Discussing Conversational Analysis: The Workd of Emanuel A. Schegloff

Discussing Conversation Analysis

Discussing Conversation Analysis: The work of Emanual A. Schegloff presents an indepth view on Schegloff’s complex and stimulating work in Conversation Analysis (CA) and offers clear insights into how it has and may be developed further as a research tool in social psychology, social science, artificial intelligence, and linguistics. What is the status of fine-grained empirical studies of human interaction in CA and how does CA relate to other approaches to linguistic interaction? What is Schegloff’s contribution to CA and how does his work relate to that of Goffman, Garfinkel, and Sacks? How does CA distinguish its own analytical tools and terms from the categories of the participants in talk? What can CA reveal about human-computer interaction? What can CA contribute to the neurosciences in the study, diagnosis, and treatment of linguistically impaired individuals? How does CA account for the socio-historical dimension of the material and semiotic resources that participants co-deploy in talk? By addressing these and other questions this volume proposes a critical guide to CA and its applications with an extraordinary interview with Emanuel A. Schegloff, and new contributions towards a debate on his work by six commentators — conversation analysts (John Heritage and Charles Goodwin), critics (Rick Iedema and Pär Segerdahl) and appliers of CA in the study of human-computer interaction (Pirkko Raudaskoski) and language disorders (Ruth Lesser). Schegloff’s Response Essay and a final discussion conclude the volume, which also features a comprehensive bibliography of his work edited by Susan Eerdmans.

Emanuel A. Schegloff is Professor of Sociology with a joint appointment in Applied Linguistics at the University of California, Los Angeles. Educated at Harvard and the University of California, Berkeley, he has taught at Columbia University as well as at UCLA. He received a Guggenheim Fellowship, and was a resident Fellow at the Netherlands Institute of Advanced Study in the Humanities and Social Sciences (1978–79) and the Center for Advanced Study in the Behavioral Sciences, Stanford (1998–99).

Discussing Conversation Analysis The work of Emanuel A. Schegloff Edited by

Carlo L. Prevignano University of Bologna

Paul J. Thibault University of Venice/Lingnan University, Hong Kong

John Benjamins Publishing Company Amsterdam
/
Philadelphia

8

TM

The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences – Permanence of Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data Discussing conversation analysis : The work of Emanuel A. Schegloff / edited by Carlo L. Prevignano and Paul J. Thibault. p. cm. Includes bibliographical references and indexes. 1. Schlegloff, Emanuel A.--Contributions in conversation analysis. 2. Conversation analysis. I. Prevignano, Carlo. II. Thibault, Paul J. P85. S3D57 2003 302.3’46-dc21 isbn 90
272
2599
0 (Eur.) / 1
58811
354
X (US) (Hb; alk. paper)

2002043639

© 2003 – John Benjamins B.V. No part of this book may be reproduced in any form, by print, photoprint, microfilm, or any other means, without written permission from the publisher. John Benjamins Publishing Co. · P.O. Box 36224 · 1020 me Amsterdam · The Netherlands John Benjamins North America · P.O. Box 27519 · Philadelphia pa 19118-0519 · usa

Contents

Preface Chapter 1 Presenting Emanuel A. SchegloŸ John Heritage

vii

1

Chapter 2 On conversation analysis: An interview with Emanuel A. SchegloŸ Svetla Cmejrková and Carlo L. Prevignano

11

Chapter 3 The power of SchegloŸ’s Work Charles Goodwin

57

Chapter 4 Putting SchegloŸ’s principles and practices in context Rick Iedema

65

Chapter 5 Conversation analysis as rigorous science Pär Segerdahl

91

Chapter 6 Users’ interpretations at a computer tutorial: Detecting (causes) of misunderstandings Pirkko Raudaskoski Chapter 7 When conversation is not normal: The role of conversation analysis in language pathology Ruth Lesser Chapter 8 Response Emanuel A. SchegloŸ

109

141

157

vi

Discussing Conversation Analysis

Chapter 9 Continuing the interview with Emanuel A. SchegloŸ Carlo L. Prevignano and Paul J. Thibault

165

Chapter 10 A bibliography of Emanuel A. SchegloŸ Edited by Susan L. Eerdmans

173

Subject index Author index

183 189

Preface

Our aim in this book is to present a number of new contributions towards a debate1 on Conversation Analysis, as developed by Emanuel A. SchegloŸ. The present work is very much in the spirit that we propose in its companion, Language and Interaction: Discussions with John J. Gumperz (Eerdmans, Prevignano and Thibault 2003). Before presenting the individual contributions to the present volume, we would like to say a few words concerning its preparation. Carlo Prevignano, when he was Secretary of the International Association for Dialogue Analysis, had the idea, during a congress of the Association in Prague (April 1996), of an interview with SchegloŸ, which was realized thanks to the collaboration of Svetla Cmejrková, who also took part in it. Paul Thibault transcribed the interview, which was then revised, updated and furnished with bibliographical citations by SchegloŸ himself. As a result, the original oral interview was transformed into a dense and extraordinary essay by SchegloŸ.2 At this point, we decided to place the revised text in the hands of a number of scholars for a peer commentary from three diŸerent perspectives: (i) those who support and collaborate with SchegloŸ’s Conversation Analysis; (ii) critics of Conversation Analysis from diŸerent theoretical points of view; and (iii) those who apply Conversation Analysis in some particular domain of social practice. John Heritage gives a brilliant initial presentation of SchegloŸ’s work. Charles Goodwin, another close collaborator of SchegloŸ, provides an elegant personal commentary on his relationship to SchegloŸ’s work, as well as on the interview. In highlighting SchegloŸ’s seminal contribution to the growth and development of Conversation Analysis, Heritage calls our attention to SchegloŸ’s unstinting commitment to ªne-grained empirical studies of the details of interactional conduct. In so doing, Heritage reveals the ways in which Conversation Analysis has provided important insights into the principles of social organization which infuse the practices and behaviors out of which human social interaction is made. Heritage cautions the reader as to the particular challenges posed by SchegloŸ’s commitment to ªne-grained empirical speciªcation. In particular, he argues that the theoretical, empirical, and methodological aspects are so complexly interrelated in SchegloŸ’s work that

viii Discussing Conversation Analysis

only a “careful, attentive reading,” based on a full acquaintance with the richness and analytical reach of his work over the past thirty or more years, can yield a full appreciation of the many lessons to be drawn from that remarkable body of work. Goodwin provides a personal account of his own struggles to come to grips with the richness and complexity of multi-party social interaction. Citing his profound dissatisfaction with the research methods and theoretical tools that he found in Social Psychology at the beginning of the 1970s, Goodwin recounts how he ªrst came to the Conversation Analysis practised by Sacks and SchegloŸ.3 He found that CA illuminated many aspects of multi-party social interaction that Social Psychology, as it then stood, was unable to solve. Goodwin draws attention to the importance of showing how participants in talk jointly construct their social interaction at the same time as they develop their own understandings of what is going on. Goodwin also provides a compelling account of his research on the talk of neurologically-impaired patients. In this way, he shows how Chil, who is an aphasic patient studied by Goodwin himself, is able to produce meaningful talk and action in the contexts that make up his lifeworld in spite of the severity of his linguistic impairment. In ways which echo Gregory Bateson’s valuable reminder as to how pathological communication has much to teach us about normal communication and vice versa, Goodwin points out that the analyst of aphasic discourse “needs to keep one eye ªrmly focussed on the practices of normal conversationalists.” In the interests of balance, the editors also turned to two “critics” of Conversation Analysis, Rick Iedema and Pär Segerdahl.4 Iedema provides an alternative theoretical and analytical perspective based on systemic-functional linguistic theory and a materialist view of social meaning-making. According to Iedema, SchegloŸ’s analysis does not address a number of dimensions of social interaction such as the socio-historical dimension of the resources — semiotic and material — which are mobilized on occasions of social interaction. These processes of resource mobilization also involve the dynamic deployment of bodily resources such as neuromuscular movement in order to materially structure and transform their interactions (see also Cowley 1998; Thibault 2003; Thibault in press). Segerdahl, in his contribution, raises doubts concerning the claim of Conversation Analysis that its tools of analysis can be used to describe not only what we do when we converse, but above all what makes us do what we do. Thus, he critiques the notion that mechanisms in the form of “devices,” “methods” or “rules” both underlie and can be used to explain conversational events.

Preface

Speciªcally, Iedema focuses on three aspects of SchegloŸ’s work which he ªnds problematic. The ªrst of these is the assumption, which he attributes to SchegloŸ, that analysts have unproblematic and transparent access to the relevance of talk for participants. Iedema raises the question as to whether analysts “inevitably” construct partial narratives, informed by their own standpoints and interests. He also questions the notion of the ‘ordinariness’ and ‘naturalness’ of casual talk, and whether this assumption is a useful starting point for the analysis of talk. Finally, Iedema asks whether it is useful for CA to concentrate on what he characterizes as the “constants of talk”, without also attending to the dynamic, the changing, and the ephemeral. This last question puts the focus on the individuating character of occasions of discourse. The assumption of the “constants of talk” usefully draws attention to the typical or generic characters of the resources which participants co-deploy in their interactions with each other. On the other hand, Iedema’s focussing on the ¶eeting, the dynamic and the ephemeral asks us to consider the question of change and how we can hone our tools of social analysis to observe and understand the emergence of the new in the social making of meanings. Furthermore, it leads to a renewed attention to the individuating characteristics of discourse. In this perspective, we see how contextual contingencies contribute to the elaboration of semiotic complexity in systems as a result of these contingencies, leaving their marks on the individuals who engage in discursive activity. Segerdahl shares with Iedema a concern with the relevance of the technical terms of Conversation Analysis to the participants to occasions of talk. While acknowledging the ways in which CA has made us more self-re¶exive about our own language use, in contrast to the mechanistic and objectivist metaphors that have prevailed in much of linguistics, Segardahl claims that the assumption on the part of SchegloŸ and co-workers that participants in occasions of talk orient to the normative constructions of CA is an illusion. He argues that the solution to this dilemma lies in the need to distinguish the analytical tools of CA from actual conversation, rather than trying to show that they coincide. Segardahl’s appeal requires us, as analysts, to take ourselves into account in our own theoretical constructions. Such a move would foster an increased awareness of the necessarily incomplete nature of all our models, along with the always partial and historically-situated character of the observer perspectives in terms of which our models are always — implicitly or explicitly — constructed. In this way, we can see that “actual conversation,” as Segardahl calls it, is not a given, but is itself a sign that we construct on the basis of our (theoretical) interpretations. That is, we must recognize that CA, for example, is itself a discourse for contextualizing ‘actual conversation.’

ix

x

Discussing Conversation Analysis

Part of Segardahl’s critique centres on his interpretation of SchegloŸ as assuming rules or devices that function as causal explanations of how participants behave in conversation.5 Generally speaking, the tendency to look for underlying devices, rules, or procedures as causal explanations of conversational activity goes hand-in-hand with a view that complex systems such as human social meaning-making can be predicted in ways that satisfy the requirements of the scientiªc view of the world as something that can be measured, predicted, and controlled. In this view, the devices, procedures, and rules critiqued by Segardahl, in ways not unlike the “constants of talk” discussed by Iedema, raise questions about a simplifying process of isolating such devices, ascertaining their constancy, and then seeing if conversation will yield to prediction. But, as Iedema also suggests, the ¶eeting, the dynamic, and the ephemeral mean that complex self-organizing systems like human discourse do not easily yield to the simplifying requirements of prediction without sacriªcing their semiotic complexity. The contributions of Pirkko Raudaskoski and Ruth Lesser present applications of Conversation Analysis in the ªelds of human-computer ‘interaction’ and neuropsychology, respectively.6 Raudaskoski raises interestingly controversial questions concerning the applicability of Conversation Analysis in her approach to both human-computer interaction and human-human interaction in the context of a word processor tutorial. Lesser shows how Conversation Analysis is necessarily adapted and modiªed when applied to the ªeld of language disorders. In this way, she shows how ‘surface’ features of conversations involving aphasia and dementia patients will be aŸected by their brain damage. Raudaskoski adapts CA to the detailed analysis of a single case study, consisting of the interaction between two novice users of Microsoft Word 5.0 and the Learning Microsoft Word program, in which written linguistic text and visual semiotic resources are co-deployed. The episode, which was videotaped so as to permit the analysis of participants’ gaze, gesture, and other bodily activities, as well as the computer screen, consists of what Raudaskoski refers to as “two participation frameworks,” viz. the human-computer interaction and human-human interaction. Her analysis, which focuses on a misunderstanding on the part of one of the interactants, provides valuable insights into the ways in which conversational routines are integrated into activities which involve the cross-coupling of both semiotic and material resources. She shows how the computer functions as an artefact whose materiality aŸords semiotically-mediated forms of activity. Such activities are themselves based on the interactants’ understandings and misunderstandings of the semiotic-material aŸordances of

Preface

the computer-cum-artefact. These aŸordances are adaptively appropriated by the two novices in the context of their negotiations of the computer-generated texts which are designed to assist in the learning of MS Word. In this way, Raudaskoski’s analysis suggests how CA may play its role in revealing how processes at the level of the activity, rather than the individual participants per se, can be linked to broader socio-cultural patterns of learning. Lesser’s contribution shows how CA can help to reverse the longstanding focus in the study of language-disordered individuals on the individual mind/ brain. For Lesser, CA is an important tool in the description of the “missing dimension” (Lesser, this volume: 142) of the ways in which neurological impairments adversely aŸect everyday conversational interaction. Her chapter reports on the progress she and her collaborators have made in adapting CA to this descriptive end with a view to developing more reªned diagnostic and therapeutic tools. She shows how assessments of the speciªc ways in which neurological deªcits such as aphasia and dementia produce disrupted talk can be derived from the ªne-grained analytical tools and descriptive techniques of CA, and in ways which complement the top-down theoretical constructs that have often prevailed in previous attempts in psycholinguistics and cognitive neuropsychology. While cautioning that we still know very little about the relations between neurological processes and conversational activity, Lesser’s discussion helps us to focus on the fact that what happens in the brain is closely coupled to the contextual activities in which the body-brain is embedded as a co-participant, rather than as their prior cause. The structures and patterns of conversational routines are integral parts of and extensions of neural activity, rather than being caused by it. Lesser’s discussion therefore implies that brain and conversational practices are integral components of a wider system of relations which is not reducible to the isolated biological organism. While her focus is on the descriptive e¹cacy of CA in the context of speech and language pathology and therapy with respect to this larger enterprise, her chapter can be seen as a valuable and stimulating contribution to the development of the analytical tools and theoretical frameworks that are needed to close the gap between the individual biological organism and transindividual conversational practice. SchegloŸ, in turn, selectively comments on the contributions proposed in this volume in his “Response,” where he revisits his own earlier contribution as well as replies to and takes issue with the individual commentators in ways that we hope will encourage readers to respond to both CA and the other approaches that are invoked with new interpretations, new questions, and new lines of

xi

xii Discussing Conversation Analysis

further research and inquiry. SchegloŸ also responds to a small number of ªnal questions which we put to him under the rubric of “Continuing the Interview.” In particular, he clariªes the analytical and historical status of the notion of “turn” with respect to CA, thereby revealing a number of likely misunderstandings of this notion. This initial clariªcation paves the way for further illuminating observations on the organizational issues that are central to CA and without which “talk-in-interaction” could not proceed. In particular, SchegloŸ singles out the organization of turn-taking, of sequences, of repair, and the overall structural organization of conversational episodes. For SchegloŸ, “sustained episodes of talk-in-interaction” (Prevignano and Thibault, this volume: 165) constitute solutions to these organizational issues. In “Continuing the Interview,” SchegloŸ also makes a number of thoughtful observations on the relationship of CA to a number of neighbouring approaches to conversation, while at the same time rea¹rming the theoretical and analytical distinctiveness of CA. At the conclusion of the volume, we have added a list of SchegloŸ’s publications, kindly provided by SchegloŸ himself, and edited by Susan L. Eerdmans. A few ªnal words of thanks are in order. We are deeply indebted to Emanuel A. SchegloŸ for his constant support and collaboration throughout all the stages of this project from its inception in Prague to the production of this volume. Without his prompt responses to our many requests, this project would not have been possible. We wish to thank Susan Eerdmans for the assistance she provided us in the preparation of the text for publication. We are also indebted to Bertie Kaal of John Benjamins Publishing Company for believing in our project from the outset, for supporting it from beginning to end, and for providing valuable suggestions and guidance which greatly helped us to improve the original manuscript. Carlo L. Prevignano Paul J. Thibault Bologna (Italy) and Hong Kong, July-August 2002

Notes 1. Among the contributions to this debate over the past decade, we recall Searle (1992a, 1992b) and SchegloŸ (1992), all of which are found in the volume, (On) Searle on Conversa-

Preface xiii

tion; also see Wetherell (1998) as a reaction to SchegloŸ (1997), and SchegloŸ’s reply (1998); ªnally, we would like to mention the exchange between Michael Billig and Emanuel A. SchegloŸ in Discourse and Society 10 (4), introduced by Van Dijk (1999). The relevant citations are Billig (1999a), SchegloŸ (1999a), Billig (1999b) and SchegloŸ (1999b). See also Mey 2001. 2. In the meantime, another interview with SchegloŸ has appeared (see Wong and Olsher 2000). 3. See also Sacks and Schegloff (2002[1975]). 4. Among critics of Conversation Analysis, we refer readers to some pertinent observations made by Gumperz (see Prevignano and Di Luzio 2003: 11; Gumperz 2003: 117–118). Gumperz has argued that conversational involvement rests on conversational inferences, rather than on “structural givens” such as the sequential ordering of turns in conversation. According to Gumperz, sequential analysis alone cannot account for situated interpretation: for Gumperz, sequential analysis is just one of the indexical resources that aŸect inferencing (see Gumperz 2003), where inferencing is to be understood as members’ procedures for the situated interpreting of conversation on the basis of the contextualization cues which interact with and co-contextualize the symbolic dimensions of lexicogrammatical forms in discourse. 5. Segerdahl’s interpretation of the causal status of these presumed rules and devices is explicitly rejected by SchegloŸ in his “Response.” 6. For a discussion of CA and applied linguistics, see SchegloŸ et al. 2002.

References Billig, Michael. 1999a. “Whose terms? Whose ordinariness? Rhetoric and ideology in Conversation Analysis”. Discourse and Society 10 (4): 543–558. Billig, Michael. 1999b. “Conversation Analysis and the claims of naivety”. Discourse and Society 10 (4): 572–576. Cowley, Stephen J. 1998. “Of timing, turn-taking, and conversations”. Journal of Psycholinguistic Research 27 (5): 541–571. Eerdmans, Susan L., Prevignano, Carlo L. and Thibault, Paul J. (eds.). 2003. Language and Interaction: Discussions with John J. Gumperz. Amsterdam and Philadelphia: John Benjamins. Gumperz, John J. 2003. “Response essay”. In Language and Interaction: Discussions with John J. Gumperz, S. L. Eerdmans, C. L. Prevignano and P. J. Thibault (eds.), 105–126. Amsterdam and Philadelphia: John Benjamins. Lesser, Ruth. “When Conversation is not Normal: the Role of Conversation Analysis in Language Pathology”. This volume. Mey, Inger. 2001. “The CA/CDA controversy”. Journal of Pragmatics 33: 609–615. Prevignano, Carlo L. and di Luzio, Aldo. 2003. “A discussion with John J. Gumperz”. In Language and Interaction: Discussions with John J. Gumperz, S. L. Eerdmans, C. L. Prevignano and P. J. Thibault (eds.), 7–29. Amsterdam and Philadelphia: John Benjamins.

xiv Discussing Conversation Analysis

Prevignano, Carlo L. and Thibault, Paul J. “Continuing the interview with Emanuel A. SchegloŸ”. This volume. Sacks, Harvey and SchegloŸ, Emanuel A. 2002 [1975]. “Home position”. Gesture 2(2). Includes 16 videoclips. SchegloŸ, Emanuel A. 1992. “To Searle on conversation: A note in return”. In (On) Searle on Conversation, H. Parret and J. Verschueren (eds.), 113–128. Amsterdam and Philadelphia: John Benjamins. SchegloŸ, Emanuel A. 1997. “Whose text? Whose context?” Discourse and Society 8 (2): 165– 187. SchegloŸ, Emanuel A. 1998. “Reply to Wetherell”. Discourse and Society 9 (3): 413–416. SchegloŸ, Emanuel A. 1999a. “` SchegloŸ’s texts’ as ‘Billig’s data’: A critical reply”. Discourse and Society 10 (4): 558–572. SchegloŸ, Emanuel A. 1999b. “Naivete vs. sophistication or discipline vs. self-indulgence: A rejoinder to Billig”. Discourse and Society 10 (4): 577–582. SchegloŸ, Emanuel A. “Response”. This volume. SchegloŸ, Emanuel A., Koshik, Irene, Jacoby, Sally and Olsher, David. 2002. “Conversation Analysis and applied linguistics”. Annual Review of Applied Linguistics 22 [Discourse and Dialogue], M. McGroaty (ed.), 3–31. Searle, John R. 1992a. “Conversation”. In (On) Searle on Conversation, H. Parret and J. Verschueren (eds.), 7–29. Amsterdam and Philadelphia: John Benjamins. Searle, John R. 1992b. “Conversation reconsidered”. In (On) Searle on Conversation, H. Parret and J. Verschueren (eds.), 137–147. Amsterdam and Philadelphia: John Benjamins. Thibault, Paul J. 2003. “Body dynamics, social meaning-making, and scale heterogeneity: Re-considering contextualization cues and language as mixed-mode semiosis”. In Language and Interaction: Discussions with John J. Gumperz, S. L. Eerdmans, C. L. Prevignano and P. J. Thibault (eds.), 127–147. Amsterdam and Philadelphia: John Benjamins. Thibault, Paul J. In press. Brain, Mind and the Signifying Body: An Ecosocial Semiotic Theory of Meaning-Making, Embodiment, and Consciousness. London and New York: Continuum. Van Dijk, Teun. 1999. “Critical Discourse Analysis and Conversation Analysis”. Discourse and Society 10 (4): 459–460. Wetherell, Margaret. 1998. “Positioning and interpretative repertoires: Conversation analysis and post-structuralism in dialogue”. Discourse and Society 9 (3): 387–412. Wong, Jean and Olsher, David. 2000. “Re¶ections on conversation analysis and nonnative speaker talk: An interview with Emanuel A. SchegloŸ”. Issues in Applied Linguistics 11 (1): 111–128.

Chapter 1

Presenting Emanuel SchegloŸ John Heritage

Emanuel SchegloŸ is a co-originator (with the late Harvey Sacks and Gail JeŸerson) and the leading contemporary authority in the ªeld which has come to be known as ‘conversation analysis’ (henceforth, CA). This ªeld, once described as “perhaps the only completely new sociological research methodology developed in the United States since World War II,” ªrst found published expression in SchegloŸ’s “Sequencing in Conversational Openings” (1968). Since that time, the ªeld has developed very substantially. Many hundreds of research papers and monographs in conversation analysis have now been published, and, in addition to North America, CA is practiced in many countries in Europe, Scandinavia, Asia and Australasia. As this list suggests, CA is ªnding application in a substantial number of languages. Its use has also broadened from its original and primary domain — ordinary conversation — to increasingly diverse social interactions ranging from those in medical, educational and legal settings, to those involving the deployment of complex communication technology, and from studies with a focus on the acquisition of language and communicative competence to those which focus on interactional and pragmatic aspects of their loss. Both the specialty and its ªndings are very widely recognized among the practitioners of such cognate social science ªelds as anthropology, linguistics, psychology, cognitive science and communication, and are ªnding applications in ªelds such as medicine, neuroscience, and human-computer interaction. During the past thirty or so years, SchegloŸ has pursued a long series of studies which have established many of the major concepts and ªndings that are now treated as axiomatic in the ªeld. The outstanding insight, cogency and ingenuity of his conceptualization of conversation analysis and of his research output are internationally acknowledged as a major intellectual achievement. SchegloŸ has developed the ªeld of CA, not by means of theoretical manifestos, but rather and exclusively through a series of ªne-grained empirical studies of

2

Discussing Conversation Analysis

the details of interactional conduct. These studies have shown the truly remarkable degree to which social organization — a ‘syntax’ of action — inhabits the practices and behaviors that make up human social interaction. They have constituted an extended demonstration that and how the empirical details of human interactional conduct can be brought under precise analytic control, and have served as an inspiration to the specialty which SchegloŸ has cofounded. In developing conversation analysis, SchegloŸ has established a major sociological input into a domain — linguistic behavior — which was previously treated largely as the province of philosophy, psychology and linguistics. A number of signiªcant sociological theorists — for example, Marx, Durkheim, Mead, Parsons, Bourdieu and Habermas — have commented, in some cases extensively, on the social character of language and its signiªcance as a social institution and as a medium of action. None of these writers, however, has been able to develop a conceptually coherent framework for the sociological analysis of interactional conduct, let alone develop that framework into an empirical discipline through a cumulative and interlocking series of empirical investigations. SchegloŸ’s development of CA has involved a major reconceptualization of extant perspectives on the nature of language and social interaction, of the kinds of data which are relevant and appropriate for the study of language, and of the analytic procedures through which empirical investigation may best be forwarded. This reconceptualization is based on the recognition that social interaction is, as he puts it, “the primordial site of human sociality,” and that the demands of social interaction are central in shaping the development and use of language. At the same time, it embodies the recognition that the sociological study of interaction cannot be developed as a coherent discipline without detailed attention to the ways in which the properties of language are systematically exploited in the prosecution of interactional tasks. The hallmark of SchegloŸ’s work, then, is its treatment of language behavior as situated social action and interaction, whose details may be proªtably investigated from a speciªcally sociological perspective. As a sociological approach to the study of language and social interaction, CA embodies a synthesis of the perspectives of two highly original social scientists, Erving GoŸman and Harold Garªnkel, both of whom directly in¶uenced Sacks and SchegloŸ. From GoŸman (1955, 1964, 1983) came the notion that social interaction is not merely a medium through which other aspects of the social world — for instance, social status, gender or personality — are

Presenting Emanuel Schegloff

manifested. Instead, GoŸman insisted that social interaction is to be conceived as a social institution in its own right, with its own normative organization and moral obligations, which, in turn, are linked to other aspects of the social world through face, role and identity. Drawing from Durkheim and RadcliŸeBrown, GoŸman conceived social interaction as the product of a set of moral rights and rituals — a “syntax” as he once put it (GoŸman 1967: 2) — irreducible to individual psychology. It was this conception which licensed, indeed mandated, the study of social interaction — what GoŸman (1983) later termed the “interaction order” — as a subject matter in its own right. Harold Garªnkel’s (1967) ethnomethodology contributed a diŸerent, but complementary, perspective. Emerging from the phenomenological tradition for which the problem of intersubjectivity and shared understanding had been a major stumbling block, Garªnkel’s researches developed the proposition, whose main progenitors were Alfred Schutz and Ludwig Wittgenstein, that shared understanding and mutual intelligibility among humans are possible only through approximate, revisable and, above all, practical and shared methods of reasoning whose results are unavoidably inscribed in courses of social action. This approach provided the basis for the notion, developed in Sacks and SchegloŸ’s earliest papers, that actions are a resource through which the parties to an interaction can see, and see in common, “where they are” in a given interaction, and how each is positioned relative to the other. Building from these perspectives, CA focusses on the competencies which persons use and rely on to co-construct orderly and mutually understandable courses of action. Accepting John Austin’s supposition that we “do things with words,” CA has developed as a program of research by mapping the resources with which members of the social world produce, recognize, understand and manipulate spoken interactions. Its basic assumption is that while the resources for the construction of conduct are highly institutionalized (in a Durkheimian or Parsonian sense), they also serve as the building blocks for highly particularized courses of conduct, and for speciªcally meaningful activities ªtted to the singular characteristics of particular persons and contexts. The creative fusion of perspectives derived from GoŸman and Garªnkel is present from the earliest papers that launched conversation analysis. For example, in “Opening Up Closings” (SchegloŸ and Sacks 1973), there is an analysis of a two-sequence process through which telephone conversations are ordinarily closed, comprising: (i) a pre-closing exchange (e.g., Okay-Okay) and (ii) a terminal exchange (e.g., Goodbye-Goodbye). A GoŸman-style analysis might deal with these sequences as an interactional ritual, departures from which

3

4

Discussing Conversation Analysis

would be face-threatening and sanctionable. The CA treatment acknowledged the moral requiredness of the sequences, while also giving an analysis of their functional necessity as resources whereby the participants can show their readiness (and negotiate) to: (i) move in the direction of jointly terminating an interaction, and (ii) actually achieve a complete termination. In this analysis, the two sequences are shown to be the parties’ practical means for analyzing and showing “where we are” in the conversation, and this, in turn, is yoked to the moral considerations that bear on their participation within them. The result is a single integrated analytic framework for the analysis of social action that transcends the separate insights of GoŸman and Garªnkel. As the analysis in “Opening Up Closings” and other early CA papers clearly showed, because actions embody both re¶ections on the past and projections into the future, and do so re¶exively and co-constructively, they are resources whereby the participants’ analyses of the past are expressed in current actions which deªne the present and, step-by-step, progress the construction of the interaction’s future. It is one thing to describe these basic characteristics of interaction and to promulgate their centrality to any analysis of its organization, it is quite another to embody these general principles in conceptualizations and analyses of speciªc interactional phenomena. The growth and development of CA has been substantially contingent on this achievement to which SchegloŸ has made a substantial, indeed latterly, an overwhelming, contribution. Paper after paper, whether single authored or co-authored with Sacks and JeŸerson, identiªes major structural axes of interaction: turn-taking (Sacks, SchegloŸ and JeŸerson 1974), sequence organization (SchegloŸ 1968, 1972, 1995, SchegloŸ and Sacks 1973), repair (SchegloŸ, JeŸerson and Sacks 1977), overall structural organization of conversation (SchegloŸ 1968, SchegloŸ and Sacks 1973), word selection (SchegloŸ 1996b), turn organization (SchegloŸ 1996c), and so on. Each one of these papers establishes a domain of study by identifying absolutely fundamental choices that participants in conversation must make, and isolating elements of the functional architecture through which these choices are made. None of these papers is intended to be deªnitive of its topic; each sketches a terrain. Yet, remarkably, while most of these areas have undergone extensive expansion over the years, whether in the form of progressive internal reªnements in the empirical territory staked out, or through wholesale additions to it, the basic frameworks that were developed in the initial papers as much as thirty years ago have remained remarkably intact. For example, the initial work on turn-taking, which speciªed that a participant was initially entitled to a single turn constructional unit, and that speaker

Presenting Emanuel Schegloff

transition can be achieved without signiªcant gap or overlap, has come to accommodate collaborative utterances (Lerner 1991, 1996), and a machinery of overlap management (SchegloŸ 2000). It has also easily accommodated a range of ªndings about the role of gaze, gesture and body deployment in the management of turn entry and turn exit (Goodwin 1981, 1984, 1986). It has been integrated with work on prosody (Local and Kelly 1986, Couper-Kuhlen and Selting 1996, SchegloŸ 1998). It helped to motivate work aimed at understanding how turn-taking is initiated (SchegloŸ 1968), terminated (SchegloŸ and Sacks 1973) and suspended (Sacks 1974). It has also served as a powerful resource when examined comparatively with other non-conversational turntaking systems, such as those which are characteristic of courtroom questioning, or interviews (Atkinson and Drew 1979, Heritage and Greatbatch 1991, Button 1992). Similarly, the initial speciªcation of sequential organization in terms of the notion of conditional relevance and the adjacency pair concept (SchegloŸ 1968, 1972, SchegloŸ and Sacks 1973) has been expanded to include pre-sequences (SchegloŸ 1980, 1988b, 1988c) that occur prior to the base pair and which may be directed either at the upcoming ªrst-pair part (for example, to have the other oŸer something, rather than the speaker to request it [SchegloŸ 1984, 1995]) or directed at the second-pair part to shape the likelihood that a particular secondpair part (e.g., an acceptance) will be forthcoming. It also accommodates insertion sequences, and post-expansions (SchegloŸ 1972, 1995). Or again, an initial paper on repair organization (SchegloŸ, JeŸerson and Sacks 1977) has been enriched with papers on self-repair in various places in a speaker’s turn (SchegloŸ 1979, 1987b), papers on the origins and corrections of misunderstanding (SchegloŸ 1987c, 1992b, 1997). These lists could be elaborated. The conclusion I draw from this concerns the extraordinarily far-sighted theoretical conceptualization of the original papers, which quite deliberately left gaps where the evidence was incomplete, and developed the basic concepts in ways that did not foreclose on the possibilities of further discovery. The result is that the major conceptual maps which SchegloŸ has sketched have not had to be redrawn and that there has been, and remains, tremendous scope for others to contribute to the emerging body of work. The formality and generalizability of much of SchegloŸ’s work (and indeed of work in CA more generally) has led some to the conclusion that his work embodies an indiŸerence to the contexts in which interaction occurs — what GoŸman referred to as “sins of noncontextuality” (GoŸman 1981: 32, see also Cicourel 1987, Duranti 1997). This criticism has always ba§ed those of us

5

6

Discussing Conversation Analysis

who know SchegloŸ’s work well. Leaving aside the fact that much of what we mean by the ‘context’ of an utterance consists of the prior turn and sequence in which it is produced, and that commonly the most relevant ‘distal’ elements of social context are embedded in those prior turns and sequences, and setting aside also the broader point that the whole of CA can be construed as an attempt to explicate the meaning of ‘context’ in interaction, there is the evidence concerning his attitude towards context that comes from SchegloŸ’s papers. Surely there are few more nuanced and detailed accounts of the contexts of utterances than those rendered by SchegloŸ himself, especially in those of his papers that are largely devoted to the explication of a single episode of interaction. And his ability to link context with the details of utterances in informal settings like data sessions is legendary. It is di¹cult for us to recognize the SchegloŸ of these subtle analyses in the comments of those critics who assert that SchegloŸ’s work takes little account of context. A key to this misperception may arise from looking at the diŸerent research objectives that SchegloŸ sets himself in diŸerent papers. Some of these papers (for example, SchegloŸ 1987a, 1988a, 1992a) pursue what might be termed a particularizing objective, and are aimed at grasping what was accomplished in a singular course of action. For these papers, the explication of the context that is relevantly invoked and ‘in play’ for the participants must be, and is, central to the analysis he develops. However, the majority of SchegloŸ’s papers aim at maximizing the level of generalization that can be achieved about the role and signiªcance of particular practices of interaction. In developing these papers, the concern with context must be primarily ‘negative’. The analyst must be concerned with investigating the contextual ‘boundaries’ of the practice in question, and with examining ways in which it can be particularized. For example, in his paper on “Conªrming Allusions” (SchegloŸ 1996a), dealing with the ways in which a repeat can be used to conªrm another’s understanding of what was previously conveyed as an allusion, SchegloŸ points to the range of activities which can be achieved through the practice, but without compromising the basic generalization that this kind of repeat recurrently does “allusion conªrming.” Thus, although, in research pursuing this second, generalizing kind of objective, ‘context’ will tend to be downplayed relative to the generalizations that the analyst is in search of, this downplaying should not be confused with an ignorance of the contexts of each data extract, or a failure to take them into account, or to factor them into consideration in constructing the generalizations that are arrived at.

Presenting Emanuel Schegloff

In the last analysis, CA constitutes the most sustained attempt yet mounted to build a natural history of human interaction; a natural history variously in¶ected by culture, but likely embodying a substantial mass of commonality in human interactional practices, and possibly a good measure of universality. This eŸort at a natural history of interaction is one that has had many obstacles. Perhaps most fundamentally there is what SchegloŸ once described as “implacable familiarity of the materials with which we work.” This familiarity of the ‘already known’ is one that can lead observers to doubt the point of what is being done, and delude researchers into the belief that there is nothing there to be discovered. Then there is the belief, common to linguists (Chomsky 1965), philosophers (Searle 1969) and sociologists (Parsons 1937), that the empirical world of social action and interaction is too chaotic to sustain serious empirical research, as if the organization of social action were just a kind of random Brownian motion. And then there is the view that conversation is just idle chit-chat, and that one would be better spending one’s time studying serious, consequential interactions, or looking at the ‘real business’ (for example, status advancement, or group maintenance) that the actions implement. Or that one might as well study invented actions as embodied in theatrical scripts, or even scenarios one made up oneself. Against all of this (and many, more discipline-speciªc obstacles as well), SchegloŸ’s research has repeatedly and painstakingly demonstrated that behind the ‘implacable familiarity’ of everyday actions lie fascinating and exact orders of organization of great generality and scope. Again and again he has shown that, in social interaction, it is order rather than chaos that is the norm; precise, speciªc order; order that the participants use and rely on to achieve their interactional objectives. Above all, almost every paper he has written underwrites the notion that because “language is the vehicle for living real lives,” the primary research site for CA must be the ‘real life’ of ordinary conversational interaction. These are demonstrations that cannot be imagined or theorized; they can only be done empirically. SchegloŸ’s manifesto-free method of articulating theory through detailed empirical speciªcation poses a particular challenge for his audience. Only careful, attentive reading can fully extract the conceptual, empirical and methodological lessons to be drawn from his work. Here it should be noted that all of SchegloŸ’s papers, even those which seem to be created just for the occasion, are designed to be interlinked and to make a contribution to the overall architectonic structure of CA ªndings. Each of them conceptualizes a

7

8

Discussing Conversation Analysis

practice, or an organization of practices, or elaborates on the details of an organization, thus contributing structure or texture (and commonly both) to the overall architecture of the domain under investigation. Though free-standing, his papers should not, in the end, be read as if they were fully independent of one another. They are written as parts of a whole fabric. Each one is a brick in an open analytical ediªce which invites further additions. It is SchegloŸ’s privilege to have contributed more than a few of the foundation stones of this structure, just as it is ours to celebrate that contribution, and to build on it.

References Atkinson, J. Maxwell and Drew, Paul. 1979. Order in Court: The Organisation of Verbal Interaction in Judicial Settings. London: Macmillan. Button, Graham. 1992. “Answers as interactional products: Two sequential practices used in job interviews”. In Talk at Work: Interaction in Institutional Settings, P. Drew and J. Heritage (eds.), 212–231. Cambridge: Cambridge University Press. Chomsky, Noam. 1965. Aspects of the Theory of Syntax. Cambridge, Mass.: MIT Press. Cicourel, Aaron V. 1987. “The interpenetration of communicative contexts: Examples from medical encounters”. Social Psychology Quarterly 50 (2): 217–226. Couper-Kuhlen, Elizabeth and Selting, Margret. (eds.) 1996. Prosody in Conversation: Interactional Studies. Cambridge: Cambridge University Press. Duranti, Alessandro. 1997. Linguistic Anthropology. Cambridge: Cambridge University Press. Garªnkel, Harold. 1967. Studies in Ethnomethodology. Englewood CliŸs, N. J.: Prentice-Hall. GoŸman, Erving. 1955. “On face work”. Psychiatry 18: 213–231. GoŸman, Erving. 1964. “The neglected situation”. In The Ethnography of Communication, J. J. Gumperz and D. Hymes (eds.), American Anthropologist 66 (6) pt. II: 133–136. GoŸman, Erving. 1967. Interaction Ritual: Essays in Face to Face Behavior. Garden City, New York: Doubleday. GoŸman, Erving. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press. GoŸman, Erving. 1983. “The interaction order”. American Sociological Review 48: 1–17. Goodwin, Charles. 1981. Conversational Organization: Interaction between Speakers and Hearers [Language, Thought, and Culture: Advances in the Study of Cognition]. New York: Academic Press. Goodwin, Charles. 1984. “Notes on story structure and the organization of participation”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J.M Atkinson and J. Heritage (eds.), 225–246. Cambridge: Cambridge University Press. Goodwin, Charles. 1986. “Audience diversity, participation and interpretation”. Text 6 (3): 283–316. Heritage, John and Greatbatch, David. 1991. “On the institutional character of institutional

Presenting Emanuel Schegloff

talk: The case of news interviews”. In Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis, D. Boden and D.H Zimmerman (eds.), 93–137. Berkeley: University of California Press. Lerner, Gene H. 1991. “On the syntax of sentences-in-progress”. Language in Society 20: 441–458. Lerner, Gene H. 1996. “On the ‘semi-permeable’ character of grammatical units in conversation: Conditional entry into the turn-space of another speaker”. In Interaction and Grammar [Studies in Interactional Sociolinguistics], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 238–276. Cambridge: Cambridge University Press. Local, John and Kelly, John. 1986. “Projection and ‘silences’: Notes on phonetic and conversational structure”. Human Studies 9 [Special Issue on Interaction and Language Use], G. Button, P. Drew and J. Heritage (guest eds.), 185–204. Parsons, Talcott. 1937. The Structure of Social Action. New York: McGraw-Hill. Sacks, Harvey. 1974. “An analysis of the course of a joke’s telling in conversation”. In Explorations in the Ethnography of Speaking, R. Bauman and J. Sherzer (eds.), 337–353. Cambridge: Cambridge University Press. Sacks, Harvey, SchegloŸ, Emanuel A. and JeŸerson, Gail. 1974. “A simplest systematics for the organization of turn-taking for conversation”. Language 50 (4): 696–735. SchegloŸ, Emanuel A. 1968. “Sequencing in conversational openings”. American Anthropologist 70 (6): 1075–1095. SchegloŸ, Emanuel A. 1972. “Notes on a conversational practice: Formulating place”. In Studies in Social Interaction, D. Sudnow (ed.), 75–119. New York: Free Press. SchegloŸ, Emanuel A. 1979. “The relevance of repair for syntax-for-conversation”. In Syntax and Semantics 12: Discourse and Syntax, T. Givón (ed.), 261–288. New York: Academic Press. SchegloŸ, Emanuel A. 1980. “Preliminaries to preliminaries: ‘Can I ask you a Question’”. Sociological Inquiry 50 (3–4): 104–152. SchegloŸ, Emanuel A. 1984. “On some questions and ambiguities in conversation”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 28–52. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1987a. “Analyzing single episodes of interaction: An exercise in conversation analysis”. Social Psychology Quarterly 50 (2): 101–114. SchegloŸ, Emanuel A. 1987b. “Recycled turn beginnings: A precise repair mechanism in conversation’s turn-taking organisation”. In Talk and Social Organisation, G. Button and J. R. E. Lee (eds.), 70–85. Clevedon, England: Multilingual Matters, Ltd.. SchegloŸ, Emanuel A. 1987c. “Some sources of misunderstanding in talk-in-interaction”. Linguistics 25: 201–218. SchegloŸ, Emanuel A. 1988a. “GoŸman and the analysis of conversation”. In Erving GoŸman: Exploring the Interaction Order, P. Drew and A. Wootton (eds.), 89–135. Cambridge: Polity Press. SchegloŸ, Emanuel A. 1988b. “On an actual virtual servo-mechanism for guessing bad news: A single case conjecture”. Social Problems 35 (4): 442–457. SchegloŸ, Emanuel A. 1988c. “Presequences and indirection: Applying speech act theory to

9

10

Discussing Conversation Analysis

ordinary conversation”. Journal of Pragmatics 12: 55–62. SchegloŸ, Emanuel A. 1992a. “In another context”. In Rethinking Context: Language as an Interactive Phenomenon, A. Duranti and C. Goodwin (eds.), 193–227. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1992b. “Repair after next turn: The last structurally provided defense of intersubjectivity in conversation”. American Journal of Sociology 97 (5): 1295–1345. SchegloŸ, Emanuel A. 1995. Sequence Organization. Department of Sociology, UCLA, ms.[an updated version is to be published as: A Primer in Conversation Analysis: Sequence Organization. Cambridge: Cambridge University Press.] SchegloŸ, Emanuel A. 1996a. “Conªrming allusions: Toward an empirical account of action”. American Journal of Sociology 104 (1): 161–216. SchegloŸ, Emanuel A. 1996b. “Some practices for referring to persons in talk-in interaction: A partial sketch of a systematics”. In Studies in Anaphora, B. Fox (ed.), 437–485. Amsterdam: John Benjamins. SchegloŸ, Emanuel A. 1996c. “Turn organization: One intersection of grammar and interaction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 52–133. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1997. “Third turn repair”. In Towards a Social Science of Language: Papers in Honor of William Labov. Volume 2: Social Interaction and Discourse Structures, G. R. Guy, C. Feagin, D. SchiŸrin and J. Baugh (eds.), 31–40. Amsterdam: John Benjamins. SchegloŸ, Emanuel A. 1998. “Re¶ections on studying prosody in talk-in-interaction”. Language and Speech 41 (3/4): 235–263. SchegloŸ, Emanuel A. 2000. “Overlapping talk and the organization of turn-taking for conversation”. Language in Society 29 (1): 1–63. SchegloŸ, Emanuel A., JeŸerson, Gail and Sacks, Harvey. 1977. “The preference for selfcorrection in the organization of repair in conversation”. Language 53: 361–382. SchegloŸ, Emanuel A. and Sacks, Harvey. 1973. “Opening up closings”. Semiotica 8 (4): 289–327. Searle, John R. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.

Chapter 2

On conversation analysis An interview with Emanuel A. SchegloŸ * Svetla Cmejrková and Carlo L. Prevignano

Carlo L. Prevignano: I’d like to start by asking you to tell us something about your present research programs as you conceive of them. Emanuel A. SchegloŸ: Well I have three major sorts of research and writing undertakings to which I am committed. One is a book that I am working on; a second consists of a few substantial research projects which have been under development for varying amounts of time; the third is composed of a large collection of research ‘seeds’ or ‘buds’, and I’ll have to come back to that to explain what I mean. One of the things I’m trying to work on now, at the urging of a number of colleagues, is a kind of synthetic manuscript that could provide something of an overview of CA work, at least as I understand it, and maybe could be used as a text for teaching purposes as well. It is largely based on the course sequence which I have been teaching for quite a few years at UCLA. This is no small undertaking, though I found that, until I got into it, I had no serious idea of what it would require. During a sabbatical leave a few years ago, I developed an overall plan for the work, and started writing text. I ended up with what I thought of as one chapter of this book. It came to over 250 pages on “sequence organization”, so you can imagine the scope of the book that is in the o¹ng; it almost seems as if each of the chapters might be a short book or something like that in its own right. So one of the projects is to produce a work that will be synoptic or synthetic, in the sense that it brings together the product of studies over the last twenty to thirty years in what, from my point of view, are the major topical areas that we work in. So there will be probably a chapter on turn-taking, and one on turn organization; one on action formation, and one on sequence organization; one on repair, one on word selection, and one on overall struc-

12

Discussing Conversation Analysis

tural organization of conversation.1 But there will have to be other chapters — and other kinds of chapters — as well, for example relating the work in conversation analysis to the half dozen conventionally-bounded disciplines that surround us — so, its relationship to sociology, to linguistics, to anthropology, to communications, to psychology, to philosophy, or to some elements of these disciplines. There will have to be discussions of methodological commitments, theoretical background and the contributing streams of prior work on which CA has drawn. I’ll try to have a chapter on doing a piece of work, including transcription, making observations, making collections, and so on. And there should probably be a chapter whose title — this probably could be a whole book in its own right — whose title is the frame sentence, “The trouble with conversation analysis is …”, and of course there are many ways of ªnishing that sentence and many replies to each of them. So, as you can see, this is a considerable undertaking, and I have to do it in a way that will make it accessible to students, while at the same time providing a level of sophistication for already working professionals and scientists. So that’s one ongoing project, and I don’t know how long it will take to ªnish it. OK, ªne. Another ongoing project is this. Several years ago, I had a grant from the National Science Foundation to study what we call “other-initiated repairs”. These are repairs initiated by the hearer of some utterance, who has had, or at least claims to have had, some problem in hearing or understanding it. The project had a number of analytic goals: one set of goals focussed on the variety of forms which other-initiated repair can take, what consequences these have and how we are best to understand the circumstances of their selection. A second set of analytic issues concerned the use of other-initiated repair sequences as a kind of prototype case of sequence organization. But the project was designed to speak to other themes as well, even if less centrally. One of the “troubles” which gets mentioned in the frame, “The trouble with CA is …”, is the absence of quantiªcation in CA work, and the claimed disinclination among conversation analysts to deal with large amounts of data. Now, as with many such “troubles with CA,” there are prima facie counters to the complaint. My own ªrst published work in CA (SchegloŸ 1968) dealt with some 500 instances of what I was dealing with, and subsequent work (both my own and that of many colleagues; for example, I think of Gail JeŸerson2 and John Heritage,3 among others) has also often dealt with substantial collections of instances. I’ve also now written a paper on quantiªcation and the study of interaction (SchegloŸ 1993). But the “other-initiated repair” study was also designed to work with a very large number of instances, and ended up with

Interview: On conversation analysis

over thirteen hundred of them. Each one of them, of course, requires careful analysis as an episode in its own right, so that project has taken a very long time to develop, even after the formal end of the grant some years ago. So work continues on that project as an enduring preoccupation (though there have been publications from it, for example, SchegloŸ 1997a). By the way, that project was designed to address another of the “troubles with CA . . .,” and that is the complaint that conversation analysis is thoroughly anglophone, or exclusively English in orientation. So when, several years ago, I realised that I was working with seven or eight graduate students who were natives of seven or eight diŸerent societies and cultures, native speakers of seven or eight diŸerent languages — German, Finnish, Swedish, Hebrew, Korean, Japanese, etc., — all of whom were quite far along in their training, we launched a project on other-initiated repair across languages and cultures which was very exciting but, unfortunately, was unable to attract the research support necessary to underwrite a seriously sustainable research undertaking. (Still, there have been results of that project as well, for example, Egbert 1996, 1997b; Kim 1993.) I might mention that this business about CA being exclusively anglophone is something of a historical accident. CA work has been done on materials from cultures and in languages quite diŸerent from American English — as diŸerent as Finnish (Sorjonen 1996), German (Egbert 1997a and op. cit.), Japanese (Hayashi 1999; Hayashi and Mori 1998; Hayashi, Mori and Takagi 2002; Lerner and Takagi 1999; Tanaka 1999), Korean (Kim 1999 and op. cit.; Park 1998), Mandarin (Wu 1997), Swedish (Lindström 1994), Thai (Moerman 1977, 1988), and others (to cite only language/culture complexes, and writers, with all but one of whom I have been associated, only work published in reasonably accessible places, and only a single reference to each, else there would be a great many more citations, languages, etc.). So the work is not diŸerentially suited to English, nor are there languages that we know about that resist analysis along conversation-analytic lines. My own belief is that the best way to have this work done in other languages is to have native speakers of those other languages and native members of those cultures learn how to do the analysis and then go to work on materials in the culture and in the language that they have a native control over. That is what has been done in the languages I just mentioned. So it’s just a question of getting people from other cultures to come and learn how to do the work and then go to do it (my apologies to my anthropological colleagues; in our area, it seems to me, we need to explore increasingly the virtues of developing, if I may paraphrase Virginia Woolf, “anthropologies of our own”).

13

14

Discussing Conversation Analysis

So that’s one of the larger-scale, continuing projects. I should add that the work on other-initiated repair is part of an ongoing series of studies which have appeared over the years about repair in various “positions.” This started with the overview presented in the paper with JeŸerson and Sacks on “The Preference for Self-Correction” (1977), which sketched an organization of repair in various positions around the “trouble-source” turn or “repairable.” There are papers then about “same turn” repair (such as JeŸerson 1974, and SchegloŸ 1979) and what I call “third turn repair” (SchegloŸ 1997b) and “third and fourth position repair” (SchegloŸ 1992c; some “special cases” are discussed in SchegloŸ 1991a). The work on other-initiated repair is, in eŸect, about “second position” repair. So gradually we get more and more detailed studies and get them in more languages and cultures, and so I want to participate in that, ªlling in the picture, so to speak. That’s part of this second ongoing work commitment. Another project that comes to mind, of quite a diŸerent sort, is maintaining a lively and hopefully convincing dialogue with a number of disciplines and subdisciplines which ostensibly work in the same area, or partially intersect the sort of work which my colleagues and I do. I think it’s useful to try to discriminate what we do respectively, not in a pejorative way, but in a way that makes clear where the diŸerences of opinion and commitment are, where it looks like either one or the other is going to be most productive, where they can both be working. So I’ve been writing some things in the last several years in particular directed at a ªeld that I think is more widespread and has more vitality in the United States than in Europe (though exceptions in Europe immediately spring to mind!). It’s called “communications” or “speech communications” in the United States and people in that ªeld have taken a lot of interest in conversation analysis in the last ten years or so. So I’ve been trying to work to build bridges to that ªeld and join forces with people in that ªeld who came across CA work and found it fruitful for their own interests. A lot of the work in the ªeld of Communications emerged from information theory in the ªfties and some of it from social psychology, and so there’s still some clarifying to be done about the diŸerence between “communication” as an idea and “interaction” as an idea, and the diŸerence between more traditional social psychological work on language and interaction and conversation analytic work. So that’s another project. Of course this interaction with other, more established disciplines continues with linguistics, anthropology, sociology, and so forth. Another research preoccupation for the last several years surfaced in the talk I gave at the conference which is the venue for the present interview, and

Interview: On conversation analysis

concerns the analysis of interaction with “neurologically-compromised” participants (SchegloŸ 1999a, 2002; Heeschen and SchegloŸ 1999, 2002). Even though my venture into dealing with people with neurological “problems” was actually quite accidental, there’s at least one important message I have wanted to get out, especially to people who are in the neurosciences. It mainly concerns the area that’s called now “the neurobiology of behavior.” The main point, just to say it brie¶y and informally, is this: it’s clear that it is the neuroscientists who have to describe the brain and what about the brain underlies whatever behavior they’re trying to explain. The question is, who is going to describe the behavior? Right now most of the behavior being dealt with is of a relatively simple sort: small muscular movements, sensory experience, and the like, and these things can for now be described in pretty much commonsense or vernacular terms. But even now more complicated behavior — for example, involving “rational” calculations of comparative value and risk (Damasio 1994) — is being brought under examination, and this will surely continue and expand. As it does so, a descriptive apparatus of appropriate sophistication and relevance will increasingly be needed, especially (but not exclusively) for conduct in interaction, and commonsense terminology will not do. So, even though it will be quite some time before we and they get to that point, it will be useful for neuroscientists to understand early on what’s “on the other side of the river.” If you are going to build a bridge with the brain on one side and with ordinary human behavior on the other side, it’s a good idea to know roughly how you’re going to be describing human behavior so that you can build your neuro-discipline with an eye to that. That’s really most of what I want to get out of this neurologically-oriented work I’ve been involved in. Of course, if I can help alleviate some of the misunderstanding of the folks who are beset by these problems, that would be most welcome, but theoretically the point is to open a dialogue with neuroscientists so we can see how the meshing of their concerns and ours might occur some day. (In the meantime, really outstanding work in this area is being done by my colleague Chuck Goodwin (1995, for example), much of which has yet to appear.) There are other substantial projects in much earlier stages of gestation — for example, one I call (after the title of Schutz’s 1964 essay) “Making Music Together,” for which I videotaped a string quartet’s series of rehearsals preparing a concert and then the concert itself. The initial motivating idea was to examine several distinct orders of interactivity which supply the infrastructure for making music together: the embodied interactive conduct of the playing itself, the interaction at rehearsal through which the playing is developed, and

15

16

Discussing Conversation Analysis

the interaction written into the score by the composer. At this point, I must say, I have no clear idea of what advances to conversation analysis are to be found here, but I have high hopes for my own enhanced appreciation of music and its realization, and perhaps that of others as well. In any case, the data are still being transcribed and I have no idea when I will be able to work on the project seriously. I should say, however, that most of my research “growth points” are not in such large-scale projects, nor do most of them have that sort of “on-anagenda-of-work” status. The way my research work is organized is much more under the control of the data which I encounter — in literature which I read in journals or which people send me, in the work of my students and colleagues, in the regular data sessions which we hold at UCLA or at conference venues, etc.. The way this works, brie¶y, is this: some observation made about some data prompts me to open a folder — formerly on paper, now on the computer — about the observation and the phenomenon it seems to exemplify, the practice which it appears to instantiate, etc.. As I encounter other candidate instances in other data which I happen to encounter, I add them to the folder. These folders grow by gradual accretion, then, and (in the ªrst instance) not by any systematic search. At irregular intervals, I have a look at some of these folders, and seeing what has accumulated there may prompt a spurt of writing about what seems to be going on, and that may prompt a systematic search for all the additional instances that I can ªnd in some set of data corpora. And sometimes this may lead to writing up a paper, sometimes a little one to satisfy an invitation to do a paper which must ªt into a twenty-minute slot at a convention panel, sometimes a more major oral presentation, sometimes a written product which far exceeds what can be done in even a plenary address (as for example with SchegloŸ 1996a, which followed just the trajectory described here, as is recounted in that paper). And sometimes it is the invitation to participate in a panel with a twenty-minute paper that sets oŸ a search through my directory of “collections” to ªnd something suitable in content, potential length and interest for me and for the audience — which may, after an investment of time and work, turn out to have been a misjudgement. This is not best understood by reference to the phrase in the question to which I’m responding, which asked about “your present research programs,” except insofar as one might say I have one research program — developing our understanding of how it is with humans in talk- and other-conduct-in-interaction, and how that relates to other disciplines whose activities intersect this domain. Within that research program, there are lots of “seeds” and “buds,”

Interview: On conversation analysis

growing at diŸerent rates, at diŸerent stages of development, some of which will come to full ¶ower while others do not (because they are/were wrong, because I lack the wit, because we do not yet know the things one must know ªrst before we can understand them, etc.). There is, of course, no way of conveying what is included in the array of collections, but the book, if/when I get it done, will convey something of the domain within which they fall or which they are meant to expand. Oh, there is one other very major undertaking; maybe the most major one, certainly the one with the biggest claim on my time. We have quite a vibrant community of inquiry, however we deªne it. I mean, whether you think of it in the most narrowly-circumscribed terms, as conversation analysts, or in terms which include each of the larger concentric circles that you can build up around that: linguistic anthropologists, students of dialogue, however you want to deªne it. But at the moment I’m thinking about the more narrowlydeªned group of conversation analysts. It’s quite vibrant and it has been growing in spite of a largely unfavorable academic environment, I think. Somehow we’ve survived and thrived. And it seems to me the most critical project for me right now, and for other colleagues who’ve reached relatively senior positions in the universities, is to help train a new and expanded generation of students who can then train students of their own. We need not only to produce work, but to reproduce workers. And that’s happening. But organizing and providing good training and helping people ªnd secure positions is very time-consuming. But it’s at least as important as the writing and the research itself, because there’s a natural and necessary end to that for each of us. But the way communities and disciplines develop depends entirely on the capacity to transcend an individual scholar’s life. C.L.P.: How did you come to the enterprise called conversation analysis? E.A.S.: Well, I didn’t really, because there was no such thing as “conversation analysis” to come to — at least not in the sense of what has developed over the last thirty-ªve years or so. C.L.P.: I mean, how did conversation analysis come into being? ™ Svetla Cmejrková: Was it the use of data from tape recordings that initiated your interest, or did your general idea of turn-taking come at the beginning? E.A.S.: There’s no question that without tape-recording it would not have thrived. It’s just improbable that it would have thrived as it did, and taken on

17

18

Discussing Conversation Analysis

the character that it has. On the other hand, I’m not a technological determinist. Tape-recording had already existed for at least ªfty years. In fact, social scientists had used tape recorders, including (perhaps even especially) students of interaction. For example, there was a famous social psychologist at Harvard, where I had my undergraduate education, named Robert Freed Bales, who in the 1940’s and 1950’s studied small groups in the experimental, social psychological tradition. When he started to do his work, he had graduate students coding the behavior of these small groups as it happened in real time, but it became obvious at some point in his work that this was really not adequate. So Bales then began tape-recording these experimental sessions and the research assistants would code the behavior from the tapes into the analytic categories of the research project … and then they erased the tapes and re-used them. For Bales and a great many other social psychologists (and other students of conduct-in-interaction), “the data” were the coded categories, the statistical frequency distributions in them, and the variables they represented, not the actual talk and conduct. So, the fact of actually having the tape recorder as an available technology didn’t determine anything. But it was almost certainly the case that without it we could never have had a ªeld; and we can talk a little bit more in a moment about why that’s so. So, how did I come to be doing this (kind of) work? Well, the question is, how much tape do you have. I’ve actually written a little bit about some of the story in my “Introduction” to Volume I of Sacks’ Lectures on Conversation (1992: xii–xxx) and in an introduction to a posthumous publication of an early paper of Sacks (SchegloŸ 1999b). Institutionally, the two most important converging intellectual backgrounds come from GoŸman and Garªnkel, and this sort of background is discussed in the “Introduction” to Interaction and Grammar (SchegloŸ, Ochs and Thompson, 1996: 11–16). But if you’re asking the question biographically, I’ll tell it biographically rather than institutionally. You said: “how did you come to it?”. I came to it in a way that is plausible and orderly only in retrospect. In real time, of course, it felt quite disjunctive. As an undergraduate at Harvard I had been interested in the sociology of knowledge, in Wissensoziologie, and pursued that interest under the guidance ªrst of Talcott Parsons and then of Barrington Moore, Jr.. After I wrote an Honors thesis in that area in 1957–58, it occurred to me — I’m sure I didn’t think it as clearly at the time as I can say it now, but I sensed in an inarticulate way — that the things that were most studied by the “sociology of knowledge” included everything except knowledge. That suggested that there would be a separation between what was called sociology of knowledge and the sociology

Interview: On conversation analysis

of science, and that to succeed as a sociologist of science you had better know some science and some mathematics; and I didn’t. So this was a problem, and when I got to Berkeley for graduate school, this converged with something that was just beginning as an intellectual development in the United States; perhaps in Europe as well. That was the development of what came to be called about ten to ªfteen years later the sociology of culture and/or cultural sociology. During my ªrst years of graduate school, I worked with (among others) Leo Lowenthal, a German emigré who had been one of the original members of the so-called “Frankfurt school” of critical theorists, who was among the pioneers of the sociology of literature, and also, with Reinhardt Bendix, deeply immersed in continental social theory of a somewhat diŸerent sort. I ended up writing a Masters thesis (1960) in the sociology of literary criticism (a bit of it is described in SchegloŸ 1997c). What was important about the thesis for the present story was its leading me to understand in a diŸerent way than I had previously how context could have a “bearing” on the form and substance of social life. In particular, in coming to understand the rise to predominance of a formalist style of literary criticism that (in the then canonical understanding of the social bases of ideas) ought to have been receding in in¶uence at just that time, I was led to focus not on the overall political/economic structure as the relevant “social context,” but on the much more immediate circumstances and practical exigencies of literary people — their increasing concentration in colleges and universities awash in the post-war democratization that brought to their classrooms students with little background in the sophisticated reading of great literature. The key was to be found in a more narrowly-drawn, more proximate, sense of context. So that’s more or less where I came from, academically speaking. I was trained as a classical sociologist; when I took my Ph.D. exams, I was examined in social theory, in the sociology of culture/knowledge, in social stratiªcation (or class analysis), and in studies of deviance. But in my third year of graduate school, I encountered this other graduate student named Sacks, who had come to Berkeley two years after I did (having spent several years in law school and its aftermath). We were “auditing” the same course (in the U. S. that means attending the lectures but not enrolled for credit), and he would ask what seemed to me very unusual questions. One day we ran into each other on campus, we went to have coŸee together, and we had (we both agreed subsequently) an amazing conversation; to each of us it was amazing, even though in diŸerent ways. It was from him … he had encountered Garªnkel … I won’t give you Sacks’ story; I’ve written some about that

19

20

Discussing Conversation Analysis

elsewhere (1992b, I: xii-xvii; 1999b). Brie¶y, after he had ªnished Yale Law School, he went to Cambridge (in the U. S.) for a while to try to ªgure out how the law worked, and he tried ªrst to do it at Harvard with Talcott Parsons, but gave it up after a year. But, as it happened, the year that he was in Cambridge, Garªnkel was on sabbatical leave there. So Sacks encountered Garªnkel, found his thinking serious in a way in which a lot of sociology was not serious, formed a relationship with him and became familiar with his writing. And so Sacks had in manuscript form a lot of Garªnkel’s work, and I got that from Sacks — an important new contribution to my own thinking. But of course in talking about the issues which were preoccupying him, Sacks had his own quite distinctive views, which in some ways overlapped with Garªnkel’s and in other ways were expressed quite diŸerently; altogether, an eye-opening encounter for me. What it was for Sacks would have been for him to tell, but that is no longer possible. Anyway, we became very good friends, talked together a lot, and worked together as much as we could until he was killed in 1975. But when we ªrst met, he was coming from quite a diŸerent academic commitment. He had ªnished law school, and he came to Berkeley interested in industrial relations and collective bargaining. But through a friend of Garªnkel’s, he had been alerted to GoŸman’s work, and so we went to GoŸman’s classes together. This is how the GoŸman and the Garªnkel connections got made. GoŸman ended up being my dissertation supervisor (and Sacks’ too; cf. SchegloŸ 1992b: xxiii-xxiv and note 18). With me he was a very nice combination of tolerance and discipline, in the sense that he didn’t supervise the substance of the dissertation in any serious sense, or at least did not require much change in what I had written. In part, this was because he was surprised by what I was doing. Because he knew the work I had done in my ªrst years at Berkeley, he thought of me as a theorist, as a critical theorist, as a Luftmensch of sorts, and the notion that I would actually be analyzing data was, so he once told me, completely a shock to him. I had taken a job in Ohio in order to get access to the data I hoped to work on. I would come back to Berkeley, show him what I had written, he would go oŸ and read it while I waited in his study, we would discuss it, and he would pretty much leave it alone. There was one exception. He said to me at one point that it was a responsibility of writing a dissertation to survey the literature of the ªeld in which the dissertation was being written. But, he said, there was no ªeld in the area in which I was writing; there was no literature to survey. But that did not mean that I didn’t have to survey the literature. Rather, he said, I had to survey the literature of all the

Interview: On conversation analysis

ªelds that were contiguous to what I was working on — and he speciªed some nine ªelds for me to survey the literature of. That was another six months of my life — to review the literatures of all these areas. In the end this turned out to be very valuable. At the time, of course, I resented it deeply. But, the exposure to all this literature added to my prior training a resource that was invaluable to building an academic career. In my ªrst several years at Berkeley, I had done almost every kind of sociology there was: I was a survey researcher for a year, I did historical research, political sociology, etc. etc.. It turned out to be very important, because the way I earned my way and found a place in the universities in which I taught as a junior faculty member was not because people necessarily appreciated or understood what I did — the work on conversation was pretty much an enigma to sociologists in those days (and to many sociologists these days as well). I was able to make my way in the University because I could talk sociology or philosophy or psychology or anthropology with my colleagues on their terms, to their satisfaction and so they were willing to tolerate this crazy thing that I said I was doing. And I think GoŸman’s insistence that I know all these other literatures contributed to my ability to earn my way in ways distinct from my own work. This contribution aside, I learned from GoŸman of the very possibility of studying interaction per se, and of the possibility of description as a serious disciplined undertaking. If the M. A. thesis had helped me focus on a narrower sense of social context than the earlier macro-sociological orientation to which I had been exposed from the perspectives of both the right/centrist sociology of Parsons and the left-oriented sociology of Moore and of many at Berkeley, GoŸman brought into view a much more proximate sense of social context … by several orders of magnitude. There’s one other piece to this puzzle (and we haven’t yet gotten to the conversation analysis!). And that is that, when Sacks came to Berkeley, he came in the ªrst instance to work with a sociologist named Philip Selznick, who had been a student of bureaucracy and organizations in the 1940’s and 1950’s, and had gotten interested in law, and had just founded an Institute called The Center for the Study of Law and Society. Selznick arranged to bring a number of graduate students into the Center, essentially as junior fellows. For the 1962– 63 academic year, Sacks was one of them, I was one of them and there was a third — one of a triumvirate of graduate students who used to hang out together — named David Sudnow. And so we were all at the Center that year, all working together and at this point we were all to varying degrees, as we say

21

22

Discussing Conversation Analysis

in America, ‘into’ ethnomethodology, pretty much of a Garªnkelian sort (I have written a bit about that year in SchegloŸ 1999b). There was still no conversation analysis in the sense that that term later came to have. I think in many ways Sudnow was into the Garªnkelian version of ethnomethodology the most, Sacks had a distinctive stance in that area, and I was halfway in and halfway out, and I think they recognized that. I started a dissertation in Berkeley that was concerned with a question in the sociology of law, at least I was treating it as that. The question was how a/ the society decides whether its members are responsible for their own conduct or not. What I undertook to study was the plea of “insanity” as a defense to criminal charges. In the United States (as an inheritor of British common law), if someone has been accused of a felony, one thing they can do is claim to be “not guilty by reason of insanity” — because they were insane at the time of the felonious act, they are/were not responsible for their own conduct. My plan was ªrst to study how this is dealt with legally, and then to examine how this was dealt with psychiatrically. At that time, in California, if a defendant pleaded “not guilty by reason of insanity,” two psychiatrists were appointed by the court, they interviewed the defendant in the jail and from the exchange of talk between them, they oŸered an opinion about whether this person was insane or not, and therefore responsible or not. My plan was to tape-record the interview, obtain the psychiatrist’s informal notes and formal report, as well as any testimony that might be subsequently oŸered in court, and then track the series of transformations which began with some talk in the initial interview and ended with a ªnding concerning “responsibility.” It became obvious very quickly that to do both the legal side and the psychiatric side was impossible. Since my father was a psychiatrist, and since that was where the talk was, I quickly decided to work on the psychiatric side. But there were so few cases of insanity pleas in the local courts that I simply could not do the project in the Berkeley area. So at the end of that academic year, I moved to Los Angeles because it had a vastly larger court system and I expected there to be many cases of people pleading insanity. Now, as it happens, because there was a vastly larger court system, there was an administrator who ran the court system, and he was suspicious about any sociologist poking around in “his” system, and he eventually blocked my access to the data, so after a year and a half, I had no dissertation. But in the meantime, Sacks had also moved to Los Angeles to be (with Garªnkel) a Fellow of the Center for the Study of Suicide. So we were both living in Los Angeles, and it was during that year that work of the sort now recognized as conversation analysis got started.

Interview: On conversation analysis

As it happened, the Suicide Prevention Center received telephone calls from people who were suicidal or who were with suicidal people in search of help, and the Center tape-recorded those calls and had someone transcribe them — stenographically and badly, as it turned out, but they were transcribed and that somehow made them accessible to examination in a diŸerent way. Sacks got hold of some of those tapes and it was a windfall. For years, Sacks had had the habit of attending to conversations going on around him — in cafés, at bus stops, in supermarkets, and so forth — and often jotting down bits and fragments of what he heard in a little notebook he always had with him. But the taped and transcribed calls did not have to be overheard, did not have to be jotted down on a single hearing. The material made available that way supplied the raw material for the start of this work. I count the start to have been in an exchange which I described in the introduction to Sacks’ Lectures. We were at the UCLA campus one day and he proposed to try out a conjecture he had about some data from one of the suicide calls. This was a particular call to the suicide center in which someone “didn’t hear” what the answerer at the Suicide Prevention Center had said and by the time the “repair” was accomplished (we weren’t calling it repair at the time, of course; it was just an observation), somehow the caller had managed to avoid identifying himself. Sacks connected that observation with discussions that the personnel of the Suicide Prevention Center were preoccupied with because they needed to get the names of the callers to the Center (to document their service function for their sources of ªnancial support), and they too often couldn’t get them. It seemed that if they couldn’t get the caller’s name at the beginning of the call, they couldn’t get it at all; and the easiest way of getting the caller’s name was that the answerer — the psychological volunteers who answered the phone — would give their name and they would often get the caller’s name back in return. But when the answerer on the phone said, “hello, this is so-and-so. Can I help you?”. And the other person said, “I’m sorry, I didn’t get your name,” “this is so-and-so,” “oh,” and they didn’t give their name in exchange, there was trouble. So it was at that point that Harvey said, “Do you think that could be systematic?”. C.L.P.: Would you say that it was that discovery that initiated conversation analysis? E.A.S.: Ok. So, it’s hard to say at what moment conversation analysis “started,” but if I had to pick a point, that’s the point I would pick. Harvey started from then to work intensely on the suicide calls, and then on other data he managed

23

24

Discussing Conversation Analysis

to tape-record — in particular a number of group therapy sessions with adolescents, conducted by a psychologist also a¹liated with the Suicide Prevention Center (though the group therapy sessions themselves were unrelated to the Suicide Center’s activities). For my part, I learned a few months later that my access to the data for the dissertation I had been working on for 18 months was blocked. I had no dissertation, I had a wife, I had no income, I had to ªnd something else to do. It seemed to be a disaster, but it turned out to be a fortunate accident. I found out about a research center in Ohio which had telephone calls to the police. I asked if I could get them. They said, “we don’t give the data to people who don’t work for us. However we have a job as a research associate.” They were paying $9,000 a year. This was three times as much money as I had ever earned in my life, so we went to Ohio, and there I got the data from which I wrote my dissertation. The cost of that “fortunate accident” was, however, that Harvey and I were no longer together and for the next seven years we could only work together sporadically, during holidays and for brief spells in the summer. Finally in 1972, when I got a job at UCLA, I went back to the West coast and we had just about three years of working together more sustainedly, and then he was killed. About that ªrst episode of CA, and the work that followed it, I must say that we had no idea, no sense of what lay ahead. I can only speak for myself. I had no idea what all this was going to amount to. I doubt that Harvey did and we had very diŸerent kinds of minds. It turned out that they were peculiarly complementary; we thought the same in some ways and very diŸerently in others. So maybe Harvey had an idea of what might develop from the outset; I don’t think so. Later on, of course, it became clear — at least to us — that something substantial might well be involved. There’s a place in his diaries where he writes about us as two little boys. There we are wandering around really having no idea the depths that this would go to, the extensiveness of it. We’ll never know what discipline it would have turned into had he still been alive. Anyway, I’ve gone into a lot of detail here.4 The upshot is that, intellectually, I came to conversation analysis via these way stations. It started at Harvard with an interest in the sociology of knowledge and a classical sociological canvas of largely macrosociological shape. Several things happened to that. First, recognizing the imminent divergence and separation of the sociologies of science and culture, I took the path of sociology of culture. Second, I found myself dealing with the puzzle of literary criticism in the U. S. in the period 1930–60, and ended up with a “solution” at a diŸerent level of social context than the macrosociological one with which I had started — more proximate,

Interview: On conversation analysis

more practically engaged in the thinker’s life, more “real,” even if not entirely disengaged from larger social contexts. Third, Garªnkel gave me resources that consolidated my critiques of the several sociologies I had tried — Parsonian, survey, political, etc.. Looking as I was for honest, defensible, engaging work, Garªnkel made it impossible for me to continue doing the received sociology. Fourth, GoŸman made interaction a viable topic of inquiry, in a fashion diŸerent from the social psychology I had previously been exposed to. Finally, the interest of Harvey’s mind, and our “clicking” together, provided the context for exploring what might be doable instead. ¦ Earlier, you mentioned the description of behavior, especially in connecS.C.: tion with the use of video recordings. I think it was the background to the paper you gave at the IADA congress here in Prague. So, do you have any idea of how this direction could continue in the future? You also mentioned that you had short strips of behavior and that now it is possible to study larger complexes of behavior. Could you tell us something more about that? E.A.S.: I’m not sure I’ve understood the question properly, but let me answer the one I think you’re asking and if that’s not the one you’re asking, you’ll correct me. There are two ways of extending out from a little bit of material. One is to have many instances of such bits (by “such bits” I mean bits which have the feature(s) being examined), and the other is to have larger bits. ¦ Larger bits, yes. S.C.: E.A.S.: And I think there’s an interest in both of these ways of extending the basis for analysis. But before talking a bit about each of these, let me just say as a matter of general principle that it seems to me that directions of development in this work are driven by two forces, and they are of quite unequal and asymmetrical weight, in my judgement. The most important consideration, theoretically speaking, is (and ought to be) that whatever seems to animate, to preoccupy, to shape the interaction for the participants in the interaction mandates how we do our work, and what work we have to do. One of the reasons there has been a focus for many years now on relatively small bits of conduct is because we can show that the participants are oriented to constructing the talk and other conduct in detail, and that makes that level of detail — with those facets of detail — matter for the participants, and that is the warrant for our focussing on them. It is not just that it appears “clever” or “insightful,” or that most persons — including most professional students of human conduct — are not aware of seeing these details

25

26

Discussing Conversation Analysis

(though they must be doing so if they are making their way through life in the company of others), and we can elicit an “ahah!” experience in them by describing in detail what goes on and how it gets done. It’s that we think that this level of detail in such small chunks of interaction demonstrably matters to the participants. Not that they could tell us that if we asked them; it is not a matter of self-conscious awareness, of what Giddens termed “discursive consciousness,” but that they in fact appear to construct — and “take care” to construct — their conduct in these ways, and to understand the conduct of others by reference to them. So, the primary consideration that theoretically justiªes this aspect of our work — this level of focus — is the demonstrable orientation and conduct of the participants in the interaction which we study, that is, it is grounded in, and warranted by, the data as we understand it. To the degree that we can progressively become aware of, and show the orientation by the participants to, larger stretches of the talk as organizational units for the participants in constructing and interpreting talk-in-interaction, we can ªnd methodological resources for capturing those and studying them as well. That’s a direction in which the work will develop, and has already developed to some degree (see for example, JeŸerson 1988; JeŸerson and Lee 1981; SchegloŸ 1980, 1990, 1992a, 1995b). So, that’s one of the things that drives the direction of work, and that can drive it from small to larger bits of data (but also from small bits to smaller bits). The other thing that shapes the focus and development of research is interaction with our academic colleagues. Now, that is a much more problematic matter, because often our academic colleagues are motivated by considerations other than the demonstrable relevance to the participants. In particular, they are most often motivated by the traditional or contemporary preoccupations of their discipline, by its current theoretical commitments or controversies, by the methodological paradigms currently in favor or seeking to be, by the apparent political tenor or implications of various directions of work, and the like. These often have as much or more to do with the situation of inquiry for the investigators than with the situation of interaction for the participants. Now arguably inquiry can never be free of the contexts in which it is framed and pursued, and it would be naive and pointless to pretend otherwise. But the impossibility of de-contexted inquiry is no excuse for analytical libertinism — for abandoning the eŸort to make the terms and practices of research as responsible as possible to the demonstrable features of the data, at the very least to avoid as much as possible making the terms of inquiry incompatible with the internal features of what is being studied, and not superseding them

Interview: On conversation analysis

(for example, theorizing as if every action in interaction was an independent “atomic particle,” rather than conditioned by its position in a stream of interaction). And nowhere is this more in point than with sentient actors who bring their own orientations and their own understanding of what is transpiring to the arena of action, understandings and orientations which are the critical formative input on which is based the construction by them of the next bit of the data which is being studied. I would like my own work to be motivated virtually exclusively by what is demonstrably relevant to the participants in the way they construct and understand the conduct which they build together. Obviously most people working in the social and human sciences are not as exclusively driven by those preoccupations. In interacting with them and the analytic terms of their own work, as well as their critiques of, and recommendations for, our work, we have had to talk about other things that are not demonstrably relevant to the interactants whose lives we study. And some of the interest in conversation analysis expanding the range of the units which it addresses is grounded in such considerations, in eŸorts to make CA commensurate with other undertakings in the social and human sciences on grounds other than its relevance to the materials being studied. So, what I try to do, to the degree that I can and I’m sure that this does not win us any friends, is I try to defer as long as I can answering academic colleagues who insist that we speak to these issues. So many people (this is actually something I welcome the chance to talk about) complain a lot about conversation analysis — maybe not all conversation analysts, but certainly they complain about me — that I don’t cite lots of other work, for example, that seems ostensibly to be in the same area. This is something I feel really bad about in some cases; the texts in question are in fact the product of engagement with repeatably examinable, naturally-occurring materials examined with diŸering interests in mind and arriving at diŸerent results; and too often I just can’t read all of it, and/or have failed to do so. But in a great many instances, even though work in other ªelds and styles of inquiry seems to be about the same subject matter, it is not about the same subject matter. It’s about common-sense knowledge of, or supposition about, what goes on in the empirical mundane world (as often in some variants of linguistics and philosophy), or accounts that are based on other methodologies, which, however respectable their histories are, seem to me no longer the state-of-the-art in the study of naturallyoccurring human interaction. These days, only such work as is grounded in tape (video tape where the parties are visually accessible to one another) or other repeatably (and intersubjectively) examinable media can be subjected to seri-

27

28

Discussing Conversation Analysis

ous comparative and competitive analysis. So, even though people seem to have very robust concepts and analytical tools, if they are grounded in very diŸerent kinds of materials (as for example in ethnographic observation based on one exposure in real time, yielding remembered, necessarily selective, ªeld notes which supply the basis for subsequent thinking and writing about the episode in question), from my point of view they are ordinarily not about the things that I study. Nonetheless, there is a pressure to speak to those literatures and those preoccupations, and that regularly includes a pressure to examine diŸerent — and larger — units of interaction than have been central in the past. Where the two converge, where our academic colleagues want us to deal with longer stretches of talk, for example, and that converges with a demonstrable orientation by the participants in the interaction to such larger trajectories, it is of course an inviting thing to do and some will take up that invitation. But I think it’s also important to recognise that we do not start to work on longer stretches of talk because we have pretty much exhausted the shorter ones. Frequently people get this impression. Students especially talk as if all the work on turn-taking has been done and there’s nothing left for them to do. It’s a terrible misconception! Just because there’s a lot of literature, it doesn’t mean it’s all correct. It doesn’t mean that everything has been “covered.” In part this misperception is an artifact of people learning what the problems are from the little literature that there is. Until they become more competent and autonomous investigators, they’re not in a position to see for themselves that there’s a wide open empirical domain, only little islands of which have actually been explored. This comes only from looking at data with an open mind both to the relevance and adequacy (when merited) of past work and to the relevance of that domain (e.g., turn-taking, repair, etc.) for other observable features of the data along lines not previously registered in the literature. So, I think almost certainly there will be people who try to deal with longer stretches of talk and with more instances of stretches of various sizes but that’s not because other areas have been used up. It’s just because we’ve increasingly developed the analytic tools to do that and it seems in fact relevant to the participants in the data being examined. For example, in this long chapter on sequence organization that I have drafted for the book I am writing (and also in SchegloŸ 1990), I try to show just this — that there are exceptionally long stretches of talk that can be shown to be constructed on the armature of a single underlying unit of sequence construction; that is, there can be very long sequences indeed — some of them running four, ªve, six pages of transcript and longer. That means segments as long as twelve minutes or longer, in which

Interview: On conversation analysis

all of the talk is really built on a single adjacency pair with multiple expansions, and one doesn’t really understand the coherence of that stretch of interaction without seeing that it’s based on a teeny little thing. So “teeny little” and “great big” are not really necessarily alternatives to each other. Often the way of understanding “great big” is to understand “teeny little.” I think I’ll skip talking about extending analysis from a single little extract by examining many little extracts; I have written about that in various places (inter alia, SchegloŸ 1996a: 174–81; 1997a: 501–2 et passim), and I think you were mainly interested in the possibility of expanding the size of the targets of inquiry. Did I speak to your question? ¦ Yes, yes, I think that you did. S.C.: C.L.P.: Some people comment with concern about the formalism of some of the work in conversation analysis, especially in view of the reaction against much formalism in other approaches to language. This is a common reaction, for example, to the paper on turn-taking in 1974. Can you say something about formalism and alternatives to it in this area of work? E.A.S.: One of the most puzzling reactions to the turn-taking paper for me is the claim that it is merely formalistic, concerned only with forms and rules and structure, and not with action or “meaning.” As puzzling is the extension of this characterization to other conversation analytic work, for example work on sequence organization or repair, and in some instances to conversation analytic work generally. Leaving aside the implicit theoretical and analytic antinomies which underlie the expressed concerns which might themselves merit discussion, let me instead respond by considering brie¶y why it was in point to have a “systematics” for turn-taking at all, how it related to other work at the time, and how that juxtaposition may have partially prompted the reaction to the work as “formalism.” So why was it in point to have a systematics for turn-taking? Here is one view, brie¶y put. From early on in conversation-analytic work, a great many analyses of discrete bits of talk-in-interaction seemed to prompt, and then be shaped by, observations about the construction of utterances in turns. These were analyses otherwise largely directed to what some utterance was doing or how some activity was constructed, and yet they required reference to turn-oriented practices. Sacks’ Lectures (1992) are full of such discussions, ones which involve only truncated observations about turn-taking organization — just enough to

29

30

Discussing Conversation Analysis

return to the preoccupation on whose behalf they were undertaken. I oŸer just one case in point out of many. Much of Sacks’ treatment of story-telling in conversation and its sequential organization (aside from 1992, passim, cf. Sacks 1974) is launched from two observations. First, that units like clauses and sentences can constitute possibly complete turns, on whose completion transition to a next speaker may become relevant; and, second, that virtually in the nature of the case, stories take more than one such unit to tell. This pair of observations leads to the recognition and formulation of the problem for prospective tellers of getting to tell the whole story — namely, that at the ªrst possible completion of a turn unit, or any subsequent one, a recipient may start talking along lines which frustrate a continuation of the telling. They lead as well to one solution to that problem for prospective tellers — the story-preface and the sequence which it initiates (e.g., “A funny thing happened on the way to the forum”), and the place of that sequence in the larger organization of story-telling. The focus here was story-telling in conversation, but it required an incursion into turn-taking organization to explicate important parts of its structuring. There are many such discussions in the Lectures, including ones addressed to even more narrowly circumscribed “actions.” So also in JeŸerson’s work around the same period. Those familiar with the so-called “precision-placement” paper (JeŸerson 1973) may recall how multi-faceted were the ways in which what someone was doing was contingent on where in the developing structure of a turn some bit of talk was placed. And this theme ªgured in my own early work as well — on sequence structure, on overlapping talk, on conversational openings, and the like. All these analytic exercises had, however, a scent of the ad hoc about them. They articulated only those observations about turn-taking which were prompted by, and were needed for, the exigencies of the “other” analytic project in progress, whatever it happened to be. They were, in that sense, opportunistic. They pointed to a larger domain of organization, and were parasitic on it, but always turned as quickly as possible to the project for which they were borrowing. But if that more extensive turn-taking organization was there, and if so often the elucidation of other particular practices, devices, phenomena, activities, etc. relied on facets of that turn-taking organization, it was virtually mandatory that our understanding of it be not limited to those aspects we were directed to by what were, strictly speaking, exogenous interests. At some point, turn-taking had to be examined as a domain in its own right, so as to make explicit the fund, the resource, on which we were so often drawing.

Interview: On conversation analysis

Of course, that meant that there would be (in that undertaking) no quick return to a more limited action/activity/device or practice as the topical preoccupation and analytic payoŸ. And it is that juxtaposition — between the terms on which turn-taking had previously ªgured in conversation-analytic work, and the way in which it ªgured in this, systematic, undertaking — which I think engendered in many readers of the turn-taking paper a sense of desiccated formalism, of “the clacking of ‘turns’ over their ‘possible completion points’,” as Michael Moerman (1988: xi) so graphically and disapprovingly put it several years ago. It appeared as if the situated substantive analysis of discrete actions and discrete episodes of interaction and their interactional import had been severed from the explication of the formal organization of turn-taking itself. However understandable as a narrative line, I think this is a deeply ¶awed understanding of the place of formal and systematic analysis in the larger enterprise of studies of talk-in-interaction — whether the formal analysis is of turn-taking, of sequence organization, of repair, or of any other organizational domain of practices of talk-in-interaction. In my view, such formal resources are like a reservoir of tools, materials and know-how from which particular academic analytic undertakings can draw in inquiry, because practising interactants draw on them in concertedly constructing what transpires in interaction. That is why disciplined control of these analytic resources should be part of any competent analyst’s tool-kit — not necessarily particular terminologies, only the actual phenomena and practices which such work has in the past brought to attention. Only now they have been explored and described more systematically as an ordered set of practices — a domain of organization with determinate internal shape. I can’t, however, give the most eŸective response to your question within the context of an interview. That would be to exemplify the claim I have just made about the role of formal work by examining several bits of data and their explication to show the role which the resources provided by formal analysis of the sort exempliªed by work on turn-taking or sequence organization can play in examining stretches of talk-in-interaction, including the action import of their components. For that I will have to refer interested readers to various papers (among my own, readers may ªnd particularly suitable: SchegloŸ 1987, 1995a, 1996a, 1996b, 1997a, 1997c) which I hope embody the opposite message, which is this. It is ill-considered to fault a focus of formal inquiry (like turn-taking or sequence structure or repair or vernacular poetics) simply for not taking “meaning” or “action” as its o¹cially central pre-occupation; for it may be by reference to just such formal features of the talk that action, and

31

32

Discussing Conversation Analysis

what is vernacularly termed “meaning,” are constituted and grasped in the ªrst instance. The upshot is that analytic resources which were developed as part of formally-oriented inquiry into what can be called “generic” organizations for talk-in-interaction serve as tools in explicating the action and interactional import of particular episodes of interactional conduct. But here I can give only a promissory note. The payoŸs are to be found in the papers — by various workers in the ªeld — which bring these resources to bear on other data with results which people must be ªnding worthwhile, else there wouldn’t be the interest in this ªeld which has prompted this very interview. C.L.P.: I’d like to insist a little more on historical, autobiographical matters. Not so much about Sacks; rather, would you like to say more about two other ªgures you mentioned, Garªnkel and GoŸman? I think as a younger researcher, you tried to ªnd your own answers, to put some distance between yourself and them. E.A.S.: OK. Let me talk a bit about GoŸman. I’ll try to avoid repeating some things I’ve written about GoŸman and my (and CA’s) relationship to him elsewhere (SchegloŸ, 1988). GoŸman was a shock to me. As I remarked earlier, I had been educated as an undergraduate and trained in graduate school up to that point as a classical sociologist (though I’m reminded that as an undergraduate I took a course with Roger Brown on the “psychology of language.” Now why I did that I really don’t know. So, there were apparently some concerns way back, perhaps as an oŸshoot of my interest in “knowledge”). I came to GoŸman at Sacks’ suggestion, and I reacted the way most conventionally-trained American sociologists would. A great many graduate students regularly reacted the same way, although the reservations were only occasionally articulated in class. When they were articulated, the other students would suck in their breath and wait expectantly for a nasty response, which GoŸman was quite capable of delivering. I remember only a few of these episodes, in one of which I raised an issue which would ironically later come to be directed to me. One of the things that many American sociologists would ordinarily think about GoŸman in those days (this was about 1960–61) was that it wasn’t “explanatory,” but “merely descriptive.” And I remember putting this to him in class during the only lecture course I took from/with him. He was a wonderful teacher, though not necessarily in the conventional way. He taught his own work (in his graduate courses, that is), and the term that I took the course with him he was writing the book that later appeared as Stigma (1963) and the

Interview: On conversation analysis

course was ostensibly about “deviance.” What he did was this. In the ªrst three or four weeks of the course, he gave us a very compressed introduction to studies of deviance in sociology, casting the broadest net imaginable and by no means constrained by contemporary understandings of what might be relevant, but delivered in a familiar academic format. I don’t think I ever took more extensive and detailed notes in my life than in those ªrst weeks. But then, when he started to talk about his stuŸ on stigma, we got the very characteristic GoŸmanian mode of delivery, often a simple listing of a series of “issues” posed by observations he had made or prompted by an excerpt from some book or magazine or diary etc.:“…and then there’s the issue of XYZ, as when someone does ABC.” And, being exposed for the ªrst time to that kind of work, I remember at some point saying to him, “how is this diŸerent from journalism?” A gasp went out across the room because, of course, this was one major concern — that it was “merely descriptive” when there was no obvious technical terminology, and so on. It was a concern both of the students inclined in a descriptivist direction but still without a way of formulating a rationale for such work in the face of conventional critiques, and of students with conventional commitments who were reluctant to voice their reservations in an open arena. But such anecdotes aside, much of what I wrote about the role GoŸman played in sociology in my paper on GoŸman (op. cit.) is surely true for his eŸect on me. He opened my eyes to a domain of inquiry that I just had no idea existed, even though I had been exposed to lots of social psychology as an undergraduate. I had not had much exposure as a graduate student because I was interested in “big issues,” Wissensoziologie, and so on and so forth. But the notion that there was a world here (that is, in these little scenes of interaction), that it was accessible to inquiry (I didn’t have the same concerns for precision and rigor at that point, or, rather, I understood them diŸerently), this was revelatory. It was not an “ahah!” experience; he only had to say it and I saw it. It took a while to cultivate an understanding of adequate breadth and depth and articulate it with my previous training and education, and there were lots of impediments; but its impact on me was that it just opened a whole possible domain of work that I had not understood the existence of as a ªeld of inquiry before. I don’t know that there’s a whole lot more that I can say. GoŸman supported me in various ways. There are a lot of bad stories, nasty stories, about GoŸman, but I must say that he never conducted himself in a bad way with me. There were some eccentricities, but we always got on very well and, if anything, I was overly paranoid about him. I mean, he really did once give me cause for that, after my degree, when I was already in a very good job, though

33

34

Discussing Conversation Analysis

still a junior faculty member. His paper “Replies and Responses” (1976) was a kind of massive attack on CA, and we were after all still “kids,” but in some ways it just showed his respect (in fact, a great deal of his writing after Frame Analysis (1974) — and even in that book — was addressed in some fashion to CA work, concerns, etc.). I regret that I didn’t respond to it while he was still alive. In any case, he was a really smart man and an extraordinarily careful and perspicuous observer of the social world, however refracted through his own prism. There’s no question: he had a distinctive vision but I think that was important. I don’t think a person with a conventional vision could have come to do what he did. In the end, GoŸman provided a point of departure for the direction our work took, and our work seemed increasingly in tension with his. Much of that was a function of generations and of technology. I am told that in much of his teaching and occasional lecturing after he moved from Berkeley to Pennsylvania, he conceded that working from tape had become the state-of-the-art way of working, though he never committed himself to that view in print. He tried to work with such materials in some of his writing (for example, his paper “Radio Talk” in GoŸman, 1981: 197–327), and I have been told by former students at Pennsylvania that he taught seminars based on videotape there, but the fundamental anchoring of his work was in extensive observations of the world in single exposures in real time and in his collections of fragments from written material, ranging from ethnography to confessionals, from ªction to memoirs, from training manuals to case reports. Our work started from the domain he had shown to be there, but was built on diŸerent foundations. With Garªnkel the story was diŸerent. I think he would probably be most unhappy at the form his initial in¶uence on me took, because he often speciªcally denies intending any critical stance toward conventional sociology. It is for him a form of practical theorizing, to be studied and understood together with other embodiments of practical theorizing, not to be criticized as a competing way of working. (Sometimes I think this was only ironic; sometimes I think it was meant seriously when formulated, though at other moments one could have heard Garªnkel speaking of conventional sociology and other social sciences in an unmistakably derisive idiom.) But a good part of Garªnkel’s initial impact on me was what I took to be — however naively or mistakenly — its critical import. I had migrated from one kind of sociology to another trying to ªnd, as I thought of it at the time, “honest work.” What I mean by that is that I would encounter some kind of sociology — some substantive sub-ªeld or some

Interview: On conversation analysis

methodological stance — and work at it for a while. There would then be colleagues who would ask challenging questions about it, or I would read a critical literature that found trouble with that way of working, and those critiques would seem to me compelling, and I found that I just couldn’t do that (kind of) work any more. It wasn’t that I was determined to do the perfect inquiry; I just couldn’t do the work if I already felt that I knew it — the genre — was wrong, I couldn’t practise doing it, I just couldn’t — whatever I thought the particular problem with that genre was. The ªrst major impact that Garªnkel’s work had on me was of a critical sort, even though he forever denies that the point of ethnomethodological studies is some sort of ironic critique of sociology. Nonetheless, however wrongheaded it was, it allowed me to consolidate all the separate critiques I had of all the separate kinds of sociology I had tried to learn how to do. Whether correctly or incorrectly, all of a sudden I could see, for example, in the relationship between “indexical and objective indicators,” or in the studies of “good organisational reasons for bad organisational records,” or in the coding study, all those themes of Garªnkel’s work allowed me to subsume under a single overarching “critique” what had previously been a whole series of separately specialized critiques. And that was a tremendous burden lifted oŸ my shoulders because I didn’t have to carry all of that critical baggage around with me. I could see a way of consolidating and having some sort of homogeneous grasp of the ªeld (and, indeed, much more than the ªeld), and there were all sorts of other sociologies I didn’t have to “subject myself to,” only then to learn what was wrong with them. So it allowed me to see almost in prospect that other areas were going to be just like things I already knew. What wasn’t clear to me, and never became clear to me, was what to put in its place within Garªnkel’s way of working. My own mind does not work well in the phenomenological idiom. To this day, every time I have taken responsibility for teaching some of Garªnkel’s work, I have had to read Studies in Ethnomethodology (1967) again, and each time it has been “news” to me all over again. I would stop at various places after having read something and think, “Gee that’s clever,” and then I would vaguely remember every other time I had read that essay, and when I got to that point I would say “oh, that’s clever,” and I had said it again this time. That’s just not the natural idiom of my own mind, and though Garªnkel was clearly something quite distinct from a phenomenologist, it’s clear also that his work and his world view are very much cast in that idiom. And in many ways his undertaking was deployed as fundamentally a “critical discipline.” For me, it fairly quickly became not satisfying; I

35

36

Discussing Conversation Analysis

just couldn’t ªnd the a¹rmative program there. In any case, whether the eŸect Garªnkel’s work had on me was intended on Garªnkel’s part or not, whether based in misunderstanding or misinterpretation on my part or not, for me his writing worked to consolidate a critical stance toward a great deal of conventional sociology, and to alert me to some issues which have remained continuing analytic constraints — ones which I got nowhere more forcefully than from Garªnkel’s writing (though not necessarily in the form in which he expressed it or in ways he would any longer accept). For example, I don’t think of it as “commonsense knowledge,” but as “vernacular knowledge;” the relationship between vernacular knowledge and technical inquiry is something that certainly was not invented as a topic by Garªnkel, but Garªnkel introduced it into contemporary sociology in a form which, at least at the time that it intersected my life, was much more compelling than other ways in which I had encountered it before. I had gotten it obviously from Parsons with whom I had had a series of private reading courses at Harvard years earlier. It didn’t make any big impression on me with Parsons. So, I learned a lot from Garªnkel, and spent a lot of time explaining and defending his work in various sociological venues. But I think Garªnkel sensed from early on — I’ve never actually asked him this — that I was not as much taken with ethnomethodology as Sacks and Sudnow were. Garªnkel held a number of conferences on ethnomethodology in Los Angeles at the time I was still in Berkeley. I wasn’t invited to the ªrst couple of them. When I ªnally did go to one, the experience was exhilarating. Among Garªnkel’s many distinctive characteristics was (and is) an amazing capacity to listen in a perspicuous way to what others say, and to hear in what they say something they never dreamed of saying themselves, and to appreciate it and applaud it. He did that for me, and it was an extremely heady experience. My presentation was given pride of place. It was a Saturday evening in his living room, a group of some forty people crowded around, the tape recorder on, and I had this paper which I’d published in the Berkeley Journal of Sociology (1963). It was on psychiatric theorising as part of that “insanity” project that I had been doing. With Garªnkel leading it, the reception was, of course, intoxicating. Nonetheless, I felt I wasn’t entirely of that group, but its eŸects on me were there, and Harold and I have always had a relationship alternating — and combining — tension and support. He is, after all, primarily responsible for my being at UCLA. I’ve made contributions of my own, I think and I hope, to his side of the ledger. Three people — GoŸman, Garªnkel, and Sacks — made a critical diŸerence to my scholarly development. And I think in each instance there has been

Interview: On conversation analysis

some mutuality of eŸect. The fact of the matter is that GoŸman’s last work in Forms of Talk was largely a dialogue with conversation analysis, and CA has I think been of consequence for Garªnkel. And Harvey and I, of course, went in an entirely diŸerent direction once we encountered each other, though most of what Harvey got from me went to the grave with him because it wasn’t written. C.L.P.: Conversation analysis is concerned with the discovery and analysis of verbal procedures in human interaction. There are conversational procedures as object of study, but also procedures of discovery and analysis of conversational procedures. What are your ideas about the cognitivist interpretation of human cognition and interaction as corresponding to some kind of procedure? And what do you think about the idea of social cognition? E.A.S.: Well, it seems to me that there are two questions in search of responses here, if I understand the question properly. One concerns the relationship between conversation analysis and more cognitively-oriented undertakings, perhaps even cognitive science. The other concerns the relationship between the practices of ordinary conduct in mundane settings of social life and the practices of our inquiry into those practices. Both questions present “tall orders,” so I’ll try to be brief at the risk of being unsatisfying. Although conversation analysis was once taken to be part of the nascent larger development called “cognitive science,” there are contrasting presuppositions which underlie them and render such a “merger” problematic. I can only sketch a few. In general, a cognitivist stance begins with the broad cultural presuppositions of the so-called Judeo-Christian stream of European culture. That cultural tradition (and cognitivist and other “psychologically-oriented” disciplines emerging from it) takes the single, “minded,” embodied individual person as the basic, enduring, integrally-organized reality to be studied. The setting such a “person” is virtually always in, the complement of other persons in that setting, etc. are taken to be contingent, transient, ephemeral contextual properties. Settings are treated, in eŸect, as composed of an aggregate of such “individual person realities,” perhaps adding something (something “social,” which is thus treated as external and subsequent to the constitutive reality of the individuals) to the given features, capacities, resources, predilections, etc. of those individual persons, rather than shaping or even engendering those features, capacities, resources, and predilections, and therefore, in a sense, as constituting the eŸective actors/participants in those settings. So when a little group or conversational cluster breaks up — like the one composing the present

37

38

Discussing Conversation Analysis

interview occasion — each of the embodied named individuals who composed it will be taken to continue to exist, even if not accessible to perception, but the group that has (as we say) “dissolved” is taken not to continue to exist. The episodic setting, the little interaction system, as GoŸman might have called it, is taken not to have perduring reality. But, as GoŸman (1967: 3) conveyed in his telling contrast between “men and their moments” on the one hand and “moments and their men” on the other, there is an alternative way of conceiving matters. We can understand “the situation” as the reality, and the individuals who happen to compose the situation on any particular occasion as what is transient. A scholar of classical Greece named John Jones some years ago (1962) wrote a book called On Aristotle and Greek Tragedy, in which he argued that it is mistaken, or simply a subsequent cultural imposition, to treat the Oedipus myth as involving a tragic hero. That grows out of a tacit ontology in the Judeo-Christian stream of western culture that it is the single, “minded” and embodied individual that is the locus of social reality — here realized in the notion that the person named Oedipus is the locus of the play’s action and import, and its “tragic hero.” The alternative view is that there are certain sorts of recurrent situation that are the locus of tragedy (as well as of other “narratives,” as the current parlance would have it), and the point of putting Oedipus into one such situation is to make the point that if a king, who is the son of a king, could be battered by the world by being caught up in this situation, how much more so is it the case for “lesser” individuals. But it is the situation which is the relevant reality, the eŸective source of Oedipus’ — and any person’s — story and fate. The individuals who are caught up in it at any given moment are what is transient. Well, when you juxtapose these two ways of seeing what’s fundamental and what’s transient and relatively epiphenomenal, and especially when you see that the second view is clearly the minority view in western culture and in the contemporary academic scene, it becomes increasingly important for those who have found a way to study matters, human and social, in the second way to insist on studying them that way. Fundamentally, cognitive science is a thoroughly psychological enterprise, and saying that it’s a thoroughly psychological enterprise is to say that it falls in step with, rather than resisting or giving us any leverage on, the otherwise inbuilt cultural presuppositions that a great many of us share as members of western culture. I should add one further point, though I can’t go into it in detail. What I have suggested above about the focus on the single individual gets carried further in cognitively- and psychologically-oriented inquiry by a focus on the

Interview: On conversation analysis

single sentence, the single act or action, etc. as the target of study and the fundamental locus of reality. We see this not only in contemporary linguistics, but in enterprises like speech act theory. The very conception of action having its origins in the acting individual’s “intention” treats the single action as the unit to be analyzed, and the single individual as the proper locus of its analysis. Thought about in the abstract, this may sound unexceptional to academicians trained in a scientiªc culture grounded in the dominant strand of western culture. But if you look not at imagined actions but at actual ones, it becomes not only unviable, but almost peculiar. And here again the availability of taperecorded, repeatably inspectable material, is deeply consequential. If one is committed to understanding actual actions (by which I mean ones which actually occurred in real time), it is virtually impossible to detach them from their context for isolated analysis with a straight face. And once called to attention, it is di¹cult to understand their source as being in an “intention” rather than in the immediately preceding course of action to which the act being examined is a response and to which it is built to address itself. So an approach to work that starts from the individual as the real — whether the individual person, or action, or utterance, or sentence — which treats that individual entity as designed for integrity as a free-standing object with its context as an extrinsic environment, can hardly avoid being characterized by atomism, atemporality, ahistoricism, and asociality. And the study of interaction and of humans in it would do well to avoid such a path. Such a view is not incompatible in principle with an interest in studying cognitive matters, but it places cognitive issues, processes, etc. within the framework of a world which is social and interactional from the outset, within which cognition is to be understood not necessarily by reference to the individual cut oŸ from a world around, but by reference to an individual engendered and constituted by the world around in the ªrst instance. A “cognitivism” or “cognitive science” along such lines, and responsible to details of naturallyoccurring interaction in ordinary-for-the-participants settings, would be of considerable potential interest. ¦ I think that not all branches of philosophy of mind assume the “single S.C.: individual” as their basis and ground. There are also disciplines of mind that are based on the assumption that the pair of persons is primary, not one individual, but two people interacting or trying to understand each other. E.A.S.: Give me a name or two.

39

40

Discussing Conversation Analysis

¦ Don’t know. Habermas, for example. S.C.: E.A.S.: Ahh, I think not. Habermas, it seems to me, made a fatal misstep very early on (e.g., Habermas 1970) when he incorporated what is essentially Searlean (or Austinian-Searlean) speech act theory (in the key respects that matter here the diŸerences are of little moment).5 That’s one of the problems, because I think it is very di¹cult to recover a socially or historically sensitive view of action once you’ve started that way. There are other related problems with the tack which Habermas takes which make it an unpromising alternative, at least for work which means to be “empirically capable.” It’s critical to the larger program of Habermas’ studies to have a pre-analytic conception of rational discourse as the model, the critical leverage, with which to critique the “distortions” introduced into actual communicative action by malformations of social structure. But this presupposes that we have or can develop in full measure what Giddens calls “discursive consciousness” for our conduct in interaction, if we are to have a pre-analytic, pre-empirical grasp of its rational character and possibilities. In a recent paper (SchegloŸ 1996a) I describe what I think is a “new” social action, that is, an action that I did not know existed (and, as far as I can make out, that other people didn’t know existed either). It bodes ill for the possibility of a pre-empirical, preanalytic pragmatics; it seems to entail that you have to have an empirical grounding to understand what this form of human communicative action is all about, and if that’s so, you can’t have it pre-empirically, pre-analytically and use that as the critical leverage for a vision of rational communication. ¦ In European philosophy, perhaps Buber, for example, or Levinas in S.C.: French philosophy. E.A.S.: Yes, I read Buber a long time ago, but I haven’t for many years, and for that reason perhaps, I never approached him seriously as a ªgure in the investigative enterprise we are discussing, but rather as a ªgure in theological discourse. That’s an interesting suggestion. I will go back and look at Buber again. Levinas, I’m afraid I don’t know and cannot comment on. But I’m happy to hear that there are others working along such less individualistic and atomistic lines. Recall though that it was not philosophy that I had my reservations about, it was psychology, and those forms of philosophy which adopt an empirical-sounding diction without having done the sort of work which would warrant it. But if there is hope to be had on any front, I’m perfectly happy to have it.

Interview: On conversation analysis

The other question lurking in this portion of our discussion may have been more in order under a more cognitivist understanding of conversation analysis; it would then have been the re¶exive question: “If you are studying processes or practices of knowing, then what do you have to say about your own practices of knowing?” But there may be something of interest to be said here nonetheless, and that is how the practices of understanding and describing conduct in academic/professional inquiry relate to the indigenous practices of understanding and describing conduct in ordinary interaction itself. Discussion in this area could get very complicated indeed, so let me limit it in the following way. Sacks pointed out (Sacks 1972b, 1992 (I): 236–266) about ordinary or vernacular or common sense description that it is recognizable as such without inspection of the circumstances or objects being described. In that discussion, Sacks sketched the importance of this feature and the economies which it aŸords the conduct of ordinary aŸairs. There are domains, however, in which the practices of describing and taking up descriptions are diŸerent — in which descriptions are in the ªrst instance to be juxtaposed with what is ostensibly or purportedly being described, and description grounded in these practices operates diŸerently — it does not aŸord the same economies, but delivers outcomes potentially diŸerent in kind from those of ordinary description. What may appear a merely stubborn insistence in conversation analysis in grounding all work in the details of actual data, ideally with the recorded version present, but with at least the transcript if that is not possible, has a grounding not only in our past experience with the productivity of proceeding in this way, but in the commitment to a diŸerent enterprise than the practices and forms of description that characterize mundane description. Papers take the form they do to maximize the opportunity for readers to immediately juxtapose every bit of description with the data of which it claims to be a description. One basis for reservations about other forms of inquiry which appear to intersect on the same subject matter but use diŸerent research methods is grounded here. To take but one example: in ethnographic work, the investigator gets to observe occurrences once in real time. Even the best ethnographer or ethnographic team will register only “the take” possible under this constraint. Under ideal circumstances, ªeld notes are made as soon as possible, but are grounded in the ethnographer’s memory of the events that were seen. The text of the ethnographic report draws together those notes and memories into an (ideally) coherent account of the object of description. The upshot is that the reader must essentially take up the description with no access

41

42

Discussing Conversation Analysis

at all to the object of description itself, but at best with access to an account reconstructed from notes grounded in memories of the sorts of observation that can be made with one exposure in real time (and under the constraints of the operative “social and cultural organization of seeing” in that context). The practices of description and description-uptake underlying these two approaches to a “same subject matter” are su¹ciently diŸerent to call into question whether this is really the same enterprise and the same subject matter taken up by two diŸerent approaches or methods, or whether two quite diŸerent sorts of undertaking are involved. This may not be quite what you had in mind by asking about the bearing of our research targets on our own research practices (if you were indeed asking about that), but it is for me a compelling outcome of re¶ection about just that issue. C.L.P. : Before coming to your research on pragmatic deªcits, I have just one further question: what about the “interaction order” according to SchegloŸ? E.A.S.: Oh, GoŸman’s article? What about GoŸman’s article? C.L.P.: How do you situate yourself in relation to the idea of the “interaction order” and to that article, now, ªfteen years after its publication? E.A.S.: I haven’t read it for a long time. There is a lot of vitality to the idea of the interaction order. To the degree that GoŸman is one of the main feeder streams to the sort of work that we do now, it’s in part his calling attention to the existence of that domain of organization that is responsible. But if I have to imagine what reservation(s) I might have, if I sat down and read it right now, it would be whether one would still or would want to retain the severity of autonomy and disjunction which I recall when ªrst reading that paper. And the reservation would extend in two directions. Given GoŸman’s own earlier writing, it was not surprising that he would have taken this stance in much the way that he did in “The Interaction Order.” I don’t know if you’re familiar with a paper of his that was published in 1961. It was in a little book published by the Bobbs Merrill Advanced Studies in Sociology; the book was called Encounters and the ªrst paper in it was called “Fun in Games.” In that paper he coined the term, “the membrane around the interaction” (or something like that), and one of the things he was concerned with there we would now speak of, in contemporary parlance, as the diŸerence between discourse identities and other identities. One of his points was that the membrane that surrounds an interaction — that marks how and where it is

Interview: On conversation analysis

bounded oŸ from the surrounding setting and world — can serve to ªlter out a lot of the things that are in some sense “objectively” true about the individuals who compose the interaction. Many — perhaps most — of the identities, memberships, debilities and strengths, achievements and stigmata which are in fact “the case” about a person do not permeate the membrane; they do not necessarily aŸect the conduct that goes on within it; they do not survive the test of relevance. (Of course, he is overtly concerned with “fun in games”, but the “game” he is ultimately interested in is all of interaction, and much of the text is about that. That’s a common strategy for GoŸman; he begins ostensibly talking about a very particular and limited phenomenon, but by the time he is done, it is everyone’s contingent reality. Just think about “face,” “demeanor,” “stigma,” etc..) One of the points I think is most important — whether it is recognizable already in GoŸman’s take on “the interactional membrane” and what gets through it or not — is that this is not for analysts to decide. This issue has ªgured centrally in subsequent conversation-analytic thinking about various identities or categories of membership of persons in society (as for example in Sacks 1972a, 1972b; and see also SchegloŸ 1991b, 1997c). As with everything else, it is the participants who embody in their conduct which features of their co-participants they are oriented to as relevant and which not (though, to be sure, eŸorts at concealment and camou¶age can be at work as well); that is a contingent matter. And features of co-participants which are “macrostructural” in the terms of social science theorizing are subject to that contingency as well. This seems to me, at the very least, the appropriate default position from which analysis must begin. Someone could undertake to show that some identities — for example, gender identity, to cite one which is often urged in this regard — are “omni-relevant,” and are never fully ªltered into irrelevance by the “interactional membrane” (though what a fully satisfactory demonstration would look like is not entirely clear). This then could be one possible reservation about GoŸman’s account of the interaction order — that one cannot theoretically legislate out of existence the prerogatives of participants in interaction to treat as relevant features of their co-participants ones which are macrosociological in character, which would compromise the “separation of orders” which many take GoŸman to have asserted. But it is quite possible that if I re-read GoŸman’s paper, I might well ªnd that this problem does not actually arise. In any case, my only objection to the conventionally claimed interfaces of the so-called micro-social with macrosociology is the insistence on the inescapable and often exclusive relevance of, to use the terms that are

43

44

Discussing Conversation Analysis

most powerful in contemporary American sociology, the intersection of race, class, and gender. My objection is only to people’s insisting that the only exclusive, centrally important thing is whether someone is a woman or a man, this or that ethnicity, and this or that social class. But that the co-participants can treat those on any given occasion, or some moment in it, as relevant (and potentially consequential) seems to me to be beyond question. As with everything else, it seems to me we have to put our analysis at the disposal of what the participants are actually doing. Now, that may be compatible with the autonomy of the interaction order or it may not be. So that would be one possible reservation. And that is, so to speak, a reservation about the ªltering down of the more macrostructural into the more “microstructural.” The other reservation is going in the other direction, conventionally speaking. Here, I suppose, it’s more a diŸuse scepticism than it is a determinate reservation. This is I suppose a leftover for me of the Garªnkelian “critique.” As a member of the society, I share in the vernacular culture, and mine is American and sociological and upper middle-class and Jewish and all the other sorts of things that frame one’s “social location” in vernacular terms. As a member of society, I perfectly well understand about social classes and all the rest of a moderately sophisticated citizenship; but the fact that I understand and see the world — or can see the world — in those terms as a member of a society is not the same as qualifying all those ways of seeing it technically, let alone subscribing to it and underwriting it as part of one’s technical apparatus for understanding the world. In fact, it’s just the opposite. The more they recommend themselves to my vernacular understanding, the more suspect they ought to be for me as part of my technical apparatus. The common or vernacular culture is, after all, a sort of “propaganda arm” of the society, serving to undergird the cultural component of the more or less smooth functioning of the society itself, not to advance or enhance a rigorous understanding of society. And so there’s a question here, because what GoŸman in eŸect does is, by implication, to ratify all of macrosociology as not “his business,” but he appears to stipulate that there surely are all these economic structures and political structures and bureaucratic structures and so on and so forth. I don’t know whether he did that as a vernacular member of the society or as a technical sociologist. He did it in his Presidential address to the American Sociological Association. He surely was aware of the fact that he was at risk of being understood to be saying it as a sociologist. And it seems to me that there are enough reasons to be uneasy about that. It’s not that there aren’t ample things

Interview: On conversation analysis

that you could point to within the domain of rigorous inquiry to make them serious things to entertain, but entertaining something seriously and seeing how actually to work it up as a robust part of one’s technical understanding of the organisation of social life are two very diŸerent things. And for me, as I say, the fact is that what we know as part of our vernacular knowledge is part of having the society work properly. And that’s very diŸerent from how to have a discipline of the society work properly. Wherever those things look like they’re bumping into each other, I think — especially because we can be treated to be experts about the matter — we have to be specially careful. Again, if I re-read GoŸman’s “Interaction Order,” I might ªnd he was way ahead of me on this and that he’s anticipated all of these concerns; but that’s certainly not the message that has ªltered down to us in the ªeld about the interaction order. C.L.P.: Can you say something now on your work on pragmatic disorders, and give some suggestions for analysts in that domain? E.A.S.: Well, ªrst I have to say that I obviously have no competence to help people therapeutically; we are talking here only about helping people who would like to do research work in this area along conversation-analytic lines to situate themselves better for doing that with some success. I think that the key thing — and there are historical grounds for saying this — the key thing is for people to get themselves properly trained in analyzing materials of talk and interaction of whatever sort. In the past, when people have used some particular, predeªned, practical interest to inform or constrain general training and the general course of research, there have been problematic outcomes. I’m going to just repeat, if I may, some things I have written about this (SchegloŸ 1991b: 66–7) because they may be of interest in this connection. For a very long time, you could not get a societal “license” to study ordinary interaction closely — either in the educational or in the research sense of “study.” These were things that presumably we all knew because that’s what the meaning of “commonsense” was; why would you waste your time, why should the university waste its money, in supporting a degree of this sort, or research of this sort. So the only way such inquiry on “ordinary, everyday behaviour” was done was typically under one of two conditions. Either the participants in the interaction to be studied were formulated as “defective” in some way — and so you could study people who had had strokes or who were mentally retarded or were schizophrenic and so on and so forth; or the activity to be studied was so strategically positioned in social life that the activity could

45

46

Discussing Conversation Analysis

be made more “proªtable” in some sense to the society if we understood how it worked better — so bargaining, negotiation, con¶ict resolution, “salesmanship” and the like could be examined closely. You couldn’t study ordinary interaction just in order to study ordinary interaction. A consequence of that situation of inquiry historically has been that people would study a particular corner of the world under the warrant of one of these “licenses,” and, whatever the auspices were under which the inquiry was conducted, the results were taken to be speciªc to that object of inquiry. So when people studied schizophrenics, that was an acceptable thing to study because of the promise of therapeutic payoŸs. When the focus turned to thought and language (I am thinking here of work in the ªrst half or so of this century), much of what was found out was taken to be characteristic of “schizophrenic thought and language.” In the last thirty-ªve years or so, when some of us were able to make a little space for studying ordinary talk in interaction, mostly without substantial research support but with our own funds and on our own time, it would often enough turn out that things that had been ªgured to be specially characteristic of “schizophrenics” or “retarded people” or other such “special” categories of person were actually quite common in ordinary conversation. That is not to say there is no diŸerence between schizophrenics and ordinary people. It is to say that, if you have not studied ordinary people (and been trained properly to study ordinary talk in interaction), you have no way of ªguring out what is speciªc to schizophrenics and what is the case about conversation per se, except that, like so much else that is “common,” it generally falls below the threshold of ordinary observability. So, somebody who’s interested in doing research on pragmatic deªcits and neurologically-impaired people has to begin, it seems to me, with understanding how ordinary interaction is organized among people without respect to neurological status. What I am saying seems to me obvious enough a point: anybody who wants to practise something in particular in a domain (playing Bach toccatas, doing cardiac surgery, training retrievers, writing sonnets, etc.) needs ªrst to become adept at practise in that domain in general (playing the piano, doing surgery, training dogs, writing poetry). And if you want to study some particular thing, you need to know how to do research in that domain, for example, how to analyze talk and interaction and body behavior and so on and so forth among humans. So, the ªrst thing is to learn how to do good work and bring it to bear on any data. If somebody was contemplating working in this area of the pragmatics of neurologically-impacted people, and was in search of additional components

Interview: On conversation analysis

of distinctively relevant training, it might be interesting to have some people who were both competent and sensitive analysts of talk and interaction and who knew something about neurology as well. Surely there are some things that can get done when all the relevant input and knowledge are controlled by the same person and not “distributed” amongst several individuals; there are things you can conceive of if various facts that have never been brought together are in the same mind. But that’s a very big order for someone who’s undertaking to be trained. In most (if not all) respects, however, concentrating on pragmatic deªcits or neurologically-specialized data is just one instance among many of focussing one’s work on some sub-domain of data, whether deªned by technical criteria or by commonsense categories of the society at large. So the next suggestion would be that, for the health, vitality, quality of work, and continuing growth both of individual researchers and of the ªeld as a whole, students in this area (and I use “students” here in its most general sense, and not referring to a stage of life and career) need to continuously play back and forth between the specialized domain that they study, whatever it is, and the ordinary run of human interaction. Many workers in the ªeld, both senior and junior, have cultivated specialized domains of work which articulate well with the current organization of universities — with traditional disciplines and recognized subªelds within them (for example, institutional sectors such as medicine, law, education, communications within sociology, or subªelds such as pragmatics, discourse and the like within linguistics) — and other organizational centers of research in order to maximise their own individual chances in the employment markets as well as the institutionalization of this type of research in the society. But from my point of view, it is absolutely critical — and I think my colleagues at UCLA John Heritage and Steve Clayman agree with this — that students should ªrst start getting trained on ordinary conversation, not on particular institutions. But even more important is that, even if they have developed a specialized knowledge in some particular institutional sector, they nonetheless keep working on two fronts — both in that specialist institutional sector and on ordinary conversation. There are several reasons it seems to me for this. The ªrst is that no institutional domain is totally segregated from general social life. Everybody knows that when they go to see the doctor or mechanic or salesperson, the talk slips in and out of the institutional framework. Not all the talk in a work setting is of that work setting, and this contingent character of conduct is not speciªc to work settings. If you lock yourself entirely into the institutional domain being studied and assume that once the episode has

47

48

Discussing Conversation Analysis

started as a professional interaction it will be a professional interaction, you simply are not in a position to understand what’s really going on. Not only do you become un-alert about the non-professional aspects and potentials of the talk; you disattend how the very professional character of what is professional is produced as professional. A second reason is that, because the society is prepared to support the work in particular professional domains deªned by its vernacular culture, there is a serious danger that we would begin no longer to get general analytic tools being developed because we would less and less be having general interactional practices being studied. We would keep learning more and more about doctors and patients, teachers and students, cooperative work groups, news interviews and so on, but we would not be developing our understanding of the generic practices of talk-in-interaction and the tools for analyzing them. All of that could become really stagnant. We could have the pool of analytic resources that we developed in the ªrst ªfteen or twenty years which could just stop growing. But it has to grow, and the same people have to help it grow as are making the separate more specialized substantive areas grow. So, what I would say about and to people who are interested in neurologically-compromised participants in interaction isn’t unique to them. Here, as in any place else where there’s a knowledge-based skill, whether it’s the practise of medicine or of music, you have to have your craft. You have to know how to do the basic work. In this area: –

–

–

you need to know how to collect data and have recurrent experience collecting it yourself, because you often enough have to be on the scene where it was collected to know what that scene was like; you need to be transcribing it yourself, all the time, and not just hiring others to do it, because then you don’t know what the data sound like. You don’t give yourself the best possible opportunity to hear something entirely new. It’s you, after all, who will over time come to have ten or ªfteen or twenty years of experience, who are now in a position, if only you listened to the raw data, to hear things you could never have heard before you had that experience. Your assistant almost certainly doesn’t yet have that experience. You have to make the basic observations and ground them ªrmly in the observable details of the material. That’s the basic craft. If you don’t do that, however fancy the written papers look, they’ll be based on water and eventually somebody will come along and actually look at the material, and the ungrounded, clever writing will all collapse.

Interview: On conversation analysis

So, people have to have their craft under control and they have to keep it updated. Then it’s a question of intuition, skill, learning, and luck. There’s always luck, right? You’re lucky you get the right data, you happen to be sharp on the day that something comes up, so you see something which on a day when you were dull you would never have seen. But that, of course, we all live with, no matter what we do. So as for the “luck” part, there’s nothing to be done. The other parts you can do something about. C.L.P.: I don’t know if you would like to add something concerning future directions for research, I mean some suggestions for younger researchers … E.A.S.: I suppose only this: this kind of work is right for some people and it’s not right for other people, and I have given up trying to ªgure out in advance which is which. I wish I could, because it would save them and me a lot of time and pain if I were to spot it early enough. There are some people who decide wrongly that it’s not for them, and they decide that because they’ve always been ‘A’ students and I don’t give ‘A’s just because people are attentive or loyal. If they do good work, they get ‘A’s and if they’re not doing good work yet, they get ‘B’s, and some students ªgure that the latter grades are telling them they’re not wanted here. That isn’t the message; the message is that they haven’t “got it” yet — not that they are incapable of getting it. So, I would urge students who feel themselves drawn to this kind of work, who feel that it gives them a kind of insight and access to interaction and culture, or think that it might, to stick with it for a while. In my experience, it requires of most students — and most colleagues, who have come to terms with it after a previous professional training — a tremendous wrench, a tremendous transformation in the way they see the world. I forget about that from time to time. The students remind me of this almost painful reorientation, and other students have to know this; that this seems often to be quite diŸerent from simply taking on a new academic subject and absorbing it like one has absorbed the previous ones. So I think the thing to say to students is, ªrst of all it’s a long, hard road. If they are not prepared for this, if they must get big payoŸs early on, this may not be the way to go. They have to take stock and decide whether they can stay the long course. Some people ªnd the exposure transformative once they get into it (I hope this will not be taken in the wrong way). I can’t tell you the number of students who’ve told me later on how this work has changed their lives, that they see the world in a wholly diŸerent way, that they found themselves with a commitment to working that’s just of a diŸerent order — that’s grounded in the world in a diŸerent way — than their previous academic commitment had

49

50

Discussing Conversation Analysis

been, and these were all obviously successful people to begin with. They survived to the point of graduate studies with ªnancial support; they haven’t ever been “bad” students. It’s easy for me to forget that, I suppose, I’m now part of the establishment. That’s how we come to be doing this interview. Thirty-ªve years ago this was a brand new venture, and I ªnd it’s still exciting, and I still ªnd that students who come into it feel that way. It’s not easy to keep that spirit alive in a world that treats you as the older generation and the establishment. But it seems to me that it’s this spirit — a sense that this work is providing a diŸerent kind of access to what it is to be human — that somehow still inhabits the work as an enterprise. It is not passed on from person to person. It is passed on by the nature of the enterprise to those who come to participate in it, to be stewards for it. I think I’d better shut up. Thank you. ¦ and C.L.P.: Thank you. S.C.

Notes * This interview took place on 20th April 1996 in Prague on the occasion of the 6th Congress of the International Association for Dialogue Analysis (IADA). Transcribed and edited by Paul J. Thibault. Additional editorial work to smooth the transition from conversational exchange to written text was undertaken by Emanuel SchegloŸ while he was the beneªciary of a Guggenheim Fellowship and a Fellowship in Residence at the Center for Advanced Study in the Behavioral Sciences, Stanford, CA, under support provided to the Center by The National Science Foundation through Grant # SBR-9022192. Revised version received October 1998; subsequent revisions and bibliographical additions received in March 1999, January 2001, and October 2001 [Editors’ note]. 1. Actually, I recently had occasion to draft about half of the chapter on word selection in preparing a paper for a volume on anaphora which focuses on person reference (SchegloŸ 1996b), and this is to be taken up in the chapter on word selection. 2. For example, JeŸerson 1989 and 1993, among many others. 3. For example, Heritage 1992; Heritage and Greatbatch 1986; Heritage and Roth 1995, among others. 4. And have nonetheless left a lot out, for example my parallel education in Jewish studies, through my college years; my ªrst degree is a Bachelor of Jewish Education — B. J.Ed., 1957 — the year before my Harvard B. A.. When I cast my lot with the study of interaction, the path I did not take was the study of the expected demise of the Yiddish language and the culture revolving around it, a project I was going to call (borrowing from the British literary

Interview: On conversation analysis

critic, Christopher Caudwell (1938), who meant something quite diŸerent by it) “Studies in a Dying Culture.” The sustained engagement with topics in which language ªgures seems quite likely related to my growing up with three of them as a child (English, Hebrew and Yiddish) and three more in school (French, German and Latin), not to mention music. 5. I discuss this form of speech act theory in greater detail in SchegloŸ 1992b: xxiv-xxvii and 1992d. The bearing on Habermas is brie¶y addressed in 1992c: 1139–1141. Other problems with Habermas’ stance towards communication and its place in social life and inquiry into it are taken up in SchegloŸ 1996a: 209–212.

References Caudwell, Christopher. 1938. Studies in a Dying Culture. London: John Lane. Damasio, Antonio R. 1994. Descartes’ Error: Emotion, Reason and the Human Brain. New York: G. P. Putnam. Egbert, Maria. 1996. “Context-sensitivity in conversation analysis: Eye gaze and the German repair initiator ‘bitte’”. Language in Society 25 (4): 587–612. Egbert, Maria. 1997a. “Schisming: The collaborative transformation from a single conversation to multiple conversations”. Research on Language and Social Interaction 30 (1): 1– 51. Egbert, Maria. 1997b. “Some interactional achievements of other-initiated repair in multiperson conversation”. Journal of Pragmatics 27: 611–634. Garªnkel, Harold. 1967. Studies in Ethnomethodology. Englewood CliŸs, N. J.: Prentice-Hall. GoŸman, Erving. 1961. “Fun in games”. In Encounters: Two Studies in the Sociology of Interaction, 15–81. Indiannapolis: The Bobbs-Merrill Company. GoŸman, Erving. 1963. Stigma: Notes on the Management of Spoiled Identity. New York: Simon and Schuster. GoŸman, Erving. 1967. Interaction Ritual: Essays in Face to Face Behavior. Garden City, New York: Doubleday. GoŸman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. New York: Harper and Row. GoŸman, Erving. 1976. “Replies and responses”. Language in Society 5: 257–313. (Reprinted in GoŸman 1981: 5–77). GoŸman, Erving. 1981. Forms of Talk. Philadelphia: University of Pennsylvania Press. Goodwin, Charles. 1995. “Co-constructing meaning in conversations with an aphasic man”. Research on Language and Social Interaction 28 (3): 233–260. Habermas, Jürgen. 1970. “Toward a theory of communicative competence”. In Recent Sociology No. 2, H. P. Dreitzel (ed.), 114–148. New York: Macmillian. Hayashi, Makoto. 1999. “Where grammar and interaction meet: A study of co-participant completion in Japanese conversation”. Human Studies 22 (2–4): 475–499. Hayashi, Makoto and Mori, J. 1998. “Co-construction in Japanese revisited: We do ‘ªnish each others’ sentences’”. In Japanese / Korean Linguistics, N. Akatsuka, H. Hoj, and S. Iwasaki (eds.), (7): 77–93. Stanford: CSLC.

51

52

Discussing Conversation Analysis

Hayashi, Makoto, Mori, J. and Takagi, T. 2002. “Contingent achievement of co-tellership in a Japanese conversation”. In The Language of Turn and Sequence [Oxford Studies in Sociolinguistics], C. E. Ford, B. A. Fox, and S. A. Thompson (eds.), 81–122. Oxford: Oxford University Press. Heeschen, Claus and SchegloŸ, Emanuel A. 1999. “Agrammatism, adaptation theory, conversation analysis: On the role of so-called telegraphic style in talk-in-interaction”. Aphasiology 13 (4–5): 365–405. Heeschen, Claus and SchegloŸ, Emanuel A. 2002. “Aphasic agrammatism as interactional artifact and achievement”. In Conversation and Brain Damage, C. Goodwin (ed.), 231– 282. New York: Oxford University Press. Heritage, John and Greatbatch, David. 1986. “Generating applause: A study of rhetoric and response at party political conferences”. American Journal of Sociology 92 (1): 110–157. Heritage, John. 1992. “Dilemmas of advice: Aspects of the delivery and reception of advice in interactions between health visitors and ªrst time mothers”. In Talk at Work, P. Drew and J. Heritage (eds.), 359–417. Cambridge: Cambridge University Press. Heritage, John C. and Roth, A. L. 1995. “Grammar and institution: Questions and questioning in the broadcast news interview”. Research on Language and Social Interaction 28 (1): 1–60. JeŸerson, Gail. 1973. “A case of precision timing in ordinary conversation: Overlapped tagpositioned address terms in closing sequences”. Semiotica 9: 47–96. JeŸerson, Gail. 1974. “Error correction as an interactional resource”. Language in Society 2: 181–199. JeŸerson, Gail. 1988. “On the sequential organization of troubles-talk in ordinary conversation”. Social Problems 35 (4): 418–441. JeŸerson, Gail. 1989. “Preliminary notes on a possible metric which provides for a ‘standard maximum’ silence of approximately one second in conversation”. In Conversation: An Interdisciplinary Perspective, D. Roger and P. Bull (eds.), 166–196. Clevedon: Multilingual Matters. JeŸerson, Gail. 1993 [1983]. “Caveat speaker: Preliminary notes on recipient topic-shift implicature”. Research on Language and Social Interaction 26 (1): 1–30. JeŸerson, Gail and Lee, John R. L. 1981. “The rejection of advice: Managing the problematic convergence of a ‘troubles-telling’ and a ‘service encounter’”. Journal of Pragmatics 5: 399–422. Jones, John. 1962. On Aristotle and Greek Tragedy. London: Chatto and Windus. Kim, Kyu-hyun. 1993. “Other-initiated repair sequences in Korean conversation as interactional resources”. Japanese/Korean Linguistics, S. Choi (ed.), (3): 3–18. Stanford: CSLI. Kim, Kyu-hyun. 1999. “Phrasal unit boundaries and organization of turns and sequences in Korean conversation”. Human Studies 22: 425–446. Lerner, Gene H. and Takagi, T. 1999. “On the place of linguistic resources in the organization of talk-in-interaction: A co-investigation of English and Japanese grammatical practices”. Journal of Pragmatics 30: 49–75. Lindström, Anna-Karin Benedicta. 1994. “Identiªcation and recognition in Swedish telephone conversation openings”. Language in Society 23 (2): 231–252. Moerman, Michael. 1977. “The preference for self-correction in a Tai conversational cor-

Interview: On conversation analysis

pus”. Language 53 (4): 872–882. Moerman, Michael. 1988. Talking Culture: Ethnography and Conversation Analysis. Philadelphia: University of Pennsylvania Press. Ochs, Elinor, SchegloŸ, Emanuel A. and Thompson, Sandra A. (eds.). 1996. Interaction and Grammar [Studies in Interactional Sociolinguistics 13]. Cambridge: Cambridge University Press. Park, Y.-Y. 1998. “A discourse analysis of contrastive connectives in English, Korean and Japanese conversation: With special reference to the context of dispreferred responses”. In Discourse Markers, A. Jucker and Y. Ziv (eds.), 277–300. Amsterdam: John Benjamins. Sacks, Harvey. 1972a. “An initial investigation of the usability of conversational materials for doing sociology”. In Studies in Social Interaction, D. N. Sudnow (ed.), 31–74. New York: Free Press. Sacks, Harvey. 1972b. “On the analyzability of stories by children”. In Directions in Sociolinguistics: The Ethnography of Communication, J. J. Gumperz and D. Hymes (eds.), 325– 345. New York: Holt, Rinehart and Winston. Sacks, Harvey. 1974. “An analysis of the course of a joke’s telling in conversation”. In Explorations in the Ethnography of Speaking, R. Bauman and J. Sherzer (eds.), 337–353. Cambridge: Cambridge University Press. Sacks, Harvey. 1992. Lectures on Conversation. Two volumes. G. JeŸerson (ed.), with Introductions by E. A. SchegloŸ. Oxford: Basil Blackwell. SchegloŸ, Emanuel A. 1960. The Moral Temper of Literary Criticism, 1930–1960. Unpublished Masters Thesis. Department of Sociology, University of California, Berkeley. SchegloŸ, Emanuel A. 1963. “Toward a reading of psychiatric theory”. Berkeley Journal of Sociology 8: 61–91. SchegloŸ, Emanuel A. 1968. “Sequencing in conversational openings”. American Anthropologist 70 (6): 1075–1095. SchegloŸ, Emanuel A. 1979. “The relevance of repair for syntax-for-conversation”. In Syntax and Semantics 12: Discourse and Syntax, T. Givón (ed.), 261–288. New York: Academic Press. SchegloŸ, Emanuel A. 1980. “Preliminaries to preliminaries: ‘Can I ask you a Question’”. Sociological Inquiry 50 (3–4): 104–152. SchegloŸ, Emanuel A. 1987. “Analyzing single episodes of interaction: An exercise in conversation analysis”. Social Psychology Quarterly 50 (2): 101–114. SchegloŸ, Emanuel A. 1988. “GoŸman and the analysis of conversation”. In Erving GoŸman: Exploring the Interaction Order, P. Drew and A. Wootton (eds.), 89–135. Cambridge: Polity Press. SchegloŸ, Emanuel A. 1990. “On the organization of sequences as a source of ‘coherence’ in talk-in-interaction”. In Conversational Organization and its Development, B. Dorval (ed.), 51–77. Norwood, NJ: Ablex Publishing Co. SchegloŸ, Emanuel A. 1991a. “Conversation analysis and socially shared cognition”. In Perspectives on Socially Shared Cognition, L. Resnick, J. Levine and S. Teasley, 150–171. Washington, D. C.: American Psychological Association. SchegloŸ, Emanuel A. 1991b. “Re¶ections on talk and social structure”. In Talk and Social Structure, D. Boden and D. H. Zimmerman (eds.), 44–70. Cambridge: Polity Press.

53

54

Discussing Conversation Analysis

SchegloŸ, Emanuel A. 1992a. “In another context”. In Rethinking Context: Language as an Interactive Phenomenon, A. Duranti and C. Goodwin (eds.), 193–227. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1992b. “Introduction, Volume 1”. In Harvey Sacks: Lectures on Conversation, G. JeŸerson (ed.), ix-lxii. Oxford: Basil Blackwell. SchegloŸ, Emanuel A. 1992c. “Repair after next turn: The last structurally provided defense of intersubjectivity in conversation”. American Journal of Sociology 97 (5): 1295–1345. SchegloŸ, Emanuel A. 1992d. “To Searle on conversation: A note in return”. In (On) Searle on Conversation, H. Parret and J. Verschueren (eds.), 113–128. Amsterdam and Philadelphia: John Benjamins. SchegloŸ, Emanuel A. 1993. “Re¶ections on quantiªcation in the study of conversation”. Research on Language and Social Interaction 26 (1): 99–128. SchegloŸ, Emanuel A. 1995a. “Discourse as an interactional achievement III: The omnirelevance of action”. Research on Language and Social Interaction 28 (3): 185–211. SchegloŸ, Emanuel A. 1995b. Sequence Organization. Department of Sociology, UCLA, ms. SchegloŸ, Emanuel A. 1996a. “Conªrming allusions: Toward an empirical account of action”. American Journal of Sociology 102 (1): 161–216. SchegloŸ, Emanuel A. 1996b. “Issues of relevance for discourse analysis: Contingency in action, interaction and co-participant context”. In Computational and Conversational Discourse: Burning Issues — An Interdisciplinary Account, E. H. Hovy and D. Scott (eds.), 3–38. Heidelberg: Springer Verlag. SchegloŸ, Emanuel A. 1997a. “Practices and actions: Boundary cases of other-initiated repair”. Discourse Processes 23 (3): 499–545. SchegloŸ, Emanuel A. 1997b. “Third turn repair”. In Towards a Social Science of Language: Papers in Honor of William Labov [Volume 2: Social Interaction and Discourse Structures]. G. R. Guy, C. Feagin, D. SchiŸrin and J. Baugh (eds.), 31–40. Amsterdam and Philadelphia: John Benjamins. SchegloŸ, Emanuel A. 1997c. “Whose text? Whose context?”. Discourse and Society 8 (2): 165–187. SchegloŸ, Emanuel A. 1999a. “Discourse, pragmatics, conversation, analysis”. Discourse Studies 1 (4): 405–435. SchegloŸ, Emanuel A. 1999b. “On Sacks on Weber on Ancient Judaism: Introductory notes and interpretive resources”. Theory, Culture and Society 16 (1): 1–29. SchegloŸ, Emanuel A. 2002. “Conversation analysis and ‘communication disorders’”. In Conversation and Brain Damage, C. Goodwin (ed.), 21–55. New York: Oxford University Press. SchegloŸ, Emanuel A., Ochs, Elinor and Thompson, Sandra A. 1996. “Introduction”. In Interaction and Grammar, E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 1–51. Cambridge: Cambridge University Press. Schutz, Alfred. 1964. “Making music together”. In Collected Papers, Volume II: Studies in Social Theory, 159–178. The Hague: Martinus NijhoŸ. Sorjonen, Marja-Leena. 1996. “On repeats and responses in Finnish conversations”. In Interaction and Grammar, E. Ochs, E. A. SchegloŸ and S. A. Thompson, (eds.), 277– 327. Cambridge: Cambridge University Press.

Interview: On conversation analysis

Tanaka, Hiroko. 1999. Turn-Taking in Japanese Conversation. A Study in Grammar and Interaction. Amsterdam and Philadelphia: John Benjamins. Wu, R.-J. R. 1997. “Transforming participation frameworks in multi-party Mandarin conversation: The use of discourse particles and body behavior”. Issues in Applied Linguistics 8: 97–118.

55

Chapter 3

The power of SchegloŸ ’s work Charles Goodwin

My ªrst encounter with the ªeld of research that Emanuel SchegloŸ, in collaboration with Harvey Sacks and Gail JeŸerson, helped to co-found, absolutely and irrevocably changed my life. In 1970 I had just completed my coursework for a Ph.D. in Communications at the University of Pennsylvania and was profoundly unhappy about the inability of the methods and theories of Social Psychology, which constituted one main emphasis of my program, to come to terms with how people in the primordial site of human communication, faceto-face interaction, in fact communicated with each other. Methodologically, the actual lived social world was considered hopelessly chaotic and unstructured, so that reliable data could only be collected in laboratory settings carefully constructed and controlled by the experimenter. Theoretically, all phenomena of interest were presumed to reside within the mental life of the individual, and what was being sought was not the detailed organization of actual social interaction, but instead quite removed hypothetical variables, such as underlying attitudes. A more fruitful approach seemed to be provided by anthropology and cybernetics. To pursue such interests, I obtained a position videotaping and analyzing family therapy sessions at the Philadelphia Child Clinic, an organization that was continuing the cybernetic analysis of communication in family systems originally developed by Gregory Bateson. As SchegloŸ notes in his interview, therapeutic settings, presumably characterized by pathology in need of a remedy, constituted one of the few places where the analysis of human interaction was o¹cially supported. My new job gave me the opportunity to record and examine in detail a substantial amount of actual face-to-face talk and conduct. However, I became quite disappointed with the institutional assumption that the category “therapy” should be used as a point of departure for all subsequent analysis. This made it impossible to come to terms with how the participants were actually organizing their interaction in concert with each other. For example, it was assumed that one person in the

58

Discussing Conversation Analysis

room, the therapist, not only had special knowledge that the others lacked, but actually understood the family’s interaction with each other better than they did themselves. At this same time, Candy (Marjorie Harness) Goodwin was audio-recording the talk of kids at play on a neighborhood street (later published as He-SaidShe-Said (Goodwin 1990)). At some point, we looked over the transcripts she was making, and decided to ask the question of how the kids got from one utterance to the next. The world of phenomena this simple question opened up was astounding, and data that until that point had seemed dense and intractable suddenly became subtle, intricate and rich with structure she hadn’t even imagined, such as the He-Said-She-Said gossip dispute. I began to explore similar questions about the multi-party construction of action through talk in my videotapes. However, the theoretical frameworks for analysis of discourse then available oŸered almost no purchase on the phenomena we were investigating. Someone suggested that Candy use Searle’s analysis of Speech Acts (1970), which had just appeared, to analyze the He-Said-She-Said. This framework, with a theoretical focus restricted to a timeless moment in the mental life of the speaker, and a participant structure that encompassed only the speaker and the speaker’s assumptions about a decontextualized, idealized hearer (an addressee never treated as an actual co-participant in the construction of action in his or her own right), oŸered no analytic insight into the way in which HeSaid-She-Said accusations created a context for speciªc action in the future by moving three structurally diŸerent kinds of participants (including a former instantiation of the current addressee) through an ordered sequence of stages in a past culminating in a culturally-organized, semiotically-structured present. In short, with the exception of frameworks being developed by GoŸman and Labov which oŸered crucial insights and points of departure, most of the resources then available for the analysis of talk as action were hopelessly inadequate for coming to terms with the intricate and diverse forms of action we were ªnding in our data. After working on her transcripts for several months, Candy showed what she was doing to Labov, who not only responded enthusiastically, but also told us about two people we’d never heard of, Harvey Sacks and Emanuel SchegloŸ, who were also looking at conversation. Indeed, Labov had a collection of mimeographed class lectures by Sacks on conversation that he suggested Candy read. She xeroxed them, and brought them home where I would get a chance to read them every night after I returned from work. This began one of the most intense, stimulating intellectual experiences of our lives. Our own

The power of SchegloŸ’s work

ability to not only ªnd instances of phenomena described by Sacks and his colleagues in our data, but to use those insights independently and in collaboration with each other to discover new phenomena, demonstrated to us that what this work revealed was not just the brilliant insights of a single scholar, such as Erving GoŸman, but a new research ªeld. DiŸerent scholars could independently probe in rich, robust data, a host of phenomena central to what it means to be a human being (language, social organization, the social production of meaningful action, etc.) in a way that permitted their separate proposals to be compared and tested, and ªndings to accumulate. The social nature of not only the phenomena being investigated, but the work itself was vividly demonstrated to us when Gail JeŸerson came to the University of Pennsylvania and we began a series of seminars together. This approach to the study of talk and action provided a radical solution to my unhappiness with existing perspectives in the social sciences. Consider for example the phenomenon of collaborative utterances, in which a ªrst speaker brings an utterance, and the sentence visible within it, to a point of possible completion, and then the next speaker adds a new increment to that sentence (Sacks 1992: 144). Joe: Henry: Mel:

We were in an automobile discussion, discussing the psychological motives for drag racing on the streets

In so far as the sentence that eventually emerges here is constructed through the talk of three separate participants, it provides an example of elementary human social organization, i.e., an instance of a single course of action built through the diŸerentiated work of multiple parties. However, the resources used to build that social organization are drawn from the deªning subject matter of a quite diŸerent ªeld: linguistics. Subsequent speakers use grammatical structure (e.g., Mel’s placement of a noun phrase after the preposition that ends Henry’s talk) to make visible the continuation of a phrase already in progress. Simultaneously, this entire process displays to its addressee (a new boy joining the group) a culturally meaningful action, indeed a construal of events that formulates the speakers as members of a speciªc social mileau: 1960’s teenage hot-rodders. Such practices of building cultural meaning were typically taken to be the special domain of anthropology. In brief, what one ªnds in Conversation Analysis is an analytic framework that cuts across the solid boundaries policed by the existing social sciences, by analyzing social organization, the details of grammatical use, and culture as integrated, seam-

59

60

Discussing Conversation Analysis

less components of a common process: the production of meaningful action within talk-in-interaction. Rather than distilling the processes being investigated into a category scheme predetermined by the investigator, and lumping together quite diverse events as manifestations of a hypothetical variable hidden in the mental life of a subject (an attitude, for example), this approach uses the actual details of what people are saying and doing to investigate how they build the events that constitute their lifeworld. In light of this synthetic perspective, which simultaneously illuminated the details of social organization and grammatical choice, my unhappiness with existing analytic frameworks seemed to emerge from a history in which the social sciences had constituted themselves by each claiming unique control over particular kinds of phenomena (language, social organization, culture, etc.), while trying to diŸerentiate themselves as much as possible from neighboring ªelds with competing claims. The unfortunate outcome was that central phenomena not only got lost between disciplines, but were actively kept in the shadows by the walls between disciplines that emerged through this process. Thus sociologists could comfortably ignore language structure, precisely because that was the domain of linguistics (Sacks 1963), while for its part, linguistics could turn more and more to a view of language as a quintessentially psychological phenomenon. This partitioning, which created the geography of the contemporary social sciences, now began to resemble the Pope dividing Latin America between Spain and Portugal in a way that not only ignored, but actively distorted, the endogenous organization of the territory so divided. Language structure is far too central an element of human social organization to be left exclusively to linguistics. SchegloŸ’s work constitutes a sustained analysis of the relationship between language structure and the practices used to organize human interaction (e.g., SchegloŸ 1979, 1996). One indicator of the power of this approach is its combination of robustness (arising I think from its insistence on looking in detail at what people actually do, combined with a focus on the description and analysis of general, pervasive interactive practices) and ability to accumulate ªndings about, and understanding of, the domain being investigated. Consider for example SchegloŸ’s “Sequencing in Conversational Openings” (1968). This was, I believe, the ªrst example of Conversation Analysis that was published in a major journal. In many ªelds such an “ancient” document would be of no more than historical interest, and its relevance would be restricted to the phenomenon named in its title, here Conversational Openings. However, in every course I’ve taught on conversation, this article has occupied a very prominent place, in large part

The power of SchegloŸ’s work

because the structures described there not only remain relevant, but actually formed the basis for later developments in the ªeld. For example, in order to explicate analytically how openings produced by diŸerent participants were genuine sequences, and not simply individual acts that happened to occur in close proximity to each other, SchegloŸ developed the notion of conditional relevance, roughly the way in which a ªrst action makes relevant a speciªc kind of next action, creating a ªeld in which not only the occurrence of that second action can be seen to be a next-to-the-ªrst, but equally the absence of such an action (e.g., someone who fails/refuses to return a greeting or answer a question) can also be publicly witnessed. This provided what Heritage (1984) has called an architecture for intersubjectivity, as well as a powerful and unique form of context that was incorporated into many later analytic developments, such as Adjacency Pairs and Turn-Taking. The analytic resources used to explicate conversational openings thus had a general power and relevance that extended far beyond the particular phenomenon for which they were ªrst described, and indeed this is a concept that I still cite and use when developing new analysis of my own. SchegloŸ (1968) contains tools that I still use today. More generally, Sequential Organization, the extensive description and analysis of how social action is organized through talk within what SchegloŸ has called the primordial site for human sociality — talk-in-interaction — constitutes one of the most novel and original theories about human social organization, and its relationship to the detailed structure of language, developed in the second half of the twentieth century. Though it provides the primary framework for the emergence of language structure in the lived social world, seminal theorists such as Saussure and Sapir seem to have had no inkling of its relevance or even existence as an ordered domain of structure and action. The power of this framework to provide insightful analysis of phenomena not initially described by it, was vividly demonstrated to me in my earliest research. In looking at my videotapes of talk in a variety of settings (family dinners, picnics, the back room of a meat market, etc.), it immediately became clear that not only the talk, but also the body was being organized through systematic interactive practices. Thus a turn-at-talk is constituted not by a speaker alone, but instead through ongoing processes of mutual orientation between speaker and hearer. Changes in gaze between the participants led to systematic changes in the emerging structure of utterances and sentences, as speakers both worked to get the gaze of disattending hearers, and redesigned the emerging structure of a sentence as the speaker moved from one type of addressee to another (Goodwin 1979, 1980, 1981). Similarly, a story was not

61

62

Discussing Conversation Analysis

simply an extended strip of talk produced by a single speaker, an organized text that an anthropologist could collect somewhere and bring back on paper to add to a card ªle in his or her o¹ce, but an interactive ªeld being sustained through the visible embodied work of structurally diŸerent kinds of participants, e.g., teller, principal character in the story, focal addressee, etc. (Goodwin 1984). In short, the analytic apparatus being developed by Sacks, SchegloŸ and JeŸerson for the study of talk-in-interaction provided a general framework for the investigation of how human beings build temporally unfolding action within the lived social world, one that encompasses not only talk but also the body. More recently I have been extending this framework to encompass how participants make use of tools and other features of the setting to build action. Once again, analytic frameworks developed by SchegloŸ and his colleagues provide crucial points of departure. Thus, in the process of mapping an excavation (see Goodwin 1994 for more detailed analysis), a senior archaeologist gives a directive to a graduate student measuring points in the dirt. Before the addressee has provided any answer whatsoever, the senior archaeologist strongly corrects her. How is it possible to ªnd an answer wrong before it’s even been spoken? By virtue of the ªeld opened by the conditional relevance of the initial directive, the senior archaeologist can inspect the embodied movements and tool use of the addressee, and ªnd that she is not performing the actions that will lead to the production of a correct answer. The radically new approach to the endogenous production of context oŸered by SchegloŸ now encompasses not only subsequent talk, but embodied action and the setting that encompasses it. In the accompanying interview, SchegloŸ discusses the issues raised by analysis of institutional, or other specialized forms of discourse, noting in particular the need “to continuously play back and forth between the specialised domain that they study […] and the ordinary run of human interaction.” Some of my current research focuses on Chil, a man able to say only three words, Yes, No, and And, because of a massive stroke in the left hemisphere of his brain (Goodwin 1995, 2000). Such neurologically-impaired individuals have been the subjects of intense study by physicians, neurologists and psycholinguists interested in what such deªcits can tell us about how language is organized within the brain. To achieve rigor and precision in this very di¹cult project, most research has examined the language abilities of a person with aphasia in carefully controlled laboratory settings. While recognizing the great importance of these eŸorts, in my own research I’ve been tracking a diŸerent animal: the ability of this man to produce meaningful talk and action in the

The power of SchegloŸ’s work

consequential settings that make up his lifeworld, despite the severity of his linguistic impairment. Clearly, both my methodologies and my analytic tools have been strongly in¶uenced by the work that has accumulated within Conversation Analysis over the past thirty-ªve years (including my own research on how the body is organized within interaction). Instead of focussing my camera on a single person in a laboratory carrying out an experimental protocol, I have videotaped Chil talking with his family in their home as they carry out the activities that constitute their daily lives (meals, visits to the hardware store, etc.). Despite the extraordinary sparseness of his vocabulary, Chil is in fact a ¶uent, successful participant in conversation. What makes this possible is the way in which he ties his speech to the talk and action of others, who, for example, state proposals about what he might want to do or say and which he a¹rms or rejects, leading at times to quite extended sequences. Note that all three of his words presuppose links to other talk. And ties other units of talk, such as clauses, to each other. Yes and No are prototypical examples of secondpair parts used to build a response to something that someone else has said. In essence, the features of sequential organization elucidated by Conversation Analysis provide a framework that enables Chil to use the intact language abilities of others to say what he wants. Instead of focussing on the uniqueness of his situation, this approach uses as a point of departure its continuity with the conversational practices of normal speakers. Simultaneously, his situation sheds light on talk-in-interaction in general, by forcing the analyst to focus on phenomena that might be taken for granted, or treated as ephemeral, if accompanied by rich talk. Thus, both participants and analysts frequently take for granted their ability to see what a speaker’s gesturing hand is doing. However, uncovering the meaning of one of Chil’s gestures frequently requires an extended sequence as his interlocutors oŸer a series of possible glosses. This process highlights the way in which the ability to spontaneously, almost transparently, see the sense of a conversational gesture is in fact a dense, interactive accomplishment as talk and gesture build temporally unfolding meaning by mutually elaborating each other. Though revealed with glaring clarity in Chil’s conversation, this rupture between gesture and the talk which makes it understandable is not restricted to interaction with aphasics, but can also be found in normal conversation, though typically in quite special activities such as heckling (Goodwin and Goodwin 1992). In short, as argued by SchegloŸ here, the analyst of aphasic discourse needs to keep one eye ªrmly focussed on the practices of normal conversationalists. More generally, the approach that SchegloŸ helped create for the inves-

63

64

Discussing Conversation Analysis

tigation of how human beings build elementary forms of social organization through talk embedded within situated action, has extraordinary power.

References Goodwin, Charles. 1979. “The interactive construction of a sentence in natural conversation”. In Everyday Language: Studies in Ethnomethodology, G. Psathas (ed.), 97–121. New York: Irvington Publishers. Goodwin, Charles. 1980. “Restarts, pauses, and the achievement of mutual gaze at turnbeginning”. Sociological Inquiry 50: 272–302. Goodwin, Charles. 1981. Conversational Organization: Interaction Between Speakers and Hearers. New York: Academic Press. Goodwin, Charles. 1984. “Notes on story structure and the organization of participation”. In Structures of Social Action: Studies in Conversation Analysis, J. M. Atkinson and J. Heritage (eds.), 225–246. Cambridge: Cambridge University Press. Goodwin, Charles.1994. “Professional vision”. American Anthropologist 96 (3): 606–633. Goodwin, Charles. 1995. “Co-constructing meaning in conversations with an aphasic man”. Research on Language and Social Interaction 28 (3): 233–260. Goodwin, Charles. 2000. “Gesture, aphasia and interaction”. In Language and Gesture, D. McNeill (ed.), 84–98. Cambridge: Cambridge University Press. Goodwin, Charles and Goodwin, Marjorie Harness. 1992. “Context, activity and participation”. In The Contextualization of Language, P. Auer and A. di Luzio (eds.), 77–99. Amsterdam and Philadelphia: John Benjamins. Goodwin, Marjorie Harness. 1990. He-Said-She-Said: Talk as Social Organization among Black Children. Bloomington: Indiana University Press. Heritage, John. 1984. Garªnkel and Ethnomethodology. Cambridge: Polity Press. Sacks, Harvey. 1963. “Sociological Description”. Berkeley Journal of Sociology 8: 1–16. Sacks, Harvey. 1992. Lectures on Conversation: Volume I. Edited by Gail JeŸerson, with an Introduction by Emanuel A. SchegloŸ. Oxford: Basil Blackwell. SchegloŸ, Emanuel A. 1968. “Sequencing in conversational openings”. American Anthropologist 70: 1075–1095. SchegloŸ, Emanuel A. 1979. “The relevance of repair for syntax-for-conversation”. In Syntax and Semantics 12: Discourse and Syntax, T. Givón (ed.), 261–288. New York: Academic Press. SchegloŸ, Emanuel A. 1996. “Turn organization: One intersection of grammar and interaction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 52–133. Cambridge: Cambridge University Press. Searle, John R. 1970. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.

Chapter 4

Putting SchegloŸ’s principles and practices in context Rick Iedema

1.

Introduction

The present chapter will dialogue with selected issues raised in “On Conversation Analysis: An Interview with Emanuel A. SchegloŸ” (hereafter, Interview), which provides an up-to-date overview of SchegloŸ’s past and present endeavors, and of his analytical and ontological appreciation of interaction. The aim here is to home in on three dimensions of his work in the order in which they are addressed in the course of the interview: SchegloŸ’s views on CA analytical method, his statements on the prime object of study, ordinary interaction, and, related to this, his views on the contexts of talk. Before starting out, I will provide a brief summary of the interview with the aim of situating these three points with respect to it. The interview starts with a listing of current projects, after which it revisits some of the history of SchegloŸ’s and Sacks’ original conceptualization of the kind of enquiry that came to be called Conversation Analysis. The interview then pans in on the kind of data CA focuses on, and elaborates the notion ‘chunks of behavior,’ and addresses the issues of their size, their context, and their analytic treatment. Following on from that, CA categories are discussed in light of their ‘formalist’ assessment by a variety of commentators. In defending his concern with interactive formalisms, SchegloŸ takes the opportunity to outline what he sees as CA’s project: to build up an ‘inventory of interactive tools’ or constants, which “interactants draw on […] in concertedly constructing what transpires in interaction” (Interview: 31). The orientation of the interview then shifts towards considering GoŸman’s and Garªnkel’s in¶uence on SchegloŸ’s work. While GoŸman is credited as having ‘opened [SchegloŸ’s] eyes’ to the possibility of analyzing the details of interaction, his “massive” critique of CA (GoŸman 1976 [1981]) is

66

Discussing Conversation Analysis

framed as being essentially “a function of generations and of technology” (Interview: 34). For his part, Garªnkel is discussed as having provided SchegloŸ with a critical impetus towards traditional sociology, while SchegloŸ regards Garªnkel’s analytical work nevertheless as being constrained by the traditional premises and methods of hermeneutic phenomenology (cf. Heritage 1984); that is, ultimately an interpretive exercise which does not make use of ‘objective data’ (Interview: 34–36). The focus of the exchange then shifts towards considering CA’s stance on the centrality of the individual in cognitivist traditions of research. Taking the individual’s ‘single, minded, embodied […] person as the basic, enduring, integrally-organized reality to be studied’, SchegloŸ asserts, is essentially a stance which is Judeo-Christian in origin, and whose presuppositions have given rise to what are now known as ‘psychologically-oriented’ disciplines such as cognitive science. The problem with this stance is that life often shows itself to be structured on the basis of a logic which transcends the individual and his/ her actions. Therefore, rather than explaining tragic events, for example, by reference to the actions of the ‘tragic hero’, we could say that “there are certain sorts of recurrent situation that are the locus of tragedy” (Interview: 38). For SchegloŸ, it is the situation “which is the relevant reality”, while the individual’s intentions are “transient and relatively epiphenomenal” (Interview: 38). While this seems straightforward, the discussion below will show that SchegloŸ’s stance on the agentive status of the individual speaker is not as simple. It is the interactive situation, for SchegloŸ, which furnishes us with “actual data” (Interview: 41). Without those ‘actual data’, the analyst will fall victim to pre-established and pre-conceived ‘social and cultural organization[s] of seeing’ (Interview: 42). It is in this connection too that the notion ‘context’ is addressed, which, for SchegloŸ, remains a hypothetical and conjectural matter unless grounded in the evidence of such ‘actual data.’ For GoŸman, he notes, the analyst can conceive of a ‘membrane’ as surrounding an interaction in so far as that certain ‘objective truths’ (about the interactants or the situation) may or may not ªlter through into the interaction itself. SchegloŸ takes issue with this formulation of how contextual factors are to be judged relevant to what goes on or not. In his view, we cannot decide on the nature of such a ‘membrane’ and of the issues or ‘facts’ which it ªlters out without close analysis of what interactants themselves do and say. Hence, only if dimensions such as race, class and gender become ‘demonstrably’ manifest in talk, can they be classed as interactionally ‘relevant’. Failing the test of demonstrability, such dimensions and their imputed relevance will remain at the level of analytical

Putting SchegloŸ’s principles and practices in context

pre-conception, or social and cultural organisation of seeing: they will lack ‘proven’ relevance for interactants. The motivation for engaging in analysis of the conversation analytical kind, SchegloŸ conªdes (Interview: 44), derives from the view that the “common or vernacular culture is, after all, a sort of ‘propaganda arm’ of the society”. SchegloŸ’s enquiry, then, is ultimately a critique of understandings grounded in everyday common sense: a démasqué of false consciousness which normally permeates the common and everyday. This of course does not only bear out a strong normative element in SchegloŸ’s orientation to what he does, but also sets up a marked distinction between analytical-technical and everyday practice. Put in somewhat diŸerent terms, SchegloŸ views speakers as engaging in everyday interaction understanding and knowing what is relevant, while not necessarily knowing why, unless they have a ‘technical’ way of achieving a “rigorous understanding of society” (Interview: 44). Signiªcantly, this distinction between ordinary experience and technical analysis is an important one for SchegloŸ. CA, “as in any place else where there’s a knowledge-based skill” (Interview: 48), embodies a ‘craft’ such that “we can be treated to be experts” (Interview: 45). CA’s expertise, following this line of argument, is buttressed in two ways: by ‘what is observable’ on the one hand, and by its ability to make ‘basic observations’ on the other: You have to make the basic observations and ground them ªrmly in the observable details of the material. That’s the basic craft. If you don’t do that, however fancy the written papers look, they’ll be based on water and eventually somebody will come along and actually look at the material, and the ungrounded, clever writing will all collapse (Interview: 48).

The last section of the interview deals with SchegloŸ’s work on interactional disorders, and provides the opportunity for him to conªrm that students of any ªeld, including that of behavioral disorders, should ªrst of all build up a solid and intimate knowledge of ‘ordinary interaction’. Ordinary interaction constitutes the ‘conversational baseline’ for ‘more specialized’ as well as ‘deviant’ kinds of talk, from which these latter two are interactive variants. SchegloŸ is of the view that it is ordinary talk “through which the work of most, if not all, the major institutions of societies […] gets done” (cf. SchegloŸ 1996a: 112; 1996b: 54; 1987a: 222). Moreover, ordinary interaction is what occurs ‘naturally’ (Interview: 27), SchegloŸ 1991b: 47; also Sacks, SchegloŸ and JeŸerson 1974: 697), while institutional interaction (much like that of behavioral disorders) imposes specialized and idiosyncratic techniques and requirements on

67

68

Discussing Conversation Analysis

how interactants behave and talk. It is in this way that ordinary, casual talk takes on the status of ‘generic standard’ against which alternative forms of interaction are to be measured (Interview: 45–46, SchegloŸ 1996a: 112, 1996b: 54, 1991: 158, 1987a: 221–2). Having provided my account of the Interview, the chapter now proceeds with a triple focus. First, because the issue is prominent in the Interview and of more general critical interest, we will consider the statements made both in the Interview and in SchegloŸ’s broader body of work in relation to CA’s method of analysis. This ªrst section makes explicit the assumptions underpinning the major claims put forward in this regard, and proceeds to problematize them. Second, and focussing on another prominent aspect of the Interview, the chapter will address the issue of ‘ordinary interaction’ and how it relates to SchegloŸ’s view of ‘the context of talk’. Again, this is done in order to consider the assumptions in which the distinctions are grounded, and to ask whether they are justiªed. Third, the chapter will relate these points to SchegloŸ’s views about context, with speciªc reference to how ‘internal context’ is distinguished from ‘external context’. This revisiting is done with three questions driving the investigation. First, do analysts have unproblematic and transparent access to ‘the relevance of talk’ or do they inevitably construct a but partial narrative about what went on, informed by their discourses, interests and concerns? Second, are we wellserved, as analysts of social life, by working with assumptions about the status of casual talk as ‘ordinary’, ‘natural’ and ‘basic’? Put diŸerently, how appropriate are these terms to describe ‘casual’ interactions if these are rarely if ever free of very delicate positioning strategies, most of which realize highly contestable, if not in fact contested, outcomes for diŸerent speakers (Eggins 1990; Eggins and Slade 1997)? Thus, how possible is it to make a distinction between ordinary and non-ordinary interaction, if ‘ordinary’ means very diŸerent things to diŸerent people, and if even what might seem ‘ordinary interactions’ are in fact highly ‘complicated’ in the way they rely on and construe time/space relations (Latour 1996)? Third and last, how important is it really to restrict our analytical focus to the ‘constants of talk’, and not address more ¶eeting, dynamic and ephemeral aspects? As analysts of social life, we are also likely to be interested in understanding the complex, and often hard to grasp, details and meanings informing social processes; in the multimodal nature of their outcomes,1 and in the contextual conditions which make such processes and outcomes possible. As analysts of social interaction, these kinds of issues and questions are likely to be of some

Putting SchegloŸ’s principles and practices in context

importance, and their answers require considerably more than an investigation into ‘interactive constants’. Finally, as part of the aim to restrict its scope to the Interview and to relevant aspects of SchegloŸ’s work more generally, the chapter uses the acronym ‘SCA’ as a shorthand for ‘Conversation Analysis as deªned by SchegloŸ’.2

2.

SchegloŸ and CA analytical methodology

The central focus of SchegloŸ’s analysis of interaction is a question CA considers to share with ordinary speakers: “why that now?” (SchegloŸ 1980: 147, 1991: 55; SchegloŸ and Sacks 1973). For SchegloŸ, answers to this why-question locate the reasons for particular aspects of talk and behavior not in individuals’ intentions and motivations, as they are aspects of interaction that are too ephemeral and ad hoc to be of analytical interest. Instead, such reasons are to be found entirely within the structural logic of interactive practice itself. In this view, the structural logic of interactive practice has a very diŸerent ontological status from the transient and ad hoc purposes, intentions and expectations which speakers bring to interaction. On the ground of their apparent persistence, then, the structures of interactive logic are the ultimate touchstone for explanations regarding “why that now?”. Importantly, these structural patterns of interaction are thereby accorded both ontological primacy (as the prime source and cause of what people do and say) and analytical primacy (as the focus of our explanations about what people do and say). These interactive patterns or structures represent what SchegloŸ calls the ‘theoretical constants’ of day-to-day interaction. The deªning feature of a theoretical constant is that it structures interaction independent of content; that is, it is “context-free” (Sacks, SchegloŸ and JeŸerson 1974: 699). One example of such a constant is the minimal order which occurs in turn-taking: the ‘adjacency pair’ (Sacks, SchegloŸ and JeŸerson 1974). Optional ‘sidesequences’ provide another example (JeŸerson 1972, 1979). Yet other constants are to be found in the systematicity centering around various ‘forms of repair’ (SchegloŸ 1987b, 1991a). Being primarily concerned with describing theoretical constants, SCA aims to build up “an inventory of sorts, a catalog of recognizable social actions in this culture” (SchegloŸ 1996a: 167), a “behavioral palette” (SchegloŸ 1996a: 209). Such an inventory is to re¶ect the “reservoir of tools, materials and know-how from which particular academic analytic undertakings can draw in inquiry,

69

70

Discussing Conversation Analysis

because practising interactants draw on them in concertedly constructing what transpires in interaction” (Interview: 31). SCA’s inventory of constants is thus meant to provide a map of speakers’ turn-taking resources. This mapping, it is claimed, is an exercise in capturing speaker-centered or ‘emic’ relevances and understandings, and in re¶ecting those relevances and understandings without interference of ‘theory’. It is its concern with ‘emic’ relevances which sets SCA apart from other approaches (SchegloŸ 1987b: 214, Interview: 25–26). DiŸerent […] ‘approaches’ come to the materials of everyday life with a theoretical ªlter that separates the sociological wheat from the chaŸ, the gold from the dross, the important from the trivial, the real from the apparent, the enduring from the transient. Generally this has resulted in accounts that formulate the actions being characterized not by reference to the projects of the actors who enacted them and were their recipients, but only by reference to the terms of the theory that was ‘processing’ them, the only terms by which they were taken to matter […] (SchegloŸ 1996a: 165–6).

It is through close mapping of the analytic (‘the inventory’) onto the empirical (‘natural-occurring interaction’) that SCA claims methodological exactitude (Interview: 27), SchegloŸ 1991a: 191, 1991b: 57; cf. Drew and Heritage 1992a: 20): The aim is to ªnd it [the inventory] and provide an account of it empirically and precisely, not imaginatively or typically or hypothetically or conjecturally or experimentally, and to use actual, situated occurrences of it in naturally-occurring social settings to control its description (SchegloŸ 1996a: 167).

The rigor of SCA’s analytical method is thus buttressed by two arguments: ªrst, SCA focuses not on individuals’ transient concerns, but on the ‘theoretical constants’ of interaction, and second, SCA grounds its ªndings in how interactants themselves assess the status of what is at issue. Deciding what something ‘is’, then, is not a matter of invoking ‘theory’, but of determining what that something means to the speakers themselves. The analyst’s treatment of an utterance as “a possible X” is then grounded in a claim about its having such a status for the participants (SchegloŸ 1996b: 117).

There is a methodological question that arises at this point, however. How does it become analytically evident that “X has such and such a status for the participants”? What are our criteria for deciding that something is [an X] for speaker Y? SCA has answered these questions in the past by referring to the talk’s ‘sequential implicativeness’, which should make evident whether some-

Putting SchegloŸ’s principles and practices in context

thing is [an X] (SchegloŸ 1984b, 1987b). The ‘sequential implicativeness’, or the talk’s sequential architecture, provides information about how the speakers themselves understood the talk in question (Taylor and Cameron 1987: 105). More speciªcally, the talk’s architecture has traditionally been seen then to make ‘demonstrably’ evident both interactants’ accountability to the rules that preside over the talk (SchegloŸ 1991a: 151; Heritage 1984: 291), and the relevance that inheres for them in its various sequencing aspects. Thus, the analyst is ostensibly concerned with the formal features of the talk as evidence for what is perceived as relevant, and s/he ediªes this analytical work on grounds of the claim that these features are representative of how speakers themselves understand what is going on. At one time, sequential-interactive placement was the only analytical clue to determining the relevance of a feature of interaction: Finding an utterance to be an answer, to be accomplishing answering, cannot be achieved by reference to phonological, syntactic, semantic, or logical features of the utterance itself, but only by consulting its sequential placement, e.g. its placement after a question (SchegloŸ and Sacks 1973: 299; my emphasis).

SchegloŸ and Sacks’ approach was able to determine the analytical relevance of some aspect of talk, then, purely on the strength of sequential placement of aspects of talk (although conclusions about what were, for example, initiating questions appeared to be drawn not on the basis of sequential placement alone!). We might conclude, putting these more dated assertions in the context of SCA’s claim that it is circumscribed “by what is demonstrably relevant to the participants” (Interview: 27), that sequential placement and its theoreticalconstant analysis fully delimits the relevance which talk can have. In fact, SCA claims that, beyond the analysis of interactive structure and sequencing, further analysis is unable to shed any more light on such talk: Then, after a ‘technical’ analysis, for example, of repair formats occurring in a segment, the question is asked, “But what is this doing interactionally?” — as if the ‘technical’ analysis supplied a mere substrate or armature to carry the real payoŸ, which is something else. […] In the materials explored in this exercise, the ‘technical’ sequential account is not a substrate on which an ‘interactional what is really going on’ is to be placed. […] An analysis of an action projection as a ‘prepre’ does not occasion the question, “And what is it doing interactionally?”; that is what it is doing interactionally (SchegloŸ 1980: 149–150).

Matters have become more complicated, however, with SchegloŸ recently remarking that a “claim is made, and can be defended, independent of whether

71

72

Discussing Conversation Analysis

the actual recipient on this occasion has treated it as [an X] or not, and independent of whether the speaker can be shown to have produced it for recognition as such on this occasion” (SchegloŸ 1996b: 117). To me as novice, this suggests that SCA’s analytical method is more complex than set out above. The relevance of talk is now not entirely delimited by recipients’ sequential ‘uptake’, and it appears there can be alternative grounds for deciding on whether to treat something as [an X]. While in earlier work the occurrence of theoretical constants was explicated with reference to interactive rules,3 more recently SCA appears to have become less hesitant about making reference to interactants’ motives. One such example is: “Were such a characterization oŸered by Shane, it could be seen as selfexculpating and self-serving. Its production by Michael is precisely a redirection of the blame — which he had appeared to be directing at Shane — now to Vic […]”(SchegloŸ 1996a: 194, also 1980: 143). This practice would appear to compromise SCA’s position: the grounds for deciding whether to treat something as [an X] can thus depend on sequential architecture, but the practice of explication may also rely on hypotheses about speakers’ unexplicit motives.4 It is important to acknowledge that SchegloŸ is right in saying that the conventional practice of explaining the unfolding of interaction purely by reference to individuals’ intentions and motivations is an analytical strategy severely lacking in sophistication: aspects of context and practice equally play a role in (over-)determining speakers’ actions. On the other hand, however, if we conceptualize interaction as a surface phenomenon merely serving to make manifest ‘theoretical-interactive constants’, we fail on two counts. First, we fail to account for the role of alternative resources that constrain and enable (i.e. ‘structure’) interaction: discursive/clausal/intonational resources; bodily resources; gestural resources; sartorial resources; architectural resources, and so on. Second, as formal structuring resources, interactive constants may enable and constrain — but they never fully determine — what transpires. It is this point that relates to GoŸman’s remark about how human interaction unfolds on the strength of not only formal ‘system requirements’ (GoŸman 1981: 16), but also draws on what GoŸman terms (inter)personal or ‘ritual requirements’ (GoŸman 1981: 17). Put simply, the formal systems and structures which we have available as resources for interaction are only theoretically — but not interactively — prior to the informal, interpersonal, dynamic and transient aspects of interaction (GoŸman’s ritual requirements).5 The view taken here is that SCA’s focus on the structural mechanics of interaction is biased against two things. First, for SCA, the situated, creative

Putting SchegloŸ’s principles and practices in context

and often unpredictable ways in which people exploit these structural resources is the dross which hides the real gems: conversational constants. Second, SCA sees structured and structuring resources other than those which organize turn-taking as secondary. While interactive patterns are an undeniably important resource for structuring communication, resources providing structuration at the level of the clause, for example, or intonation, discourse semantics, text-type or genre (see Martin 1992) play an equally signiªcant part in how members of a culture communicate. In our view, a focus on interactive constants is an appropriate analytical end only for those kinds of enquiry whose ultimate objects of concern are their own self-generated formalisms. We say this because no formalism, articulated with whatever level of sophistication and detail, can fully or even satisfactorily account for what transpires between people. Put diŸerently, interaction occurs at the interface between the predictable and the unpredictable, and neither perspective in isolation will ever be able to comprehensively answer the question ‘why that now?’.

3.

Casual talk as ‘ordinary talk’

The second part of this discussion aims to tease out the provenance of and raise questions about a related set of notions central to SCA: ‘naturally occurring human interaction’, ‘mundane talk’ and ‘ordinary conversation’. This set of terms encapsulates what SCA sees as being the primary mode of human communication (Interview: 25–28, SchegloŸ 1987a: 221, 1996b: 54; also highlighted in Levinson 1983: 284; Drew and Heritage 1992a: 19; Segerdahl 1998). We shall argue that attributes such as ‘natural’, ‘mundane’ and ‘ordinary’ are far from transparent. This will involve, ªrst, showing that the implied opposition between terms like ‘natural’ and ‘non-natural’ derives from a pre-empirical assumption about human action and agency. Second, it will involve showing that the privileging of any one particular aspect of human communication as ‘natural’ (i.e. ‘mundane talk’) disregards the ‘complicatedness’ (Latour 1996) which inheres in all human/social semiotic processes. To begin, SchegloŸ warns analysts not to focus exclusively on professional interaction for fear of becoming “un-alert about the non-professional aspects and potentials of the talk” (Interview: 48). This comment is made in the spirit of a two-way distinction imposed on interaction — ordinary talk versus nonordinary or professional talk, with the former favored as “the basic form of organization for talk-in-interaction”:

73

74

Discussing Conversation Analysis

I take it that, in many respects, the fundamental or primordial scene of social life is that of direct interaction between members of a social species, typically ones who are physically co-present. For humans, talking in interaction appears to be a distinctive form of this primary constituent of social life, and ordinary conversation is very likely the basic form of organization for talk-in-interaction. Conversational interaction may then be thought of as a form of social organization through which the work of the constitutive institutions of societies gets done — institutions such as the economy, the polity, the family, socialization, and so on. It is, so to speak, sociological bedrock (SchegloŸ 1995: 187, my italics; also SchegloŸ 1987a: 222).6

SchegloŸ has characterized ‘ordinary conversation’ in the past as follows: in ordinary conversation determination of both who shall speak next and when that one should speak (i.e. when the current turn should end) is accomplished in a local, turn-by-turn manner and not by some pre-determined pattern (SchegloŸ 1987a: 222).

Thus, the ‘ordinary’ is a singularly local or ‘micro’ matter not aŸected or in¶uenced by ‘some pre-determined pattern’, even if sensitive to interactive patterns or ‘constants’. The ‘non-ordinary’, SchegloŸ appears to argue, consists in speakers’ enactment of interactive patterns that exceed the turn-byturn, moment-to-moment enactment of talk. This is clearly a complex issue. By saying that ordinary talk is not aŸected or in¶uenced by ‘some pre-determined pattern’, do we conclude that ‘ordinary conversation’ is the realm where social actors deploy interactive resources whose structural nature restricts their ‘determining in¶uence’ to, for example, the adjacency of turns only, or to the structuring of self-repair? In other words, is ordinary interaction deªned as that dimension of social life where social actors realize SCA’s ‘constants’, but not other kinds of patternings? And is SCA implying that there is a cut-oŸ point between ‘ordinary’ and ‘non-ordinary interaction’, where ‘interactive constants’ shift from being at the behest of speakers to being at the behest of ‘institutional forces’? Put in more generalizing terms, what does it mean to imply, and what are our grounds for implying, that ‘non-ordinary’ talk answers to conventions that are not locally negotiated and is a distal or ‘macro’ matter (Mehan 1987: 293)? Following SchegloŸ’s deªnition cited above, does this distinction mean that in ‘non-ordinary interactions’, social actors’ control over the talk is minimal, or perhaps even non-existent? While perhaps not a common strategy for ªnding the answers to these kinds of questions, we propose to explore whether they are partly to be found by analyzing SchegloŸ’s own discourse. In this discourse, we ªnd that the semantic

Putting SchegloŸ’s principles and practices in context

distinction between ordinary and non-ordinary is in fact realized in the lexicogrammar on the basis of a recurrent syntactic pattern or ‘reactance’ (Whorf 1953: 70).7 This reactance represents a patterned allocation of agency to either individual actors or to social structures. More speciªcally, SchegloŸ’s distinction between ‘ordinary’ and ‘non-ordinary’ is recontextualized at the level of the lexicogrammar in such a way as to co-pattern ‘ordinary talk’ with ‘personcentered agency’, and ‘non-ordinary talk’ with ‘structure-centered agency’. The grammar of SchegloŸ’s discourse, then, already embodies a view as to the agency that individuals have access to in a particular context: individuals are agentive in ordinary interaction’, and they are non-agentive in ‘non-ordinary interaction’. As will become clearer below, the point of this analysis is to show that SchegloŸ’s distinction between ‘ordinary’ and ‘non-ordinary’ renders impossible questions about individuals’ personal socialization into, identiªcation with, or ‘role distance’ from an institutional position (GoŸman 1972). By setting up a radical, unbridgeable (analytical and ontological) divide between how individuals act in ‘casual’ situations and how they act in formal situations (or, for that matter, how they act casually in formal situations and vice versa), SCA pre-decides and dichotomizes the origin and nature of individuals’ interactive contributions, and ultimately, of interactive agency. A prototypical realization of this patterned distribution of agency is the above-mentioned deªnition itself: both who shall speak next and when that one should speak (i.e. when current turn should end) is accomplished [by interactants, RI] in a local, turn-by-turn manner [‘interactants’=implied individual agent, RI] and not by some pre-determined pattern [‘pre-determined pattern’=structural agent, RI] (SchegloŸ 1987a: 222).

This example construes individual speakers’ agency as it is enacted in ‘ordinary talk’ as a local, turn-by-turn accomplishment by interactants of the interaction. In the next instance of discourse, by contrast, agency is ªrst construed as residing with ‘stances’ (stances which require ‘access to the ¶oor and the length of speaking’). Stances here is a structural, institutional entity, and this ‘entity’ (not the speaker) controls the unfolding of the interaction. The last clause of this example also points to the impossibility of speakers taking up particular interactive positionings in formal, institutional situations. Clearly, for SchegloŸ, non-ordinary interaction denies speakers interactive agency. Especially for non-presenters, certain stances vis-à-vis some presentations will not be interactionally feasible (or will entail substantial reputational costs) because of the access to the ¶oor and the length of speaking turn they require

75

76

Discussing Conversation Analysis

[stances=structural (modulating) agent, RI], and the impossibility or unsuitability of accomplishing [(by) non-presenters=(implied) individual agent, RI] those ¶oor requirements in these interactional circumstances (SchegloŸ 1987a: 227).

As the following two examples show, agency is mapped onto ‘interactional contexts’ and ‘courtroom-ness’, and both are credited with control over interaction patterns when ‘non-ordinary’ or institutional contexts are addressed: interactional contexts can demonstrably control [contexts=structural agent, RI] what participants in conversation think to say (SchegloŸ 1987a: 227); it is the ‘courtroom-ness’ of courtrooms in session which seems in fact to organize [courtroom=structural agent] the way in which the talk is distributed among the persons present (SchegloŸ 1991b: 54).

The examples and their analyses show how contextual aspects (‘structures’, not actors) are accorded lexicogrammatical agency when they are (considered) ‘institutional’, and how individuals are accorded agency outside of such contexts (or within those contexts, but interacting ‘ordinarily’). In this way, the lexicogrammar of SchegloŸ’s own discourse aligns individuals with agency when addressing ‘ordinary’ talk, and aligns structural factors with agency when dealing with institutional or ‘non-ordinary’ talk. Unfortunately, this discursive reactance of course pre-empts the empirical question as to whether a speciªc stretch of talk is enacted voluntarily or not. Moreover, the empirics of this matter are not guaranteed to be found in individuals’ turn-taking tactics, but in individuals’ levels of identiªcation with what they do and say, and in their views about whether they are forced to act and speak in a particular way or whether they do so of their own choosing. Thus dutifully replayed at the level of the lexicogrammar, the semantictheoretical distinction between ‘ordinary interaction’ as individual-centered, and ‘non-ordinary’ interaction as pattern-centered, is rendered natural and necessary. SchegloŸ’s selective and pre-determined ascription of agency, then, leaves the analyst no room for asking questions about the extent to which speakers have control over their everyday and institutional roles. More speciªcally, SchegloŸ’s discourse eŸectively rules out of court the question of speakers’ level of identiªcation with or ‘role distance’ (GoŸman 1972) from speciªc aspects of their social and institutional practices. SchegloŸ’s position is highly problematic for researchers interested in the issues of professionalization and the accomplishment of professionalism, especially because people’s enact-

Putting SchegloŸ’s principles and practices in context

ments in institutional practices are not at all necessarily a matter of ‘subjugation to institutional forces’: subjects’ […] sense of self-identity is invested in the reproduction of [institutional] practices — not simply to achieve material rewards or avoid punishment but to gain and conªrm a (self-disciplining) sense of their own normality as sovereign subjects (Willmott 1994: 106).

Hence, the voluntary reproduction of institutional practices can be a source of subjectiªcation for interactants, as well as constituting a condition for their participation. Put diŸerently, interactants frequently willingly identify with institutional practices of doing and talking, even if that delimits how they as actors can construct themselves discursively and practically. Thus, and even if diŸerent kinds of resources are mobilized, institutional practice or discourse, like any other form of practice and discourse, is ultimately a function of interactants’ own agency8 (cf. Schatzki 1996; Schatzki and Natter 1996). In our view, agency is an ongoing form of self-construction which, linguistically speaking, can range from ‘I’ as single person to the individual being conceived as “the position” (Iedema 1998). In Kress’s words: ‘If I am to become the incumbent I will be responsible for…’. This ‘I become/am the incumbent…’ is, in my terms, a crucial, although unarticulated step in the training of social and linguistic subjects. It may lead from ‘(as the incumbent) in this position I am responsible for…’ to ‘As the position, I am responsible for…’, and ªnally to ‘The position is responsible for…’ (Kress 1995: 124).

Hence, agency is not an ‘eŸect’ of what others and circumstances allow us to do, but a matter of how we construct ourselves through saying and doing. In this view then, institutional discourse and practice is where the: social and linguistic agent has come to see him or herself as this kind of subject, who has not just colluded in and ‘accepted’ this construction of him or herself, but has actively participated in his or her (re-)production, so that this is now not just a possible but an unremarkable, habitual, and ‘natural’ speaking position (Kress 1995: 124).

Equally, as in Drew and Sorjonen’s view (1997: 97), “participants’ institutional identities can be viewed, not as exogenous and determining variables, but as accomplished in interaction”. For them, too, “a key focus of research into institutional dialogue is to show how participants’ orientation to their institutional identities is manifest in the details of the verbal conduct through which they manage institutional tasks” (Drew and Sorjonen 1997: 97). The question

77

78

Discussing Conversation Analysis

then to what extent “institutional context can demonstrably control what participants think to say” is therefore an empirical one, not a theoretical one. As Kress’ argument above bears out, participants can willingly adopt and manipulate institutional resources, and thereby identify (‘become identical with’) with the institutionality whose interactive resources they are deploying (Iedema 1998). Before moving to the last section of this chapter, we should like to make a brief comment about the likely provenance of the dichotomy which SchegloŸ sets up between ‘ordinary’ interaction and individual agency on the one hand, and ‘non-ordinary’ interaction and structural agency on the other. Rather than originating with SchegloŸ, this dichotomy harks back to a distinction popular in traditional sociology which opposes social actors to inert social structures (for discussion see Giddens 1981, 1984; Bourdieu 1990; Manicas 1993; Archer 1995 for speciªc positions on this; see Knorr-Cetina and Cicourel 1981; Alexander et al. 1987; Bryant and Jary 1991 for overviews). Brie¶y, this literature sees two things as the more serious shortcoming of this paradigm. First, the paradigm reduces social theorizing to two dimensions only (individual action and social structure). Second, the incommensurability which the paradigm (often implicitly) ascribes to individual action and social structure deªes attempts to explain how it is that action and structure mutually aŸect one another in other than all or nothing terms. We might take some time to ¶esh out these two last-mentioned issues, because they have implications for SCA as an analytical project. With reference to the ªrst problem, a dichotomous perspective on social process prevents us from appreciating that there are multiple resources through which we act and which act through us, whether interaction takes place in ‘ordinary’ or in ‘nonordinary’ settings. The important point here is that these resources harbour diŸerent levels or modalities of structuredness. As a continuum, these modalities of structuredness range from relatively ephemeral and temporized (such as certain kinds of neuronal activity, speech, or gesture), to others which are more tangible and durable (such as architecture). On the one end of this continuum, think of dreams, thought and cognition materialized as neuronal ªrings; think of the human voice, intonation and pitch materialized as larynx and muscle movements and airwaves; or think of interactive gaze, gesture and posture materialized as bodily behaviors. Towards the other end of the continuum, think of dress and other forms of body adornment; the ‘produced space’ of spatial, architectural and infrastructural meanings materialized through exosomatic resources (Lefebvre 1991), and so on, and so on. As Latour puts it:

Putting SchegloŸ’s principles and practices in context

The interactionists are right when they say we should never [stop focussing on] interactions — but if one follows human interactions then one never stays in the same place, nor even in the presences of the same actors and never in the same temporal sequence. Herein lies the complete mystery that made their [interactionists’] adversaries say that they did not take ‘structural eŸects’ or the ‘macro’ into account. By dislocating interaction so as to associate ourselves with nonhumans, we can endure beyond the present […] (Latour 1996: 238–239).

These ‘non-human’ modalities enable and constrain to diŸerent degrees, and each imposes diŸerent levels of structural complexity and constraint aŸorded by their respective materialities. With regard to the argument we are developing here, we note that individual action is never possible without relying at least on some kinds of ‘pre-determined patterns’, or on a number of modalities of structure.9 Thus, much as it does not help to oppose action to structure, distinguishing ‘ordinary’ from ‘non-ordinary’ interaction along the lines suggested obscures more than it clariªes. Hence, the way SCA privileges speech means the diŸerent aŸordances of talk, gaze, gesture, tools, and spatial structure and their role in structuring the interaction are not addressed, and thus naturalized. The way SCA a priori locates the source of interactants’ agency means questions about their mode of subjectiªcation (e.g. their professionalization) fall by the wayside. The second problem is about how action and structure mutually constitute one another. Put simply, a dichotomous action/structure paradigm makes it hard for us to address how diŸerent structural modalities relate and invites us to privilege one modality (e.g. speech) at the expense of others. Questions that become impossible to ask within the conªnes of this paradigm, for example, are how and why some modalities are preferred over others at times, and how and why we see interaction transform some modalities into other modalities. For example, we could ask why and how some talk transforms into movement, silence or touch (McInnes 1998; Lebaron and Streeck 1997; Streeck 1996), or talk into forms of writing, architectural design and built construction (Mehan 1993; Iedema 1997a). One guiding question here might be how the realization of some meanings in materialities other than the communicating human body (meaning realized in ‘exo-somatic materialities’) makes them less changeable, and therefore less negotiable. A related question turns on why meanings realized in such exosomatic materialities often become taken-as-given by speakers in interaction — think of the variety of material objects which surrounds us (and why these are often relegated to static ‘context’ by social analysts rather than included in

79

80

Discussing Conversation Analysis

the active meaning-making domain!). All these are in fact questions about how meaning is constantly rearticulated or ‘resemiotized’ as interaction unfolds (Iedema 1997a, 1999, 2000, 2001), and they aim to respect the semiotically rich, ¶uid and complex nature of human interaction. While acknowledging the unique role of spoken language in interaction, the essence of these questions is that they neither ascribe to speech an a priori or pre-empirical prominence, nor pre-judge levels of control or agency on the part of human or other participants (e.g., written directives, conventions, and so on). Our analysis of interaction is not served by constructing ideal-types of human interaction centered around linguistic exchange and assumptions of what drives those exchanges. Rather, in this age of increasingly complex and time-distantiated forms of communication and social organization, we might be better served regarding social interaction from the perspectives of multimodality and dynamic semiotic transformation.

4.

Two types of context

This section of the paper asks two related questions. The ªrst is about the distinction SchegloŸ draws between types of contexts, and the second is about the claimed status of SCA as “technical analysis”. Running parallel with SCA’s a priori opposition between ordinary and ‘situationally pre-determined’ talk, a distinction is made between contexts which are ‘external’ and those which are ‘intra-interactional’ (SchegloŸ 1992a: 195) or ‘macro-structural’ (Interview: (43)). This distinction is again clearly related to the separation SchegloŸ imposes between ‘common social understanding’ and ‘technical understanding of the organization of social life’ (SchegloŸ, Interview: 44). We might be tempted to draw all kinds of macro-social conclusions about an interaction, such as whether someone interacts the way s/he does because of a certain gender, race or class background. SchegloŸ’s point is that SCA’s technical analysis only can show whether aspects of the ‘external’ context (Interview: 44) are demonstrably enacted or whether they remain “exogenous”. As SchegloŸ elaborates in an earlier paper, the ‘external’ comprises: aspects of social life long central to the social sciences — the class, ethnic, and gender composition of an interaction, each of these understood either as a distinctive source of ordering of and constraint on social life, or as an embodiment of more general properties such as ‘power’ … . Here as well are found the various institutional matrices within which interaction occurs (the legal order, economic

Putting SchegloŸ’s principles and practices in context

or market order, etc.), as well as its ecological, regional, national, and cultural settings, all of which may be taken as ‘shaping’ what goes on under their auspices or in arenas of social life on which they have a bearing (SchegloŸ 1992a: 195).

In contrast to these ‘external’, contextual issues, the ‘intra-interactional’ includes: the sort of occasion or genre of interaction which participants, by their conduct, make some episode be an instance of, the sorts of sequences of talk or courses of conduct in which particular events may occur […], the capacity in which participants act relative to the episode in progress […] (SchegloŸ 1992a: 195).

In response to this distinction, and perhaps repeating some of the things we said earlier, we would argue that any human interaction is embedded in multiple and complexly-related contexts of relevance (Latour 1996: 232). Latour formulates this point as follows: We say, without giving the matter too much thought, that we engage in ‘face-toface’ interactions. Indeed we do, but the clothing we are wearing comes from elsewhere and was manufactured a long time ago; the words we use are not formed on this occasion; the walls we have been leaning on were designed by an architect for a client, and constructed by workers — people who are absent today, although their action continues to make itself felt. The very person we are addressing is a product of history that goes far beyond the framework of our relationship. If one attempted to draw a spatio-temporal map of what is present in the interaction, and to draw up a list of everyone who in one form or another were present, one would not sketch out a well-demarcated frame, but a convoluted network with a multiplicity of highly diverse dates, places and people (Latour 1996: 231).

Latour here points to how all modes of human interaction have complex sociohistorical links and dimensions, and that neither social actors, nor analysts, are able to satisfactorily separate the now from either its past and future or from its immediate spatial-cognitive background. Thus, since interaction is both dynamically and complexly intertwined with both cognitive-embodied and exosomatic aspects of interaction (Fairclough 1995; Goodwin and Duranti 1992; Streeck 1996; Lebaron and Streeck 1997), these ‘contextual factors’ cannot be treated post hoc in analysis, not even in the ªrst instance. A further implication of Latour’s statement is that SchegloŸ’s distinction between ‘what is technically there now’ and ‘what we interpretively mobilize to understand or explain the situation’ is tenuous at best, which leads us to our second point. Where our ªrst point questioned the distinction between types of context, this second point queries SCA’s self-description as technical analysis’ (Interview: 36, 44–45). Again in parallel with other distinctions made in SchegloŸ’s

81

82

Discussing Conversation Analysis

work (‘ordinary’ versus ‘non-ordinary’; ‘intra-interactional’ versus ‘external’ contexts), an opposition is put forward between social analysis fueled by ‘exogenous interests’ and that based on ‘technical analysis’ (Interview: 36, 44–45). This distinction is drawn with the aim to favor interaction as it is standardly captured (tape-recorded), represented (transcribed)10 and analyzed by SCA, and to marginalize ‘exogenous’, ‘analyst-centered’ concerns such as power, gender, age and ethnicity. The reasoning behind the distinction is that ideas about power, gender, age, ethnicity and so on derive from abstract kinds of sociology (which rely on what Sacks called ‘unexplicated resources’ and as referred to in the Interview). Due to their abstract nature, these conceptions do not standardly or necessarily have relevance for speakers in mundane interaction. Their very ‘sociologicalness’ situates them within the specialized discourses of sociology, discourse analysis, or feminist theory; their a priori relevance to ordinary talk is unwarranted, and their relevance or applicability to stretches of interaction needs to be demonstrated. For SchegloŸ, and as shown above, to show the relevance of exogenous issues such as gender, race, class and ethnicity is to engage in a secondary kind of description and argumentation, both of which remain contingent upon the outcomes of SCA’s ‘technical analysis’. While it cannot be denied that some statements describe better or more technically a given reality compared to others, it would be unwise to claim that any particular analysis can technically, comprehensively and objectively account for what transpires in interaction. While we can of course point out the advantages of one kind of analysis or description over another (as I do in this paper), no one analysis or description can serve all human interests and perspectives. Analysts are not born and raised with similar assumptions about and views on the world. It follows that while some might be interested in pursuing what they see as being ‘constant’ in interaction and thereby formalize a status quo, others might be devoting themselves to understanding the dynamics of communicative positioning and semiotic transformation, and to devising methods and strategies for in¶uencing or challenging those dynamics. Finally, and perhaps most problematically, to claim that SCA’s ‘technical analysis’ is categorically Other to ‘analyst-centered interpretation’ shows evidence of an attitude which regards ‘demonstrable proof’ as the foundation of knowledge (cf. Taylor 1995 on scientiªc ‘empiricism’), and which is unable to contextualize its own practices and relate them to its own speciªc and situated social and moral understandings and objectives (Billig and SchegloŸ 1999).11 In contrast to this, we promote the view that our own interpretive frameworks

Putting SchegloŸ’s principles and practices in context

cannot be divorced from our analysis at will; they cannot be absolute and therefore transparent, and they arise from and serve a speciªc set of situated interests and concerns. The distinction made between ‘technical analysis’ and ‘interpretive description’ can thus ultimately not be a categorical one, and remains one which is theoretically imposed and politically legitimated.

5.

Conclusion

This chapter has engaged with three issues which were addressed in the Interview, and which are central to SchegloŸ’s social scientiªc project: analytical methodology, types of talk, and types of context. In each case, our argument has centered on showing that SchegloŸ constructs analytical-interpretive distinctions as ‘real’ or ontological distinctions, and that in each case (analytical methodology, types of talk, types of context), a theoretical, discursive construal was elevated to scientiªc premise. We also elaborated on those dimensions of social interaction which remain unaddressed as a result of the analytical delimitations put forward in SchegloŸ’s work. We mentioned the complex network of socio-historical meanings that characterizes interaction and which ‘spreads’ across the whole gamut of semiotic-material resources which interaction mobilizes. Here we referred to the dynamic, productive nature of interaction, which is evident from how actors materially structure and transform their exchanges — as muscular movement and airwaves in talk, as imprinting in writing, drawing, painting and design, as manipulating electronic signals in computerized communication, or as building structures in the production of our social spaces. In closing, SchegloŸ’s work centers around the requirement that something has to be part of human interaction to become socially relevant and meaningful (cf. Garªnkel 1967; Butler 1996; Cameron 1997: 29 on ‘performativity’). While this is an important corrective to those sociological constructions which are unable to relate phenomena such as globalization to the realm of human interaction (McKenna and Graham 2000), social science is not wellserved by essentializing human interaction as spoken exchange, by reifying sources of human agency, and by disregarding the time-space distantiating dimensions of social life. In the view advanced here, human interaction is both multi-modal and transformative, and the analysis of interaction from those alternative perspectives enables a rich appreciation of its complex and dynamic nature.

83

84

Discussing Conversation Analysis

Notes 1. Multimodality aims to capture a shared perspective on meaning-making which acknowledges that a speciªc mode of meaning-making (such as gesture, language, or image) rarely occurs on its own, and is generally integrated with, and sometimes even dependent on, other modes of meaning-making. The concept multimodality, therefore, constitutes a way to formulate an approach to the analysis of such co-occurrent modes. As such it answers a recent re-orientation in both teaching and research towards addressing such con¶uences of various modes of meaning-making (Kress and van Leeuwen 1996; Iedema 2000, 2001). 2. Work such as Goodwin and Duranti 1992, Ford and Thompson 1996, Pomerantz and Fehr 1997, and several others, while inspired by, if not deriving from, CA, present at times clearly very diŸerent views on the methodological and analytical issues under focus here. 3. “Analysis of the [deviant] case reveals that the distribution rule, while it holds in most cases, is in fact best understood as a derivative of more general rules […]”(SchegloŸ 1972: 356, cited in Coulter 1983: 373). 4. Other commentators, too, have, after “close inspection of the practice of […] conversation analysis, revealed considerable motive attribution” (Bruce and Wallis 1985: 467; cf. also Sharrock and Watson 1984; Dubois 1993). 5. To make the problem referred to here more tangible, we might draw the parallel between interaction and phenotype and between theoretical constant and genotype: never can the genotype be said to be entirely isolated or independent from the surface experiences of the phenotype. Both are inextricably intertwined, and neither is prior to the other (Edelman 1992; Dennett 1991, 1995; Deacon 1997). 6. Elsewhere, SchegloŸ positions ordinary talk as the “primary constituent of social life” (SchegloŸ 1996b: 54). Note the divergence between SchegloŸ’s and Berger and Luckmann’s views on the status of ordinary conversation: “It is important to stress however that the greater part of reality-maintenance in conversation is implicit, not explicit. Most conversation does not in so many words deªne the nature of the world. Rather, it takes place against a background of a world that is silently taken for granted” (Berger and Luckmann 1966: 173, my emphasis). For SchegloŸ, the implicit has no analytical status if not demonstrable within the framework of SCA’s transcriptional and analytical methodology. 7. A reactance is deªned as “not the analog of a word, but of a rapport-system, and awareness of it has an intuitive quality […]. It is a submerged, and elusive meaning, corresponding to no actual word, yet shown by linguistic analysis to be functionally important in the grammar” (Whorf 1953: 70). Whorf applies this notion to diŸerent languages; here we apply the term to characterize patterns in a particular person’s discourse. 8. Agency, and thus the power to act, are not “acquired, seized, or shared” (Foucault 1978: 94); they are the very condition for our partaking in social (semiotic) interaction in the ªrst place. 9. While simian societies manifest relatively simple forms of co-presence, human societies exploit a whole spectrum of symbolic-material means to frame goings-on. This means human interaction embodies and mobilizes the eŸects and products of diŸerent times, spaces and

Putting SchegloŸ’s principles and practices in context

processes. It is for that reason that no form of interaction, however ordinary it may appear to us, can be regarded as manifesting nothing but itself. Latour, in reporting on his research in the domain of simian sociology, takes up the question of whether it is justiªed to regard human interaction as a purely local matter (SchegloŸ’s “ordinary”), and whether social analysis should proceed on the assumption that contextual factors, if relevant, will become evident from aspects of the interaction itself. In taking up these questions, Latour argues that for interaction among monkeys, these assumptions may indeed be appropriate: “each interaction must, locally and on its own account, test all over again the set of relationships without being able to sum, nor to enter into a determined role or function that would hold by itself without the aid of physical bodies” (Latour 1996: 233). In contrast to this, human interaction is not only contextualized but also historicized: what transpires must be considered in the light of what went before and of what will follow, as well as of the multiple causes (agencies) that come into play: “every time we go from the complex life of monkeys to our own, we are struck by the multiple causes that at once come into play to dislocate co-presence from social relationships” (Latour 1996: 233). Human interaction inevitably implicates everything which the interactants rely on for enacting the exchange: language and the registerial, textual, grammatical and phonological resources which interactants draw on; space and its aŸordances such as acoustics and possibilities for human movement and separation; tools and modes of communication (a mobile phone? an on-line chat?) and all the levels of social capital inscribed into those diŸerent modes, and so on. 10. See Thibault 1994 on the status of transcription in social analysis. 11. My re-analysis of the transcript of an extended stretch of talk pubished in SchegloŸ 1991a suggests that diŸerent views on “why that now” are in fact not uncommon, and that SchegloŸ’s analysis is informed by a very speciªc view of male-female relations (Iedema mimeo).

References Alexander, JeŸrey C., Giesen, Bernard, Münch, Richard and Smelser, Neil J. (eds.). 1987. The Micro-Macro Link. Berkeley, CA.: University of California Press. Archer, Margaret S. 1995. Realist Social Theory: The Morphogenetic Approach. Cambridge: Cambridge University Press. Berger, Peter and Luckmann, Thomas. 1966. The Social Construction of Reality: A Treatise in the Sociology of Knowledge. Harmondsworth: Penguin Books. Billig, Michael and SchegloŸ, Emanuel A. 1999. “Critical discourse analysis and conversation analysis: An exchange between Michael Billig and Emanuel SchegloŸ”. Discourse and Society 10 (4): 543–582. Bourdieu, Pierre. 1990. The Logic of Practice. Cambridge: Polity Press. Bourdieu, Pierre. 1991. Language and Symbolic Power. Edited and introduced by J. B. Thompson. Cambridge, Mass: Harvard University Press. Bruce, Steve and Wallis, Roy. 1985. “‘ Rescuing motives’ rescued: A reply to Sharrock and Watson”. The British Journal of Sociology, xxxvi (3): 467–470.

85

86

Discussing Conversation Analysis

Bryant, Christopher G. A. and Jary, David (eds.). 1991. Giddens’ Theory of Structuration: A Critical Appreciation [International Library of Sociology]. London: Routledge. Butler, Judith. 1996. “Performativity’s social magic”. In The Social and Political Body, T. Schatzki and W. Natter (eds.), 29–48. New York: The Guilford Press. Cameron, Deborah. 1997. “Theoretical debates in feminist linguistics: Questions of sex and gender”. In Gender and Discourse [Sage Studies in Discourse], R. Wodak (ed.), 21–36. London: Sage Publications. Chomsky, Noam. 1988. Language and the Problem of Knowledge: The Managua Lectures. Cambridge, Mass: MIT Press. Christie, Frances, and Martin, James R. (eds.). 1997. Genre and Institutions: Social Processes in the Workplace and School. London: Cassell. Collins, Randall. 1981. “Micro-translation as a theory-building strategy”. In Advances in Social Theory and Methodology: Toward an Integration of Macro- and Micro-Sociologies, K. Knorr-Cetina and A. V. Cicourel (eds.), 81–108. Boston: Routledge Kegan Paul. Coulter, JeŸ. 1983. “Contingent and a priori structures in sequential analysis”. Human Studies 6: 361–376. Deacon, Terrence W. 1997. The Symbolic Species: The Co-Evolution of Language and the Brain. Harmondsworth: Penguin Books. Dennett, Daniel Clement. 1991. Consciousness Explained. Harmondsworth: Penguin Books. Dennett, Daniel Clement. 1995. Darwin’s Dangerous Idea: Evolution and The Meaning of Life. Harmondsworth: Penguin. Drew, Paul and Heritage, John. 1992a. “Analysing talk at work: An introduction”. In Talk at Work: Interaction in Institutional Settings [Studies in Interactional Sociolinguistics 8], P. Drew and J. Heritage (eds.), 3–65. Cambridge: Cambridge University Press. Drew, Paul and Heritage, John (eds.). 1992b. Talk at Work: Interaction in Institutional Settings [Studies in Interactional Sociolinguistics 8]. Cambridge: Cambridge University Press. Drew, Paul and Sorjonen, Marja-Leena. 1997. “Institutional dialogue”. In Discourse as Social Interaction (Vol. 2), T. A. van Dijk (ed.), 92–118. London: Sage Publications. Du Bois, John. 1993. “Meaning without intention: Lessons from divination”. In Responsibility and Evidence in Oral Discourse [Studies in the Social and Cultural Foundations of Language 15), J. H. Hill and J. T. Irvine (eds.), 48–71. Cambridge: Cambridge University Press. Duranti, Alessandro and Goodwin, Charles (eds.). 1992. Rethinking Context: Language as an Interactive Phenomenon. Cambridge: Cambridge University Press. Edelman, Gerald H. 1992. Bright Air, Brilliant Fire: On the Matter of the Mind. Harmondsworth: Penguin Books. Eggins, Suzanne. 1990. Keeping the Conversation Going: A Systemic-Functional Analysis of Conversational Structure in Casual Sustained Talk. Ph.D. Thesis, Linguistics Department, University of Sydney. Eggins, Suzanne and Slade, Diana. 1997. Analysing Casual Conversation. London: Cassell. Fairclough, Norman. 1992. Discourse and Social Change. Cambridge: Polity Press. Fairclough, Norman. 1995. Critical Discourse Analysis: The Critical Study of Language [Language in Social Life Series]. London: Longman.

Putting SchegloŸ’s principles and practices in context

Ford, Cecilia E. and Thompson, Sandra A. 1996. “Interactional units in conversation: Syntactic, intonational, and pragmatic resources for the management of turns”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 134–184. Cambridge: Cambridge University Press. Foucault, Michel. 1978. The History of Sexuality, Vol. I. Harmondsworth: Penguin Books. Garªnkel, Harold. 1967. Studies in Ethnomethodology. Englewood CliŸs, N. J.: Prentice-Hall. Giddens, Anthony. 1984. The Constitution of Society. Cambridge: Polity Press. GoŸman, Erving. 1972. Encounters: Two Studies in the Sociology of Interaction. Harmondsworth: Penguin Books. GoŸman, Erving. 1976. “Replies and responses”. Language and Society 5: 257–313. GoŸman, Erving. 1981. Forms of Talk. Oxford: Basil Blackwell. Goodwin, Charles. 1995. “Co-constructing meaning in conversations with an aphasic man”. Research on Language and Social Interaction 28 (3): 233–260. Goodwin, Charles and Duranti, Alessandro. 1992. “Rethinking context: An introduction”. In Rethinking Context: Language as an Interactive Phenomenon, A. Duranti and C. Goodwin (eds.), 1–42. Cambridge: Cambridge University Press. Harvey, David. 1996. Justice, Nature and the Geography of DiŸerence. Oxford: Basil Blackwell. Heritage, John. 1984. Garªnkel and Ethnomethodology. Cambridge: Polity Press. Heritage, John and Atkinson, J. Maxwell. 1984. “Introduction”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 1–15. Cambridge: Cambridge University Press. Iedema, Rick. 1997a. Interactional Dynamics and Social Change: Planning as Morphogenesis. Unpublished Ph.D. Thesis, University of Sydney. Iedema, Rick. 1997b. “The language of administration: Organising human activity in formal institutions”. In Genres and Institutions: Social Processes in the Workplace and School, F. Christie and J. R. Martin (eds.), 73–100. London: Pinter. Iedema, Rick. 1998. “Hidden meanings and institutional responsibility”. Discourse and Society 9 (4): 497–516. Iedema, Rick. 1999. “The formalisation of meaning”. In special issue of Discourse and Society on organizational research, R. Wodak and R. Iedema (eds.), 10 (1): 49–66. Iedema, Rick. 2000. “Bureaucratic planning and resemiotisation”. In Discourse and Community: Doing Functional Linguistics, E. Ventola (ed.), 47–70. Tübingen: Narr Verlag. Iedema, Rick. 2001. “Resemiotization”. Semiotica 137 (1/4): 23–39. Iedema, Rick. Mimeo. “A discussion with Emanuel SchegloŸ: The relevance of structure and the signiªcance of interactive dynamics”. University of New South Wales. JeŸerson, Gail. 1972. “Side sequences”. In Studies in Social Interaction, D. Sudnow (ed.), 294–338. New York: The Free Press. JeŸerson, Gail. 1979. “A technique for inviting laughter and its subsequent acceptance/ declination”. In Everyday Language: Studies in Ethnomethodology, G. Psathas (ed.), 7996. New York: Irvington Press. Knorr-Cetina, Karin D. and Cicourel, Aaron V. (eds.). 1981. Advances in Social Theory and Methodology: Toward an Integration of Micro and Macro Sociologies. Boston: Routledge

87

88

Discussing Conversation Analysis

Kegan Paul. Kress, Gunther R. 1995. “The social production of language: History and structures of domination”. In Discourse in Society: Systemic Functional Perspectives Vol. L: Meaning and Choice in Language: Studies for Michael Halliday [Advances in Discourse Processes], P. Fries and M. Gregory (eds.), 115–140. Norwood, N. J.: Ablex. Kress, Gunther R. and van Leeuwen, Theo. 1996. Reading Images: The Grammar of Visual Design. London: Routledge and Kegan Paul. Latour, Bruno. 1996. “On Interobjectivity”. Mind, Culture and Activity 3 (4): 228–245. Lebaron, Curtiss D. and Streeck, Jürgen. 1997. “Built space and the interactional framing of experience during a murder investigation”. Human Studies 20: 1–25. Lefebvre, Henri. 1991. The Production of Space. Oxford: Basil Blackwell. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Manicas, Peter T. 1993. “The absent ontology of society: Response to Juckes and Barresi”. Journal for the Theory of Social Behaviour 23 (2): 217–228. Martin, James R. 1992. English Text: System and Structure. Amsterdam: John Benjamins. McElhinny, Bonny. 1997. “Ideologies of public and private language in sociolinguistics”. In Gender and Discourse [Sage Studies in Discourse], R. Wodak (ed.), 106–139. Newbury Park: Sage Publications. McInnes, David. 1998. Attending to the Instance: Towards a Systemic-Based Dynamic and Responsive Analysis of Composite Performance Acts. Unpublished Ph.D. Thesis, Department of Performance Studies and Department of Linguistics, University of Sydney. McKenna, Brian and Graham, Phil. 2000. “Technocratic discourse: A primer”. Journal of Technical Writing and Communication 30 (3): 219–247. Mehan, Hugh. 1987. “Language and power in organizational process”. Discourse Processes 10 (4): 291–301. Mehan, Hugh. 1993. “Beneath the skin and between the ears: A case study in the politics of representation”. In Understanding Practice: Perspectives on Activity and Context, S. Chaiklin and J. Lave (eds.), 241–268. Cambridge: Cambridge University Press. Ochs, Elinor, SchegloŸ, Emanuel A. and Thompson, Sandra A. (eds.). 1996a. Interaction and Grammar [Studies in Interactional Sociolinguistics 13]. Cambridge: Cambridge University Press. Pomerantz, Anita and Fehr, B. J. 1997. “Conversation analysis: An approach to the study of social action as sense making process”. In Discourse as Social Interaction, Vol. 2, T. van Dijk (ed.), 64–91. London: Sage Publications. Sacks, Harvey E., SchegloŸ, Emanuel A. and JeŸerson, Gail. 1974. “A simplest systematics for the organization of turn-taking for conversation”. Language 50 (4): 696–735. Schatzki, Theodore R. 1996. “Practiced bodies: Subjects, genders and minds”. In The Social and Political Body, T. Schatzki and W. Natter (eds.), 49–78. London: The Guilford Press. Schatzki, Theodore R. and Natter, Wolfgang (eds.). 1996. The Social and Political Body. London: The Guilford Press. SchegloŸ, Emanuel A. 1972. “Sequencing in conversational openings”. In Directions in Sociolinguistics: The Ethnography of Communication, J. J. Gumperz and D. Hymes (eds.), 346–380. New York: Holt, Rinehart and Winston.

Putting SchegloŸ’s principles and practices in context

SchegloŸ, Emanuel A. 1980. “Preliminaries to preliminaries: ‘Can I ask you a question’”. Sociological Enquiry 50 (3–4): 104–152. SchegloŸ, Emanuel A. 1984a. “On some gestures’ relation to talk”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 266–296. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1984b. “On some questions and ambiguities in conversation”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 28–52. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1987a. “Between micro and macro: Contexts and other connections”. In Micro-Macro Link, J. Alexander, B. Giesen, R. Münch and N. Smelser (eds.), 207–234. Berkeley, CA: University of California Press. SchegloŸ, Emanuel A. 1987b. “Some sources of misunderstanding in talk-in-interaction”. Linguistics 25: 201–218. SchegloŸ, Emanuel A. 1990. “On the organization of sequences as a source of ‘coherence’ in talk-in-interaction”. In Conversational Organization and its Development, B. Dorval (ed.), 51–77. Norwood, NJ: Ablex. SchegloŸ, Emanuel A. 1991a. “Conversation analysis and socially shared cognition”. In Perspectives on Socially Shared Cognition, L. Resnick, J. Levine and S. Teasley (eds.), 150–171. Washington, D. C.: American Psychological Association. SchegloŸ, Emanuel A. 1991b. “Re¶ections on talk and social structure”. In Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis, D. Boden and D. H. Zimmerman (eds.), 44–70. Cambridge: Polity Press. SchegloŸ, Emanuel A. 1992a. “In another context”. In Rethinking Context: Language as an Interactive Phenomenon, A. Duranti and C. Goodwin (eds.), 193–227. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1992b. “Introduction”. In Sacks’ Lectures on Conversation, Vol. I, G. JeŸerson (ed.), ix-lxii. Oxford: Basil Blackwell. SchegloŸ, Emanuel A. 1992c. “On talk and its institutional occasions”. In Talk at Work: Interaction in Institutional Settings [Studies in Interactional Sociolinguistics 8], P. Drew and J. Heritage (eds.), 101–134. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1995. “Discourse as an interactional achievement III: The omnirelevance of action”. Research on Language and Social Interaction 28 (3): 185–211. SchegloŸ, Emanuel A. 1996a. “Conªrming allusions: Toward an empirical account of action”. American Journal of Sociology 102 (1): 161–216. SchegloŸ, Emanuel A. 1996b. “Turn organization: One intersection of grammar and interaction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 52–133. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1997. “Whose Text? Whose Context?”. Discourse and Society 8 (2): 165–187. SchegloŸ, Emanuel A. and Sacks, Harvey. 1973. “Opening up closings”. Semiotica, VIII (4): 289–327.

89

90

Discussing Conversation Analysis

SchegloŸ, Emanuel A., Ochs, Elinor and Thompson, Sandra A. 1996. “Introduction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 1–51. Cambridge: Cambridge University Press. Segerdahl, Pär. 1998. “Scientiªc studies of aspects of everyday life: The example of conversation analysis”. Language and Communication 18 (4): 275–323. Sharrock, Wes W. and Watson, D. Rodney. 1984. “What’s the point of ‘rescuing motives’?”. British Journal of Sociology 35 (3): 435–451. Streeck, Jürgen. 1996. “How to do things with things”. Human Studies 19: 365–384. Taylor, Charles. 1995. Philosophical Arguments. Cambridge, Mass.: Harvard University Press. Taylor, Talbot J. and Cameron, Deborah. 1987. Analysing Conversation: Rules and Units in the Structure of Talk. Oxford: Pergamon Press. Thibault, Paul J. 1994. “Text and/or context: An open question”. Editorial for The Semiotic Review of Books, P. Bouissac (ed.), 5 (2): 10–12. Willmott, Hugh. 1994. “Theorizing human agency: Responding to the crises of (post)modernity”. In Towards a New Theory of Organizations, J. Hassard and M. Parker (eds.). London: Routledge.

Chapter 5

Conversation analysis as rigorous science* Pär Segerdahl

1.

The phenomenological pathos of truth in SchegloŸ’s work

“My own mind does not work well in the phenomenological idiom,” Emanuel A. SchegloŸ says in the interview made for this volume (this volume: 35). Phenomenology, however, was more than an idiom. It was to a large extent a reaction against the unconditional faith in the sciences and the tendency to think that the perspectives of the sciences mirrored the basic structures of the world. Edmund Husserl claimed that this faith had resulted in a cultural crisis. Considering some of the remarks that SchegloŸ makes in the interview, his work in conversation analysis has its motivation in a reaction that is at least comparable with Husserl’s. Consider, for instance, SchegloŸ’s remark that: […] often our academic colleagues are motivated by considerations other than the demonstrable relevance to the participants. In particular, they are most often motivated by the traditional or contemporary preoccupations of their discipline, by its current theoretical commitments or controversies, by the methodological paradigms currently in favor or seeking to be, by the apparent political tenor or implications of various directions of work, and the like. These often have as much or more to do with the situation of inquiry for the investigators than with the situation of interaction for the participants (this volume: 26).

SchegloŸ’s writings make a strong impression on the reader. His work appears to be motivated by severe demands of intellectual honesty. Recall how he describes his early attempts at ªnding an honest way of doing sociology: I just couldn’t do the work if I already felt that I knew it — the genre — was wrong, I couldn’t practice doing it, I just couldn’t (this volume: 35).

It is not di¹cult to hear in this confession a moral reaction connected with the ideal of truthfulness, and it is this aspect of SchegloŸ’s work that interests me,

92

Discussing Conversation Analysis

and that makes it meaningful to compare his notion of conversation analysis with phenomenology. Normal scientiªc work tends to be bound by loyalties to already established aims, perspectives, methods and results. SchegloŸ’s work in conversation analysis, however, must be described as a more intense moral phenomenon, since he reacts against the complacent scientiªc morality, experiencing it as being itself a moral problem, and considering the fresh empirical methodology of conversation analysis as the living solution to this problem. SchegloŸ’s conversation analysis is therefore not an ordinary form of scientiªc work, something that SchegloŸ acknowledges at the end of the interview, when he remarks that studying conversation analysis: […] seems often to be quite diŸerent from simply taking on a new academic subject and absorbing it like one has absorbed the previous ones. […] I can’t tell you the number of students who’ve told me later on how this work has changed their lives, that they see the world in a whole diŸerent way, that they found themselves with a commitment to working that’s just of a diŸerent order — that’s grounded in the world in a diŸerent way — than their previous academic commitments had been […] (this volume: 49–50).

The most important phenomenological trait that is discernible in SchegloŸ’s work is perhaps the one that he describes as: the eŸort to make the terms and practices of research as much responsible as possible to the demonstrable features of the data (this volume: 26).

SchegloŸ’s emphasis on detailed examination of empirical data is not an habitual adherence to positivism, but an active and innovative attempt to ground research in the phenomenon under study. The pathos reminds one of Husserl’s “zur Sachen selbst,” and to some extent, perhaps, of Ludwig Wittgenstein’s emphasis on close study of examples, where the examples are somehow allowed to speak for themselves. The diŸerence between the real world and its representations has become evident to SchegloŸ, and his emphasis on “the demonstrable relevance to the participants” is an attempt to put all representations aside, and instead unravel for us “actual reality, the reality in which we live, move, and are” (to use Husserl’s words, 1965: 140).1 I ªnd this phenomenological stance expressed in SchegloŸ’s determination to “show that the raw data of everyday conversational interaction can be subjected to rigorous analysis,” and in his conviction that conversation analysis indicates “the possibility of direct analysis of the ‘stuŸ of everyday life’.”2 There is in SchegloŸ’s emphasis on “rigorous analysis”

Conversation analysis as rigorous science

and “direct analysis” something that reminds one of Husserl’s frequent declarations that he has discovered a form of investigation that opens up the most immediate aspects of experience for detailed examination. In “Philosophy as Rigorous Science,” he writes triumphantly: That one can here investigate and make statements, and do so on the basis of evidence, adapting oneself to the sense of this sphere of ‘experience,’ is absolutely evident (Husserl, 1965, p. 109).

“That one can here investigate”: what vital importance the word ‘Untersuchung’ must have had for Husserl. In my view, some of that importance is expressed in SchegloŸ’s use of ‘rigorous analysis.’ Moreover, when SchegloŸ says, “I would like my own work to be motivated virtually exclusively by what is demonstrably relevant to the participants,” he seems to emphasize what Husserl calls adapting oneself to the sense of a sphere of experience. SchegloŸ’s work, it seems to me, is characterized by a phenomenological pathos of truth.

2.

The phenomenological architecture of SchegloŸ’s work

From SchegloŸ’s conversation analytical perspective, ordinary social science is perceived as naïve to the extent that it takes social reality for granted, and merely attempts to mirror this given structure in a theoretical framework. In opposition to this stance, SchegloŸ concludes one of his articles, “Discourse as an Interactional Achievement,” with the following statement: If certain stable forms appear to emerge or recur in talk, they should be understood as an orderliness wrested by the participants from interactional contingency, rather than as automatic products of standardized plans. Form, one might say, is also the distillate of action and/in interaction, not only its blueprint. If that is so, then the description of forms of behavior, forms of discourse (such as stories) included, has to include interaction among their constitutive domains […] (SchegloŸ 1982: 89).

SchegloŸ writes that familiar forms of behavior, such as telling a story, include other forms of behavior — i.e., interaction — “among their constitutive domains,” as if two levels of human action were involved in a conversation. At the superªcial level, conversation is the everyday activity as we already are aware of it and its characteristic features. At the deeper level, conversation is the activity of ‘achieving’ those familiar features interactionally. Although this more basic form of action is, in some sense, always before our eyes, it is unnoticed by

93

94

Discussing Conversation Analysis

everyone except by the trained conversation analyst: it is, in the words of Heritage and Watson (1979), “seen-but-unnoticed.” The essential feature of the fundamental, interactional aspect of human action, according to SchegloŸ, is its sequential nature. To ‘produce’ an utterance in a conversation is not simply to utter it; it is to utter it as a response to a previous utterance, and as a constraint on subsequent contributions to the conversation. It is, basically, this sequential organization of conversational interaction that allows the claim that conversation analytical studies are guided by “what is demonstrably relevant to the participants.” The idea seems to be that when a participant in a conversation answers a possible question, he displays his present understanding that the previous utterance was a question both to the ªrst speaker and to the professional analyst. The third turn in such sequences is of vital importance, for here the ªrst speaker may correct the second speaker’s ‘analysis,’ and if the ªrst speaker continues without objecting, this supports the professional’s analysis of the ªrst utterance as being a question. The sequence of utterances just mentioned — a question followed by an answer — is what SchegloŸ would call an adjacency pair (other examples of adjacency pairs are greeting-greeting and invitation-acceptance/declination). Adjacency pairs are often adduced as one of the most central examples of how one action in a sequence may constrain the next action. Given the production of the ªrst-pair part of an adjacency pair, its speaker should stop, and the next speaker should produce a second-pair part of the same adjacency pair. The production of a question, for instance, requires that the next speaker produces an utterance which is recognizable as an answer to the question. Adjacency pairs are sometimes described by SchegloŸ as devices used by participants in conversations to produce their own behavior and deal with that of other participants in a methodical manner. How is such an assumption justiªed on conversation analytical grounds, that is, how can SchegloŸ demonstrate that adjacency pairs are not only e¹cient research tools, but also ‘relevant to the participants’? An additional analytical resource of conversation analysis, described by Heritage (1988), is deviant case analysis. One might, for instance, analyze data in which the production of the ªrst-pair part of an adjacency pair is not immediately followed by the production of an appropriate second-pair part. Common features of such data are: the producer of the ªrst-pair part does not continue speaking, and his contribution is followed by a silence which can be attributed to the participant addressed in the ªrst-pair part; the producer of the ªrst-pair part makes repeated attempts to elicit a second-pair part; the participant addressed in the ªrst-pair part utters expressions of hesitation, or

Conversation analysis as rigorous science

explains why he cannot produce a (preferred) second-pair part. Considering such data, it appears that the question-answer adjacency pair is not only a convenient way of formulating an observed regularity (that answers follow questions), for the data seem to show how the participants themselves treat the adjacency pair as a norm of conduct: their own behavior, and their inferences about other participants’ conduct, appear to be oriented to the rule that questions require answers from their addressees. The adjacency pair is for SchegloŸ not only an e¹cient research tool, then, it is taken by him as operative in the conversation itself. Interaction according to adjacency pair rules could be described as an aspect of “how the conversation makes itself happen and how it provides itself with those features that it exhibits,” to paraphrase a formulation by Sharrock and Anderson, which accentuates that “interaction” in conversation analysis is used in the sense of constitution.3 The familiar feature that a silence following a question is heard as a particular participant’s silence, for instance, would be analyzed in terms of the adjacency pair rule for questions and answers, which allocates this participant as the next speaker. The analysis of this feature seems to describe its interactional achievement — or constitution. SchegloŸ evidently makes a fundamental distinction between his own work and ordinary social and linguistic theory. As conversation analyst, SchegloŸ is not only aware of the constitutive relation of sequentially organized human interaction to everyday conversation as we ordinarily experience it, he can even examine this relation in detail, in a rigorous empirical way. This resembles the separation that Husserl made between phenomenology and the sciences, and one may therefore speak of SchegloŸ’s conversation analysis as an attempt to be rigorous science in a sense that is at least analogous to Husserl’s “Philosophie als strenge Wissenschaft.”

3.

Must a formal device make us do what we do or hear what we hear?

SchegloŸ’s aim is to achieve a technical understanding of conversation, where “technical,” judging from what he says in the interview, implies a more fundamental or rigorous understanding: As a member of society, I perfectly well understand about social classes and all the rest of a moderately sophisticated citizenship; but the fact that I understand and see the world — or can see the world — in those terms as a member of a society is not the same as qualifying all those ways of seeing it technically, let alone subscrib-

95

96

Discussing Conversation Analysis

ing to it and underwriting it as part of one’s technical apparatus for understanding the world. In fact, it’s just the opposite. The more they recommend themselves to my vernacular understanding, the more suspect they ought to be for me as part of my technical apparatus. The common or vernacular culture is, after all, a sort of “propaganda arm” of the society, serving to undergird the cultural component of the more or less smooth functioning of the society itself, not to advance or enhance a rigorous understanding of society (this volume: 44).

This statement expresses a further aspect of SchegloŸ’s desire to achieve truthfulness in his work: the ideal of truth can be realized only through the development of a technical apparatus for understanding the world. In the case of the study of conversation, this apparatus consists of the sequential perspective, and the technical notions, adjacency pair, preference organization, turn-taking, presequence, repair, account, and so forth. Let me ªrst of all make clear that I consider the development of this technical apparatus a tremendous intellectual achievement, and that its spreading in linguistic circles (not least in Scandinavia) is in the highest degree motivated. The conversation analytical literature makes us aware of previously neglected aspects of our actual language use, in contrast to common expectations having to do with the prevailing grammatical perspective in linguistics — in fact, in our entire literate culture. Conversation analysis should be an important eye-opener to linguists and to everyone whose education has made him more or less blind to features of our linguistic behavior as it is when we do not write, read or speak according to a grammatical standard. The conversation analytical perspective not only presents these features in an instructive manner, making it easier to establish and accept them as facts, it also facilitates the description of many of these features in a surprisingly detailed way.4 As I want to show in the remainder of this paper, however, the most fundamental features of SchegloŸ’s technical apparatus do not correspond to the actual facts of conversation. His technical notions can only be understood as research tools used by conversation analysts in their professional work, and with good results. These tools are not, however, ‘relevant to the participants.’ The appearance that it can be demonstrated that participants in conversations orient their behavior towards these normative constructions is, as I hope the investigation below shows, an illusion. If I am right, SchegloŸ’s problem of truthfulness is solved by distinguishing the tools of conversation analysis from our actual conversations, rather than by trying to prove that they coincide.5 Let us look more closely at an example of how SchegloŸ tries to demonstrate that one of his technical descriptions is ‘relevant to the participants.’

Conversation analysis as rigorous science

Consider the fact that we do not repeat a summons without limitation when it is not answered, and are annoyed when children repeat our name until we attend to them. That fact could be described in terms of a rule. In SchegloŸ (1986) such a “terminating rule” is assumed, but not merely as a description of the fact I mentioned, but as a deeper explanation of the fact that it is a fact, so to speak. The idea seems to be that this particular fact would not belong to the world of facts, if the rule was not “used by members of the society to limit the number of repetitions of an S”: […] that some terminating rule is normally used by adult members of the society can be noted by observing their annoyance at the behavior of children who do not employ it (SchegloŸ 1986: 365).

What we can begin to suspect when we consider this simple example, is that the rule which describes the mentioned feature of summonses can be conceived as ‘demonstrably relevant to the participants’ only if this kind of relevance is in a certain sense demanded. The implicit presumption seems to be that people who converse would not act as they do unless some formal device or rule made them act as they do. How else could the trivial fact that we are annoyed when children repeat a summons without limitation evidence the existence of a rule “used by members of the society”? It was with some hesitation that I previously compared SchegloŸ’s work with Wittgenstein’s close study of examples, for in many of the philosopher’s investigations, the aim is to show that, if we consider the facts closely, we see that nothing must make us do what we do: And the mistake which we here and in a thousand similar cases are inclined to make is labelled by the word ‘to make’ as we have used it in the sentence ‘It is no act of insight which makes us use the rule as we do,’ because there is an idea that ‘something must make us’ do what we do (Wittgenstein 1969: 143).

Wittgenstein does not deny that sometimes, or very often, something makes us do what we do. What he is ªghting is the compulsion to think that something must make us do what we do: There is a kind of general disease of thinking which always looks for (and ªnds) what would be called a mental state from which all our acts spring as from a reservoir. Thus one says, ‘The fashion changes because the taste of people changes.’ The taste is the mental reservoir. But if a tailor to-day designs a cut of dress diŸerent from that which he designed a year ago, can’t what is called his change of taste have consisted, partly or wholly, in doing just this? (Wittgenstein 1969: 143).

97

98

Discussing Conversation Analysis

The tendency to look for mental causes is not characteristic of SchegloŸ’s work. He does, however, treat the rule indicated above as if it were a kind of interactional reservoir of behavior. The manner in which he uses our annoyance when children repeat a summons too many times as if this were obvious evidence of such “reservoirs,” indicates that he takes it for granted that there must be some reservoir: a device, a procedure, or a rule — otherwise we would presumably not act at all, or behave chaotically, without meaning and understanding. It would be comical if we approached someone and said, “Have you noticed that we use a terminating rule for summonses?”, “Why, what do you mean?”, “Well, we are annoyed when children repeat a summons without limitation.” Still, this kind of proof procedure is very common in linguistic theorizing, not to mention cognitive science, and in SchegloŸ’s work it is an important component of what he means by calling conversation analysis “rigorous,” or “relevant to the participants.” Our everyday forms of behavior are presented in these proofs as if they were crackled, they seem fragile and in need of support. Fortunately, help is not far away, for since there must be an underlying support (otherwise there would be no orderly forms of human behavior), we can scrutinize the details in the crackle (e.g., our annoyance when a child repeats a summons ten times), and they will indicate to us the devices used to achieve and sustain those fragile forms interactionally. Look, there they are, just beneath the surface of our everyday behavior: the tools of conversation analysis are operative in our own conversations. We cannot converse orderly without, in some sense, doing conversation analysis!6 When an imposed intellectual perspective is so intensely felt to constitute the essence of the facts, there arises a tendency to paraphrase everyday descriptions into the technical language of the perspective. The result has the appearance of a jargon, but it is not meant as a jargon, but as the discourse most in line with the fundamental aspect of things highlighted by the new perspective. In an attempt to illustrate the concept “conditional relevance” (summonses make answers relevant), SchegloŸ gives the following example of what he calls a “no answer-no person” inference: A person returning home seeking to ªnd out if anyone else is already there may call out the name of his wife, for example, and upon not receiving an answer, may typically take it that she is not home or, while physically home, is not interactionally ‘in play’ […] (SchegloŸ, 1986: 368).

Describing a person who is asleep, for example, as “not interactionally in play,” or as “interactionally absent,” might be an economical manner of indicating

Conversation analysis as rigorous science

the many ways in which a person who is physically present still is unable to participate in a conversation. The person may be reading an extremely interesting book, solving a di¹cult mathematical problem, or she may have turned oŸ the hearing aid. These diŸerent factors typically explain why someone does not answer a summons: “she cannot answer because she is absorbed by the new computer game.” The picture of not being interactionally in play, however, seems to mean more than these everyday explanations do. When we hear the technical phrase, we tend to have an ahah experience, as if it explained the explanations. The person who is “interactionally absent” appears to be insensitive to the technology of conversation, and that is why she cannot answer when she is absorbed by the computer game.7 The regulated conversation analytical outlook is employed here not only as an imposed intellectual perspective, useful and instructive for certain practical purposes. It has slipped into the facts themselves, it has become their ‘body.’ This is not a coincidence: it is the result of attaching the phenomenological pathos of truth to the perspective of a scholarly form of work. SchegloŸ’s feeling of dishonesty connected with his early attempts to pursue ordinary sociological research was, I suppose, the feeling that he was imposing intellectual schemes, settled beforehand by interests external to the facts under study, as if he was thereby hiding the facts rather than revealing them. SchegloŸ’s pathos of truth demands a perspective that is so true to the facts that it is scarcely a perspective at all, since it coincides with the facts: a literally true perspective. SchegloŸ claims that rigorous empirical work can live up to this ideal, and demands that a proper conversation analytical study can be shown to be relevant to the participants. There is an empirical procedure of demonstrating that relevance, but that procedure appears to be contingent upon the presumption, mentioned above, that people who converse would not do what they do, and would not hear what they hear, unless some scientiªcally analyzable factor (a procedure, a device etc.) made them do what they do, or hear what they hear — so that the tools employed in the analysis can be considered constitutive of the conversation, too. This presumption indicates how the phenomenological architecture works as a way of interpreting the techniques of conversation analysis as if they were literally true: the relation “is constitutive of” mediates between the technical perspective and ordinary behavior, and makes it possible to schematize the imposed perspective as if it were the seen-but-unnoticed fundamental aspect of these forms of behavior. The idea of extremely rapid, unconscious calculations in cognitive science appears to have the same archi-

99

100 Discussing Conversation Analysis

tectonic function of establishing a link between the imposed perspective and life itself — as if the perspective was not imposed at all, but rather, in some sense, already there. SchegloŸ is more cautious, more “rigorous,” when he establishes this link. A distinguishing quality of SchegloŸ’s work is the amount of empirical data it contains. “Papers take the form they do to maximize the opportunity for readers to immediately juxtapose every bit of description with the data of which it claims to be a description,” SchegloŸ explains in the interview (this volume: 41). No doubt, substantial beneªts can be (and are) drawn from this important practice, but it does seem to have a seductive aspect too. It appears, to a certain extent, to be a way of cultivating the experience of the literal truth of the imposed perspective. If you look at the data with the presumption that some formal device must make people do what they do, then you have already made up your mind to interpret the data as evidence of such devices, and you tend to react: “Look, the participants actually treat the adjacency pair rule as a norm of their conduct!” Each analysis, together with the data it describes, constitutes (in part) a metaphysical proof procedure of a kind that most treatises in linguistic theory provide only in the lofty part of the introduction, to explain to the reader in what frame of mind he should continue the reading.8 I believe that one reason why SchegloŸ’s conversation analysis has attracted many linguists is that formal grammarians have stopped cultivating the experience of the literal truth of their perspectives. It is not so much that one has discovered the limitations of these perspectives: rather, the perspectives are no longer stimulating, and one yearns not only for new instructive outlooks on language, but also for fresh proof procedures that intensify the feeling of being in contact with the essence of linguistic phenomena. SchegloŸ’s conversation analysis provides such metaphysical proof procedures in abundance, with the agreeable taste of rigorous empirical inquiry. The manner in which data appear in articles is an important aspect of these metaphysical proof procedures.9 If the experience of truth connected with linguistic theorizing, or with the speculations of cognitive scientists, is an occasional “Is that how it is,” the experience of truth created by SchegloŸ’s papers is an almost constant “Look, that is how it is.” Let us return to SchegloŸ’s notion of the question-answer adjacency pair as a device employed by participants in conversations when they ask and answer questions. The idea that we employ this device when we converse is the

Conversation analysis as rigorous science 101

idea that conversation is constituted by a more basic form of activity: the activity of achieving questions and answers interactionally by orienting our behavior towards the adjacency pair rule for questions and answers. This architecture is expressed in formulas that are characteristic of the conversation analytical literature: asking and answering questions is described as “doing questioning and answering,” laughing is described as “doing laughter,” and ordinary conversation is described as “doing ordinary conversation.” The result of perceiving this architecture is that our attention is drawn towards its most basic level, the making of a conversation. Our primary task in conversation now appears to be that of taking a number of subtle measures which are necessary if our behavior is to become a characteristic conversation, and not something else — as if such derailment was a hidden threat to the smooth functioning of everyday life activities. Why this tendency to interpret the tools of conversation analysis as if they were operative in our actual conversations? Is it not su¹cient that they are good tools for analyzing conversations? That this new form of analysis is superior in many respects is convincingly shown by SchegloŸ (1984), where he presents data which demonstrate the immense di¹culty, or even the impossibility, of analyzing questions in naturally-occurring conversations by focussing on the linguistic form of isolated utterances. If you want to analyze the various contributions to a conversation, you have all to gain by focussing on how these contributions are positioned in a sequence of actions: A: Why don’t you come and see me sometimes B: I would like to

A’s contribution has the linguistic form of a question, but when we consider the sequence of utterances, we notice that it is not of the question-answer kind, but rather of the invitation-acceptance kind. If you are on the lookout for questions in data of naturally-occurring conversations, then, look for the adjacency pair instead of the interrogative form of isolated utterances, at least if you are interested in the act of asking someone something, rather than the linguistic formula. Linguists and ordinary language philosophers have tended to treat the linguistic form of individual utterances as the determining factor of linguistic meaning and communication, and SchegloŸ rightly objects that “in the real world of conversation, it is not. Most centrally, an utterance will occur someplace sequentially.”10 His similar remark in the interview that,

102 Discussing Conversation Analysis

if one is committed to understanding actual actions (by which I mean ones which actually occurred in real time), it is virtually impossible to detach them from their context for isolated analysis […] (this volume: 39)

is understandable when we consider the simple example above. Sequential placement is the key to the status of an utterance if you do conversation analysis, and a good key too. Whenever you ªnd an utterance in your data that can be heard as a question, another utterance is immediately relevant to you as analyst, namely, one that can be heard as an answer to the question. When SchegloŸ describes this methodological procedure, however, he presents it as if it were a fundamental assumption about the conversation, rather than a principle that professional analysts have freely assumed as an imposed guideline of work: The tool I shall use initially is one based on, and fundamental to, a great deal of prior work in conversation analysis. It is that coparticipants in conversation operate under the constraint that their utterances be so constructed and so placed as to show attention to, and understanding of, their placement (SchegloŸ 1984: 37).

SchegloŸ describes his tool as if it was not a tool at all, but rather a general fact about conversations. He speaks from within the intellectual perspective of conversation analysis, as if it was impossible to transcend this and similar regulated outlooks and see them as perspectives. Whatever characterizes work within the perspective is interpreted as a fundamental trait of the facts under study. And since the conversation analytical techniques of analysis focus on the sequential placement of utterances, it seems that people who converse must do so too: our primary task when we converse, according to SchegloŸ, is to construct and place our utterances so as to “show attention to, and understanding of, their placement.” The conversation analyst’s primary task is to construct his analyses so as to show attention to the sequential placement of utterances. The requirement is that he can analyze our utterances as questions and answers, invitations and acceptances, according to the adjacency pair rules guiding his analysis. Must he justify this fruitful technique of analysis by assuming that people who converse ‘place’ their utterances so as to show the same kind of attention to the conversation analytical outlook? If that is so, then we will have to put up with ªnding our primary tasks when talking with each other vary with the intellectual perspectives imposed on our behavior in various university departments. Was not this just the situation which not only Husserl, but SchegloŸ too, found unacceptable and dishonest?

Conversation analysis as rigorous science 103

But what about deviant case analysis? Does it not demonstrate that the conversation analytical perspective is literally true, in the sense that the rules guiding intellectual analysis of data coincide with procedures employed by the participants? Consider the following example, where a teenager B gives a younger child A the cold shoulder by not answering A’s questions: A: what are you doing with my bike (1.4) A: hey, what are you doing (1.9) A: can I borrow yours then B: no way

The child A is at ªrst in a weak position, since he is incapable not only of making B stop using his bike, but also of making B answer. A is ingenious, however, and asks a question that he knows B will answer, although the answer will be in the negative. By asking a question that B is almost compelled to answer, and in fact does answer, A succeeds in putting an end to the game that B is playing. A has made his position stronger, and perhaps he is now able to make B stop using his bike too. Does not this sequence demonstrate that A actually treats the rule that questions require answers as a relevant rule of conduct? And how can B’s silence be interpretable as “giving A the cold shoulder” unless the participants juxtapose the fact that B is present and the fact that the adjacency pair rule allocates B as the next speaker? How can B give A the cold shoulder at all, unless he utilizes these facts in juxtaposition as a resource? He cannot maltreat A by, for instance, being silent while A is still busy asking one of his questions; or by being silent after A has replied “bye” to a parting greeting by B; or by being silent in a situation where it is not obvious that B can hear A’s questions. It appears that B must be very observant when he decides the placement of his silences, and the game that A ªnally manages to put an end to seems to utilize the adjacency pair rule that questions require answers from their addressees. When I consider this example closely, however, I must confess that all I can perceive is that A requires an answer when he does not get one, and that B gives A the cold shoulder by not answering. Nothing in the example indicates that A would not require answers unless he oriented his behavior towards a rule according to which questions require answers, and nothing indicates that this rule provides the basis for our hearing that B is giving A the cold shoulder.

104 Discussing Conversation Analysis

Wittgenstein said that the di¹culty in philosophy is “to say no more than we know.”11 This remark seems to me to apply to the study of conversation too.12 The appearance that we actually can perceive that people who converse orient their behavior towards adjacency pair rules arises when we consider their behavior as if it would disintegrate at any moment, unless they constantly employed various rules or devices making their behavior orderly — and thereby analyzable according to the same rules. This is an imposed way of looking at conversational behavior, and it is not di¹cult to guess the paradigm of this conception of the essence of conversation. It is the situation of intellectual analysis of empirical data according to formal rules. Assume that you have a collection of data of naturally occurring conversations, and that you want to study sequences where one participant gives another participant the cold shoulder. You moreover want to do this work in a systematic fashion, so you try to formulate for yourself the determining factor of these sequences. Assume that you ªnd sequences such as the one above — what is their determining factor? It is not easy to ªnd a satisfactory answer that can guide you in your formal and systematic search for further instances of this form of maltreatment. B’s silence, for instance, cannot be the determining factor, for normally when we are silent we are not giving someone the cold shoulder, and B is not acting neglectfully when he is silent while A is still speaking. Neither is B’s silence after A has spoken the determining factor, otherwise a conversation could not be ended in any other way than by maltreating the last speaker. It is essential that the previous utterance by A is a question, a summons, an invitation etc., but this feature must not only be provided with the amendment that B is present and is able to hear and respond to A’s utterances, it must also be expressed in more general and formallydeªned terms. A formal account of this kind can be achieved by using the concepts adjacency pair and conditional relevance. If we are successful, our account will work as a research tool that can be employed systematically to trace and formulate in detail relevant features of sequences where this kind of activity occurs. The value of such a tool is evident when we consider the di¹culty of pinpointing the determining factor of these sequences. The intellectual attempt to establish a formal order is constantly on the brink of failure, and is constantly in need of guidance. A satisfactory formal account will constitute technical knowledge which can be employed over and over again to produce sure comments on conversational data: it will indeed make our judgements superior to, and more orderly than, the scattered remarks made by someone

Conversation analysis as rigorous science 105

unprepared for this task. But is this valuable tool utilized when we give someone the cold shoulder, too? The formal account will indeed provide a systematic basis for intellectual analysis, but does it provide a basis for A’s hearing B’s silence as a kind of maltreatment too? SchegloŸ’s answer to these questions is evidently a¹rmative: […] the resource that members of the society draw upon in doing such activities as ‘cold shouldering,’ ‘insulting,’ ‘sulking,’ etc., involves the joint observability of the physical presence, social presence (that is, consciousness and awareness) and the absence of an answer to a summons, indicating or claiming unavailability for interaction (SchegloŸ 1986: 373).

The technology of intellectual analysis is no longer doing work, and the data are used to illustrate the technology, as if it were inherent in our conversations. Wittgenstein once remarked: “The confusions which occupy us arise when language is like an engine idling, not when it is doing work.”13 Similar confusions arise when the perspective of a scholarly form of work is like an engine idling, not when it is doing work.

4.

Concluding remarks

SchegloŸ has contributed to the development of extremely valuable techniques for describing what we do when we converse, a completely new and original outlook on linguistic interaction that reveals the limitations of previous ‘grammatical’ perspectives on language use. My doubts concern the claim that these tools can be used to describe not only what we do when we converse, but above all what makes us do what we do: certain ‘devices,’ ‘methods’ or ‘rules’ that are supposed to coincide with the tools of intellectual analysis. The ‘rigorous’ part of SchegloŸ’s work in conversation analysis, his attempt to demonstrate that the sequential outlook on conversational behavior is relevant to the participants, is, if I am right, the speculative part of his methodology: it is the part which is not relevant to the participants. SchegloŸ’s pathos of truth — his reaction against imposed perspectives and the demand for a literally true perspective — leads to its own opposite when combined with professionalized and technical research activities. The result is a more thoroughgoing reconstruction of the situation that was originally found unacceptable.

106 Discussing Conversation Analysis

Notes * I wish to thank Sören Stenlund and Sven Öhman for valuable comments on a draft of this paper. 1. Roy Harris expresses a similar kind of sensitivity when he remarks that “the pronouncements of linguistic theorists make little if any contact with the communicational world of daily life” (Harris 1997: 242). Harris’ integrational linguistics can be interpreted as his reaction to the contrast of which Husserl speaks. The way out of the unacceptable situation, according to integrational linguistics, is said to consist in a form of theoretical self-awareness: “You cannot seriously set up in the business of ‘describing language(s) scientiªcally’ unless you are prepared to come clean about the status of your own descriptive language” (Harris 1997: 276). And: “the integrationist will insist that an important priority for linguistics is to re-examine and re-evaluate the discourse of linguistics itself” (Harris 1997: 298). 2. SchegloŸ 1986: 349. 3. Sharrock and Anderson 1986: 70. 4. Just consider the way that SchegloŸ, JeŸerson and Sacks (1977) were able to describe what we often would be inclined to dismiss as just careless speech in terms of various forms of “repair.” 5. Similar themes are discussed in Segerdahl (1995, 1998). 6. The confusion of life with imposed intellectual perspectives is so common that one is amazed when one happens to ªnd an exception. In a very original book, The Experiencing of Musical Sound, F. Joseph Smith discusses the di¹culty of analyzing our musical experience of a certain chord (created by Stravinsky) using the techniques of harmonic analysis — he says: “It is very simple: there is no harmonic analysis called for; it is simply a question of how the composer placed his hands on the keyboard. It was this bodily positioning of the hands which gave birth to the complex of sound, and not some theoretical idea that made it possible” (Smith 1979: 178). Recall Wittgenstein’s remark about the tailor’s new taste, which might simply be a question of how he designs a cut of dress, rather than its underlying cause. 7. We sometimes want to object to the utterance, “That cup of coŸee really made me alert,” by saying that it does not mention the determining factor of having a cup of coŸee and perking up, namely, caŸeine and the way this substance aŸects the physiological organism: that is what makes us perk up when we have a cup of coŸee. There are indeed contexts where the latter explanation is the appropriate one, but it would be a mistake to treat it as the literally true explanation, and reject “I’m alert because I just had a cup of coŸee” as vacuous (see Segerdahl 1994). 8. Linguistic creativity and the fact that sounds can transmit thoughts are two standard miracles of language that speculative linguists and cognitive scientists use to convince us why we must take their theoretical constructs seriously as important hypotheses about the nature of language. 9. Talbot Taylor (1992: 227) remarks that conversation analysis “draws rhetorical strength from its association to intellectual practices of empirical justiªcation.”

Conversation analysis as rigorous science 107

10. SchegloŸ 1984: 34. 11. Wittgenstein 1969: 45. 12. Someone greets me by waving, and I wave to greet him back. Am I careful to place my waving immediately after the other person’s greeting, so that what I do can be understood as a greeting too, and as a display of my understanding of his waving? If he passes by in a car, I might be careful to wave back quickly, as soon as I see him raise his hand, otherwise I will fail to greet him back and show him that I recognized that he was greeting me. But is the same kind of consideration relevant in general, for instance, when we recognize each other at diŸerent tables in a restaurant, and greet each other by casually waving and waving back? We know that the consideration is relevant in the particular case, but we want to say that it is relevant in general. 13. Wittgenstein 1974: §132.

References Harris, Roy. 1997. “From an integrational point of view”. In Linguistics Inside Out. Roy Harris and His Critics, G. Wolf and N. Love (eds.), 000–000. Amsterdam and Philadelphia: John Benjamins. Heritage, John C. and Watson, D. R. 1979. “Formulations as conversational objects”. In Everyday Language: Studies in Ethnomethdology, G. Psathas (ed.), 123–162. New York: Irvington Publishers. Heritage, John C. 1988. “Explanations as accounts: A conversation analytic perspective”. In Analyzing Lay Explanation: A Casebook of Methods, C. Antaki (ed.), 127–144. London: Sage Publications. Husserl, Edmund. 1965. Phenomenology and the Crisis of Philosophy, (Q. Lauer, transl.). New York: Harper and Row. SchegloŸ, Emanuel A., JeŸerson, Gail and Sacks, Harvey. 1977. “The preference for selfcorrection in the organization of repair in conversation”. Language 53 (2): 361–382. SchegloŸ, Emanuel A. 1982. “Discourse as an interactional achievement: Some uses of ‘uh huh’ and other things that come between sentences”. In Analyzing Discourse: Text and Talk [Georgetown University Roundtable on Languages and Linguistics 1981], D. Tannen (ed.), 71–93. Washington D. C.: Georgetown University Press. SchegloŸ, Emanuel A. 1984. “On some questions and ambiguities in conversation”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 28–52. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1986. “Sequencing in conversational openings”. In Directions in Sociolinguistics: The Ethnography of Communication, J. J. Gumperz and D. Hymes (eds.), 346–380. Oxford: Basil Blackwell. Segerdahl, Pär. 1994. “Critique of pure capacity. Searle and the background”. Philosophical Investigations 17 (3): 507–535.

108 Discussing Conversation Analysis

Segerdahl, Pär. 1995. “Linguistic theory and actual language”. Language and Communication 15 (1): 31–42. Segerdahl, Pär. 1996. Language Use. A Philosophical Investigation into the Basic Notions of Pragmatics. Houndmills, Basingstoke: Macmillan / New York: St. Martin’s Press. Segerdahl, Pär. 1998. “Scientiªc studies of aspects of everyday life: The example of conversation analysis”. Language and Communication 18 (4): 275–323. Sharrock, Wes and Anderson, B. 1986. The Ethnomethodologists. Chichester: Ellis Horwood Limited / London: Tavistock Publications. Smith, F. Joseph. 1979. The Experiencing of Musical Sound. Prelude to a Phenomenology of Music. London: Gordon and Breach. Taylor, Talbot J. 1992. Mutual Misunderstanding. Scepticism and the Theorizing of Language and Interpretation. Durham and London: Duke University Press. Wittgenstein, Ludwig. 1969. The Blue and Brown Books. Oxford: Basil Blackwell. Wittgenstein, Ludwig. 1974. Philosophical Investigations, second edition, (G. E. M. Anscombe, transl.). Oxford: Basil Blackwell.

Chapter 6

Users’ interpretations at a computer tutorial Detecting (causes) of misunderstandings Pirkko Raudaskoski

1.

Introduction

Conversation analysis (CA) has turned out to be a powerful method to analyze meaning-making or interpretation as a sequential phenomenon in authentic face-to-face or telephone conversations. At the same time as conversation analysts (or ethnomethodologists) have shown sociology the importance of researching situated language use to understand social order, they have given researchers of language meaning and interpretation a sound method to investigate contextual participants’ understandings. Emanuel SchegloŸ is one of the founders of CA, and his interview in this volume displays his commitment to researching further the sequential organization of mundane human conversations. SchegloŸ stresses the importance of continuous investigation of the basic features of everyday face-to-face and telephone interactions. Because of his emphasis, the research interest I pursue in this study, i.e., what makes electronic texts easy or di¹cult to understand, could be regarded as lying clearly outside the ªeld of conversation analysis, or at least looming at its outer margins. However, texts are one type of language use, and their production or reception is often an essential part of a “(situated) activity system” (e.g., Goodwin and Goodwin 1987) or “multi-party interactive activities” (Goodwin 1996), i.e., in collaborative action in which semiotic ªelds are an important resource. But the role of atemporal texts as part of ongoing action has not been of primary interest in the conversation analytic or ethnomethodologic work: While utterances in the course of a conversation are treated in ethnomethodology as inseparable from the ordering the analyst ªnds in the sequencing of utterances, texts are not. They should be (Smith and Whalen 1995: 31).

110 Discussing Conversation Analysis

Smith and Whalen research a fairly short relay between the produced text and its interpretation: they examine how emergency calls are transformed into electronic texts which are again reproduced as talk. In the ªeld of humancomputer interaction, ethnomethodologists and CA researchers have undertaken studies in how people understand language use with a longer relay, namely from the programmer(s)’ s design to the system’s messages or instructions (e.g., Suchman 1987). Also in my data, the text encountered is a highly asynchronous, commercial product that is meant to be used and understood by an unlimited number of user-readers. However, the text clearly aims at a certain eŸect on the reader’s interpretation. In my doctoral dissertation (Raudaskoski 1999), I reported on four case studies in which electronic or paper-based texts were instructing the reader about what to do next. In this paper, I shall introduce one of them: two persons using a computer tutorial. My interest is not in how or whether computer programs can represent conversational interaction (cf. Button et al. 1995), but in how participants in a situation which involves texts interpret them, and how the texts appearing on paper or on a computer screen as static or shown due to the user’s actions, aŸects the user-reader’s interpretation. I ªnd that conversation analysis with its appreciation of interactional details provides an excellent method to investigate the semiotic materiality in these encounters that cannot be categorized strictly as either symbolic or instrumental, thus blurring the division between language and tool use (cf. Engeström and Middleton 1996). So, when for the researchers of the interaction order the most fundamental resource is that “the participants are oriented to constructing their talk and other conduct in detail” (SchegloŸ in his interview: 25), in my research, this same resource is used not only as a way to study how the participants orient to and construct their mutual interaction as an intelligible one, but also as a ‘lens’ to ªnd out how the other signifying elements in the participants’ surroundings are interpreted. The emphasis is therefore not so much on talk, but on “humans in talk-andother-conduct-of-interaction” (Interview: 16), where the ‘other-conduct’ part consists of a complex encounter with a language-using artefact. And, in this paper, the main interest is not in how the speaking and other action is organized [c.f. the topical areas of turn-taking, turn organisation, etc. that SchegloŸ lists in his interview (Interview: 12)], but in what a CA analysis can reveal about user-readers’ interpretations of electronic texts. This speciªc setting means that “what is demonstrably relevant to the participants in the way they construct and understand the conduct which they build together” (Inter-

Users’ interpretations at a computer tutorial

view: 27) is based also on other types of action or activity than talk. Therefore, in the data analysis, those other types are also regarded as an interpretation (i.e. as a turn), and not just of the other participant’s turn, but also of the computer screen. Thus, the concept of turn-taking is expanded from being only a matter of conversational action. With the following case study and analysis, I want to show the strength of CA as a method for ªnding out how text users understand what is going on. The context-dependency of understanding language and action is acknowledged by virtually all the ªelds of reading research. However, I ªnd that CA oŸers a powerful method for getting to the actual process of interpretation as a public phenomenon which is sequential, the sequentiality of which is realized by embodied people. This means that language is not just a symbolic system, but is also heard or spoken, read or written. For instance, when texts are encountered, they are not just read as symbolic systems, but also seen as material objects: texts are not as transient as sounds, and that material feature can in¶uence how they are understood in the sequential activity. The case study will be introduced ªrst, then some important methodological aspects, with the bulk of the paper consisting of expanding the analysis from where a misunderstanding has occurred to the wider interactional context of the encounter. The data analysis goes thus from “a little bit of material” (Interview: 25) to an extended data extract. Thus, rather than searching for patterns of language use, the analysis concentrates more on the indexicality of interpretation, something that has been stressed by ‘radical CA’ or ethnomethodologists (cf. the followers of Garªnkel, whose research SchegloŸ discusses (Interview: 34–37). The outcome of the analysis will help understand the users’ expectations of routine during the human-computer interaction. In the case study, a local abbreviation (i.e. not found outside the context of the text) that two user-readers learned in a word processor tutorial is used successfully. However, at one point, the interpretation of the agreed convention changes to something else for one of the participants. A careful conversation analytical investigation of the local history of the encounter reveals the contingent reasons for the wrong interpretation: (1) the misleading semiotics of the visual details of the text on one screen, and (2) the general orientation of the users to the program as logical in its performance.

111

112

Discussing Conversation Analysis

2.

The electronic text of the case study

The Learning Microsoft Word program (for Word 5.0) dating back to 19891 provided text-based help for users (though some elementary graphics and animation were sometimes used to illustrate a point). In its design, simple interactional devices, including some corrective ones, were implemented. Two fairly novice users explored the program. One (A) had been using Word 5.0, but not the mouse, nor was the participant acquainted with some special features of the program; the other (B) had used text processors before only a little, and had once been shown how Word 5.0 works. As there were two people using the program, meaning-making was a cooperative eŸort, thus there was more to rely on in the analysis than just the script of what happened on the screen and the actions of one user (cf. Frohlich et al. 1994; Suchman 1987). Thus, the situation had two participation frameworks: human-computer interaction and human-human interaction. The encounter was videotaped with two cameras so that both the participants’ gaze direction, gestures and other activities were captured together with what was happening on the screen. This was important in order to undertake an investigation that was grounded ªrmly in the observable details of the material. The observable nature of my data means that not only must any claims about it be detectable in the data, as SchegloŸ emphasizes, but also that the embodied, visual nature of both the participants’ actions and the computer screen is essential for the analysis.2

3.

Adjusting interpretation: Repair work

Correcting (meaning) in talk, i.e. repair work, has been researched extensively in the CA research tradition (SchegloŸ et al. 1977, and many other works by SchegloŸ). In the present data, meaning-adjusting is mainly done between the human participants, but also the tutorial program prompted the user if they did something wrong. There seems to be an order of importance or preference in how people do repairing, and this order is closely connected with who starts the repair (self or other), and who is the target of the corrective work (self or other). Selfinitiation is preferred over other-initiation, and the space for both types of initiation are usually within three turns (the present, at the transition relevance place and in the next turn, or after the next turn) (SchegloŸ et al. 1977).

Users’ interpretations at a computer tutorial

Anything in the conversation can be a repairable. Following from the importance of embodiment, in my data the concept ‘repairable’ extends to the physical actions of the participants, as the movements, clickings or pressings can exhibit their interpretation of the screen or the other participant’s talk. As is evident from his interview, SchegloŸ has a special interest in repair work, and especially in so-called other-initiated repair. The situation of computer use analyzed below is abundant with repairs: the participants repairing each other’s interpretation of the situation, and the system (or, rather its designers), attempting to mimic second-position repair initiations. As mentioned above, the embodied actions of the users are important in the analysis, not just what they say. When they repair each other’s actions or intended actions (as seen by hand or ªnger movements, for example), language use becomes much more ‘external,’ or concerning physical action (which itself is an interpretation of the situation). So, though relying heavily on the conversation analytic premises and ªndings of human interaction that SchegloŸ discusses throughout his interview, my analysis orients strongly to language being used in a cooperative physical activity, rather than the interpretative work being done in a purely conversational encounter. This extension from talk only results in interesting formations of other-initiated repairs, which often orient to the uncertainty of the participants about the relevant next step to take with the program. So, what is being ultimately repaired is the other’s interpretation of the computer screen, which again is displayed in the participants’ talk and actions.

4.

The setting of the case study

This study diŸers from SchegloŸ’s ideals in that the data was not collected in an absolutely authentic, but in a semi-experimental, situation, in which the participants — though genuinely interested in trying out the program — were asked to go through the computer tutorial in a specially arranged setting. However, the analysis and its results concern two users trying to decipher a computer system, and in that work the experimental nature of the setting is not a problem. In using the tutorial, the users could choose which module to do, and then go through it by reading about it and trying the functions out on a replica of a Word 5.0 screen. At the very beginning of the session (before the start of the actual tutorial), the users could select to do a separate mouse tutorial to learn how to use the mouse. In using the program, the human participants had to

113

114

Discussing Conversation Analysis

decide whether a screen was only for reading, for reading and acting upon reading (which could be bypassed), or having to act upon reading. As can be seen below, usually the last option, or rather obligation, accompanied an interactive message from the computer. Talking, gesturing and other meaning-making was undertaken by two participants side-by-side, ‘ear-to-ear’ in front of a computer. The program exhibited information and requests, the user-learner’s action was acceptable or not to the program. In addition to the overt acceptability of doing the practice correctly, a more covert interaction was also going on. Namely, the users’ choice of Next Screen 3 or End Practice 4 at the lower right corner of the screen implied that they had read/seen the screen they were leaving. In fact, these two options for the users to choose in order for the ‘tutor’ to continue were also indicating what kind of screen was in front of them, whether they should read (Next Screen) or practise (End Practice). However, sometimes a problem arose when they could not see the diŸerence of action space created for the userreader; as a result, the users would try to do the action described in the text. This mistake would then result in a corrective message from the computer, an attempt at mimicking a next turn repair initiator (SchegloŸ et al. 1977), which behaved a bit diŸerently from the corrective messages in the testing or practice phase. When the user’s action showed a wrong understanding of the content of the direction, a repair ensued that would tell the user how to do the action correctly; the repair initiator would disappear immediately after correct action. If the user attempted to do what the text was informing him about when the screen was not yet a practice one, a repair initiator would appear, and disappear from the screen without any action required from the user. The consequences of having various types of repair initiators can be quite intricate; however, through careful data analysis, they can be found to be reasons for some problematic understandings. It is in situations of asynchronous instruction following like this that the inseparable and embodied nature of turn-taking, repair work, action, and meaning becomes highly accentuated. Thus, I would not have been able to analyze how the participants interpreted what was going on, i.e. to investigate ‘meaning,’ had I not been equipped with an understanding of the basic ‘structural’ or ‘formalistic’ ideas of CA. I therefore fully agree with SchegloŸ that generic organizations “serve as tools in explicating the action and interactional import of particular episodes of interactional conduct” (Interview: 32). To extrapolate quite heavily from SchegloŸ’s interest in spoken sequential interpretation, the whole encounter between a text and a reader can be ana-

Users’ interpretations at a computer tutorial

lyzed as interaction unfolding in time and space which, in addition to the local history of the encounter, also forms an integral context of understanding or misunderstanding. The text on the computer screen can be regarded as an important resource in the sequential organization of the situation. Most often, the user-reader is in control of the encounter, choosing the relevant next in the text. However, sometimes the program corrects the user in an interactive fashion, and in these moments the user-reader is meant to decipher the ensuing turn/action by him/her as part of a side sequence in which the previous action should be repaired. Thus, the repair initiator by the program always concerns the user’s interpretation of the program’s instruction.5 A careful analysis6 of the encounter below will show that di¹culty in making the diŸerence between commands to do and reformulated directions (repair) can be a cause of some problems in maintaining an interpretation. 4.1. Visuality and materiality7 The written language that the users encountered on the screen formed a visual context that was surrounded by the icons, arrows, (moving) graphics and other visible surroundings of the program. Thus, the interactive space was diŸerent from a text-only format prevalent in most ªctional texts (also electronic ones; cf. Aarseth 1997). The textual area was not just to be read, but to be acted upon (on the basis of what was read).

5.

Data analysis

The tutorial was given by a software program, which the designers had programmed to take into account that the context of computer use might not be known to every user. Consequently, a potential learner could choose to use the interface to ªnd out, for instance, how to use the mouse. Indeed, this option was available for the users at the very beginning of the learning program, preceding the actual tutorial. Most of the mouse tutorial concerned which buttons to click on the mouse and how to proceed to the next screen with the mouse. Thus, local (i.e. not to be found outside the tutorial program) abbreviations were introduced to make referring to clicking the left or right button (or both) of the mouse more e¹cient by not using much of the screen space. One of the most used ones was Click-L to refer to clicking the left button.

115

116

Discussing Conversation Analysis

5.1. Click-L In the mouse tutorial, Click-L “means to press and release the LEFT mouse button,” or, in the words of the summary: When you see this: Do This: __________________________________________________ _______ Click-L press and release the Left mouse button

So, Click-L was a local hybrid with features of symbol, icon and index at the same time. It clearly was artiªcial and conventional (symbol), its design iconic in that the acronym was a symbolic metaphor of the activity (pressing and releasing: Click) and the object (the left mouse button: L). All in all, Click-L was an abbreviation typical of written language, a trouble source, a complex token whose value (Halliday 1994) was given in the deªnition. When it was used without the deªnition, it was a two-morpheme construction, and, according to Thibault, composite structures “require relatively more constructive and/or interpretative work on the part of the language user so as to construe the new joint meaning which results from the combination of their constituent parts” (Thibault 1997: 283). Click-L as an instruction worked quite well in the tutorial, but at one point, one of the users clearly had an incorrect notion of what Click-L refers to. B’s interpretation is available for the analyst in the video recording, and in the transcript in line 114, which is marked by an arrow, at which point B is going to press key l on the keyboard (the transcription conventions can be found in Appendix): (1) 108 109 110 111 112 113 114 115 116 117 118 119

B: oh [(move the mouse pointer)] ((whistle)) (2) y[eah (1) ] [((moves mouse))] [((gaze: sc, kb))] click l ((hand to l on kb, gaze up, hand to rest; knits eyebrows)) A: did you click l? B: [no] → [((hand towards l on the keyboard))] A: [the left button?] [((gaze to B))] B: ((hand to rest, gaze to mouse, hand to mouse)) ([ )]

Users’ interpretations at a computer tutorial

120 121

[((↓↑))] (( ))

Line 114 in Extract (1) is a very strong indicator of what Click-L means to B at that point: it refers to clicking 1 on the keyboard, in other words, the Click (press and release) part of the acronym was ‘correctly’ interpreted, as click is a transitive verb, meaning, according to The New Penguin English Dictionary, “to strike, move, or produce with a click.” It was the L part which for B at this point did not refer to the left mouse button but the key l. It could be possible to leave this misunderstanding sorted out through a repair device by A at that, just as a description of what happened. But as there had been no problem with Click-L before, i.e. the users had demonstrably learnt the meaning of this acronym at the beginning of the lesson, this brief moment in the encounter oŸered an exception, a deviance that should be accounted for.8 A detailed analysis of the data below will show that the users were trying to ªnd a logic in how the program worked, and that one of the complicated reasons for why in Extract (1) B misunderstood A’s direction/repair initiator was that the routine was not there. Line 114 in Extract (1) and the ensuing repair by A give the analyst enough data to conclude that in this particular point in the encounter for B, “click l” meant clicking the key l on the keyboard. There follows a closer look into the locally wider context in order to track the possible reasons for this confusion. Already Extract (1) showed that the diŸering semiotic systems of writing and speaking might have contributed to the misunderstanding, i.e. when the acronym Click-L was spoken aloud (“click l”), the potential ambiguity of it referring to clicking the key l was enforced. However, to make the point clearer, the interaction preceding Extract (1) is needed, and is shown in Extract (1′). (1′) 102 103 104 105 106 107 108 109 110 111

C: [[=Move the mouse pointer to the ”d” in “due”. Click-L]] | B: [o-oh] A: ‘move the (.) mouse’ B: mouse A: ªrst B: oh [(move the mouse pointer)] ((whistle)) (2) y[eah (1) ] [((moves mouse))] [((gaze: sc, kb))] click l ((hand to l on kb, gaze up, hand to rest; knits eyebrows))

117

118

Discussing Conversation Analysis

112 A: did you click l? 113 B: [no] 114 [((hand towards l on the keyboard))] 115 116 A: [the left button?] 117 [((gaze to B))]

Extract (1′) begins with the repair initiator that appeared at the lower part of the screen after a wrong action by B. Now there is a clear connection between the computer’s ‘turn’ in line 102, and what B says in lines 109 (“move the mouse pointer”) and 111 (“click l”): he is quoting what he sees on the screen (though he is muttering the words to himself, rather than reading aloud to A). In line 111, B’s actions and facial expression show that his uncertain analysis at that point is that Click-L refers to the key l on the keyboard. In line 112, A asks “did you click l?”, which B treats both as a question about his previous unªnished action (“no”) and a request to ªnish it (line 114). However, A’s question comes as a part of a longer check-up procedure than concerning just B’s previous action. A had been going through B’s interpretation of the computer’s repair initiator, checking ªrst that B moves the mouse, and now, with “did you click l”, A other-repairs B’s misunderstanding of Click-L, making thus the repair initiator that is visible on the screen relevant. But as Click-L is spoken aloud, the interpretation of it as meaning that he/they click the key l is enforced for B. The consequences of the diŸerence between “click l” as the spoken form of the written Click-L re¶ect what Halliday has said about the diŸerences between written and spoken modes of language: “Writing brings language to consciousness; and in the same process it changes its semiotic mode from the dynamic to the synoptic: from ¶ow to stasis, from choreographic to crystalline, from syntactic intricacy to lexical density” (1993: 118). The lexical density of written text in this moment of spoken intertextuality creates problems rather than solves them. This is the reverse of the cases in Hutchins’ (1995) discussion about the materiality of signs (e.g., Light2 and LightZ can be mixed up as visual signs but not when spoken aloud: “light two” and “light z”). After B starts moving his ªnger towards the key l on the keyboard, A glances at him, which is a very distinctive action in an ‘ear-to-ear’ situation. That is, when the persons are sitting shoulder to shoulder at a computer screen, it is not very often that they turn to look at each other. A’s gaze occurs with his next other-initiated repair (“the left button?”) of B’s action (which is B’s interpretation of A’s previous turn), where the referent of

Users’ interpretations at a computer tutorial

L is made explicit (cf. SchegloŸ et al. 1977: 369: “if more than one otherinitiated sequence is needed, the other-initiators are used in order of increasing strength”). With the question intonation, A produces the other-initiated repair as if it was a self-initiated self-repair of his own turn (i.e. clarifying “did you click l?”). So here we have two examples of how much the formulations of other-repair can vary according to “the circumstances of their selection” (Interview: 12).9 An analysis of Extract (1″), which extends the example to show what happened next, reveals that although the problem seems to have been resolved and the side sequence that the repair initiator started has ªnished, B seems to still be confused. (1″) 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127

C: [[=Move the mouse pointer to the “d” in “due”. Click-L]] | B: [o-oh] A: ‘move the (.) mouse’ B: mouse A: ªrst B: oh [(move the mouse pointer)] ((whistle)) (2) y[eah (1) ] [((moves mouse to d))] [((gaze: sc, kb))] click l ((hand to l on kb, gaze up, hand to rest; knits eyebrows)) A: did you click l? B: [no] [((hand towards l on the keyboard))] A: [the left button?] [((gaze to B))] B: ((hand to rest, gaze to mouse, hand to mouse)) ( [ )] [((↓↑))] ((
)) B: ((gaze: screen; moves back sharply)) ((gaze: down screen, up screen; moves mouse pointer to the right)) okay so (I want to move my) mouse pointer [gaze: down screen; moves mouse pointer))] |

119

120 Discussing Conversation Analysis

128 129 130 131 132 133 134 135

A: [yeah] B: to the d ((gaze up screen)) [((hands palms up))] | A: [and now you can type] B: [do you] think? [((hands palms up))]

In line 120 above, B does the required clicking of the mouse button, and what was designed as a corrective message disappears from the bottom of the screen (line 121). That B moves sharply (line 123) conveys a surprised reaction to the disappearing of the message. What B then does seems to be very counterintuitive: he is quoting the directions which had just vanished, as if they were the relevant next action to be performed in the unfolding situation. This means that for B, the episode of clicking l (lines 112–120) was not part of following what the message at the bottom of the screen told them to do (we do not normally act upon an instruction twice, unless we think that our ªrst attempt was not successful). It is impossible to know whether B in line 135 (hands palms up) has forgotten what the rest of the instruction said or whether he is just perplexed about Click-L still. (That he actually types due, though the instructions of typing past are still on the screen, demonstrates that he is aligning to the disappeared message and in it the next word after to the d, namely, due). In order to discover plausible reasons for why the repair initiator is not comprehended as such, even more of the preceding context of the discourse is needed: (1′′′) 93 94 95 96 97 98 99 100 101 102

((√ entering text — practice)) B: heh [he] | A: [oh] there we go now ‘move the highlight to the d in due (.) type the word past’ B: that’s what’s (…) gonna do. (move the) [(highlight)] [((moves mouse; pointer to the lower right corner))] let’s press this ←= C: [[=Move the mouse pointer to the “d” in “due”. Click-L]]

Users’ interpretations at a computer tutorial

103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125

B: A: B: A: B:

A: B:

| [o-oh] ‘move the (.) mouse’ mouse ªrst oh [(move the mouse pointer)] ((whistle)) (2) y [eah (1) ] [((moves mouse to d))] [(gaze: sc, kb))] click l ((hand to l on kb, gaze up, hand to rest; knits eyebrows)) did you click l? [no] [((hand towards l on the keyboard))]

A: [the left button?] [((gaze to B))] B: ((hand to rest, gaze to mouse, hand to mouse)) ( [ )] [((↓↑))] ((
)) B: ((gaze: screen; moves back)) ((gaze: down screen, up screen; moves mouse pointer to the right” okay so (I want to move my) mouse pointer

121

122 Discussing Conversation Analysis

126 127 128 129 130 131 132 133 134 135

[“gaze: down screen; moves mouse pointer))] | A: [yeah] B: to the d ((gaze up screen)) [((hands palms up))] | A: [and now you can type] B: [do you] think? [((hands palms up))]

The directive (quoted on lines 95 and 96, and prefaced with “oh there we go now”, in which “oh” conveys noticing something (Heritage 1984)) did not explicitly mention that the highlight should be moved to by the mouse. B’s “that’s what’s ( ) gonna do” (line 97) makes it clear that his attention is drawn to the same piece of text. He does not use the mouse, however, but leaves the mouse pointer at the lower right corner (line 99) and tries to use the backspace key (line 100), a logical step in the sense that the use of the backspace key was taught a bit earlier, and also in that the visual ‘meaning’ of the wordings to move the highlight and to the left does take place when backspace key is hit. B’s “o-oh” in line 103 indicates that he noticed the message appearing at the bottom of the screen and that therefore his pressing of the backspace key was not sequentially correct. A quotes the instruction with a pause which makes “mouse” foregrounded (line 105). By repeating “mouse” (line 106), B makes it interactionally relevant as well, but does not do anything. It is only after A’s added “ªrst” to his quote that B seems to get what A means, and produces an “oh” receipt (Heritage 1984: 319) to acknowledge he understands what A proposed, namely that the ªrst thing to adhere to is the instruction at the bottom of the screen. In other words, B did not treat the message appearing at the bottom of the screen as a repair initiator that needs to be dealt with immediately. In lines 109 and 110, B demonstrates that he is following the ªrst part of the new instruction, but Click-L seems to be a problem to be locally solved, as explained above. The question still remains: why would B have this new referent for Click-L, or rather L? B is clearly hesitant to press key l (line 114), and this could be an indication that he still has the mouse click as the other option. However, a closer look at what had happened earlier in this Entering Text module of the tutorial gives evidence that what B might be perplexed about is not a choice between these two interpretations as such, but,

Users’ interpretations at a computer tutorial 123

more speciªcally, between whether L refers to clicking key l on the keyboard or clicking l in the word bill on the screen. The Practice module was preceded by Overview and Steps screens that might give an explanation for B’s problematic interpretation of Click-L. The following extract starts with the participants quitting the practising of entering text with and without using the ‘enter’ key for a new line. Click-L appeared on that screen within the instruction to go on (When you’ve ªnished, point to End Practice and Click-L), which was carried out with no di¹culty (see lines 1–5 in Extract (2)). So by this time in the encounter, there had been no problems establishing Click-L as referring to press and release the left mouse button. (2) 1 B: okay. [(we’ve ªnished with that)] 2 [((glances at the mouse))] 3 end of practice (here we go) ((glancing at the mouse, ↓↑)) 4 5 [tsh]hh 6 [((≡))] 7 ((pushes the keyboard back to the middle, right hand to the mouse)) 8 9

124 Discussing Conversation Analysis

10 11 12 13 14 15 16 17

B: mm (where do we) paragraphs [((gaze: screen; moves the mouse pointer to Next Screen))]  A: [((gaze: screen))] ‘later in the course’ B: ((looks at mouse)) u:h so we go (we go) to t[he next screen]  A: [°go next screen°] B: ((↓↑))

18 ‘(you can always add text to a document)’ 19 (4) 20 (oh) ((knits eyebrows)) o:h ((g:scr)) so! [((g:scr))] 21 | 22 A: [((g:scr))] (‘to the left of the 23 high[light’)] 24 | 25 B: [so you] put (.) 26 27 [( )] 28 [((hand to the side of neck))] 29 A: ‘what you [type will appear to the left] of the highlight’

Users’ interpretations at a computer tutorial

30 | 31 B: [a:h (.) so that adds a space] 32 [((hand to point to A, cheek, rest))] 33 puts a space in (( )) ((gaze: screen)) ((animation on 34 the screen ‘typing’ (with a rattling sound) some[times])) 35 | 36 A: [( )]

37 38 39 40 41 42 43 44 45 46 47 48 49 50 51

A&B: ((gaze: screen)) B: ‘move the highlight to the place you want to add new text. type the new text’ [‘hard work will sometimes lead to riches’] [((gaze to lower part of screen with head movement down))] ((gaze: upper part of screen)) A: [is that ( )?] [((moves closer to screen))] B: ((moves his left hand to the keyboard; looks up)) A: oh [there there] [((smiles; left h. to screen))] h:a(h). B: we [need to put the] l (.) to this [((index ªnger to press))] A: do we do that now?

125

126 Discussing Conversation Analysis

52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

B: ((hand away from kb)) we wanna add the sometimes? [((tongue smack))] [((moves to left))] [( ) last screen] [((whispering, smiling))] ((right index ªnger pointing screen)) [uhm:] [((scratches chin))] [((ªnger to kb; gaze: screen))] | A: [°’type the new text’°] B: [(yeah) we will put it to the left (2) or shall I shall I ªgure out sometimes (on there) (2)] [((gaze: screen; right index ªnger on backspace key))] ((right hand to point screen)) [do you think?] [((moves to writing position))] A: give it a try B: (to) see what happens [s=] [((gaze to keyboard))]

Users’ interpretations at a computer tutorial 127

75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92

C: [[=Just read now. You’ll have a chance to try it later. To go on, point to “Next Screen” and click. To use the Course Controls, point to “Course Controls” and click.]] A: [(.) EH ha ha] | B: [((gaze to screen))] wh[at happened?] | A: [‘just] read now’ ((laughing voice)) B: [huh?] [((gaze: screen; hands on kb))] A: impolite thing ((
)) B: [oh:] [((moves away from keyboard))] ((sigh)) A: so I guess we just go to the [nex- next screen=] | B: [=go[to the next scre]en.] [((↓↑))]

A and B encounter a new module (Entering Text) on the screen reproduced between lines 9 and 10. Lines 10 to 17 display one of the objects of continuous meaning negotiation in the encounter: is the screen for reading only or for reading and doing? B’s words in line 10 seem to orient to the latter interpretation (before quoting the screen in “paragraphs?”, B says, “where do we?”, which orients to doing something, as does the text before what he quotes: You’ll be working with). This is how A interprets his turn (line 13), producing an other-initiated other-repair by transforming the screen text from neutral information to an account of why B should not proceed with the topic or action (“later in the course”). Here we have another concrete example of how text becomes an important part of the sequential action, this time being used to repair the other participant’s interpretation, and therefore adding to the “variety of forms which other-initiated repair can take” (Interview: 12). The next screen (between lines 17 and 18), which appears as a result of the mouse click on Next Screen (line 17), is a complicated one: there are deictical references within the text and as it is describing the steps to take to add text to a document, the imperative mode is used (in contrast with the Overview screen between lines 9 and 10). This screen also introduces Click-L for the ªrst time in the Entering Text module instructions, and here Click-L is not used to refer to clicking the right-hand lower corner of the screen.

128 Discussing Conversation Analysis

The following two subsections point out two types of complexities that language-in-use creates in this case, ªrstly because of its materiality, and secondly because of the read/act dichotomy instructional texts produce. 5.2. Potential hazards created by the materiality of the text On the screen (between lines 17 and 18), the ªrst sentence (You can always add text to a document) is a declarative one, and always makes it a description of a general state of aŸairs; the following one introduces what has to be done: These are the steps you will follow. Numbering the step as 1. (‘one’ or ‘ªrst’; it can be understood as a number, and a number in a sequence of numbers yet forthcoming) and attaching a (red) arrow next to the instruction, can be understood to indicate that what A and B see in front of them is a ªrst step. However, the sentence in the next line, To do this, point to the spot. Then Click-L, in this particular screen allows for more than one referent for L. Firstly, the reference of the anaphoric expression To do this is not quite clear: does it refer to move or to add? If the referent is Move the highlight to the place, and it is compared with the example at the lower part of the screen, one plausible interpretation is that the spot has already been pointed to and the letter l of the word lead clicked; after all, there is an arrow pointing at the letter l. Also on the original screen, both the highlight, the arrow and the box frame, together with text, were red in colour, and the arrow and the box appeared on the screen a couple of seconds later than the rest of the screen, thus being foregrounded as important. If to do this refers to add, then the interpretation of the rest of the instructions could mean: point to the spot and click key l on the keyboard (so ead becomes lead); a similar interpretation would result if To do this referred to the whole of the preceding sentence (i.e. Move and add). To complicate the meaning potential, the word left appears in the box text, and L has been used to refer to the left mouse button earlier. Complex intertextuality and meaning potential might be the cause of A and B’s perplexity at this point in their use of the program. A seems to be concentrating on the (red and thus foregrounded) text in the box (lines 22–23 and 29). B’s interpretation of what the text in the box says also reveals that he considers the example text as depicting what has happened when the instruction in 1. is carried out (and To do this refers to the whole of the preceding sentence). To give an explanation to the problematic sentence in the box, B ªnds a logical solution: “that adds a space, puts a space in” (lines 31–33); i.e. after the letter l is highlighted/pressed, a space has been added to the left of the highlight (by the program).

Users’ interpretations at a computer tutorial 129

5.3. To do or not to do The next screen (between lines 36 and 37) shows an animation of the (red) word sometimes being typed to the text in the example, something which A misses at ªrst, even if the animation had a sound eŸect of typing (the clattering sound of which does not exactly coincide with the appearing of the letters). In line 49, what was previously hypothesized from B’s action now becomes visible: B interprets L to refer to the key l, and number 2. instruction is an instruction to carry out the action (as number 1. was interpreted to be an example-to-read about selecting the spot and typing there). A stops him by questioning, at least on the surface, the sequential placement of the activity of pressing l (but not the activity itself, i.e. A does not directly accuse B of a wrong interpretation of the screen as one to practice on): “do we do that now?” (line 51). In fact, A’s turn is ambiguous. B takes the other-initiated repair of his interpretation of the screen to be about ‘doing what,’ rather than about ‘reading, not doing.’ B suggests another text to add, which is the same text as was typed last by the program (“the sometimes,” line 52). There is no acknowledgement by A of this understanding check/suggestion; A’s next contribution is in line 64, which shows that A is still pondering over the instruction type the new text. For B, who is already convinced that they have to do something, and who is waiting for A’s evaluation of what to do, A is producing a request rather than quoting a repairable. B then self-repairs his pressing of key l to moving the highlight “to the left” (his ªnger ready to press the backspace key), but A, again, does not comment on B’s suggestion (2 seconds pause in line 66), which results in B reformulating his earlier non-commented suggestion “we wanna add the sometimes?” as “shall I shall I ªgure out sometimes on there.” B is taking the responsibility and is showing that the last proposal is his intent about which he is asking A’s desire/opinion (see Boyd 1992). A is still silent, so B elicits A’s assessment again while moving to the writing position (“do you think?”). A’s reply makes it clear that he is not sure whether this is what should be done (“give it a try”), nor is B (“to see what happens”). However, because B is in a typing position and looking at the keyboard, he misses the repair initiator that is promptly produced by the computer, and to which A immediately strongly reacts (line 78). B initiates repair (“what happened?”), which shows that he missed the program’s ‘relevant next.’ As a repair, A quotes the computer (just read now), but he transforms the repeated text, because, in the process of quoting it, he ªts it in his own action of laughter (line 82). The computer’s turn is diŸerent from the other repair initiators in that it disappears from the screen

130 Discussing Conversation Analysis

without A or B using the keyboard or the mouse. B did not necessarily understand what was going on by the time the fairly long message vanished (even if, in line 84, B displays some sort of change of state). However, though the message was temporary, it was still visible for some time and available to be used as a resource.

Repair initiators and the screen types encountered If the local history of the encounter is to contribute to the meaning-making, then it is important to know what repair messages the users had come upon and in which circumstances before those in Extracts (2) (Just read now…) and (1) (Move the mouse pointer…). The very ªrst screen that the users encountered (after the copyright title of the program) was one that had an error message visible at the bottom of the screen. This is where the users learnt that a message at the bottom of the screen has to be dealt with ªrst to be able to continue using the program. The next encounter with a repair initiator occurred in the mouse tutorial. The corrective message appeared because they did not click the words Next Screen on the right lower corner of the screen, but the words in the instruction that referred to the phrase. Their correct action led to a praise and a direction to do the same action again (EXCELLENT! Try it once more.), which replaced the instructions on the screen. Thus, the history of the encounter revealed an inconsistency in the (outlook of the) screens from the mouse tutorial and that for reading: clicking Next Screen in the Overview mode would make the next screen appear, instead of a remark on the same screen. 5.4. Click-L revisited Extract (2) ends with the users entering the practice stage, which we have seen in Extracts 1–1′′′, to go through the local details to ªnd possible reasons for B’s misunderstanding of the reference of Click-L. The limited history of the users’ encounter with the Overview and Steps modules showed that B never actually got to try his wrong interpretation of L as key l, and therefore he was never repaired by the program. Originally, the wrong ‘pointer’ seems to have been caused by the ambiguous visual layout of the Steps screen (see Example (2)), the picture between lines 17 and 18), which allowed for L to be connected with the letter l in the word lead. When the reference of To do this in the same screen also allowed for several candidate directions, one possibility was to link L and letter l of lead, and, at the same time, to key l on the keyboard. Thus, the

Users’ interpretations at a computer tutorial

semiotic materiality of the screen was strong enough to override the carefully taught and successfully practised connection between L and the left mouse button: the lexical metaphor (e.g., Martin 1992) L in the technical term ClickL was unbound, and a less complicated connection (L = l) established. In other words, the new referent of Click-L was an example of how, also from the receiver’s point of view, “speciªc associative patterns are evoked from the virtual associative series that constitute ‘an entire latent system’ in response to speciªc contextual contingencies” (Thibault 1997: 274). The Click-L episode gives a concrete example of how the actual context of interpretation is crucial for meaning-making (and its professional analysis). The wrong interpretation was later maintained through the possibilities created by stepwise meaning-making. The repair message, Move the mouse pointer to the “d” in “due”. Click-L, was an example of a typical turn-analysis repair; it was meant to correct the user’s previous action. However, when the users encountered this repair message, they had just dealt with the turn-taking repair initiator of Just read now. You’ll have a chance to try it later. To go on, point to “Next Screen” and click. To use the Course Controls, point to “Course Controls” and click. This message disappeared without the user’s doing anything with the gadgets, which was logical in the sense that it was only meant to be read, not acted upon. However, the latter part of the message gave two options of doing, and thus did deal with ‘a right action.’ B treated the Move the mouse pointer… repair initiator as a message that also disappeared by itself rather than after the correct action. This is why B did not seem to connect the side sequence of the reference of click l being corrected by A’s talk and the repair message itself: B orients to the disappeared message as a relevant next. Part of this foregrounding is due to the higher rank of interactive computer messages in the order of ‘mustness’ — it was always necessary for the user to do/stop doing something after an interactive message, whereas the practice instructions could always be bypassed. 5.5. Summary The design of the program was very modular, and the interactive features which were ‘true’ for one module, were not in another one. So when B had just learnt that the ‘requests’ of repair messages can be oriented to, even after they disappear (just read now — go to next screen), he applied this latest knowledge to the present case as well. The problem of wrong reference or disruption in the meaning of the acronym that the program had managed to create was therefore

131

132 Discussing Conversation Analysis

caused and maintained because the users treated the program as functioning with a particular logic, usually adhering to how the program had behaved ‘the last time.’ Also, the wrong interpretation of Click-L was carried on partly because the program was unable to make clear for B which messages were repair initiators and which mere instructions. The indistinctiveness of the two types of messages was increased by the fact that the repair initiators were not clearly connected to the action of an instruction they were correcting. So, in Example (1′′′), the screen between lines 9 and 10, the wordings of the instruction (Move the highlight to the d in due and type the word past.) and that of the repair initiator (Move the mouse pointer to the “d” in “due”. Click-L) start identically. The repair initiator did not mark the repairable, the trouble source, distinctively (for example, by To move the highlight, move the mouse…, or even, No, to move the highlight you have to move the mouse pointer…) and the only hint that it is a repair initiator is that it surfaced after an (erroneous) action. The design of the program relied totally on the users’ understanding that interactive messages from the computer were other-initiated repairs. The interactive force created by the repair messages was understood to be the strongest: they could not be ignored. However, this did not mean that the ‘repairness,’ the message as beginning a corrective side sequence, was understood as such. Strate regards the metatime of cyberspace as a reason for “the breakdown of the signifying chain” (Strate 1997: 373). The example of Click-L shows that also the sequential time of semiosis, when it is represented in a computer program, can be part of the breakdown. Another feature of instructions seems to be their double nature: they can be taken as general information, or as requests to act promptly. Roy Harris (1996) calls the two types of action sequences that can result from communication “assimilative” and “enactive.” This distinction is visible also in telephonemediated human-human interaction, of which an example can be found in Whalen and Vinkhuyzen’s data from a large corporation’s customer support telephone center (2000: 127–128): CSSR: Cust: CSSR: Cust: → CSSR: → Cust:

it:’ s so simple here’s what you need to do okay the access panel ((pause)) where the dry ink cartridge is? uh huh open it ((longer pause)) you want me to through it right now?

Users’ interpretations at a computer tutorial

The customer service and support representative (CSSR) instructs the customer to open the dry ink cartridge (“open it”). The customer’s question, “you want me to through it right now?”, orients to the inherent ambiguity of the CSSR’s instruction as an assimilative or an enactive one: is the CSSR describing an action for future reference or telling the customer to act on his instructions during the telephone conversation?

6.

Discussion

The purpose of this paper was to demonstrate, with SchegloŸ’s interview as a special resource, and with my case study as a practical example, how the impressive conversation analytic research agenda about the organization of social action in conversational settings can be fruitfully adopted and developed to investigate meaning-making in computer (or electronic text) use settings. For instance, the notions of turn-taking and repair can be extended to cover physical action when that action is part of the interpretative sequence. And, as exempliªed by the observation made about the telephone conversation extract above, sometimes ªndings from investigating situations in which (stable or interactive) texts are interpreted can actually help detect similar phenomena in conversational practice. The study was undertaken by a non-native speaker of English, but an ‘enculturated’ user of various language technologies (cf. SchegloŸ’s recommendation of native speakers researching their own language and culture). In the case study, I demonstrated that for one of the participants, the material/ visual semiotics, embedded in the sequential sense-making, created another meaning for the recently learnt Click-L. Thus, a conversation analytic investigation could give a concrete example of “the cross-coupling of the material and the semiotic domains, which is a necessary condition of all meaning-making” (Thibault 1997: 287). SchegloŸ emphasizes how “[o]ften the way of understanding ‘great big’ is to understand ‘teeny little’” (Interview: 29). Though SchegloŸ refers to longer stretches of interaction being built on, for instance, a single adjacency pair, I ªnd this quote also a ªtting description of my present study: detecting a ¶eeting misunderstanding (i.e. observable only from the participant’s physical action simultaneous to his talk) of a textual formation, a ‘tiny’ abbreviation, helped me — after detective (CA) work about its causes — illustrate how accessibility of computerized (and other) texts is not a once-and-for-all phenomenon, but

133

134 Discussing Conversation Analysis

in ¶ux. When a speciªcally agreed convention turns out to be indexical, we have evidence for the creativity of the interpreter (as discussed by, e.g., Streeck 1980), aided by the visual (cf. reading) and the sequential (cf. interaction) resources oŸered by the program, the latter being diŸerent from traditional (printed) text-user encounters. Thus, the question of what makes texts easy or di¹cult to understand is an empirical one. In his interview, SchegloŸ stresses his commitment to a “lively and hopefully convincing dialogue with a number of disciplines and subdisciplines which ostensibly work in the same area.” In the same vein, I would like this paper to assure also other than CA researchers that the act of signifying can actually be studied empirically: when two persons cooperate, their interpretations of each other and of texts and other signs can be studied as a publicly available event of sequential interpretation. What was shown was that the meaning stratiªcation — the layers of meaning, the interactive accomplishment of meaning-making, the dialogic alternative to meaning negotiation, to name a few perspectives on meaning-making — can be subtle and intricate. In instructional contexts, in which the intent surely is to be as clear as possible, indeterminacy of meaning exists. Meanings are transient: the interactive contingencies disrupt the good intentions of the author(s) in ways that they might not have been anticipated. This is why close inspections of empirical materials of the kind reported here are needed to connect the social semiotic to social interaction and to appreciate the complexity of the ‘context of use,’ not as an additional separate ‘factor’ of meaning-making, but as intertwining text, context, and intertext into aspects of sequential interpretation. When the aim is to channel the user of an electronic (instructional or otherwise interactive) text, the interpretative framework has to be researched. Towards that goal, conversation analysis as a method is needed to reveal the sensitivity of the situated and sequential meaning-making process to the material and semiotic context. SchegloŸ criticizes with good reason cognitive science and speech act theory’s focus on individuals and their individual actions or sentences, usually in “imagined actions” instead of “actual ones” (Interview: 39). In reading research, the ‘coherence’ of text as a bigger unit has been appreciated, and, especially in the latest theories, the reader’s interpretations and the context of reading have been accentuated. In more clearly socially-oriented ªelds like social semiotics, a general recognition of the circumstances of interpretation prevails. The lesson to learn for reading research from CA is that there are situations in which the focus has to move from text-internal coherence and imagined reading situations to the reader’s actual sequential interpretation of

Users’ interpretations at a computer tutorial

the text as part of practical action. And for social semiotics, CA oŸers a sound way of investigating actual interpretations, an undertaking which can help ªnd new, sometimes even surprising, aspects of the interaction and interpretation in question. The present paper also showed that simply imitating a teaching process in computer-aided instruction does not take into account that the basic interactive features of the process (‘positive feedback,’ i.e. evaluation, and ‘explanation,’ i.e. repair work) in human-human interaction are seen to conªrm the participants’ intersubjective understanding of what is being talked about (e.g. SchegloŸ 1992). This has two kinds of consequences for human-computer interaction in which the system does not have the same capacities as we do: either the computer’s evaluative ‘turns’ are oriented to as funny (i.e. out of place; cf. Extract (2), lines 75–78), or interactionally more consequential (partly due to the complexities of the semiotic surroundings); when the system’s next turn repair initiators are misinterpreted, the system has no capacity to continue the “routine grounding for intersubjectivity” (SchegloŸ 1992: 1295), i.e. to repair the user’s misunderstanding of the ªrst repair initiation. In his interview, SchegloŸ discusses his ongoing projects on the next turn (or second position) and third position repair initiators. In another connection, he spells out the limits of the research: “We did not mean to include within the scope of ‘repair’ all practices addressed to problems of understanding (like understanding exactly how the Internet works), only the narrower domain of ‘understanding what someone has just said’ — though there can on occasion be a fuzzy boundary between these.” (SchegloŸ 2000: 207). My paper dealt with understanding exactly how a tutorial program works — not in principle, but in a concrete encounter with it. I ªnd that in this speciªc case, not even a fuzzy boundary was there as the participation framework of the users interacting with the text on the screen dissolved in or at least became detectable through the participation framework of them interpreting each other in the cooperative situation.

Notes 1. Though the technology in question is outdated as such, the data is still very much usable for investigating how asynchronously produced language is interpreted. 2. Unfortunately, in his interview, SchegloŸ does not enter into a more detailed discussion about video recordings and the “description of behavior” (as the question was put forward),

135

136 Discussing Conversation Analysis

because SchegloŸ himself has done research on the embodied nature of interaction, for instance on gestures as accompanying talk (eg., SchegloŸ 1984). To me, the interest lies in physical action as a more independent interpretative unit. 3. E.g., the screen in Extract (1′′′) between lines 100 and 101. 4. E.g., the screen in Extract (2), lines 9-10. 5. The active role of the program in marking the problematic actions of the user makes it diŸerent from Suchman’s (1987) photocopier, which relied much more on the user to ªnd the trouble: the photocopier showed the user instructions of the relevant step on the basis of what the user was modeled to be doing. The display had a Help button which was meant to be selected in case of problems. Thus, even if the photocopier’s instruction would have in practice been a repairing one (e.g., repeating an earlier instruction because the user’s action caused the program to backtrack to an earlier stage), the instruction was exactly the same as always in that phase: it was not designed as a repair initiator, but as an appropriate response to the user’s detectable action(s). 6. “Careful” meaning that the analysis will not be a vernacular description of what I see in the data, but a rigorous CA investigation (c.f. SchegloŸ’s discussion about descriptions as mundane action or as CA practice). 7. In his interview, Emanuel SchegloŸ mentions Chuck Goodwin’s neurologically-oriented CA work with appreciation. For myself, both Chuck and Candy Goodwin’s work has been inspiring because of their interest in the embodied nature of the participants’ actions. Especially their analyses of interaction in various technology-induced work settings provide excellent examples of how talk and the visual and material environment encompass each other. 8. In the spirit of conversation analysis that nothing in the data is irrelevant a priori. 9. However, the self-repairing aspect of “the left button?” reminds us of an alternative way of analyzing the sequence: “the left button?” is a self-repair that was other-initiated by B’s erroneous interpretation (line 114) of A’s “did you click l?” (i.e. “the left button?” is a case of a “third position repair” SchegloŸ mentions in his interview). For me, this would be too ‘formalistic’ an analysis (i.e. it is only technically that A does repair himself), and I therefore prefer analyzing “the left button?” as orienting to B’s erroneous interpretation of A’s previous talk (i.e. as a second position other-repair).

References Aarseth, Espen J. 1997. Cybertext: Perspectives on Ergodic Literature. Baltimore: John Hopkins University Press. Boyd, J. 1992. “The act in question”. In (On) Searle on Conversation, J. Searle et al. (H. Parret and J. Verschueren (eds.)), 31–34. Amsterdam and Philadelphia: John Benjamins. Button, Graham, Coulter, JeŸ, Lee, John and Sharrock, Wes (eds.). 1995. Computers, Minds and Conduct. Oxford: Polity Press.

Users’ interpretations at a computer tutorial 137

Engeström, Yrjo and Middleton, David. 1996. “Introduction: Studying work as mindful practice”. Cognition and Communication at Work, Y. Engeström and D. Middleton (eds.), 1–14. New York: Cambridge University Press. Frohlich David, Drew, Paul and Monk, Andrew. 1994. “Management of repair in humancomputer interaction”. Human-Computer Interaction 9 (3/4): 385–425. Goodwin, Charles. 1996. “Transparent vision”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 370–404. New York: Cambridge University Press. Goodwin, Charles and Goodwin, Marjorie Harness. 1987. “Concurrent operations in talk: Notes on the interactive organization of assessments”. IprA Papers in Pragmatics 1 (1): 1–55. Halliday, Michael A. K. 1993. “Language and the order of nature”. In Writing Science: Literacy and Discursive Power, M. A. K. Halliday and J. R. Martin (eds.), 106–123. London: Falmer Press. Halliday, Michael A. K. 1994. An Introduction to Functional Grammar. 2nd edition. London: Arnold. Harris, Roy. 1996. Signs, Language and Communication: Integrational and Segregational Approaches. London: Routledge. Heritage, John. 1984. “A change-of-state token and aspects of its sequential placement”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 299–345. Cambridge: Cambridge University Press. Hutchins, Edwin. 1995. Cognition in the Wild. Cambridge: The MIT Press. Martin, J. R. 1992. English Text, System and Structure. Amsterdam and Philadelphia: John Benjamins. Raudaskoski, Pirkko. 1999. “The use of communicative resources in language technology environments. A conversation analytic approach to semiosis at computer media”. Doctoral dissertation. Department of English, University of Oulu. Åbo: Åbo Akademi tryckeri. SchegloŸ, Emanuel A. 1984. “On some gestures’ relation to talk”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 266–296. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1992. “Repair after next turn: The last structurally provided defense of intersubjectivity in conversation”. American Journal of Sociology 97 (5): 1295–1345. SchegloŸ, Emanuel A. 2000. “When ‘others’ initiate repair”. Applied Linguistics 21 (2): 205– 243. SchegloŸ, Emanuel A., JeŸerson, Gail and Sacks, Harvey. 1977. “The preference for selfcorrection in the organization of repair in conversation”. Language 53 (2): 361–382. Smith, D. E. and Whalen, Jack. 1995. Texts in Action. Paper presented at the Symposium on Conversation, Linguistic Society of America Summer Institute, Albuquerque, July 1995. Strate, Lance. 1997. “Cybertime”. In Communication in Cyberspace: Social Interaction in an Electronic Environment, L. Strate, R. Jacobson and S. Gibson (eds.), 351–377. Cresskill:

138 Discussing Conversation Analysis

Hampton Press. Streeck, Jürgen. 1980. “Speech acts in interaction: A critique of Searle”. Discourse Processes 3: 133–154. Suchman, Lucy A. 1987. Plans and Situated Actions. The Problem of Human Machine Communication”. New York: Cambridge University Press. Thibault, Paul J. 1997. Re-Reading Saussure: The Dynamics of Signs in Social Life. London: Routledge. Whalen, Jack and Vinkhuyzen, E. 2000. “Expert systems in (inter)action: Diagnosing document machine problems over the telephone”. In Workplace Studies: Recovering Work Practice and Informing System Design, P. LuŸ, J. Hindmarsh and J. Heath (eds.). Cambridge: Cambridge University Press.

Users’ interpretations at a computer tutorial 139

Appendix Transcription Conventions lower case (bold) UPPER CASE x °word° .word >word< ‘‘ () = ! . , : th: (N) (.) (( )) (( ; )) [] [] | italics ↓↑ ← [[italics in double square brackets]]
√ ≡ Abbreviations: kb sb twds btw

what was actually said loud voice stressed (part of) word word delivered quieter than the surrounding talk word produced with an inbreath speech item delivered quicker than other talk quoting the screen analyst not sure what was said talk/action latches on another exclaiming tone of voice falling intonation ¶at intonation prolonged articulation hissing sound length of pause in seconds pause shorter than one second an activity or comment on the delivery of speech simultaneous activities simultaneous speech/activity simultaneous speech/activity by two persons what is typed on the screen mouse click pressing the backspace key C’s turn appearing at the bottom of the screen C’s turn disappears from the screen new screen (vs. minor changes) screen black for 1 second

keyboard spacebar towards between

Chapter 7

When conversation is not normal The role of conversation analysis in language pathology Ruth Lesser

It would have been di¹cult during the last decade (the “decade of the brain”) not to be aŸected by some enthusiasm for trying to link language with its neurological basis. In expressing his interest in this aspect of the “neurobiology of behavior,” SchegloŸ (Interview: 15) asks the important question, “who is going to describe the behavior?”. This echoes the concern three decades ago of another discipline, psycholinguistics, which found that the ªrst attempts to examine the neural basis of language used a simplistic description of language which was diŸerentiated only by its media of operation. Since then, psycholinguistically-based models have provided a more elaborate framework for mapping mental onto neural activity. Advances in the technology of brain imaging have made it possible to record activity in diŸerent parts of the brain while people are performing various language-related activities. Localization of responses to the human voice (speaking, laughing, sighing etc.), for example, has been reported in functional magnetic resonance imaging to be in the superior rim of the temporal lobes (Belin, Zatorre, Lafaille, Ahad and Pike 2000). The left posterior inferior temporal gyrus is said to be critically involved in lexical-phonological retrieval, and a network of areas including the left anterior temporal region is said to be involved in semantic operations (Mummery, Patterson, Wise et al. 1999). To date, the activities which underlie such ªndings have been formal rather than naturalistic, for example, tasks which required the subject to ªnd verbs associated with given nouns, or to select from a pair of words a close associate of a given word. The emergence of new methods of describing language behavior such as

142 Discussing Conversation Analysis

conversation analysis may seem to promise another paradigm shift in moving towards relating language and the brain. It is not clear, however, how the naturalistic analysis of everyday conversation could make any contribution to this burgeoning aspect of neurolinguistics. Indeed, there may be inherent di¹culties in attempting to map in any formal way the brain activities which underlie everyday spontaneous speech, since practiced and automatic behaviors seem to draw on diŸerent cognitive, and possibly, neural structures from novel tasks (Lum and Ellis 1999). The link between neurolinguistics and conversation analysis which SchegloŸ anticipates therefore rests at the moment on the study of how acquired brain damage (from stroke or dementia, for example) aŸects the use of everyday language. Over the last decade, conversation analysis has become increasingly applied in the study of aphasia and dementia, particularly in English and Finnish, and especially by speech and language pathologists, not least for its potential use in therapy. For them it has provided a similar revelation to that oŸered by psycholinguistics or “cognitive neuropsychology” in the 1980s — then a new way of systematically interpreting the nature of disordered language processing. Conversation analysis provides a missing dimension, the description of the consequences of language impairment on everyday interaction. It substituted for early rough-and-ready attempts to examine “functional communication” with a rigorous system for capturing the subtle nuances of the pragmatics of neurologically-disordered language. For language pathologists, aphasiologists and other clinicians, applying conversation analysis has involved both plusses and problems. Its main positive contribution has been to provide a structure for describing the essential characteristic of aphasia — the disturbance of spoken language or “talk-ininteraction” through which it is initially identiªed. It has shown how judgements of the smooth or disrupted uses of language do not need to be based on presumptions of abnormality or inappropriacy, but can be derived internally from the ¶ow of conversation. Its data-driven approach brought a refreshing alternative to the theory-based constructs such as distinctive neurological syndromes which have dominated the study of aphasia for several decades. It incorporated into analysis the minutiae of language, such as ers and ums, which permeate much aphasic speech, but which had traditionally been discarded in grammatical examinations. It complements the psycholinguistic description of language errors (such as paraphasia, agrammatism) with uncommitted observations of the pragmatic consequences of such presumed impairments. On the problem side (perhaps), it has revived the question of what is normal and what is abnormal in language. Psycholinguistic models were origi-

When conversation is not normal 143

nally derived from experimental studies of normal subjects, since elaborated by evidence from brain-injured people. They were based on the presumption that brain damage reveals normal mental processes, slowed down and therefore more clearly exposed by the injury. The focus of analysis of some conversations between aphasic people and their unimpaired interlocutors, however, has been on the extent to which such conversations deviate from “normal” exchanges. It has highlighted the necessity of studying the contexts in which the conversations take place, and of studying sequences rather than isolated examples. Of its essence is the examination of at least two individuals in a conversational partnership, thus introducing variability depending on the interlocutor — and also requiring the collection of extensive data rather than neat summaries. Its emphasis on qualitative analysis rather than the quantitative, which would lend itself to statistical interpretation, makes it di¹cult to get a handle on this variability across partners in any reproducible way, and to understand longitudinal changes. Below I will attempt to show how researchers have tackled these features. The conversational parameters which have been used include turntaking, topic management, cohesion, repairs and collaborative sequences.

What is diŸerent about aphasic conversations? The non-ordinary consequences of linguistic impairments can feature, on the part of the aphasic speaker, prolongation of pauses, circuitous attempts to self-repair phonemic paraphasias, frequent circumlocution, reduction of turns to minimal contributions, increased use of other-repair and greater use of non-oral media such as writing, “air spelling,” drawing, gesture and laughter. Conway (cited in Lesser and Milroy 1993) observed that aphasic speakers with word-ªnding di¹culties diŸer from the non-aphasic in preferring selfinitiated other-repair to other-initiated self-repair. On the part of the interlocutor, there can be misreadings of unrepaired semantic paraphasias or pauses, and detours to the conversation when the aphasic person’s di¹culties are drawn attention to by prolonged other-repair when the meaning is already clear. A frequent feature is long sequences of collaborative repair which can extend over minutes rather than the three-turn span proposed by SchegloŸ, JeŸerson and Sacks (1977). Some of the earliest papers to apply CA to aphasia took up this theme of collaborative repair of the trouble spots to which aphasic impairments give rise (Lubinski, Duchan and Weitzer-Lin 1980; Milroy and Perkins 1992; Lesser and Milroy 1993 (Chap 9); Goodwin 1995;

144 Discussing Conversation Analysis

Klippi 1996 (Chap.8)). Milroy and Perkins, for example, used Clark and Schaefer’s (1989) principle of least collaborative eŸort to show that there are complex types of repair sequences where acceptance phases are recursively embedded within higher-level acceptance phases, and which can incorporate the non-aphasic partner’s collaborative use of paraphrase to check understanding of a referent. They note that long collaborative sequences often involve the exchange of two sets of “yeah” acknowledgement tokens, such as only feature in normal conversational preclosing sequences where mutual understanding is conªrmed. Laakso and Klippi (1999) propose four phases in collaborative sequences when word-ªnding troubles occur in aphasia: establishing a problem; establishing a collaborative co-participation phase; hint and guess; and a long conªrmation phase. Goodwin (1995) describes a situation where the person with aphasia has only three words at his disposal (yes, no, and). The interaction is managed in this case by the participants making a series of guesses, facilitated by the degree of information provided in the setting. Goodwin emphasizes the aphasic man’s active role through intonation and body behavior. People with such a restriction on their output vocabulary are often reported to confuse yes and no. Goodwin points out a reason for these possible misunderstandings in that the use of no may refer to rejection of a sequence or a situation rather than being denial of an immediately preceding item. It may in fact show frustrated disengagement. Particularly where the aphasia is severe, the majority of the studies of aphasia which have employed conversation analysis have used videorecordings, where this appeared not to be too intrusive or unfamiliar to the participants, as in Goodwin’s study. Body behavior has therefore frequently been incorporated in the analysis, although the role of prosody (Couper-Kuhlen and Selting 1996) has so far been virtually unexplored, despite its critical role for people with a global aphasia or severe output restrictions. Unlike people whose output is limited to stereotyped nonsense or repeated phrases, the limited vocabulary retained by Goodwin’s case consisted of words which could be used productively in social interaction. Others may have to rely more heavily on intonation and body language.

Sequentiality and topic management In these collaborative exchanges, the participants move towards eŸecting repair where the linguistic impairment has left a gap in information which needs

When conversation is not normal 145

to be ªlled. By focussing on the sequential context of a series of turns, Wilkinson (1999) has shown that problems may arise where the non-aphasic partner has di¹culty in grasping the relationship of an utterance to its context. This can occur not only when an uncorrected semantic paraphasia misleads the partner or severe articulatory di¹culties make fragments of speech unintelligible, but where the aphasic person shows lack of awareness of the need (or lack of ability) to make the context clear. Wilkinson gives an example where an aphasic husband follows his wife’s report of arriving at a supermarket just as it was closing with “well you’ve got enough now haven’t you.” The ensuing conversation reveals that this remark was intended as a pre-request for her to turn oŸ the videotape recorder. Wilkinson insightfully points out that such contextual misunderstandings can be exacerbated by an aphasic tendency to use words which have already been used by another speaker — in this case “got” referring to the shopping failure. This kind of subtle pragmatic di¹culty in an aphasic speaker becomes magniªed in the condition of dementia, where unexpected topic shifts can be a major characteristic of conversation. Watson, Chenery and Carter (1999) found that the speech of people with dementia of the Alzheimer type was characterized by frequent failure to maintain and elaborate a topic, sometimes accompanied by metalinguistic comments such as “Oh I can’t think of it at the moment”. Garcia and Joanette (1997) also noted that Alzheimer conversationalists demonstrated more unexpected topic shifts than normal elderly controls.

Qualitative or quantitative analysis? The importance of considering situational context in sequences of turns in analyzing conversation raises the question of the legitimacy of using quantitative statistics, where elements may be pooled out of context and critical information for their interpretation be lost. This is a controversy of relevance to the study of aphasia, which traditionally has relied on quantitative measures for assessing the type and severity of the disorder, the comparison of individuals across and within groups and across time and, not least, the eŸectiveness of any intervention. Although SchegloŸ has claimed (Interview: 12–13) that he uses quantiªcation in his analyses, he uses this term in the sense of employing large quantities of examples and other data in seeking a systematic structure to conversation. No-one can dispute that conversation produces vast amounts of data, requir-

146 Discussing Conversation Analysis

ing extensive transcription — a disadvantage in any applied ªeld where a quick interpretation is required. In aphasiology, however, quantiªcation is a useful statistical technique for making comparisons across situations, individuals and times, and generally involves the grouping of data in its management. When the local context of an utterance is of critical signiªcance to the structures which it comprises, the pooling of tokens from their surface representations presents some problems. SchegloŸ (1993) has drawn attention to the dangers of classing together as a homogenous group phenomena of the same surface form which actually serve diŸerent conversational functions in diŸerent contexts. For clinicians who wish to study changes in their patients over time, including testing the eŸects of any intervention they may have received, qualitative comparisons unsupported by statistical measures of signiªcance may seem too subjective. Consequently, some researchers in language pathology have treated conversational data as if it consisted entirely of quantiªable experimental data. Towards the extreme of this continuum, Ripich, Carpenter and Ziol (2000), for example, tested the longitudinal changes in the use of cohesion devices in the speech of Alzheimer patients using multivariate analyses of variance (MANOVA). Although they found that the patients showed a signiªcant decline in the use of ellipses and conjunctions over a period of 18 months, the researchers concluded that their method was of limited use in distinguishing normal and demented use of cohesion devices, possibly because they used a structured task to elicit conversation, focussed on the oŸer of tea or coŸee. A more common practice in the literature which has applied conversation analysis to aphasia is to combine some quantiªcation with a qualitative ampliªcation. Oelschlaeger (1999), for example, in her study of a wife’s collaboration in her aphasic husband’s word searches, compared the number of turns each contributed to eight sampled conversations, and found a discrepancy of only 5% between the two partners. For before-and-after therapy comparisons, Boles (1998) used four measures: words per minute, words per utterance, relative contribution of words to the conversation (i.e. ratio of words between the speakers) and number of self-repairs, predicting an increase in all for the aphasic participant and a decrease in her partner. Watson et al. (1999) used a compilation of trouble-indicating behaviors, repair trajectories and repair types, including those described by SchegloŸ (1992), in post hoc statistical analysis in their comparison of the contributions to the natural conversations of ten Alzheimer patients paired with unfamiliar students. Others (Crockford and Lesser 1994; Perkins 1995) have used the ratio of minimal turns to non-

When conversation is not normal 147

minimal as an index of changing conversational behaviors over time. Perkins’ study of two aphasic women and one aphasic man, recorded in conversations at home with their spouses, employed two measures: number of turns, as sensitive to the facility with which the impaired speaker responds to a turn transition relevance point, and ratio of minimal turns as an indication of the impaired person’s reliance on passive recipiency in maintaining the conversation. This statistical analysis was accompanied by a qualitative examination which revealed the partners’ roles in tolerating pauses and attempts at selfrepair. Perkins was able to relate these also to the peculiar nature of the linguistic impairments in each individual.

Variability between dyads One of the major questions which arise in attempts to compare conversational data over time concerns their dependence on situational characteristics and the dynamics between the conversational partners. It has frequently been demonstrated in clinical studies that success in conversation depends critically on the nature of the aphasic — nonaphasic interaction, a not-too-surprising observation given the mutual reliance on collaboration. Perkins’ study referred to above noted a marked reduction in collaborative sequences when one of the aphasic women was talking with her husband rather than the researcher. She attributed the diŸerence to the discrepancy between shared knowledge of the interlocutors, individual discourse styles and the relative severity of the woman’s word-ªnding di¹culties in accessing items in the phonological output lexicon. Lindsay and Wilkinson (1999) made a direct comparison between the conversational repair sequences of two aphasic people with their spouses and their speech and language therapists. Repairs in the conversations with the spouses showed a predominance of revision sequences which made the repair itself the focus of attention, and continued beyond the point where the target of the word search became known. These were absent in the therapists’ interactions, possibly due to a reluctance to expose incompetencies of the aphasic person. It is interesting to compare these observations with that of Heeschen and SchegloŸ, who felt that the behavior of a friend in conversation with an agrammatic speaker in using other-repair to emphasize an error was institutional-type, “like a therapist” (1999: 384), an interpretation not consistent with the more sensitive reactions of therapists reported elsewhere.

148 Discussing Conversation Analysis

Wilkinson (1995a) has developed this question of therapists’ sensitivity to interactions, following SchegloŸ’s (1988) drawing attention to the inadequacy of considering conversation as simply concerned with the transmission of information. Wilkinson looked at the trouble points in ªve conversations between therapists and aphasic people, where “the person’s identity as an aphasic has become relevant and [that] the participants can be seen to be orienting to this as an interactional problem” (op. cit.: 136). The particular points he selected were ones where, following some display of incompetence, the aphasic conversationalist laughs. In no case did the therapists share in the laughter to an equal degree, as might be expected in normal conversations as a sign of social solidarity, but either ignored it or responded only with a smile. This reduces the likelihood of the aphasic lapses in competence themselves becoming the conversational business, generating sequences of their own. The reinforcement of the aphasic person’s contribution to the conversation as being “non-ordinary” is thus avoided.

Variability within dyads across time In addition to the variability which occurs when people with aphasia are talking to diŸerent partners, an issue important when comparing conversations across time is the question of how consistent they are within the same dyad. Perkins, Crisp and Walshaw (1999) investigated this issue both quantitatively and qualitatively through analyses of four conversations from each of eight conversational partners, recorded at home at the same time of day over four weeks. The transcription and coding of episodes of collaborative repair was computerassisted. Collaborative repair was quantiªed as the proportion of major turns involved in this activity out of all major turns, given that all major turns in aphasia carry the potential of leading to collaborative repair. The statistical analysis investigated whether the variation between conversations by the same dyads obscured the predicted variation between diŸerent dyads. The results showed that there was signiªcant variation in collaborative repair within the dyads over the four episodes, but that that between the diŸerent dyads was much greater. Qualitative analysis revealed consistent patterns of trouble sources and resolutions of repair. Four couples consistently achieved fast resolutions of repairs. Three consistently used demonstration that understanding had been achieved as their commonest repair strategy, one used initiators such as “what?”, while four others showed use of a range of repair initiators. The

When conversation is not normal 149

authors comment that an extended baseline of repeated analyses could be used to strengthen the reliability of a quantitative analysis of changes over time to measure deterioration or improvement following therapy. Heeschen and SchegloŸ (1999) came to a rather diŸerent conclusion about consistency in their study of a woman with agrammatic aphasia in conversation with a friend on two occasions. The diagnosis of agrammatism was established through her responses to a clinical task of describing a series of cartoon pictures, during which she showed omissions and substitutions of function words and break-oŸ and incompleteness of sentence constructions. The measure of conversation used was not collaborative repair, but the degree of “telegraphic” speech employed by the patient. SchegloŸ and his clinical colleague speculate that the telegraphic style might be being deployed by the patient speciªcally as a resource to mobilize help from the friend. The authors claim that this supports the “adaptation theory” of agrammatism, in that such speakers can resort to a preverbal simpliªcation at the cognitive beginning of speech production which does not stress the patient’s impaired grammatical encoder. This study is a preliminary attempt to marry the top-down theorybased clinical analysis of an aphasic syndrome with the essentially bottom-up atheoretical approach on which conversation analysis is based. The authors describe this as “ ‘CA-assisted aphasiology’ diŸering from straightforward CA on the same materials by taking prior aphasiological understanding as part of its point of departure and as part of its problematic” (op.cit.: 378). In this it diŸers from the approach currently being taken in the application of conversation analysis to data from aphasic conversations in the UK, where the individuals are not classed by putative syndromes, and the interest is in applying conversation analysis to the peculiar abnormal circumstances which aphasic impairments characteristically impose. The motivation is not to test theories of syndromes but to use this type of analysis in assessment of individual interaction and, eventually, therapy.

Conversation analysis in assessment and therapy Given the insights it can oŸer into talk-in-interaction, conversation analysis is being increasingly used as one component of the clinical assessment of aphasic people. Its naturalistic setting oŸers a necessary complement to traditional ways of examining aphasia, which focus on deªning the nature of the impairment rather than its interactive consequences. Crockford and Lesser (1994)

150 Discussing Conversation Analysis

contrasted conversation analysis with some other published methods of examining “functional communication” in aphasia, comprising role play of social situations and questionnaire responses. Despite the amount of time data transcription and analysis take, even for ten-minute samples, they concluded that conversation analysis gave a clinically more useful result in identifying the communicative consequences of aphasia than did the structured methods. In an attempt to provide therapists with a practical instrument for capturing the clinical relevance of conversational analysis, Perkins, Whitworth and Lesser (1998) and Whitworth, Lesser and McKeith (1999) have produced a conversation analysis proªle which can be used for aphasic people and, with extensions, people with cognitive impairment. It combines an interview with the caregiver and an analysis of a ten-minute sample of natural conversation between caregiver and patient. Using the parameters of turn-taking, repair and topic management, its aims are fourfold: to determine the caregiver’s perception of the partner’s current conversational abilities; to identify what strategies are being used in the interactions, their success and the extent to which they are considered to be a problem; to assess changes from premorbid interaction; and to test the relationship between the caregiver’s perceptions and what actually happens in the recorded conversation. Despite the obvious restrictions in using a ten-minute audio-taped sample to capture the nature of the conversational interactions between the partners, Whitworth, Lesser and McKeith (1999) reported in a study of 12 people with Parkinson’s Disease and cognitive impairment that there was a 67.8% agreement between the caregivers’ reports and the ªndings of the conversation analyses, with the discrepancy being principally due to lack of evidence in the limited sample. The amount of quantiªcation permitted by the proªle also makes it relevant to assessing changes which occur due to deterioration or improvement, the latter perhaps linked with a program of intervention based on the ªndings. Using conversational analysis as a basis for therapy is still in its infancy, although stimulating natural conversation through educating partners and advising volunteer helpers has a longer history (e.g., Lesser and Watts 1978). Kagan (1995) describes a well-established program for training helpers and professionals in skills of revealing the underlying competences in aphasia by modifying their talk and using supporting non-verbal communication in conversations. She (Simmons-Mackie and Kagan 1999) supports with evidence her thesis that the interactions between aphasic and non-aphasic conversationalists need to reinforce the image of the person with aphasia as competent,

When conversation is not normal

trustworthy, interesting and sincere. The evidence comes from a study of four dyads, two rated by several judges as “good” communicators and two as “bad.” The good communicators used more acknowledgement tokens, more congruent overlaps of laughter, fewer disjunct markers (e.g., well), more willingness to accommodate to the aphasic person’s use of writing etc., and more completion of clariªcation sequences (whether successful or not) with mutual agreements. The researchers note that these observations do not provide a prescription for “good” communication, but were “contextually dependent behaviours that evolved out of collaboration between both participants in each interaction” (op. cit.: 818). Conversation analysis oŸers the opportunity to move from global advice to the partners of people with aphasia to individually-tailored observations. Lesser and Algar (1995) reported some success with an initial foray into applying individualized advice to two dyads based on conversation analysis combined with identiªcation of linguistic impairment. The dyads were given advice and booklets which cited speciªc examples of successful interactions taken from the recordings the two had made at home. Booth and Swabey (1999) used a pilot version of the conversational proªle described above as the basis for their intervention with the caregivers of two aphasic men and two aphasic women. Although the intervention took the form of six weekly meetings for the caregivers as a group, the advice given was individualized and related to examples from the tape recordings made in the individuals’ homes. Booth and Perkins (1999) provide a more detailed account of one of the couples, an aphasic man and his brother. The non-aphasic brother showed the feature described earlier of extensive use of other-repair, bringing the aphasic person’s troubles to the surface of the conversation and pursuing them, even to the extent of withholding collaboration in providing a known word. Correction of the aphasic person’s troubles thus dominated the conversation. The brother attended the group training sessions, during which he learned to understand the nature of the psycholinguistic consequences of aphasia and the individual’s speciªc linguistic impairments better, with more productive strategies such as using more facilitatory phonemic cues. After the six weeks of group training, a reassessment with the conversation proªle showed a marked reduction in the quantity of major turns involved in repair sequences, from 78% to 29%, with the average length of repair decreasing from 35 to 8 turns. The aphasic brother was now reported as occasionally starting up a conversation himself and introducing new topics, something not reported before the intervention.

151

152 Discussing Conversation Analysis

Wilkinson, Bryan, Lock, Bayley, Maxim, Bruce, Edmundson and Moir (1998) describe a similar case where the husband of a young woman with aphasia characteristically brought her phonemic di¹culties to the fore in extended repair sequences. A conversation analysis of a videotape made by the two at home incorporated, for example, a sequence of about 50 seconds where the husband was repeatedly trying to get his wife to produce /t/ at the end of the word left, which had occurred in the conversation. The attempt was unsuccessful, and caused distress to the wife. The therapy comprised four two-hour sessions, including a three-way discussion of the videotape with husband, wife and therapist, during which the husband acknowledged the interruption to conversation which these sequences caused and expressed the wish to discontinue them. The questionnaire section of the conversation proªle described above revealed topics on which it was suggested the couple could focus rather than repair. A videotape made after the therapy program showed no episodes of other-repair as identiªed before. These studies, therefore, show some promise in the application of conversation analysis in aphasia therapy, although this approach to intervention is still in its infancy.

Conclusion SchegloŸ’s ideas have provided the rootstock for most of the work described above. With few exceptions, his name features extensively in the reference lists of all the papers in which the work is described. The in¶uence of the methodology of conversation analysis is such that the undergraduate and graduate programs leading to qualiªcations in speech and language pathology and therapy necessarily include a training in the principles of such an analysis. But adaptations have inevitably been made for its application to the ªeld of language disorder after brain damage. Conversation analysis is essentially grounded in surface data, without theoretical assumptions, and the nature of language disorders is that they essentially aŸect such surface behaviors. Conversations which include aphasic and demented speakers will always therefore be inherently abnormal. Conversation analysis applied to the speech of the brain-damaged cannot, therefore, claim to provide a window on the normal functions of the brain such as is claimed for psycholinguistic interpretations. The conversation analysis of disordered language is an outgrowth of the discipline which is thus developing its own rules. While sensitive to the importance of locating each utterance in its sequential context, it incorporates a cautious

When conversation is not normal

component of quantiªcation. It includes a more important role for non-verbal contributions and body language than conversation analysis has traditionally done, with its origins in the study of telephone calls. It is also distinctive in its attempts to interfere with conversation in order to eŸect changes in interaction. Indeed it is its potential for improving the lives of those who suŸer the limitations of their language impairments which excites the burgeoning numbers of workers in this ªeld.

References Belin, Pascal, Zatorre, Robert J., Lafaille, Philippe, Ahad, Pierre and Pike, Bruce. 2000. “Voice selective areas in human auditory cortex”. Nature 403: 309–312. Boles, Larry. 1998. “Conversational discourse analysis as a method for evaluating progress in aphasia: A case report”. Journal of Communication Disorders 31: 261–274. Booth, Susan and Perkins, Lisa. 1999. “The use of conversation analysis to guide individualised advice to carers and evaluate change in aphasia: A case study”. Aphasiology 13: 283–303. Booth, Susan and Swabey, Donna. 1999. “Group training in communication skills for carers of adults with aphasia”. International Journal of Language and Communication Disorders 34: 291–309. Clark, Herbert H. and Schaefer, Edward F. 1987. “Collaborating on contributions to conversations”. Language and Cognitive Processes 2: 19–41. Couper-Kuhlen, Elisabeth and Selting, Margret (eds.). 1996. Prosody in Conversation: Interactional Studies. Cambridge: Cambridge University Press. Crockford, C. and Lesser, Ruth. 1994. “Assessing functional communication in aphasia: Clinical utility and time demands of three methods”. European Journal of Disorders of Communication 29: 165–182. Garcia, Linda J. and Joanette, Yves. 1997. “Analysis of conversational topic shifts: A multiple case study”. Brain and Language 58: 92–114. Goodwin, Charles. 1995. “Co-constructing meaning in conversations with an aphasic man”. Research on Language and Social Interaction 28: 233–260. Heeschen, Claus and SchegloŸ, Emanuel A. 1999. “Agrammatism, adaptation theory, conversation analysis: On the role of so-called telegraphic style in talk-in-interaction”. Aphasiology 13 (4–5): 365–405. Kagan, Aura. 1995. “Revealing the competence of aphasic adults through conversation: A challenge to health professionals”. Topics in Stroke Rehabilitation 2: 15–28. Klippi, Anu. 1996. Conversation as an Achievement in Aphasics. Helsinki: Suomalaisen Kirjallisuuden Seura. Laakso, Minna and Klippi, Anu. 1999. “A closer look at the ‘hint and guess’ sequences in aphasic conversation”. Aphasiology 13: 345–363. Lesser, Ruth and Algar, L. 1995. “Towards combining the cognitive neuropsychological and

153

154 Discussing Conversation Analysis

the pragmatic in aphasia therapy”. Neuropsychological Rehabilitation 5: 67–92. Lesser, Ruth and Milroy, Lesley. 1993. Linguistics and Aphasia: Psycholinguistic and Pragmatic Aspects of Intervention. London: Longman. Lesser, Ruth and Watts, M. 1978. “Untrained community help in the rehabilitation of stroke suŸerers with language disorders”. British Medical Journal ii: 1045–1048. Lindsay, Jayne and Wilkinson, Ray. 1999. “Repair sequences in aphasic talk: A comparison of aphasic-speech and language therapist and aphasic-spouse conversations”. Aphasiology 13: 305–325. Lubinski, R., Duchan, D. and Weitzner-Lin, B. 1980. “Analysis of breakdowns and repairs in aphasic adult conversation”. In Clinical Aphasiology Conference Proceedings, R. Brookshire (ed.), 111–116. Minneapolis: BRK Publishers. Lum, Carmen and Ellis, Andrew W. 1999. “Why do some aphasics show an advantage on some tests of nonpropositional (automatic) speech?”. Brain and Language 70: 95–118. Milroy, Lesley and Perkins, Lisa. 1992. “Repair strategies in aphasic discourse: Towards a collaborative model”. Clinical Linguistics and Phonetics 6: 27–40. Mummery, C. J., Patterson, K., Wise, R. J. S., Vandenbergh, R., Price, C. J. and Hodges, J. R. 1999. “Disrupted temporal lobe connections in semantic dementia”. Brain 122: 61–73. Oelschlaeger, Mary L. 1999. “Participation of a conversation partner in the word searches of a person with aphasia”. American Journal of Speech-Language Pathology 8: 62–71. Perkins, Lisa. 1995. “Applying conversation analysis to aphasia: Clinical implications and analytic issues”. European Journal of Disorders of Communication 30: 372–383. Perkins, Lisa, Crisp, Jenny and Walshaw, David. 1999. “Exploring conversation analysis as an assessment tool for aphasia: The issue of reliability”. Aphasiology 13: 259–281. Perkins, Lisa, Whitworth, Anne and Lesser, Ruth. 1998. “Conversing in dementia: A conversation analytic approach”. Journal of Neurolinguistics 11: 33–53. Ripich, Danielle N., Carpenter, Brian D. and Ziol, Elaine W. 2000. “Conversational cohesion patterns in men and women with Alzheimer’s Disease”. International Journal of Language and Communication Disorders 35: 49–64. SchegloŸ, Emanuel A. 1988. “Discourse as an interactional achievement II: An exercise in conversation analysis”. In Linguistics in Context: Connecting Observation and Understanding, D. Tannen (ed.), 135–158. New Jersey: Ablex. SchegloŸ, Emanuel A. 1992. “Repair after next turn: The last structurally provided defense of intersubjectivity in conversation”. American Journal of Sociology 97 (5): 1295–1345. SchegloŸ, Emanuel A. 1993. “Re¶ections on quantiªcation in the study of conversation”. Research on Language and Social Interaction 26: 99–128. SchegloŸ, Emanuel A., JeŸerson, Gail and Sacks, Harvey. 1977. “The preference for selfcorrection in the organization of repair in conversation”. Language 53 (2): 361–382. Simmons-Mackie, Nina and Kagan, Aura. 1999. “Communication strategies used by ‘good’ versus ‘poor’ speaking partners of individuals with aphasia”. Aphasiology 13: 807–820. Watson, Caroline M., Chenery, H. J. and Carter, M. S. 1999. “An analysis of trouble and repair in the natural conversations of people with dementia of the Alzheimer’s type”. Aphasiology 13: 195–218. Whitworth, Anne, Lesser, Ruth and McKeith, Ian. 1999. “Proªling conversation in Parkinson’s Disease with cognitive impairment”. Aphasiology 13: 407–425.

When conversation is not normal

Wilkinson, Ray. 1995a. “Doing ‘being ordinary’: Aphasia as a problem of interaction”. In Work in Progress, Vol. 5, M. Kersner and S. Peppe (eds.), 134–149. Department of Human Communication Studies: University College London. Wilkinson, Ray. 1995b. “Aphasia: Conversation analysis of a non-¶uent aphasic person”. In Case Studies in Clinical Linguistics, M. Perkins and S. Howard (eds.), 271–292. London: Whurr. Wilkinson, Ray. 1999. “Sequentiality as a problem and resource for intersubjectivity in aphasic conversation: Analysis and implications for therapy”. Aphasiology 13: 327–343. Wilkinson, Ray, Bryan, K., Lock, S., Bayley, K., Maxim, J., Bruce, C., Edmundson, A. and Moir, D. 1998. “Therapy using conversation analysis: Helping couples adapt to aphasia in conversation”. International Journal of Language and Communication Disorders 33: 144–149.

155

Chapter 8

Response* Emanuel A. SchegloŸ

The interview presented in Chapter 2 of this volume has given me more than ample opportunity to express myself (to “sound oŸ,” in the American idiom) on a range of matters related to my work and that of my closest colleagues, and to take up issues raised by some of the reactions to that work in neighboring disciplines. It therefore behooves me to exercise discretion and self-control when presented with yet a further opportunity to sound oŸ. In what follows, then, I try to suppress my garrulous side in favor of my more terse one. My good fortune in having colleagues like John Heritage and Chuck Goodwin is exempliªed not only in the quality of the original research work that they do, but also in the generosity displayed in their contributions to this volume. The synoptic overview of the disciplinary nexus and intellectual ancestry of the work in conversation analysis provided by John Heritage is tellingly complemented by Chuck Goodwin. His is an intersecting account of much of the same terrain, as encountered from the trenches of a then-graduate student’s astute and insistent eŸorts to ªnd sensible and eŸective resources for coming to terms with the naturally-occurring events of people’s lives in the world, at the level of detail to which we had long been alerted by work in the arts and letters, and which was no longer to be denied to observant workers in the social and human sciences. It has been my good fortune to ªnd readers and colleagues who were attracted by the same things that intrigued me, who were able to ªnd something valuable in what I have done, and who, by making something of it and going beyond it, have enhanced whatever value it may turn out in the longer term to have had. If conversation analytic work ends up having made the contribution I think it can make, it will because of the convergent and cumulative work of a community of inquiry; here, as elsewhere, the central note is the shaping and realization of an enterprise as a collaborative and interactional outcome. John and Chuck have been important ªgures in that enterprise, and I am grateful to them for the generosity of their comments.

158 Discussing Conversation Analysis

The other contributors to this volume I have never met, and I thank them for their willingness to invest time and eŸort in it. The stances they take up are quite diŸerent from one another, of course, and re¶ect the very diŸerent scholarly, professional, scientiªc and academic contexts and commitments characterizing their work. In keeping with my above-taken pledge, my responses to their several contributions will be brief, but, I hope, to the point. There are two elements in Ruth Lesser’s contribution which I would like to address. One is the richly informative account of some of the ways in which work in conversation analysis has been taken over by language pathologists, aphasiologists and other clinicians, which has enriched their capacity to do their work — both research and therapeutic — and has been subjected to modiªcations in coming to address the contingencies of work on so-called “disordered language.” The work which Lesser reviews goes well beyond that with which I was familiar, and it is both exciting and gratifying to know of the appeal of CA to workers in this historically quite separate area, and of its anticipated e¹cacy for improving the lives of those who suŸer from the limitations of language impairments. To my mind, this is a vindication of our hope that coming to understand the underlying organizations of practice which inform the organization of any talk-in-interaction should contribute to our understanding of any setting in which talk-in-interaction occurs. Even though this may initially serve to introduce “problems” for researchers and practitioners in this ªeld — as Lesser terms the areas which serve to organize her review of the literature — it will serve (one hopes) to bring the ªeld into more intimate contact with the naturally occurring home of the conduct and troubles which supply the ªeld’s mandate. Lesser’s discussion vindicates as well my sense that workers using CA work and methods ought, as a matter of course, to try to stay in touch with both work in their special area — aphasia, stuttering, etc. — and with ongoing research and the data of “ordinary conversation.” She remarks, for example, on a ªnding which registers “an aphasic tendency to use words which have already been used by another speaker” (this volume: 145). In work which I have been doing and teaching for some time, but which remains unpublished, such a tendency has turned up as a recurrent feature of ordinary talk-in-interaction by persons with no known neurological compromise. It may make quite a diŸerence for both students of “normal” talk-in-interaction and of talk which is impacted by untoward neurological events to know that this practice is common to both, and is not itself (for example) a “symptom” of some neurological disease process. Perhaps some day it will be part of the expected profes-

Response 159

sional/scientiªc activities of language pathologists and aphasiologists to take a continuing active research interest in ordinary talk-in-interaction as part of their growing leverage on the characteristics of “disordered” talk and how it can be addressed. I will be briefer in my second remark on Professor Lesser’s contribution. The point of departure in the interview for her discussion of CA’s bearing on work in language pathology and aphasiology is my interest in the “describing the behavior” part of the recent surge of interest in the “neurobiology of behavior.” I take her point about the di¹culty, at the current stage of work and of technology, of bringing brain-imaging methods to bear on the sorts of things conversation analysts describe. Still, one important aspect of CA work is its commitment to getting at the practices and resources of ordinary conversation in the terms in which they are oriented to by participants. That is, that they are natural categories being discovered, and not analytic ones being imposed. If that is the case, then these categories of resources and practices must inform behavior or conduct in its very course of production, on the one hand, and uptake, on the other. And this is just the sort of thing which should “interface” with a neurological substrate, if there is in fact an interface to be speciªed. Attending such a possibility may help to encourage the development of a technology appropriate to the implied inquiry, even if it does not now exist. If, as Lesser reports at the very outset of her contribution, functional magnetic resonance imaging can capture such categories as “lexical-phonological” and “semantic,” or “practiced/automatic” behaviors and “novel” ones, then why not “turns” and their possible completion, or “ªrst-pair parts” and “secondpair parts” — categories of conversational events which demonstrably have extremely robust recurrent consequences for the generation of behavior in talk-in-interaction? Rick Iedema is coming from an altogether diŸerent direction. He seems to wish to treat my work as a foil for his own commitments and preoccupations. He has collated extracts from the interview and various past papers of mine to construct a supposed representation of my “position” on some matters on which he has a diŸerent position. I say a “supposed representation,” because he has selected his quotations carefully and refracted them through a prism of his own making to project a position on to my work which is designed more for its vulnerability to his critique than for its adequacy as a representation of the object of the critique. How or why else would he come to the astonishing (to me, at least) assertion that “SchegloŸ’s inquiry […] is ultimately a critique of understandings grounded in everyday commonsense: a démasqué of false con-

160 Discussing Conversation Analysis

sciousness which normally permeates the common and everyday” (Iedema,, this Volume 67)?!?! To rectify only the misreadings or misinterpretations of the interview in the ªrst several pages of his contribution would exceed the limits of sensible investment in this exchange. Readers so inclined can see for themselves by (re-)reading the interview in juxtaposition with Iedema’s account of it. Pirkko Raudaskoski’s contribution, like Ruth Lesser’s, considers the possibility that CA may have some implications for domains of event diŸerent in a greater or lesser degree from that which gave it rise and which supplies its core subject matter. I share her hopes that this may be the case. However, her enterprise is rather further removed from ordinary conversation than the disordered talk implicated in Lesser’s ªeld, and is, I fear, rather more problematic. As it happens, the material which Raudaskoski has examined — humans engaged with a computer — does not really involve “conversation” or “interaction” in the usual sense in CA at all. Rather, it involves dealing with a “text” whose “sender” is not present. Raudaskoski means to investigate “how asynchronously produced language is interpreted;” she is concerned with misunderstanding, and undertakes to bring the resources developed in CA under the rubric of “repair” to bear on this problem. The entire undertaking, however, rests on several assumptions which seem to me problematic for the sort of work I take to be at the heart of the interests which I bring to this volume. Raudaskoski writes: […] the whole encounter between a text and a reader can be analyzed as social interaction unfolding in time and space which, in addition to the local history of the encounter, also forms an integral context of understanding or misunderstanding. (Raudaskoski, this volume: 114–115)

And following up on this premise, she ascribes to the computer the taking of turns, the initiation of repair, etc.; for example: Extract (1′) begins with the repair initiator that appeared at the lower part of the screen after a wrong action by B. Now there is a clear connection between the computer’s turn in line 102, and what B says in lines 109 (“move the mouse pointer”) and 111 (“click l”) [:..] (Raudaskoski, this volume: 118).

And so on. Proceeding in this fashion may make sense in the domain in which Raudaskoski works. To me it is deeply problematic. The encounter between a text and a reader, or between a computer and a user, can not properly be analyzed as social interaction, etc., in the same sense

Response

as ordinary interaction between humans can. To assert that is to reveal either a) a fundamental misunderstanding of text-reader/computer-user processes, ordinary human interaction, or both, or b) a willingness to settle for analogic analysis prematurely, that is, before exhausting the eŸort to describe the object of interest in its own terms. Whenever my students say about some fragment of data, while working with me, “this is like an X,” I always counsel against it. To say it is “like an X” is to concede that it is not an X — else one would have said, “this is an X.” And it is to give up the search for what it is, and settle for what it is like. From what I know about ordinary human interaction, human/computer “interaction” is not seriously “like it.” At least, that has not been shown, to my knowledge. And considerable mischief and errors can be incurred by proceeding on that basis. And the text/computer can not be seen as taking turns, at least not in the sense in which conversation analysts properly speak of taking turns (and I must assume that that is what is here intended, given the auspices of this volume, and this article being a contribution to it). Again, analogical analysis is premature. In settling for the computer acting like it was taking a turn by writing, “[…] can be seen as taking turns […],”or by putting ‘turn’ in scare-marks, we are discouraged, even stopped, from specifying exactly what the computer may actually be understood to be “doing” in the kind of interactional phenomenology which Raudaskoski appears to be pursuing. There is now a substantial literature on what is involved in taking a turn in conversation and in timing such a taking, or delaying it, etc., or declining to take it; how that is contingent on a running analysis (i.e., synchronously at a ªne level of granularity) by the prospective turn-taker of the turn in progress before — and issuing in — that “taking of a turn”; and so on and so forth — none of which a computer demonstrably does (and cannot do if the whole process is asynchronous). And since the organization of repair, which is also invoked here, is organized in substantial measure by reference to turn-taking organization and the relationship between one turn and that which follows or precedes it (cf. SchegloŸ et al. 1977; SchegloŸ 1992, inter alia), the relevance of the CA literature on repair in this context is at most analogical. And to so deploy it is to risk misleading readers who lack their own independent access to that literature about what its claims and ªndings are, and on what they rest. I must enter one more general demurrer here: it is a central feature of those actions we term repair-initiations that they put the ongoing course of action — whether turn or sequence — “on hold” in order to address some trouble in speaking, hearing or understanding the talk, with resumption of that activity

161

162 Discussing Conversation Analysis

following solution of the trouble or abandonment of the eŸort to deal with it. When the main activity in progress is instructional, special care is needed to distinguish actions which promote or advance that mainline activity and are parts of it from those which put it into temporary abeyance in order to initiate dealing with trouble in speaking, hearing or understanding the talk. Di¹culties in making this discrimination appear relevant to Raudaskoski’s project, whatever else may be said about the invocation of “turns” and “repair” to the understanding of its very diŸerent “universe of discourse.” I hope I can be forgiven for not entering into any greater detail into this analysis. I understand that when authors put a piece of work into the public domain, they largely cede control over what will be made of it. It may well be that Raudaskoski’s work makes an important or useful contribution to the ªeld in which she works. If so, I hope its eŸects continue to be of beneªt. But they do not serve well a proper grasp of CA, at least as I understand it, even if aspects of CA work have served as a model or inspiration. Pär Segerdahl’s contribution reminded me of a striking utterance in a fragment of conversation presented in the turn-taking paper (Sacks et al. 1974: 715). A young married couple are in the car on the way to her parents’ home for Sunday brunch, and she says to him (presented here in regularized orthography), “That’s a really nice sweater; that’s my favorite sweater on you; it’s the only one that looks right on you.” Although introduced in that paper for a diŸerent reason, I am always struck on encountering this utterance by the rapidity with which what started out as a compliment ends up as a critique. Segerdahl starts me oŸ in the company of Husserl, but ends me up as yet another naive believer in the omnipotence of science — no, of his own science; he starts me as a noble rebel against the deadening constraints of academic disciplines, and ends me up as a reproducer of that same prison — indeed a worse one. What can I say? None of these assessments is warranted, in my view. I cannot here go through all the misunderstandings which seem to me to underlie Segerdahl’s discussion, from his belief that I believe that people do things because of rules like a terminating rule which makes them do such things, to his belief that we impute features of our research stance to the world under study rather than using the latter to constrain and shape the former, to the relevance of people’s reactions if we presented them with an account of ordinary behavior framed in technical terms, etc.. For a while, I thought that a reading of some of my work written after 1981 would make a diŸerence (as their respective initial notes make clear, the “1984” paper which Segerdahl cites was written in 1972; the “1986” one which he cites is a reprint of a 1968 paper;

Response 163

he has apparently read nothing more recent), but it became clear to me that nothing would make a diŸerence; that it is a principled philosophical stance that is being exercised here, and is being brought to bear on this work. Naive as it may be, I have not given up on the accessibility of the world to empirical inquiry; on our capacity to frame accounts in ways that are more or less correct, in ways that capture better and worse the ways that some things are organized; and, most striking of all, on the special and distinctive leverage for such inquiry that ordinary human interaction, and especially talk-in-interaction, and conversation in particular, aŸords us because of its very interactivity, because of the ways in which subsequent utterances and other forms of action give us access to the endogenous understanding of what is going on, on which the participants build the ensuing trajectory of the interaction — in other words, which give us access to the terms of human practice in interaction itself. Pursuing inquiry along these lines, we have been led to ªnd things that appear real, that are fresh once found, that seem to give us access of a sort we did not have to how it is with humans. Philosophy has often told empirical inquiry what it could not do, only to have to revise its view once that was done — successfully. We should try once again.

Note * Second, revised version received 1st January 2001.

References Iedema, Rick. This volume. “Putting SchegloŸ’s principles and practices in context”. Sacks, Harvey, SchegloŸ, Emanuel A. and JeŸerson, Gail. 1974. “A simplest systematics for the organization of turn-taking for conversation”. Language 50 (4): 696–735. SchegloŸ, Emanuel A. 1968. “Sequencing in conversational openings”. American Anthropologist 70 (6): 1075–1095. SchegloŸ, Emanuel A. 1984. “On some questions and ambiguities in conversation”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J. M. Atkinson and J. Heritage (eds.), 28–52. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1990. “On the organization of sequences as a source of ‘coherence’ in talk-in-interaction.” In Conversational Organization and its Development [Advances in Discourse Processes, Vol. XXXVIII], B. Dorval (ed.), 51–77. Norwood, NJ: Ablex.

164 Discussing Conversation Analysis

SchegloŸ, Emanuel A. 1992. “Repair after next turn: the last structurally provided place for the defence of intersubjectivity in conversation.” American Journal of Sociology 95 (5): 1295–1345. SchegloŸ, Emanuel A., JeŸerson, Gail and Sacks, Harvey. 1977. “The preference for selfcorrection in the organization of repair in conversation”. Language 53 (2): 361–382.

Chapter 9

Continuing the interview with Emanuel A. SchegloŸ * Carlo L. Prevignano and Paul J. Thibault

Carlo L. Prevignano: CA is founded on the idea of turn. What were your reasons for not adopting the GoŸmanian idea of “move,” or, if you like, what diŸerences did and do you see between “turn” and “move”? Emanuel A. SchegloŸ: First of all, to my mind, CA is not “founded on the idea of ‘turn’,” either analytically or historically. Analytically, there are several organizations of practice that can be treated as “generic.” By that I mean that they are addressed to organizational issues, without which conversation, or talk-in-interaction more generally, cannot proceed in the manner in which it observably does proceed. Sustained episodes of talk-in-interaction imply, rest on, and embody ongoing solutions to those issues. Among these are the organization of turn-taking, the organization of sequences, the organization of repair, and the overall structural organization of single episodes of conversation (or other forms/occasions of interaction); there are others. The unit “turn” (and “turn-constructional unit”) is basic to turn-taking organization, and can in that sense be treated as one building block foundational to conversation, and consequently to Conversation Analysis. But no more so than the adjacency pair is basic to the organization of sequences, with the same ‘foundational’ import. Historically, I suspect that the turn has been accorded the centrality you refer to because of its intersection with the discipline of linguistics. The publication of the turn-taking paper (Sacks et al. 1974) in the journal Language had something to do with that — with its invitation to linguists to contribute to this undertaking. And, as I have argued elsewhere (SchegloŸ 1989, 1996), the turn at talk seems to be the natural, “ecological” home for units of language production, whether thought of as “sentences” or in other ways and terminologies, and so, when some linguists were gradually attracted to this work, it was to what happens in turns — and the interactional contin-

166 Discussing Conversation Analysis

gencies bearing on turns — that they were most naturally drawn. But this has more to do with history and traditions of linguistics (and linguists) than with CA. If you look at the earliest work, Sacks’ earliest lectures and papers and my own, they are not in the ªrst instance about turns. If they “feature” anything, they feature action, “sequences,” and word selection (i.e., formulations or identiªcation or reference). Secondly, we did not abandon the notion of “move,” although we use the term itself only infrequently. But central to the bulk of CA work is “action,” and what is getting done by some feature of the talk or other conduct. As far back as “Opening Up Closings” (SchegloŸ and Sacks 1973) in published papers, and earlier in Sacks’ Lectures, we understood the basic issue for parties to interaction to be “why that now,” and the default answer for participants has to do with what a speaker or other participant is doing by talking or conducting themselves otherwise (i.e., in physically-realized conduct) in the way they are. Its centrality for interactants mandates the centrality for CA of understanding how that works. In a global sense, I’m not sure how much diŸerence there is between this preoccupation in CA and what GoŸman had in mind by “move.” And I’m not sure how much diŸerence it makes; the issue is how best to get at what is going on in interaction and how it comes to be going on, and do so in an empirically grounded way. C.L.P.: A much longer question. Suppose what you call “Italian conversation” (SchegloŸ et al. 1996: 30) shows also a really “massive overlapping,” a parallelism of turns or moves by the conversationalists, i.e., a “diŸering organization.” Then a turn-centered idea of interaction even as a default turn-after-turn pattern, in that case, might no longer be su¹cient. In order to cope with “the multi-vocality of Italian conversation” (SchegloŸ et al. 1996: 32), when Italian conversationalists leave or don’t adopt the turn-following mode, I’m thinking about a more general model, a multi-track or n-track model of conversation (where n = the number of conversationalists). Every conversationalist moves on his/her track, but how does a bundle of tracks or trajectories make a conversation? Beyond the problems of track-taking, track-leaving, etc., conversationalists have also that of following the other conversationalists moving on their own tracks. According to this perspective, turns could be reconsidered and generalized as track units or constituents, with new problems concerning contacts or relationships between tracks and between track constituents. What do you think of this?

Continuing the interview with Emanuel A. SchegloŸ 167

E.A.S.: First of all, there is no particular investment, at least on my part, in the account of turn-taking provided in Sacks et al. 1974 and extended in SchegloŸ 2000. In the 1974 paper, we said explicitly, and more than once, that the paper was almost certainly wrong in various places and respects, and would in retrospect appear overly coarse and unspeciªc. The investment is in getting it right, or as right as we can. Should someone provide an empirically-grounded account of a turn-taking organization in operation under naturally-occurring circumstances which is diŸerent from what has been described (for conversation; we already know that turn-taking is diŸerent in other speech-exchange systems), then that will be not a “defeat,” but an advance. It should allow us to formulate a more general or formal account of turn-taking which could be shown to subsume the previous (Sacks et al.) account and the new one as alternative speciªcations. Second, the description you put forward in your question of a putative turn-taking organization for “Italian conversation” (a notion which was not of my devising, but one summoned up by others, on which I was commenting) is not unfamiliar. It has quite a bit in common with the account oŸered some years ago by Karl Reisman (1974) of so-called “contrapuntal conversation” in Antigua. However, when talk in what is arguably the same genre in the same “speech community” or family of speech communities was examined in recorded and transcribed form, which allowed more detailed, rigorous and veriªable examination, this account of turn-taking was called into question, to say the least (Sidnell 2001). It seems as improbable to me for conversation in Italy, but that is an empirical question. To take the matter up seriously, someone should assemble a corpus of conversational material in state-of-theart form (i.e., with video if the talk is co-present, etc.), get it properly transcribed as a method for mobilizing appropriately detailed examination, and provide an account of a turn-taking organization being deployed by the parties in that talk that is diŸerent than what has previously been described — whether like the one suggested in the question or not. Then one can seriously ask what the import of these ªndings is. Paul J. Thibault: From your perspective, could you clarify the bi-directional relationship between the levels of grammar, or linguistic forms, and turns in conversation and their organization? In other words, how do you see constraints emanating from the level of grammar operating in relation to turntaking in conversation and the construction of turns, and, secondly, how does turn-taking aŸect or impinge on grammar and its speciªc contribution?

168 Discussing Conversation Analysis

E.A.S.: This is not a question for me to answer, but a research program to be pursued by many investigators (several contributions to the Ochs, SchegloŸ and Thompson volume (1996) are addressed to such a research program, including my own (SchegloŸ 1996)). It is a matter not of “perspective,” but of analysis and the ªndings of sustained careful examination of data. A substantial number of people have begun addressing themselves to such questions, and doing so across a range of languages. It seems to me too early to oŸer “answers” to these questions yet. In fact, it is probably too early to know yet whether these are the right questions, or ones formulated in the most productive way. In general, it has in the past proven best to start not with questions, but with observations. If one starts with questions — especially ones not grounded in data but in “the literature” or “theoretical imagination” — and they are not formulated quite right, much time can be wasted in a search for answers which fails and leaves one with nothing enduring. If one starts with observations about data, one is already “ahead of the game” — there is already “news” in hand. And the questions one may use the data to pose have the advantage that they are grounded in observable reality, and so the questions too may turn out to be about observable reality, and that enhances the chances that the answers will be too. C.L.P.: How do you consider the relations between CA and interactional linguistics? E.A.S.: I suppose that this is a “work in progress,” and it remains to be seen what the relationship between the two will turn out to be. At present, “interactional linguistics” seems to be quite close to CA, adopting its premises, grounding much of its research and many of its problems in past CA work, etc.. The diŸerentiation in name at the present time is, I suspect, to some degree an eŸort to insist on the relevance of this work to Linguistics as a discipline within the contemporary university structure, indeed, to its proper membership in the canon of the ªeld, and thereby to legitimate the conduct of this work within Departments or Faculties of Linguistics — in teaching, in research, in the training of students and the granting of degrees, and the hiring of faculty. It seems to me outrageous that the sort of work in question, addressed as it is to topics which have traditionally been part of Linguistics — whether in grammar, in phonetics and phonology, in pragmatics, etc. — should have its legitimacy called into question for turning attention to new sorts of data, new sorts of ªndings about it, new modalities of research, all of which are clearly preoccupied with language, its structure, its organization, and its realization and deployment

Continuing the interview with Emanuel A. SchegloŸ 169

in the natural world. Often such work is dismissed as “untheoretical,” but in fact it is only responsive to diŸerent theoretical initiatives than are for the moment predominant in the ªeld. But there are several other sorts of issues that may be involved in the diŸerentiation, and these invite thought and discussion. One involves the sociological roots of CA and linguists’ lack of interest in them. Another concerns matters of traditional interest to linguists which have not found a place in CA. These complementary divergences may appear to invite resolution by pursuing a diŸerently framed discipline or sub-discipline. I must confess some misgivings about this solution, and I’ll try to sketch them brie¶y. As for the sociological roots of CA, it seems to me worth the eŸort for colleagues trained in other ªelds who ªnd CA of serious relevance to their interests to give some attention to its origins. The critical stance which CA embodies towards Sociology’s intellectual legacy (some of it taken over from Ethnomethodology, some distinctive to CA) has left an important mark on CA itself, and underlies why CA is set up to work the way it does, and why we resist the lines of working and theorizing which we resist. It underlies the centrality of the very data that supply our focus and the theoretical import of such data (beyond their mere accessibility). These choices and avoidances are not arbitrary or conventional or historical artifacts; they are the product of reasoned engagement with serious past practices in sociology and other social sciences. Those who want to build on CA work and to extend it, or, for that matter, to alter it, would be well served by knowing its roots, its background, and its rationale. Indeed, it would be useful to subject the legacy of Linguistics to the same sort of critical scrutiny that informed the emergence of CA as a distinct undertaking, rather than to simply accept the legacy of the ªeld — including its conventional sub-divisions — as the shaper of contemporary work. Or will resources and analytic discriminations be carried over from a Linguistics grounded in very diŸerent data into the study of the varieties of naturally-occurring talk-ininteraction which set our contemporary agenda? It is not only the data which are very diŸerent, but what is taken to feature most centrally in our understanding of the talk and its resources: action — what the parties are doing — comes to ªgure centrally, in a way that had little place or precedent in the Linguistics of the past. If there is any merit to this suggestion, then some interests derived from training in traditional linguistic concerns might be better set aside as no longer central or even viable, rather than becoming the raison d’être for a separate subdiscipline.

170 Discussing Conversation Analysis

It seems to me that one key desideratum now is the prospect of colleagues coming from Linguistics trying increasingly to situate their past training within a broader analytical preoccupation with the basic, formative sources and structures of interaction as we understand them. These focus most centrally on action in interaction, and the social occasions and social relationships in which they are embedded and which they in turn shape, more than on propositions, information, sentences, etc.. As conversation analysts have tried to be responsive to initiatives taken by linguists in exploring, for example, prosody in conversation, so we look forward to linguists developing their research orientation to feature the action and interactional practices which are at the ground level of the data of conversation and other talk-in-interaction. Is this prospect enhanced or endangered by a separate interactional linguistics? As I said at the outset, it is a work-in-progress; we shall see what develops.

Note * To conclude this volume, Emanuel A. SchegloŸ kindly agreed to respond to the following questions on the part of the editors; received 11th February 2001.

References Ochs, Elinor, SchegloŸ, Emanuel A. and Thompson, Sandra A. (eds.). 1996. Interaction and Grammar [Studies in Interactional Sociolinguistics]. Cambridge: Cambridge University Press. Reisman, Karl. 1974. “Contrapuntual conversations in an Antiguan village”. In Explorations in the Ethnography of Speaking, R. Bauman and J. Sherzer (eds.), 110–124. Cambridge: Cambridge University Press. Sacks, Harvey, SchegloŸ, Emanuel A. and JeŸerson, Gail. 1974. “A simplest systematics for the organization of turn-taking for conversation”. Language 50 (4): 696–735. SchegloŸ, Emanuel A. 1989. “Re¶ections on language, development, and the interactional character of talk-in-interaction”. In Interaction in Human Development, M. Bornstein and J. S. Bruner (eds.), 139–153. Hillsdale, NJ: Lawrence Erlbaum Associates. SchegloŸ, Emanuel A. 1990. “On the organization of sequences as a source of ‘coherence’ in talk-in-interaction”. In Conversational Organization and its Development [Advances in Discourse Processes, Vol. XXXVIII], B. Dorval (ed.), 51–77. Norwood, NJ: Ablex. SchegloŸ, Emanuel A. 1992. “Repair after next turn: The last structurally provided defense of intersubjectivity in conversation”. American Journal of Sociology 97 (5): 1295–1345. SchegloŸ, Emanuel A. 1996. “Turn organization: One intersection of grammar and interaction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13],

Continuing the interview with Emanuel A. SchegloŸ

E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.): 52–133. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A., Ochs, Elinor and Thompson, Sandra A. 1996. “Introduction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E. A. SchegloŸ and S. A. Thompson (eds.), 1–51. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 2000. “Overlapping talk and the organization of turn-taking for conversation”. Language in Society 29 (1): 1–63. SchegloŸ, Emanuel A., JeŸerson, Gail and Sacks, Harvey. 1977. “The preference for selfcorrection in the organization of repair in conversation”. Language 53 (2): 361–382. SchegloŸ, Emanuel A. and Sacks, Harvey. 1973. “Opening up closings”. Semiotica VIII (4): 289–327. Sidnell, Jack. 2001. “Conversational turn-taking in a Caribbean English Creole”. Journal of Pragmatics 33 (8): 1263–1290.

171

Chapter 10

A bibliography of Emanuel A. SchegloŸ Edited by Susan L. Eerdmans

Introduction When I accepted to edit this bibliography of Emanuel SchegloŸ, I had no idea it would be such an enriching experience. Having been interested in Conversation Analysis for some time, I was naturally already familiar with SchegloŸ’s work; however, I hadn’t realized to what extent his research, re¶ected so eloquently through his publications, had delved with such systematicity into the intricacies of interactional behavior. During my editing, I have certainly learnt a great deal about the way and the directions in which his work has been developed over the past thirty-ªve years, and I feel stimulated to expand my still very incomplete knowledge of the fascinatingly detailed organization of naturally-occurring talk-in-interaction. In keeping with the general methodological stance of conversation analysts, I have tried to proceed in a systematic way in order to present a clear, orderly picture of the wealth of ªndings published by Emanuel SchegloŸ since his ªrst ground-breaking article in 1968. This has involved making certain decisions regarding the details to be included in the presentation, as well as the order to be adhered to; in general, my policy has been to be as transparent as possible, with the aim of enhancing the accessibility of the information supplied. Of course, my work would have been impossible without Emanuel SchegloŸ’s collaboration in providing the fundamental bibliographical material, and in supplying further information when necessary.

174 Discussing Conversation Analysis

Publications SchegloŸ, Emanuel A. and Kruytbosch, Carlos. 1961. “Some comments on working class authoritarianism”. Berkeley Journal of Sociology VI (1): 99-105. SchegloŸ, Emanuel A. 1963. “Toward a reading of psychiatric theory”. Berkeley Journal of Sociology VIII: 61-91. SchegloŸ, Emanuel A. 1968. “Sequencing in conversational openings”. American Anthropologist 70 (6): 1075-1095. Reprinted in: Fishman, Joshua A. (ed.). 1972. Advances in the Sociology of Language, Volume 2: Selected Studies and Applications [Contributions to the Sociology of Language 2], 91-125. The Hague: Mouton and Co.; Gumperz, John J. and Hymes, Dell (eds.). 1972. Directions in Sociolinguistics: The Ethnography of Communication, 346-380. New York: Holt, Rinehart and Winston; Laver, John and Hutcheson, Sandy (eds.). 1972. Communication in Face to Face Interaction [Penguin Modern Linguistics Readings], 374–405. Harmondsworth, U.K.: Penguin Books; Coulter, JeŸ (ed.). 1990. Ethnomethodological Sociology, 271–305. Aldershot: Edward Elgar Publishing Limited. Reprinted, with editorial changes, as: Schegloff, Emanuel A. 1973. “Opening conversations”. In Urbanman: The Psychology of Urban Survival, J. Helmer and N.A. Eddington (eds.), 142–169. New York: The Free Press/London: Collier-MacMillan Ltd. SchegloŸ, Emanuel A. 1972. “Notes on a conversational practice: Formulating place”. In Studies in Social Interaction, D.N. Sudnow (ed.), 75-119. New York. The Free Press/ London: Collier-MacMillan Ltd. Reprinted in: Coulter, JeŸ (ed.). 1990. Ethnomethodological Sociology, 306–352. Aldershot: Edward Elgar Publishing Limited. Excerpts reprinted in: Giglioli, Pier Paolo (ed.). 1972. Language and Social Context [Penguin Modern Sociology Readings], 95-135. Harmondsworth, U.K.: Penguin Books. SchegloŸ, Emanuel A. and Sacks, Harvey. 1973. “Opening up closings”. Semiotica VIII (4): 289-327. Reprinted in: Baugh, John and Sherzer, Joel (eds.). 1984. Language in Use: Readings in Sociolinguistics, 69-97. Englewood CliŸs, N.J.: Prentice-Hall.

A bibliography of Emanuel A. SchegloŸ

Slightly abridged version in: Turner, Roy (ed.). 1974. Ethnomethodology: Selected Readings [Penguin Modern Sociology Readings], 233-264. Harmondsworth, U.K.: Penguin Books. Sacks, Harvey, SchegloŸ, Emanuel A. and JeŸerson, Gail. 1974. “A simplest systematics for the organization of turn-taking for conversation”. Language 50 (4): 696-735. Reprinted in: Hodge, Robert (ed.). 1989. Readings in Language and Communication for Teachers, 165175. Melbourne: Longman Cheshire House; Kasher, Asa (ed.). 1998. Pragmatics: Critical Concepts, Volume 5 (10): Talk-in-Interaction, 193-242. London: Routledge. Variant version (prior to editorial emendations for publication in Language) in: Schenkein, Jim (ed.). 1978. Studies in the Organization of Conversational Interaction, 755. New York: Academic Press. Abridged Italian translation in: Giglioli, Pier Paolo and Fele, Giolo (eds.). 2000. Linguaggio e Contesto Sociale, 97-135. Bologna: Il Mulino. SchegloŸ, Emanuel A. 1976. “On some questions and ambiguities in conversation”. Pragmatics Microªche 2.2: D8-G12. Abridged version reprinted in: Dressler, Wolfgang U. (ed.). 1978. Current Trends in Textlinguistics [Research in Text Theory, Vol. 2], 81-102. Berlin and New York: Walter de Gruyter; Atkinson, J. Maxwell and Heritage, John C. (eds.). 1984. Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], 2852. Cambridge: Cambridge University Press. JeŸerson, Gail, Sacks, Harvey and SchegloŸ, Emanuel A. 1976. “Some notes on laughing together”. Pragmatics Microªche 1 (8): A2-D9. SchegloŸ, Emanuel A. 1977. “Identiªcation and recognition in interactional openings”. In The Social Impact of the Telephone, I. De Sola Pool (ed.), 415-450. Cambridge, MA: M.I.T. Press. Expanded version published as: SchegloŸ, Emanuel A. 1979. “Identiªcation and recognition in telephone conversation openings”. In Everyday Language: Studies in Ethnomethodology, G. Psathas (ed.), 23-78. New York: Irvington Publishers. SchegloŸ, Emanuel A., JeŸerson, Gail and Sacks, Harvey. 1977. “The preference for selfcorrection in the organization of repair in conversation”. Language 53 (2): 361-382.

175

176 Discussing Conversation Analysis

Reprinted in: Psathas, George, Coulter, JeŸ and Frankel, R. (eds.). 1990. Interaction Competence, 3161. Washington, D.C.: University Press of America; Kasher, Asa (ed.). 1998. Pragmatics: Critical Concepts, Volume 5 (10): Talk-in-Interaction, 243-272. London: Routledge. SchegloŸ, Emanuel A. 1979. “The relevance of repair to syntax-for-conversation”. In Syntax and Semantics, Volume 12: Discourse and Syntax, T. Givón (ed.), 261-286. New York: Academic Press. Sacks, Harvey and SchegloŸ, Emanuel A. 1979. “Two preferences in the organization of reference to persons in conversation and their interaction”. In Everyday Language: Studies in Ethnomethodology, G. Psathas (ed.), 15-21. New York: Irvington Publishers. Published simultaneously as: Sacks, Harvey and SchegloŸ, Emanuel A. 1979. “Zwei Präferenzen in der Organisation personaler Referenz in der Konversation und ihre Wechselwirkung”. In Sprachstruktur - Socialstruktur, U. QuasthoŸ (ed.), 150-157. Berlin: Skriptor Verlag. SchegloŸ, Emanuel A. 1980. “Preliminaries to preliminaries: ‘Can I ask you a question’”. Sociological Inquiry 50 (3-4): 104-152. SchegloŸ, Emanuel A. 1980. “What type of interaction is it to be”. In The 18th Annual Meeting of the Association for Computational Linguistics and Parasession on Topics in Interactive Discourse. Proceedings of the Conference, June 19-22, 1980, University of Pennsylvania. Philadelphia, Pennsylvania: 81-82. SchegloŸ, Emanuel A. 1982. “Discourse as an interactional achievement: Some uses of ‘uh huh’ and other things that come between sentences”. In Georgetown University Roundtable on Languages and Linguistics 1981: Analyzing Discourse: Text and Talk, D. Tannen (ed.), 71-93. Washington D.C.: Georgetown University Press. SchegloŸ, Emanuel A. 1984. “On some gestures’ relation to talk”. In Structures of Social Action: Studies in Conversation Analysis [Studies in Emotion and Social Interaction], J.M. Atkinson and J.C. Heritage (eds.), 266-296. Cambridge: Cambridge University Press. JeŸerson, Gail, Sacks, Harvey and SchegloŸ, Emanuel A. 1984. On Laughter in Pursuit of Intimacy. Working Papers and Pre-Publications 135, Series C. Centro Internazionale di Semiotica e di Linguistica, Università di Urbino, Italy. Expanded version published as: JeŸerson, Gail, Sacks, Harvey and SchegloŸ, Emanuel A. 1987. “Notes on laughter in the pursuit of intimacy”. In Talk and Social Organization [Intercommunication Series 1], G. Button and J.R.E. Lee (eds.), 152-205. Clevedon, England: Multilingual Matters, Ltd.. SchegloŸ, Emanuel A. 1986. “The routine as achievement”. Human Studies 9 [Special Issue on Interaction and Language Use], G. Button, P. Drew and J.C. Heritage (guest eds.), 111-151.

A bibliography of Emanuel A. SchegloŸ 177

SchegloŸ, Emanuel A. 1987. “Analyzing single episodes of interaction: An exercise in conversation analysis”. Social Psychology Quarterly 50 (2): 101-114. SchegloŸ, Emanuel A. 1987. “Between macro and micro: Contexts and other connections”. In The Macro-Micro Link, J.C. Alexander, B. Giesen, R. Münch and N.J. Smelser (eds.), 207-234. Berkeley and Los Angeles: University of California Press. Abridged French translation in: Sociétés 14: 17-22, 1987. SchegloŸ, Emanuel A. 1987. “Recycled turn beginnings: A precise repair mechanism in conversation’s turn-taking organisation”. In Talk and Social Organisation [Intercommunication Series 1], G. Button and J.R.E. Lee (eds.), 70-85. Clevedon, England: Multilingual Matters, Ltd.. SchegloŸ, Emanuel A. 1987. “Some sources of misunderstanding in talk-in-interaction”. Linguistics 25: 201-218. SchegloŸ, Emanuel A. 1988. “Description in the social sciences I: Talk-in-interaction”. Papers in Pragmatics 2: 1-24. SchegloŸ, Emanuel A. 1988. “Discourse as an interactional achievement II: An exercise in conversation analysis”. In Linguistics in Context: Connecting Observation and Understanding: Lectures from the 1985 LSA/TESOL and NEH Institutes [Advances in Discourse Processes, Vol. 29], D. Tannen (ed.), 135-158. Norwood, N.J.: Ablex. SchegloŸ, Emanuel A. 1988. “GoŸman and the analysis of conversation”. In Erving GoŸman: Exploring the Interaction Order, P. Drew and A.J. Wootton (eds.), 89-135. Cambridge: Polity Press. SchegloŸ, Emanuel A. 1988. “On an actual virtual servo-mechanism for guessing bad news: A single case conjecture”. Social Problems 35 (4): 442-457. SchegloŸ, Emanuel A. 1988. “Presequences and indirection: Applying speech act theory to ordinary conversation”. Journal of Pragmatics 12: 55-62. SchegloŸ, Emanuel A. 1988/1989. “From interview to confrontation: Observations on the Bush/Rather encounter”. Research on Language and Social Interaction 22: 215-240. SchegloŸ, Emanuel A. 1989. “Harvey Sacks - Lectures 1964-1965: An introduction/memoir”. Human Studies 12: 185-209. Italian translation in: Sacks, Harvey. 1995. Lezioni 1964-1965. La Conversazione: Procedure e Metodi, A. Marcarino (ed.), 43-68. Urbino: Editrice Montefeltro. SchegloŸ, Emanuel A. 1989. “Re¶ections on language, development, and the interactional character of talk-in-interaction”. In Interaction in Human Development [Crosscurrents in Contemporary Psychology], M.H. Bornstein and J.S. Bruner (eds.), 139-153. Hillsdale, N.J.: Lawrence Erlbaum Associates. SchegloŸ, Emanuel A. 1990. “Discussion of Suchman and Jordan, ‘Interactional troubles in face-to-face survey interviews’”. Journal of the American Statistical Association 85: 248250.

178 Discussing Conversation Analysis

Reprinted as: SchegloŸ, Emanuel A. 2002. “Survey interviews as talk-in-interaction”. In Standardization and Tacit Knowledge: Interaction and Practice in the Survey Interview [Wiley Series in Probability and Statistics], D.W. Maynard, H. Houtkoop-Steenstra, N.C. SchaeŸer and J. van der Zouwen (eds.), 151-157. New York: John Wiley. SchegloŸ, Emanuel A. 1990. “On the organization of sequences as a source of ‘coherence’ in talk-in-interaction”. In Conversational Organization and its Development [Advances in Discourse Processes, Vol. 38], B. Dorval (ed.), 51-77. Norwood, N. J.: Ablex. SchegloŸ, Emanuel A. 1990. “(Re)analyzing Applied Linguistics”. Special Feature Roundtable. Deªning Our Field: Unity in Diversity. Issues in Applied Linguistics 1 (2): 166. SchegloŸ, Emanuel A. 1991. “Conversation analysis and socially shared cognition”. In Perspectives on Socially Shared Cognition, L.B. Resnick, J.M. Levine and S.D. Behrend (eds.), 150-171. Washington, D.C.: American Psychological Association. SchegloŸ, Emanuel A. 1991. “Re¶ections on talk and social structure”. In Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis, D. Boden and D.H. Zimmerman (eds.), 44-70. Cambridge: Polity Press. SchegloŸ, Emanuel A. 1992. “In another context”. In Rethinking Context: Language as an Interactive Phenomenon, A. Duranti and C. Goodwin (eds.), 193-227. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1992. “Introduction”. In Harvey Sacks: Lectures on Conversation, Vol. 1, G. JeŸerson (ed.), ix-lxii. Oxford: Basil Blackwell. SchegloŸ, Emanuel A. 1992. “Introduction”. In Harvey Sacks: Lectures on Conversation, Vol. 2, G. JeŸerson (ed.), ix-lii. Oxford: Basil Blackwell. SchegloŸ, Emanuel A. 1992. “On talk and its institutional occasions”. In Talk at Work: Interaction in Institutional Settings [Studies in Interactional Sociolinguistics 8], P. Drew and J.C. Heritage (eds.), 101-134. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1992. “Repair after next turn: The last structurally provided defense of intersubjectivity in conversation”. American Journal of Sociology 97 (5): 1295-1345. SchegloŸ, Emanuel A. 1992. “To Searle on conversation: A note in return”. In (On) Searle on Conversation, H. Parrett and J. Verschueren (eds.), 113-128. Amsterdam and Philadelphia: John Benjamins. SchegloŸ, Emanuel A. 1993. “Re¶ections on quantiªcation in the study of conversation”. Research on Language and Social Interaction 26 (1): 99-128. SchegloŸ, Emanuel A. 1994. “Telephone conversation”. In The Encyclopedia of Language and Linguistics, Vol. 9, R.E. Asher and J.M.Y. Simpson (eds.), 4547-4549. Oxford: Pergamon Press. SchegloŸ, Emanuel A. 1995. “Parties and talking together: Two ways in which numbers are signiªcant for talk-in-interaction”. In Situated Order: Studies in Social Organization and Embodied Activities, P. ten Have and G. Psathas (eds.), 31-42. Washington, D.C: University Press of America. SchegloŸ, Emanuel A. 1996. “Conªrming allusions: Toward an empirical account of action”. American Journal of Sociology 102 (1): 161-216. SchegloŸ, Emanuel A. 1996. “Issues of relevance for discourse analysis: Contingency in

A bibliography of Emanuel A. SchegloŸ 179

action, interaction and co-participant context”. In Computational and Conversational Discourse: Burning Issues - An Interdisciplinary Account [NATO Asi Series. Series F, Vol. 151: Computer and Systems Sciences], E.H. Hovy and D. Scott (eds.), 3-38. Heidelberg: Springer Verlag. Adaptation of part of this paper published as: SchegloŸ, Emanuel A. 1995. “Discourse as an interactional achievement III: The omnirelevance of action”. Research on Language and Social Interaction 28 (3): 185211. Slightly revised version of above reprinted in: SchiŸrin, Deborah, Tannen, Deborah and Hamilton, Heidi E. (eds.). 2001. The Handbook of Discourse Analysis [Blackwell Handbooks in Linguistics], 229-249. Oxford: Basil Blackwell. SchegloŸ, Emanuel A. 1996. “Some practices for referring to persons in talk-in-interaction: A partial sketch of a systematics”. In Studies in Anaphora, B. Fox (ed.), 437-485. Amsterdam: John Benjamins. SchegloŸ, Emanuel A. 1996. “Turn organization: One intersection of grammar and interaction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E.A. SchegloŸ and S.A. Thompson (eds.), 52-133. Cambridge: Cambridge University Press. Reprinted (in part) in: Candlin, Christopher N. and McNamara, Tim F. (eds). 2003, in press. The Applied Linguistics Reader. London: Routledge. SchegloŸ, Emanuel A., Ochs, Elinor and Thompson, Sandra A. 1996. “Introduction”. In Interaction and Grammar [Studies in Interactional Sociolinguistics 13], E. Ochs, E.A. SchegloŸ and S.A. Thompson (eds.), 1-51. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 1997. “‘Narrative analysis’ thirty years later”. Journal of Narrative and Life History 7 (1-4): 97-106. Reprinted in: Paulston, Christina and Tucker, Richard (eds.). 2003, in press. Sociolinguistics: The Essential Readings, 105–113. Oxford, U.K. and Cambridge, U.S.A.: Basil Blackwell. SchegloŸ, Emanuel A. 1997. “Practices and actions: Boundary cases of other-initiated repair”. Discourse Processes 23 (3): 499-545. SchegloŸ, Emanuel A. 1997. “Third turn repair”. In Towards a Social Science of Language: Papers in Honor of William Labov. Volume 2: Social Interaction and Discourse Structures, G.R. Guy, C. Feagin, D. SchiŸrin and J. Baugh (eds.), 31-40. Amsterdam: John Benjamins. SchegloŸ, Emanuel A. 1997. “Whose text? Whose context?” Discourse and Society 8 (2): 165187.

180 Discussing Conversation Analysis

SchegloŸ, Emanuel A. 1998. “Body torque”. Social Research 65 (3): 535-596. SchegloŸ, Emanuel A. 1998. “Re¶ections on studying prosody in talk-in-interaction”. Language and Speech 41 (3/4): 235-263. SchegloŸ, Emanuel A. 1998. “Reply to Wetherell”. Discourse and Society 9 (3): 413-416. SchegloŸ, Emanuel A. 1999. “Discourse, pragmatics, conversation, analysis”. Discourse Studies 1 (4): 405-435. SchegloŸ, Emanuel A. 1999. “Naivete vs. sophistication or discipline vs. self-indulgence: A rejoinder to Billig”. Discourse and Society 10 (4): 577-582. SchegloŸ, Emanuel A. 1999. “On Sacks on Weber on Ancient Judaism: Introductory notes and interpretive resources”. Theory, Culture and Society 16 (1): 1-29. SchegloŸ, Emanuel A. 1999.“‘SchegloŸ’s texts’ as ‘Billig’s data’: A critical reply”. Discourse and Society 10 (4): 558-572. SchegloŸ, Emanuel A. 1999. “What next?: Language and social interaction study at the century’s turn”. Research on Language and Social Interaction 32 (1/2): 141-148. Heeschen, Claus and SchegloŸ, Emanuel A. 1999. “Agrammatism, adaptation theory, conversation analysis: On the role of so-called telegraphic style in talk-in-interaction”. Aphasiology 13 (4/5): 365-405. SchegloŸ, Emanuel A. 2000. “Das Wiederauftauchen des Unterdrückten”. Psychotherapie und Sozialwissenschaft; Zeitschrift für qualitative Forschung 2 (1): 3-29. Also published as: SchegloŸ, Emanuel A. 2003. “The surfacing of the suppressed”. In Studies in Language and Social Interaction: In Honor of Robert Hopper, P.J. Glenn, C.D. LeBaron, and J.S. Mandelbaum (eds.), 241–262. Mahwah, N.J.: Lawrence Erlbaum Associates. SchegloŸ, Emanuel A. 2000. “On granularity”. Annual Review of Sociology 26: 715-720. SchegloŸ, Emanuel A. 2000. “Overlapping talk and the organization of turn-taking for conversation”. Language in Society 29 (1): 1-63. SchegloŸ, Emanuel A. 2000. “When ‘others’ initiate repair”. Applied Linguistics 21 (2): 205243. SchegloŸ, Emanuel A. 2001. “Accounts of conduct in interaction: Interruption, overlap, and turn-taking”. In Handbook of Sociological Theory, J.H. Turner (ed.), 287-321. New York and Dordrecht: Kluwer Academic/Plenum Publishers. SchegloŸ, Emanuel A. 2001. “Getting serious: Joke → serious ‘no’”. Journal of Pragmatics 33 (12): 1947-1955. SchegloŸ, Emanuel A. 2001. “Overwrought utterances: ‘Complex’ sentences in a diŸerent sense”. In Complex Sentences in Grammar and Discourse: Essays in Honor of Sandra A. Thompson, J. Bybee and M. Noonan (eds.), 321-336. Amsterdam: John Benjamins. SchegloŸ, Emanuel A. 2002. “Beginnings in the telephone”. In Perpetual Contact: Mobile Communication, Private Talk, Public Performance, J.E. Katz and M. Aakhus (eds.), 284300. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 2002. “On ‘opening sequencing’: An introductory note”. In Perpetual Contact: Mobile Communication, Private Talk, Public Performance, J.E. Katz and M. Aakhus (eds.), 321-325. Cambridge: Cambridge University Press.

A bibliography of Emanuel A. SchegloŸ

SchegloŸ, Emanuel A. 2002. “Opening sequencing”. In Perpetual Contact: Mobile Communication, Private talk, Public Performance, J.E. Katz and M. Aakhus (eds.), 326-385. Cambridge: Cambridge University Press. SchegloŸ, Emanuel A. 2002. “Re¶ections on research on telephone conversation: Issues of cross-cultural scope and scholarly exchange, interactional import and consequences”. In Telephone Calls: Unity and Diversity in Conversational Structure across Languages and Cultures, K.K. Luke and T.S. Pavlidou (eds.), 249–281. Amsterdam: John Benjamins. SchegloŸ, Emanuel A., Koshik, Irene, Jacoby, Sally and Olsher, David. 2002. “Conversation analysis and applied linguistics”. In Annual Review of Applied Linguistics, 22 [Discourse and Dialogue], M. McGroarty (ed.), 3-31. SchegloŸ, Emanuel A. 2003. “Conversation analysis and ‘communication disorders’”. In Conversation and Brain Damage, C. Goodwin (ed.), 21–55. New York: Oxford University Press. SchegloŸ, Emanuel A. 2003. “On ESP puns”. In Studies in Language and Social Interaction: In Honor of Robert Hopper, P.J. Glenn, C.D. LeBaron, and J.S. Mandelbaum (eds.), 531– 540. Mahwah, N.J.: Lawrence Erlbaum Associates. Heeschen, Claus and SchegloŸ, Emanuel A. 2003. “Aphasic agrammatism as interactional artifact and achievement”. In Conversation and Brain Damage, C. Goodwin (ed.), 231– 281. New York: Oxford University Press. Sacks, Harvey and Schegloff, Emanuel A. 2002 (1975). “Home position”. Gesture 2(2). Includes 16 videoclips. SchegloŸ, Emanuel A. Forthcoming. A Primer in Conversation Analysis: Sequence Organization. Cambridge: Cambridge University Press.

Interviews Wong, Jean and Olsher, David. 2000. “Re¶ections on conversation analysis and nonnative speaker talk: An interview with Emanuel A. SchegloŸ”. Issues in Applied Linguistics 11 (1): 111-128.

181

Subject index

A Action 166 as an atomic particle 27 centrality in CA 169–170 course of 39, 59 multi-party construction of 58 physical (when part of interpretive sequence) extension of turn-taking and repair to cover 135 syntax of 2 vs. structure 79 Activity instructional 162 main or mainline 162 Adjacency pair 5, 94–95 as a norm of conduct 95, 100 rules 95, 100ff Affordances 79 semiotic-material x–xi Agency 75ff, 85 n. 8, n. 9 Agrammatism (agrammatic aphasia) 142, 149 adaptation theory of 149 and telegraphic speech 149 Allusion 6 confirming 6 Analysis analogical 161 qualitative vs. quantitative, see Qualitative vs./& quantitative analysis sequential, see Sequential analysis Analytical libertinism 26–27 Anthropologies of our own 13 Aphasia 62–63, 142ff agrammatic see Agrammatism

assessment and therapy and CA 149ff See also Conversation and aphasia; Conversation analysis profile Aphasiology CA-assisted 149 quantification in, see Quantification in aphasiology B Behavior chunks (bits, complexes, strips) of 25ff, 65 interactional reservoir of 98 ordinary, everyday 45 See Neurobiology of behavior Body human 61, 62, 79 Body-brain and conversational practices xi C Center for the study of suicide/Suicide prevention center 22, 23, 24 Closings of telephone conversations 3 See also Exchange; termination Cognitive science 37 39 Cold shouldering 103–105 Collaborative utterances 59 See Sequence, collaborative Computer-aided instruction interactive features of See CA and encounters with tutorial programs Consciousness discursive 26, 40

184 Discussing Conversation Analysis

Constants conversational 73 interactive 69, 72, 73, 74 of talk ix, 68 theoretical 69, 70, 72 Context 5, 6 downplaying 6 external 68, 80, 82 indifference to 5 institutional 78 internal 68 intra-interactional 80, 82 macro-structural 80 Conversation 7 abnormal (with aphasic and demented speakers) 141ff actual ix, 96 and aphasia viii, x, xi, 45 49, 62, 63, 142ff between therapists and aphasic people 147–148 contrapuntal in Antigua 167 multi-track or n-track model of 166 normal 63 ordinary 1, 158, 159, 160 as CA original and primary domain 1 telephone 3 usual CA sense of 160 See also Italian conversation Conversation analysis (CA) vii, viii, ix, x, xi, xii, 1, 2, 3, 4, 5, 6, 7, 11, 12, 13, 14, 17, 22, 24, 34, 37, 59ff, 65, 105, 109ff, 141ff, 152–153, 157ff, 165–166, 168–170 absence of quantification in 12 and applied linguistics xiii n. 5 and cognitive science 37 and Communications/Speech Communications 14 and encounters with tutorial programs 109ff, 160ff and interactional linguistics 168ff and language pathology 141ff and linguistics 168–170 and neurolinguistics 141ff

and neurosciences 15 and phenomenology 91ff and pragmatic disorders, see Pragmatic disorders as a discourse for contextualizing ‘actual conversation’ ix as a natural history of human interaction 7 as rigorous science 91ff as a study of interactive consequences of aphasia 149–150 categories of resources and practices according to capturable by functional magnetic resonance imaging? 159 natural and discovered vs. analytical and imposed 159 critics of vii, viii x, xiii n. 3 debate on vii, xiii n. 1 formalism in 29ff Goffman’s attack on 34 in aphasia assessment and therapy 149ff methodology 69ff sociological roots of 169 troubles with 12–13 See also Aphasiology, CA-assisted; SCA Conversation analysis profile for aphasic people 150, 151 for people with cognitive impairment 150 Conversation analysts as a community of inquiry 17, 157 Co-participant 58 D Dementia 142, 145 Alzheimer-type and failure in topic maintaining and elaborating 145 unexpected topic shifts in 145 use of cohesion devices in 146 Deviant case analysis 94, 103

Subject index 185

E English 13, 51 n. 4 Ethnomethodology 3, 22, 35, 36, 109, 169 Exchange Pre-closing 3 terminal exchange 3 Family therapy and analysis of human interaction 57– 58 F Finnish 13 French 51 n. 4 G Gaze 61 German 13, 51 n. 4 Gesture conversational 63 H Hebrew 13, 51 n. 4 I Identities 42–44 discourse vs. other identities 42 institutional 77 omni-relevance of some 43 Implicit no analytical status of in SCA 84 n. 6 Individual focus on the 38–39 vs. situation as the locus of social reality 38ff, 66 Inference conversational xiii n. 3 no answer-no person 98 Interaction computer-user, see human-computer human-computer x, 110ff, 160–161 See also CA and encounters with tutorial programs

human-human 2, 3, 4, 7, 72, 73, 85 n. 9 according to Goffman 3 and genotype/phenotype 84 n. 5 and its contexts of relevance see Relevance aphasic-nonaphasic 63, 147 casual 68 ideal-types of 80 institutional 67 moral obligations of 3 multi-modal and transformative 84 multi-party viii multi-track or n-track model of 166 non-ordinary 68 normative organization of 3 not a kind of random Brownian motion 7 ordinary 46, 65, 67, 68, 161 ordinary vs. non-ordinary 75ff resources for 72 termination of joint, complete 4 turn-centered idea of 166 simian (among monkeys) 85 n. 9 See also Constants, interactive; Rules, interactive; Interactionists and interaction Interaction order 3, 7, 42–45 See Separation of orders Interactionists and interaction 79 Interactional membrane 42–43, 66 as a filter 43 Interactional reservoir of behavior See Behavior Involvement conversational xiii n. 3 Italian conversation 166–167 J Japanese 13

186 Discussing Conversation Analysis

K Korean 13 L Latin 51 n. 4 Linguistics 60, 168–170 integrational 106 n. 1 interactional 168ff systemic-functional viii Logic interactive 69 M Mandarin 13 Materialities exosomatic 79 semiotic 110, 131 somatic 79 Meaning-making social a materialist view of viii predictable? x Meaning potential 128 Misunderstanding 5 and users of tutorial programs 109ff Motives of interactants 72 Move 165–166 Multimodality 84 n. 1 Music making music together 15–16 N Neurobiology of behavior 15, 141, 159 O Omni-relevance of participants’macrostructural features 43 See Identities, omni-relevance of some Openings conversational 60f as genuine sequences 61 Overlap

management 5 Overlapping massive in Italian conversation 166 P Paraphasia 142 semantic 143, 145 Participant kinds of 58, 62 structure 58 See also Co-participant; Understandings, participants’ Participation framework 135 and users of tutorial programs 109ff in encounters with tutorial programs x Phenomenology and SCA 91ff hermeneutic 66 Philadelphia Child Clinic 57 Post-expansions 5 Practices institutional 76 77 Pragmatic disorders suggestions for doing CA research on 45–49 Procedures conversational 37 discovery procedures of 37 or methods, rules employed by the participants vs. CA constructs viii, x, 103 See Rules Professionalization 77, 79 Prosody 5 its critical role in global aphasia 144 Q Qualitative vs./& quantitative analysis 145ff and conversations between aphasic and unimpaired interlocutors 143

Subject index 187

See Quantification Quantification and the study of interaction 12 in aphasiology 146 See also Conversation analysis, absence of quantification in; Qualitative vs./ & quantitative analysis R Reactance discursive 75, 76, 84 n. 7 Reading research and CA 134 Reasoning methods of and social action 3 Relevance analytical 71 conditional 5, 61, 62, 98 emic or speaker-centered 70 interactional, as relevance for interactants 66–67, 92, 94, 96ff multiplicity of contexts of and human interaction 81 test of 43 See also Omni-relevance; Identities, omni-relevance of some Repair 4, 14, 23, 106 n. 4, 133, 135 across languages and cultures 13 and physical action, see Action, physical collaborative 143, 148 see also Repair sequences; Sequence, collaborative initiation 161–162 initiator 114ff, 148 in aphasic-nonaphasic conversation 147ff in various positions 14 organization 5, 161 other-initiated 12–14, 113ff self-repair 5, 147 sequences in aphasic-nonaphasic conversation 144, 147

work 112ff Repairable 113 Repeat 6 and allusion confirming 6 Requirements ritual 72 system 72 Reservoir See Behavior, interactional reservoir of Resources viii, 72, 73, 78 cross-coupling of both semiotic and material x exosomatic 78 79 for interaction, see Interaction, resources for institutional 78 interactive 74, 78 modalities of structuredness of 78 semiotic-material 83 somatic (bodily) viii, 72 unexplicated (Sacks) 82 Role distance 75, 76 Rules as causal explanations of participants’ behavior x explicitly rejected by Schegloff xiii n. 4, 162 interactive 72 See Procedures; Summons repetition, terminating rule for S SCA or Conversation Analysis as defined by Schegloff 69ff as technical analysis 80, 82–83 vs. interpretive description 83 Separation of orders 43 Sequence collaborative, in aphasic-nonaphasic conversation 143–144, 147 phases of 144 insertion 5 moral requiredness of 4

188 Discussing Conversation Analysis

organization 4, 11 Sequential analysis xiii n. 3 Sequential implicativeness 71 Sequential organization 61 Situation 38, 66 Sociology of knowledge 18 Speech act theory 39, 40, 51 n. 5 Stances 75–76 Story telling in conversation 30, 61–62 Structure, see Action vs. structure Summons 97–98 repetition 97–98 terminating rule for 97–98, 162 Swedish 13 T Talk and gesture 63 constants of 68 control over 74 ordinary, casual as generic standard 68 ordinary vs. non-ordinary 73ff sequential implicativeness of 71 Talk-in-interaction 61 Telephone conversation See Conversation Terminating rule See Summons Termination, see Interaction Thai 13 Transcription in social analysis 85 n. 10 Turn 165–166 and functional magnetic resonance imaging 159 as track unit or constituent 166 at talk 61 entry 5 exit 5

organization 4 ratio of minimal to non-minimal turns in aphasiology 146–147 Turn-constructional unit 4, 165 Turn-following mode 166 Turn-taking 4, 5, 111, 133, 161, 167 and levels of grammar 167–168 and physical action, see Action, physical and track-tacking 166 as a synchronous process 161 systematics for 29ff systems non conversational 5 Tutorial programs their and their users’ turns and repairs See CA and encounters with tutorial programs understandings and misunderstandings of their users 109ff U Understandings participants’ 99 Uptake sequential 72 V Variability in aphasic-nonaphasic conversation between dyads 147–148 within dyads across time 148–149 W Where we are/where they are 3, 4 Why that now? 69, 73, 85 n. 11 as the basic issue for parties to interaction 166 Word selection 4 Y Yiddish 51 n. 4

Author index

A Aarseth, Espen 115 Ahad, Pierre 141 Alexander, Jeffrey C. 78 Algar, Louise 151 Anderson, Bob 95, 106 n. 3 Archer, Margaret S. 78 Atkinson, J. Maxwell 5 Austin, John L. 3, 40 B Bales, Robert Freed 18 Bateson, Gregory 57 Bayley, Kate 152 Belin, Pascal 141 Bendix, Reinhardt 19 Berger, Peter 84 n. 6 Billig, Michael xiii, n. 1, 83 Boles, Larry 146 Booth, Susan 151 Bourdieu, Pierre 2, 78 Boyd, Julian 129 Brown, Roger 32 Bruce, Carolyn 152 Bruce, Steve 84 n. 4 Bryan, Karen 152 Bryant, Christopher G.A. 78 Buber, Martin 40 Butler, Judith 83 Button, Graham 5, 110 C Cameron, Deborah 71, 83 Carpenter, Brian D. 146 Carter, Michelle S. 145 Caudwell, Christopher 51 n. 4

Chenerey, Helen J. 145 Chomsky, Noam 7 Cicourel, Aaron V. 5, 78 Clark, Herbert H. 144 Clayman, Steve 47 Conway, N. 143 Coulter, Jeff 84 n. 3 Couper-Kuhlen, Elizabeth 5, 144 Cowley, Stephen viii Crisp, Jenny 148 Crockford, Catherine 146, 149 C¦ mejrková, Sv¦etla vii, 17, 25, 29, 39, 40, 50 D Damasio, Antonio R. 15 Deacon, Terrence W. 84 n. 5 Dennett, Daniel Clement 84 n. 5 Di Luzio, Aldo xiii, n. 3 Drew, Paul 70, 73, 77, 78 DuBois, John 84, n. 4 Duchan, Judith 143 Duranti, Alessandro 5, 81, 84 n. 2 Durkheim, Emile 2, 3 E Edelman, Gerald H. 84 n. 5 Edmundson, Anne 152 Eerdmans, Susan L. vii, xii, 173 Egbert, Maria 13 Eggins, Suzanne 68 Ellis, Andrew W. 142 Engeström, Yrjo 110 F Fairclough, Norman 81 Fehr, B. J. 84 n. 2

190 Discussing Conversation Analysis

Ford, Cecilia 84 n. 2 Foucault, Michel 85 n. 8 Frohlich, David 112 G Garcia, Linda J. 145 Garfinkel, Harold 2, 3, 4, 18, 19, 20, 22, 32, 34-36, 37, 44, 65, 66, 83, 111 Giddens, Anthony 26, 40, 78 Goffman, Erving 2, 3, 4, 18, 20, 21, 32-34, 36, 37, 38, 42, 43, 44, 58, 59, 65, 66, 72, 75, 76, 165, 166 Goodwin, Charles vii, viii, 5, 15, 57, 61, 62, 63, 81, 84 n. 2, 109, 136 n. 7, 143, 144, 157 Goodwin, Majorie Harness 58, 63, 109, 136 n. 7 Graham, Phil 83 Greatbatch, David 5, 50 n. 3 Gumperz, John J. xiii n. 3 H Habermas, Jürgen 2, 40, 51 n. 5 Halliday, Michael A. K. 116, 118 Harris, Roy 106 n. 1, 132 Hayashi, Makoto 13 Heeschen, Claus 147, 149 Heritage, John vii, 5, 12, 47, 50 n. 3, 61, 66, 70, 71, 73, 94, 122, 157 Husserl, Edmund 91, 92, 93, 95, 102, 162 Hutchins, Edwin 118 I Iedema, Rick viii, ix, 65, 77, 78, 79, 80, 84 n. 1, 85 n. 11, 159-160 J Jary, David 78 Jefferson, Gail 1, 4, 5, 12, 14, 26, 30, 57, 59, 67, 69, 106 n. 4, 143 Joanette, Yves 145 Jones, John 38

K Kaal, Bertie xii Kagan, Aura 150 Kelly, John 5 Kim, Kyu-hyun 13 Klippi, Anu 144 Knorr-Cetina, Karin D. 78 Kress, Gunther 77, 78, 84 n. 1 L Laakso, Minna 144 Labov, William 58 Lafaille, Philippe 141 Latour, Bruno 68, 73, 79, 81, 85 n. 9 Lebaron, Curtiss D. 79, 81 Lee, John R. L. 26 Lefebvre, Henri 79 Lerner, Gene H. 5, 13 Lesser, Ruth x, xi, 141, 143, 149, 150, 151, 158-159, 160 Levinas, Emmanuel 40 Levinson, Stephen C. 73 Lindstrõm, Anna-Karin Benedicta 13 Local, John 5 Lowenthal, Leo 19 Lubinski, Rosemary 143 Luckmann, Thomas 84 n. 6 Lum, Carmen 142 Lindsay, Jayne 147 Lock, Sarah 152 M Manicas, Peter T. 78 Martin, James R. 73, 131 Marx, Karl 2 Maxim, Jane 152 McInnes, David 79 McKeith, Ian 150 McKenna, Brian 83 Mead, George Herbert 2 Mehan, Hugh 74, 79 Mey, Inger xiii n. 1 Middleton, David 110 Milroy, Lesley 143, 144

Author index

Moerman, Michael 13, 31 Moir, Diana 152 Moore, Barrington Jr. 18, 21 Mori, Junko 13 Mummery, Catherine J. 141 N Natter, Wolfgang 77 O Ochs, Elinor 18, 168 Oehman, Sven 106 n * Oelschlaeger, Mary L. 146 Olsher, David xiii n. 2 P Park, Yong-Yae 13 Parsons, Talcott 2, 3, 7, 18, 20, 21, 36 Patterson, K. 141 Perkins, Lisa 143, 144, 146, 147, 148, 150, 151 Pike, Bruce 141 Pomerantz, Anita 84 n. 2 Prevignano, Carlo L. vii, xii, xiii n. 3, 11, 17, 23, 29, 32, 37, 42, 45, 49, 50, 165, 166 R Radcliffe-Brown, Alfred R. 3 Raudaskoski, Pirkko x, xi, 109, 110, 160162 Reisman, Karl 167 Ripich, Danielle N. 146 Roth, Andrew L. 50 n. 3 S Sacks, Harvey viii, 1, 2, 3, 4, 5, 14, 18, 19, 20, 21, 22, 23, 24, 25, 29, 30, 32, 36, 37, 41, 43, 57, 58, 59, 60, 62, 67, 69, 71, 82, 106 n. 4, 143, 162, 165, 166, 167 Sapir, Edward 61 Saussure, Ferdinand de 61 Schaefer, Edward F. 144 Schatzki, Theodore 77 Schegloff, Emanuel A vii, viii, ix, x, xi, xii,

xiii n. 1, n. 2, n. 4, n. 5, 1, 2, 3, 4, 5, 6, 7, 8, 11, 12, 13, 14, 15, 16, 18, 19, 20, 22, 26, 28, 29, 31, 32, 40, 42, 43, 45, 49, 50 n *, n. 1, 51 n. 5, 57, 58, 60, 62, 63, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 78, 80, 81, 82, 83, 84 n. 3, n. 6, 85 n. 9, n. 11, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 105, 106 n. 2, n. 4, 107 n. 10, 109, 110, 112, 113, 114, 119, 133, 134, 135, 135 n. 2, 136 n. 2, n. 6, n. 7, n. 9, 141, 142, 143, 145, 146, 147, 148, 149, 152, 157, 159, 161, 165, 166, 167, 168, 170 n. *, 174 Searle, John R. xiii n. 1, 7, 40, 58 Segerdahl, Pär viii, ix, x, xiii n. 4, 73, 91, 106 n. 5, 106 n. 7, 162-163 Selting, Margret 5, 144 Selznick, Philip 21 Sharrock, Wes W. 84 n. 4, 95, 106 n. 3 Sidnell, Jack 167 Simmons-Mackie, Nina 150 Slade, Diana 68 Smith, D. E. 109, 110 Smith, F. Joseph 106 n. 6 Sorjonen, Marja-Leena 13, 77, 78 Stenlund, Sören 106 n * Strate, Lance 132 Stravinsky, Igor 106 n. 6 Streeck, Jürgen 79, 81, 134 Suchman, Lucy A. 110, 112, 136 n. 5 Sudnow, David 21, 22, 36 Swabey, Donna 151 T Takagi, Tomoyo 13 Tanaka, Hiroko 13 Taylor, Charles 82 Taylor, Talbot J. 71, 106 n. 9 Thibault, Paul J. vii, viii, xii, 50 n. *, 85 n. 10, 116, 131, 133, 165, 167 Thompson, Sandra 18, 84 n. 2, 168 V Van Dijk, Teun xiii n. 1

191

192 Discussing Conversation Analysis

Van Leeuwen, Theo 84 n. 1 Vinkhuyzen, Erik 132 W Wallis, Roy 84 n. 4 Walshaw, David 148 Watson, Caroline M. 145, 146 Watson, D. Rodney, 84 n. 4, 94 Watt, M. 150 Weitzner-Lin, Barbara 143 Wetherell, Margaret xiii n. 1 Whalen, Jack 109, 110, 132, 138 Whitworth, Anne 150

Whorf, Benjamin Lee 75, 84 n. 7 Wilkinson, Ray 144, 147, 148, 152 Willmott, Hugh 77 Wise, R. J. S. 141 Wittgenstein, Ludwig 3, 92. 97, 104, 105, 106 n. 6, 107 n. 11, n. 13 Wong, Jean xiii n. 2 Woolf, Virginia 13 Wu, Ruey Jiuan Regina 13 Z Zatorre, Robert J. 141 Ziol, Elaine 146