentative resources a n d strategies. T h e exclusion is articulated o n two axes: D I V I D I N G , that is, establishing the categories which will b e opposed i n the conflict - us, with several different referents, vs them, or, nearly always, him; a n d REJECTING, that is, segregating, marginalizing, creating a negative image. T h e creation of an e n e m y , them or him, makes it possible to establish, as the o p ponent, a n inclusive subject, we, which is defined simply b y opposition. I n the same way, the construction of a negative image of the e n e m y , using the fact that he flouts a series of established values a n d is given negative features like irration ality or madness, provides the inclusive we, b y opposition, with a positive image. The result is a toe which is indefinite b u t which evokes a set of shared values that remain implicit. Both the subject a n d the values are thus underdefined, b u t despite this the reader absorbs t h e m a n d b e c o m e s part of t h e m . By contrast, i n the enemy c a m p there appears a concrete rival w h o embodies all the anti-values, all the crimes. Discourse brings about a process of categorization, of ordering and normalizing, which generates knowledge about ourselves, a n d about the events we are living. I n this case, a vision of the w a r is created, transmitted a n d consolidated. All this fits into w h a t Lakoff (1992) has referred to as 'the fairy tale of the just war'. This plot brings together all the metaphors Lakoff has examined, and illuminates the mechanisms a n d strategies I h a v e found in the analysis. I n this p a p e r I refer to these three processes: the establishing of subjects, t h e clarifying of values a n d the transmission of a specific vision of the events. I have taken as m y corpus the so-called 'objective' press a n d , in particular, chosen from it o n e of the most prestigious Spanish newspapers: El Pais. T h e style of 'objectivity', b y m e a n s of m o d e r a t e , emotionally controlled language,

98

DISCOURSE STUDIES

close to the standard register, seems to b e a neutral style and to represent 'the ideology of consensus' (see Fowler, 1991; Norton, 1983; Sandell, 1977; Sherer a n d Giles, 1979 for different insights o n formal style). I have chosen to e x a m i n e the 'objective' press because m y aim is to expose the m o r e sophisticated and less explicit mechanisms used to control a n d bias the information given. To this effect, this analysis goes h a n d in h a n d with v a n Dijk's (1991) work o n 'elite' or inhibited racism. I will try to show that, apart from s o m e unique linguistic strat egies a n d resources, exclusion procedures that appear in what we might call the 'elite press' d o not essentially differ from those u n c o v e r e d in the British tabloid newspaper the Sun or elsewhere b y other researchers (see I m b e r t a n d Vidal Beneyto, 1986, for similar conclusions).

Foucault a n d E x c l u s i o n P r o c e d u r e s T h r o u g h discourse, according to Michel Foucault, the n e w rituals of power are exercised. These include the definition of what is usual, habitual a n d expected, as opposed to what constitutes a deviation, an exception. This type of power creates subjects a n d self-knowledge. Self-regulation, normalization, is produced w h e n 'external' forces bring 'interior' self-discipline. I n this case, it is seen both in the construction of a h o m o g e n e o u s us a n d in the exclusion of S a d d a m Hussein b y m e a n s of this construction of what it m e a n s to b e h u m a n . T h e exclusion procedures permit, as we shall see, the cognitive categorization of individuals into groups, which is a way of ordering perception a n d judgement, a way of m a k i n g sense of social processes. T h e m e a n s of these categories, their function a n d their social and psychological significance is established inside the discourse context, and changes when the context changes. This means that the categories are n o t constructed in the m i n d from a process of perception, and only later receive a linguistic label, but, rather, they are built in and b y discourse 'as part of a collective d o m a i n of negotiation, debate, argumentative a n d ideo logical struggle' (Wetherell a n d Potter, 1992: 77). Michel Foucault analyses three different external exclusion procedures in his b o o k L'Ordre du discours (1971), a n d these can also b e seen in the w a y infor mation o n the Gulf War was presented. Prohibition. All through the Gulf War there was a prohibition, given that there was press censorship. T h e fact that it was the military w h o controlled and filtered the information o n the war explains partially h o w pieces of news were selected a n d events were omitted. Division and rejection. T h e e x a m p l e of these c o m b i n e d activities of dividing and rejecting that Foucault studied m o s t is the opposition between reason and madness. This opposition appears in discourses dealing with the Gulf War, and is accompanied b y others like 'ours vs theirs', 'order vs brutality' and, ultimately, 'good vs evil'. 'Division' m e a n s establishing a n inclusive us a n d an exclusive them; in this case, t h e m was just him, because the conflict was embodied in S a d d a m Hussein. Beneath this dual vision lies a m o d e l of society based on the idea of consensus (Fowler, 1 9 9 1 : 1 6 - 2 4 , 4 9 - 5 4 ) . I n the foreground there are two

BTATITTN ROJO


99

opposing groups: those w h o share the accepted values a n d behaviour, a n d those w h o flout them. This dual vision entails a personalization of the conflict. Personalization is o n e of the essential factors formulated b y j o h a n n Gaining and Mari R u g e in their analysis of socially constructed values in news (1973: 62-72). Its functions, as Fowler proposes a n d as our analysis shows, are 'to p r o mote straightforward feelings of identification, e m p a t h y or disapproval a n d to effect a m e t o n y m i c simplification of complex historical a n d institutional pro cesses' (1991: 15). O n c e the two camps are established, the field of conflict (them) and that of consensus (us), 'the ideology of consensus' a n d ethnic prejudices are used to reinforce exclusion or 'rejection'. W h e n they are invoked, an image of S a d d a m Hussein is constructed in which h e plays the stranger, the irrational being, the m a d m a n , the beast, a n d in this w a y he, ultimately, personifies evil. These are the villain's attributes within the fairy tale Lakoff (1992) discusses. Rejection allows us n o t only to identify the enemy, but to give h i m a n answer as well. As a result of an implicit or explicit contrast, s o m e b o d y else's madness makes us rational, civilized h u m a n beings. It is b y opposition that each one's charac teristics can b e delimited a n d typological differences can b e established. Identity is, then, relative, since it is constructed b y m e a n s of contrasts with an external otherness. It is a matter not only of fixing limits - which seems to b e a constant in bourgeois mentality (Subirats, 1977) - b u t also of the projection of our fear, of our own irrationality. I n the case in question, the exclusion of the 'other' gives direction to the reaction, and, m o r e importantly, justifies the war: the m a d m a n is dangerous, h e m u s t b e eliminated - if not, h e m a y impose his irrationality o n the world. This p a p e r focuses on this particular procedure. The opposition 'true vsfalse'. ElPaisknows that the assumption that objectivity is possible is deeply rooted. Discourse tends to b e considered as a reflection a n d not as an interpretation of the state of affairs. I t also knows that it is highly valued. C o n s e q u e n t l y , El Pais gives t h e i m p r e s s i o n of b e i n g a n objective newspaper, a j o u r n a l which seeks for 'truth'. This pretence produces a style in which mitigation, implicitness a n d inferencing play a major role. M y analysis highlights this; the process b y which S a d d a m Hussein is demonized is specially efficient w h e n a discourse is n o t considered to b e an interpretation, m a d e b y a particular ideology, b u t is taken to b e true. As v a n Dijk points out, the perlocutionary or persuasive dimension in news lies 'in the formulation of meanings in such a way that they are n o t merely understood but also accepted as the truth or at least as a possible truth' (1988:83). Rhetorical structure accompanying assertive speech acts enhances beliefs assigned to the asserted propositions in the text. If we do not believe w h a t the other says, we can hardly b e expected to change our opinions. A n y acceptance is a c o m p l e x process which presupposes a minimal degree of coherence of the n e w propositions with the knowledge a n d beliefs w e already have. Otherwise we d o not discard fundamental beliefs built u p during years of understanding, experience a n d action (van Dijk, 1988:83). I n this sense, the demonization of S a d d a m Hussein b e c o m e s convenient a n d necessary to overcome the ideological opposition to the war: establishing a n individual agent of aggressive actions, the collective representations of abnormality, irrationality 5

k I I B I I

100

DISCOURSE STUDIES

a n d madness are evoked. These representations act as principles of organization, structuring the social experience, in this case the perception of the conflict (see REJECTION).

F r o m n o w on, I develop m y analysis of the linguistic and argumentative procedures of division a n d rejection.

D i v i s i o n : T h e Personalization of t h e Conflict A m o n g the procedures used to divide which I illustrate in this p a p e r are: (1) the discourse strategy of intertextuality, which makes it possible to reproduce the ideology of consensus, personalize a n d affirm the existence of two conflicting groups, avoiding any responsibility for the newspaper in the discourse; (2) cognitive models, which underlie the discourse quoted and that of the p a p e r itself, and which permit the construction of the two camps: a m o n g t h e m are categorization, m e t a p h o r a n d m e t o n y m y ; (3) finally, the contrast between actions a n d agents, which organizes the opposition us vs them o n the lexical, syntactic a n d discourse levels, and gives coherence to the discourse. This discourse strategy delimits with clarity b o t h subjects (us and him), and strengthens the process of division a n d demarcation. Intertextuality and the Establishment of the Rival Groups T h e persuasive content of news requires organizing, as v a n Dijk shows, 'in order to b e understood, represented, memorized, and finally believed a n d integrated' (1988: 84; see also 9 6 - 1 0 0 a n d 114-19). Van Dijk points out that one m e a n s of increasing the appearance of truth in news is to resort to evidence from other reliable sources (authorities, respectable people, professionals). I n this way, ElPais systematically quoted authorities in the period just before the war. These quotations were often presented as the topic of the news, appearing in headlines a n d leads (categories w h i c h summarize the content of the text and express t h e semantic macrostructure). Nevertheless, as v a n Dijk claims, a n d as w e will see below, the social hierarchy is r e p r o d u c e d in the rhetorical hierarchy of reliability assumed b y the newspaper. I call this argumentative resource intertexuality. Its study is b a s e d o n the the ories of Bakhtine (1981), rCristeva (1986a, 1986b) a n d D u c r o t (1984) (see also Fairclough 1992, for an appraisal of works b y Bakhtine a n d Fvristeva). For Bakhtine, any utterance is m a d e u p of multiple voices, it is built u p of fragments of other utterances which are m o r e or less explicit a n d complete. T h e speaker m a y give the floor momentarily to o n e of the characters; in the case in question, El Pais shows its preference for George Bush a n d J a m e s Baker. I n this way, the discourse b e c o m e s p a r t of history a n d of society - from the m o m e n t that the text updates, responds to or reworks previous texts. Consequently, it contributes to making history a n d bringing about m o r e transcendental changes. At the same time, t h e discourse lets history a n d society into it - since the text absorbs and is m a d e u p of p r e v i o u s or c o n t e m p o r a n e o u s texts (see Rristeva, 1986a, and

MARTIN ROJO


101

Fairclough, 1992: Ch. 4). All the examples found respond to the field Foucault (1969) calls 'presence' - that is, they evoke fragments of discourse outside the text, either to show agreement with them, because they are considered trans cendental, or to criticize t h e m or to discredit t h e m . This resource entails a personification of the conflict, which produces two different results: (1) the construction of the other; (2) the construction of us. The construction of the other: saddam hussein is Iraq. From the very beginning, El Pais put the b l a m e o n S a d d a m Hussein for the invasion a n d subsequent crisis. In the days before the war started, this accusation was m a d e m o s t frequently b y means of quoting George Bush's statements or discourses. This technique cannot b e understood b y m e a n s of the alleged 'register of truth' the newspaper claims to have, because it allows readers to involve a n d share a particular conception, without apparently stating it explicidy. As a result, b y including multiple voices, by m e a n s of a n explicit polyphony, the most radical views can b e articulated while maintaining the style of objectivity (moral ambivalence): (1) 'Si Sadam no abandona Kuwait incondicionalmente se arriesga a sufrir unas consecuencias devastadoras para el pais.' (5.1.1991: 3) (1) [Quoting James Baker 'If Saddam does not leave Kuwait unconditionally, he will expose his country to devastating consequences.' (5.1.1991: 3) S a d d a m Hussein appears in the semantic roles of agent of the invasion a n d the beneficiary of the retaliations. I n this w a y the first t e r m in the opposition is established. George Bush and J a m e s Baker are the m a i n agents here. I n all these examples of intertextuality (see also (7), (10), (11)), the paper does not paraphrase the discourse - the speakers are all quoted directly. T h e y are not m e n tioned with the p u r p o s e of criticizing t h e m , but, rather, the words of these two politicians seem to constitute in themselves the information, a n d there are n o unguistic signals to show that the author of the article ('locuteur', in Ducrot's terms [1984: C h . 8]) does n o t share the vision of the events transmitted b y these discourses (that is, with the point of view of the 'enunciateur', in Ducrot's terms). By using preformulated information, the journalist presents a coded version of a situation m o d e l , which m a y include, as in this case, 'opinions, attitudes, perspective, a n d other information about the m o d e l of the speaker, as well as specific meanings, hierarchies (topics, schemata), a n d stylistic and rhetorical forms' (van Dijk 1988: 115). I n this sense, the fact that several headlines focus on George Bush's declarations (for example: 'Bush r e c o m m e n d s S a d d a m to o p t for peace', 'Bush declares ...', 'Bush claims ...', 6.1.1991) shows a preference for this source a n d for a particular m o d e l of the event. T h r o u g h George Bush's a n d James Baker's words, the personification of the war is achieved a n d a hierarchy of participants is established. Quotations r e p r o d u c e the different lexical proce dures (see, for instance, taxemes, in example (7)), syntactic procedures (contrast expressed through agents a n d actions, as in examples (10) a n d (11)), a n d prag matic devices (see, for instance, the maintenance of the illocutionary force in (12)), used b y A m e r i c a n diplomacy. All these resources seek to emphasize the

DISCOURSE STUDIES

102

differences in power, status a n d possibilities of control, and reproduce a particu lar m o d e l of the situation. By contrast, and as a consequence of the hierarchy of sources and the different degrees of reliability assumed b y the newspaper, El Pais reproduces neither S a d d a m Hussein's conceptualization of the conflict n o r his rhetorical procedures. T h e presentation of S a d d a m Hussein's vision of the conflict could h a v e m a d e clear h o w m u c h simplification was d o n e in order to m a k e h i m the only person responsible for the conflict, a n d the only agent in it. Nevertheless, in accordance with the exclusion process, S a d d a m Hussein's w o r d s are very rarely quoted and, if they are, they are often paraphrased, framed a n d so discredited - either b y extension (2) or explicitly (3). (2) Las fuerzas armadas estan preparadas para una larga batalla', segiin elpresidente iraqui. (7.1.1991: 3; titular pequefio) (3) Estas soflamaspropagandisticas se subrayaron con l a s . . . . (5.1.1991: 2) (2) 'The armed forces are prepared for a long battle', according to the Iraqi president. (7.1.1991: 3; small headline) (3) These inflammatory speeches-were emphasized b y . . . . (5.1.1991: 2) I n (2), although the words of the President are quoted, the presence of the discourse m a r k e r 'according to' points to the distance the paper ('locuteur', in Ducrot's terms) wants to establish, a n d the fact that it does not identify with the opinion quoted - that is, with the point of view of Saddam Hussein ('enunciateur'). T h e use of 'according t o ' relativizes his vision, so that it does not appear as the words of authority (as in (1), (7), (10), (11)), b u t as an utterance which is questioned b y the journalist, a n d only S a d d a m himself is responsible for it. I n (3) we see h o w S a d d a m Hussein's words are discredited b y lexical m e a n s : the distinction between 'objective discourse' a n d 'inflammatory speeches' is called u p and S a d d a m Hussein's words are classified as 'propaganda'. This distinction is based on an objectivist conception of knowledge a n d language, which has d e e p roots in our culture. O n l y o n one occasion, o n 7 J a n u a r y , was a sentence quoted from S a d d a m Hussein's speech that m a d e it possible to guess his vision of the conflict: (4) . . . la lucha contra 'la corruption, la hegemonia [extranjera] y la injusticia econdmica a escala del mundo y regional.' (7.1.1991: 3) (4) . . . the struggle against 'corruption, [foreign] hegemony and economic injustice at a regional and global scale.' (7.1.1991: 3) I n opposition to the personalization of the conflict (everyone against Saddam) which we find in the declarations of George Bush a n d J a m e s Baker a n d in the discourse of the newspaper itself, S a d d a m Hussein in (4) uses abstract nouns referring to processes, to actions which take o n the category of states. Thus, he presents the war as a n act of rebellion against the established order, an order which prevails thanks to a participant which is n o t mentioned. Examples (4) and (5) do not give us e n o u g h data to k n o w the sort of discourse used b y Saddam Hussein o n the conflict, though they do share a preference for the presentation ™ i'lonlnoical motivation, rather than personalization. A

MARTTN ROJO


103

T h e road which takes us from personalization to other m o r e complex motives based o n economics, politics a n d ideology is, as w e shall see, longer. Personal ization is a simplification and makes it possible to leave the 'real' causes for the conflict unmentioned. I n (2), (3), (4) a n d (5), the words of S a d d a m Hussein a p p e a r e d a m o n g a list of details that r e d u c e d their credibility: for example, the fact that his speech lasted 3 0 minutes whereas George Bush's - delivered the same day - was only 6 minutes long, or the fact that in his speech S a d d a m Hussein repeatedly used words such as 'battle', 'fight', etc. All these details reinforce the categorization of Saddam Hussein's discourse as propaganda. Such a presentation prevents the contrast b e t w e e n the opposing points of view a n d makes S a d d a m Hussein's rationale a p p e a r irrational. (5) ' . . . la batalla es entre Dios y Satan.' (31.1.1991: 6) (5) [Quoting Saddam Hussein]'... this is a battie between God and Satan.' (31.1.1991:6) By quoting the speeches of George Bush a n d S a d d a m Hussein, delivered the same day (7.1.1991), El Pais n o t only increases the appearance of truth but also produces an effect of factuality a n d dramatization: a dialogue b e t w e e n the two leaders. Nevertheless, b y treating b o t h discourses in a different way, the paper confers superiority to o n e interlocutor over another, which results in a n asymmetrical dialogue. In editorials, w h e r e quotations are n o t used, a n d in some columns as well, it is the newspaper itself that contributes to the process of personalizing a n d estab lishing the two camps them/him vs us, assigning S a d d a m Hussein the agent role in the case of negatively valued actions a n d the beneficiary role in the case of retaliations. T h e m e c h a n i s m b e c o m e s m o r e pervasive as the conflict intensifies. This increases exclusion a n d p r o d u c e s a b a c k g r o u n d of'absolute otherness', b y which S a d d a m Hussein is presented as the negative image of us: (6) Los incendios de pozos de petroleo en Kuwait y, posteriormente, los vertidos al golfo Persico refuerzan desgraciadamente los argumentos de quienes consideraron que Sadam recurriria a una politica de tierra quemada antes de soltar su presa . . . (1.2.1991: 16; editorial) (6) Oil wells being burnt and oil being poured in the Gulf afterwards unfortunately are new arguments for those who thought that Saddam would follow a scorched earth policy before letting his prey go . . . . (2.1.1991: 16; editorial) Identity: the construction of us. Facing S a d d a m Hussein there was an inclusive us which was often called u p b y m e a n s of quoting the words of J a m e s Baker, that is, using the same discourse strategy: intertexuality. 'Sadam sigue reforzando susposiciones defensivas en Kuwait, y cuanto mas esperamos mas alto sera el riego que corremos, mas alto sera el riego que correran nuestros soldados.' (6.1.1991: 3) Saddam continues building up his defensive works in Kuwait and the longer we wait, the more we expose ourselves, the more our soldiers expose themselves.' (6.6.1991:3)

104

DISCOURSE STUDIES

T h e first-person plural creates groups of people, b u t it m a y refer, in fact, to different groups (see, for example, Guespin, 1985). A m o n g the values of us which have b e e n pointed out is the national us, found in the quotations of Bush and Baker (as in (7)). However, what is most characteristic of this strategy is the fact that its value lies in designating a collective which has o n e n a m e b u t varies as to content a n d functions (see Geffroy, 1985). T h e p a p e r ' s reproduction of the us of political discourses a n d the progressive use of other co-referential terms - like the U N , the West, the civilized world, the whole world - shows that it widened its reference, as the war went on, to a transnational group. This group is presented as sharing ideological values which are never m a d e explicit - we m u s t n o t forget w h o m a d e u p the 'anti-Saddam coalition', in which it would b e difficult to find c o m m o n values. T h u s the referent is n o longer national b u t ideological. I n this way, b y dividing - segregating through its discourse - the p a p e r not only allows a set of categories to b e established, but it also strengthens the illu sion of a consensus. As Fowler says, 'the we of consensus narrows a n d hardens into a population which sees its interest as culturally a n d economically valid, but as threatened b y a them comprising a m o d e y of antagonistic sectional groups' (1990:53). T h e construction of us implies, then, a simplification, differences are smoothed over, and in this case it m a d e it possible for El Pats and the Government to a n n o u n c e that Spain belonged to the West as one of the most developed countries, and shared the structure of values attributed to this group. However, this process of identification a n d construction of us took place slowly in El Pais, and was m a d e u p of two stages. Stage One. At the beginning of the war, El Pais did not use the generalizing us in its o w n discourse. A b o v e all, it was George Bush a n d the U S w h o faced S a d d a m Hussein. T h e newspaper was against the war a n d supported diplomacy. T h e encompassing us was then m a d e u p of those w h o c o n d e m n e d the invasion a n d wanted a peaceful agreement. George Bush, in the other group, wanted war. Examples (8) and (9) illustrate this stage. I n (9) what is at stake is a questioning of U S leadership. (8) Elpresidente George Bush - no sin fuerte oposicion interior... - ha dejado claro que estd dispuesto a emplear la fuerza..., Sin embargo, existe un anhelo cast una nime en la opinion publico, mundial..., de que se evite le guerra (12.1.1991: 10; editorial) (9) This is not an issue between the United States and Iraq..., but one that puts the UN, almost every country in the world, against the government which has seized Kuwait, so violating international law. (12.1.1991: 10; editorial) (8) President George Bush - not without strong domestic opposition... - has unambi guously said that he is prepared to useforce However, there is an almost unanimous outcry within world public opinion... asking for war to be avoided . . . . (9) This is not an issue between the United States and Iraq..., but one that puts the UN, almost every country in the world, against with the government which has seized Kuwait, so violating international law. (12.1.1991: 10; editorial) Examples (8) a n d (9) show that divergences a n d opposed attitudes are related to the search for the most appropriate solution (diplomacy vs war), but d o not affprt the conceptualization of the conflict. El Pais assigned to S a d d a m Hussein

MARTIN ROJO


105

the agent role in it a n d e m p l o y e d the identical argumentative a n d linguistic re sources for creating a negative image of him. T h e only difference is that during this first stage El Pais did not resort to an inclusive us, which was, however, always implied b y the use of intertexuality. Stage Two: As soon as the allies attacked Iraq, El Pais changed its attitude and started to support the d o m i n a n t discourse: S a d d a m Hussein's m a d n e s s has blocked an agreement a n d has inevitably led to war. F r o m that m o m e n t on, this idea pervaded editorials a n d news reports, and was argued b o t h b y reinforcing exclusion, stressing the rejection a n d the negative image of S a d d a m Hussein that is, b y showing that h e was a n irrational beast - a n d b y inclusion - b y pre senting h. united us that encompassed the w h o l e world. Geffroy (1985: 87) has shown h o w war forces society to form a single group, with a unique principle, us, whose definition and delimitation comes from what Poliakov (1980) calls 'devil-dialectic' (see Howlett, 1982: 15). I n this way, the reader feels appealed to, and absorbed into a unique a n d all-embracing we, in which references are superimposed and change throughout the discourse, or even in a single sentence. W h a t the existence of two different stages shows is h o w the divergences over the potential solution of the conflict disappear, h o w the newspaper's identi fication with the d o m i n a n t discourse - represented b y declarations of George Bush a n d J a m e s Baker - b e c o m e s complete. This identification entails the de crease of the use of intertextuality, and, as a consequence, the use of the firstperson plural p r o n o u n , we, b e c o m e s less frequent. Despite the similarities in the two discourses, El Pais prefers the use of collective subjects, such as 'the inter national community', 'the U N ' , ' E u r o p e ' , or 'the civilized world', in which the existence of a united MS remains implicit. This tendency is m o r e evident in reports about aggressive actions at the beginning of the war (especially in headlines from 17.1.1991: ' U S b o m b s Baghdad'; 'Bush is certain of victory'). At the same time as the subjects which m a d e u p us were left inexplicit, a n acceptable picture of the conflict which faced t h e m was drawn. Both personal ization and setting the conflict within a global background (the 'civilized world against Saddam') were priorities for American diplomacy. This was a result of strategic aims and was linked with face m a n a g e m e n t , but there was a n even more important motive b e h i n d it all, to suppress the economic a n d geopolitical dimension of the conflict and to situate it in a n e w context: on the b o u n d a r y b e tween reason a n d madness, an opposition which brings together all the discourses made b y the West o n the Gulf War. These boundaries, as well as the conflict it self, only allowed for two options: either you enter the game or you are out. Cognitive Models In the previous section, I h a v e shown h o w El Pais, b y m e a n s of explicit inter textuality, contributed to sorting people into categories; later (see the discussion of the contrast b e t w e e n agents a n d actions below), w e will see how different values are put on each of them. However, the personalization of the conflict, with the establishment of two opposing camps, that of the 'ideological we* a n d that of the enemy, e m b o d i e d in S a d d a m , would have b e e n impossible even

DISCOURSE STUDIES

without having recourse to cognitive models built as metaphors a n d metonymies (Lakoff, 1987). Despite t h e existence of some divergences related to t h e solution of the crisis, El Pais conceptualized the conflict b y using t h e same examples of m e t a p h o r s a n d m e t o n y m i e s present i n d o m i n a n t discourse, a n d i n particular in George Bush's discourse. George LakofT argues that there exists a series of relatively fixed a n d extended conceptual metaphors and metonymies which organize our thoughts. If a concept is created b y a m e t a p h o r o r a m e t o n y m , some aspects of it are emphasized and others which are inconsistent with t h e m concealed. I n the cases I examine, using these procedures makes any other interpretation of the facts impossible a n d works as a m e a n s to justify t h e w a r (Lakoff, 1992). I review, first of all, procedures which highlight S a d d a m Hussein a n d contri bute considerably to t h e division process: THE STATE AS A PERSON METAPHOR

A state is conceptualized as a person engaging i n social relations within a world community. Its landmass is its h o m e . It lives in a neighborhood, and has neighbors, friends a n d enemies. States are seen as having inherent dispositions: they can b e peaceful or aggressive, responsible or irrespon sible, industrious or lazy. (Lakoff, 1992: 465) This m e t a p h o r helps sustain u n a n i m o u s j u d g e m e n t s o n t h e state as a whole because it conceals differences b e t w e e n social classes or groups, nationalities, religions or concerns. I n news a n d editorials on the Gulf War t h e same metaphor is used for the two contenders (9). I h a v e noticed, however, that it is n o t used in the same way for both. Given this metaphor, the U S , a n d even the m i x e d 'antiS a d d a m coalition' or t h e West or t h e so-called 'international community', are pictured as a united block i n which concerns, attributes, values, n o r m s and ideology are shared. I n t h e case of Iraq, this m e t a p h o r combines w h a t is really a conglomerate of ethnic groups a n d religions a n d unifies their behaviour, so that, since they are represented in the figure of a single m a n , they can b e made responsible a n d easily c o n d e m n e d . I r a q is seen as a person w h o , from t h e start, is also an e n e m y w h o m a y attack us, hurt us, m a y b e destroy us; so w e have got to face h i m , set u p targets, ask for support against him. T h e m e t a p h o r discovers the e n e m y , creates h i m a n d points to t h e expected response: confrontation o n t h e field - that is, war. THE RULER STANDS FOR THE STATE METONYM. This m e t o n y m is systematically used

b y El Pais w h e n S a d d a m Hussein is involved ((1), (6) a n d (7)), b u t m u c h less fre quently if George Bush is m e n t i o n e d (8). I t has different readings depending on the case. I n the case of Iraq, editorials a n d columns do n o t talk about any abstract entity, b u t expliciuy isolate Saddam Hussein as the agent responsible for the whole conflict as well (see (1): Tf Saddam does not leave Kuwait u n c o n d i t i o n a l l y , . . . ' ) . A m o n g the consequences of this procedure there is t h e selection and justification of the actions u n d e r t a k e n against h i m (against Iraq), which a r e viewed as an un avoidable response because of some particular features of his personality. These t have b e e n carefully chosen a n d highlighted, or even invented, through the use I

MARTIN ROJO


107

of the underlying m e t a p h o r a n d m e t o n y m a n d b e c o m e attributes of this state. U n d e r these circumstances, the process of demonization b e c o m e s m u c h easier. Responses are explicidy portrayed as collective responses against an individ ual. Saddam Hussein seems to b e the beneficiary, w h e n there is in fact a n embargo and a n attack going o n ((1) a n d (12)). T h e m e t o n y m comes t h e n to: S A D D A M H U S S E I N IS I R A Q .

By m e a n s of this m e t o n y m a particular sort of violence is evoked: the o n e which is exercised b y individuals and not b y groups or states. This k i n d of violence is linked to the social imagery of irrationality a n d m a d n e s s (Pintos de Cea-Naharro, 1994). This m e t o n y m acts as an argumentative resource, as a per suasive device, which makes acceptable the rationalizations b y which the domin ant discourse tends to b e legitimized (under REJECTION, we will see the role played b y the imaginary dimension in the conceptualization of the war as a fairy tale, with a h e r o , a villain a n d a victim - that is, its role as a principle of or ganization of perception and social experience). Within the acts of dividing a n d rejecting, the concept of oris n o t constructed in the same way as them or himis. I n the first case, we have a process of encompass ing, whereas in the second w e single s o m e b o d y out. This stresses S a d d a m Hussein's isolation, his irrational a n d impossible stand. T h e m o v e emphasizes the contrast b e t w e e n the two sides a n d efficiendy triggers patriotism. T h e R U L E R STANDS F O R T H E STATE M E T O N Y M is in the case of George Bush far less pervasive. Instead of it, we find collective subjects. T h e y are 'the international community', the U N , or 'the civilized world' a n d are also personified. T h e y are given values, attributes a n d m a d e responsible for actions, all of which contrast with, and often are even the direct opposite of, those assigned to S a d d a m Hussein (see below the discussion of contrast expressed through agents and actions). T h e p r o n o u n we refers to a set of values a n d the content of the consensual assumption could b e spelt out as a long series of propositions about social, political, economic a n d cultural world order. T h e a p p a r e n t homogeneity strengthens the use of we a n d supports what is merely a vision of the events which harmonizes with the interests of some governments a n d powerful groups. This vision grows from the 'ideology of consensus' and, in its turn, articulates this ideology (see Fowler, 1991: 4 9 - 5 3 ; Hall, 1973; H a r d e y , 1982: 1-3). 'Our' ideology implies the assumption that the whole population - in this case, the international c o m m u n i t y - shares the same interests a n d beliefs. T h e use of we and collective subjects produces, then, an over-simplification rather like that of T H E STATE A S P E R S O N METAPHOR. This process combines a conglomerate of ethnic groups, interests, cultures a n d religions a n d unifies their behaviours. Thus it presents the interests a n d values of the d o m i n a n t groups as universal. Fowler uses the expression 'homocentrism' for this overgeneralization: ' a pre occupation with countries, societies a n d individuals perceived to b e like oneself; with boundaries; with defining groups felt to b e unlike oneself, alien, threatening' 1991: 16). 1 now present some examples showing this m e t a p h o r (THE STATE A S P E R S O N METAPHOR) a n d m e t o n y m (THE R U L E R S T A N D S F O R T H E STATE METONYM) at work and how the contrast strengthens the dichotomy and, as a consequence, division. 6

108

DISCOURSE STO

The Rivals: Contrast Expressed through Agents and Actions T h e act of dividing leads to contrast. Exclusion is the establishment of a borderline b e t w e e n two camps: 'our c a m p ' (inside), w h e r e 'our' allies and friends are in cluded; a n d 'their c a m p ' (outside), w h e r e 'others', 'our enemies' are excluded This establishment implies the definition o f ' o u r ' a n d 'their' land. Thus, contra* will give sense to words like us or him. Contrast has a n effect not only on the agents, b u t also o n their behaviour a n d attributes. S a d d a m Hussein acts as 'oar e n e m y ' - that is, as the agent of aggressive actions, and as the subject/agent in transitive or intransitive sentences e n d o w e d with negative values (see example (6), (7), (10), (11), (20) a n d (26)): 'to offend', 'to threaten', 'to invade', 'toattack'. * build up defensive works', 'to bomb', etc. T h e U S a n d the international communirf are presented, however, as agents of actions such as ' to present a peace profaxT, 'to defend Kuwait', 'to defend international order', 'to order a ceasefire', etc. jsee, far instance, 'peace offering' a n d ' p r a y for p e a c e ' , in (10) a n d (11)].

Different m e t h o d s are used to build u p this contrast (see (10) and (li)): Lexical procedures. Lexical categorization. As regards the agents, the contrast! h a v e referred to is m a d e b e t w e e n : (a) all-encompassing agents (like "the wrHf a n d 'the international c o m m u n i t y ' - in (10) a n d (11) - or 'civilized world' (26)) vs very concrete agents ('Saddam Hussein', or ' h i m ' - in (10) and 'Iraq' - in (26)); (b) a united a n d ideological we (7) vs a concrete agent, hif a n d (11) or ' S a d d a m ' (7); (c) two concrete agents, 'Bush' (several headluu ^ toco o n George Bush's declarations, for e x a m p l e 'Bush r e c o m m e n d s Saddan :••> opt for peace', 'Bush d e c l a r e s . . . ' . 'Bush c l a i m s . . . ' , 6.1.1991) vs 'Saddam' ('b idaa intends to keep Kuwait a n d liberate Palestine', 7.1.1991). T h e second •••rraej the contrast is always him, which implies the presentation of Saddam f •.atm as the unequivocal agent of the following actions. Vocabulary establishes a division b e t w e e n peaceful actions and aggre. ^ e ittions ((10) a n d (11)). I n this sense, euphemisms are used to present agj? ^ actions as legitimate, peaceful a n d m o d e r a t e ('defend Kuwait' vs 'attac 'peace offering' a n d 'pray for p e a c e ' in (10) and (11), at a moment wi Americans w e r e preparing for war - in fact already sending troops ou contrast is m o r e evident once war was declared, and, as different analy* shown (see Fowler, 1991; H o d g e a n d Kress, 1992) is a very common str«;^); western newspapers. D u r i n g the period studied in this p a p e r (that is, jus* bilam the war), this p r o c e d u r e presented S a d d a m Hussein as an antagonisi M M actions as refusals, to 'peace offerings'. I T h e m a i n results of this p r o c e d u r e a r e first to link S a d d a m Hussein vm the use of violence. This evocation - b y using T H E R U L E R STANDS FOR TKL stM M E T O N Y M - of a particular k i n d of violence, exercised b y a n individual. playxi w e will see, a n essential role in the rejection m o v e . T h e second main result al show that there could n o t b e a c o m m o n 'rhetorical g r o u n d ' and hen.e the) could not b e a diplomatic solution to the conflict. T h e use of force is justified Syntactic procedures. By these procedures, the role played b y participanu different actions is emphasized or h i d d e n . i

•OJO


109

Syntactic transformations of the clause. This is achieved b y blurring the focus o n agents, or suppressing t h e m b y m e a n s of impersonal structures, passive verbs, owninalizatii : etc. I n examples Uke: 'Waves o f b o m b s fell o n selected objectives in the Iraqi c o i t a l ' or 'Black clouds go u p from B a g h d a d ' (both 17.1.1991), the «Weubn o f tin agent is complete. T h e use of this procedure, studied b y 'critical Sngnistics' (si Fowler. 1991; H o d g e a n d Kress, 1992), increases remarkably frotn 17 Janiiii- on - that is, once w a r was declared a n d attacks b e g a n to b e reported. When the agents of 'our' actions are m e n t i o n e d , El Pais shows its pre ference for oilier forms of detachment: the use of some concrete agents like the US and the 'ailies'. from which Spain seems to b e excluded: 'Los aliados dicen haber cestruidu . . . ' ('Their allies claim . . . ' ) . In this wav- a friendly i m a g e is c r e a t e d t h r o u g h parallel structures a n d antonyms, especially those related with their respective behaviours and attributes (Hodge and Kress. 1992), which are used to reinforce the contrast as in the following examples: (10) 'Mientwelmundoesperaba.', dijo, 'SadamHuseinrespondioa.cada.gestodepazpor parte de lacomunidad internacional, conun desafio' ill) 'Mimtra'elmundo rezabapor lapaz, Usepreparabapara laguerra.' (16.1.1991: 1) flO) Quoting George Bush 'While the world waited, Saddam Hussein respondedto any ptaa offering made by the international community with a challenge' 11) 'WkiU tht world was prayingfor peace, he was preparing war.' (16.1.1991: 1) El Pais reproduces in examples (10) and (11) the rhetorical devices employed by 'George Bush: parallel structures (While + all-encompassing agents + peaceful action/concrete agent (he) + aggressive actions) articulate a n d reinforce the contrast between agents and actions. This contrast articulates the text a n d m a k e s it coherent in news reports, a n d even more so in editorials. M o r e o v e r it is efficient in creating a positive face for o and a negative one for them or him. I n o w focus o n this point - the second move of the exclusion process: REJECTION. I n this case, the m a i n strategy is the evocation of the imaginary d i m e n s i o n , as a m e a n s to structure the social experience and the perception of the conflict. T h e 'script' of the event activated by (he newspaper seems to b e w h a t Lakoff (1992) calls 'the fairy tale of the just war*. The aim of most of the linguistic a n d argumentative resources used b y El Pais could be seen as an attempt to force the m a i n characters a n d their actions to respond to this script.

Rejection: T h e D e m o n i z a t i o n of S a d d a m H u s s e i n la this section, I focus on the role of the imaginary dimension in the organization of perception and social experience. T h e concept of the imaginary dimension, developed by sociologists like Duvignaud, Balandier a n d Rochberg-Halton, is based on Durkheim's concept of 'shared representations' (see, for instance, Uduit, 198": 45; Rochberg-Halton, 1986: 299). This d i m e n s i o n c a n n o t b e

D I S C O U R S E D ; 'Dir.*

110

identified with ideologies. It is considered as a principle of organization, whose m a i n function is to m a k e ideologies reliable a n d acceptable. As this analysis shows, El Pais, resorting to social imagery, looks for a particular interpretation of the events. T h e reader is asked to understand the Gulf War, recalling and u p d a t i n g a n old m o d e l , which, as Lakoff claims, is like a fairy tale. (See vats Dijk, 1 9 9 3 : 9 2 - 1 3 4 ; and v a n Dijk and Kintsch, 1983 for a theoretical framework which features the 'sociocognitive interface' of both personal mental models and socially shared mental representations.) George Lakoff gives an account of the process b y which Saddam Hussein b e c o m e s a d e m o n through the conceptualization of the war in terms of a fairy tale. T h e m a i n characters of the story are the villain, the victim and the hero This plot is reduced to an offence committed b y the villain on an innocent vktim. T h e crime is rooted in a n imbalance of p o w e r and leads to a moral imbal ance. T h e h e r o must h a v e a h a r d time a n d fight the villain in order to rescue tht victim. His victory restores morality, a n d h e himself, because of his unselfish actions, is glorified. Contrast a n d asymmetry between hero and villain are essen tial c o m p o n e n t s in the tale: the hero's morality is perfect whereas the villain » immoral; the h e r o is valiant a n d rational; the villain is a n irrational being, though h e might b e v e r y cunning a n d calculating; u n d e r these circumstances, tbefe i> n o w a y a deal can b e m a d e with h i m : h e has to b e challenged (LakofT. 19S2[. I will n o w m a k e clear h o w the different linguistic and argumentative resource contribute b o t h to the conceptualization of the war following this script and to the presentation of the political authorities a n d their actions in terms of th» main characters of the story and their stereotypical behaviour. My analysis foam o n S a d d a m Hussein and o n the social imagery of irrationality and madness that is, o n the creation of the villain. A m e t a p h o r emerges from the tale - T H E E N E M Y A S A D E M O N METAPHO* which lies b e h i n d whatever descriptions, qualities a n d behaviours are attributed to S a d d a m Hussein. Amongst the villain's alleged attributes, Lakoff points te irrationality. We m a y w o n d e r w h y this is the most outstanding and distinctive feature assigned to the villain w h e n an event is conceptualized in terms of ik fairy tale of the just war'. T h e opposition between sanity a n d m a d n e s s is linked, so Foucault ihinki with power: T h e realm of light is opposed to that of darkness in a way that reminds us of the w a y the world of the powerful is opposed to that of the weak. Light is only owned b y official powers, whereas darkness belongs to the demon ized classes. (Alvarez-Uria, 1983: 21-4). 7

concern

T h e image of a dissident is, Foucault says, the result of a series of izations m a d e u p b y the law a n d ruling classes u p o n practices that violet k social n o r m (see Foucault, 1964; and Alvarez-Uria, 1983, for analyses of this ps cess in Spain). That is to say, if you want to establish a n o r m , a power domsis' either in a particular country or at an international level - a n d maintain it. ra .i c—„ - " i n t i - t h n s R w h o do n o t obey it. If such forms;

ROJO


111

domination are not accepted b y b o t h sides - the dominant and the dominated as a 'natural order', aggression b e c o m e s the m o s t c o m m o n w a y to k e e p a n d reinforce the established order. (See Pintos d e Cea-Naharro, 1995: 41, for the mental operations which are at the base of violent events: exclusion, discrimin ation, domination and aggression.) T h e evocation of a kind of violence, which o exercised bv one individual (instead of groups or states), presents the rebellion of the dominated as an isolated gesture of greed a n d as the effect of irrationality and madness. As a consequence, p u n i s h m e n t is justified. The Gulf War takes place at the very m o m e n t w h e n the old world order a n d traditional \ alues are changing. These changes are often received with anguish. Within this context. S a d d a m Hussein does not react according to his status; his logic is insurrectionary. D u e to the slu-inking of the planet, the product of this logic becomes threats o n our v e r y doorstep. T h e features of the presentation of ibe Gulf War were a paradigmatic example of h o w the media can take us instandy into the farthest corner of the world. Images, information, goods, n o w h a v e freedom of movement far b e y o n d all national frontiers, which produces a feeling of there being too much space, so that w e h a v e the impression that everything that happens affects us directly and immediately (Auge, 1993: 25). T h e reports of the Gulf War constitute a perfect e x a m p l e of h o w t h e mass m e d i a 'project us in seconds into the farthest corners of the earth' a n d h o w the 'far other' w h o in timates all our fears (aggressiveness, cruelty, inhumanity, madness) appears as an immediate menace (Auge, 1993: 25ff.). 8

El Pais - and some other newspapers in the West as well - does n o t deKribe the conflict as a result of a n e w world order which is harmful to some of its members. This picture is in fact drawn b y the Iraqi press a n d b y that of other Arab countries, like J o r d a n . El Pais describes it, however, as a gesture of greed, as a threat to 'our interests'. T h e global solution to the conflicts in the M i d d l e East desired by Iraq is n e v e r contextualized, is n e v e r linked with the n o n acceptance of those borders i m p o s e d b y Western Empires which split the 'Arab nations' between rich and poor. Following this logic, any proposal m a d e b y Iraq is turned into nonsense a n d the Gulf Crisis is understood n o t as a conflict of interests, but as a challenge to a n e w world order, ruled through reason, peace and justice; however the extent to which this order benefits us is n e v e r made explicit. Globalization m u s t b e accomplished in o u r o w n image a n d likeness, and in agreement with our interests. All dissidents must b e expelled from the realm of humankind. W h a t the West sets u p as ' n o r m a l ' must b e 'uni versal', and this vision draws its legitimacy from concepts such as 'world order', reason', 'justice', 'balance' a n d ' p e a c e ' (Gillaume, 1993: 45ff.). The fact that the news is focused o n S a d d a m Hussein, George Bush, Iraq, the US, the West or the international community is irrelevant. Given that madness implies a separation between rationality and ^rationality, what matters is that this contrast, either implicitiy or explicitly, is always there: y o u h a v e to choose e side or the other, you h a v e 'to b e within the group or r e m a i n out of it' oucault, 1964: 260). In this sense, emphasizing some features and not others, buting some particular actions or behaviours to the 'other', implies the abce of these features and forms of behaviour in one's o w n errmn W P h a w

112

DISCOURSE STUDIES

then e n d e d u p with a kind of mirror image: n o w the qualities a n d actions at tributed to S a d d a m Hussein d o not m a k e u p a n illogical or chaotic list, but re p r o d u c e the idea we h a v e of madness, and, ultimately, draw the image of evil. I n this section, we will look at two sorts of processes: (1) discrediting b y means of differences in power; (2) the construction of the 'absolute other'. T h e personal ization of the conflict a n d the division m o v e makes b o t h m e t h o d s possible.

While the concept us is based u p o n the solidarity axis, that of him relies o n the p o w e r axis. O n l y from the point of view of a h o m o g e n e o u s g r o u p which is identified as the voice of reason can the stranger b e excluded; for that purpose, the group will h a v e to ensure its h e g e m o n y , will h a v e to construct a n d under line asymmetries. I n editorials a n d in the news, different linguistic items are used that m a y b e interpreted as taxemes (Kerbrat-Orechioni, 1987); they outline the differences in power that exist between the two rivals: (12) Elpresidente deEEUUadvieite a Sodom sobre 'terribles consecuencias' si continua en Kuwait. (6.1.1991: 3; titular) (12) The President of the US draws Saddam's attention to the 'terrible consequences' if he does not leave Kuwait. (6.1.1991: 3; headline) T h e pervasive imbalance w e find in the way that forms of address are used for b o t h contenders legitimizes the threats m a d e b y the 'President' of the United States. E u p h e m i s m s are used to refer to those threats ('initiatives', 'talk offers'), b u t syntax a n d taxemes show what they really are: directive acts (for example, 'Bush r e c o m m e n d s S a d d a m to opt for peace'). T h u s , El Pais, in its o w n discourse a n d in the discourse quoted, assumes that what is at stake is a process of domin ation, in which S a d d a m Hussein plays the deserved role of the dominated. How ever - a n d m o r e importantiy - this imbalance delegitimizes S a d d a m Hussein, within the background of the ideology of consensus. This ideology says a presi dent represents his country; h e has, we suppose, b e e n elected a n d should respect the public's will. O n the other h a n d , a n individual w h o is being addressed by his first n a m e only (Saddam) represents himself alone. I n the picture drawn of the two leaders o n 17 J a n u a r y the contrast b e c o m e s self-evident. T h e column o n S a d d a m Hussein (see (17)) starts b y saying: ' S a d d a m Hussein is 5 3 years old. H e is a well-built m a n , almost 1.90 metres tall', etc. (i.e. a list of his physical characteristics). O n the other h a n d , the portrait of George Bush goes like this: 'George H e r b e r t Walker Bush, 41st President of the U n i t e d States a n d Com m a n d e r in Chief of its a r m e d forces', etc. A n d so w e e n d u p with the 'in-group' b e i n g glorified a n d the 'other' being isolated, degraded a n d delegitimized. His rebellion is n o t legitimate whereas the use of force against h i m is justified. Participants are categorized through the use of language in the discourse. As we shall see in the following examples, and as Fowler points out, besides differeni styles of naming, conventionally associated with different social values, 'those w h o are disfavoured a n d discriminated against are likely to b e associated with pejorative or at least low-status verbs a n d adjectives' (see Fowler, 1991: Ch. 6).

MARTIN ROJO


113

By analysing the discourse, it is possible to identify what is considered negative, both socially a n d b y certain groups in power. This is w h y the appearance of ethnic prejudice as a way of discrediting S a d d a m Hussein is important. Implicit Judgements: The Selection and Omission Principle Rejecting is reinforced in El Pais basically through the insertion of details which are associated with strong feelings a n d individual a n d cultural imagery that bear, above all, o n the b r e a k of the ideology of consensus a n d o n ethnic preju dices. T h u s , we obtain a symbolic descriptive illustration which conveys a n implicit meaning a n d leads the reader towards particular judgements, even though the newspaper is not expressing t h e m explicitly. Implicitness a n d inferencing are two important characteristics in the text a n d in the reader. Meanings are frequendy conveyed without b e i n g explicitly stated. T h e strategy is to supply a sequence of details which evoke the imagery of violence. These shared repre sentations are supported b y the ideology of consensus a n d ethnic prejudices, but at the same time m a k e b o t h reliable a n d acceptable. Details usually a p p e a r in e m b e d d e d clauses or sequences of prepositional phrases a n d their function is the specification of the information in the m a i n clause. Because of the particular local organization of news, 'the reader must unscramble bits a n d pieces through out the text a n d fit t h e m into the appropriate topics a n d schematical categories' (van Dijk, 1988: 65). I n this sense, the role played b y all these, apparently ir relevant, details in the disqualification of S a d d a m Hussein is crucial. As w e will see, they serve to evoke the imaginary dimension a n d to activate the ideology of consensus as well as ethnic prejudices, b y m e a n s of which S a d d a m Hussein will b e judged. T h e prejudiced use of the imaginary dimension is always present in the reports, but increases remarkably from 17 J a n u a r y o n - that is, once war was declared. This was a crucial day also for editorials in which explicit j u d g e m e n t s started appearing a n d b e c o m i n g m o r e a n d m o r e negative as time went by. I n both cases, the purpose was to portray Saddam Hussein as a ruler whose madness made it impossible to agree with him. I n the columns, the reader was pointed in the direction of that conclusion, whereas, in the editorials, it was expressed explicitiy. This p r o d u c e d a cumulative effect b y which the reader started extracting data from various news reports and making judgements that were later confirmed by the background voice of editorials. Rhetoric a n d persuasion worked together in harmony - that is, the rhetoric of editorials appealed to prejudice and to social imagination, as did the connotations, metaphors a n d small details featured in news reports. By contrast, this h a r m o n y was absent from the Sun, where the accumulation of figures of speech that insisted o n the brutalization of S a d d a m Hussein and other devices that appealed to prejudice was so obvious that there seemed to b e n o place left for rhetoric, only for repulsion. The axes o n which the process of S a d d a m Hussein's demonization is founded are basically dual: the alleged break in the ideology of consensus a n d the invoca tion of ethnic prejudices. T h e projection of b o t h axes presented the reader with an unequivocal e x a m p l e of w h a t M a r c Auge (1993) calls 'absolute otherness':

114

DISCOURSE STt'DllP

'If necessity arises, all those defects one denies having oneself can b e attributed to the foreigner. Towards that foreigner outside our borders a r e projected, from time to time, t h e ghosts of ferocity, cannibalism, a n d inhumanity' (Auge, 1993: 18). N o t only did the beginning of the war reinforce the exclusion and inclusion processes, b u t it also involved a change of attitude a n d behaviour towards the other. N o longer would S a d d a m Hussein b e appealed to to give up his policy and his behaviour a n d to adapt to 'normality'. O n t h e contrary, the emphasis would fall o n t h e irreducible character of the differences separating him from the rest of the w o r l d a n d the subsequent n e e d to m a k e h i m disappear. The d*course b e c a m e emotional, impassioned, even i n the editorials of a so-called 'im partial' newspaper such as El Pais (in particular, o n 17.1.1991). The break in the ideology of consensus. For Fowler, 'articulating the ideology of consensus is a crucial practice in t h e Press's m a n a g e m e n t of its relahooi with government a n d capital, o n t h e o n e h a n d , a n d with readers on the other" (1991:49). T h e ideology of consensus appears as a set of shared beliefs (everyone thinks that . . .). T h e r e is n o suggestion that they might b e useful to certain interests, since society is n o t considered to b e a place of conflict. The press cootributes, transmitting this idea persuasively. However, this is, in fact, a linguistic practice, which has not b e e n studied m u c h y e t Of course, this does not mean thi the journalist is clearly conscious of the values a n d beliefs h e or she is using, or that t h e readers are helpless, a n d necessarily persuaded b y the vision they are given of the events. T h e first-person plural p r o n o u n we, with its multiple reference, is used is, political discourse a n d in t h e press to create t h e illusion of the existence of i c o m m u n i t y which is ideologically one. T h u s , t h e existence of differing iniere* a n d opposing ideology a m o n g t h e population is hidden. However, in the a s of the Gulf War, what exacdy this consensus consists of is never made expliot This h a p p e n s n o t only because it might turn o u t to b e controversial, but abe because of cultural, religious a n d ideological differences - as well as those d p o w e r a n d status, which clearly existed in t h e allied group, the group rt-fernfl to as us. I I n spite of this, the ideology of consensus is the cognitive base on which I n categorization of subjects a n d actions takes place (styles of naming: Saddam •m the President; t h e words of Bush vs the p r o p a g a n d a of Saddam Hussein; tttfl invasion of Kuwait vs its liberation, etc.). It explains the choices of details aM the foundations for certain value j u d g e m e n t s in t h e process of discrediting tfcfl President of Iraq. By slipping into the reports different details not really r e l e v j to a particular piece of news (the clothes S a d d a m Hussein was wearing, f o r a l ample), journalists were able to evoke t h e ideology of consensus, and foicfl readers to use it to interpret the text. T h e r e were two aims hidden under this uffl o n the o n e h a n d , to m a k e it clear that S a d d a m Hussein does not share this logy, which leaves h i m outside the group; o n the other, to consolidate the unr of we, a n d project a positive image of the group. Readers infer, using the irairwork of contrast which h a s b e e n built u p , that t h e values which are shojj negative are n o t accepted b y the allied group.

JO

DEMONIZATION O F SADDAM

115

As regard-. ;he break in the ideology of consensus, S a d d a m Hussein's con ceptualization runs along t h e following lines. H e is presented as: (1) a soldier and a dictator; 2) the m a n w h o is manipulating information; (3) the m a n w h o is manipulating religious fanaticism; a n d (4) a n instigator of terrorism. All these characteristics are attributed to h i m through the selection of details, the insertion of convenient comments and through omissions, such as forgetting, for instance, that all these transgressions w e r e also taking place within the so called 'antiSaddam coalition'. Examples of this p h e n o m e n o n are extracts (13), (14), (15) and (16) below. These examples show ways of attributing to S a d d a m Hussein values which from the point of view of the ideology of consensus are considered negative and are rejected. T h e y are nearly all o n the list published b y Chibnall •~ !°77. Among the linguistic strategies are: the insertion into the text of details h call up the transgression of the ideology of consensus; the inclusion of .; .stive judgments: and significant choices from the lexicon, especially of adtvuves and verbs. Tkt soldier and the dictator. El Pais insists o n the fact that S a d d a m Hussein is a soldier by continuously mentioning the tides which link h i m with the army, through the use of possessives that m a k e it clear that the Iraqi a r m y is S a d d a m ' s this linguistic resource is linked with the underlying m e t o n y m : T H E RULER STANDS fOR THE STATE, and also through the selection of apparently poindess details, such as his clothing):

#.

til Sad am con una gorra verdey traje de faena, se dirigio al pais para rendir tributo a su rjerctio de un millon de hombres. (7.1.1991: 3) 11J) Saddam, in a green cap and battledress, addressed the country honouring his one million army. (7.1.1991: 3) In (13) and (14) we find examples of ways in which journalists include items of information attached to the m a i n proposition, with the apparent aim of de scribing or presenting Hussein a n d his actions. T h e s e are expressed b y m e a n s of sequences of prepositional phrases a n d subordinate clauses (for this procedure, see van Dijk, 1988: 66ff.). I n this sense, what is relevant in (13) a n d (14) - as in most of the following examples - is h o w explicit descriptions ((13): in a green cap aibattUdrttf; 'honouring his one million armf; (14): 'who hardly ever delegates') are, in fact, implicit qualifications. By choosing to m e n t i o n all these apparently ^relevant details, the writer is evoking the imagery of violence a n d , as a conse quence, acti vating a set of tacit assumptions a n d beliefs - in this case the ideology of consensus, which serves to c o n d e m n S a d d a m Hussein. As in other examples, the moral ambivalence of the n e w s p a p e r is revealed. The imagery of violence, linked to the villain of the tale, is evoked here b y pointing out his clothing a n d the proportions of the army. Furthermore, it is precisely this imaginary dimension that explains the selections of details and refers to the ideology which gives coherence to the text. T h e emphasis o n the towards all things military presented S a d d a m Hussein implicitly as l

IK)

DISCOURSE STUDIES

a dictator, helped b y ghosts of World War I I and of Latin American dictatorships. However, explicit j u d g e m e n t s could also b e found in t h e newspaper: (14) La ultima palabra esta en manos del lider supremo, Sadam Husein, que rara vez delega; todo lo mas consulta con su cfrculo de intimos. (10.1.1991: 4) (14) The last word always comes from the supreme leader, Saddam Hussein, who hardly ever delegates, if anything he asks his closed circle of intimates. (10.1.1991: 4) Monopolizing power, illegality and violence are h e r e the negative values. It is easy to infer, given the constant contrast b y m e a n s of w h i c h the reports and leading articles are structured, that the opposites, freedom of choice, legality and desires of peace, are present in the anti-Saddam coalition, a n d that the pur pose of the conflict is defence of these values. The manipulator of information. Dogmatism, intolerance, biased reporting, monopolizing of information are crimes of S a d d a m Hussein. T h e opposites h e r e would b e compromise, tolerance, impartiality a n d freedom of choice. (15) Leprensa de Sadam es triunfalista, porque as! lo exige la omnipresentepropaganda. (16.6.1991: 4). (15) Saddam Hussein's press is triumphalist because that is what is required by the omnipresent propaganda system. (16.6.1991: 4) I n (15), linguistic procedures used are: possessives (see e x a m p l e (14)) and lexical categorization (see example (3)). H e r e , the subordinated clause, presented as an explanation (though it is not clear w h y anything n e e d s to b e explained), is in fact a w a y of establishing t h e equivalence b e t w e e n the Iraqi press and a p r o p a g a n d a machine. The manipulator of religiousfanaticism: Allusion to G o d a n d religion, as well as to the spirit of evil, w e r e frequent in S a d d a m Hussein's speeches as in George Bush's. H o w e v e r , they w e r e treated very differently in b o t h cases. I n Saddam Hussein's case, they were disqualified b y invoking ethnic prejudices a n d b y re sorting to linguistic devices that deprived t h e m of credibility. I n (16), since S a d d a m Hussein was presented as agent of to raise to the rank of, the implication for the reader is that the war lacked that position. T h e r e w e r e n u m e r o u s details in news reports that contributed to creating this impression. For example, Saddam Hussein's religious faith was described as 'sudden' (10.1.1991: 4) a n d implicitly presented as bogus a n d essentially manipulative. (16) Como ya viene siendo habitual desde el comienzo de la crisis, Sadam elevd al rango de yihad [guerra Santa) el combate que podrfa librarse en la region. (7.1.1991:3) (16) As had become customary from the beginning of the crisis, Saddam has raised the war that may take place in the region to the rank ofholy war (jihad). (7.1.1991:3) An instigator of terrorism. Besides warning against Iraq's military strength, El Pais asserted, in the days before the war, that the true front line would b e in Europe, in a deliberate overestimating of S a d d a m Hussein's p o w e r to spread terror in the West. However, the direct participation of S a d d a m Hussein is always kepi

MARTIN ROJO


117

implicit. T h e negative values attributed indirecdy to h i m are: extremism, chaos, violence, as o p p o s e d to moderation, order a n d the desire for peace. I n this sense, news about a p p a r e n d y peripheral events (like Europeans beginning to store food [16.1.1991] or police action against 'false pacifists' [16.1.1991]) plays a role in the evocation of the imagery of violence. Ethnicprejudices. While the fact of breaking the ideology of consensus presents Saddam Hussein as an antagonist, it is the use of ethnic prejudice which plays the strongest role i n transforming h i m into the 'absolute other'. Such prejudices work in the same w a y as the ideology of consensus - that is, as a sort of backdrop, as a cognitive base, which is n o t m a d e explicit, b u t which explains the choice of details included in the news. Readers must have recourse to these prejudices in order to interpret the t e x t T h e y are n e e d e d to give it its coherence (see (10)). Ethnic prejudices are invoked b y m e a n s of selecting minor details which are inserted h e r e a n d there in different reports. T h e basic lines for this invocation are the following oppositions: civilization vs barbarism; rationality vs irrationality, including the image of the beast; and, ultimately, good vs evil. T h e portrait of Saddam Hussein in El Pais o n 17 J a n u a r y (in examples (17) and (18)) contains all these ingredients. T h e detailed description of his appearance clearly shows ethnic features a n d cultural differences: colour, his penetrating gaze a n d so on. Nothing is mentioned about the positive or negative aspects of having olive skin, or a penetrating gaze. However, for readers to find any reason for their inclusion in the text, they have to activate knowledge, which the writers present as shared, and which links these sorts of physical characteristics with certain sets of values. O n the other h a n d , George Bush's picture o n the same page begins b y enum erating his tides and merits. T h e implicit contrast turns t h e m into a b o d y versus mind opposition: ours against theirs. (17) Sadam Husein tiene 53 anos. Es de complexion atletica, de casi 1, 90 mStros de estatura, cabello muy negro y piel aceitunada. Sus ojos son verde clans y su mirada interna. (18) . . . casado con ST, prima suya, su primogenito, Udai, mato a golpes y en publico, a uno de los guardaespaldas de su padre. Juzgado y condenado, masas enfervorecidas recorrieron las calles de Bagdad pidiendo la clemencia paterna. La tuvo. Sadam, licenciado en Derecho por la Universidad de Bagdad, no mostro nunca, en cambio clemencia con sus rivales y adversarios. Centenares de ellos desfilaron ante la horca o los pelotones de fusilamiento. (19) Licenciado en Economicas por la Universidad de Yale, que ha llegado a la presidencia con una experiencia polftica nacional e internacional poco habitual. Diputado por Tejas en dos legislaturas, Presidente del Comite Nacional de Partido Republicano, embajador ante la ONU, primer jefe de mision en Pekfn, director de la Agencia Central de Inteligencia (CIA) y vicepresidente durante ocho anos. (17.1.1991: 6) (17) Saddam Hussein is 53 years old. He is a well-built man, almost 1.90 metres tall. He's got very black hair and is olive-skinned. His eyes are green and he has a pene trating gaze. (18) . . . he married Sadija Talfa, a cousin of his, and his elder son, Udai, beat, in a public place, one of his father's bodyguards to death. He was tried and found

118

DISCOURSE STUDIES

1

guilty, but, arousing the masses, walked the streets of Baghdad asking for pa ternal benevolence. He got it. Saddam, who has a degree in Law from the Uni versity of Baghdad, was never merciful, however, with his rivals and opponents. Hundreds ofthem were hanged or shot dead. (19) He graduated in Economics in Yale University and has become president after acquiring unusual political expertise, both national and international. He has been a Congressman for Texas for two terms, President of the National Com mittee of the Republican Party, ambassador to the UN, first ambassador in Peking, director of the CIA and vice-president for eight years. (17.1.1991: 6) W h a t is obvious in these examples is the w a y the selection a n d omission principle works. I n the case of George Bush (19), titles come first; Hussein's are just mentioned to discredit h i m (18). T h e reference to S a d d a m Hussein's physical characteristics (17) a n d his family relationships (18) is relevant. It is interesting that this sort of detail is hardly ever included in reports about m e n . Instead, it is usual to include information about profession a n d qualifications. I n (18), such details call u p the imagery of incest, a n d point to negative characteristics of the regime, like arbitrary decisions, illegality a n d corruption. W i t h regard to texi coherence, what allows specific sentences to b e connected into a coherent se quence - such as ' H e married Sadija Talfa, a cousin of his, and his elder son. Udai, beat, in a public place, one of his father's b o d y g u a r d s to death' in (18) - is the resort to some shared knowledge about different ethnic groups, and in par ticular some ethnic prejudices, which underlie the opposition 'civilization vs barbarism'. T h e function of the propositions included in ' H e married Sadija Talfa, a cousin of his', in (18), is to provide an explanation or at least to suppon the final judgement, 'was never merciful, however, with his rivals a n d opponents. H u n d r e d s of t h e m were hanged or shot dead', b y which the paradigm of indi vidual violence a n d the image of the beast are again evoked. T h e Sun newspaper published a portrait of the two leaders o n 17 Januan 1991. I n this case, the description/evaluation was carried out b y a psychologist who selected features that coincided with those highlighted b y El Pais. However, besides giving these details, the psychologist diagnosed S a d d a m Hussein as: i megalomaniac, d u e to his monopolizing of power; a 'charmer', since he could persuade his people to die for h i m ; a psychotic a n d p a r a n o i d schizophrenic T h e Sun concluded that his extreme mental disorders, together with his powe as a dictator, m a d e h i m a lethal figure. T h e image of the beast was b a s e d u p o n two features which were always stressed: his fondness of w e a p o n s a n d violence, a n d his errors of judgement With respect to the former, I have found n u m e r o u s examples; but outstanding a m o n g t h e m are those featured in the portrait in (20) because they lead readers to infer that this aspect of his personality was responsible for the position Saddam Hussein occupies today.

(20) Ha sido siempre un hombre de accion. Amante de las armas y burn tirador it revolver, atentd contra el General Abdel Karim Kasem, en 1959. Fue herido, pero h u v o . . . . Desde 1963, su carrera hacia la cuspide ha sido metedrica. (17.1.1991:6)

MARTfN ROJO


119

(20) He has always been a man of action. He loves guns and is a good shot. He attempted to assassinate General Abdel Karim Kasem in 1959. He was wounded but managed to escape.... Since 1963 his career has been a meteoric climb to the top. (17.1.1991: 6) T h e function of the accumulation of details ('He loves guns and is a good shot. He attempted to assassinate') seems to b e t h e explanation for w h a t is affirmed i n the last sentence: 'his career has been a meteoric climb to the top'. Implicitiy, a rela tionship of causality is established b e t w e e n facts denoted b y propositions. This argumentative flow p r o d u c e d the required effect, as can b e seen b y the accumu lation of negative information a n d animalizing metaphors. T h e following are examples of this latter strategy: there was speculation that S a d d a m Hussein would follow a scorched earth policy before 'letting his p r e y g o ' (see (2): t h e biography published o n 31.1.1991 was entitled 'A wolf a m o n g gazelles'. However, El Pais was m o r e cautious than the Sun w h e n it came to using images that evoke violence. This m a y have been due to the fact that, despite their persuasive effect for they reinforce S a d d a m Hussein's negative image with strong convincing strokes - they went too far into t h e realm of emotion a n d therefore did n o t suit the rhetoric of El Pais. With respect to irrationality and errors ofjudgement, S a d d a m Hussein does not fit into metaphors like WAR A S POLITICS, POLITICS AS B U S I N E S S (Lakoff, 1992), and the analysis of w a r as a cost a n d benefits balance: 'defining beneficial "•objectives' , tallying the "costs", a n d deciding whether achieving the objectives is "worth" the costs' (Lakoff, 1992: 464). This is the reason w h y at the beginning there was talk about S a d d a m Hussein's bluff, whereas later o n it was referred to 9s madness, for h e was defying a p o w e r h e would b e u n a b l e to defeat. 1

9

(21) 'Sadam cree que Bush esta bromeando... pero esta equivocado.' (5.1.1991: 3) •22) 'Nuestro temor es que, en su habitual estilo, [Sadam] calcule ma/donde se halla el limite.' (12.1.1991:5). ;23) Sadam triunfalista y convencido de ganar la guerra. (12.1.1991: 2; titular) 24) Sadam volvio a insistir en la capacidad de sus fuerzas armadas para una 'larga batalla'. (7.1.1991:3) 2l) [Quotingjames Baker] 'Saddam believes that Bush is joking.. .but he is wrong.' (5.1.1991:3) (22) [Quotingjames Baker] 'We fear that, as he normally does, [Saddam] will miscalculate where his limits are.' (12.1.1991: 5) (23) Saddam smug and convincedhe can win the war. (12.1.1991: 2; headline) (24) Saddam insisted again that his armed forces are able to fight a long battle'. (7.1.1991:3) The use of 'convinced', 'insisted again' a n d , in other examples, t h e use of verbs of speech, such as 'said' a n d 'asserted' underline the subjective character of Saddam Hussein's opinions. These are modality markers, used to present the writr r's position as regards t h e proposition. T h e readers, w h o knew about t h e capaury of the American war machine, inferred that S a d d a m Hussein was falling into :n> error of perception. After all, h e was just S a d d a m , a m a n with n o tides and ;lti anrhnrii-ti- Ac 4-1

n-

•

_

DISCOURSE STUDIES

120

opinions that described S a d d a m Hussein as a m a d m a n , b u t in its editorials, in particular in those that argued that war was unavoidable, presented h i m as such. T h e process of symbolic descriptive illustration b e c a m e explicit judgements or metadiscursive illustration: (25) Respecto a la responsabilidad ultima de este conflicto, la historia pedira cuentas al dictadoritaaxa, cuya locura e intransigencia han lanzado al mundo a una aventura de consecuencias imprevisibles. (17.1.1991: 18; editorial) (25) As for the ultimate responsibility for this conflict, History will ask the Iraqi dic tator whose madness and intransigence have sent the world on a course of unpre dictable consequences. (17.1.1991: 18; editorial) Both devices contributed to construct a portrait of S a d d a m Hussein as an unjust m a n , cruel, impious a n d wild. A s his portrait b e c a m e m o r e a n d more despicable, his adversary's b e c a m e m o r e a n d m o r e magnificent. George Bush was described as just, humanitarian, pious a n d civilized. George Bush himself often formulated explicitly the content of this contrast: (26) 'En los ultimos cinco afios, Sadam Husein ha demostrado su desprecio por las normas de comportamiento civilizado. En los proximos dias, Irak Ilega a la fech a lfmite que colma lapaciencia del mundo civilizado' (9.1.1991: 3) (26) 'In the last five years Saddam Hussein has showed his contempt towards civilized norms of behaviour. In the next few days Iraq will reach the deadline when the civilized world's patience will run out.' (9.1.1991: 3) F r o m the m o m e n t El Pais stopped asking S a d d a m Hussein to change his m i n d a n d adopted the strategy of emphasizing the contrast - that is, adopting George Bush's position - communication a n d m u t u a l understanding were no longer possible, there could n o t b e a c o m m o n 'rhetorical ground', a n d hence there could not b e a diplomatic solution to the conflict. Both the editorials, which labelled Saddam Hussein as a m a d m a n , a n d the news reports, b y appealing to emotion through invocation, implicitness and inference, presented an 'absolute otherness' to which the only possible response was its exclusion, rejection and destruction. H o w e v e r , u n d e r n e a t h the m e t a p h o r of the stranger, the irrational mad m a n , lies the image of evil a n d its incarnation. This is the key point that shows how, in westerners' views, evil encompasses all the features m e n t i o n e d above: S a d d a m Hussein 'has b e c o m e the angel of war w h o prefers desolation to life' (1.1.1991: 16).

Conclusions A system of rhetoric that aims to create a n d maintain the e n e m y is not only based o n exclusion, it also produces knowledge: the m e n a c i n g presence of the e n e m y creates the conditions for a p e r m a n e n t discourse that provides coherence for the system itself. T h u s , S a d d a m Hussein b e c o m e s convenient a n d necessary

MARTTN ROJO


121

T h e image constructed b y the press of S a d d a m Hussein, incarnating the e n e m y and i n c o r p o r a t i n g features such as injustice, pitilessness, irrationality a n d wildness, gave the illusion that this world, which is becoming m o r e homogeneous, is ruled b y the principles of reason, justice a n d respect for nations. This m e a n s that the U N were able to play their role in accordance with these same principles. This illusion palliated the 'crisis of space' (Auge, 1993) which appears associated with the continual extending of geographical space: increased facilities for the movement of goods, information and people seem to work against the under standing a n d control of space, a n d this, in turn, against the understanding a n d organization of social groups. Mixing of cultures a n d individualizing of con science a±e for Auge the signs of a n age in which the image of the other is m o r e diffuse and unstable than ever before, since the other can n o longer b e associated with a particular place (emigration has b r o k e n the links between the other a n d a place of origin). T h e Gulf War p r o d u c e d an absolute other, which the mass m e d i a m a d e seem close, but which continued to incarnate all our nightmarish fears (bestiality, madness, evil). S a d d a m Hussein, therefore, contributed to reducing the blurring of signs of identity b r o u g h t about b y globalization, at the same time as his image reinforced our collective references. This study has a i m e d to reveal n e w rituals of power, including the con cept of normalization, and to see h o w normalization creates subjects a n d selfknowledge. I h a v e b e e n especially interested in personification a n d its effect. It is the personification of the conflict which allows the conceptualization of the protagonists as the m a i n characters of the tale: the villain a n d the hero. But it is also personification which permits the a p p e a r a n c e a n d consolidation of a sole but all-embracing we, which is used to cover the way its referents change as a piece of discourse unfolds. As readers, we are spoken to, included in a n d absorbed by a unique concept of us, which turns into a n 'ideological u£. T h e aim of all this is to strengthen the ideology of consensus a n d m a k e it possible to recreate the illusion that it is these positive principles which unite the world against Saddam Hussein, a n d that it is in the defence of the ideology of consensus that the war is being waged. T h a n k s to the role played b y Saddam Hussein, and which he was forced to play, for a time the world was freed from irrationality, con tradictions, divisions, conflict, economic a n d political injustice, and, of course, from the difference which was e m b o d i e d in h i m . Now the d r e a m is over, what r e m a i n s is the discourse built o n ethnic and cultural difference. JE/Paw took two positions: that of calling S a d d a m Hussein to "become like us', and, later on, that of exclusion, showing the unyielding character of the differences separating both positions and advocating the destruction of die other. His otherness was radical, almost unthinkable in a world which con siders itself m o r e and m o r e h o m o g e n e o u s , a n d that is w h y it could only b e incarnated b y an individual. O n c e the war began, El Pais changed its discourse, since, then, its opposition could b e understood as treason. As Luis Martin Santos !>aid. 'defencelessness incapable of resisting the seduction of p o w e r is m u c h ^ ^ ^ B more serious at times of crisis, w h e n w e let fear o v e r c o m e us and we are afraid

DISCOURSE STUDIES

122

to reveal our true selves' (Martin Santos, 1988: 21). T h e Sun and other repre sentatives of the British press opted from the very beginning for a discourse of exclusion. T h e s e two type of discourses are to b e found, m o r e a n d m o r e often, in the treatment of immigration (Landowski, 1993). T h e first o n e is partly due to universalism a n d the second o n e is a consequence of a tendentious interpretation of relativism. O t h e r potential discourses seem to h a v e disappeared: that of respect towards differences, the idea of the cultural melting-pot or that proposed b y Balandier: 'considering differences to the point w h e n they stop being separating factors a n d b e c o m e multiple manifestations of the "unity of m a n " ; a unity grad ually discovered, not postulated, which is what the philosophies of the Universal d o ' (1993: 36). Exclusion is a response to the desire for particularization, a force opposed to globalization, at least w h e n we are n o t certain that this is going to b e according to our model. It is n o t coincidental that it b e c a m e activated at this time, since anti-Arab racism is the least inhibited of all racisms. M a y b e it is h e r e where the d r e a m turns into a nightmare. O u r j o u r n e y has taken us from the h u m a n b o d y to h u m a n territory: from the metaphors a n d m e t o n y m which m a k e it possible for states to b e conceived as people with h u m a n attributes, defects a n d virtues, to the anthropological space 'in which we try to find points of reference for our identity, our relations, our history' (Auge, 1993: 21). It is this real space, geographically far, b u t made close b y the work of the mass media, which is defended b y m e a n s of the appro priation of discourses. T h e same thing h a p p e n s with the imaginary space, in which the image of the beast, of'absolute otherness', so useful to some, incarnates all those qualities a n d b y its m e r e existence frees us from them. By allowing us a tamed taste for fear, the Gulf War m a d e possible a generalized discourse. But the fate of words a n d images can b e changed. If w e really want to, we can recover t h e m a n d continue their work. ' T h e word which has been m a r k e d b y power can b e dealienated b y our opposition, a n d this will go further ... find n e w forms, n e w words, n e w images, that is, n e w meanings which renew those words which have b e e n alienated' (Martin Santos, 1988:220). This is, in essence, what I have tried to do in this paper. W h a t I really wanted to do was go b e y o n d the statement that discourse is not transparent, that it is possible to take it over a n d m a k e it opaque. We n e e d n o t leave matters here - in 'their' hands. As readers, we should react, questioning the existence and relevance of that 'united us'. We should oppose the expulsion of irrationality a n d the illusion which n o t only transformed us into heroes b u t m a d e the world we live in into the incarnation of g o o d (see Naur, 1994, for a very similar proposal).

Notes I would like to thank Luis Eguren, Angel Gabilondo, Conception Gomez Esteban, Carmen Valle, Teun van Dijk and Rachel Whittaker, who were all able to read between the lines and suggest ways to improve the analysis. A first outline of this research was

MARTIN ROJO


123

presented by Luis Eguren and the author at the 4th International Pragmatics Conference (Kobe). Thanks also to Juan Manuel Delgado for his assistance in gathering data. The English translations in all the extracts are translations from Spanish rather than the original wording. 1. '[Discourse] surfaces as an asset - a finite, limited, desirable, useful one - that has its own rules of appearance, but also its conditions on appropriation and use; an asset that, consequently, brings out, from its very existence (and not only in its "practical applications") the issue of power; an asset which is, by nature, an object of struggle, of political struggle' (Foucault, 1969: 58; see also Gabilondo, 1990; 89-106). 2. In French in the original: 'dans toute societe, la production du discours et a la fois controlee, selectionee, organisee et redistribuee par un certain nombre de procedures qui ont pour role d'en conjurer les pouvoirs et les dangers, d'en maitriser Pevenement' aleatoire, d'en esquiver la lourde, laredoutable materialtie' (Foucault, 1971: 11). 3. This simply means that there is nothing to be interpreted. There is nothing primary to interpret, because, essentially, everything is already an interpretation. Each sign is in itself not the object offered to be interpreted but the interpretation of other signs' (Foucault, 1967: 35-6). 4. As El Pais (17.1.1991; editorial) said: 'In the midst of that confusion, the Spanish government's response from the beginning of the invasion has tried to be coherent with the country's new responsibilities as a member of the European Community and of the West European Union, as well as with the resolutions of the UN.' ('En medio de ese desconcierto, la respuesta dada por el gobierno espanol desde el comienzo de la invasion ha pretendido ser coherente con las nuevas responsabilidades de nuestro pais como miembro de la Comunidad Europea y de la Union Europea Occidental, asf como con los mandates de las Naciones Unidas'.) 5. 'In order to act, we need at least to identify' (Canguilhem, 1966: 63). 6. See Blommaert and "Verschueren (1994) for a similar process of overgeneralization in the conceptualization of the relationships with migrants and in the justification of discriminatory politics and behaviours. 7. 'I, Lucifer, mighty and independent emperor, free and absolute lord of the whole of the "Sulano" kingdom, despotic lord in all my formidable jurisdiction, forbidding and very noble, under whose orders everything moves, arbiter of all fortunes and all certainties' (Barreiro, 1973: 283-4). 8. 'With respect to these researchers sometimes bring up the idea of a crisis of identity. But, if such a crisis exists, in my opinion, it is nothing more than the natural conse quence of two previous crises: that of space (how can I imagine, at one and the same time, the planet as if it were a canton, and my canton as if it were a complete world?) and that of otherness. The stability of the other was what made it possible to conceive of and identify her/him with ease. When it was a question of a faraway other, there was no problem: one only saw her/him if one travelled or, at least, if one went to visit the Colonial Exhibition' (Auge, 1993:33). 9. It is worth observing the features in common with the definition of madness in the Encyclopedic: 'To get away from reason without knowing it, because one lacks ideas, is to be an imbecile. To get away from reason knowingly because one is enslaved by a violent passion is to be weak. However, to get away from it confidently and being convinced that one is following it, that is my view what is called madness' (Encyclopedic, entry 'Madness', quoted by Foucault, 1964).

DISCOURSE STUDIES

124

References Alvarez-Uria, F. (1983) Miserables y locos: Medicina mental y orden social en la Espana del siglo XIX Barcelona: Tusquets. Auge, M. (1993) 'Espacio y Alteridad', Revista de Occidente 140: 13-34. Bakhtine, M. (1981) Le principe dialogique. Paris: Seuil. Balandier, G. (1993) 'La aprehension del otro: antropologia desde fuera y antropologia desde dentro', Revista de Occidente 140: 35-42. Barreiro, B. (1973) Brujos y astrohgos de la Inquisition de Galiciay elfamoso libro de San Cipriano. Madrid: Akal. Blommaert, J. and Verschueren, J. (1994) 'La retorica de la tolerancia o lo que se ensefia a los agentes de policia sobre los immigrantes', in L. Martin Rojo, C. Gomez Esteban, F. Arranz Lozano and A. Gabilondo (eds) Hablar y dejar hablar (sobre racismo Y xenofobid) pp. 27-44. Madrid: Universidad Autonoma de Madrid. Canguilhem, G. (1966) Le normal et le pathologique. Paris: PUF. Chibnall, S. (1977) Law and Order. London: Tavistock. Deleuze, G. (1986) Foucault. Paris: Minuit. Ducrot, O. (1984) Le Dire et le dit. Paris: Minuit. Fairclough, N. (1985) 'Critical and Descriptive Goals in Discourse Analysis'./oanwZ of Pragmatics 9: 733-63. Fairclough, N. (1989) Language and Power. London: Longman. Fairclough, N. (1991) 'Critical Linguistics', in W. Bright (ed.) Oxford International Encyclo paedia ofLinguistics. Oxford: Oxford University Press. Fairclough, N. (1992) Discourse and Social Change. Cambridge: Polity Press. Foucault, M. (1961) Historic de la folic a I'age classique, Vol. 1. Paris: Plon. Foucault, N. (1967) Nietzsche, Freud, Marx. Paris: Minuit. Foucault, M. (1969) L'archeologie du savoir. Paris: Gallimard. Foucault, M. (1971) L'ordre du discours. Paris: Gallimard. Fowler, R. (1988a) 'Notes on Critical Linguistics', in T. Threadgold and R. Steele (eds) Language Topics: Essays in Honor of Michael Halliday. Philadelphia, PA and Amsterdam: John Benjamins. Fowler, R. (1988b) 'Critical Linguistics', in K. Malmjaer (ed.) The Linguistic Encyclopaedia. London: Routiedge & Kegan Paul. Fowler, R. (1991) Language in the News. London: Routiedge. Fowler, R., Hodge, B., Kress, G. and Trew, T. (1979) Language and Control. London: Routiedge & Kegan Paul. Gabilondo, A. (1990) El Discurso en action. Barcelona: Anthropos. Galtung, J. and Ruge, M. (1973) 'Structuring and Selecting News', in S. Cohen and J. Young (eds) The Manufacture ofNews: Social Problems, Deviance and the Mass Media, pp. 62-72. London: Constable. Geffroy, A. (1985) 'Les nous de Robespierre', Mots 10: 63-90. Gillaume, M. (1993) 'El otro y el extraflo', Revista de Occidente 140: 43-58. Guespin, L. (1985) 'Nous, la langue et l'interaction', Mots 10: 45-63. Hall, S. (1976) 'A Word at One with Itself, in S. Cohen andj. Young (eds) The Manufacture of News: Social Problems, Deviance and the Mass Media, pp. 85-94. London: Constable. Hartley, J. (1982) Understanding News. London: Methuen. Hodge, R. and Kress, G. (1992) Language as Ideology (2nd edn). London: Roudedge. Howlett, M.W. (1982) 'Contr'Un', L'Ane 6: 15. Imbert, G. and Vidal Beneyto, J., eds (1986) El Pais o la referenda dominante. Barcelona:

MARTfNROJO


125

Jex, W. (1983) The Language ofthe News: A Comparative Semantic Analysis ofthe Language of Network Television and New York City Newspaper Reports. New York: New York University Press. Kerbrat-Orecchioni, C. (1987) 'La mise en places', in J. Cosniere and C. KerbratOrecchioni (eds) Decrire la conversation, pp. 319-52. Lyon: PUL. Kress, G.R. (1991) 'Critical Discourse Analysis', in W. Grabe (ed.) AnnualReview of Applied Linguistics, Vol. 2, pp. 84-100. New York: Cambridge University Press. Kristeva, J. (1986a) 'Word, Dialogue and Novel', in The Kristeva Reader (ed. T. Moi). Oxford: Blackwell. Kristeva, J. (1986b) 'The System of the Speaking Subject', in The Kristeval Reader (ed. T. Moi). Oxford: Blackwell. Lakoff, G. (1987) Women, Fire andDangerous Things. Chicago/London: Chicago University. Press. Lakoff, G. (1992) 'Metaphors and War: The Metaphor System Used to Justify War in the Gulf, in M. Putz (ed.), Thirty Years ofLinguistic Evolution, pp. 463-81. Philadelphia, PA and Amsterdam: John Benjamins. Landowski, E. (1993) 'Ellos y nosotros: notas para una aproximacion semiotica a algunas figuras de la alteridad social', Revista de Occidente 140: 98-118. Ledrut, R. (1987) 'Societe reelle etsociete imaginaire', Cahiers Internationaux de Sociologie 82: 41-56. Marafioti, R, ed. (1991) 'Un ejemplo de argumentacion peridistica', in R. Marafioti (ed.) Temas de argumentacion, pp. 109-43. Buenos Aires: Biblos. Martin Santos, L. (1988) Diez lecciones de sociologia. Madrid and Mexico D.F.: Fondo de Cultural Economica. Nai'r, S. (1994) 'El otro como enemigo', in L. Martin Rojo, C. Gomez Esteban, F. Arranz Lozano and A. Gabilondo Pujol (eds) Hablar y dejar hablar (sobre racismo y xenofobia) pp. 231-41. Madrid: Universidad Autonoma de Madrid. Norton, R. (1983) Communicator Style. Beverly Hills, CA: Sage. Pintos de Cea-Naharro,J.L. (1994) 'Orden social e imaginarios sociales', Papers (in press). Poliakov, L. (1980) La Causalitediabolique.-Essaisur I'origine despersecutions. Paris: CalmannLevy. Rochberg-Halton, E. (1986) Meaning and Modernity: Social Theory in the Pragmatic Attitude. Chicago, IL: Chicago University Press. Sandell, R. (1977) Linguistic Style and Persuasion. London: Academic Press. Sherer, K.R. and Giles, H., eds (1979) Social Markers in Speech. Cambridge: Cambridge University Press. Subirats, E., ed. (1977) 'La distincion de sujeto', in S. Freud and G. Groddeck, Corres pondent pp. 7-25. Barcelona: Anagrama. van Dijk, T.A. (1988) News as Discourse. Hillsdale, NJ: Erlbaum. van Dijk, T.A. (1991) Communicating Racism: Ethnic Prejudice in Thought and Talk. Philadelphia, PA and Amsterdam: John Benjamins. van Dijk, T.A. (1993) 'Analyzing Racism through Discourse Analysis. Some Methodological Reflections', in J.H. Stanfield II and R.M. Dennis (eds) Race and Ethnicity in Research Methods, pp. 92-139. Newbury Park, CA: Sage. van Dijk, T.A. and Kintsch, W (1983) Strategies of Discourse Comprehension. New York: Academic Press. Wetherell, M. and Potter, J. (1992) Mapping the Language of Racism: Discourse and the Legitimation ofExploitation. New York: Harvester Wheatsheaf.

. ' . I I

-

.

52 Multimodal Resources for Turn-Taking: Pointing and the Emergence of Possible Next Speakers Lorenzo, Mondada

Introduction

I

1

n this p a p e r , 1 study a specific gestural practice through which a recipient gradually establishes h e r or himself as next speaker in talk-in-interaction: b y using a pointing gesture, while current speaker's turn is still going on and has n o t yet reached its completion. T h u s , pointing gestures will b e studied h e r e n o t as a resource primarily devoted to the accomplishment of referential or deictic actions b u t as a resource locally mobilized for self-selecting and for organizing the emergent character of a «transition space » {Schegloff, 1996,96). This practice a n d resource will b e explored in its orderly a n d systematic char acter in a specific setting, an informal work session where participants - a multidisciplinary t e a m of agronomists a n d computer scientists - are involved in a discussion a r o u n d a table, whilst writing notes a n d looking at m a p s . Over the last years, the «turn-taking madiinery »introduced b y Sacks, Schegloff a n dJefferson (1974) has b e e n extensively studied, b o t h in its turn-constructional component, allowing the identification of relevant spaces for turn transition, a n d in its turn-allocational component, constituted b y selection techniques by which either current speaker selects next or next speaker self-selects. Explorations of the turn-allocational componenthave highlighted various prac tices for selection (such as turn-sharing a n d other choral performances, Lerner 2002; various m e t h o d s for addressing the n e x t speakear, Lerner, 2003) and their eventual specialization a n d diversification in conversation, institutional a n d professional contexts (namely in talk at work, Drew & Heritage, 1992).

.. i T „ i„t,j r

nt!

i

n

Rnirlish)

Cahiers de Linguistique Francaisevdi. 26,2004, pp. 269-292.

MONDADA

MULTIMODAL RESOURCES FOR TURN-TAKING

127

The consideration of various settings has implied a diversification of the resources studied (verbal a n d vocal resources for telephone conversations, multimodal resources for face-to-face interactions, artefacts and technologies for m e d i a t e d interactions, etc.). Explorations of the turn-constructional components h a v e b e e n carried out with the collaboration of linguists over the last years (within interactional linguistics: see Ochs, Schegloff, T h o m p s o n , 1996, Selting & Couper-Kuhlen, 2001, Ford, Fox & T h o m p s o n , 2002, HakuUnen & Selting, 2005): they m a d e possible the identification of Turn-Constructional Units (TCUs) as interactive organizational units formatted b y exploiting linguistic resources in ways that are b o t h shaped by the contingencies of the step-by-step incremental organization of interaction and b y the constraints a n d possibilities of g r a m m a r , o n e being reflexively struc tured b y the other. T C U s are a specific kind of unit, praxeologically defined b y the local achievements of the participants, interactively negotiated, emerging within the temporal m o m e n t - b y - m o m e n t unfolding of talk as it is joinly, situatedly, contingently produced b y the participants. Opportunities to co-participate to and interactively produce emerging T C U s are centrally provided b y their pro jective potential, responsible for the predictible features of a turn's emergent construction. Projectabuity (Sacks, 1992, II, 4 0 ; Sacks, Schegloff, Jefferson, 1974: 702; Schegloff, 1984:267) is a central feature that provides for n e x t speakers the possibility n o t only to identify turn completion b u t also to predict it before it oc curs, to anticipate transition-relevance points a n d to locate the u p c o m i n g place where to begin to speak (see Ford, Fox & T h o m p s o n , 1996; Selting 2000; Tanaka 2001, Auer, 2005 for a description of linguistic resources for projectability): The sentence is a great packaging technique for a series of reasons [...]. It has a struc ture which can at all points be seen as to whether it is possibly complete or not possibly complete, and people are able to deal with it in such a way as to see, on its occurence that it's possibly complete. And also, from its beginning it can be looked at to see what it will take to complete it. (Sacks, 1992, II, 40) Sentential constructions are capable of being analysed in the course of their production by a party/hearer able to use such analyses to project their possible directions or completion loci. (Sacks, Schegloff, Jefferson, 1974, 709) These features are responsible for the flexible, m o m e n t - b y - m o m e n t , dynamic, revisable a n d e x p a n d a b l e definition of turns b y the participants; in this sense, turns and T C U s are «interactively determined » units (cf. Ford, 2 0 0 5 emphasizing this praxeological a n d interactive aspect): The turn-unit is f a sort which a) employs a specification of minimal sizes, but b) pro vides for expansion within a unit, c) is stoppable (though not a any point), and d) has transition places discretely recurring with it, e) which can themselves be ex panded or contracted; all of these features except the first are locii of interactional determination. By virtue of this character, it is misconceived to treat turns as units characterized by a division of labor in which the speaker determines the unit and its boundaries, with other parties having as their task the recognition of them. Rather, the turn is a unit whose constitution and boundaries involve such a distribution of

128

DISCOURSE STUDIES

tasks as we have noted: that a speaker can talk in such a way as to permit projection of possible completion to be made from his talk, from its start, allowing others to use its transition places to start talk, to pass up talk, to affect directions of talk, etc.; and that their starting to talk, if properly placed, can determine where he ought to stop talk. That is, the turn as a unit is interactively determined. (Sacks, Schegloff, Jefferson, 1974: 726-727) Therefore, a central task for conversation analysis has b e e n to describe methods a n d resources b y which recognizable units in action are interactively a n d reflexively p r o d u c e d a n d m o n i t o r e d . For e x a m p l e , previous research has shown h o w turns are interactively constructed (Goodwin, 1979, 1981), h o w T C U s can b e e x t e n d e d b e y o n d the initially projected completion allowing for the timely production of assessements b y the other party within the unit's boundaries (Goodwin & Goodwin, 1987); h o w they can h a v e a « semi-permeable » character and their second p a r t can b e predicted for the production of collaborative sentences (Lerner, 1991, 1996). T h e dates of these studies show that the import ance of multimodal resources for these collaborative a n d reflexive practices has long b e e n recognised (see also Hayashi, 2 0 0 5 , Schmitt, 2005). Analyses of the methods involved in projection a n d in the online incremental construction of T C U s were first focussed o n linguistic resources, but w e r e soon e x p a n d e d in order to take into consideration gestures, gazes a n d bodily postures. Linguistic r e s o u r c e s c o n s i d e r e d h a v e b e e n p r i m a r i l y syntactic (Sacks, Schegloff, Jefferson, 1974; Sacks, 1992,1,642 referred to « sentences »,« clauses >, etc. although pointing to the practical a n d emic character of these units, n o t cor responding to their definition in classical linguistics) a n d syntax is still considered as one important resource for the « packaging » of turns (see Schegloff, 1996 a b o u t syntactic materials characterizing initial, m i d d l e a n d t e r m i n a l T C U positions; Auer, 2005 o n the projective potentials of syntax). But its interplay with prosodic resources has b e e n largely explored since (see already Sacks, 1992, 1,651; Auer, 1996, Couper-Kuhlen, 1996,2001,2004). Moreover, b y the nineteen eighties, French & Local (1983), Local, Kelly, Wells (1986), Local & Kelly (1986) were showing the import of phonetic resources for turn projection. Instead of claiming the relevance of o n e u n i q u e dimension, studies focus m o r e o n the con tributions of syntactic, prosodic, a n d pragmatic resources for achieving comple tion (Ford & T h o m p s o n 1996; Ford, Fox, T h o m p s o n , 1996 take into consideration gaze and b o d y m o v e m e n t s too). M o r e recendy, multimodal resources h a v e been strongly taken into account, including gestures (Schegloff, 1984, Streeck & Hartge 1992; Ford, Fox, T h o m p s o n , 1996; Hayashi, Mori, Takagi 2002), gazes (Goodwin, 1981; L e r n e r , 2 0 0 3 , R o s s a n o , 2005) a n d m o r e generally t h e entire b o d y (Goodwin, 2000). Establishing Speakership as a Practical Accomplishment T h r o u g h the analysis of multimodal practices for establishing speakership in a corpus of video recordings of professional meetings, this p a p e r aims at con tributing to the growing literature o n turn-taking o n the following aspects:

MONDADA


129

• the orderly a n d systematic way in which multimodal resources are exploited for the organization of turn-taking and m o r e particularly within practices for self-selecting contingendy defining-and-exploiting transition spaces; • the reflexive organization of turns-at-talk, relying b o t h o n the production of recognizable emergent structures b y current speaker a n d o n the local interpretation, monitoring a n d online analysis d o n e b y a n d i m p l e m e n t e d in recipients/next possible speakers actions. Multimodal practices con tribute crucially to the visibility for participants - a n d for analysts alike of public displays a n d recognitions of « in-course parsing of a turn in progress » (Jefferson, 1984: 14). • the interactive organization, identification a n d exploitation of « p r e beginnings », i.e. of actions taking place before the turn properly starts, in pre-initial t u r n positions, a n d projecting the beginning of a possible n e x t turn: this position can b e occupied b y h e a d movements, gaze redirections, onset of gestures, incipient facial expressions, in-breaths,« u h ( m ) » t o k e n s (Schegloff, 1 9 9 6 , 9 2 - 9 3 ) , i.e. actions that are not yet realized as a turn b u t which strongly project it. We will show that multimodal actions play a crucial role in this position, allowing for their simultaneous production with ongoing current-speaker's talk. Analysis of these actions e x p a n d s our focus of study b e y o n d the limited «transition-relevance p o i n t » t o the more conceptually extensive «transition space », characterized b y its shape, duration, i m p o r t and limits (Schegloff, 1996: 96-97). • the very notion of speakership, a n d other talk-related categories (« next speaker », «incipient speaker », « current speaker », etc.), as well as their rights a n d obligations, as products of emergent joint accomplishments: the d y n a m i c , locally organized, contingent emergence of a « speaker » provides for the occasion to respecify this category as it is recognized a n d accomplished b y participants themselves in the course of their interaction. • the articulation between the context-free a n d the context-sensitive dimensions of turn-taking: although they recognized the situated order of turn-taking, Sacks, Schegloff & Jefferson were m o r e concerned with the abstraction of the machinery rather than the particularities of content, context or identity (1974:699) a n d abstained from precisely defining « t h e scope of reference of 'context' that is relevant » (1974: 6 9 9 n 8). This p a p e r will show h o w situated features of the material and spatial environment are methodically mobilized for achieving self-selection a n d in turn shed some light o n m o r e general practices for turn-taking.

D i s c o v e r y of t h e P h e n o m e n o n In what follows, w e will analyze turn-taking practices observable in a particular corpus, dealt with as a « perspicuous setting »(Garfinkel & Wieder, 1992,184), i.e. a setting that in its specificity a n d uniqueness allows us to highlight methodic and systematic features. This setting provides for specific multimodal resources, related to the spatial a r r a n g e m e n t of the nartif-mantc t r . ~-*~r~~i- ^—

130

DISCOURSE STUDIES

manipulate a n d to their peculiar activity. T h e s e resources are contingent, asso ciated with the specificity of the activity; nevertheless, they are mobilized within systematic practices which can shed some light o n other turn-taking and turntransition practices. T h e professional activity concerned is a series of work meetings (the corpus is c o m p o s e d of 6 sessions, for about 15h of video data) in which agronomists a n d computer scientists collaborate to develop a c o m m o n cartographic lan guage for modelling agricultural land. T h e agronomists collected a series of geographical a n d land-survey m a p s of several farms a n d created a new, more abstract, topological representation of these territories; the c o m p u t e r scientists w e r e engaged in producing a further formalization, in the form of graphs. These activities involve several reformulations a n d re-descriptions of the initial m a p representation, as well as a n u m b e r of controversial discussions about the descrip tive categories each discipline is using (for detailed analysis of this « multimodal semantics in action » see M o n d a d a , 2005a, 2005b). T h e m a p s a n d other visual izations lay o n the w o r k table, around which three or four participants are seated; their attention is focused on the documents that are discussed, described, pointed at, a n d even modified in the course of the talk. T h u s , the activity considered is characterized b y a specific ecology, w h e r e a limited work space full of artefacts mobilized b y talk a n d gesture is at the centre of the collective attention. In contrast to other w o r k settings w h e r e multi-activity occasions a constant dis persion of attention in fragmented spaces a n d w h e r e mutual attention has to be constantly re-achieved t h r o u g h intense interactional w o r k (see G o o d w i n & Goodwin, 1992, Goodwin, 2 0 0 3 , H e a t h & Luff, 1992, H e a t h et alii, 2002 for examples), the activity studied h e r e is characterized b y a sustained focus of col lective attention on the documents at the centre of the table. This peculiar ecology of action provides for specific resources for the organization of interaction, both in a indexical a n d systematic way.

Figure 1: Perspective view

Figure 2: Vertical view

T h e w a y in which this activity has b e e n videotaped is reflexively sensitive to the features of action: a perspective view, allowing to see the u p p e r part of the par ticipants' bodies as well as their mutual orientations, is completed b y a vertical view, allowing the detailed description of the material and spatial environment

[ONDADA


131

involved, as well as participants' gestures (for a detailed analysis of recording choices as reflexively shaping analytical possibilities, see M o n d a d a , 2006a). Looking at the video recording of this activity, a recurrent action is noticeable: pointing gestures d o n e b y participants t o w a r d t h e m a p s a n d other docu ments o n the table are highly frequent - either in the form of one person point ing or even in the form of various participants pointing at the same time to the same item. Pointing gestures within talk h a v e b e e n mainly investigated in the literature in relationship with deictical reference (Hanks, 1992; K e n d o n , 2 0 0 4 ; Kita, 2003): they are a pervasive resource for identifying a n d m a k i n g recognizable an object (even if it is n o t visible for the participants, Haviland, 1996), within a referential practice which m a y or m a y not co-occur with other spoken resources, e.g. deixis. Studies of pointing-in-interaction (Goodwin, 2 0 0 3 ; H i n d m a r s h & Heath, 2000, M o n d a d a , 2005c) h a v e s h o w n that the articulation b e t w e e n deicticals a n d pointing gesture is not so simple as it might appear in previous accounts: pointing gestures are precisely timed, b e i n g synchronised with the m o m e n t - b y - m o m e n t organization of talk-m-interaction, with recipient-oriented talk a n d bodily con ducts, with appropriate arrangments of bodies a n d objects in space, with the progressive (re) disposition of the d o m a i n of scrutinity, with contingent d e m a n d s of the activity, a n d with timely organized coordination. Moreover, pointing gestures are not always and exclusively devoted to deic tical reference. I n our corpus, at the first sight, these gestures seem to b e related to the descriptive practices characterizing this work activity, as well as to the common attention towards shared referents. However, detailed examination show that these gestures are n o t systematically associated with deictical forms and are m u c h m o r e frequent t h a n referential expressions. Pointing gestures d o not seem to h a v e just a referential use. T h a t p r o m p t s a closer analysis of the cir cumstances of their finely-timed unfolding in interaction. This noticing has o p e n e d another perspective o n the corpus, raising another kind of issue: speakers do point to the m a p s n o t only in order to refer to t h e m but also to m a k e visible their engagement a n d participation in interaction. T h e occurrence and temporality of pointing gestures is a key resource for the organ ization of turn-taking. Pointing practices a n d turn-taking practices are h e r e deeply e m b e d d e d , in a w a y that exploits the specificities of the situated action such as the fact that it takes place around a table covered with artefacts, the fact that participants' attention is focused m o r e o n these objects rather t h a n o n coparticipants, a n d the fact that topical activities are centred o n visible materials. The specificity of the situation is thus locally exploited b y the interactional prac tices, although their exploitation remains related to m o r e general principles governing action a n d interaction. O u r analysis will b e organized a r o u n d two basic questions: the first concerns the sequential e n v i r o n m e n t at which pointing begins, characterizing the context in which the recipient/possible n e x t speaker is emerging a n d establishing as a speaker. T h e second concerns the sequential e n v i r o n m e n t w h e r e pointing ends, showing that interestingly this does n o t always occour at the established speaker's turn completion. T h u s , analysis will show a range of ordered sequential positions

132

DISCOURSE STUDIES

at which pointing as a technique for self-selecting is used b y participants, and will reflect u p o n the w a y in which this technique reflexively shapes sequentiality as it is practically m a n a g e d b y co-participants.

W h e n d o e s Pointing Begin?: Practices for Self-Selecting I n what follows, w e will focus on the w a y in which pointing is exploited b y par ticipants as a practice for self-selecting a n d for projecting current speaker up coming turn completion. This descriptive aim makes some strict d e m a n d s of the data used (Mondada, 2006b): o n the one h a n d , it requires videorecords that m a k e relevant details accessible for the analyst: this is the reason w h y we will use multiple video sources. O n the other h a n d , it requires transcripts that repre sent the finely tuned aspects of multimodal details as they sequentially unfold in interaction - making t h e m available for analysis. A rough notation of gesture would erase the orderly organization of the very p h e n o m e n a we try to study. In order to preserve them, we will adopt a specific notation system for gestures, complementing Jefferson's conventions used for talk. T h e system, inspired by Goodwin's gaze notation (1981) a n d b y Schegloff s gesture notation (1984), aims at capturing the m a i n phases of a gesture's trajectory a n d to synchronize them with talk: we will describe the preparation phase (represented b y dots ....), the culmination or stroke, which can b e held for a while (represented b y dashes — ) , a n d the retraction (represented b y commas,,,,) (see the e n d of this p a p e r for a full account of conventions used). Turn-initial Pointings: Displaying Incipient Speakership I n the following excerpt, Pierre-Alain (PAL) is engaged in a long multi-unit turn about the w a y in which farmers use their land over the year. W h e n it is brought to completion, Viviane (VIV) self-selects a n d relates Pierre-Alain's explanation to its cartographical representation. She is overlapped b y a third speaker, Laura (LAU), asking for a confirmation about the object represented o n the m a p . (1) (e9/agro 1-47.00) 1 PAL ben suivant le cas euh: ben on tra- (.) on est la que pour le well in certain cases ehm well one wor- (.) one is there only for the 2 champ, et puis a d'autres moments:, ben on va echouer, (0.3) field, and at other times, well one ends up, (0.3) 3 en paturage. ,h sur I'assemblage sans parcours. .h je pense in pasture, .h on rough ground without any path, .h i think 4 que dans le cas du gaec du pradou, .h c'est tout Pun, that in the case of the ((region's name)), .h it's either one, 5 tout l'autre. or the other. 6 VIV +.hh#oui. #parce# que:i#m' *sem#+ble: ehi*- ici# c'etait s::- ce .hh yes. because it seems to me eh he-here it was w::-what fig #fig.l#fig.2 #fig.3 #fig.4#fig.5 Mg.6 + +PPP p e n — » w

MONDADA

7 qui: 8 LAU


133

ce qu« 9 a voulait representor, [c' etait this wanted to represent, [was [ *c'est les am*andes ca? [ are these almonds? ((=kind of field)) * *PPP finger-» w

Figure 4

que : i#

Figure 5 'm sem-#ble:

Figure 6

eh i- ici#

W h e n Pierre-Alain has reached a recognizable turn completion (syntactically, prosodically, pragmatically), Viviane, line 6, self-selects: she begins h e r turn with two turn-entry-devices: with an in-breath (« .hh ») and with her a r m beginning to stretch over the table (« ....»). A pointing gesture emerges from this stretching of her arm a n d h e r b e n d i n g over the table, showing a direction with h e r p e n extending the h a n d m o v e m e n t . Excerpt (1) shows the occurrence of a second instance of turn-initial pointing: Laura self-selects overlapping Viviane (line 9) a n d begins to point at the beginning of the turn too. This occurrence of turn-initial pointing is d o n e with the finger and not with a p e n (the p e n is b e i n g used b y L a u r a writing notes). Some remarks can b e m a d e at this stage: • for the description of the gesture, the sequential placement of the beginning of its preparatory phase is crucial, and allows the identification of a relevant sequential environment for the p h e n o m e n o n . • the pointing gesture is not limited to the h a n d : it engages the whole b o d y (the u p p e r b o d y b e n d i n g over the table, the a r m crossing the table). More over, the e m b e d d e d n e s s of pointing in other activities a n d in the ecology of the ongoing action shapes the form a n d the resources of pointing: point ing with a pencil or pointing with a finger allow different kinds of m o v e ments, of ways of being precise, of parallel simultaneous activities. • Viviane's pointing trajectory is n o t a m e r e individual isolated gesture b u t a finely coordinated one, synchronized with the format of the turn in pro gress a n d adjusted to other's actions-

134

DISCOURSE STUDIES

(2) = (1) line 6

6 V I V -Khh oui. parce que: im' *sem+ble: eh i*- ici + lau

w

e n

>

+PPP P — ' ' *opens folder*

>

Viviane's turn is formatted in a way that delays (with the insertion of « i m'semble » a n d the self-repair of « i- ici » - which are thus not just hitches in h e r talk but resources for the production of a timed action) the position of the deictical«ici» until the point where the pointing gesture has reached its maximal extension (when she says « i m ' s e m b l e »). This in turn is reflexively adjusted to Laura's action: Laura is holding a closed folder where the referent of Viviane's utterance is h i d d e n a n d which therefore constitutes a n obstacle for her referential action. Laura opens it as the pointing gesture reaches it. So, Viviane's gesture a n d turn o n the one h a n d , a n d Laura's manipulation of the referred artefact o n the other h a n d are finely adjusted a n d reflexively organized. Pre-initial Turn Pointings: Projecting Self-selection, Claiming Speakership T h e first case e x a m i n e d h e r e , although it exhibits two occurrences of the same p h e n o m e n o n - turn-initial pointing -, is relatively less frequent than another p h e n o m e n o n : pre-initial turn pointing. In most of the cases observed in the corpus, pointing projects self-selection well before the completion of current speaker's turn a n d well before the beginning of n e x t speaker's turn. H e r e is a n occurrence: (3) (e3/agro 1-16.48) 1 PAL 2 3 4 5 la 6 LAU

7 PAL 8 LAU

et done on voit la loglque, avec cet cet amenagement du: de and so one sees the logic, with this this settlement of: of l'espace, .hh qui revient a obtenir, (.) euh: des des de- des space, .hh which would make (.) ehm paturages, (.) RElativement plus productives, qu'elles ne pasturelands, (.) RElatively more productive, than what they l'etaient avant, grace notamment a la: (.) < l a were before, thanks among other things to the (.) * redistribution of common goods. ((decrescendo))> * * *et done le: ce qui est en (.) orange ici la, terre assolee and so the: what is coloured in (.) orange there, farmland *PPP - » hum e'est, (.) des prairies, that's, (.) grassland,

I n this case, Laura begins to point well before her turn beginning: the prepara tion of h e r gesture takes place before, b u t reaches its maximal extension exactly

MONDADA


135

in turn-initial position. If we consider the sequential position in which the preparation begins, w e notice that it is reflexively adjusted to Pierre-Alain's ongoing turn production: his last utterance is syntactically quite complex, but could be syntactically complete with the end of the comparison (« qu'elles ne Petaient avant » 3-4). Pierre-Alain adds another constituent (beginmng with « grace a » 4). This last add-on is characterized by some hitches (« a la: (.) la redistribution » 4-5) and b y a decrescendo low voice until the end of the turn. Laura initiates the preparation of her pointing gesture just after the beginning of the descrescendo, projecting an imminent turn completion. Therefore, her move ment anticipates the upcoming transition-relevance place and projects her as possible next speaker. Pointing begins properly at the same time as her estab lishment as self-selected current speaker. This occurrence shows some more general points: • pointing orients to transition-relevance points, even if it can occur much earlier (see below). • pointing appears as a method fox projecting self-selection, being part of an emergent movement for establishing upcoming speakership; • pointing makes visible for all participants a pre-beginning, taking the visual floor but not yet the vocal-spoken turn; • this pre-beginning phase anticipates the possible completion of the current speaker's unit in progress: a) it exhibits the monitoring of the structure in progress b y the possible next speaker, and more precisely his finely tuned online analysis for all practical purposes, b) it projects the next action i.e. the turn to be taken by the recipient/incipient speaker who is pointing. • thus, pointing has more to do with the temporality of the interactive con struction of TCUs in progress than with referential practices perse- since pointing begins well before deictics occur and even in absence of deictics, that is, well before the relevance to deictically indicate something. Never theless, if w e observe where deictics are positioned in these utterances we can say that deictical reference in turn-initial position can be a resource for enhancing the legitimacy and accountability of pointing as a method for establishing speakership. Deployment of gesture at transition-place has been described by Streeck & Hartge (1992), who analyze two gestural techniques being used in this position as gestural turn-entry devices in Ilokano conversations: the [a]-face, preparing the articulation of [a], even when no sound is produced, making visible for all participants the speaker's gearing up to utter the vowel; and the « palm up » gesture employed not only for claiming the floor but also for providing co-participants with a preview of the type of projected talk, prefiguring a list. These cases confirm that gestures have a potential of anticipation: not only iconic gestures precede their speech-affiliates; more generally, gestures are oriented to the organization of interaction and project relevant sequential pos itions, such as transition-points. In our corpus, pointing gestures do not just precede their deictical affiliates, but are more generally exploited for 2

DISCOURSE STUDIES

136

projecting self-selection (in another paper, Streeck, 1995 speaks of « action projectors »); they occur at transition places too, b u t m o r e generally they t e n d to start even before. T h e aim of this p a p e r is precisely to explore the kind of « projection space » (Schegloff, 1984) they o p e n u p . 3

If w e turn b a c k to the first example, a n d reconsider it, as well as its transcript, w e notice that although the preparation of the pointing gesture takes place at the beginning of the turn - that is, relatively late if c o m p a r e d to except (2) VTV accomplishes other actions before she self-selects. T h e y are represented in a second version of the transcript: (4) = (1) lines 2-7 2 PAL et puis a d'autres moments:, and at other moments, 3 ben on va echouer, (.) en paturage, .h well one ends up, (.) in pasture, .h 4 +sur l'assemblage +sans parcours. .h +je pense que+ on an rough ground without any path, .h i think that viv + +rnoves paper 15 +dans le cas du gaec du pr+adou, .h c'est tout Pun, in the case of the ((region's name)), .h it's either one, viv -Rums wrist, comes in h 6 tout 1'autre. or the other. 7 VIV + .hh oui. parce que: i'm'sem+ble: eh i- ici c'fitait .hh yes. because it seems to me eh he- here it was + +PPP p e n — » w

V i v i a n e ' s m o v e m e n t s e x h i b i t a v e r y s h a r p a n d timely u n d e r s t a n d i n g of Pierre-Alain's ongoing turn. T h e latter is characterizing o n e form and use of pasture: his T C U could b e syntactically c o m p l e t e line 3 b u t prosodically, continuing intonation projects m o r e to come, that is a prepositional syntagm (« sur l'assemblage sans parcours. » 4). Towards the e n d of this added-on con stituent, Viviane begins to m o v e , anticipating a n u p c o m i n g transition-relevance place. But shortiy after Viviane m o v e d the p a p e r o n which they are working, what comes next is n o t h e r self-selection b u t a n e w T C U initiated b y PierreAlain, continuing with a n in-breath a n d a stance m a r k e r (« .h j e pense que » 4). At that point, Viviane withdraws. T h e « Gestalt» of the n e w T C U has a n argu mentative binary form, stressed b y a contrastive prosody, projecting clearly its u p c o m i n g completion. At its beginning, Viviane's h a n d comes closer again to the centre of the table, in a position that enables a n d facilitates the pointing gesture which follows. Thus, the initial-turn position of her pointing gesture has been p r e p a r e d a n d projected b y other actions before, showing h e r anticipatory on going understanding.

MONDADA MULTIMODAL RESOURCES FOR TURN-TAKING

137

Recipients' predictions can have various scopes, and can exhibit a very early orientation to the next possible transition-relevance place: (5) (agro 1 /19.26-19.56 = 20.15 persp) 1 PAL ,h et done elles rentrent en production a::m: e'est-a-dire elles .h and so they go into production at m: that is to say they 2 son- elles mettent bas, .h (.) a quatorze mois, e'est-aare- they give birth, .h (.) at fourteen months, that is to 3 *dire ou trei- oui qua*torze mois, e'est-a-dire deux mois apres say or thirt - yes forteen months, that is to say two months after Iau * *ppp-> 4 les brebis, .hJ_ et done, (.) pour arriver a. cette performance the ewes, .h and so, (.) in order to secure this performance lppp-> 5 e'est-a-dire etre aptes a m i - * a etre mises a la reproduction that is to say be able to re-to be put to reproduction lau —>*suspends pointing—> 6 . h a leur premier automne, .hh euh elles sont a un regime assez soigne,* ,h for their first autumn, .hh ehm they are given a careful diet lau —->* 7 *et elles sont notamment *hebergees la. and they are specifically housed there. lau * *ppp » 8 LAU c'est 9 a quLL est r'presente ici? and that's what is represented here? pal — ->1„„ 9 PAL voila. there it is.

Pierre-Alain is engaged in a long multi-unit turn about sheep reproduction. Laura begins to point at line 3, i.e. in a position where Pierre-Alain is reformula ting the age at which sheep can be reproduced. Pierre-Alain's syntactical con struction projects the age from line 1 on (with the preposition « a ») but before telling this information he inserts a reformulation of the process (introduced by « e'est-a-dire » 1); another reformulation, concerning the age, is inserted im mediately after its mention (2). In this way, although various insertions and expansions are provided by Pierre-Alain, the syntactical construction of his turn is projected very early. Laura orients to this feature: she ends the preparation phase and begins to point properly during Pierre-Alain's reformulation of the sheep age (3). But at the end of this TCU, i.e. at a possible transition-relevance place, the current speaker initiates a new TCU (« .h et done, » 4) and begins to point on the map. He projects a longer chunk to come, both by his pointing and his syntactical construction, which initiates a dependent clause (« pour arriver a cette production » 4) projecting a main clause, then delaying it by the insertion of a reformulation of the former (« e'est-a-dire etre aptes » 5).

138

DISCOURSE STUDIES

Laura exhibits her responsiveness to these delayed projections b y suspending her pointing (1. 5); she also exhibits her anticipation of the completion at the end of the main clause b y pointing again (1. 7) when Pierre-Alain adds a second paratactic clause. Thus, Laura's pointing gesture exhibits her syntactic online analysis of Pierre-Alain's turn, done for the practical purpose of looking for possible transition-relevance places. The suspension of her gesture shows her recognition not just of projection spaces but also of their spans: when the pro jected point is delayed, the gesture is relaxed and suspended; when the projected point comes nearer, the gesture is reactivated. This demonstrably shows that Laura is orienting to the relevance and legitimacy, in certain sequential positions, of her claims of speakership, as well as to the fact that, at other moments, these claims cannot be accountably sustained and have to be withdrawn or abandoned. The very fact that she suspends her gesture but leaves her hands ready for ser vice at short distance, shows her readiness to pursue again the same gesture, projecting a future moment where this will be relevant again. In this context, gesture, incremental syntax and rights and obhgations related to speakership are deeply interwoven and made mutually visible, enacted through the unfolding of gestures in time. Pre-initial Turn Pointings, Overlaps and Concurrent Claims of Speakership Pre-initial pointings can be found in the very same sequential environments where next speakers/turn claimants produce serial starts of an incipient turn in overlap with current speaker/turn occupant (Jefferson, 2004: 46): (6) (e7/agl-22'50) 1 PAL

2 3 4 5

devient une unite annexe, (.) qui: sert, (.) uniquement, (.) et en becomes an annex unit, (.) which is used, (.) exclusively, (.) and »ppp-> continu, (.) a la generation de renouvellement continuously, (.) for the new generation (of sheep). *(0.5) lau *....-> PAL [elles y restent [jusqu'a [*toussaint au moment de la: lutte. [they (the sheep) stay [until [All Saints at the moment of the fight. LAU [#et [#et [±#*c[and [and, [ thpal —>1,„„ >*ppp w p e n - »

fig #fig. 1 #fig.2 #fig.3 6 LAU ca, # la. (.) y a une difference entre ca et ca?= this, there. (.) is there a difference between this and this?= fig #fig.4

'

MONDADA

Figure 3


139

Figure 4

At the end of line 2, Pierre-Alain has reached a completion point a n d has e n d e d up his turn. T h e g a p that follows (3) is interpreted b y Pierre-Alain as a n oppor tunity to talk not taken b y his co-participants: h e continues to speak (4), b u t at the same m o m e n t Laura begins h e r serial attempts to begin h e r turn. Laura sees the gap as a n opportunity for self-selecting: this is exhibited b y h e r h a n d approaching the m a p before to point (3). This projection of a n imminent selfselection is materialized verbally a few m o m e n t later, within the serial start (« e t » « e t » « c- » 5) overlapping Pierre-Alain. At the third try, Laura points with h e r pen, strengthening h e r position as incipient speaker. Gestures - as fragments of incipient turns or as in-breaths - are multimodal resources which m a k e pre-begmnings visible-and-audible. T h e s e resources are particularly interesting in a multi-party participation framework, w h e r e there is a pressure for current non-speakers w h o might self-select, to do it as soon as pos sible and at least at the first possible transition point (Sacks, Schegloff, Jefferson, 1974,712-713). Pre-beginnings are a technique for securing precise coordination of prior turn ending with n e x t turn beginning, as well as for securing self-selection lor would-be n e x t speakers. W h e r e a s verbal a n d other acoustic resources are vulnerable to overlaps in these early starts, pointing gestures are n o t a n d can b e produced simultaneously with the terminal segment of the ongoing turn. This allows at the same time the opportunity to achieve an early self-selection a n d to display an orientation to t h e minimization of g a p a n d o v e r l a p , i.e. to the

DISCOURSE STUDIES

140

preservation of the « o n e party at a time » normative principle (see Schegloff, 1987 for a similar analysis of turn-beginning recycles). Nevertheless, it can b e noticed that some of these gestural pre-beginnings are dealt with as having an « i n t e r r u p t i v e » potential or effect. So, even if gestures can b e p r o d u c e d simul taneously with talk without overlapping it, pointing gestures as practices for claiming speakership and for imposing self-selection are oriented to as exhibiting concurrent practices of turn taking (cf. Jefferson w h o notes about overlapping laughter t h a t « recipient/prior speaker a p p e a r to b e attending to the ' p r e m a t u r e ' a n d turn-incursive character of their talk » 1984:30). This m a y b e consequential for the ongoing turn production b y current speaker, as w e will see below. This situation is particularly clear in environments w h e r e concurrent point ings b y several persons occur. (7) (e9/ ag 1-47.52) 1

VIV

pal 2 3

4

lau fig PAL

viv lau fig

(0.4)

5

6

+puisque: ici on on est sur du d u paca:ge, Xdonc since here we we are on (sort of) park, so +ppp--> JLhead forth-> c'est c'est des bonnes euh c'e- s:- c'est des surfaces there are there are good ehm th- s- there are good areas a *bon a b o n potentiel,# done *euh= with a good a good potential, so ehm= * *ppp-> #fig.l = l * e t c'est* l c l o s # and it's closed ->± lppp-> PPPP > ->*„„,suspends her pointing—> #fig.2 #fig.3

VIV

et *c'est clos, et [* (est-ce)# l p a [ s be+soin d e : and it's closed, and (there is) no need to; 7 LAU [*oui mais #(est-[ce que)+ yes but ((interrogative)) i ppp >+ pal -ppp->l„„„„ lau ->* *ppp » fig #fig.4 8 PAL [(a part) [(aside from) 9 LAU iz y: i vont p a s i vont pas dans l'champ apres?= they they don't go they don't go in the field afterwards?= -ppp . » 10 PAL = n o n . =no. v

v

MONDADA


Figure 3

Ml

Figure 4

In this excerpt, we notice that two recipients attempt to self-select while Viviane is still speaking. If we were to restrict our analysis to verbal conduct only, w e would only see that Pierre-Alain's self-selection does not overlap with Viviane (4), a n d is pre sented as a collaborative completion of Viviane's turn, ratified b y her. But, if we consider co-participants' gestures, we get another picture: we can notice that Laura comes in with h e r h a n d as early as line 3 (fig. 1), pointing before « e u h » which concludes Viviane's turn (3). This projects a possible self-selection b y Laura. However, after Viviane, it is Pierre-Alain w h o speaks first, beginning to point in turn-initial position (fig. 2). Pierre-Alain has b e n t his h e a d over the table m u c h earlier (line 1), at the end of Viviane's first T C U a n d before she continues with her first« d o n e ». Laura withdraws her pointing during Pierre-Alain's comple tion (fig. 3) - recognizing it as concurrent with, a n d a challenge to, h e r o w n claim of speakership - b u t prepares to point again as soon as Viviane repeats Pierre-Alain's collaborative item. L a u r a then fully points w h e n she self-selects in overlap with Viviane (7). Thus, here, three participants are pointing at the same m o m e n t (fig. 4 ) : Viviane points during h e r turn, as current speaker; Pierre-Alain points w h e n h e engages in his turn, whereas Laura begins to point earlier, projecting and making visible h e r claim of incipient speakership. We can observe that pointing as a visible and public action projection can b e initiated b y various co-participants modifying their participation status as turns-at-talk unfold; they can manifest concurrent a n d mutually exclusive projections a n d claims of speakership.

142

DISCOURSE STUDIES

As such, they axe monitored by others as persisting or withdrawing claims, as related to yet-relevant or no-more-relevant contributions to the ongoing conver sation. In this sense, publicly displayed pointings as practices for projecting im minent speakership are not only seen but oriented to and exploited b y others, who can consequendy adjust and modify their conducts. Thus, pointings re flexively shape the ongoing incremental productions b y other participants, who mutually respond and adjust to them. In the following excerpts, we will see that mutual and visible availability and recognition of participants' online interpretation of turns at talk is consequential for the emergent organization of their embodied conduct. (8) (e8/agl-23.50) 1 PAL

-Lau cadastre il est carrement inscrit a leur nom. on the land-survey map it ((=the field)) clearly belongs to them. Xppp—> 2 *(0.S) lau *....-> 3 LAU [oui. oui oui, * (.) oui oui] [yes. yes yes (.) yes yes ] 4 PAL [done euh la se] *curite d e : l du foncier, .h est n o u l y e l l e , _L [so ehm the se]curity of of the property, .h is new, — » l o o k s at LAU1 looks at maps ±at LAU1 lau —>*ppp—> 5 l . h et * et p- certainement explique .h que on on defrirche, on .h and and p- surely explains .h that one one clears the land, one Hooks at maps >1.11 lau *„„suspends her pen at close distance—> 6 cloture, on amenage, et done on .h [on LREnd fonctionnelle encloses, one settles, and so one .h [one makes functional ppp—> -Lgesticulates-> 7 LAU [hum 8 PAL cette sous-unite qui assure done,* .h le renouvellement* du this sub-unit which secures then .h the regeneration of lau ->* *ppp—> 9 important, ->_L -LpPP - distance w p e n — i . 10 parce que *pour pour * donner un chiffre, .hh euh: on d- on because in order in order to give an estimation, .hh ehm one h- one lau -->*,,„„„» *chews pen >1. 15 11 doit (1.1) on garde une agnelle pour: ±cinq brebis euh auhas (1.1) one keeps one lamb for five ewes ehm at —-±>looks at LAU—> 12 au minimum quoi.i. at least ((particle)). -->! 13 1(0.5) ™l Hooks at VIV - > ar

MONDADA

MULTIMODAL RESOURCES F O R TURN-TAKING

143

14 PAL *.h d o n c l [eu*h .h so [ehm —XL

lau * * com ((VIV doesn't look at PAL)) 15 LAU [euh* non mais c'est p- c'est [ehm no but that's r- that's

16

—>*ppp--> plutot par rapport a: (0.5) tsk enfin a la legende. rather with regard to (0.5) tsk well to t h e caption.

At the e n d of line 1, Pierre-Alain produces a first possible turn completion, fol lowed b y a silence (2). D u r i n g this gap, L a u r a comes in pointing with h e r p e n and produces a n a g r e e m e n t (3) overlapping with Pierre-Alain w h o expands his turn (4) (cf. excerpt (6), lines 1-5). This T C U (4) begins with « d o n e » a n d can b e interpreted as the conclusion of the previous one: this seems to b e Laura's version, who holds h e r pointing. But Pierre-Alain does not just add some short conclusion: he projects a continuation (« et e t » 5). Consequently, Laura, at the very beginning of this n e w T C U , suspends h e r gesture, although n o t returning to h e r h o m e position b u t leaving h e r h a n d in place. Later, L a u r a points again in a peculiar sequential position within Pierre-Alain's turn: line 8, she anticipates the e n d of the relative clause as a possible turn completion. Line 9, Pierre-Alain deals with it as completed (namely b y lowering his voice) b u t again adds a n e w T C U , initiated with a n in-breath a n d louder voice. This initiation projects a longer argument (« alors c'est important, parce que... » 9-10), a n d is r e s p o n d e d to b y Laura suspending again her pointing a n d beginning to chew h e r p e n - an action which n o t only visibly withdraws h e r gesture b u t also immobilizes h e r h a n d and occupies h e r m o u t h . Pierre-Alain completes his long argument after a l o n g pause (11), which is n o t dealt with b y Laura as a transition-relevance point. O n this completion, h e successively looks at Laura and then at Viviane: this technique for other-selection is vulnerable to t h e absence of gazes of t h e selected coparticipants (Lerner, 2003). A s a matter of fact, Viviane does n o t look at h i m and Laura comes back gesturally only o n the terminal particle « quoi », beginning to speak as Pierre-Alain, after a gap w h e r e n o b o d y speaks, tries to continue. Thus, in this excerpt w e see various attempts b y Laura, i m p l e m e n t e d i n a n embodied way, to initiate turns. T h e s e attempts m a k e visible for the analyst, o n the basis of Laura's online analysis, t h e identification of turn-relevance points within Pierre-Alain's t u r n a n d h e r projection of possible self-selections. Gesture make visible a n d recognizable Laura's projections for other participants as well namely for Pierre-Alain. Their public character h a s a reflexive effect o n turn formatting: for instance, Pierre-Alain organises his turn i n a way that minimizes transition-relevance points in order t o continue to speak. This is n a m e l y t h e case of the transition-relevance point of line 9: Pierre-Alain accelerates his speech tempo at the very beginning of the T C U ; m o r e o v e r , at that m o m e n t , h e does not glance at other participants b u t looks at the m a p s in front of h i m . This min imization is even clearer after several Laura's serial pointings. By contrast, h e looks at L a u r a at t h e beginning of the excerpt, a n d at the e n d of his long multiunit turn (line 11-12), in pre-completion nositions. onenino- n n riio ^ r i ^ r - n ^ u .

144

I

DISCOURSE STUDIES

Thus, pointing gestures as practices for projecting incipient speakership re flexively configure the other's conduct, w h o progressively integrates within his turn-format the emergent relevancies publicly manifested b y gestures. This reflex ive adjustment is responsible for the flexible, incremental, collective achievement of turn's units (cf. Goodwin, 1979,1981). T h e public display of projections is par ticularly effective in cases of concurrent self-selections. (9) (e2/ag2-16.18) 1 PAL

la la pour moi y- pour moi y avait de la surface (.) et done there there for me th- for me there was some surface (.) and so 2 je disais dens, (.) le quartier. e'est un quartier, i said look, (.) the block, it's a block, 3 BRU ((coughts)) 4 PAL (.) qui%: * e'est un %J_k- qui* designe un ensem°/oble % (.) which: that's a k- which defines a set Lcircular hand gesture-> bru %pulls document°/o °/otakes°/o towards him his pen lau * *ppp—> 5 %de surfaceslutilise dans cer°/ota[ins (.) domaines, of areas used in some (.) domains -->! bru °/o "/otouches and reads his docs-> 6 LAU [ +comes back w pen—t-ppp 1bru —>%pulls H down°/oputs H on p a p e r - > 8 -HdTier+be pour faire manger les: of grass to feed the 9 BRU d'accord. % (0.2) % y a beaucoup plus d'informations okay. (0.2) there is a lot more information ((on this mapj) —->% %ppp—» A s the excerpt begins, Viviane is producing an explanation and points to Bm • paper. Bruno displays h e is listening to her, looking at her, his hand on m o u t h . After Bruno's continuer, line 6, Viviane continues to speak but pre;, sively withdraws her p e n , pointing again only in a limited way, when sht repaired a p r o b l e m of expression (7). So, she stops to point in a continuous from line 7 on, in a position where « qui e n plus est»(7) initiates a last expanclosing her turn. Interestingly, this projection of turn's closure is nor unilaiL b u t comes just after Bruno has pulled his h a n d - which was held on his moutti until then - d o w n o n the table, thus m a k i n g free his m o u t h for speaking and hn h a n d for pointing. Actually Bruno puts his h a n d o n his document and sh«r points at the beginning of his turn. T h e end of Viviane's turn is thus an inters achievement, p r o d u c e d through the alignment of her projecting the cor completion of h e r turn a n d h e r co-participant projecting his upcoming catt. of « possible n e x t speaker ».

Pointing Gestures Persisting after Turn Completion Nevertheless, not all pointing gestures end with the turn they are affiliated! i n =. nrprnmnletion Dosition. Some gestures are held after turn completion*

MONDADA

M U L T I M O D A L R E S O U R C E S FOR TURN-TAKING

147

post-completion persistence seems to contradict the idea that gestures are a speaker' - characteristic a n d thus delimit the rights and obligations of speakership. Detailed analysis shows that this is not the case, but that speakership is p r o b ably a phenomenon that is not stricdy limited within turn (verbal) boundaries, but that is interactively shaped b y c o m p l e m e n t a r y participants' relevancies a n d orientations Here is a first occurrence of persisting pointing: 1

(II) (e9/ag I -47.00) (= at the end of exc. 1)

v

1 VTV

3

4 5 6

+.h oui. parce que: i'm'sem+ble: eh i- ici c'eteit s::- ce .h yes. because it seems to me eh he- here it was wh- what ~ +ppp-> 1.6 qui: ce que ca voulait representer, (* c'etait this wanted to represent, [ it was LAU [ * c'est les am*andes ca? [ are these almonds? lau * *PPP - > VTV [.) oh ca c'etait (.) des amandes,* [ c'etait aussi 1'idee que: une • oh these were (.) almonds, [ it was also the idea that: during „>• lau LAU [°ouais° [°yeah° VIV journee,+ (.) les: i- i restaient dans une en- une une < me day, (.) the: th- they stood within a wi- one one ~>+iconic g e s t u r e s — » seu- une seule entite. ca suffisait pour la nourriture <ml- one only unit, this was enough for their food

In this excerpt. Viviane is trying to explain w h a t she sees o n the m a p (1-2). Laura asks a confirmation question in overlap (3), Viviane answers (4) a n d then continues her previous explanation. If we look at the pointing gestures in this excerpt, we can first notice that iaura begins to point at the beginning of h e r question (3) a n d keeps pointing .nti! the end of the answer (4), s t o p p i n g just before h e r a c k n o w l e d g m e n t • ouais »;">,. As far as Viviane is concerned, we can notice that she keeps pointing »hen the question is asked, w h e n she gives a response a n d then continues h e r xplanation The continuity b e t w e e n fragments of h e r explanation (« ce que ca oulait representer. c'etait » 2), suspended b y the question/answer adjacency pair, is secured through the recycling of « c'etait» after the second pair part (4). Similarly, ihi- fact that she maintains h e r gesture during the adjacency pair p r o daces an online practical categorization of this sequence as being a « punctual insertion * ur a « m o m e n t a r y suspension ». By holding their pointing gestures, b o t h participants are p r o d u c i n g the ifflelligibtlity and accountability n o t only of the current turn organization b u t more generally of the current sequence organization. T h e y are displaying their onentation to the way in which sequence is structured, to the sequential implicarjveness oi their turn and to the span of the ongoing sequential actions. Pointings ihow that Laura is orienting to the adjacency pair as the sequential unit relevant

DISCOURSE STUDIES

148

for her action; whereas Viviane is orienting to the continuity of her explanatory action b e y o n d the insertion of the pair. T h e sequential scope considered b y participants can b e variable, as shown b y the following excerpt: (12) (e5/agl-21.04) 1 PAL 2 3 LAU 4 PAL 5 LAU

6 PAL lau 7 8 LAU 9 PAL 10 LAU 11 PAL lau 12 LAU

ca veut dire qu'on se deplace, (.) pour rentrer et this means that one has to travel, (.) in order to go back and sortir les agnelles, (.) tou[s les jours, (.) mais to take out the lambs, (.) everfy day, (.) but [mhm ell[es sont effectivthey are effectivpointing begins fet: * et *elles restent pas sous* l'abri,* FPP 1 (Q) [and and they don't stay under the shelter, *left H comes in-*ppp w finger *»,„•»,* *(0.5) non non voila elles sont abritees la SPP 1 (A) no no well they are protected during the *PPP finger at distance—> nuit, (.) [et elles sont dans le pare le jour. (.) et 9 a night, (.) [and they are in the paddock during the day. (.) and this [oui [yes veut di- elles sont pas en plein air integral mean- they are not outdoor all the time , ii est ou Pparc? FPP2(Q) where is the paddock? il est le* (.) le pare il est autour SPP 2 (A) it is the (.) the paddock is all around >* pointing stops °ahoui° °oh yes° w

Pierre-Alain is describing the temporal organization of the sheep farmer's work: Laura asks a first question (5), with a first start in overlap a n d then a second star in the clear (5), beginning to point just after the former. She is writing note with her right h a n d , a n d this constrains her m o v e m e n t s (she comes back wit' the h a n d holding h e r piece of paper, line 5, a n d continues to point at distana holding the p a p e r from line 6 to line 11). T h u s , her pointing gesture ends rn just after the first adjacency pair is closed, but at the e n d of the second one, it dated b y her with a second question. Both questions pick u p a referential eleme in Pierre-Alain's previous turn a n d operate a « t y i n g » with his turn b y means a p r o n o m i n a l a n a p h o r a (« elles » 5, « i l » 10 within a right dislocation, possit orienting to the fact that another p r o n o m i n a l reference has b e e n used b y Pien Alain after his mention of « l e pare » 7). ttv holding her pointing gesture, the speaker designs the span of h e r actf: initiated b y her.

MONDADA


149

These data suggest that participants, a n d m o r e particularly speakers, organize the public accountability a n d visibility of the scope of their turns, covering not only the length of their turns at talk, b u t the sequence they initiate - typically, an adjacency pair - or m o r e subsequent sequences. I n this way, speakers' gestures suggest that rights a n d obligations characterizing speakership d o n o t stop at turn boundaries but continue within the normative expectancies related to the sequence initiated b y that turn. I n this sense, the speaker remains « responsible » of the sequence she has initiated a n d she continues to control, n a m e l y b y dis playing if it has b e e n completed or answered in an adequate way. Pointing Gestures Persisting over Extended Sequences This practical delimitation of the scope of the speaker's rights a n d obligations across the sequence can e x t e n d b e y o n d adjacency pairs. This shows that par ticipants d o orient to the complex sequential organization of talk-and-embodiedconduct-in-interaction, taking into detailed consideration b o t h the finest a n d the most complex aspects of this organization. 5

(13) (e8/agl-24.17) (end of exc. 8) 1 PAL com 2 PAL

lau 3 LAU

4 5

VIV 6 PAL

7 8 9 LAU 10 PAL 11

1(0.5) Hooks at VIV-> ((VIV is not looking at her co-participants)) *.h d o n c l [eu*h .h so ehm ->1 * * [euh* non mais c'est p- c'est plutot par rapport [ehm no but that's r- that's rather in relation *ppp -> a: (0.5) tsk enfin a la legende. (0.2) mais c'est p't-et with (0.5) tsk well with the caption. (0.2) but it's maybe Vivian*ne qui: (1.+2) *+euh[:: Viviane who: ehm[:: ->*pushes paper tow. VIV*ppp—>1.20 -fshrinks H back+ [°nJ.on° mais la des JLterres [no but there some farmland 1 ±ppp->1.17 assolees, with crop rotation, (0.5) ouais yeah done euh on y fait une recolte ou on le reserve la pature so ehm either one does a harvest or one allots the pasture en particulier au mois d'juille:t, [les repousses on va particularly in the month of July, the new growth will be

DISCOURSE STUDIES

150

12 LAU 13 PAL 14 LAU 15

PAL

16 VIV

17

PAL

18 LAU 19 PAL 20 LAU

[ouais [yeah eventuellement les d o n n e r aux [agnelles, [(.) .h et et le reste eventually given to the lambs .h and the remaining [et [etla, [and [and there, ca c'est du du parcours a m e n a g e . qui a ete euh [: this is a an established path, which has been ehm[: [amelioreA [improved com ((VIV does any gesture nor movement)) amellio[re± improv[ed ^L,„,„-l[done laboure et: reseme? [°ou non° [so cropped and: sowed again? "or not" [dans l'temps oui. (.) oui oui [in the old times yes. (.) yes yes ah oui* d'a[ccord* oh yes oka [y •>* *

Line 3, L a u r a asks a question about the caption of the m a p , overlapping PierreAlain w h o is expanding a previous long multi-unit turn (see exc. 8). Laura mentions Viviane as b e i n g the relevant speaker for that topic, but she uses the third person, referring to her m o r e than addressing her. L a u r a does not glance at Viviane, although she pushes a relevant document towards her, while creating a long g a p (5), giving h e r an opportunity to respond. O n her side. Viviane is not glancing at h e r co-participants, a n d the only m o v e m e n t she does is slightly to pull b a c k h e r h a n d during t h e long gap, displaying h e r unavailability for talk. I n this case, w e notice that pointing, pushing a d o c u m e n t towards some body, retracting a h a n d even minimally are movements that can radically reshape participation frameworks and mutual engagement displays. I n this context, Pierre-Alain answers to the question formulated b y Laura (6-7). D u r i n g t h e p r o d u c t i o n of this second pair part, Pierre-Alain points, displaying his speakerhip. But Laura holds h e r pointing too. This continues to b e the case through the except, even line 14, w h e n L a u r a reorients Pierre-Alain's explanation with a c o m p l e m e n t a r y question. Interestingly, w h e n Pierre-Alain p r o d u c e s a word-search, line 15 (« qui a ete euh:»), Viviane prompfly proposes a candidate (« ameliore » 16) which is repeated b y the h i m (18). During this short collaborative completion of Pierre-Alain's turn, Viviane m a k e s n o m o v e m e n t or gesture: she displays that she is speaking for her co-participant - animating him - but she is not assuming properly a speaker ship position - of author. W h e n ratifying the item closing the w o r d search, Pierre-Alain withdraws his pointing (17), projecting his turn completion. H e holds this position, even if Laura asks a last confirmatory question (18), which gets a short answer (19).

MONDADA


151

Laura withdraws only after she closes the w h o l e stretch of interaction (« ah oui d'accord « 20). Closure with an agreement and a change-of-state-token (« ah » in French, corresponding to the « o h » studied b y Heritage, 1984) displaying the completion of the sequence from b o t h an interactional a n d an informational perspective - , together with the withdrawal of a gesture held for a long period, exhibits the boundaries of a n extended segment of talk as prac tically and incrementally defined b y the pointing participant w h o initiated the sequence a n d projected a space for its completion.

Conclusions The cases we studied in this paper outline a systematic use of pointing as a m e t h o d for projecting self-selection, designing a transition space, a n d m o r e generally denning speakership within a range of sequential positions. Pointing as a practice for turn taking can b e initiated either before the actual completion of current speaker's turn, (thereby projecting its end), or at the begin ning of the incipient speaker's turn. I n both cases, pointing displays a participation shift, the pointer initiating, often before even saying a word, h e r transition from the category of « n o n current speaker » to the category of «incipient speaker », through the category of « possible next speaker »: in this sense, pointing gestures manifest the temporal, situated, e m b o d i e d emergent process of the establishment of speakership. Moreover, this process shows publicly a n d visibly the way in which a « recipient» scrutinizes, for all practical purposes, an ongoing turn a n d produces an online parsing of this turn in T C U s . I n this way, n o t only a local understanding of the ongoing T C U s is exhibited, but also its e m b o d i e d use oriented to the turn-taking machinery. I n turn, this public manifestation of an online analysis is visible for other participants, w h o reflexively adjust their own turn design a n d conduct to w h a t is m a d e recognizable b y the pointer - e.g. in concurrential turn at talk. Pointing can end either in a pre-completion position or at the end of the turn. But it can also e n d m u c h later in the sequence: at the end of one or m o r e adjacency pairs initiated b y the pointing speaker, as well as at the e n d of an ex tended sequence controlled b y her. T h e latter case shows a n extension of the local display of speakership: speakers hold their posture across turns, manifesting an extension of their rights a n d obligations b e y o n d the turn a n d b e y o n d their talk. In this way, pointing across sequences shows the situated online sequential interpretation and production b y participants, displaying the span of their actions and of their sequential implicativeness. Thus, pointing as it has b e e n analyzed in this particular corpus highlights very general features of turn-taking a n d sequential organization. At the same time, it is interesting to notice that this methodical a n d systematic practice exploits the context sensitive specificities of the ecology of action. Pointing as a practice for projecting turns a n d organizing sequences is n o t available in every setting: in the case studied, pointing gesture is a resource m a d e available b y the peculiar spatial and material environment of the interaction, constituted b y a table covered

DISCOURSE STUDIES

152

with documents, maps and other visualizations, massively used within an activity focused on reading, writing and correcting cartographic representations. In this particular setting, interaction is not primarily organized as a face-to-face exchange of talks and glances but rather as a side-by-side c o m m o n focalization of attention on objects. In this sense, pointing gesture have a prominence and a salience that is locally achieved through the very activity participants are involved in. Moreover, pointing gestures as resources for interactional practices are made available by the high focus on referential and deictic activities going on in this praxeological setting. The referential dimension of pointing works here as a resource at the service of the interactional organization: deicticals at turnbeginnings are a resource which can enhance the accountability and legitimacy of pointing gestures for taking the turn. In this sense, both represent not exclusive but rather complementary uses of the same resource. The efficiency of singular pointing gestures is reflexively shaped b y their embededdness in this peculiar activity: pointing with a pen or pointing with a finger not only have different accuracies, but offer different opportunities for their integration within activities. For example, pointing with a pen excludes other uses of the pen, such as taking notes; pointing with the left hand allows one to write with the right hand. Chewing the pen versus placing the hand holding the pen close to the centre of the table constitute two home positions which enable different preparation phases and rely on different kinds of reactivity and tempor ality for pointing. In this sense, pointing involves an arrangement of the material environment as well as a disposition of bodies, all eventually preparing and projecting an upcoming-speaker position. Therefore pointing is here the indexical resource at hand, made available by a specific ecology of action. In similar ecologies, the same resource can be mobilized. In other settings, pointing may be used in the same way, but it is also possible that other multimodal resources, best suited to the specificities of the local context and material environment, may be used instead, although never theless respecting the same sequential and interactional constraints. In summary, then, this paper aims to be a contribution to the description of both systematic and indexical features of multimodal turn and sequence organization.

Transcript C o n v e n t i o n s Data have b e e n transcribed according to conventions d e v e l o p e d b y Gail Jefferson. A indicative translation is provided line per line. Multimodal details have b e e n transcribed according to t h e following conventions:

* * ++

delimitate Laura's gestures and actions descriptions, delimitate Viviane's gestures and actions descriptions, delimitate Pierre-Alain's gestures and actions descriptions.

MONDADA


153

Glances are described by the same conventions, but in a light-coloured character. >

* ~-*— » —>* » ~ — „„, lau fig * com ppp w H

gesture or action described continue across subsequent lines. gesture or action described continue until and after excerpt's end. gesture or action described continue until the same symbol is reached. gesture or action described begins before the excerpt's beginning. gesture's preparation. gesture's apex is reached and maintained. gesture's retraction. participant doing gesture is identifyied when (s)he is not the speaker, the exact point where screen shot (figures) has been taken is indicated, with a specific sign showing its position within turn at talk, commentary pointing gesture with (pen/finger) hand

Notes 1. This paper has been presented and discussed at The Nijmegen Gesture Centre Lecture Series on April 8, 2005, and at the 9" Colloque de Pragmatique de Geneve on May 26-28,2004 and has benefited from comments of the participants. I am most grateful to Charles Antaki for revising my English text. 2. Thus, this gesture done by a current non-speaker seems only apparently contradicting the idea that « hand gesturing is largely, if not entirely, a speaker's phenomenon > (Schegloff, 1984, 271): Schegloff notes that an exception, which is itself in keeping with this general principle, is provided by the fact that « current non-speakers who initiate a hand gesture may show themselves thereby to be intending, and incipient, speakers, and the gestures may thus be used as a way of making a move for a turn at talk next» (1984, 271). It is precisely this exception that is explored in this paper. 3. « The notion of a 'projection space' is concerned with both the span in which some element of talk is 'in play' before being produced, and with the evidence of that which a speaker's turn may make available to its recipient. For example, it appears central to the organization of tightly coordinated turn-transfer from one speaker to a next that aspects of some current turn are projected, and are available to analysis by a recipient/potential-next-speaker before their actual occurrence > (Schegloff, 1984, 267). 4. The continuous transformation and negotiation of the participant's categories have been pointed out by Goodwin & Goodwin (2004) revisiting and criticizing the goffmanian notion of « participation ». Narratives in interaction have been studied from the perspective of the progressive establishment of « teller » as well as « cotellers » (Goodwin, 1984, 1987; Hayashi, Mori & Takagi 2002). But more generally this movement is observable within every next turn. It is interesting to look at the categories used for dealing with « not yet speakers » within the current literature: in her analysis of systematic positionings of overlaps, Jefferson (1984) speaks of « recipient/next speaker », as well as, in concurrent overlaps, of Turn Occupant vs Turn Claimant, showing that these two categories are not always easy to differentiate (2004, 49). When analyzing pre-beginnings, Schegloff (1996, 97) uses the category

DISCOURSE STUDIES

154

« current-recipient-of some talk/potential-next-speaker » showing through this com plex label the transitional dynamic character of the category. 5. This is consistent with the analyses provided by Fox (1987) on long distance anaphoras, manifesting the way in which speakers deal for all practical purposes with long stretches of talk as being homogeneous and continuous from a praxeological and a topical point of view.

Bibliography Auer, Peter. (1996). On the prosody and syntax of turn-taking. In: Elizabeth CouperKuhlen and Margret Selting (eds.), Prosody and conversation, 57-100. Cambridge: Cambridge University Press. Auer, Peter. (2005). Projection in interaction and projection in grammar. 7a:/25.1: 7-36. Couper-Kuhlen, B. (1996). Intonation and clause combining in discourse: The case of because. Pragmatics, 6/3, 389-426. Couper-Kuhlen, E. (2001). Interactional prosody: High onsets in reason-for-the-call turns. Language in Society, 30,29-53. Couper-Kuhlen, E. (2004). Prosody and sequence organization: The case of new begin nings. In: Elizabeth Couper-Kuhlen & Cecilia E. Ford (eds.). Sound Patterns in Interaction. Amsterdam: Benjamins, 335-376. Drew, P. & Heritage, J. (eds.) (1992). Talk at Work. Cambridge: Cambridge University Press. Ford, C.E. (2005). Contingency and units in interaction. Discourse Studies, 6.1: 27-52. Ford, C.E. & S.A. Thompson. (1996). Interactional units in conversation: syntactic, intonational, and pragmatic resources for the projection of turn completion. In: E. Ochs, E.A. Schegloff, & S.A. Thompson (eds.), Interaction and grammar, 135-184. Cambridge: Cambridge University Press. Ford, C.E., Fox, B.A., & S.A. Thompson. (1996). Practices in the Construction of Turns: the 'TCU' revisited, Pragmatics 6.3, 427-454. Ford, C.E., Fox, B.A., & S.A. Thompson (eds.) (2002). The Language of Turn and Sequence. Oxford: Oxford University Press. Fox, B.A. (1987). Anaphora and the structure ofdiscourse. Cambridge: Cambridge University Press. French, P. &John Local. (1983). Turn competitive incomings. Journal of Pragmatics, 7, 701-715. Garfinkel, H., & Wieder, D.L. (1992). Two incommensurable, asymmetrically alternate technologies of social analysis. In: G. Watson, & R.M. Seiler (eds.). Text in context: studies in ethnomethodology (pp. 175-206). Newbury Park: Sage. Goodwin, C. & Goodwin, M.H. (1987). Concurrent Operations on Talk: Notes on the Interactive Organization of Assessments. Pragmatics 1(1): 1-55. Goodwin, C. & Goodwin, M.H. (2004). Participation. In: A. Duranti (ed.). A Companion to Linguistic Anthropology. Oxford: Blackwell. Goodwin, C , Goodwin, M.H. (1992). Context, Activity and Participation. In: P. Auer. A. di Luzio (eds.). The Contextualization of Language. Amsterdam: John Benjamins. 77-99. Goodwin, C. (1979). The Interactive Construction of a Sentence in Natural Conversation. In: G. Psathas (ed.). Everyday Language: Studies in Ethnomethodology, New York: Irvington, 97-121.

MONDADA


155

Goodwin, C. (1981). Conversational Organization: Interaction Between Speakers and Hearers. New York: Academic Press. Goodwin, C. (1984). Notes on Story Structure and the Organization of Participation. In: M. Atkinson & J. Heritage (eds.). Structures ofSocial Action. Cambridge: Cambridge University Press, 225-46. Goodwin, C. (1987). Forgetfulness as an Interactive Resource. Social Psychology Quarterly, 50(2): 115-31. Goodwin, C. (2000). Action and Embodiment Within Situated Human Interaction./oumaZ of Pragmatics, 32: 1489-522. Goodwin, C. (2003). Pointing as Situated Practice. In: Kita, S. (ed.). Pointing: Where Lan guage, Culture and Cognition Meet. Mahwah, NJ: Lawrence Erlbaum, 217-41. Hakulinen, A. & Selting, M. (eds.) (2005). Syntax and lexis in conversation. Amsterdam: Benjamins. Hanks, W. F. (1992). The indexical ground of deictic reference. In: A. Duranti & C. Goodwin (eds.). Rethinking Context: Language as an interactive phenomenon. Cambridge: University Press. Haviland, J. (1996). Projections, transpositions, and relativity. In: J. J. Gumperz & S.C. Levinson (eds.). Rethinking Linguistic Relativity. Cambridge: Cambridge University Press. Hayashi, M. (2005). Referential problems and turn construction: An exploration of an intersection between grammar and interaction. Text, 25/4, 437-468. Hayashi, M., Mori, J., Takagi, T. (2002). Contingent achievement of co-tellership in a Japanese conversation: An analysis of talk, gaze and gesture. In: C. Ford, B. Fox, S. Thompson (eds.). The Languge of Turn and Sequence. Oxford: Oxford University Press. Heath, C. & Luff P. (1992). Media space and communicative asymmetries: Preliminary observations of video mediated interaction. Human Computer Interaction, 7:315-346. Heath, C , Svensson, M., Hindmarsh, J., Luff, P., Vom Lehn, D. (2002). Configuring Awareness. Computer Supported Cooperative Work, 11: 317-347 Heritage, J. (1984). A Change of State Token and Aspects of Its Sequential Placement. In: J.M. Atkinson &J.C. Heritage (eds). Structures of Social Action. Cambridge: Cambridge University Press, 299-345. Hindmarsh, J., Heath, C. (2000). Embodied reference: A study of deixis in workplace interaction. Journal ofPragmatics, 32, 12, 1855-1878. Jefferson, G. (1984). Notes on some orderlinesses of overlap onset. In: V. d'Urso and P. Leonardi (eds.). Discourse analysis and natural rhetoric. Padua: Cleup, 11-38. Jefferson, G. (2004). A sketch of some orderly aspects of overlap in natural conversation (1975). In: G.H. Lerner (Ed.) Conversation Analysis: Studies from the first generation. (pp. 43-59). Amsterdam: Benjamins. Kendon, A. (2004). On pointing. In: Gesture. Visible Action as Utterance. Cambridge: Cambridge University Press, 199-224. Kita, S. (ed.) (2003). Pointing: where language, culture, and cognition meet. Mahwah, NJ: Lawrence Erlbaum. Lerner, G.H. (1991). On the syntax of sentences in progress. Language in Society, 20: 441-458. Lerner, G.H. (1996). On the 'semi-permeable' character of grammatical units in con versation: conditional entry into the turn space of another speaker. In: E. Ochs, E.A. Schegloff, & S.A. Thompson (eds.). Interaction andgrammar, 238-271 Cambridge: Cambridge University Press.

156

DISCOURSE STUDIES

Lerner, G.H. (2002). Turn-sharing: the choral co-production of talk-in-interaction. In C. Ford, B. Fox, & S.A. Thompson (eds.). The language ofturn and sequence. 225-256. Oxford: Oxford University Press. Lerner, G.H. (2003). Selecting next speaker: The context-sensitive operation of a contextfree organization. Language in Society, 32(2): 177-201. Local,J. & Kelly,J. (1986). Projection and 'silences': Notes on phonetic and conversational structure. Human Studies, 9:185-204. Local, J., Kelly,J. & Wells, B. (1986). Towards a phonology of conversation: Turn-taking in Tyneside English. Journal ofLinguistics, 22, 411-437. Mondada, L. (2005a). Visions controversies de la carte: construire le visible par les gestes et la parole en interaction. In C. D'Alessandro, F. Charvolin, V. November, E. Remy (eds). Espaces, savoirs et incertitudes. Paris: Belin, 15-31. Mondada, L. (2005b). L'analyse de corpus en linguistique interactionnelle: de 1'etude de cas singuliers a l'etude de collections, In. A. Condamine (ed.), Semantique et corpus, Paris: Hermes, 76-108. Mondada, L. (2005c). La constitution de 1'origo deictique comme travail interactionnel des participants: une approche praxeologique de la spatialite. Intellectica.. No special surEspace, Inter-action & Cognition, 2/3, 41-42, 75-100. Mondada, L. (2006a). Video Recording as the Preservation of Fundamental Features for Analysis, in Knoblauch, H., Raab,J., H.-G. Soeffner, Schnettler, B. (eds.). Video Analysis, Bern: Lang. Mondada, L. (2006b). La pertinenza del dettaglio: registrazione e trascrizione di dati video per la linguistica interazionale. In Y. Biirki, E. de Stefani (eds.). Trascrivere la lingua. Dalla fihlogia all'analisi conversazionale. Bern: Lang, 313-344. Rossano, F. (2005). When it's oveer is it really over? On the effects of sustained gaze vs. gaze withdrawal at sequence possible completion. Paper presented at the International Pragmatic Association, Riva del Garda,July 2005. Sacks, H. (1992). Lectures on Conversation. Oxford: Blackwell, 2 vols. Sacks, H., Schegloff, E. A.,Jefferson, G. (1974). A Simplest Systematics for the Organization of Turn-Taking for Conversation. Language, 50, 4, 696-735. Schegloff, E.A. (1984). On Some Gestures' Relation to Talk. In: J.M. Atkinson & J. Heritage (eds.). Structures of Social Action. Cambridge: Cambridge University Press, 266-298. Schegloff, E.A. (1987). Recycled Turn Beginnings. In: G. Button &J.R.E. Lee (eds.). Talk and Social Organization. Clevedon: Multilingual Matters, 70-85. Schegloff, E.A. (1996). Turn Organization: One Intersection of Grammar and Interaction. In: E. Ochs, E.A. Schegloff & S.A. Thompson (eds.). Interaction and Grammar. Cambridge: Cambridge University Press, 52-133. Schmitt, R. (2005). Zur multimodalen Struktur von turn-taking. Gesprdchsforschung Onlinezeitschrift zur verbalen Interaktion, 6,17-61 (www. gespraechsforschung-ozs.de). Selting, M. & Couper-Kuhlen, E. (eds.) (2001). Studies in interactional linguistics. Amsterdam: Benjamins Selting M. (2000). The construction of units in conversational talk. Language in Society 29: 477-517. Selting, M., Couper-Kuhlen, E. (eds.). (2001). Studies in interactional linguistics. Amsterdam: Benjamins. Streeck, J. (1995). On projection. In: E.N. Goody (ed.). Social intelligence and interaction. 87-110. Cambridge: Cambridge University Press.

MONDADA


157

Streeck, J. & U. Hartge. (1992). Gestures at the transition place. In Peter Auer and Aldo di Luzio, eds., The contextualization oflanguage, 135-157. Amsterdam: John Benjamins. Tanaka, H. (2001). Adverbials for turn projection inJapanese: Toward a demystification of the "telepathic" mode of communication, Language in Society. 30, 559-587. Thompson, S.A., E. Ochs, & E.A. Schegloff (eds.). (1996). Interaction and Grammar. Cambridge: Cambridge University Press.

53 The Political Function of Narrative in Organizations Dennis K. Mumby

U

n d e r the rubric of interpretive theory, a great deal of current research in organizational c o m m u n i c a t i o n has focused o n the sense-making practices of organization m e m b e r s , as manifested in their use of organ izational symbolism. Such research is motivated b y the belief that organization members, as social actors, actively participate in the construction of organizational reality t h r o u g h organizational discourse (Frost et al., 1985; P o n d y et al., 1983; Putnam & Pacanowsky, 1983; Smircich, 1983). This "organizational culture" a p p r o a c h recognizes the emergent, ongoing, a n d sometimes precarious nature of organizational reality, a n d thus represents a significant shift away from the static, m o r e functionalist approaches of traditional organizational theory. M y purpose in this p a p e r is to extend these recent developments b y demon strating that cultural theory, despite its significant contributions to organizational research, articulates a rather limited a n d theoretically naive view of the rela tionship b e t w e e n organizational symbolism and organizational reality. While such research adequately describes organizational sense-making at a surface level, it does little to explicate the deep structure process through which certain organizational realities c o m e to hold sway over competing world-views. This article redresses this imbalance b y focusing o n narrative (as o n e form of organizational symbolism) a n d addressing its role as a legitimating device in formal organizations. Narratives provide m e m b e r s with accounts of the process of organizing. Such accounts potentially legitimate d o m i n a n t forms of organiza tional reality, and lead to discursive closure in the sense of restricting the interpret ations and meanings that can b e attached to organizational activity. Closure frames discourse in a distortive fashion, a n d often misrepresents the interests of particular groups within organizations (Deetz & M u m b y , 1985). Source: Communication Monographs vol. 54,1987, pp. 113-127.

MUMBY

NARRATIVE IN ORGANIZATIONS

159

T h e notion of narrative closure within organizations is u n p a c k e d b y laying out the complex a n d interdependent relationship that exists b e t w e e n narrative, power, a n d ideology in organizations. From this perspective, narratives n o t only evolve as a product of certain p o w e r structures, b u t also function ideologically to produce, maintain, a n d reproduce those p o w e r structures. Because of their embeddedness in the everyday practices of organizations, narratives are n o t easily perceived as legitimating devices - they often articulate a n organizational reality that is accepted as "the natural order of things.'' By adopting a political reading of narrative, however, it is possible to demonstrate that story-telling is not a simple representing of a pre-existing reality, b u t is rather a politically motivated production of a certain w a y of perceiving the world which privileges certain interests over others. T h e fundamental premise for this article is based o n J a m e s o n ' s edict that argues for . . . the priority of the political interpretation of literary texts. It conceives of the political perspective not as some supplementary m e t h o d , not as an optional auxiliary to other interpretive m e t h o d s current today - the psychoanalytic or the mym-critical, the stylistic, the ethical, the structural b u t rather as the absolute horizon of all reading a n d all interpretation. (1981, p . 17) Such a position confronts narrative politically, as a n ideological force that articulates a system of m e a n i n g which privileges certain interests over others. A political reading of narrative draws attention to the relationship between narrative structure a n d the process of interpretation and, as such, focuses o n the process by which d o m i n a n t m e a n i n g systems arise. Narratives "address" readers/listen ers b y creating a "fit" b e t w e e n the narrative a n d the individual's world-asexperienced. Such a fit is b y n o means perfect, however; otherwise, the individual would b e completely subjugated to the narrative's ideologically constructed "reality." As Nakagawa (1983, p . 16) points out, this margin of difference between narrative a n d reader/listener translates into a tension between what is highlighted (present) a n d what is h i d d e n (absent). It is this difference that a political read ing of narrative exploits, seeking to create a disjuncture b e t w e e n the privileged (dominant, ideological) reading a n d that constructed b y the reader/listener. In a very real sense it is this disjuncture that this p a p e r exploits, for it is clear that the issue of legitimation is a double-edged sword. Although it is argued that narratives in organizations can function as an ideological device to legitimate the meaning systems of dominant groups, narratives can also potentially delegitimate dominant m e a n i n g systems. I n this context, p o w e r becomes m o r e than a p h e n o m e n o n that is i m p o s e d o n subordinate groups; it involves a "dialectic of control" (Giddens, 1979, p . 72) in which even the ostensibly powerless can utilize organizational structure to their advantage. This dialectic will b e c o m e clearer in the analysis of narrative in the final section of the paper.

160

T h e b a l a n c e of this p a p e r will therefore a d d r e s s t h e following issue p o w e r a n d its relationship to organizing; ideology and organizational structuring; a n d the relationship between narrative a n d the social construction of organ izational reality. i _

P o w e r i n Organizations T h e issue of p o w e r maintains a s o m e w h a t tenuous status in organizational research. O n the o n e h a n d , organizations themselves h a v e a vested intei est is presenting a n image which emphasizes efficiency, productivity, and a cwicem for the well-being of personnel. Power does not fit very comfortably into this image. I n addition, managerial lore relies heavily o n the "myth of rationality* (Conrad, 1985; Weick, 1979), i.e., the belief that organizations operate n the basis of decisions m a d e through the objective assessment of available informa tion a n d the setting of appropriate, carefully selected goals. To suggest - at least publicly - that decisions in organizations are m a d e o n any other bases power and political expediency, for example) is heretical to m a n y organization managers. O n the other h a n d , organizational researchers h a v e clearly established "power" as an important p h e n o m e n o n in organizations, although it is bard to find a consistent articulation of its nature (Pfeffer, 1981). Much of the early research o n organizational p o w e r is strongly influenced b y Weber's (1947) seminal work. I n particular, his distinction between p o w e r (Macht) and authority (Herrsckaft) has b e e n largely co-opted b y m o r e traditional theorists in a way that legitimates organizational hierarchy a n d domination. Dahrendorf (1959, p 166j clearly articulates this distinction: " T h e important difference between power and authority consists in the fact that whereas p o w e r is essentially tied to the personality of individuals, authority is always associated with social positions« roles." Within organizations, the exercising of authority is viewed as the rational extension of a social actor's legitimate role, i.e., as legitimate power. M o r e recentiy, however, traditional conceptions of organizational power h a v e b e e n challenged in order to d e b u n k the m a n a g e m e n t bias that has charac terized m u c h organizational research. In particular Clegg's (1975; 1981; Cieggl Dunkerley, 1980) work has shown how Weber's concept of power and bure» cracy can b e utilized to provide a m o r e radical interpretation of organizatiod power. Clegg (1975, p . 59) reinterprets Hemchaft&s "rule" to demonstrate "W interpretative work that people engage in, w h e n they m a k e sense of the wotii All authority, h e suggests, is d e p e n d e n t u p o n this conception of rule-govew behavior because rules orient people toward structuring their behavior in ata tain way. Different rule systems therefore provide different structures for orient^ behavior. Domination involves getting people to organize their behavior arwa a particular rule system. I n this context, "power is about the outcomes of issues enabled by tfce ri of a substantive rationality which is temporally a n d institutionally located" lid p . 77). Power, then, is achieved b y establishing a n organization's mode of ratio ality through controlling the d e e p structure rules of organizations. It is not s m ip a possession of individuals within organizations. By adopting this concept!

framework, Clegg is able to show h o w structures of domination can b e c o m e ^naturally" and "rationally" integrated into organizations. Power, in this sense, B legitimate only because a particular " m o d e of rationality" d e e m s it to b e so; the group that maintains this m o d e of rationality (the organizational rule system) therefore controls the organization's system of power. Several authors have sought to question the legitimacy of organizational models baseri on the notion that organizations are intrinsically rational in their modus operandi. These authors often characterize organizations as complex, ambiguity-ridden structures w h o s e m e m b e r s engage in a constant struggle to reduce the equivocal nature of the communication environment that they inhabit. Most notably, Cohen, March a n d Olsen's (1972; M a r c h & Olsen, 1976) "garbage can" model of organizing has d o n e m u c h to u n d e r m i n e traditional perspectives based on rational decision-making. An organization is a collection of choices looking for problems, issues and feelings looking for decision situations in which they m i g h t b e aired, solutions looking for issues to w h i c h they m i g h t b e t h e answer, a n d decision makers looking for work. (1972, p . 2) Building on this t h e m e , Weick (1979) points out that organizational activity is often made sense of. i.e., m a d e to seem rational, only in retrospect; there is frr-quently only a limited and incomplete a n t e c e d e n d y existing rationale which ities and directs organizational behavior. T h u s "organizations are often • ;

Discourse Studies (Sage Benchmarks Discourse Studies)

Gender and Discourse (Sage Studies in Discourse)

Discourse in Ritual Studies

Discourse, of Course: An overview of research in discourse studies

Discourse

Workplace Discourse (Continuum Discourse)

Discourse Strategies (Studies in Interactional Sociolinguistics)

Ireland and Postcolonial Studies: Theory, Discourse, Utopia

Tamhwa yŏnʼgu ŭi kichʻo Discourse Studies Korean

Discourse and Identity (Studies in Interactional Sociolinguistics)

Studies in Interactional Linguistics (Studies in Discourse & Grammar)

Studies in Interactional Linguistics (Studies in Discourse & Grammar)

Discourse in Ritual Studies (Empirical Studies in Theology)

Reason Discourse

Professional Discourse

Discourse Processing

Discourse Processing

Investigating Classroom Discourse (Domains of Discourse)

Using Corpora in Discourse Analysis (Continuum Discourse)

Confusing Discourse

Discourse, Figure

Discourse Strategies

Investigating Media Discourse (Domains of Discourse)

Discourse, Figure

Gendered Discourse in Professional Workplace (Palgrave Studies in Professional and Organizational Discourse)

Discourse of Blogs and Wikis (Continuum Discourse)

Discourse on the Move: Using corpus analysis to describe discourse structure (Studies in Corpus Linguistics)

Discourse and Practice: New Tools for Critical Discourse Analysis (Oxford Studies in Sociolinguistics)

Discourse and Practice: New Tools for Critical Discourse Analysis (Oxford Studies in Sociolinguistics)

Discourse as Structure and Process (Discourse Studies: A Multidisciplinary Introductio) (v. 1)

Discourse Studies (Sage Benchmarks Discourse Studies)

Gender and Discourse (Sage Studies in Discourse)

Discourse in Ritual Studies

Discourse, of Course: An overview of research in discourse studies

Discourse

Workplace Discourse (Continuum Discourse)

Discourse Strategies (Studies in Interactional Sociolinguistics)

Ireland and Postcolonial Studies: Theory, Discourse, Utopia

Tamhwa yŏnʼgu ŭi kichʻo Discourse Studies Korean

Discourse and Identity (Studies in Interactional Sociolinguistics)

Studies in Interactional Linguistics (Studies in Discourse & Grammar)

Studies in Interactional Linguistics (Studies in Discourse & Grammar)

Discourse in Ritual Studies (Empirical Studies in Theology)

Reason Discourse

Professional Discourse

Discourse Processing

Discourse Processing

Investigating Classroom Discourse (Domains of Discourse)

Using Corpora in Discourse Analysis (Continuum Discourse)

Confusing Discourse

Discourse, Figure

Discourse Strategies

Investigating Media Discourse (Domains of Discourse)

Discourse, Figure

Gendered Discourse in Professional Workplace (Palgrave Studies in Professional and Organizational Discourse)

Discourse of Blogs and Wikis (Continuum Discourse)

Discourse on the Move: Using corpus analysis to describe discourse structure (Studies in Corpus Linguistics)

Discourse and Practice: New Tools for Critical Discourse Analysis (Oxford Studies in Sociolinguistics)

Discourse and Practice: New Tools for Critical Discourse Analysis (Oxford Studies in Sociolinguistics)

Discourse as Structure and Process (Discourse Studies: A Multidisciplinary Introductio) (v. 1)

Recommend Documents