Вы находитесь на странице: 1из 554

Studies in Language

and Social Interaction

In Honor of Robert Hopper


Jennings Bryant/Dolf Zillmann, General Editors
Selected titles in Language and Discourse (Donald Ellis, Advisory Editor) include:
Ellis From Language to Communication, Second Edition
Haslett/Samter Children Communicating: The First Five Years
Locke Constructing The Beginning: Discourses of Creation Science
Ramanathan Alzheimer Discourse: Some Sociolinguistic Dimensions
Sigman Consequentiality of Communication
Tracy Understanding Face-to-Face Interactions
For a complete list of titles in LEAs Communication Series, please contact Lawrence Erlbaum Associates, Publishers

Studies in Language and Social

In Honor of Robert Hopper
Edited by

Phillip J.Glenn
Emerson College
Curtis D.LeBaron
Brigham Young University
Jenny Mandelbaum
Rutgers University


Mahwah, New Jersey London

Copyright 2003 by Lawrence Erlbaum Associates, Inc.

All rights reserved. No part of this book may be reproduced in any
form, by photostat, microform, retrieval system, or by any other means,
without the prior written permission of the publisher.
Lawrence Erlbaum Associates, Inc., Publishers
10 Industrial Avenue
Mahwah, New Jersey 07430
This edition published in the Taylor & Francis e-Library, 2009.
To purchase your own copy of this or any of
Taylor & Francis or Routledges collection of thousands of eBooks
please go to www.ebookstore.tandf.co.uk.
Cover design by Kathryn Houghtaling Lacey
Library of Congress Cataloging-in-Publication Data
Studies in language and social interaction/edited by Phillip J.Glenn, Curtis D.LeBaron,
Jenny S.Mandelbaum.
p. cm.
Festschrift for Robert Hopper.
Includes bibliographical references and index.
ISBN 0-8058-3732-9 (alk. paper)
1. Sociolinguistics. 2. Interpersonal communication. 3. Social interaction.
4. Conversation. I. Glenn, Phillip J. II. LeBaron, Curtis D. III. Mandelbaum, Jenny S.
IV. Hopper, Robert.
P40.E93 2001
ISBN1-4106-0696-1Master e-book ISBN

ISBN0-8058-3732-9(Print Edition)

To Robert Hopper (19451998)
Scholar, Teacher, Colleague, Friend

Descriptions are the gifts observers give:

Refraining patterns message bearers live.1

From poem by Robert Hopper, Observer: Steps to an Ecology of Mind, Communication Theory,
1991, 1, 267268.


1. An Overview of Language and Social Interaction Research

Curtis D.LeBaron, Jenny Mandelbaum, and Phillip J.Glenn




2.Extending the Domain of Speech Evaluation: Message Judgments

James J.Bradac

3.Designing Questions and Setting Agendas in the News Interview

John C.Heritage

4.Taken-for-Granteds in (an) Intercultural Communication

Kristine L.Fitch

5. So, What Do You Guys Think?: Think Talk and Process in Student-Led
Classroom Discussions
Robert T.Craig and Alena L.Sanusi

6.Gesture and the Transparency of Understanding

Curtis D.LeBaron and Timothy Koschmann




7. Utterance Restarts in Telephone Conversation: Marking Topic Initiation and

Charlotte M.Jones

8.Recognizing Assessable Names

Charles Goodwin

9.Interactional Problems With Did You Questions and Responses

Susan D.Corbin


10.Managing Optimism
Wayne A.Beach

11.Rejecting Illegitimate Understandings

Samuel G.Lawrence

12.Interactive Methods for Constructing Relationships

Jenny Mandelbaum

13.A Note on Resolving Ambiguity

Gail Jefferson


viii Contents
14. The Surfacing of the Suppressed
Emanuel A.Schegloff


15. Sex, Laughter, and Audiotape: On Invoking Features of Context to Explain

Laughter in Interaction
Phillip J.Glenn

16. Gender Differences in Telephone Conversations

Hanneke Houtkoop-Steenstra





17. Comparative Analysis of Talk-in-Interaction in Different Institutional

Settings: A Sketch
Paul Drew

18. Conversational Socializing on Marine VHP Radio: Adapting Laughter and

Other Practices to the Technology in Use
Robert E.Sanders

19. Law Enforcement and Community Policing: An Intergroup Communication

Jennifer L.Molloy and Howard Giles

20.Preventatives in Social Interaction


21. The Interactional Construction of Self-Revelation: Creating an Aha


22. A World in a Grain of Sand: Therapeutic Discourse as Making Much of

Little Things
Kurt A.ruder

23.Modeling as a Teaching Strategy in Clinical Training: When Does It Work?

Anita Pomerantz






24. Indeterminacy and Uncertainty in the Delivery of Diagnostic News in Internal

Medicine: A Single Case Analysis

Douglas W.Maynard and Richard M.Frankel

25. Body Movement in the Transition From Opening to Task in Doctor-Patient
Daniel P.Modaff


Contents ix


26.The Body Taken for Granted: Lingering Dualism in Research on Social

Jrgen Streeck


27.Action and the Appearance of Action in the Conduct of Very Young Children
Gene H.Lerner and Don H.Zimmerman
28.Speech Melody and Rhetorical Style: Paul Harvey as Exemplar
John Vincent Modaff

29.The Body Present: Reporting Everyday Life Performance

Nathan P.Stucky and Suzanne M.Daughton

30. Ethnography as Spiritual Practice: A Change in the Takenf or-Granted (or an

Epistemological Break with Science)
Mara Cristina Gonzalez

31.The Tao and Narrative

Mary Helen Brown

32.Conversational Enslavement in The Truman Show

Kent G.Drummond

33.On ESP Puns

Emanuel A.Schegloff





34. Robert Hopper: An Intellectual History

Jenny Mandelbaum


35.The Scientist as Humanist: Moral Values in the Opus of Robert Hopper

Sandra L.Ragan

36.The Great Poem

Leslie H.Jarmon

37.Phone Openings, Gendered Talk, and Conversations About Illness

Wayne A.Beach

38.Nothing Promised
James J.Bradac


x Contents
39.The Last Word
Robert Hopper








An Overview of Language and Social Interaction
Curtis D.LeBaron
Brigham Young University
Jenny Mandelbaum
Rutgers University
Phillip J.Glenn
Emerson College
This book is an edited collection of empirical studies and theoretical essays about human
communication in everyday life. The primary focus is on small or subtle forms of communication that are easily overlooked and too often dismissed as unimportant. Authors examine various features of human interaction (e.g., laughter, vocal repetition, hand gestures)
occurring naturally within a variety of settings (e.g., at a dinner table, a doctors office, an
automotive repair shop), whereby interlocutors accomplish aspects of their interpersonal
or institutional lives (e.g., resolve a disagreement, report bad medical news, negotiate a
raise), all of which may relate to larger social issues (e.g., police brutality, human spirituality, death and optimism). The present collection is bound together by a recognition that
social life is largely a communicative accomplishment, that people constitute the social
realities experienced everyday through small and subtle ways of communicating, carefully
orchestrated but commonly taken for granted.
This volume represents Language and Social Interaction (LSI) perspectives on human
communication. LSI is a popular umbrella term for scholarly work carried out within and
across a number of academic disciplines. The label covers an array of assumptions, methods, and topics, which draw unity from certain family resemblances (discussed later). LSI
research includes studies of speech, language, and gesture in human communication; studies
of discourse processes, face-to-face interaction, communication competence, and cognitive
processing; conversation analytic, ethnographic, microethnographic, ethnomethodological, and sociolinguistic work; dialect and attitude studies, speech act theory, and pragmatics. Within the field of communication, scholarship in LSI has flourished in recent years.
There are large and active LSI divisions within the National Communication Association
(NCA) and the International Communication Association (ICA); the journal Research on
Language and Social Interaction (originally called Papers in Linguistics) is now a mainstay within the field; LSI research appears regularly in books (e.g., Ellis, 1999a) and a host
of mainstream disciplinary journals (e.g., Leeds-Hurwitz, 1992); and a growing number of
communication departments at major universities emphasize LSI in their curricula.
The present volume originated as a Festschrift celebrating the intellectual career of the
late Robert Hopper, a leading LSI researcher and an extraordinary teacher. Hopper completed his doctoral studies in 1970 at the University of Wisconsin and joined the faculty in
Speech Communication at the University of Texas at Austin, where he remained until the
end of his career. As author of eight books and dozens of published essays, he was known

2 Studies in language and social interaction

for his innovative thinking, lucid writing, and ability to bring together diverse scholars
and perspectives. He taught more than 60 graduate courses and supervised more than 30
doctoral dissertations1. He received many awards2 for his research and teaching. Over the
course of three decades, Hopper (and his students) pursued a rigorous speech science that
led him to the forefront of approaches to LSI, as they were new to communication. He
worked first with techniques for measuring language attitudes, then with discourse analysis, then conversation analysis, and finally explored microethnographic techniques for
analyzing videotaped data. Each of these research traditions helped to shape the field of
LSI, and each continues to make robust contributions to a rigorous science of speech in
the communication field. By soliciting papers from Hoppers former students and close
colleagues, therefore, we have collected a cross-section of cutting-edge LSI research. This
volume, then, arises out of two interrelated rationales. One, it is designed to showcase the
diversity of contemporary LSI research, altogether allowing for reflection on LSI as an
established and expanding area of study. Two, it celebrates Robert Hopper and the trajectory of his intellectual career, which in many ways paralleled developments in the field of
LSI, for which he provided impetus. To the extent that this volume forwards his ideas and
interests, it will make important contributions to the study of human communication and
social interaction.
The remainder of this chapter explicates these two interrelated themes. First, we describe
the emergence and influence of LSI within the field of communication3. The work of Robert Hopper embodies both the diversity of LSI research and the eclecticism of the communication field. Second, we describe the current state of LSI and discuss seven points
of commonality and contention within the areathat is, seven points around which LSI
scholars tend to rally in one way or another. Third, we preview the main sections of this
book and comment on its organization.
LSI is a relatively recent area within the field of communication, which has been dominated by rhetorical and psychological approaches for almost a century. The field of communication traces its beginnings to 1914, when a group of speech scholars met in Chicago
A chronological list of Robert Hoppers doctoral students appears in the Appendix to Chapter 34.
For example, in 1983 Hopper became the Charles Sapp Centennial Professor of Communication
at the University of Texas. In 1990 he was honored as one of three Outstanding Graduate Teachers
at the University of Texas. In 1994 he received ICAs B. Aubrey Fisher Mentoring Award. In 1996
he received the Outstanding Scholarly Publication Award (from the LSI Division of NCA). In 1998
he was first to be honored by NCAs newly established Mentor Fund. Over the years, Hopper made
an impressive collection of audio and video recordings of everyday interaction, known as the University of Texas Conversation Library, which in 1998 was officially named in his memory.
By focusing specifically upon the field of-communication, we risk de-emphasizing LSI colleagues
in other disciplines. As the terms language and social interaction suggest, LSI represents a
convergence of concerns originating in linguistics, sociology, and anthropology. Nevertheless, LSI
is especially strong within the field of communication, which is located at the crossroads of these
interdisciplinary movements.

An overview of language and Social interaction research 3

to officially break from their English (and theater) departments at various U.S. universities by organizing the National Association of Academic Teachers of Public Speaking (see
ONeill, 1915). Early publications show a division within the field: Many speech scholars advocated standards of positivistic science, with a psychological rather than a sociological bent (e.g., Winans, 1915; Woolbert, 1916, 1917); many others had a humanistic
and rhetorical emphasis, mostly grounded in neo-Aristotelian philosophy (e.g., Hudson,
1923, 1924; Hunt, 1920). Within a few decades, a respectable research literature had been
established (see Simon, 1951), but it was mostly concerned with individual performers
of speech during situations of public address. After 1950, as the field matured, its domain
extended to include a broad array of communicative phenomena within a wide variety of
human activities. Several scholars have documented the unfolding history and nature of the
communication field (see Arnold & Bowers, 1984; Benson, 1985; Bitzer & Black, 1971;
Gouran, 1990; Kibler & Barker, 1969).
In the late 1970s, a series of groundbreaking publications set the stage for LSIs emergence within the field of communication (at that time called speech communication).
Bringing together interpersonal communication and the detailed study of natural language,
Nofsinger (1975, 1976) and Hawes (1976) demonstrated and advocated scientific analyses
of naturally occurring speech without the use of statistical methodsan innovative proposition for the field of communication at that time. For instance, by drawing on conversation analytic work on presequences, Nofsinger (1975) identified a commonplace speech
device he called the demand ticket (e.g., Yuh know what?), whereby a person may
initiate a topic and at the same time secure the conversational floor. Nofsinger went on
to suggest that utterances be understood according to their location within conditionally
relevant sequences of talk, rather than in terms of gross numbers of occurrences per unit
of time or whatever (p. 9). Philipsen (1975) drew on ethnographic methods pioneered by
linguistic anthropologists Dell Hymes and Ethel Albert in his ground-breaking study of
gendered patterns of speech in a blue-collar urban neighborhood (this essay won the NCA/
LSI divisions Outstanding Publication award in 1998). Two years later, in a special issue
of Communication Quarterly (Summer 1977), naturalistic approaches (Pearce, 1977) to
communication research were more thoroughly described, including ethnomethodology
(Litton-Hawes, 1977), conversation analysis (Nofsinger, 1977), discourse analysis (Jurick,
1977), hermeneutic phenomenology (Hawes, 1977), and ethnography (Philipsen, 1977).
Naturalistic methods were soon featured in other mainstream communication journals (e.g.,
Beach, 1982). Jackson and Jacobs (1980) combined detailed study of natural language with
interests in rhetoric: They analyzed the structure of naturally occurring arguments and compared these to theoretical models of argument and the problem of enthymemes (missing
or taken-for-granted premises of arguments), thereby illustrating the utility of discourse
analysis to the field of communication generally and to rhetorical theory specifically. In an
awardwinning essay, Hopper (1981b) expanded upon the issue of the taken for granted
(TFG) in everyday communication and social life. He brought together a wide variety of
linguistic approaches, showing how concern with TFGs is a communication issue. After
reviewing the difficulties that TFGs have caused scholars across a variety of disciplines
(enthymemes for rhetoricians, presuppositions for linguists, etc.), Hopper suggested that
there may exist a functional and principled incompleteness in language use (p. 205) and
he provided a schematic model for how people handle TFGs in everyday situations. In sum,

4 Studies in language and social interaction

these early publications pushed naturalistic methods into the mainstream of communication research, providing new ways of conceptualizing and analyzing communication, and
bringing attention to phenomena previously overlooked.
In the early 1980s, Robert Hopper and several other communication scholars interested in everyday language use participated in a series of conferences whereby the new
research area (LSI) took shape. The first communication conference focusing on conversational interaction and discourse processes occurred in 1981 at the University of
Nebraska (cohosted by Wayne Beach, Sally Jackson, and Scott Jacobs). The following
year, two conferences occurred: one on language and discourse processes at Michigan
State University (hosted by Don Ellis and William Donohue); the other on discourse analysis and conversational coherence at Temple University (cohosted by Karen Tracy and
Robert Craig). Participants in the Michigan State conference produced a published volume about contemporary issues in language and discourse processes (Ellis & Donohue,
1986), which represented the wide range of LSI approaches (including speech act theory,
discourse analysis, and conversation analysis) that were emerging at that time within the
field of communication. For example, Hopper, Koch, and Mandelbaum (1986) described
methods of conversation analysis, as the authors were coming to understand them. Participants in the Temple conference produced a published volume of original research (Craig
& Tracy, 1983) that evidenced a scholarly movement [with] radically different methods,
databases, and conceptual frameworks for studying human interaction (Knapp, 1983, p.
7). Each of the authors, including Hopper, examined the same data set: a careful transcription of a lengthy conversation between B and K, two female undergraduate students
who talked casually about their families, friends, food, holiday plans, horses, weather, and
whatever else happened to emerge in the course of their interaction. Authors employed
qualitative and quantitative methods to analyze the structures and strategies of B and Ks
talk, providing detailed descriptions and accounts of the orderly and meaningful ways that
competent speakers may show their talk to be coherently connected. For example, Hopper
(1983) showed that coherence is an interactive accomplishment (we can no longer rely
upon a model of communication that emphasizes the role of the speaker over that of the
listener p. 84), across turns at talk (the fundamental unit of interpretation is the pair p.
80), whereby shared meanings systematically emerge and evolve (the ordering of events
in sequential time frequently seems an important tie to the interpretive process p. 92).
During the final decades of the 20th century, LSI scholars in communication brought
together approaches and concerns from a number of related movements. Hoppers research
exemplifies the eclectic interests which contributed to the emergence of LSI as a distinct
area of study. Resonating with the fields origins in rhetorical theory, LSI research on
speech evaluation sought to gauge audience responses to speakers and their messages (e.g.,
Gundersen & Hopper, 1984). Early message research employed sociolinguistic methods
to examine the effects of speech on the listener by focusing on how listeners evaluated
speakers on the basis of characteristics of the talk or the speaker (e.g., de la Zerda & Hopper, 1979; Giles & Powesland, 1975; Zahn & Hopper, 1985). The influence of ordinary
language philosophy (e.g., Austin, 1962; Wittgenstein, 1953) prompted studies of speech
as action (e.g., Hopper, 1981a). Concurrently, sociological studies reflecting the influence of symbolic interactionists directed attention to such topics as accounts and formulations under the umbrella term alignment talk (e.g., Morris & Hopper, 1987; Ragan &

An overview of language and Social interaction research 5

Hopper, 1981). An emphasis on issues of coherence and cohesion drawn from linguistics
(Coulthard, 1977) combined with these other streams under a broader label of discourse
analysis (e.g., Ellis, 1995; Hopper, 1983). At the same time, ethnographic approaches to
communication were drawn from fields such as linguistic anthropology (e.g., Fitch & Hopper, 1983; Philipsen, 1975). Conversation analysis in the ethnomethodological tradition
(e.g., Beach, 1982) provided alternative methods for studying structures and functions of
everyday language use and, through such study, for investigating processes whereby people communicatively constitute everyday activities (e.g., Hopper & Doany, 1989; Hopper
& Drummond, 1990, 1992; Hopper & Glenn, 1994; Hopper, Thomason, & Ward, 1993).
More recently, continued technological developments (e.g., multimedia and digital video)
have opened up new opportunities for conducting detailed studies of embodied interactions, thereby creating a parallel stream to continued research on the organization and
workings of speech-in-interaction (e.g., LeBaron & Hopper, 1999). This parallel stream
furthers a tradition of ethological study and context analysis exemplified in the work of
Kendon (1990). Recent work in LSI also reflects and contributes to theory and research in
performance studies (e.g., Hopper, 1993a, 1993b). For communication researchers using
LSI methods, the essential feature of interest is human communication itself, which contrasts with scholars in related academic disciplines who use LSI methods but display ultimate preoccupation with language, society, or culture.
The relationship between LSI and the field of communication has been mutually influential and beneficial. On one hand, LSI research has increased understanding of what
communication is and how it is done. Arguably, the field of communication has been preoccupied with various factors that influence communication (such as individual dispositions,
contexts, goals, gender, etc.), and with how communication influences a variety of factors
(satisfaction, compliance, persuasion, social support, etc.), at the expense of examining
the actual processes through which communication occurs. The LSI focus on discourse
(or alternate terms such as speech, messages, talk, conversation, or interaction) has helped
shape these issues as central to the communication discipline. On the other hand, traditions
within communication studies have helped to shape LSI research. To illustrate, we identify
the following four areas of mutual influence.
First: Moving Beyond the Sender-Receiver Model
During the telecommunications boom associated with World War II, Shannon and Weaver
(1949) proposed a model of communication based on their knowledge of how the telephone works. According to their model, communication begins with a source or sender,
who encodes thoughts or feelings into a message that is then transmitted across a channel
to a receiver, who in turn decodes the message and thereby understands the information
transmitted. This model had immediate and widespread appeal as it perpetuated a psychological view and at the same time resonated with the traditional rhetorical topoi of speaker,
message, audience, and context. Although the transmission model was useful and fruitful
in many ways, and although it continues to be taken for granted by many social scientists
and laypersons, much communication research acknowledges the importance of moving
beyond the transmission model (e.g., see Goldsmith & Baxter, 1996). Arguably, too much
research on communication has tried to isolate component parts of the transmission model,

6 Studies in language and social interaction

at the cost of seeing communication as a constitutive process through which interactants
work together to construct lines of action.
Three decades of LSI research have helped the field of communication to specify the
details of the move beyond the transmission model and toward a social constructionist or
constitutive view of communication. Using an array of empirical methods, LSI researchers
have shown that:
Messages are not discrete from peoplein some ways people are the message;
Notions of self and other are constituted in and through discourse, and the\boundaries between sender, message, and receiver are not always clear;
Meaning is not solely the product of the senderrather, messages and meanings are
joint creations, even if only one person appears to be doing most of the speaking;
Meanings may remain incomplete, emergent, and subject to retrospective modification;
Messages and context are mutually elaborative;
Context is invoked, oriented to, and constituted in interaction;
And conversely, context influences the organization of interaction; and so forth.
Thus, LSI researchers have shown that human interaction is partly or largely constitutive
of the component parts that the sender-receiver model takes for granted. That is to say,
through communication participants perform and realize their relative roles, interactively
negotiating the meanings of so-called messages, orienting toward some symbol systems as
relevant and recognizable, in many ways constituting their communicative context (e.g.,
Hopper, 1992b; Hopper & LeBaron, 1998). (A constitutive view of communication is further discussed later.)
Second: Reexamining Cognitive and/or Theoretical Constructs
Through different sorts of empirical investigation (often involving analysis of audio recordings, video recordings, and/or field notes), LSI researchers have reconsidered and respecified various theoretical constructs associated with the field of communication. Sometimes
specific concepts have been the target of LSI investigation from the outset. That is, LSI
researchers have occasionally set out to examine details of the empirical world with the
express purpose of scrutinizing theoretically derived concepts. For example, researchers
with a specific interest in social identity have collected and examined discourse to learn
more about the interactive construction of identity in everyday life (e.g., Carbaugh, 1993;
Mandelbaum, 1994; Tracy, 1997). Some ethnographers have reexamined the traditional
and monolithic concept of culture, respecifying it as practices whereby culture is constructed through conduct (e.g., Fitch, 1998a). Through analyses of audiotaped and videotaped communication within classrooms and schools (e.g., McHoul, 1990; see also chap.
6, this volume), LSI researchers have shown that human minds extend beyond the skin as
people depend upon social and material worlds to acquire knowledge and display intellectual ability. Therapeutic discourse has also been an object of study (e.g., Bavelas, 1989;
Buttny, 1993, 1996; LeBaron & Hopper, 1999; Morris & Chenail, 1995; Perkyl, 1995) as
LSI researchers have sought to emphasize social aspects of patients mental or psychological states. In this way, theoretical concepts associated with the field of communication have
guided LSI research, which has in turn influenced the field at large.

An overview of language and Social interaction research 7

Other times, theoretical constructs have come under scrutiny in the course of LSI research
on a set of data already collected. Conversation analysts regularly advocate unmotivated
looking (Pomerantz & Fehr, 1997; Sacks, 1984), such as through data sessions, a process
whereby data are analyzed in order to see what is going on and how it is getting done,
which routinely leads to discovering phenomena occurring in the wild, perhaps warranting respecification of theoretical constructs in the end. For instance, practices of relationship construction or dismemberment have been respecified after examinations of data have
shown an opportunity for doing so (e.g., Hopper & Drummond, 1992; Mandelbaum 1989).
Processes through which gender becomes socially relevant have been similarly reexamined
(e.g., Hopper & LeBaron, 1998; Lawrence, Stucky, & Hopper, 1990; see also chaps. 15 and
16, this volume). Philipsen (1975) used ethnographic methods to study Teamsterville culture and discovered that (and how) the value of speaking or fighting may vary significantly
from one culture to another. In his book, Conversations About Illness, Beach (1996) noted
that he did not begin with an interest in studying eating disorders or the social construction
of illnessrather, he came across data providing a compelling entry into these issues and
allowing for respecification of them. Through close examination of empirical data, then,
LSI researchers have come upon opportunities to reconsider and respecify conceptual and/
or theoretical constructs within the field of communication.
Third: Bringing Verbal and Nonverbal Communication Together
Within the field of communication (and other social sciences), verbal and nonverbal forms
of communication have traditionally been treated as separable, distinct areas of inquiry.
Although scholars of various stripes have lamented this artificial separation (e.g., see
Streeck & Knapp, 1992, who described the separation as misleading and obsolete), the field
of communication generally has made little progress toward mending the rift. Recently,
however, LSI researchers have employed methods that bring the two modalities togetheror rather, have examined vocal and visible forms of communication without separating
them in the first place. Through methods that rely on videotaped recordings of naturally
occurring interaction, LSI researchers have been able to get at communication as it is
holistically enacted by interlocutors in the first place (e.g., C.Goodwin, 1986; C.Goodwin
& M.H.Goodwin, 1986; LeBaron & Streeck, 1997; Streeck, 1984, 1993, 1994, 1996).
The field of communication and LSI research will undoubtedly continue to be mutually
influential in this area.
Fourth: Appreciating the Poetics of Language
After separating from English (and theater) departments in 1914, scholars attempting to
establish a science of speech tried to distance themselves from the literary and theatrical
traditions. Nevertheless, scholarly interest in performance and other humanistic approaches
has flourished within the field of communication. Contemporary uses of the term performance within communication include (a) a research method for studying communication,
(b) an important feature of communication, and (c) a useful metaphor for talking about
communication. This abiding interest within the field has influenced studies of language
and social interaction. Performance methods have proven useful in sociolinguistic studies

8 Studies in language and social interaction

of speech evaluation (Lawrence et al., 1990). Methods in LSI, which are notorious for close
attention to discourse texts, invite noticing of poetic and performative features of everyday
interaction. For example, Hopper (1992b) likened his own transcriptions to stanzas of a
poem, and his scientific work was often inspirited with a poetic sense of social life (e.g.,
Hopper, 1991, 1992a, 1993a, 1995). Hopper and other LSI researchers have explored theoretical and theatrical applications of using transcripts plus recordings of naturally occurring
interactions as scripts for staged performance (e.g., Crow, 1988; Stucky, 1988; see also
chap. 29, this volume). This has led to a substantial body of performed and written scholarship on what has been called everyday life performance (ELP). Repeated applications
have shown that ELP makes for lively and insightful theatrical productions (e.g., Hopper,
1996). Furthermore, the ELP processes help practitioners learn about self and others, about
patterns of interaction, and about production nuances of everyday talk (Stringer & Hopper,
1997). Thus, LSI research has significantly benefited from and contributed to performance
studies within the field of communication (e.g., Gray & Van Costing, 1996).
To summarize, we have briefly described the historical emergence of LSI research
within the field of communication and have discussed a few areas of mutual influence
between the division and the field. Robert Hopper, as much or more than any other scholar,
has been central to this unfolding. We now turn our attention more specifically to current
trends within LSI research. In the following section, we identify and discuss seven points
of commonality and contention within the areathat is, contested points around which
LSI scholars tend to rally in one way or another, points whereby LSI studies bear a family
resemblance (Wittgenstein, 1953) to one another.
The field of communication is like a no-host party at an academic convention4. Communication scholars have come together and noisily organized themselves into various
divisions or interest groups where they talk, sometimes to be overheard by other groups.
Membership within each division fluctuates as scholars come and go, sometimes listening,
sometimes talking, arriving after the discussion has already begun and leaving before it is
complete. Although the organization of a particular division may be somewhat arbitrary,
it is nonetheless consequential for those involved: What may be stated and how, who may
state it and when, depends largely upon the participants who subtly negotiate the trajectory
of their conversation and the standards for appropriate participation.
LSI is an eclectic group, boasting various intellectual pedigrees. Not only are a variety
of research methods employedincluding ethnography, discourse analysis, conversation
analysis, sociolinguistics, micro-ethnography, and pragmaticsbut some scholars choose
to blend methods (e.g., Moerman, 1988; Tracy, 1995). Clearly, such diversity has had synergistic outcomes for the discipline, but it has also led to basic disagreements (e.g., see
Beach, 1995a; Sanders & Sigman, 1994; Tracy, 1994) and self-contemplation (e.g., Craig,
1999; Ellis, 1999b; Sanders, 1999; Wieder, 1999) on the nature of the discipline. As we
Our analogy is a crude adaptation of Burkes (1941/1973) parlor metaphor, where the human
condition is likened to an unending conversation (p. 111).

An overview of language and Social interaction research 9

privilege one way of describing here, we recognize that there are countless other ways
that the field could be describedchronologically, topically, ideologically, methodologically, demographically, logistically, and so forth. Our choices (perhaps biases) have consequences for the centers and margins of the field we depict, which may include or exclude
colleagues in odd or unfortunate ways. Nevertheless, occasional stocktaking may help to
promote synergistic outcomes and prevent or reconcile unnecessary fissures within the
field. Despite the risks, our description may help newcomers who are preparing to join the
lively conversation underway, or it may help active LSI scholars assess their discipline and
participation. In recent years, especially with the start of a new millennium, LSI scholars
have seen several stocktaking exercises in the form of papers, panels, and publications
(e.g., see special issues of Research on Language and Social Interaction, such as the Talking Culture issue in 1990, and the Millennium issue in 1999). Because our description is
only one of several, we hope that it will continue dialogue rather than discourage it, invite
and include participants rather than exclude them.
Our description is organized around key pointsor contested conceptswe think
underlie, unify, and galvanize LSI research. Specifically, we propose that LSI researchers
tend to rally around the following interrelated points, agreeing and disagreeing with them
in various ways, whereby LSI studies take on a recognizable relationship to one another:
1. LSI research privileges mundane, naturally occurring interaction within casual and
institutional settings.
2. LSI research adheres to principles of an empirical social science.
3. LSI research describes and explains.
4. LSI research is inductive and abductive.
5. LSI research treats communication as constitutive and consequential.
6. LSI research emphasizes emic, participant perspectives.
7. LSI research focuses on language in use.
Why have we approached our description of LSI in this way? Because work in LSI is
unusually eclectic and faces the ongoing challenge of holding to common ground while
exploring new and different directions for scholarship. We acknowledge that our list of
seven points may be incomplete and may at some stage become obsolete. Moreover, we
strongly emphasize that adherence to any one of the seven points listed is not required for
membership within the LSI family. Rather, each point is a contested site of commonality within the field, and we present (herein) plenty of counterexamples for each point,
showing that each has been contested by the very researchers that these points have generally brought together. As evidenced by the descriptions that follow, these seven points are
interrelatedeven overlapping, though not redundant.
Research on Language and Social Interaction Privileges Mundane, Naturally Occurring
Interaction Within Casual and Institutional Settings
A conversation between two people washing dishes in their kitchen, for example, may
warrant examination as much or more than a televised presidential speech. The term mundane refers to communication that may be commonplace regardless of setting, is usually

10 Studies in language and social interaction

uncelebrated, and is too often dismissed as unremarkable or unimportant. The term also
incorporates features of communication that are often ignored or regarded as peripheral,
such as vocal restarts and hesitations (e.g., C. Goodwin, 1980), laughter (e.g., Glenn, 1989,
1992, 1995; Jefferson, 1994), and seemingly insignificant acknowledgment tokens such as
oh (e.g., Heritage, 1984) and okay (Beach, 1993, 1995b). Communication is considered to be naturally occurring if it would have occurred whether or not it was observed
or recorded (see Beach, 1990, 1994). Participant observations, field notes, and audio or
video recordings of everyday speech events are considered premium data from which to
make conclusions about human communication and social life. Sacks (1984) criticized
a common concern among social scientists for finding supposed good data and good
problems. He observed:
Such a view tends to be heavily controlled by an overriding interest in what are in
the first instance known to be big issues, and not those kinds of objects they use to
construct and order their affairs, (pp. 2224)
Such emphasis on mundane and naturalistic communication diverges from a variety of
other research traditions. LSI research contrasts with methodologies that (a) rely upon
hypothetical or imagined exemplars of language use as a basis for linguistic claims, (b)
focus exclusively upon mass-mediated events, such as a television drama, as a basis for
conclusions about culture, (c) concentrate only upon big speech events, such as presidential speeches, which are supposed to be especially important to society, or (d) generate
data through experimental methods, perhaps under laboratory conditions where subjects
are removed from the social and material environments in which they typically interact.
Although LSI research privileges mundane interaction, considerable attention has been
given to popular and publicized speech events. For instance, Atkinson (1984) scrutinized
the behavioral patterns (both vocal and visible) of Margaret Thatcher and Ronald Reagan during public speeches, and identified devices whereby the politicians cued audience applause and interactively performed charisma. In a special issue of Research on
Language and Social Interaction, several scholars analyzed patterns of turntaking and
interruption during an explosive television interview (or rather argument) between Dan
Rather and George Bush, when Bush was campaigning for the U.S. presidency in the 1980s
(e.g., Nofsinger, 1988). Bavelas, Black, Chovil, and Mullett (1990) examined equivocal
statements that politicians use to cope with no-win situationsthat is, when all direct
messages would lead to negative consequences. Lynch and Bogen (1996) studied congressional procedure and testimony associated with the Iran-Contra hearings, showing how the
history of illegal activities was contested and interactively produced. Carbaugh (1989)
conducted an ethnographic study of the Donahue television show, depicting it as a portrait of American society. John Modaff (chap. 28, this volume) microanalyzed the speech
melody of radio personality Paul Harvey, and identified rhetorical properties of his vocal
inflections. These citations (and numerous others) notwithstanding, research on language
and social interaction is overwhelmingly concerned with mundane features of mundane
interaction. Although researchers occasionally focus on the communicative behaviors and
cultural furnishings of politicians and other public performers, it is the behaviors and the

An overview of language and Social interaction research 11

furnishings themselves that warrant the LSI studynot the celebrities, nor their histories.
Studies of the spectacular may inform us about what is commonplace.
Mundane interaction (as we defined it) occurs in both casual and institutional settings.
Beach (1996) argued that families are the primordial institutional systems (p. xi) and
that interactions between, say, a grandmother and a granddaughter might reveal patterns of
interrogation like those found in a courtroom. LSI researchers have entered an array of
social institutions and organizations to explicate the everyday behaviors whereby institutions are interactively formed and sustained (e.g., Atkinson & Drew, 1979; Atkinson &
Heritage, 1984; Drew & Heritage, 1992; Houtkoop-Steenstra & Antaki, 1997; Metzger &
Beach, 1996; Morris & Cheneil, 1995; Tracy, 1995, 1997). For example, recent research
on medical interviewing has addressed significant moments between doctors and patients
(e.g., Beach & Dixson, 2000). Conversations about health and illness also occur at home,
such as when family members discuss a loved ones diagnosis and treatment for cancer
(chap. 10, this volume).
Recently, the notion of naturally occurring has been indirectly and directly called into
question. For instance, Pratt and Wieder (1993) conducted an ethnography of public
speaking among the Osage Nation, a Native American community. Not only were public
speeches prepared or scripted in advance, these researchers asked subjects to reperform
speeches that they had given before during some prior ceremony or event of the Osage
Nation. Pratt and Wieder argued that their data were sufficiently natural because the focus
of their study was on the formal features of the original speeches and not the in-themoment contingencies (p. 358). Bavelas (1999) worked to broaden notions of naturalistic
within the field of LSI. She argued that laboratory data should not be dismissed out of hand,
because when people communicate under laboratory conditions, they necessarily employ
the sorts of vocal and visible behaviors whereby they communicate everydaythere is
no other way to interact. Moreover, Bavelas suggested that a laboratory may need to be
recognized as a special site (with its own social and material affordances), but it should not
be rejected as artificial just because it is built to serve researchers endsafter all, every
built space serves some social and micropolitical end.
The notion of naturalistic has also been stretched by literary inclinations. In his
book on gender and gender talk, for example, Hopper (in press) supplemented his tape
recordings of naturally occurring talk with exemplars from other sources, including the
Fiction. For obvious reasons, there are few candid recordings of moments involving
sexuality, sexual harassment, codependent family interaction, and so forth. Films regularly portray such dialogue in a way that resembles everyday social interaction, which
may serve as a resource for scientific inquiry.
Self-reports. Ethnographers routinely interview people about their speech practices.
Self-report data show few discourse features and they may be replete with socialdesirability biases, but participants recollections of social interaction have proven to
be a useful resource.
Hypothetical examples. In the absence of recorded data or firsthand observation, a
writer may fabricate a hypothetical example to illustrate (precisely) a particular
argument. Such fabrications often stand up through replication and critical scrutiny,

12 Studies in language and social interaction

perhaps due to the incredible overdetermined orderliness of language use and social
Hopper openly acknowledged the risk of mixing evidence types. Of course scientists must
be wary of generalizing from film to life, and self-report findings should be confirmed
by fuller discourse renderings. Nevertheless, by mixing evidence types Hopper was able
to address areas of theory and general concern for which limited data could be found. In
another study, Drummond (chap. 32, this volume) participated in the dialogue between
real and fiction: Using Hoppers (198 la; 1981b) notion of taken-for-granted, Drummond
explicated the idea of interactional enslavement within the movie The Truman Show.
Points suggested by more literary sorts of evidence may be taken as a stimulus to collect
more naturalistic examples of similar phenomena.
Research on Language and Social Interaction Adheres to Principles of an Empirical
Social Science
Research conclusions about communication, culture, and social life are properly supported
by firsthand observations of human interaction. When LSI researchers present their findings in papers or reports, they usually include examples or excerpts of the phenomenon
under investigation. Careful descriptions, field notes, transcriptions, photographs, videotapes, and other sorts of recordings are taken to represent the audible and visible behaviors
that social interactants made available to each other (in the first place) and to analysts (who
acted as overhearers and onlookers). Hence, all arguments are based on evidence that must
pass the test of intersubjective agreement among researchers and readers (see Beach, 1990,
1994; Pomerantz & Fehr, 1997). A particular phenomenon is taken to exist, to the extent
that data, analyses, and conclusions are reproducible or verifiable by others.
At the same time that most language and social interaction researchers maintain an
empirical stance toward their objects of study, we suggest that they necessarily engage in
an ongoing interpretive process. Researchers are participants in the social world they analyze, both creating and interpreting human experience, moment to moment and day to day.
Researchers do more than document patternsthey appraise the significance of behaviors
documented. Geertz(1973) wrote:
Believing, with Max Weber, that man is an animal suspended in webs of significance
he himself has spun, I take culture to be those webs, and the analysis of it to be therefore not an experimental science in search of law but an interpretive one in search of
meaning, (p. 5)
To some extent, all research on language and social interaction has kinship with the work
of Geertz, who sought to understand human cultures through thick descriptionrather
than explain them through theories of causation or natural law. Research on language and
social interaction is itself suspended within the webs of culture that it brings to light. Forms
of communication that may be empirically ascertained are also interpreted and thereby
made meaningful to participants and analysts alike.
Within the field of language and social interaction, some methods flaunt their interpretive stripes more than others. On one hand, ethnographers seek presence and participation

An overview of language and Social interaction research 13

within the speech communities they study, acknowledging their interpretive role and even
relishing the flavor of their own influence. Their basis for selecting objects of ethnographic
study is sometimes unsystematic and rather intuitiveby design. For instance, Fitch
(1994) observed that some ethnographers choose to examine cultural sites and communicative practices that contrast strikingly with their own. The best way to understand and
accurately report on a culture, the ethnographic argument goes, is to fully experience and
interpret it as do the cultural members themselves. In a study of culture within the southern
United States, Fitch (1998b) recorded a conversation in which she participated; she then
transcribed and analyzed the talk (including her own); and finally she contemplated (as part
of her ethnographic report) the difference between her in-the-moment (subjective) experience and her later (objective) microanalysis of it. Hence, to change the ethnographer would
be to alter the ethnographic outcome.
On the other hand, conversation analysts may downplay and even deny their interpretive role. They rarely appear as participants within the data they choose to examine; they
seldom rely on in-the-moment observations of speech events, choosing instead to focus
on audio or video recordings; and they present their findings as being empirically evident,
independent of the particular analyst. Hopper et al. (1986) described conversation analysis
(CA) as a search for patterns in the mode of natural science. As paleontology describes
fossils to understand geological history, CA describes recordings to understand structures of
conversational action and members practices for conversing (p. 169). Despite the empirical rigor that conversation analysts insist on (see also, Sacks, 1984, 1992), they ought to
also recognize their subtle but substantive interpretive moves. Even before recorded messages are analyzed, recording itself is an interpretive act: Cameras and tape recorders must
be placed, pointed, and turned on, which is to make decisions about what is important or
worth recording; transcripts are necessarily selective. Moreover, conversation analysts rely
on members knowledge (i.e., the interpretations that interactants show to one another in
the course of their interaction) to understand what is being displayed within data. Some
conversation analysts accept and even embrace their interpretive bent. For example, Hoppers (1992b) analysis of telephone conversation often waxed poetic. He encouraged readers to attune themselves to a primordial voicethe voice of poetry in conversation, the
great Poem, speaking us (p. 190). Thus, even the most rigorous empiricist may orient to,
listen to, and be inspired by the humanist within.
Despite these variations within the field of LSI, there is a general commitment to
empirical methods. After acknowledging the role of intuition in ethnographic research,
Fitch (1994) recommended more systematic bases for ethnographic choices. And Hoppers
(1992b) poetic treatment of telephone conversations was constantly based upon empirical
details displayed by participants to one another (p. 20). Overwhelmingly, LSI researchers treat what they are doing as meriting scientific status, affirming the need for clear and
repeatable methods to produce replicable results.
Research on Language and Social Interaction
Describes and Explains
By carefully and thoroughly describing human interaction, researchers begin to understand
and explain it. Most LSI research provides straightforward (even matter-of-fact) accounts

14 Studies in language and social interaction

of phenomena, written as if the features of human interaction exist in the social world to be
documented and interpreted. Nevertheless, description is not a neutral activity and data are
not self-explicating. The item(s) chosen for analysis represent important choices (whether
conscious or unconscious) by the researcher. For this reason, LSI researchers tend to be
reflexive about word choice, writing style, and presentation of data, recognizing that these
are in part constitutive of the social phenomena under investigation. Conversation analysts
seem especially particular about terminology. For example, when Pomerantz (1989) suggested that conversation analysts translate CA jargon into more commonsense lay terminology, so as to make it more accessible to more readers, Jefferson (1989) disagreed. Jefferson
insisted that CA terminology is not just a complicated way of saying what otherwise can
be said with lay, commonsense, interactants terminology (p. 427); rather, she insisted, CA
terms are imbued with special ways of looking at and describing the social world.
Data presentation is also an ongoing concern. Ochs (1979) observed that presentation
tools such as transcription systems are inherently theoretical and should not be regarded
as one-to-one representations of reality. Jacoby and Ochs (1995) emphasized that human
interaction is contingently dynamic and unfolding in interactional time (p. 179) and that
researchers who use recordings and transcriptions should not treat communication as a
freestanding text. Jarmon (1996a) became frustrated with the presentational constraints
of transcriptions and written descriptions, so she began using multimedia technology and
eventually produced a dissertation on CD-ROM. Her dissertation proposed an amendment
to the turn-taking model published by Sacks, Schegloff, and Jefferson (1974), who based
their model on analysis of audio recordings. Through analysis of videotaped recordings,
Jarmon concluded that embodied actions (such as facial expressions) are in some ways
similar to grammatical units and may alter the projectability of turn boundaries or even
function as a complete turn. Thus, the distinction between good description and good
analysis blurs, as description documents and characterizes phenomena, providing both the
basis and the impetus for analysis that follows in the wake. Even the term description
may prove misleading or unduly limiting, to the extent that it buys into a representative
view that there is a reality out there that may be described, in contrast to a social constructionist perspective that the act of attempting to write about something discursively
constitutes that something.
Within the field of language and social interaction, description and explanation are
regarded as worthwhile research goals or achievements in and of themselves. This contrasts with a hypothetico-deductive approach to communication research, which views
description as only a first step that is incomplete unless followed by more substantive steps
of developing theory, deriving hypotheses, and testing them experimentally. Descriptive
research also contrasts with critical research, for which description may precede and set up
a move to evaluation by practical, aesthetic, political, or moral standards. A third contrast
is with applied research, for which description provides a starting point allowing a move
to prescription, training, or pedagogy. There are plenty of examples of LSI research that
do make critical or applied turns. Van Dijks critical discourse analyses (1993, e.g.), like
Conquergoods (1991) critical ethnography, seek to apply naturalistic methods to social
problems. Likewise, LSI research on discourse within institutional settings (e.g., Drew
& Heritage, 1992; Houtkoop-Steenstra & Antaki, 1997; Tracy, 1995, 1997) either explicitly makes or leads closer to deriving prescriptive applications from research findings.

An overview of language and Social interaction research 15

Nevertheless, even in these examples of LSI research, description and explanation remain
the central tasks.
Research on Language and Social Interaction is Inductive and Abductive
There is a general commitment among LSI scholars to avoid premature theory building.
Rather than begin with a research question or hypothesis (from which data collection,
analysis, and conclusions would logically follow), LSI researchers regularly begin with
data: Naturally occurring communication is observed or recorded and analyzed, and from
this process new (sometimes revolutionary) claims and conclusions emerge. Ethnographers
have a long tradition of selecting speech communities to study without knowing in advance
what sorts of findings might arise. Sacks (1984) recommended the following bottom-up
approach to research:
When we start out with a piece of data, the question of what we are going to end up
with, what kind of findings it will give, should not be a consideration. We sit down
with a piece of data, make a bunch of observations, and see where they will go. Treating some actual conversation in an unmotivated way, that is, giving some consideration to whatever can be found in any particular conversation we happen to have our
hands on, subjecting it to investigation in any direction that can be produced from it,
can have strong payoffs, (p. 27)
Although some readers may think Sacks is being idealisticthat is, to what extent can any
examination be truly unmotivated?many LSI studies indeed begin in this way. Soon
researchers notice and take interest in some phenomenon, and unmotivated looking gives
way to directed examination and explanation. As a research project takes shape, inductive
methods tend to become more abductive. Analysts go looking for instances within naturally occurring data that may support a particular claim. The field of LSI is notorious for socalled bottom-up inquiry and inductive proof, whereby claims are consistently grounded
by reference to evidence in data.
Technology not only supports naturalistic research, it facilitates inductive inquiry and
insight. A primary challenge for LSI researchers is to recognize what is commonly taken
for granted: Because researchers are themselves embedded every day within forms of communication and culture, it may be difficult for them to look at the social world that
they are accustomed to looking through. Field notes, transcripts, photographs, audiotapes, films, videotapes, multimedia, and other forms of technology help to make the social
world strange, enabling researchers to perceive it, as Garfmkel would say, for another
first time. Bateson and Mead (1942) reported using photographs in their ethnographic
work because photographs could capture and present behavioral events better than verbal
descriptions. Sacks founded the field of CA after discovering recordings of telephone conversations, which provided the proximate source for the focused attention to talk itself
perhaps the most critical step toward the development of conversation analysis (Schegloff,
1992a, p. xvi). Kendon studied talk until 1963, when he discovered film and began to
analyze embodied interaction: It became apparent at once that there were complex patterns and regularities of behavior, and that the interactants were guiding their behavior,

16 Studies in language and social interaction

each in relation to the other (Kendon, 1990, p. 4). Using multimedia technology, LeBaron
(1998) digitized and then microanalyzed video recordings, and found recurring hand gestures that were identifiable because the computer provided a nonlinear environment within
which to work, making it possible to analyze multiple videotaped images simultaneously,
juxtaposing them on the computer screen. Moreover, technology allows for detailed and
repeated examination of messages. It also affords the opportunity to manipulate messages
so that analysts can see how the interaction changes when they slow it down or zoom in on
different features of a visual image.
Induction can serve both as a pattern for the research process and a pattern for the written research report (although these need not parallel each other). Reports tend to take shape
as either (a) claims based on a collection of occurrences, each documented and discussed,
that altogether warrant some subsuming claim about LSI within a speech community or
culture (e.g., Coutu, 2000; C.Goodwin, 1980; ten Have, 1999), or (b) a detailed explication of some single, perhaps singular, occurrence that reflects upon the language and social
interaction of a speech community or culture (e.g., C.Goodwin, 1979; Philipsen, 2000).
What occurrence(s) a researcher chooses to reportor is able to reportdepends on the
LSI method being employed. With roots in a sociological method Znaniecki (1934) called
analytic induction, conversation analysts may assume the responsibility of identifying a
structural pattern in a way that shows recurrence in the routine instances but also shows
orientation to the regularity in the deviant cases (e.g., Schegloff, 1986). The aim is to
provide an account of the phenomenon that holds beyond the particular instance, such an
account thereby being both context sensitive and context free (Sacks et al., 1974). Discourse analysts choices may be informed by a wide variety of influences, from linguistic categorizations and structures to whatever themes or beliefs subjects manifest through
their situated discourse or through interviews with the researcher (e.g., Tracy & Muller,
1994). Ethnographic choices may be guided by the researchers intuition or reflection, the
subjects disclosures or interpretations of occurrences, the community members overall
insights and reflections (as gleaned through interviews), or some universal theory (e.g.,
politeness theory) against which the ethnographer may work (Fitch, 1994). Despite obvious differences in these inductive methods, there is an abiding assumption that a priori
theorizing risks diverting attention away from the central tasks of describing and explaining phenomena based on observable details (see Sanders & Sigman, 1994). A combined
emphasis on description, explanation, induction, and abduction gives LSI work a basis for
its empirical grounding.
Not all LSI research is inductive. In his early programmatic statements about the ethnography of speaking, Hymes (1978) asserted that descriptive accounts of cultural ways
of speaking could and should be followed by subsequent research in which hypotheses
are developed and tested in the field. Sociolinguistic research on power and speaker style
often operates under a deductive framework, drawing on preceding research to generate
hypotheses for testing (chap. 2, this volume). Mulac, Bradac, and Gibbons (2001) draw on
previous research to generate (and subsequently test) research questions and a hypothesis
regarding gender-based differences in language use. Discourse analytic, conversation analytic, and ethnographic reports may make use of previous research to explicate features
within a present set of data, and as findings accumulate, opportunities increase for applying
generalized claims in making sense of newly encountered particular instances. Periodically

An overview of language and Social interaction research 17

a researcher may take stock of some line of research and make a generalized statement
about a phenomenon (e.g., Morris, White, & Iltis, 1994). Moreover, some research focuses
on a theory question that the data did not in the first place suggest; some analyses rely on
data having turned up that happen to relate to a particular question or theory or practice.
For example, Hopper and Drummond (1990) joined a theoretical discussion about romance
turning points only after they found a telephone recording that happened to include a
dating break-up. Nevertheless, the primary goal of most LSI research involves careful
description and explanation, accomplished through the inductive and abductive process of
gradually building generalized claims from analysis of particular cases of a phenomenon.
Research on Language and Social Interaction Treats Communication as
Constitutive and Consequential
The transmission model of communication (Shannon & Weaver, 1949), discussed earlier,
typifies a representative view of communication, which sees language as reflecting a preexisting and external reality. Although the transmission model was widely accepted and
continues to be taken for granted by most social scientists and laypersons, it has been
repudiated by three decades of research on LSI, which shows that human interaction is
partly or largely constitutive of the component parts that the transmission model presupposes. Even social conditions thought to be stable are contingent and constantly shifting
as interlocutors co-construct their social worlds (Jacoby & Ochs, 1995)including gender (Sheldon, 1996), ethnic identity (He, 1995), and individual competence (C. Goodwin,
1995). Setting aside the assumption that context exists a priori and that context unilaterally
shapes communication, LSI research has shown how context may be invoked, oriented
to, and constituted through social interaction at the same time that context may influence
the organization of communication (e.g., see Drew & Heritage, 1992; Tracy, 1998). The
LSI perspective that communication and context are mutually elaborative contrasts with
more representative, static, or external to message (Hopper, 1992b) approaches. According to a constitutive view, then, communication is a primary means whereby social realities, cultural contexts, and the meanings of messages are interactively accomplished and
experienced (Stewart, 1995).
Commitments to a representative or constitutive view can operate at two levels. The first
level is the extent to which researchers treat interactants as themselves constituting their
social realities. Ethnomethodology, with its focus on how people construct social order, has
informed conversation analysis and allied methods. Expounding on the work of Garfinkel,
Heritage (1984) observed that messages are not inherently meaningful, because communicative behaviors are subject to inference and open to negotiation among participants:
Utterances accomplish particular actions by virtue of their placement and participation
within sequences of action (p. 245). In an examination of a videotaped business meeting, Streeck (1996) found that material objectsnot just spoken and written messages
may become (situated) symbols through their appropriation and physical placement during
face-to-face interaction. Among the things that interaction may accomplish is the instantiation of social roles (Schegloff, 1992b), everything from sender-receiver to mother-daughter
(Hopper, 1992b). In analyses of storytelling, C. Goodwin (1984) and Mandelbaum (1987)
identified patterns of talk whereby the roles of storyteller and hearer were jointly achieved.

18 Studies in language and social interaction

Button (1992) examined recordings of job interviews and identified question-and-answer
structures of speech whereby people may perform the roles of interviewer and interviewee.
Even built spaces (i.e., physical structures made of brick and steel) are given shape and
significance through social interaction. LeBaron and Streeck (1997) examined a videotaped
police interrogation in which participants moved their bodies in strategic ways, appropriating and interpreting the physical features of their interrogation room, making possible
certain vocal arguments that eventually moved the suspect toward confession.
The second level is the extent to which researchers explicitly acknowledge or problematize how research itself represents or constitutes the social phenomena under investigation.
In other words, do researchers discover and represent the objects of their study, or does the
research process itself bring phenomena into being? It is difficult to find examples of LSI
research that take a radically constitutive stance at this second level by explicitly focusing on the researchers role in constituting the objects of study. This provides a point of
divergence for ethnographers working in the Hymesian ethnography of speaking tradition
and those engaging in autoethnography (e.g., Bochner & C.Ellis, 1995). Likewise, some
ethnomethodologists have criticized conversation analysts for failing to practice radical
reflexivity (Pollner, 1991). Many discourse analytic, conversation analytic, and sociolinguistic studies tend to employ a reporting vocabulary and posture that minimizes explicit
attention to the researcher as an active creator of meaning (see item 3, earlier).
The ethnomethodological roots of some methods could nudge researchers toward viewing their work as constitutive. Conversation analysts, for example, avoid invoking labels or
categories or contexts unless those are demonstrably relevant for participants. For instance,
Button (1992) said that in the face of multiple categorization possibilities for any person
(an interviewer may be a father as well, for instance), the warrantable use of a categorization by a researcher resides in the participants orientation to and constitution of their
activities (p. 230). Although such self-awareness among researchers has the blush of a
constitutive view, conversation analysts regard their reflexivity as a form of rigor and see
themselves as all the more accurate in their reporting. Occasionally LSI researchers turn
their cameras and recording devices on themselves. For instance, Jarmon (1996b) examined videotapes of conversation analysts at work. While participating in a data session,
the analysts performed with their bodies what they saw in their data, tailoring their performances to display specific analyses and arguments. Jarmon discussed the degree to which
performance may play a part in how research is conducted (p. 16), but her conclusions
stopped short of a radically constitutive view of research. In another study, Modaff and
Modaff (1999) talked to each other on the telephone, recorded their conversations at both
locations, transcribed both recordings, and then analyzed both transcriptions using conversation analytic methods. After finding substantive differences between the transcriptions,
the researchers questioned the accuracy of mainstream recording devices and hence the
accuracy of LSI research that depends on such devices. Thus, Modaff and Modaff took a
representative stance by arguing for more accuracy in LSI research methodsthey did not
assume a radically constitutive view of the researcher as one who more or less creates the
phenomena under investigation.
In practice, the representative view and the constitutive view are not mutually exclusive, freestanding alternatives; rather, they are ways of conceptualizing communication
that have points of convergence. Within the division of LSI, or even within a particular

An overview of language and Social interaction research 19

research report, combinations of these views may be evident (e.g., see Tracy, 1998, who
edited a special journal issue on Analyzing Context, in which LSI researchers aligned
with representative or constitutive views in various ways). To illustrate, consider the extent
to which culture determines or is determined by everyday communication. Some LSI
researchers (e.g., ethnographers) may implicitly or explicitly recognize that communication at any one moment is responsive to the history of interactional moments experienced
by participants individually and collectively over time. Others (e.g., conversation analysts)
may ignore or downplay the impact of established cultural or linguistic resources on a
particular moment of interaction or on a phenomenon under investigation unless interactants show that they take them to be relevant. Moerman, who combined ethnographic and
conversation analytic methods (e.g., 1988), observed that the work of producing ethnicity
and identity involves both durable culture and the momentary contingencies of interaction
(1993, p. 85). Sequeira (1993) conducted an ethnographic study of address terms (e.g.,
you, mom, doctor, etc.), which were used in both conventional and unconventional
ways, whereby social participants both reinstantiated their culture and constituted it anew.
Thus, the interplay between representative and constitutive views within LSI research may
be seen to resonate with the interplay among social interactants themselves.
Research on Language and Social Interaction Emphasizes Emic,
Participant Perspectives
Social scientists who study communication and culture sometimes make the distinction
between emic and etic forms of research5. The first (emic) reports the members (or
subjects) view of their communication and community; the second (etic) reports the outsiders (or researchers) view. This distinction has been important within the LSI tradition,
among scholars who avoid imposing their own theorized views on the social phenomena they examine, who strive instead to ground their descriptions and arguments within
the social displays that the participants constitute and at the same time experience. Emic
understandings may be uncovered in a number of ways. Through participant observation,
ethnographers are able to speak and move within a speech community, pursuing depth and
breadth of understanding through extended involvement, literally assuming the perspectives of those that they study. Some ethnographic work is coupled with detailed explications of small moments, whereby the many strands of members understandings may
be both teased apart and brought together within an ethnographic report. For example,
Liberman (1995) explained:
When doing studies of intercultural communication it is important to present to the
reader the looks of the world for the participants, for that is what the participants are
The terms emic and etic were derived from the linguistic words phonemic and phonetic
(Pike, 1966). When a sound difference between two words produces a meaning difference, the linguistic difference is said to be phonemic. When a sound difference between two words does not
produce a meaning difference, the linguistic difference is phonetic. Hence, emic research reports
what is meaningful to the cultural member or participant, and etic research reports what is primarily
meaningful or recognizable to the researcher or outsider.

20 Studies in language and social interaction

attending to and so are the only sociological facts worthy of the name. A faithful
recordingfaithful not to sociological (including ethnomethodological) principles
but to the looks of the world for the participants themselvesnecessitates laying out
the contingent details of interactional events to a precision that readers may find tiresome. Some readers may be presented with more detail than they care to know. But
there are no shortcuts to the lived world of social participants, (p. 119)
Ethnographers and sometimes discourse analysts choose to interview interactants about
their experiences and understandings. Tracy and Muller (1994) studied academic discourse
(e.g., during departmental meetings or colloquia) by recording and transcribing it, but
they also interviewed the participants to more fully ascertain the beliefs, attitudes, and
evaluative expectations (p. 321) that the participants brought to their social interaction.
Moreover, these researchers attended closely to discourse that occurred after a particular
speech event, because it might be especially revealing:
We would expect the beliefs to be most directly visible in peoples aftertalk, the
postmortem analyses of discussion occasions that occur in offices and hallways. That
is, beliefs about what is appropriate (or what is not appropriate) would repeatedly
be asserted, or implicitly assumed, in the criticisms and complaints people make
about actual occasions. In this sense, the language of aftertalk is more similar to the
language of interview-talk, (p. 344)
In response to Tracy and Muller, the journal editors (Sanders & Sigman, 1994) questioned
whether interviewing was an appropriate way to study social interaction. The editorial
comments displayed a preference by many LSI researchers to recover meanings and understandings as they are displayed or oriented to in situ by interactants (e.g., chaps. 14, 21, 22,
this volume). Such focus on how communicators understandings are located in specific
characteristics of talk is sometimes called the message-intrinsic view of communication
(Hopper, 1992b; Mandelbaum, 1991). In short, different notions of meaning and understanding result in different sorts of LSI research, all devoted to emic accounts of social
Some strands of LSI research do not explicitly focus on participants perspectives. In
some discourse analytic approaches, where the goal is to lay out the usage of a conversational object, recovering participants meanings may not be a principal objective. Rather,
format may be seen as somewhat independent of the local situation in which they are
found (e.g., van Dijk, 1993). Other research that does not explicitly focus on participants
perspectives nonetheless addresses issues of how the communication of one participant
impacts another. For instance, research on speech evaluation shows how characteristics of
a speakers speech may result in particular evaluations of that speaker (e.g., chap. 2, this
Research on Language and Social Interaction Focuses on Language in Use
Although different approaches to LSI research may have different agendas, virtually all
approaches regard language in use as central to communication and hence the study of

An overview of language and Social interaction research 21

communication. Ethnographers, discourse analysts, and conversation analysts typically
start from the premise that language is used in orderly ways to enact particular activities, roles, and relationships. For example, Katriels (1993) ethnographic study of Israeli
communication and culture included consideration of lefargena way of speaking that
some cultural members adopt. She wrote:
Whereas parsing out the semantic features of lefargen would in itself be an interesting analytic taskmy main interest lies in reflecting upon the larger contextual
issues associated with the adoption (through lexical borrowing) and spread of the
term as part of Israeli social semantics. I submit that in commending a person as
someone who knows how to express supportspeakers give voice to an ethnosociological model in which social relations and interpersonal patterns of a particular
kind are verbally reified and valorized, (p. 33)
What some form of communication means, Katriels study illustrates, is largely what it
is being used to do. The doing of communication is the means by which social life is constituted, moment to moment and turn by turn. Each LSI approach uses different research
strategies to uncover the orderly ways that language is used. Studies of language attitudes
take it that specific structures or features of language create certain impressions of speakers. Some approaches pay particular attention to how a given activity is undertaken. Others are more interested in why it is done. Nevertheless, a common feature of work within
the LSI rubric is that its focus is on situated language, rather than language as an abstract
commodity (e.g., Searle, 1979).
Recently, LSI researchers have extended notions of language in use to include embodied
processes, recognizing that verbal and nonverbal behaviors necessarily occur together,
providing for their mutual performance and interpretation, making suspect any isolated
examination or treatment of one (Moerman, 1990; Streeck and Knapp, 1992; chaps. 6, 25,
26, 27 and 29, this volume). Several researchers have documented peoples orchestrated
use of what have traditionally been regarded as separate channels of behavior. For example, C. Goodwin (1980) explicated subtle forms of coordination between utterance-initial
restarts and shifts in participants eye gaze (hence attention) toward the speaker. Heath
(1986) studied the organization of speech and body movement (especially shifts in posture
and eye gaze) during medical consultations, whereby patients may direct their doctors
attention toward parts of their bodies that need medical attention. Streeck (1993) showed
how hand gestures may be exposed (i.e., made an object of attention during moments of
interaction) through their coordination with indexical forms of speech (e.g., words such as
this) and eye gaze (which may perform pointing functions). C. Goodwin (1996) examined grammar as interactionally situatednot limited to phenomena within the stream of
speech, but encompassing structures and organization associated with the endogenous
activity systems within which strips of talk are embedded (p. 370). (See also Atkinson,
1984; Bavelas, 1994; Curley, 1998; C. Goodwin, 1986; Goodwin & Goodwin, 1986; Kendon, 1972, 1980, 1987; LeBaron & Streeck, 2000; Schegloff, 1984.) In an analysis of girls
playing hopscotch, C. Goodwin (2000) went beyond the human body to consider the entire
contextual configuration, which included a range of structurally different kinds of sign
phenomena in both the stream of speech and the body, graphic and socially sedimented

22 Studies in language and social interaction

structure in the surround, sequential organization, encompassing activity systems, etc.
(p. 1). In sum, recent LSI research has taken up a more constitutive and holistic view of
language in use.
It is clear that LSI has emerged over the past two decades as a lively and substantive
area within the study of human communication. There is no one principle that consistently
unites or defines LSI research in contrast to other research traditions. The seven points we
have outlined here represent recurrent and interrelated issues within LSI work. Altogether,
they are more central to LSI identity than they are for those working in other traditions,
topics, or methods. Nevertheless, plenty of counterexamples exist within LSI for each point
that we have discussed. Thus, it is most helpful to think of the seven points presented, not
as universally guiding principles within LSI, but as points of ongoing attention or concern.
Call these, if you will, prominent themes in the conversation going on within the area of
Language and Social Interaction.
This volume includes 32 original articles, which are grounded in LSI perspectives, research
questions, and methods, plus 6 short pieces in the final section reflecting on Robert Hoppers teaching and scholarship. A majority of the articles employ conceptual and methodological approaches of ethnomethodological CA. This reflects Robert Hoppers legacy, for
he as much as anyone worked to connect CA with the study of human communication. It
also reflects the prominence of CA research within LSI. Other approaches that have kinship with CA and that are represented in the book include ethnography of communication,
discourse analysis, sociolinguistic studies of language and power, and performance studies.
Most of these report research on naturally occurring interaction. Others make theoretical
or conceptual arguments.
Some edited volumes begin with a conceptual scheme then invite individual articles to
reflect component parts, resulting in a strong thematic coherence. In the present case, the
call for papers invited authors to submit work they thought fitting for a tribute to Robert
Hopper. We did not attempt from the outset to select pieces based on their relevance to a
prearranged scheme. Rather, the organization of the book arose from an inductive process
of sorting the articles by various similarities. We decided on five parts, clustering around
distinct interests and approaches that related in particular ways to LSI as a field and to
Hoppers work.
The first part includes articles we selected to represent major research traditions within
LSI. The second features studies of talk in everyday life, primarily casual discourse. The
third part features studies of institutional discourse, particularly talk concerned with health
and medical settings. The fourth part contains a relatively eclectic group of articles under
the theme of future trajectoriesin various ways, these articles move beyond current
research topics and practices to explore and advocate innovative directions. The fifth part
is a set of personal tributes to Robert Hopper.
There are other ways to group the articles in this book, and it may be useful to the reader
to consider some of these:

An overview of language and Social interaction research 23

Empirical studies: reports of new findings. Beach, Corbin, Craig and Sanusi, Fitch,
Glenn, Goodwin, Heritage, Jefferson, Jones, Lawrence, LeBaron and Koschmann,
Lerner and Zimmerman, Maynard and Frankel, Mandelbaum, Maxwell, Dan Modaff,
John Modaff, Morris, Pomerantz, Sanders, Schegloff, Wrobbel.
Review articles: summarizing areas of research, calling for new directions. Bradac,
Brown, Drew, Gonzalez, Houtkoop-Steenstra, Molloy and Giles, Streeck, Stucky and
Theory pieces: working with or developing theoretical or philosophical positions.
Brown, Drummond, Gonzalez, LeBaron and Koschmann, Molloy and Giles, Streeck,
Stucky and Daughton.
Applied research: dealing directly with practical life problems. Bruder, Maynard and
Frankel, Daniel Modaff, Molloy and Giles, Pomerantz, Wrobbel.
We thank Wayne Beach, Bob Craig, and Karen Tracy, who reviewed an earlier version of
this chapter and provided many helpful suggestions. Bob Craig, Kent Drummond, Kristine
Fitch, Anita Pomerantz, Jrgen Streeck, and Karen Tracy served as outside reviewers for
the articles in this book. Alexander Kozin, Sam Thomas, and Stephanie Poole Martinez,
doctoral students at SIU Carbondale, have been helpful in editorial assistant capacities.
Thanks to the Department of Speech Communication and the College of Liberal Arts at
Southern Illinois University Carbondale for providing office space, equipment, and staff
support. We also thank the Marriott School of Management at Brigham Young University
for providing office space, computer equipment, and support staff for this project. We are
grateful to Rachael Deceuster of BYU for her assistance, as well as Hank Marew and Paul
Kennis for their support. Our appreciation to Linda Bathgate, Marianna Vertullo, and Art
Lizza for their assistance in bringing this project to completion. We are grateful to June
Hymas (Robert Hoppers sister) for providing the photo of Robert. Thanks to Kay Hopper
for ongoing encouragement throughout the development of this book. We were pleased that
Robert Hopper was able to see an earlier version of it shortly before his death in December
1998. It is because of his ongoing influence in the lives of so many people that this book
has come to fruition.
Arnold, C.C., & Bowers, J.W. (Eds.). (1984). Handbook of rhetorical and communication theory.
Boston: Allyn & Bacon.
Atkinson, J.M. (1984). Our masters voices. London: Methuen.
Atkinson, J.M., & Drew, P. (1979). Order in court: The organisation of verbal interaction in judicial settings. London: Macmillan.
Atkinson, J.M., & Heritage, J. (Eds.). (1984). Structures of social action. Cambridge, England:
Cambridge University Press.
Austin, J.L. (1962). How to do things with words. Cambridge, MA: Harvard University Press.
Bateson, G., & Mead, M. (1942). Balinese character. A photographic analysis. New York: New
York Academy of Sciences.

24 Studies in language and social interaction

Bavelas, J.B. (1989). On integrating interactional points of view. The Counseling Psychologist,
17, 466469.
Bavelas, J.B. (1994). Gestures as part of speech: Methodological implications. Research on Language and Social Interaction, 27, 201221.
Bavelas, J.B. (1999, May). Is there a place for social psychology in LSI? Paper presented at the
49th Annual Conference of the International Communication Association, San Francisco.
Bavelas, J.B., Black, A., Chovil, N., & Mullett, J. (1990). Equivocal communication. Newbury
Park, CA: Sage.
Beach, W.A. (1982). Everyday interaction and its practical accomplishment: Progressive developments in ethnomethodological research. Quarterly Journal of Speech, 68, 314327.
Beach, W.A. (1990). Orienting to the phenomenon. In J.A.Anderson (Ed.), Communication yearbook 13 (pp. 216244). Beverly Hills, CA: Sage.
Beach, W.A. (1993). Transitional regularities for casual okay usages. Journal of Pragmatics,
19, 325352.
Beach, W.A. (1994). Orienting to the phenomenon. In F.L.Casmir (Ed.), Building communication
theories: A socio-cultural approach (pp. 133164). Hillsdale, NJ: Erlbaum.
Beach, W.A. (1995a). Maps and diggings. In S.J.Sigman (Ed.), The consequentiality of communication (pp. 223226). Hillsdale, NJ: Erlbaum.
Beach, W.A. (1995b). Preserving and constraining options: Okays and official priorities in
medical interviews. In G.H.Morris & R.Chenail (Eds.), Talk of the clinic: Explorations in the
analysis of medical and therapeutic discourse (pp. 259289). Hillsdale, NJ: Erlbaum.
Beach, W.A. (1996). Conversations about illness: Family preoccupations with bulimia. Mahwah,
NJ: Erlbaum.
Beach, W.A., & Dixson, C.N. (2000). Revealing moments: Formulating understandings of adverse
experiences in a health appraisal interview. Social Science & Medicine, 4(4/5), 121.
Benson, T.W. (Ed.). (1985). Speech communication in the 20th century. Carbondale: Southern
Illinois University Press.
Bitzer, L.F., & Black, E. (Eds.). (1971). The prospect of rhetoric. Englewood Cliffs, NJ: PrenticeHall.
Bochner, A.P., & Ellis, C. (1995). Telling and living: Narrative co-construction and the practices of
interpersonal relationships. In W.Leeds-Hurwitz (Ed.), Social approaches to communication
(pp. 201260). New York: Guilford.
Burke, K. (1973). The philosophy of literary form. Berkeley: University of California Press. (Original work published 1941)
Buttny, R. (1993). Blame-account sequences in couple therapy: Accountability for relational problems. In R.Buttny (Ed.), Social accountability in communication (pp. 6684). Newbury Park,
CA: Sage.
Buttny, R. (1996). Clients and therapists joint construction of the clients problems. Research on
Language and Social Interaction, 29, 125153.
Button, G. (1992). Answers as interactional products: Two sequential practices used in job interviews. In P.Drew & J.Heritage (Eds.), Talk at work (pp. 10134). Cambridge, England: Cambridge University Press.
Carbaugh, D. (1989). Talking American: Cultural discourse on Donahue. Norwood, NJ: Ablex.
Carbaugh, D. (1993). Soul and self: Soviet and American cultures in conversation. Quarterly
Journal of Speech, 79, 182200.
Conquergood, D. (1991). Rethinking ethnography: Towards a critical cultural politics. Communication Monographs, 58, 179194.
Coulthard, M. (1977). An introduction to discourse analysis. London: Longman.

An overview of language and Social interaction research 25

Coutu, L.M. (2000). Communication codes of rationality and spirituality in the discourse of and
about Robert S.McNamaras: In retrospect. Research on language and social interaction,
33, 179212.
Craig, R.T. (1999). Metadiscourse, theory, and practice. Research on Language and Social Interaction, 32, 2129.
Craig, R.T., & Tracy, K. (1983). Conversational coherence: Form, structure, and strategy. Beverly
Hills, CA: Sage.
Crow, B. (1988). Conversational performance and the performance of conversation. The Drama
Review, 32, 2354.
Curley, C. (1998). Teaching the body to make tea within social interaction. Issues in Applied Linguistics, 9, 151178.
de la Zerda, N., & Hopper, R. (1979). Employment interviewers reactions to Mexican-American
speech. Communication Monographs, 46, 126134.
Drew, P., & Heritage, J. (Eds.) (1992). Talk at work: Interaction in institutional settings. Cambridge, England: Cambridge University Press.
Ellis, D. (1995). Fixing communicative meaning: A coherentist theory. Communication Research,
22, 515544.
Ellis, D. (1999a). From language to communication. Mahwah, NJ: Erlbaum.
Ellis, D. (1999b). Research on social interaction and the micro-macro issue. Research on Language
and Social Interaction, 32, 3140.
Ellis, D., & Donohue, W. (Eds.) (1986). Contemporary issues in language and discourse processes.
Hillsdale, NJ: Erlbaum.
Fitch, K. (1994). The issue of selecting objects of analysis in ethnographies of speaking. Research
on Language and Social Interaction, 27, 5193.
Fitch, K. (1998a). Speaking relationally: Culture, communication, and interpersonal connection.
New York: Guilford.
Fitch, K. (1998b). Text and context: A problematic distinction in ethnography. Research on Language and Social Interaction, 31, 91107.
Fitch, K., & Hopper, R. (1983). If you speak Spanish, theyll think youre a German. Journal of
Multilingual and Multicultural Development, 4, 115127.
Geertz, C. (1973). Interpretation of cultures. New York: Basic Books.
Giles, H., & Powesland, P. (1975). Speech style and social evaluation. London: Academic Press.
Glenn, P.J. (1989). Initiating shared laughter in multi-party conversations. Western Journal of
Speech Communication, 53, 127149.
Glenn, P.J. (1992). Current speaker initiation of two-party shared laughter. Research on Language
and Social Interaction, 25, 139162.
Glenn, P.J. (1995). Laughing at and laughing with: Negotiating participant alignments through
conversational laughter. In P. ten Have & G. Psathas (Eds.), Situated order: Studies in the social
organization of talk and embodied activities (pp. 4356). Lanham, MD: University Press of
Goldsmith, D., & Baxter, L. (1996). Constituting relationships in talk: A taxonomy of speech events
in social and personal relationships. Human Communication Research, 23(1), 87114.
Goodwin, C. (1979). The interactive construction of a sentence in natural conversation. In G.
Psathas (Ed.), Everyday language: Studies in ethnomethodology (pp. 97121). New York:
Goodwin, C. (1980). Restarts, pauses, and the achievement of a state of mutual gaze at turn-beginning. Sociological Inquiry, 50, 277302.
Goodwin, C. (1984). Notes on story structure and the organization of participation. In J.M. Atkinson & J. Heritage (Eds.), Structures of social action, (pp. 225246). Cambridge, England: Cambridge University Press.

26 Studies in language and social interaction

Goodwin, C. (1986). Gestures as a resource for the organization of mutual orientation. Semiotica,
62(1/2), 2949.
Goodwin, C. (1995). Co-constructing meaning in conversations with an aphasic man. Research on
Language and Social Interaction, 28, 23360.
Goodwin, C. (1996). Transparent vision. In E. Ochs, E. Schegloff, & S. Thompson (Eds.), Interaction and grammar (pp. 370404). Cambridge, England: Cambridge University Press.
Goodwin, C. (2000). Action and embodiment within situated human interaction. Journal of Pragmatics, 32, 14891522.
Goodwin, C. & Goodwin, M.H. (1986). Gesture and coparticipation in the activity of searching for
a word. Semiotica, 62(1/2), 5175.
Gouran, D.S. (1990). Speech communication after seventy-five years. In G.M. Phillips & J.T. Wood
(Eds.), Speech communication: Essays to commemorate the 75th anniversary of the Speech
Communication Association, (pp. 132). Carbondale: Southern Illinois University Press.
Gray, P.H., & Van Costing, J. (1996). Performance in life and literature. Needham Heights,
MA: Allyn & Bacon.
Gundersen, D.F., & Hopper, R. (1984). Communication and law enforcement. New York: Harper &
Hawes, L.C. (1976). How writing is used in talk: A study of communicative logic-in-use. Quarterly
Journal of Speech, 62, 350360.
Hawes, L.C. (1977). Toward a hermeneutic phenomenology of communication. Communication
Quarterly, 25(3), 3041.
He, A. (1995). Co-constructing institutional identities: The case of student counselees. Research on
Language and Social Interaction, 28, 213231.
Heath, C. (1986). Body movement and speech in medical interaction. Cambridge, England: Cambridge University Press.
Heritage, J. (1984). Garfinkel and ethnomethodology. Cambridge, England: Polity Press.
Hopper, R. (1981a). How to do things without words: The taken-for-granted as speech-action. Communication Quarterly, 29, 22836.
Hopper, R. (1981b). The taken-for-granted. Human Communication Research, 7, 195211.
Hopper, R. (1983). Interpretation as coherence production. In R.T. Craig & K. Tracy (Eds.), Conversational coherence: Form, structure, and strategy. Beverly Hills, CA: Sage Publications.
Hopper, R. (1991). Observer: Steps to an ecology of mind [poem]. Communication Theory, I,
Hopper, R. (1992a). Speech errors and the poetics of conversation. Text and Performance Quarterly, 12, 113124.
Hopper, R. (1992b). Telephone conversation. Bloomington: Indiana University Press.
Hopper, R. (1993a). Conversational dramatism and everyday life performance. Text and Performance Quarterly, 13(2), 181183.
Hopper, R. (Ed.). (1993b). Performance and conversation [Special issue]. Text and Performance
Quarterly, 13(2).
Hopper, R. (1995). Episode trajectory in conversational play. In P. ten Have & G. Psathas (Eds.),
Situated order: Studies in the social organization of talk and embodied activities (pp. 5771).
Baltimore: University Press of America.
Hopper, R. (1996, November). Presidency on the line: Lyndon B. Johnson s first month in office.
Performance (Hopper as Director) at the Eighty-Second Annual Convention of the Speech Communication Association, San Diego.
Hopper, R. (in press). Gendering talk. East Lansing: Michigan State University Press.
Hopper, R., & Doany, N. (1989). Telephone openings and conversational universals: A study in
three languages. International and Intercultural Communication Annual, 13, 148170.

An overview of language and Social interaction research 27

Hopper, R., & Drummond, K. (1990). Emergent goals at a relational turning point: The case of
Gordon and Denise. Journal of Language and Social Psychology, 9, 3965.
Hopper, R. & Drummond, K. (1992). Accomplishing interpersonal relationship: Telephone openings of strangers and intimates. Western Journal of Speech Communication, 3, 185200.
Hopper, R., & Glenn, P.J. (1994). Repetition and play in conversation. In B. Johnstone (Ed.), Perspectives on repetition (Vol. II, pp. 2940). Norwood, NJ: Ablex.
Hopper, R., Koch, S., & Mandelbaum, J. (1986). Conversation analysis methods. In D. Ellis &
W. Donohue (Eds.), Contemporary issues in language and discourse processes (pp. 169200).
Hillsdale, NJ: Erlbaum.
Hopper, R., & LeBaron, C. (1998). How gender creeps into talk. Research on Language and Social
Interaction, 31, 5974.
Hopper, R., Thomason, W.R., & Ward, J. (1993). Demographic questions in telephone calls to the
Cancer Information Service. Southern Journal of Communication, 58, 11527.
Houtkoop-Steenstra, H., & Antaki, C. (1997). Creating happy people by asking yes-no questions.
Research on Language and Social Interaction, 30, 285314.
Hudson, H.H. (1923). The field of rhetoric. Quarterly Journal of Speech, 9, 167180.
Hudson, H.H. (1924). Rhetoric and poetry. Quarterly Journal of Speech, 10, 143154.
Hunt, E.L. (1920). Plato on rhetoric and rhetoricians. Quarterly Journal of Speech, 6, 3556.
Hymes, D. (1978). What is ethnography? Sociolinguistic Working Paper # 45. Austin, TX: Southwest Educational Development Laboratory.
Jackson, S., & Jacobs, S. (1980). Structure of conversational argument: Pragmatic bases for the
enthymeme. Quarterly Journal of Speech, 66, 251265.
Jacoby, S., & Ochs, E. (1995). Co-construction: An introduction. Research on Language and Social
Interaction, 28, 171183.
Jarmon, L. (1996a). An ecology of embodied interaction: Turn-taking and interactional syntax in
face-to-face encounters. Doctoral dissertation, published on CD-ROM, University of Texas at
Jarmon, L. (1996b). Performance as a resource in the practice of conversation analysis. Text and
Performance Quarterly, 16, 336355.
Jefferson, G. (1989). Letter to the editor re: Anita Pomerantz epilogue to the special issue on
sequential organization of conversational activities, Spring 1989. Western Journal of Speech
Communication, 53, 427429.
Jefferson, G. (1994, May). A note on laughter in male-female interaction. Unpublished manuscript.
Jurick, D.M. (1977). The enactment of returning: A naturalistic study of talk. Communication
Quarterly, 25(3), 2129.
Katriel, T. (1993). Lefargen: A study in Israeli semantics of social relations. Research on Language
and Social Interaction, 26, 3153.
Kendon, A. (1972). Some relationships between body motion and speech. In A. W. Seigman &
B.Pope (Eds.), Studies in dyadic communication (pp. 177210). Elmsford, NY: Pergamon.
Kendon, A. (1980). Gesticulation and speech: Two aspects of the process of utterance. In M.R.Key
(Ed.), The relationship of verbal and nonverbal communication (pp. 207228). The Hague,
Netherlands: Mouton.
Kendon, A. (1987). On gesture: Its complementary relationship with speech. In A.W.Siegman &
S.Feldstein (Eds.), Nonverbal behavior and communication. Hillsdale, NJ: Erlbaum.
Kendon, A. (1990). Conducting interaction: Patterns of behavior in focused encounters. Cambridge, England: Cambridge University Press.
Kibler, R.J., & Barker, L.L. (1969). Conceptual frontiers in speechcommunication. New York:
Speech Association of America.

28 Studies in language and social interaction

Knapp, M.L. (1983). Series editors forward. In R.T. Craig & K. Tracy (Eds.), Conversational
coherence: Form, structure, and strategy. Beverly Hills, CA: Sage.
Lawrence, S., Stucky, N., & Hopper, R. (1990). The effects of sex dialects and sex stereotypes on
speech evaluations. Journal of Language and Social Psychology, 9(3), 209224.
LeBaron, C. (1998). Building communication: Architectural gestures and the embodiment of new
ideas. Doctoral dissertation, University of Texas at Austin. (University Microfilms No. AAT98
LeBaron, C., & Hopper, R. (1999, February). The interactive achievement of deception detection in
a therapy group for sex offenders. Paper presented at the Western States Communication Association annual convention, Vancouver, Canada.
LeBaron, C., & Streeck, J. (1997). Built space and the interactional framing of experience during a
murder interrogation. Human Studies, 20, 125.
LeBaron, C., & Streeck, J. (2000). Gestures, knowledge, and the world. In D. McNeill (Ed.), Language and gesture. Cambridge, England: Cambridge University Press.
Leeds-Hurwitz, W. (1992). Forum introduction: Social approaches to interpersonal communication.
Communication Theory, 2(2), 131138.
Liberman, K. (1995). The natural history of some intercultural communication. Research on Language and Social Interaction, 28, 117146.
Litton-Hawes, E.M. (1977). A foundation for the study of everyday talk. Communication Quarterly,
25(3), 211.
Lynch, M., & Bogen, D. (1996). The spectacle of history: Speech, text and memory of the IranContra hearings. Durham, NC: Duke University Press.
Mandelbaum, J. (1987). Couples sharing stories. Communication Quarterly, 35, 144170.
Mandelbaum, J. (1989). Interpersonal activities in conversational storytelling. Western Journal of
Speech Communication, 53, 114126.
Mandelbaum, J. (1991). Beyond mundane reasoning: Conversation analysis and context. Research
on Language and Social Interaction, 24, 331348.
Mandelbaum, J. (1994, November). The search for self in interaction. Paper presented at the annual
meeting of the Speech Communication Association, New Orleans.
McHoul, A. (1990). The organization of repair in classroom talk. Language in Society, 19, 349377.
Metzger, T.R., & Beach, W.A. (1996). Preserving alternative versions: Interactional techniques for
organizing courtroom cross-examination. Communication Research, 23, 749765.
Modaff, J., & Modaff, D. (1999). Technical notes on audio recording. Research on Language and
Social Interaction, 33, 101118.
Moerman, M. (1988). Talking culture: Ethnography and conversation analysis. Philadelphia: University of Pennsylvania Press.
Moerman, M. (1990). Exploring talk and interaction. Research on Language and Social Interaction, 24, 173187.
Moerman, M. (1993). Ariadnes thread and Indras net: Reflections on ethnography, ethnicity, identity, culture, and interaction. Research on Language and Social Interaction, 26, 8598.
Morris, G.H., & Chenail, R.J. (Eds.) (1995). The Talk of the clinic: Explorations in the analysis of
medical and therapeutic discourse. Hillsdale, NJ: Erlbaum.
Morris, G.H., & Hopper, R. (1987). Symbolic action as alignment: A synthesis of rules approaches.
Research on Language and Social Interaction, 21, 131.
Morris, G.H., White, C.H., & Iltis, R. (1994). Well, ordinarily I would, but: Reexamining the
nature of accounts for problematic events. Research on Language and Social Interaction,
27, 123144.
Mulac, A., Bradac, J.J., & Gibbons, P. (2001). Empirical support for the gender-as-culture hypothesis: An intercultural analysis of male/female language differences. Human Communication
Research, 27(1), 121152.

An overview of language and Social interaction research 29

Nofsinger, R.E. (1975). The demand ticket: A conversational device for getting the floor. Speech
Monographs, 42, 19.
Nofsinger, R.E. (1976). On answering questions indirectly: Some rules in the grammar of doing
conversation. Human Communication Research, 2, 172181.
Nofsinger, R.E. (1977). A peek at conversational analysis. Communication Quarterly, 25(3), 1220.
Nofsinger, R.E. (1988). Lets talk about the record: Contending over topic redirection in the
Rather/Bush interview. Research on Language and Social Interaction, 22, 273291.
Ochs, E. (1979). Transcription as theory. In E. Ochs & B. Schieffelin (Eds.), Developmental pragmatics (pp. 4377). New York: Academic Press.
ONeill, J.M. (1915). The professional outlook. Quarterly Journal of Speech, 2, 5263.
Pearce, W.B. (1977). Naturalistic study of communication: Its function and form. Communication
Quarterly, 25(3), 5156.
Perkyl, A. (1995). AIDS counseling: Institutional interaction and clinical practice. Cambridge,
England: Cambridge University Press.
Philipsen, G. (1975). Speaking like a man in Teamsterville: Culture patterns of role enactment in
an urban neighborhood. Quarterly Journal of Speech, 61, 1322.
Philipsen, G. (1977). Linearity of research design in ethnographic studies of speaking. Communication Quarterly, 25(3), 4250.
Philipsen, G. (2000). Permission to speak the discourse of difference: A case study. Research on
Language and Social Interaction, 33, 213234.
Pike, K. (1966). Etic and emic standpoints for the description of behavior. InA. G. Smith (Ed.),
Communication and culture: Readings in the codes of human interaction (pp. 15263). New
York: Holt, Rinehart & Winston.
Pollner, M. (1991) Left of ethnomethodology: The rise and decline of radical reflexivity. American
Sociological Review 56, 370380.
Pomerantz, A. (1989). Epilogue. Western Journal of Speech Communication, 53(2), 242246.
Pomerantz, A., & Fehr, B.J. (1997). Conversation analysis: An approach to the study of social
action as sense making practices. In T.A.van Dijk (Ed.), Discourse as social interaction: Discourse Studies 2A multidisciplinary introduction (pp. 6491). London: Sage.
Pratt, S., & Wieder, D. (1993). The case of saying a few words and talking for another among the
Osage people: Public speaking as an object of ethnography. Research on Language and Social
Interaction, 26, 353408.
Ragan, S.L., & Hopper, R. (1981). Alignment talk in the job interview. Journal of Applied Communication Research, 9, 85103.
Sacks, H. (1984). Notes on methodology. In J.M.Atkinson & J.Heritage (Eds.), Structures of social
action (pp. 2127). Cambridge, England: Cambridge University Press.
Sacks, H. (1992). Lectures on conversation. (2 vols., G.Jefferson, Ed.). Oxford: Blackwell.
Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking for conversation. Language, 50, 696735.
Sanders, R. (1999). The impossibility of a culturally contexted conversation analysis: On simultaneous, distinct types of pragmatic meaning. Research on Language and Social Interaction,
32, 129140.
Sanders, R., & Sigman, S. (1994). An editorial caveat. Research on Language and Social Interaction, 27, 41921.
Schegloff, E.A. (1984). On some gestures relation to talk. In J.M.Atkinson & J.Heritage (Eds.),
Structures of social action (pp. 266296). Cambridge, England: Cambridge University Press.
Schegloff, E.A. (1986). The routine as achievement. Human Studies, 9, 111152.
Schegloff, E.A. (1992a). Introduction. In G. Jefferson (Ed.), Harvey Sacks: Lectures on conversation (pp. ix-xii). Oxford: Blackwell.

30 Studies in language and social interaction

Schegloff, E.A. (1992b). On talk and its institutional occasions. In P.Drew & J. Heritage (Eds.),
Talk at work (pp. 101134). Cambridge, England: Cambridge University Press.
Searle, J. (1979). Expressions and meaning. Cambridge, England: Cambridge University Press.
Sequeira, D. (1993). Personal address as negotiated meaning in an American church communication. Research on Language and Social Interaction, 26, 259285.
Shannon, C., & Weaver, W. (1949). The mathematical theory of communication. Urbana: University of Illinois Press.
Sheldon, A. (1996). Constituting gender through talk in childhood: Conversations in parent-child,
peer, and sibling relationships. Research on Language and Social Interaction, 29, 15.
Simon, C.T. (1951). Speech as science. The Quarterly Journal of Speech, 37, 281293.
Stewart, J. (1995). Language as articulate contact: Toward a post-semiotic philosophy of communication. State University of New York Press: Albany.
Streeck, J. (1984). Embodied contexts, transcontextuals, and the timing of speech acts. Journal of
Pragmatics, 8(1), 113137.
Streeck, J. (1993). Gesture as communication I: Its coordination with gaze and speech. Communication Monographs, 60, 275299.
Streeck, J. (1994). Gesture as communication II: The audience as co-author. Research on Language
and Social Interaction 27, 239267.
Streeck, J. (1996). How to do things with things. Human Studies, 19, 365384.
Streeck, J., & Knapp, M. (1992). The interaction of visual and verbal features in human communication. In F.Poyatos (Ed.), Advances in nonverbal communication (pp. 324). Amsterdam:
Benjamins B.V.
Stringer, J., & Hopper, R. (1997). Generic he in conversation? Quarterly Journal of Speech,
84(2), 209221.
Stucky, N. (1988). Unnatural acts: Performing natural conversation. Literature in Performance,
8(2), 2839.
ten Have, P. (1999). Doing conversation analysis: A practical guide. London: Sage.
Tracy, K. (1994). Boundary drawing in language and social interaction study. Research on Language and Social Interaction, 27, 423425.
Tracy, K. (1995). Action-implicative discourse anlaysis. Journal of Language and Social Psychology, 14, 195215.
Tracy, K. (1997). Interactional trouble in emergency service requests: A problem of frames.
Research on Language and Social Interaction, 30, 315343.
Tracy, K., (1998). Analyzing context: Framing the discussion. Research on Language and Social
Interaction, 31, 128.
Tracy, K., & Muller, N. (1994). Talking about ideas: Academics beliefs about appropriate communicative practices. Research on Language and Social interaction, 27, 319349.
van Dijk, T.A. (1993). Elite discourse and racism. Newbury Park, CA: Sage.
Wieder, D.L. (1999). Ethnomethodology, conversation analysis, microanalysis, and the ethnography
of speaking (EM-CA-MA-ES): Resonances and basic issues. Research on Language and Social
Interaction, 32, 163171.
Winans, J.A. (1915). The need for research. Quarterly Journal of Speech, 1, 1723.
Wittgenstein, L. (1953). Philosophical investigations. (G. Anscombe, Trans.). London: Macmillan.
Woolbert, C.H. (1916). The organization of departments of speech science in universities. Quarterly Journal of Speech, 2, 6477.
Woolbert, C.H. (1917). Suggestions as to methods in research. Quarterly Journal of Speech,
2, 1226.
Zahn, C.J., & Hopper, R. (1985). Measuring language attitudes: The speech evaluation instrument.
Journal of Language and Social Psychology, 4, 113123.
Znaniecki, F. (1934). The method of sociology. New York: Farrar & Rinehart.

Orienting to the Field of Language
and Social Interaction
The first section of this volume includes five articles that represent major research traditions
within the interdisciplinary field of language and social interaction (LSI): sociolinguistics,
conversation analysis, ethnography, discourse analysis, and microethnography. Sociolinguists typically take some aspect of the social dimensions of everyday life (class, ethnicity,
gender, etc.) and pair it with some aspect of spoken language (accent, rate, dialect, etc.),
exploring the extent to which variation in social dimensions correlates with variations in
language use (for an overview, see Fasold, 1990; for a foundational collection see Baugh
& Sherzer, 1984). James Bradacs piece (chap. 2) summarizes work on speech evaluation,
concerned with identifying features of speech that contribute significantly to hearers judgments about speaker credibility, competence, and so forth. Based on his review, Bradac
recommends that future research in this area shift from examining the evaluations hearers
make of speakers under various conditions to more direct studies of perceptions of features
of messages themselves. In this way, speech evaluation research would pay more attention
directly to messages and less to peoples perceptions thereof.
Conversation analysis, in the ethnomethodological tradition, exemplified by John Heritages article (chap. 3), treats audio- and videotaped naturalistic interactions as primary
data. Recordings and transcripts provide resources for constructing detailed accounts of
the activities interactants undertake in and through interaction. Heritages early work on
formulations opened the way for a growing body of research about the organization and
accomplishments of news interviews (Heritage & Watson, 1979). In the current essay, he
examines how interviewers employ questioning to take up particular positions vis a vis
interviewees while managing competing pressures of the interview situation. These pressures include on the one hand taking a somewhat adversarial stance, so as not to operate as a mouthpiece for the interviewee, while on the other hand maintaining a neutral
stance, avoiding making their own opinions available in the way their questions are structured. He shows how news interviewers questions are in fact neutralistic: They have the
appearance of neutrality but actually in various ways are not quite neutral.
Researchers in the ethnography of communication tradition move from thick description of communicative phenomena to identifying underlying speech codes or cultural patterns (for overview, see Saville-Troike, 1989; also Carbaugh, 1990; Fitch & Philipsen,
1995). Kristine Fitch (chap. 4) advocates grounding claims about communication and culture in details of particular interactions. This echoes Michael Moermans (1988) call for
a culturally contexted conversation analysis. Moermans proposal for a union between
ethnography and conversation analysis spawned much discussion, including a special issue

Part I: Orienting to the Field of Language 33

of Research on Language and Social Interaction (1990/1991) edited by Robert Hopper,
to which Fitch contributed an article. In the present piece, Fitch analyzes a transcript of a
family mealtime conversation. It is an everyday life dramatic moment, a child negotiating
a raise in allowance. Fitchs analysis shows that such critical moments in interaction where
culture becomes an issue for participants may provide a resource for analysts to reexamine
this elusive concept.
Although discourse analysis is a term that means many different things (Tracy, 2001),
here we use it to encompass studies that identify particular speech acts and their functions,
focus on coherence as a feature of talk, or trace the actions performed through particular
lexical items that occur commonly in everyday talk. Much contemporary LSI research
(including some studies in the ethnography of communication tradition) reflects grounding
in discourse analytic approaches and specifically in speech act theory. Why do speakers
sometimes choose to say I think that as preface to expressing an opinion? If we assume
that all speech is connected in some way to cognitive activity, then conceivably one could
precede anything one says with I think. What gets marked at moments when speakers
use the verb think? Robert Craig and Alena Sanusi (chap. 5) pursue these issues in videotaped data collected during student group discussions. Think is one of a number of items
by which speakers can indicate standpoint or footing (Goffman, 1983) in relation to the
words they are uttering. The authors show that uses of think include displaying online
thought process to others, marking transition from presentation of canned to spontaneous
material, inviting expression of online thinking from other participants, and displaying
process when sense of process seems to be threatened. Their analysis links to the study of
argument in everyday discourse.
The fifth chapter in this section represents a strain of LSI research we refer to as microethnography. By that term is meant close attention to details of embodied actions as a
means of characterizing emic, participant-grounded ways of enacting and interpreting
meaning in actions. Curtis LeBaron and Timothy Koschmann (chap. 6) examine the coordination of talk, body orientation, gaze, and gesture within small groups working toward
a shared understanding about some issue or topic at hand. For example, when a group of
medical and nursing students read and discuss the symptoms of a hypothetical patient, they
encounter new clinical terms that some members dont understand. By gesturing in relation to their own bodies, informed students explain the new terms to uninformed students,
who then perform the same gestures in the process of coming to understand. Participants
achieve shared understanding (or at least shared understanding is displayed) only after (and
arguably through) gestures repeatedly performed. The authors suggest a socially mediated
and embodied notion of humans coming to understand.
Baugh, J., & Sherzer, J. (Eds.). (1984). Language in use: Readings in sociolinguistics. Englewood
Cliffs, NJ: Prentice Hall.
Carbaugh, D. (Ed.). (1990). Cultural communication and intercultural contact. Hillsdale, NJ: Lawrence Erlbaum Associates.
Fasold, R. (1990). The sociolinguistics of language. Cambridge, MA: Basil Blackwell.
Fitch, K. (1990/1991). A ritual for attempting leave-taking in Colombia. Research on Language and
Social Interaction, 24, 209224.

34 Studies in language and social interac tion

Fitch, K., & Philipsen, G. (1995). Ethnography of speaking. In J.Verschueren, J. Ostman, &
J.Blommaert (Eds.), Handbook of pragmatics (pp. 263269). Amsterdam: Benjamins.
Goffman, E. (1983). Forms of talk. Philadelphia: University of Pennsylvania Press.
Heritage, J., & Watson, D.R. (1979). Formulations as conversational objects. In G.Psathas (Ed.),
Everyday language: Studies in ethnomethodology (pp. 123162). New York: Irvington.
Hopper, R. (Ed.). (1990/1991). Special section: Ethnography and conversation analysis after Talking Culture. Research on Language and Social Interaction, 24, 173387.
Moerman, M. (1988). Talking culture: Ethnography and conversation analysis. Philadelphia: University of Pennsylvania Press.
Saville-Troike, M. (1989). The ethnography of communication: An introduction. Cambridge,
MA: Blackwell.
Tracy, K. (2001). Discourse analysis in communication. In D.Schiffrin, D. Tannen, & H.Hamilton
(Eds.), Handbook of discourse analysis. Maiden: Blackwell.

Extending the Domain of Speech Evaluation:
Message Judgments
James J.Bradac
University of California, Santa Barbara
There is a flourishing research tradition in which the major objects of scrutiny are the kinds
of evaluations that hearers make of speakers and the factors that affect these evaluations.
Factors that have been examined include communication context, for example, formality
of the situation in which a message is delivered (Street & Brady, 1982), and (crucially)
speech style, for example, speaker accent and dialect (Cargile, 1997; Giles & Powesland,
1975; Hopper & de la Zerda, 1979). Evaluation is a basic, even primitive, psychological process, at its core entailing approach-avoidance tendencies and behaviors apparent
in humans, canines, felines, reptiles, and unicellular organisms alike. In humans (at least),
evaluation has a cognitive component in that thought, and more particularly verbalization,
is often, even typically, inextricably bound to the process of acceptance or rejection of
evaluationtriggering stimuli. Speech evaluation research has always exploited this cognitive component by using respondents who are aware of what they are doing, that is,
evaluating speakers, and by asking respondents to make their judgments via verbal, often
semantic-differential-type, scales (Bradac, 1990).
Any stimulus or imagined stimulus can activate the evaluation process. In the arena of
human communication, message recipients can evaluate speakers, their styles of speech,
their messages, specific message features such as arguments, and more specific or idiosyncratic variables, for example, physical aspects of the communication situation and
responses of other message recipients (booing or applause). Evaluations of communication
stimuli or of any stimuli are made at specific times and places; that is, evaluation has a temporal/spatial dimension. This can be important because evaluations can vary systematically
as a function of variations in occasions. For example, message recipients may be relatively
negative when they are fatigued and they may be less attentive to message details, relying on various peripheral or heuristic cues to make judgments of communicators (Petty &
Cacioppo, 1986); or a message that follows an initial message may be evaluated differently
than if it had been presented in the initial position, as a result of perceptual contrast effects
(Bradac, Davies, & Courtright, 1977). Additionally, evaluations have consequences for
both evaluators and the persons (or other organisms) evaluated. A positive evaluation made
of a communicator on one occasion may predispose the evaluator to respond positively on
a second occasion as a result of a commitment effect; the positively evaluated communicator may view the message recipients positive response as a signal to persist.
Thus, in theory, speech evaluation covers the whole communication process, that is,
any communication-related stimulus; it is affected by temporal/spatial variables; and it has

36 Studies in language and social interaction

In practice, research on speech evaluation has had a narrow focus, which has not been
entirely disadvantageous because it has allowed a good deal of concerted effort resulting
in some highly reliable findings. But there is room for expansion. Some of the earliest
pertinent studies were conducted by Lambert and associates who investigated the effects
of language and dialect differences on respondents evaluations of speakers. For example,
in an initial study Frenchand English-speaking monolingual respondents heard audiotapes
of readings of a prose passage recorded in French and English by bilingual speakers and
subsequently rated the speakers in terms of a number of traits, for example, intelligence
and sociability (Lambert, Hodgson, Gardner, & Fillenbaum, 1960). The English guises
received more positive ratings on several traits from both groups of respondents. In a
later study, Arab and Jewish respondents rated speakers who read passages in Arabic and
Hebrew (Lambert, Anisfeld, & YeniKomshian, 1965). In this case, Arab respondents evaluated the Arabic guises more positively, whereas the Hebrew guises were evaluated more
positively by the Jewish respondents, an example of ingroup favoritism. These (and other)
studies were precursors of contemporary language-attitudes research, which has continued
to investigate evaluative consequences of different languages, dialects, and accents (Giles
& Coupland, 1991; see also chap. 3, this volume).
More recently, Mulac and associates have examined the gender-linked language
effect, a relationship among gender, language, and perception that demonstrates that there
is a pervasive tendency for persons to rate womens language as high in Socio-Intellectual
Status and Aesthetic Quality and mens language as high in Dynamism (e.g., Mulac, 1998;
Mulac & Lundell, 1986). The research on this effect has used as an evaluation instrument
the Speech Dialect Attitudinal Scale (SDAS), which reflects the three general dimensions
just mentioned, uncovered through factor analysis (Mulac, 1975). Factor analysis was also
used by Zahn and Hopper (1985) in their attempt to design an instrument that would be
broadly useful in research on speech evaluation (the Speech Evaluation Instrument or SEI).
Employing a variety of communication stimuli and a wide range of evaluative items to
which persons responded following exposure to the stimuli, these researchers obtained three
general factors, which they labeled Superiority, Attractiveness, and Dynamism. Despite the
different communication stimuli and respondents used in constructing the two instruments,
the factor structures of SDAS and SEI are quite similar, although SEI exhibits a relatively
large number of items representing each factor. Both Superiority and Socio-Intellectual
Status include items such as literate/illiterate and white collar/blue collar; Attractiveness
and Aesthetic Quality include sweet/sour and nice/awful; and the two Dynamisms include
strong/weak and active/passive.
The first two factors of both instruments appear to be specific manifestations of the
highly general Evaluation factor obtained by Osgood and associates in their semanticdifferential research on the connotative meanings of a diverse array of concepts (Osgood,
May, & Miron, 1975). The two Dynamism factors appear to combine Osgood, May, and
Mirons Activity and Potency factors. The factor structures of SDAS and SEI are also
similar in some respects to factor structures obtained in early studies of communicator
credibility and attitude change; for example, Authoritativeness, Character, and Dynamism
(or variants thereof, e.g., Competence, Trustworthiness, and Dynamism) are dimensions

Extending the domain of speech evaluation 37

that emerged in factor analytic research and were used subsequently to measure attitudes
toward message sources (McCroskey & Mehrley, 1969).1
The research on speech evaluation measurement (coupled with the work on source
credibility) reveals a strong pattern: Speaker status and attractiveness (in a general sense)
are pervasive evaluative dimensions, as is perceived dynamism. The status/attractiveness
distinction is related to two basic dimensions of interpersonal relationships: power and
solidarity (Brown & Oilman, 1960). Giles and Ryan (1982) noted the importance of the
status and attractiveness dimensions and suggested that when collectivistic concerns are
salient, hearers will perceive speakers in terms of social status and group solidarity; on
the other hand, when individualistic concerns are prominent, hearers will focus on speaker
competence and attractiveness.
The similarity of the dimensions of status/competence and solidarity/attractiveness to
the major dimensions of communicator credibility call attention to the likelihood that in
the many studies of speech evaluation that have used SDAS or SEI (or related items),
respondents were making judgments of or attributions about speakers rather than evaluating speech per se: The speaker was intelligent, likable, active, and so on. It may be much
more usual for persons to judge message sources than to judge messages or message style;
there may be something like a fundamental attribution error in the realm of speech evaluation, where message sources are unduly prominent (Bradac, 1989; Nisbett & Ross, 1980).
But, on the other hand, sometimes message recipients focus on messages per se or features
of messages; in some important communication contexts, message sources are obscure or
unknown, as in the case of reading newspapers, and sometimes even where sources are
known, messages will be examined closely and judged. It may be useful to think about and
investigate message judgments in order to correct an imbalance in our research that has
tipped the scales in favor of message sources.
The bias toward source evaluation may be to some extent a product of the research paradigm exploited in speech evaluation studies. Prototypically, respondents hear one or more
audiotaped messages delivered by a speaker (or speakers) exhibiting a standard or nonstandard dialect or accent and subsequently they complete evaluative scales representing
the dimensions described in the previous section. The content of the messages processed is
bland (sometimes described as neutral) and respondents have little involvement with this
content or with its evaluation. An underlying belief seems to be that the use of neutral
message content will allow respondents to focus on the stylistic variable of interest, which
may be the case, but this use also heightens the attributional prominence of the speaker.
And specific scales representing the dimensions of status/competence and solidarity/attractiveness force a speaker attribution, for example, intelligent, friendly, and trustworthy.

Zahn and Hopper (1985) also noted similarities between Osgood and associates Activity dimension and the Dynamism dimension of speech evaluation, and between source credibility measures
and measures of speech evaluation.

38 Studies in language and social interaction

But in some communication contexts, persons are inclined to scrutinize the substance
of an utterance or utterances, to attend to how an utterance is constructed, or to make a
global judgment of message quality. The specialized context of a public-speaking class is
one example; here evaluators, for example, instructors, examine arguments, message structure, and message style as a consequence of their training. Indeed, Becker (1962) factor
analyzed 10 speech quality rating scales designed for speech classes and found evidence
of three dimensions: content, delivery, and language.2 In this case no speaker factor was
obtained. In less specialized contexts also, given particular constraints, the focus will be on
messages, not speakers.
The meaning of speaker (and the attached attributes of status, etc.) is clear, but the
meaning of message is less obvious, although this is not the place to offer a detailed discussion of definitional issues. Bradac, Hopper, and Wiemann (1989) suggested that messages,
compared to other entities, are high in symbolicity (to use Cronkhites 1986 term), coherence, and intentionality. Additionally, messages often have a point or points that are inferred
by message recipients. The notion of coherence suggests that messages are perceived as
units; they are bundles of significance. But the boundaries of these bundles shift or even
change drastically as message recipients change perspectives and purposes. For example, a
film constitutes a message for many casual viewers, and global judgments of this message
are made: The Negotiator was really good. On the other hand, for film students analyzing
a film closely, particular scenes will constitute messages, even short scenes: That visual
transition is excellentit establishes appropriate expectations. A film analysts significant scene may not even be noticed by the casual viewer. Messages are meaningful units,
the boundaries of which vary across occasions, purposes, and recipients; these units are
sometimes evaluated.
Although message judgments have been neglected compared to judgments of speakers, some exceptions and suggestive possibilities are apparent in communication research
and theory. The message variable argument strength is one example. When exposed to
persuasive messages, persons may attend to arguments that are offered and evaluate them
along a strong-weak dimension. In research on the Elaboration Likelihood Model (ELM)
of persuasion (Gibbons, Busch, & Bradac, 1991; Holtgraves & Lasky, 1999), respondents
evaluations of argument strength have constituted merely a manipulation check of strongand weak-argument messages. These messages are ultimately intersected with high- and
low-relevance conditions, for example, to create differential message processing in respondents (specifically, central and peripheral processing). Evaluations of argument strength
have been subservient to attitudes toward the speakers proposal, the theoretically important measure from the standpoint of ELM, but they demonstrate that arguments are, or at
Beckers (1962) study represents a particular tradition of speech evaluation research with a long
history, namely, research on evaluation in the communication classroom. This is applied research
designed to investigate problems pertaining to evaluating public-speaking effectiveness and effectiveness in group discussion, for example, the reliability and validity of speech ratings scales used
by communication teachers, halo effects in the rating process, and effects of order of presentation
of speeches (Becker, 1953). Most of this research involves judgments reflecting special training and
special conceptions of effective speech, in contrast to the naive social judgments that are the focus
of this essay, so this research tradition will not be discussed at length.

Extending the domain of speech evaluation 39

least can be, evaluated along the dimension of strength; strong arguments are better than
weak ones for most purposes.
Judgments of argument strength pertain to a specific message feature. Other message
judgments are globalgeneral impressions of a message as a whole. In special cases,
global judgments are made of clusters of messages: I thought the debate was uninteresting. Referring to a specific class of messages, Dillard, Wilson, Tusing, and Kinney
(1997) suggested that social actors naturally evaluate influence messages in terms of three
distinct and conceptually orthogonal features: explicitness, dominance, and argument (p.
317). Explicitness refers to the directness of the influence attempt, dominance refers to the
level of control attempted, and argument refers to the extent to which reasons are offered
in support of a proposal. Dillard et al. further suggested that [t]hese three constructs lie
midway between the relatively microscopic objective features of messages (such as word
choice) and more macroscopic evaluations of messages (such as judgments of politeness)
(p. 303). Perhaps explicitness, dominance, and argument are best conceptualized as qualities of influence messages that social actors naturally perceive, rather than as evaluations
that naturally occur, because it is difficult to think of these qualities in terms of an unambiguous good-bad criterion that is a necessary feature of all evaluations; indeed, at one
point, Dillard et al. referred to the three qualities as percepts (p. 320). It is useful to
distinguish between perception and evaluation in research on message processing (Street
& Hopper, 1982). On the other hand, politeness, which is Dillard et al.s criterion variable
in the research that they reported, is clearly evaluative. Dillard et al. found that there was
a negative association between perceived dominance in influence messages and judgments
of politeness; by contrast, there was a positive association between judgments of politeness
and perceptions of both explicitness and argument.
A dimension of message perception that is closely related to the dominance dimension
just discussed is power. There is evidence that messages exhibiting hedges, hesitations,
and tag questions are rated as relatively low in power (Bradac & Mulac, 1984; Holtgraves
& Lasky, 1999). It may be that a perception of high- or low-power messages is more
accurately described as an evaluation, because in the particular case of linguistic power
it appears that connotations of good-bad are inevitably attached; a good deal of research
on high- and low-power styles indicates that what is powerless is bad. In any case, a
high-power style appears to produce judgments of high communicator competence and
attractiveness, whereas Dillard et als dominant messages produced judgments of low message politeness. Is this contradictory? Or is it possible for a communicator to be judged
as attractive when delivering an impolite message? Probably yes to the latter. It would be
useful to obtain politeness ratings of high- and low-power styles in future research; it may
be that a high-power style will trigger perceptions of high dominance, which will reduce
politeness ratings.
Kellermann and associates have proposed two additional types of message judgments
that revolve around the meaning of messages. The first is coherence judgments, which
occur when activated knowledge structures are consonant with the perceived nature of the
discourse processed by message recipients (Kellermann & Sleight, 1989, p. 122); coherence is an evaluative judgment of meaningfulness of discourse (p. 105). In most contexts
most people expect communicators to make sense, to produce meaningful utterances, so
a judgment that takes the form That message was coherent is probably rare, occurring

40 Studies in language and social interaction

mainly when for whatever reasons persons expect an incoherent message. A judgment
of incoherence probably occurs more frequently because of the pervasive expectation of
coherence; incoherence is the marked case, so it will be noticed and evaluated negatively.
Also there are degrees of incoherence: The last part of the film was baffling or The
statement wasnt completely clear. Another type of message judgment is informativeness,
which is a judgment that is concerned with the importance or relevance of either the parts
or whole of a message (Kellermann & Lim, 1989, p. 118). It is also probably the case that
messages that are judged to be informative are substantively novel: I never would have
guessed that. So, the perceptions of importance, relevance, and novelty may contribute to
evaluations of informativeness. There may be an inverse relationship between perceptions
of novelty and judgments of coherence such that something that is radically unfamiliar may
make little sense; a particular message may be judged to be quite informative and relatively
incoherent (perhaps like this essay).
Another message judgment that seems to occur fairly frequently can be labeled stimulation-value. Some messages are arousing or exciting, whereas others are soothing or
dull. There appear to be positively and negatively valenced high arousal (exciting and
grating, respectively) and positively and negatively valenced low arousal as well (relaxing and boring), so this is a relatively complex message judgment (cf. J.K.Burgoon,
Kelley, Newton, & Keeley-Dyreson, 1989). Stimulation-value judgments are not bound to
messages uniquely in the way that politeness judgments, for example, appear to be. Both
roller coaster rides and action films are potentially exciting. Also the judgment made by a
given message recipient will depend on her preexposure level of arousal; when preexposure arousal is high, a stimulating message may be evaluated more negatively than when
preexposure arousal is low. This message judgment appears to be more clearly dependent on the cognitive and emotional states of message recipients than are the judgments
discussed previously.
As a final example, some messages may be evaluated along a sociability dimension.
This judgment corresponds to the speaker attribution of solidarity, but it is a messagecentered evaluation: That was a kind remark or That was a friendly overture. An interesting possibility is that a speaker judged to be generally low in solidarity may produce
a message judged to be extremely high in sociability. Such an occurrence may cause the
message recipient to reassess the judgment of low speaker sociability or to search for an
explanation for the discrepancy between the message judgment and the judgment of the
speaker. Probably more typically, at least in first-impression situations, a highly sociable
message will lead directly to a judgment of high speaker solidarity, unless this message is
perceived as manipulative or patronizing (Giles, Fox, & Smith, 1993).
Thus, a given supermessage may be evaluated as polite, powerful, coherent, informative, stimulating, and sociable, whereas its dark opposite may be judged to be impolite,
powerless, incoherent, uninformative, boring, and unsociable. Probably in most situations
most message recipients would approach the former message and avoid the latter, although
no doubt across the globe there are scattered individuals who generally prefer impoliteness,
powerlessness, incoherence, a lack of information, boredom, and unfriendliness. Informativeness and coherence seem to be at the base of a message-judgment hierarchy because
they are pertinent to all messages; stimulation-value is at the next level because it is pertinent to many types of messages; and at the level above that are politeness, power, and

Extending the domain of speech evaluation 41

sociability, which appear to be relevant to specific message types. There are certainly other
levels, and within levels there are other judgment types.
Messages are likely to be the objects of primary scrutiny when message recipients are
involved with message content and when this content is relevant to them, at least in the
case of persuasive messages (Petty & Cacioppo, 1986). Many specific variables are associated with relevance and involvement, for example, decisions hinging on message content
or need to transmit the content to another person. It is worth noting that relevance and
involvement have not been manipulated in studies of speech evaluation; in fact, relevance
and involvement typically have been low, which has probably led evaluators to focus more
on speakers than on messages. It is also the case that specific roles will predispose persons
to focus on messages. To give a specialized example, film critics are required to analyze
films and to make global judgments. Noncritics, that is, naive viewers, commonly eschew
analysis of films but easily offer quick judgments: It was exciting or It didnt make
sense. The film-going experience requires message evaluation, although this evaluation
may be implicit and may remain unexpressed. Particular occasions will also precipitate
message evaluation. An interesting case in point is President William Jefferson Clintons
speech to the nation about a sexual relationship, which was delivered at the time of this
writing (August 17, 1998). Many nonspecialists offered opinions about his message, its
content, and its presentation: It wasnt (was) satisfying, It was too short, It was too
general, It had a blurred focus, and so on. On this occasion, there was a great deal of
interest in what the speaker would say; opinions about the speaker himself, his power
and his character, were already well formed. Finally, sometimes particular message features will cause message recipients to focus on messages, specifically, features that violate
expectations (M.Burgoon, 1990) or features that are marked, for example, a low-power
language style (Gibbons et al., 1991).
So, message evaluation, that is, the process and structure of naive judgment, has been
seriously neglected, compared to speaker evaluation. Particular types of evaluations, for
example, coherence and informativeness judgments, barely have been investigated. The
interaction between message evaluations and message genres remains essentially unexplored: Particular dimensions of evaluation are likely to be especially, even uniquely,
relevant to specific genres or types of messages.
Apart from the issue of message judgment, it was mentioned at the beginning of this
essay that evaluations of all sorts are situated temporally and spatially, but this fact has
been largely ignored in speech evaluation research; the time and place of evaluation has
been imposed upon respondents for the purpose of coordinating an experiment. It would be
very useful at this point to discover where and when speech evaluation naturally occurs
outside of the laboratoryand to discover how different settings and temporal factors
affect evaluations of speakers and messages (cf. Duck, Rutt, Hurst, & Strejc, 1991). Additionally, as suggested earlier, evaluations have consequences for evaluators and persons
evaluated, but in the typical speech evaluation experiment, evaluations are made in a vacuum. Would respondents who evaluate a speaker as high in power, status, and competence

42 Studies in language and social interaction

choose not to interact with her? Under what circumstances will respondents self-esteem
affect this interaction decision? How will speakers react if they are evaluated as high in
competence but low in attractiveness? How will they react if their messages are judged
to be informative but incoherent? Investigating temporal/spatial factors in speech evaluation, along with evaluative consequences for communicators and message recipients, could
extend the domain of speech evaluation research in important ways. And, almost certainly,
the speech evaluation research domain will be extended fruitfully by shifting attention to
the dimensions underlying message judgments, to the interactions between dimensions
and message genres, and to the many communication variables that systematically affect
dimension-relevant evaluations.
Becker, S.L. (1953). The ordinal position effect. Quarterly Journal of Speech, 39, 217219.
Becker, S.L. (1962). The rating of speeches: Scale independence. Speech Monographs, 29, 3844.
Bradac, J.J. (1989). On coherence judgments and their multiple causes: A view from the messagevariable paradigm. In J.A. Anderson (Ed.), Communication yearbook 12 (pp. 130145). Newbury Park, CA: Sage.
Bradac, J.J. (1990). Language attitudes and impression formation. In H. Giles & W.P. Robinson
(Eds.), Handbook of language and social psychology (pp. 387412). Chichester, England:
Bradac, J.J., Davies, R.A., & Courtright, J.A. (1977). The role of prior message context in judgments of high- and low-diversity messages. Language and Speech, 20, 295307.
Bradac, J.J., Hopper, R., & Wiemann, J.M. (1989). Message effects: Retrospect and prospect. In
J.J. Bradac (Ed.), Message effects in communication science (pp. 294317). Newbury Park, CA:
Bradac, J.J., & Mulac, A. (1984). A molecular view of powerful and powerless speech styles: Attributional consequences of specific language features and communicator intentions. Communication Monographs, 51, 307319.
Brown, R., & Oilman, A. (1960). The pronouns of power and solidarity. In T.A. Sebeok (Ed.), Style
in language (pp. 253276). New York: Wiley.
Burgoon, J.K., Kelley, D.L., Newton, D.A., & Keeley-Dyreson, M.P. (1989). The nature of arousal
and nonverbal indices. Human Communication Research, 16, 217255.
Burgoon, M. (1990). Language and social influence. In H. Giles & W.P. Robinson (Eds.), Handbook of language and social psychology_(pp. 5172). Chichester, England: Wiley.
Cargile, A.C. (1997). Attitudes toward Chinese-accented speech: An investigation in two contexts.
Journal of Language and Social Psychology, 16, 434443.
Cronkhite, G. (1986). On the focus, scope, and coherence of the study of human symbolic activity.
Quarterly Journal of Speech, 72, 231246.
Dillard, J.P., Wilson, S.R., Tusing, K.J., & Kinney, T.A. (1997). Politeness judgments in personal
relationships. Journal of Language and Social Psychology, 16, 297325.
Duck, S., Rutt, D.J., Hurst, M.H., & Strejc, H. (1991). Some evident truths about conversations
in everyday relationships: All communications are not created equal. Human Communication
Research, 18, 228267.
Gibbons, P., Busch, J., & Bradac, J.J. (1991). Powerful versus powerless language: Consequences
for persuasion, impression formation, and cognitive response. Journal of Language and Social
Psychology, 10, 115133.
Giles, H., & Coupland, N. (1991). Language: Contexts and consequences. Buckingham, England:
Open University Press.

Extending the domain of speech evaluation 43

Giles, H., Fox, S., & Smith, E. (1993). Patronizing the elderly: Intergenerational evaluations.
Research on Language and Social Interaction, 2, 129150.
Giles, H., & Powesland, P.F. (1975). Speech style and social evaluation. London: Academic Press.
Giles, H., & Ryan, E.B. (1982). Prolegomena for developing a social psychological theory of language attitudes. In E.B.Ryan & H.Giles (Eds.), Attitudes towards language variation: Social
and applied contexts (pp. 208223). London: Edward Arnold.
Holtgraves, T., & Lasky, B. (1999). Linguistic power and persuasion. Journal of Language and
Social Psychology, 18, 196205.
Hopper, R. & de la Zerda, N. (1979). Employment interviewers reactions to Mexican American
speech. Communication Monographs, 46, 126134.
Kellermann, K., & Lim, T. (1989). Inference-generating knowledge structures in message processing. In J.J.Bradac (Ed.), Message effects in communication science (pp. 102128). Newbury
Park, CA: Sage.
Kellermann, K., & Sleight, C. (1989). Coherence: A meaningful adhesive for discourse. In
J.A.Anderson (Ed.), Communication yearbook 12 (pp. 95129). Newbury Park, CA: Sage.
Lambert, W.E., Anisfeld, M., & Yeni-Komshian, G. (1965). Evaluational reactions of Jewish and
Arab adolescents to dialect and language variations. Journal of Personality and Social Psychology, 2, 8490.
Lambert, W.E., Hodgson, R., Gardner, R.C., & Fillenbaum, S. (1960). Evaluational reactions to
spoken languages. Journal of Abnormal and Social Psychology, 60, 4451.
McCroskey, J.C., & Mehrley, R.S. (1969). The effects of disorganization and nonfluency on attitude
change and source credibility. Speech Monographs, 36, 1321.
Mulac, A. (1975). Evaluation of the speech dialect attitudinal scale. Speech Monographs, 42,
Mulac, A. (1998). The gender-linked language effect: Do language differences really make a difference? In D. Canary & K. Dindia (Eds.), Sex differences and similarities in communication
(pp. 127153). Mahwah, NJ: Lawrence Erlbaum Associates.
Mulac, A., & Lundell, T.L. (1986). Linguistic contributors to the genderlinked language effect.
Journal of Language and Social Psychology, 5, 81101.
Nisbett, R.E., & Ross, L. (1980). Human inference: Strategies and shortcomings of social judgment. Englewood Cliffs, NJ: Prentice-Hall.
Osgood, C., May, W., & Miron, M. (1975). Cross cultural universals of affective meaning. Urbana:
University of Illinois Press.
Petty, R.E., & Cacioppo, J.T. (1986). Communication and persuasion: Central and peripheral
routes to attitude change. New York: Springer-Verlag.
Street, R.L., Jr., & Brady, R.M. (1982). Speech rate acceptance ranges as a function of evaluative
domain, listener speech rate, and communication context. Communication Monographs,
49, 290308.
Street, R.L., Jr., & Hopper, R. (1982). A model of speech style evaluation. In E. B.Ryan & H.Giles
(Eds.), Attitudes towards language variation: Social and applied contexts (pp. 175188). London: Edward Arnold.
Zahn, C.J., & Hopper, R. (1985). Measuring language attitudes: The Speech Evaluation Instrument.
Journal of Language and Social Psychology, 4, 113123.

Designing Questions and Setting Agendas
in the News Interview
John Heritage
In news interviews, unlike speeches, lectures or other forms of monologic communication, public figures overwhelmingly give information and express opinions in response to
journalists questions. The news content that results is thus a joint construction, whether
collaborative or conflictual, that emerges from the confluence of the questions journalists
choose to put and the responses that those questions engender.1
For this reason, questioning is central to the practice of news interviewing, and skill in
question design is at the heart of the interviewers (IRs) craft. The limits of questioning
play a significant part in defining the parameters of the permissible in mass media content,
and innovations in question design often embody efforts to redefine these parameters. In
designing questions, IRs ordinarily attempt to strike a balance between two competing
journalistic norms. On the one hand, IRs are expected to be impartial, objective, unbiased,
and disinterested in their questioning of public figures. They are expected to have respect
for the facts and the perspectives that interviewees (IEs) communicate, and to work to
bring these into the public domain. On the other hand, IRs also subscribe to a norm of
adversarialness. They should actively challenge their sources, rather than being simply
mouthpieces or ciphers for them. This second norm is one that pushes IRs not to let the
interview be a kind of platform or soapbox from which public figures can get away with
their own spin on events.
In part, the management of the tension between these two norms is handled by questioning itself. Questioning is conventionally understood as an action that does not take
up a substantive positioninvolving either agreement or disagreementvis-a-vis the IE.
For this reason, IRs work hard to package their actions as questions, and may invoke this
packaging to defeat IE claims that they are pursuing some kind of personal or institutional agenda. In the following case, for example, ABC journalist Sam Donaldson defends
himself against such a claim in just this way
(1) [U.S. ABC This Week: October 1989: Barman]



Isnt it a fact, Mr. Darman, that the taxpayers


pay more in interest than if they just paid it out

of general revenues?


No, not necessarily. Thats a technical


Schudson (1994) gives a nuanced account of the emergence of the news interview as a medium of
journalistic practice, Clayman and Heritage (2002a) describe its development in British and American broadcasting.

Designing questions and setting agendas 45




Its not a-- may I, sir? Its not a technical


argument. Isnt it a fact?


No, its definitely not a fact. Because first of

all, twenty billion of the fifty billion is being

handled in just the way you wantthrough


treasury financing. The remaining




Im just asking you a question. Im not expressing



my personal views.



I understand.

This example is from an interview about alternative ways of financing losses from collapsed savings and loans companies, and the IERichard Darmanis a treasury official
in the Bush administration. Faced with insistent questioning from Donaldson (Lines 13,
56), he responds that twenty billion of the fifty billion is being handled in just the way
you wantthrough treasury financing. (Lines 810), thus implying that Donaldson is
advocating a specific policy preference. It is just this departure from journalistic norms
that Donaldson is quick to rebut at Lines 1112 with Im just asking you a question. Im
not expressing my personal views. Darman responds with an acceptance of this account
(Line 13).
As this example illustrates, questioning is a vehicle by which broadcast journalists can
sustain a neutralistic stance vis a vis interviewees, and defend themselves against charges
that they have overstepped their role as elicitors of information Clayman, 1988; Heritage
& Greatbatch, 1991).2
However, as the term neutralistic suggests, news interview questioning is not, and cannot be, strictly neutral. Because questions unavoidably encode attitudes and points of view
(Harris, 1986), IRs must still design their questions to strike a balance between the journalistic norms of impartiality and adversarialness. The particular balance that is achieved
between these two norms can be a distinctive, and even defining, characteristic of particular interviewing styles. In turn, distinctive styles of question design are an important element of the public personae of IRs ranging from Walter Cronkite to Ted Koppel to Larry
King in the United States, or Sir Robin Day to Jeremy Paxman to Jimmy Young in Britain.
The significance of question design as a signature feature extends from IRs as individuals to the news programs of which they are a part (e.g., PBSs Newshour vs. ABCs
Nightline), and ultimately to whole periods that are characterized by what may be termed
dominant styles of interviewing. This chapter discusses question design in the news interview, and addresses some of the resources through which IRs manage the balance between
impartiality and adversarialness in this context.
The term neutralistic is used in parallel with Robinson and Sheehans (1983:34) distinction
between objective and objectivistic news reporting. In their usage, objectivistic describes a
manner or style of reporting, while the term objective is treated in the conventional sense of a
judgement about balance, truthfulness and the absence of bias in the news.

46 Studies in language and social interaction

Consider the following 1951 interview of British Prime Minister Clement Attlee, who has
called a general election and just returned to London to begin his election -campaign. The
interview is conducted at the London rail station where Mr. Attlee has just arrived. The
following transcript represents the complete interview:
(2) [UK BBC Interview with Clement Attlee (British Prime Minister 19451951)]
Good mor:ning Mister A:ttlee,=We hope (.)

had a good journey,


Ye::s excellent, h

Can you:- (.) now youre ba:ck hhh having cut

short your: lecture tour::. (.) tell us [something


of how you- (0.2) vie::w the election prospects?


Oh we shall go in tgive them a good fi:ght,

very good, (0.4) very good cha:nce of >winning, =We


shall go in confidently,=We always do,<


U:::h And- (.) on wha:t will Labour take its

We:ll that we shll be announcing shortly.

What are your immediate pla:ns: Mister

immediate plans are <tgo do:wn> to a committee


tdeci:de on just that thing, .hhh (.) >soons I


can get away from here.<


hheh .hh
Uhm, hh (.) Anything else you would> ca:re

about (.) th coming election.



Designing questions and setting agendas 47




Uhm, (0.4) Uhm, ((end of interview segment))

The IRs questioning in this interview has a number of noticeable features:

First, his questions are all very open. Questions like Can youtell us something of
how you view the election prospects (Lines 69) and On what will Labour take its
stand (Line 15) permit the IE enormous latitude in developing responses.
Second, the questions are not the prefaced, multi-sentence questions that are common
today, where prefatory statements are used to establish context and background for
what follows. Rather Attlee is presented with simple inquiries that treat the immediate context of the interviewthe impending electionas the only thing necessary to
understand the questions that follow.
Third, the IR does not materially shift topic. The context of the interview is the Prime
Ministers arrival in London to strategize for national elections, and the IR does not
diverge from that. There are no shifts to discuss Britains relations with foreign powers, or disagreements within the Labour Party. The IRs questions remain tied to the
immediate context of the interviewthe election and Mr. Attlees view of it.
Fourth, even though Attlee gives noncommittal, if not downright evasive, replies to his
questions, the IR makes no attempt to pursue more specific responses. Rather he simply accepts the response that he is given and moves on.
Fifth, the design of the questions is fundamentally deferential to the power and status
of the Prime Minister. This is expressed through conventional indirectness (Brown
& Levinson, 1987; Clayman & Heritage, 2002b). Questions like Can youtell
us something of how you view the election prospects (Lines 69) and Anything
else you would care to say about the coming election (Lines 2627) evidently treat
Attlees responses as optional rather than obligatory. They indicate that Attlee will not
be pressed by this IR if he does not care to respond.
Finally, the deferential style embodied in the IRs questions is reciprocated in Attlees
brusquely, noncommittal responses. Attlee is not merely unafraid to decline the questions, he clearly feels under no obligation to respond to them. Indeed, he is quite happy
to imply (at Lines 2223) that the interview itself is preventing him from getting on
with more important election matters. No modern politician entering an election campaign today would dream of addressing an IR (or the voting public) in this way.3
Interviews like this one are a valuable historical benchmark. They tell us about the
extent to which present day broadcast interviews differ from those of the past. And they
are evidence of quite different relationships between broadcasters and politicians than exist
today. The modern political interview differs from this one in every major respect. This
chapter examines some of the ways in which IEs struggle with IRs over the terrain that is
Attlee could afford to adopt this stance because the audience for this broadcast was miniscule: less
than one per cent of the British public had access to a television set in 1951.

48 Studies in language and social interaction

constructed through news interview questioning. We begin with an exploration of some of
the basic features and objectives of question design in the news interview.
News interview questions are often very subtle and complex constructions. They express
particular aspects of the public roles of IR and IE, and they can index elements of the
personal identities of both (Roth, 1998a). They can be primarily geared to the concerns
and preoccupations of either the questioner, the answerer, or the overhearing audience
members, or all three of these to varying degrees. They can embody complex grammatical
and rhetorical constructions to engage in the widest range of tasks designed to support or
undermine the positions of public figures on issues of the moment. It is obvious, therefore,
that they can be examined from many different angles.
We can begin by observing that, at the minimum, IRs questions have the following features: First, they establish particular agendas for IE responses. Second, they tend
to embody presuppositions and/or assert propositions about various aspects of the IEs
actions, interests, opinions, and the social and political context of these. Third, they often
incorporate preferences, that is, they are designed so as to invite or favor one type of
answer over another. Similarly, IEs can formulate their responses in ways that accept or
resist (or reject altogether) any or all of these. Thus IEs responses engage (or decline to
engage) the agenda set by IRs questions, confirm (or disconfirm) its presuppositions, and
align (or disalign) with its preferences. These possibilities are displayed in Table 1:
Table 1: Dimensions of Questioning and Answering
IR Questions:

IE Responses:

Set Agendas:
(i) Topical agendas
(ii) Action agendas
Embody presuppositions
Incorporate preferences

Engage/Decline to engage:
(i) Topical agendas
(ii) Action agendas
Confirm/Disconfirm presuppositions
Align/Disalign with preferences

These three dimensions are fundamental and inexorably relevant characteristics of question design and production.4 Because it is not possible to avoid them, IRs questions can
only select between different possibilities for agenda setting, presuppositional content, and
preference design. These selections are crucial for the work that questions do, the nature of
the interview that is built through them, and the IR and news show identity that is sustained
by these means.

See Boyd and Heritage (in press) for a parallel discussion of these issues in relation to questioning
in medical interviews.

Designing questions and setting agendas 49

These three dimensions of question design are made more complex in prefaced questions.
These are questions that are preceded by one or more statements (Clayman, 1988; Heritage
& Greatbatch, 1991; Heritage & Roth, 1995). These prefaces were quite absent in Example
2, but they are very much a part of the modern news interview. Their manifest function is
often to contextualize and provide relevance for the questions that follow, sometimes for
the IE and often for the news audience. Example 3 is a clear case of this:
(3) [U.S. ABC Nightline: 22nd July 1985: South Africa]
1 IR:
P-> .hh Two- two members of your organization

Supposedly arrested today:

Q-> dyou feel in some danger when you go back

Here the prefatory statement (Lines 12) establishes a context that gives meaning and
point to the subsequent question, which otherwise might seem to come out of the blue, and
indeed be incomprehensible for many members of the news audience.
A prime difference between simple and prefaced questions concerns the degree to which
they embody initiative in establishing a context for the question to follow (Clayman &
Heritage, 2002b). Most simple questions draw on resources from the prior answer to provide for their relevance and intelligibility. The following is a case in point. Here a British
Labour politician with overall responsibility for his partys defense policy explains why he
walked out of the defense debate at his partys annual convention. In his first turn, he says
that he was angry because the person chairing the debate did not call him to speak and
allow him to reply to attacks on him. The IR then asks him whether the chairs action was
intentional (Line 8):
(4) [UK BBC TV: Nationwide: 30 September 1981: Labour Party Conference]
Well I walked out because I was ang:ry at not being

called by the chairman after two personal attacks


.hhh had been launched on me from the rostrum.=I


dont complain about those attacks. .hhh But I


think that any fair chairman would have given me an


opportunity of replying to them.


IR: -> Was it intentional not to call you?
.hhh Well i- (.) I dont think it was mali::gn,=but

It was intentional in the sense that he he referred


at the e:nd to the fact that I had put in a note


asking to be calle:d. .hh and couldnt be called.=


=So it obviously was intentional.=It wasnt .hh an


o:versight on his part.

This simple follow up question raises something that is implicit in the IEs previous answer,
especially the IEs reference to what a fair chairman would have done (Lines 56), and

50 Studies in language and social interaction

it is explicit in introducing the issue of the chairs intentions as a relevant matter to be
addressed by the IE in his next turn at talk. It does not require prefatory remarks because
it transparently draws on, and projects an extension to, the IEs immediately previous talk.
The IE responds by devoting his next turn to asserting the intentional nature of the chairs
action, beginning with a pre-emptive denial that its intent was malign.
However, journalists may often find themselves in circumstances where a simple
follow-up question that explores some dimension of a prior answer is quite undesirable.
Under these circumstances, prefaces are an essential resource for resetting the context for
the question to come. In the following case, for example, a journalist uses a prefaced question design to put a topical issue raised by the IE (about blacks against blacks violence in
South Africa) on hold. This needs prefatory statements:
(5) [U.S. ABC Nightline: 22 July 1985: South Africa]
.hhhh The: urgent an pressing: need hh the: ()
tch .hhh uh immediate one: is to stop

violence. () violence perpetrated by blacks upon


blacks. () This is what we have to end (.) to get


to: uh situation .hhh where we can start ()


talking. Where we can start n uh peaceful man6


() to haff (.) political dialogue.

tch .hhh Arright lemme get tuh that blacks against
IR: 1->
blacks question in uh minute but first lemme ask

you it seems to me nobody dispu:tes .hhh thet

power in south africa (0.2) is with thuh white


goverment. .hh An it seems to me that within thuh
rule of law: that could be do:ne.

Why duh laws haftuh be suspended in order to


thuh violence.
.hhhh Uhm (.) seems to me: uh- eh and always
been: a balance between freedom () an disor17



[35 lines of talk omitted]


=Arright lemme talk about this question then fer a

IR: 3->

moment of violence (.) of blacks against blacks.(.)
We live here in thuh United States......

In addition to using prefatory statements (1->) to place the IEs immediately preceding
statements on hold, the IR deploys additional statements (2->) to set up a question about
the necessity of suspending the rule of law in South Africa, and then further statements
(3->) to return to the blacks against blacks issue raised earlier by the IE.
Journalists may also use prefatory statements, not merely to give background for a
question (as in Example 3), but to provide a motivational context for the IEs answer. In
the following case, discussed in Roth (1998a), dealing with proposals to arm the British

Designing questions and setting agendas 51

police, the personal experience of the IEa policeman who was shot by a criminal while
unarmedis invoked to convey to the audience that the question has a special relevance
for him.
(6) [UK BBC TV Newsnight: 21 Oct 1993]
1 IR: You as I say have been shot yourself in
in thuh line of duty, ahm Lets just look at
question of arming thuh police first of all.
Is it your view that the police should now be
6 IE:
.hhh But definitely. .hhh Ahm we w- (.) have
7 rights as a society to expect young men to enter
8 situations....

Here the question preface provides that the IEs experience of being shot is the presumptive
foundation of his perspective in answering it, and may privilege that experience as having
a special weight and significance for the audiences understanding and evaluation of his
response to the question.
In sum, prefaced question give IRs room to maneuver. Whereas simple questions leave
the IEs last response as the context for the next question, prefaced questions allow IRs to
escape from this constraint and construct a context of their own choosing for the question
they are about to put into play. The shift toward the use of complex question designs has
been relatively marked in both the United States and the United Kingdom from the 1950s
to the present, and it embodies a real growth in the scope, power, and autonomy of IR
questioning. Additionally, as we see later, the manifest function of prefaced questions
providing context for the subsequent question to the news audienceprovides justification
and cover for very much more hostile and aggressive questioning strategies than were
dreamed of in the early days of news interviewing.
Questions Set Agendas
The claim that IR questions set agendas for IEs involves three features of their design that
constrain IEs.
First, questions set agendas by identifying a specific topical domain as the appropriate or relevant domain of response. As a classical form of adjacency pair, they achieve
this by making non-responses (e.g., silence) or failures to address the questions topical agenda noticeable and accountable (Schegloff, 1972). Under such circumstances,
the questioner has the right to repeat the question or to solicit an answer in other ways
(Heritage, 1984:24551). Moreover, failure to respond appropriately attracts special inferences: in particular, that the answerer is being evasive, or has something to hide. This latter
sanction is particularly important when there may be millions of people watching on TV.

52 Studies in language and social interaction

These constraints are quite compelling for IEs. Silence hi the face of news interview questioning is incredibly rare! When asked a question, IEs always try to respond in some way,
and most often attempt to look as if they are answering the question (Clayman, 1993,
Notwithstanding the fact that the term topic is loose and difficult to define (Jefferson,
1984; Sacks, 19641972/1992),5 it is plain that IEs are oriented to the fact that there are
real boundaries to the topical agendas set by questions. In Example 7, a British Labour
politician is asked about the significance of a right-wing leadership success for the future of
his party. He begins by responding to the question as put, and then adds a comment (Lines
1418) about the future actions of the losing left-wing politician.
(7) [UK BBC TV: Panorama: 28 January 1981]
Roy Hattersley .hhh is it right to interpret this

as a move back .hh to the right. =This er victory


such a narrow marg[in of Denis Healey.]

[.h h h h N o] I dont

believe it i:s. in some ways I wish I could say


that, .hhhh But I dont believe it i:s. I believe


its a mo:ve back .hhh to the broad based


tolerant representative Labour Part(h)y, .hhh the


Labour Party in which Neil Kinnock and I: who


disagree on a number of policy issue:s .hh can


argue about them .hh without accusing each other


treachery:, .hhh without suggesting that one or


the other of us is playing into the Tories ha:nds.

.hhh And let me say something about the next year

because that was your original question. .hhh I


think Tony Benn would be personally extremely


foo:lish to sta:nd for the deputy leadership



The IE explicitly marks his additional comment as distinct and as departure from the questions agenda, and he goes out of his way to justify this departure by reference to an earlier
question asked by the IR (cf. Clayman, 2001). Here, the IE is clearly oriented to the topical
domain set by the IRs question.
Second, questions not only identify the topical domain to be dealt with in a response,
they also identify actions that the IE should perform in relation to the topical domain. In
Example 8 for instance, British Prime Minister Edward Heath is asked by David Frost
if he likes his main political rival of this period, Harold Wilson. Twice in this sequence,
Heath responds by addressing the topic of the questionWilsonbut he does not respond
in terms of the action agenda that the question called fora yes/no response on whether
See in particular Sacks lectures of March 9, 1967, and April 17, 1968. See also Spring 1970,
Lecture 5; Winter 1971, February 19, and Spring 1971, April 9.

Designing questions and setting agendas 53

he likes Wilson (cf. Raymond 2000). Instead he avoids the issue by talking in terms
of dealing with him and, more evasively still, working with other people who are in
(8) [UK BBC TV Omnibus: Frost-Heath Interview]
IR: Do you quite li:ke him?

IE: .hhh .h .h We: ll I th- I thi.nk in politics you see:
i- its not a ques:tion of going about (.) li:king

people or not, hh Its a question of dealing with


people, h .h a:n::d u::h (.) Ive always been

able to deal perfectly well with Mister

indeed: uh- he has with me,

IR: <But do you like> him?

IE: .hhhh Well agai:n its not a question of uh (.)
li:kes or disli:kes. I::ts a question of wor:king

together:: with other people who are in politics,



IR: But do ylike him.

IE: .hhh (.) Thatll have to remain tbe see:n wont

Heaths avoidance of the questions action agenda licenses Frost to renew it, and he does so
in a most pointed way at Line 10, and again at Line 16. Frosts <But do you like> him?
establishes a contrast (with the but) between Heaths response and what he wants to
know, and the repetition of his original question sets aside that response and clearly indicates (both to Heath and, more important, to the television audience) that Heaths response
was inadequate, and that he has avoided the question.
Third, the agenda-setting function of questions involves decisions about how narrowly
or broadly defined the IEs response should be. In Example 8, the agenda was set pretty
narrowly by means of a yes/no question that made Heath accountable to respond in these
terms (Raymond, 2000). Yes/no questions are recurrent sites of conflict between IRs and
IEs, as in Example 9, in which a Serbian commander who is suspected of war crimes in the
Bosnian conflict is pressed about whether he will deal with United Nations personnel who
are responsible for investigating war crimes:
(9) [UK BBC TV Newsnight: 11/02/93(IR Jeremy Paxman; IE Dragoslav Bokan)]
Mister Bokan, are you prepared to make yourself
available to U N investigators?

.hhhh Ah: first of all: I: just want to say that

its you know, very strange you know, to hear all

those accuses.=And ah: .hhh ah: its v(h)ery


strange to be in thuh (passive) role:: o:f

hearing, an:d ah .hh ah not to have an opportunity

54 Studies in language and social interaction











you know to:: say anything: uh .hhh ah about

yourself or: you know your: ah go:als. In war. .hh
An:d [ah:
[Im not interested in your goals Mister Bokan.=
=Thuh question wars: are you prepared to make
yourself avai:lable to U N investigators.
.hhh You know uh- you know: the answer, you
uh maybe better than ah m:yself. .hhh Because: o::f
>you know from the beginning of war,< .hhh I: have
just uh one goal an:d thats tdefend you know my
people: from thuh (lynch.)=
=Is that a yes or a n:o?
Uh: Is it a cour:t. (.) Or: a: interview.
So- you are: prepared to make yourself available
to U N investigators or no[:t.
[Of Course.

Here, the IE repeatedly avoids the question (Lines 411, 1519, 22). As the IRs series of
pursuits (arrowed) illustrates, the significance of yes/no and alternative questions is not that
IEs are necessarily forced to say yes or no right away. Rather it is that these questions
lay down a marker, making a yes/no response accountably avoided if it is not forthcoming. This in turn establishes the IRs right to renew the question,6 and IRs can and do avail
themselves of this right (Clayman, 1993). In a notable case, the interviewer in Example
9, Jeremy Paxman, asked a question 14 times of a British cabinet minister on network
televison!7 Thus IEs know that visible evasions license an IR to press them subsequently to
answer yes or no, and that this pressure may be heard as reasonable by the TV audience if
they seemed evasive in the first place. This kind of IR pressure may be heard as particularly
relevant and appropriate when there is the suspicion of wrongdoing, and/or where there is
an issue about the public accountability of the IEs actions.
In contrast, wh- questionsespecially what, why and how questionscan set the
parameters of response more broadly.8 For instance, Example 10 sets up a very open range
of responses from General John Vessey about his trip to Hanoi to negotiate over|information
about U.S. MIAs from the Vietnam War.

In this particular instance, the IR further narrowed the agenda of the question at Lines 2021 by
renewing his question as an explicitly disjunctive yes/no question. In this way,, he sharpened the
degree of constraint on the IE, and further underscored the IEs previous evasiveness as requiring
this narrowing.
This interview took place on May 13, 1997. Paxman subsequently won an award from the British
Academy of Film and Television Arts for the interview.
Not all wh- questions are equally open. In general, what, why, and how questions can enable more
exposition than who, when, and where questions, and are, in this sense at least, more open.

Designing questions and setting agendas 55

(10) [US PBS: Newshour: 10/23/92]
1 IR:
.hhh With us no:w for a newsmaker interview: is
delegation chairman former chairman of the joint
chiefs of sta:ff retired army general John Vessey.

IR: ->




General, welcome.
Thank you.
.hhhh Sir h:ow would you descri:be thuh significance
of this: (.) agreement.
.hhhhh Thuh Vietnamese:: uh: (0.2) foreign minister
and thuh Vietnamese prime minister (0.3) described
it to me: .hhh as a turning point.(0.3) i:n (0.4)
reh-resol:ving thuh fates of our missing. (0.5)
And I think thats what it is. It- .hh in thuh las:t
uh fi:ve years::

Here, the agenda for General Vesseys response is very under-specified. Almost any ontopic response would have likely counted as a valid and appropriate answer to the question.
In general, yes/no questions are potentially more constraining to an IE, whereas wh- questions can normally be successfully answered in a wider range of ways and using a wider
range of resources.
Tightening Question Agendas: Using Prefaces
As we have already suggested, the manifest function of question prefaces normally involves
giving background information to the audience, or managing topic shifts of various kinds.
However, question prefaces can also be used to make the agenda of a question more complex, constraining, or problematic. In the following case, a British conservative politician,
Michael Heseltine, is asked about his views on closer ties with Europe, an issue that had
become a source of conflict within his party:
(11) [UK BBC TV: Newsnight: 1989]
1 IR: .hh What Missus Thatcher has been saying: is
there is a danger (.) .h of a socialist superstate
being imposed (0.5) from Brussels (0.2) and
Mister Heath and others are saying is (0.2) that
is (.) is an illusory fear.=
=Where do you: line up on that is:sue.
7 IE: Well: (eh) technically, becaus:e (.) eh these
decisions are y:et to be ta:ken, it can go either
wa:y, (continues)

56 Studies in language and social interaction

Here, Heseltine is not simply asked about his opinion on the creation of a socialist superstate. Instead, by means of the question preface, the audience is instructed about the existence of two conflicting positions on this issue that are held by two of the most senior
members of the Conservative party. Within this framework, the question was made more
pointed and newsworthy by its invitation to Heseltine to say where he lines up in that
conflict. Here the question preface describes the parameters of the dispute and its primary
movers, making the nature of his political dilemma very clear to a viewing audience that
may have known little about the then-emerging disputes within the Conservative party on
this issue.
Prefatory statements may also be used to tighten the agenda being set for an IE by
blocking certain types of answer. The following segment comes from an interview with
Margaret Thatcheralso on closer ties with Europe:
(12) [UK BBC TV: Newsnight: 1989]
1 IR: Now turning to the exchange rate mechanism
have consistently said or the government has
.hh that you will joi:n when the ti:me is right
but people are saying: .hh that that means never.
Could you defi:ne the ki:nd of conditions when
you think we would go in.
7 IE: Uh no I would not say it means never. For the

The IRs question (Lines 56) is aimed at pinning down Thatcher to a specification of circumstances in which she would agree to join the exchange rate mechanism. He establishes
the agenda for this question with a preface that contrasts vaguely worded statements by
Thatcher concerning entry when the time is right with an interpretation of that statement, attributed to unidentified people (Clayman 1992), as never (Lines 14). The
preface provides a platform from which the question itself can be launched, while blocking
a response that, like the quoted when the time is right, would be vague and anodyne.
Still more complex is the following question preface to Senate majority leader, Robert
Dole. Here three main prefatory statements, all attributed to Dole, are used to set problems
for Doles stated objectives as a budget cutter:
(13) [U.S. NBC Meet the Press: 8 Dec 1985]
1 IR:
You cant have it both ways either.=>On thi.s
program< you have said that you dont think, .hhh
that youll eliminate thirty to fifty programs,
[an] Senator Packwood says you have to,=
5 IE:
[( )]
6 IR:
=.hh Number two you say you hope you will not
uh tax increase, [.hhhh And] number
8 IE:
[But I do.]

Designing questions and setting agendas 57




=and number three you say you h:ope you can have
a:l[m o s t] three percent on: .hhh on: on defe:nse,
[( )]
.hh And yet you hafta cut fifty billion next year.
Now which othose threes gunna give Senator,

In this case, the interviewer uses a series of prefatory statements to create a complex
dilemma for Dole. The statements describe three aspects of Doles positionhis admitted
inability to eliminate programs (Lines 24/6), his desire to avoid a tax increase (Lines 67),
and his hope to increase the defense budget (Lines 910). All three are incompatible with
Doles objective of cutting $50 billion from the federal budget. These three statements are
prepared for with a fourth at Line 1 (You cant have it both ways either.) that, among
other things, projects (to Dole and the news audience) that the subsequent statements will
identify contradictions that are troublesome to his position. At the end of this lengthy preface, Dole is invited to back down from one of his stated objectives (Line 12). This kind of
agenda could not be constructed without the prefatory materials.
Questions Embody Presuppositions
In addition to setting agendas, questions often assert propositions and they embody presuppositions with varying degrees of explicitness. This is so for both simple and prefaced
Most prefaced questions incorporate explicit contextualizing propositions. Once the
prefatory proposition is in place, the subsequent question can build from it and can embody
additional embedded presuppositions (Harris, 1986). Both of these features can be clearly
seen in the next case, which concerns an election in progress in which Labour politician
Tony Benn was ultimately the loser. Here the prefatory statement guardedly asserts (with
the evidential verb seems, Chafe, 1986) two propositions: the likely result of the election is
(a) close, and (b) against Tony Benn.
(14) [UK BBC TV: DLP: Hanna-Lansman]
1. IR: The result seems t be very close but (.) on
whorle it (0.2) doesnt look very good for::
Tony Benn.
Who do you bla:me for this?

The IE, Jon Lansman, was a supporter of Tony Benns. Thus the perjorative term blame here
also indexes his affiliation with Benn as the losing party in the election. As a matter of historical
record, the question likely invites the IE to name Neil Kinnock, at that time a left-inclined Labour
party figure whose vote againt Benn (together with those of a few supporters) may have tipped the
balance. After these events, Kinnock rapidly moved to the center of the Labour Party, later becoming its leader.

58 Studies in language and social interaction

The subsequent question Who do you bla:me for this? builds from this platform to project blame and its allocation as the primary agenda for the IEs response. Quite clearly,
it embodies the presupposition that a nameable set of persons can be held responsible for
the impending election defeat, and that these persons can and should be relevantly blamed
by the IE for this.9
Presuppositions vary in the extent to which they are embedded within a question. To
assess the degree of this embeddedness, we can consider whether the respondent can
address a questions presuppositions, while still responding to its agenda. In Example 14,
the respondent could have directly answered the question by responding that no one was
to blame. In this way, he would have responded to the questions overt agenda, while also
denying its basic presupposition. Thus the presupposition that persons are responsible and
blameable for Benns defeat is relatively close to the surface of the questions design.
This contrasts with other more embedded cases in which, if respondents wish to contest
a questions presuppositions, they must depart from directly answering the question
as put.
In Example 15, for instance, this more embedded form of presupposition is present.
This interview took place during a period in which health care reform was on the U.S. congressional legislative agenda. Here an advertising professional who ran a TV advertising
campaign against the Clinton proposals is questioned about the timing of her campaign.
Embedded in the question shown is the presupposition that this campaign has been initiated
early relative to the timing of the legislative program for health care reform:
(15) [US PBS Newshour 21 October 1993; Health Care: the IR, addressed by name at Line 6, is
Margaret Warner]
1 IR: =Mizz Jenckes, let me start with you. Ah: y:ouve
started all (of) this I think, thuh health industry
association.>Health insurance association, .hhh
Why:: so early in this debate when theres not gonna
be:: a vote on it ih- fr maybe a year?
6 IE: Margaret (.) health care reform is well under way

In this case, the presupposition is buried a little deeper than in Example 14: The IE begins
her response with an initial move to deny the questions presupposition that the campaign
was started early. Subsequently, she develops this response into an answer that more
explicitly justifies the timing of the campaign (data not shown).
A similar form of embedding is found in the following two casesalso involving
wh- questions:
(16) [US ABC Nightline: 15th October 1992 (concerning Bushs attacks on Clintons
character during the 1992 U.S. election campaign)]
1 IR:
But, Mister Cicconi, >what do you what dyou make
of thuh fa::ct that (.) the audience, thuh voters,
dont seem to like that?
4 IE:
.hhh Well I- ih: I didnt get that from the
audience at all, Chris. I thought- I thought thuh
point that thuh president ma::de about .hhh who can
you trust in a crisis, who.

Designing questions and setting agendas 59

(17) [US PBS Newshour: 21 October 1993] (Simplified)]
(Let me- Let me (just) ask Mandy Grunwald one other

question.=How do you explain: that (.) public


support for thuh Presidents plan has dropped off


rather sharply since he announced it a month ago?=

=We havent seen those sharp drops, at all. In

fact wev[e seen

[So your internal p[oiling doesnt
[Our- our internal

polling has seen sustain:ed ah: support for thuh



In each case, a presupposition embedded in the questions design and treated as given
information is contested by the IE who, as a result, did not so much answer the question as respond to it. In Example 17, it is noticeable that the IR pursues the discrepancy
between her assumed information and that of the IE by asking about the IEs alternative
source of information (internal polling).
Deeply embedded presuppositions can be put to damaging effect in what have been
usefully termed quandary questions (Nevin, 1994). These are questions of the when did
you stop beating your wife variety in which highly hostile presuppositions are so deeply
embedded in the questions design, that any response that directly answers the question
will also confirm the questions presupposition(s)with damaging consequences for the
IE. Wh- questions are generally the most favorable environment for deeply embedded
quandary type presuppositions. The following is a case in point:
(18) [UK BBC Radio: World at One: 13 March 1979]
.hhh er Whats the difference between your marxism

And Mister McGarheys communism.

er The difference is that its the press that

constantly call me a ma:rxist when I do not, (.)


and never have (.) er er given that description of



Any response by the IE, left-wing miners leader Arthur Scargill, that addresses the difference between his views and those of McGahey would confirm the embedded presupposition of the question that he is a marxist. Here, although Scargill starts his response within
the frame of the question (The difference is), he subsequently moves to undercut that
presupposition. However he can do so only by failing to respond to the question as put.
Yes/no or polar alternative questions, although they offer specific propositions for direct
response, still normally contain embedded presuppositions. For instance, Example 19 presupposes that Clintons character is problematicsomething that the IE, Clinton supporter
James Carville, explicitly contests in his response:

60 Studies in language and social interaction

(19) [US ABC Nightline: 15 October 1992) ((On the 1992 U.S. Presidential campaign)]
1 IR: ->
=.hhh Mister Carville: should Governor Clintons
character now be off: limits somehow?
3 IE:
Well I dont know anything about his character

being off limits thuh man has magnificent



And in the following case, the two alternatives (arrowed a and b) that are presented for
the IE to endorse are presented as exhaustive of his motives, and presuppose that there can
be no others. The question concerns Texacos agreement to settle out of court on charges
that the company systematically discriminated against its African-American employees.
(20) [US NBC Nightly News: 11/15/96:1]
.h Mister Bijur whats pro:- what prompted this

settlement? .hh

a-> Thuh fact that you concluded your company was


a-> in fact discrimina: ting


b-> or thuh prospects of: (.) more economic losses.

To:m it was that we wanted to be f:air: to

ah all of the employees involved, were a:


wonderful: gr:oup of people and family in this


company, en we wanta be equitable with everybody.

Here, as Roth (1998b) has noted, the or construction presupposes the correctness of one or
other of the candidate answers, simply leaving it the IE to confirm whichever explanation
is appropriate. This is something that the IE, a Texaco corporate executive, understandably
resists. It is notable in this example that the IE begins his response at Line 6 by addressing
the IR by name (To:m), summoning him into recipiency (cf. Schegloff 1968). By this
means, he projects that his subsequent action will be a volunteered first action, reducing its status as a second action that should properly fall within the terms of the prior
question (see also Example 15).10
In sum, all news interview questions embody presuppositions of some kind. For the
most part, these presuppositions are clearly shared between IR and IE and, quite commonly,
they have been established in earlier interview talk. Because of this, the presuppositional
basis of many IR questions can easily be overlooked and taken for granted. The nature of
IR presuppositions becomes most visible when, as in most of the previous cases, they are
rejected by IEs. In these incidentsand especially in quandary questionsthe difficulty
or hostility of the questions presuppositional content emerges quite clearly.11
The hostility embodied in IR questioning can be further shaped by aspects of question
design that favor one type of response over another, and it is this aspect of question design
to which we now turn.
See Clayman (1998) for a general account of the use of address terms in news interviews and
Heritage (2002) for other practices for reducing the responsiveness of second-position actions.

Designing questions and setting agendas 61

Questions Prefer Particular Responses
Though many news interview questions are not designed to favor particular answers, some
evidently are. This is important because the more strongly the IR designs a question to favor
one response over another, the more nearly their neurralistic posture may be compromised.
A number of practices of question designlargely associated with yes/no questionscan
achieve this outcome. What these practices have in common is some procedure for designing questions so as to inviteor, in conversation analytic terms, prefer (Heritage 1988;
Pomerantz, 1984; Sacks, 1973/1987; Schegloff, 1988)particular responses. This practice treats alternative IE responses as nonequivalent, and establishes a higher threshold of
accountability if the IE chooses to respond with the dispreferred option. When preference
organization is mobilized against the likely position of IEs, the latter may find themselves
responding in a more defensive or self-justifying way than might otherwise be the case.
Questions can be shaped to prefer particular responses through the design of the question
itself, or through prefatory statements, or by a combination of the two.
Conveying preferences through the design of interrogates. Various aspects of questions
can be designed to favor or facilitate particular IE responses. Some of these involve features of interrogative syntax itself. Although it might be thought that interrogatives are
safe and neutral because they do not express positions, this not always the case. For
example, questions that are framed using negative interrogative syntaxsuch as Wont
you, Isnt this, and so onare routinely treated as embodying very strong preferences
about answers. Indeed IEs recurrently respond to such questions as opinion statements to
be agreed or disagreed with (Heritage, in press). The following is a case in point. Here the
IE is the U.S. Ambassador to South Africa:
(21) [US PBS Newshour: 22 July 1985]
IR: -> But .isnt this (.) d- declaration of thuh state of

emergency:: (.) an admission: that the eh=South


African goverments policies have not worked,


in fact that the um- United States (0.2)





administrations policy of constructive engagement

(.) has not worked.
-> I do not agree with you .hhhh that the approach
have taken (.) toward South Africa is- a- is an
incorrect approach...

The IRs negative formulation Isnt this... is clearly treate d by the IE as asserting an
opinion when he replies I do not agree with you... This is the only type of interrogative
to which IEs recurrently respond in this way.

See also Maynard (1985) for a discussion of how presuppositions become progressively disembedded in argument sequences involving children.

62 Studies in language and social interaction

Given that negative interrogatives are often understood as opinion statements, a return
to our first example suggests an interesting kind of disingenuousness on Sam Donaldsons
(1) [US ABC This Week: October 1989: Barman]
1 IR: -> Isnt it a fact, Mr. Darman, that the taxpayers will

pay more in interest than if they just paid it out


of general revenues?
4 IE:
No, not necessarily. Thats a technical argument
5 IR -> Its not a-- may I, sir? Its not a technical
-> argument. Isnt it a fact?
7 IE:
No, its definitely not a fact. Because first of

all, twenty billion of the fifty billion is being


handled in just the way you want--through treasury


financing. The remaining

11 IR: -> Im just asking you a question. Im not expressing
-> my personal views.
13 IE:
I understand.

Here, it can be noticed that not only is Donaldons first question a negative interrogative
of the type that is frequently treated as an opinion statement, but also that at Lines 56,
Donaldson directly disagrees with the IE (with Its not a technical argument.), and then
effectively reasserts that opinion a second time with a renewal of his earlier negatively formulated question. Under these circumstances, it is hardly surprising that the IE treats him as
having taken a position on the issue (Lines 710). And it is this that makes Donaldsons
subsequent defense (that he was just asking you a question) distinctly disingenuous!
Other aspects of interrogative syntax can also be designed to prefer particular responses.
Straightforward cases involve the [statement]+[tag] question design. The statement
describes a state of affairs and the tag invites agreement or disagreement with the statement. The use of this format is designed to promote the IEs agreement with the statement,
thus agreement with the statement is preferred. Example 22 exhibits this construction:
(22) [UK BBC Radio: World at One: 13 March 1979]
1 IR:

=Do you ascri:be to Marxist economic philosophy.=

2 IE:

=I would say that there: er: the: (.) philosphy of

Marx as far as the economics of Britain is




concerned is one with which I find sympathy.=and

would support it.=Yes.
Well that makes you a Marxist doe[snt it.]
[Not nece]ssarily

Example 23 similarly illustrates the device in reverse form:

Designing questions and setting agendas 63

(23) [UK BBC Radio: Today: 1993]





Now theres talk that thuh cabinet will announce

some sort of am:nesty for people whove committed
crimes: ah racially motivated crimes presumably.
.hh Uhm under thuh ah over thuh last few years.
That wouldnt be acceptable to thuh A.N.C. would
.hhh Question of amnestys a very difficult

Here, agreement with the statement prior to the tag is still facilitated but, because the initial
statement is negatively formulated, an agreeing No answer is preferred.
Other aspects of question design can also embody preferences of this kind. For example
negative polarity items (Horn 1989) such as any embody a preference for a No answer,
as the following case in which the journalist relays other peoples descriptions of prison
camps in Bosnia to the IE, a representative of the International Society for Human Rights
(ISHR), and then asks Do you believe theres any justification for that at all?
(24) [UK BBC Radio Today: Bosnia Camps]
1 IR:
.hhh People have u::sed thuh phrase concentration

camps: and thuh Bosnians themselves have used


-> Do you believe theres any justification for that

at all?

Here, the final question-formatted segment of the IRs turn incorporates the negative polarity terms any justification and at all. This question, asked early in the Bosnia conflict and
before Serbian war crimes had been confirmed and publicized, and directed to a representative of an organization noted for its caution and probity in making partisan accusations,12 is
cautiously designed for a negative answer.
Finally, incorporation of terms like seriously or really can also embody preferences
for particular answers. When they are used, as they normally are, in questions that prefer
responses that contrast with IEs known positions, they strongly challenge them to defend
those positions. For example, in the following case, Ross Perot is interviewed about his
candidacy in the 1992 U.S. Presidential election, and his position on the growing U.S.
federal budget deficit. Earlier in the interview he had justified his candidacy as a means of
getting the main political parties to take the deficit seriously.
(25) [US PBS Newshour: 18 September 1992]

Alright n-, lets talk about some of the things


64 Studies in language and social interaction











propose. R: raising the tax on gasoline ten cents

yea:r for the next five y[ears fifty cents.
A::h a gallon after five y [ears.
[A:fter five years.
Eh: taxing all but fifteen percent of the social
security benefits of recipients that e:arn over
twenty five thousand dollars a year.
=Now youre endorsing that.
Do you (.) s:eriously believe that President
or Bill Clinton again is going to endorse either
[one of those.
[(I thought) they feel the American people
have the stomach for fair (0.2) shared (.)
sacrifice. (1.2) The facts are the American
do=Thats the point were trying to make.

Here, after listing two potentially unpopular tax measures, the incorporation of the word
seriously into the IRs question is designed for a no answerand is thus hostile to Perots
political position. If he is to be consistent with his earlier stated position, his answer to this
question must be yes, it must be accounted for, and he must do so in competition with the
skepticism that the interviewers question conveys.
Designing Preference Through Question Prefaces. In addition to the interrogative component of question design, question prefaces can also be built to prefer particular responses.
One straightforward method of doing so is to invoke others who take a particular
view of the issue (Clayman, 1992). In Example 26, for instance, the IE (who works for a
human rights organization) is asked whether he would describe prison camps in Bosnia as
concentration camps.
(26) [UK BBC Radio: Today: Bosnia Camps]
IR: -> .hhh People have u::sed thuh phrase concentration

-> camps: and thuh Bosnians themselves have

used that

-> phrase. Do you believe theres any justification

Earlier in the interview, the IR and IE collaborated extensively in establishing that the organization that the IE represents is at independent and impartial in the way it deals with human rights


Designing questions and setting agendas 65






for that at all?

.hh I think in thuh case of some of thuh larger
camps there are, thats certainly accurate .hh ah
if you count .h torture and execution as hallmarks
.h of concentration camps .h then thuh reports
weve received ah would seem to suggest that
is an
accurate description for some of them.

This is obviously a delicate question for a human rights worker to answer. As noted previously, in an earlier part of the interview, the IE had been at pains to stress the apolitical
and nonpartisan nature of his organization. The design of the IRs question reflects an
orientation to this issue. He introduces the question by referring to anonymous people
who have used the term concentration camps, and then augments this with the assertion
that the Bosnians themselves have used the same term, thus favoring a yes answer.
The final question asks if there is any justification for the use of this term. Although the
question itself, as we have seen, is designed for a negative answer, the referencing of others who would answer affirmatively establishes a favorable environment for an affirmative
answer. Overall then, whichever way the IE responds, he will be seen to have responded to
a carefully and judiciously formulated question, and can match it with an equally judicious
answer. It is just such a response that the question receives (Lines 510).
A rather more overt mobilization of preference is exhibited in Example 27. Here the
interview concerns pending legislation to reduce the time limit for legal abortions. The IE,
British Conservative MP Jill Knight, is in favor of the proposed legislation.
(27) [UK ATV: Afternoon Plus 1979 Abortion]
1 IR:
Can we now take up then the main issues of

that bill which r- (.) remain substantially the


same. (.) and indeed (0.2) have caused great deal


of concern. (0.4) But first youll note .hhh


is the clause about (.) time limits h in which h


abortions can be .h legally=

7 IE:
8 IR:
=ha:d. And the time limit h (.) according to the

bill has now dropped .h from twenty eight weeks


(.) to twenty wee[ks.

11 IE:
12 IR: ->
=Now<a lot of people are very concerned about
[.hh How concerned are you.
14 IE:
15 IE:
.hhh uh: (.) I think this is right. I think that

urn: .hh again ones had a lot of e:uh conflicting

66 Studies in language and social interaction


evidence on this but .hh what has come ou::t h an

I think that .h the public have been concerned
about this, .hhh is that there have been thmost
distressing cases

The IRs lengthy question preface (Lines 110) shifts topic (Lines 14) and describes
the proposed reduction of the legal abortion period (from 28 weeks to 20 weeks). It culminates in the observation that a lot of people are very concerned about this (Line 12).
The final interrogative component of the IRs turn invites, or challenges, the IE to address
that concern. Here the IE is invited to address the concern of people about the reduction
in the time limit for abortions, when this is something that she herself favors. The compelling power of this hostile question preface is shown by the IEs rather convoluted effort
to harness the term concern to issues on heranti-abortionside of the argument. The
distressing cases she goes on to describe involve the destruction of wellformed fetuses.
In this way, the IE establishes a superficial lexical connection between her comments and
the agenda set by the IRs question, and thus manages to twist the terms of the question in
a fashion that is more helpful to her position.13
In these cases, preference is established by a statement prior to the IRs question. A
similar effect can be achieved by a statement positioned after the question as in Example
28, where a member of the governing Conservative Party is questioned about the upshot of
his disagreements with the Thatcher administration:
(28) [UK BBC TV Newsnight: 14 October 1981]
But wont you have to consider threatening to vote

against the government, =

IR: ->
=Thats surely what (.) what all the critics now

have to face.
We::ll I dont know, no I- I think the: the were

still at the (.) stage of the intellectual argument


which I think .hh were winning,=because what


put forward is just the same old stuff. =Which nobody


believes and it hasnt worked.

Here the initial question component of the IRs turn, a negative interrogative that is itself
strongly weighted to expect an affirmative answer (see the earlier discussion of Example
21), is further supported by the flat assertion that all the (internal) government critics now
have to consider threatening to vote against the government.
The practice of prefacing questions with statements that are designed to favor particular
responses response can be developed to the point that IRs present positions as effectively
incontrovertible, and then invite IEs to deny them. This practice is common in cases where
IEs are engaged in defensive stonewalling. The following is a case in point. Here, the
U.S. Deputy Defense Secretary is interviewed about the Gulf War Syndrome and its


See Clayman (2001) for further examples of this process.

Designing questions and setting agendas 67

possible origin in seron gas used during the conflict. The syndrome is now the focus for
claims for compensation by war veterans:
(29) [US CBS 60 Minutes: Gulf War Syndrome]
1 IR: Secretary Deutch you say there is no evidence.
.hh Youve got ca:ses where: khh then- Czechs:
that they foun:d seron.
You say they didnt, th:ey say: (.) that they did.
.hh You have soldiers say:ing: that they experienced
burning sensations after explosions in the air. That
they became nauseous, that they got .hh headaches.
.hh You have two hundred fifty gallons of chemical
agents that were found in:si:de Kuwait.
.hh You had scuds that had seron in the warheads.
12 IR: If thats not evidence what is in.

Deutchs defense is, of course, oriented to the federal governments vulnerability to medical and other damages claims, which could be very extensive. The IR contrasts Deutchs
position with the statements of Czechs, the reported symptoms of soldiers, and other observations that are presented as fact. The final interrogative simply challenges the IE to deny
the evidential status of these various reported statements and assertions. In the way that this
evidence is compiled, the IR manages to exert very strong pressure on the IEs position.
Hostile Questioning: Splits, Forks, and Contrasts
Perhaps the most hostile questioning that IRs can engage in involves constructing IEs as
some form of disagreement or self-contradiction. This can take two main forms: IEs can be
presented (a) as in disagreement with their political allies, or (b) as in a situation of inconsistency or self-contradiction in their own positions.
We have already seen the first of these maneuvers in several earlier examples (e.g., 11
and 28). It is very common in Britain where the parliamentary process places a premium
on party loyalty, and consistency in voting with the party leadership. It is less common in
the United States where congressional voting is less constrained by party loyalties. British
journalists sometimes refer to this style of questioning as split hunting. A very overt case
is the following. The context of this interview is a developing disagreement within the Conservative Party over Britains relations with the European Union. The Conservative right,
led by then-Prime Minister Margaret Thatcher, was hostile to closer relations. Her position
was opposed by ex-Prime Minister Edward Heath, who led a faction favoring closer ties to
Europe. The conflict seemed likely to impact the political succession to Thatcherif the
left prevailed, the IE in the following example, Michael Heseltine, would have been the
likely next leader of the party. In this case, the IR attempts to induce Heseltine to take up a
public position that is opposed to Thatchers (and aligned to Heaths) on three successive
occasions and, in the subsequent parts of the interview, the same topic is pursued in more

68 Studies in language and social interaction

subtle ways. In fact, the entire 7minute interview is devoted to split hunting. We begin at
the beginning of the interview, where the IRs first question refers to a filmed report that
had just been shown:
(30) [UK BBC TV Newsnight: 1989]
1 IR:

Well Michael Heseltine lets begin: with

one of the

comments towards the end of Margaret


Was Philip Stevens of the Financial
Ti:mes right
(.) to place you: (.) in this argument
closer to
Mister Heath (.) than to Missus
7 IE:

=.hhh Well you know one of the reasons

that I:

wanted to (.) come on you:r pro:gra:m

.hh is

precisely to refu:se to invo:lve the personalities:


in this issue. I think Mister Heath has

done his

own cause a disservice .hh in: EU: the

way in which

he has spoken. This is not a matter of


personalities and the conservative party

is not

going to have th- the sort of row that the


will enjoy:. .Hhh but it is impo:rtant .h

that (.)

the conservative party and the country (.)


the ideas. And I wholly reject the analysis that


this will do us harm in the po:lls. I


itll do us good (.) because we shall be


the British people what the options are,

(.) what

the alternatives are, (.) and there will be


doubt in my mi:nd they will want conservatives to

Designing questions and setting agendas 69

24 IR:


26 IE:


27 IR:
28 IE:



31 IR:






38 IR:












45 IE:



pursue: whichever one we select.

But on: the substance of the ar:gument
are you
closer to: to Mister Heath=
=No youre [back on the [sa:me
si[tuation and what
[b- [No [Im ah
youre gonna try and do and youre not
succeed if we sit here all night, you are
not going
to get me into a personality [divisive
process. .hh
I will ta:lk about the ideas of Europe.
My:- my- I
cannot overstress(f) to you (.) the European issue
is going to dominate the next deca:de,
and if we
try to conduct it on a sort of personality
basis .h we will divert the industrialn
companies away from the real challenges
they face.
Well often uh (.) politics reach: the public uh (.)
through personality, .hh what Missus
Thatcher has
been saying: is that there is a danger (.)
.h of a
socialist superstate being imposed (0.5)
Brussels (0.2) and what Mister Heath
and others are
saying is (0.2) that is (.) is an illusory
=Where do you: line up on that is:sue.
Well: (eh) technically, becaus:e (.) eh
decisions are ye: t to be ta:ken,..

In the first yes/no question (Lines 46), the IR constructs an agenda for Heseltines response
that presupposes the conflict between Thatcher and Heath as its primary reference point,
and is designed for a Yes response. When Heseltine attempts to reformulate the issue in
terms of discussing the ideas and options (Lines 1617, 1821), the IRs subsequent

70 Studies in language and social interaction

question (Lines 245) pursues the original question of Heseltines alignment. He does so,
with the but preface, and the virtual repeat of the terms of his earlier question at Line 4, in
such a way as to formulate Heseltines previous response as an evasion (see the earlier discussion of Example 8). This question is also designed for a yes. Finally, after Heseltine
again declines to respond in terms of personalities (Line 36), the IR reinstates the issue
for a third time in terms of a substantive disagreement between Heath and Thatcher (Lines
3844), albeit with a question that is neutral in preference terms. Although this case is
quite egregious, it embodies characteristic features of British political interviewing that are
applied to senior figures in all three political parties.
A close relative of split-hunting questions are those that place the IE in a dilemma
or fork. Most commonly, these are shaped as disjunctive questions. For instance, in
Example 31, a British Labour politician is discussing his partys defense policy: Across a
number of earlier turns, the IR has been pressing his respondent on the issue that the party
would like to be rid of nuclear weapons:
(31) (UK: BBC TV Newsnight: 1989)

So what will you be pushing for tomorrow, what


your: bottom line as you said earlier?


Well I think therell be a number of (0.2) proposals


put by different colleagues, but the bottom line


to be that if things go well and talks procee:d w


uh, as we would want them to, over the first two


three years, both on strategic arm:s, and on the


question of a nuclear free Europethen, of,


course wed have achieved our objective slightly


more slowly than we used to deba:te, but (.) as


part of a: an international change, which would


be welcome and would contribute to the safety

of the

world. .hh if we dont get that, then I think some


us have to sa:y in- in all credibility .hh that we


would want Britain to be able to remove those


.hhh independently, unilateral[ly if tha[ts the


[In uh- [In uh

17 IR:
18 IE:

=youd like to put it.=

19 IR:

=In other words, I dont understand the logic of


this:, uh Mr. Blunkett,

a-> if things are going well, and the, the atmosphere

Designing questions and setting agendas 71






of international detente continues (.) youre quite

happy to negotiate the weapons away,
but if things (.) go badly, and I assume by that
you mean some kind of return, to some kind of
war atmosphere, then youll (.) give them away
[Well I: I Im not talking about giving anything

Here the IEs lengthy statement about his partys nuclear weapons policy (Lines 316)
straddles policy conflicts within his party between those who wish to remove nuclear weapons as part of a negotiation, and those who would prefer to remove them unilaterally. The
IRs summary formulation (Heritage, 1985) simply sharpens this into an explicit contradiction, suggesting that the party will remove nuclear weapons under any conditions. This
implies either that the party has no coherent negotiating position or, worse, that it remains
committed to the politically unpopular policy of unilateral nuclear disarmament.
A rather different kind of fork is manifested in Example 32. Here the IEthen-Senate
leader Robert Doleis invited to explain the fact the President Reagans political programs are in trouble. In the question preface, the IR offers two anonymous and thirdparty-attributed formulations of the situation. The first is that Reagans programs, though
not the President himself, are in trouble. The second offers an explanation for that trouble
in terms of ineffective legislative leadership. The latter explanation, which engenders a
little laugh from Dole, is explicitly offered as implicating Dole himself.
(32) [US NBC Meet the Press: December 1985]

Senator (0.5) uh President Reagans elected


thirteen months ago: an enormous landslide.


It is s::aid that his programs are in trouble,

though he seems to be terribly popular with the

American people.

It is said by some people at thuh White House

could get those programs through:: if only we
perhaps more: .hh uhffective leadership up on

hill an I [suppose] indirecly that might ( )
12 IE:

13 IR:
relate tyou as well:.

15 IR:

Uh whaddyou thi.nk thuh problem is rilly. Is it

72 Studies in language and social interaction



(0.2) thuh leadership as it might be claimed up

thuh hill, er is it thuh programs themselves.

In the final formulation of the question, the IR draws on this extensive question preface
and explicitly invites Dole to identify the problem in terms of either the (de-)merits of
the programs, or ineffective legislative leadership. These were presented as exhausting the
possible explanations for Reagans legislative difficulties. As in Example 20, neither option
can possibly commend itself to a Republican Senate leader, and Doles response avoids
these options in favor of a response that cites the weakness of his majority in the Senate
(data not shown).
Finally, in a convergence of the split and the fork formats, IRs may contrast the
conduct of the IE with the conduct of another individual who is allied to the IE. In these
kinds of contrasts, the conduct of the second individual is normally used as a kind of
moral template for appropriate conduct (Smith, 1978). A notable use of this kind of
question occurred when then Vice-President (and presidential candidate) George Bush was
interviewed by Dan Rather live on CBSs Evening News.14 The film report preceding
the interview focused heavily on the Iran-Contra scandal, and ended with a description
of contacts between Bushs long-serving national security aide Donald Gregg and Contra
middleman Felix Rodriguez. Rathers opening question took up this topic.
(33) [CBS Evening News: 1/25/88 Bush-Rather]
1 IR: Mister Vice President, thank you for being with us
tonigh:t, .hh Donald Gregg still swerves as your
trusted advisor, He was deeply involved in running
arms to the contras, n e didnt inform you.
.hhh Now when President Reagans (0.3) trusted
advisor: Admiral Poindexter: (0.5) failed to inform
him::, (0.7) thuh President (0.4) fired im.
Why is Mister Gregg still:: (0.2) inside thuh White
House n still a trusted advisor.
11 IE: Because I have confidence in him, .hh n because
matter Dan:, as you well know:,

Here the IR, building from the film report, begins by asserting that Gregg still serves
Bush as a trusted advisor. He continues by depicting Greggs conduct as untrustworthy:
running arms to the Contras without informing Bush. This state of affairs is then contrasted
with the morally appropriate action that President Reagan took when his trusted advisor
Admiral Poindexter engaged in actions that breached that trust (Lines 57). The contrast
between Reagans and Bushs conduct is clearly drawn. The similarities between the advisors are established point for point, and Bushs conduct is presented as clearly differing
See Clayman and Whalen (1988/1989), Schegloff (1988/1989) and Pomerantz (1988/1989) for
other treatments of this interview.

Designing questions and setting agendas 73

from Reagans. This contrast is particularly pointed. Not only is Reagan Bushs political
ally and superior, he is also President of the United States, and a role model for the position
that Bush is currently campaigning for. Bush can thus be directly asked to explain the contrast between his conduct and that of his superiorthe occupant of the supreme position to
which he aspires. This is, of course, what the IRs question (Lines 910) proceeds to do.
This chapter has argued that, although questioning may generally be understood as a
neutralistic activity in the news interview context, neutralism is not to be confused with
neutrality. News interview questioning is very far from being a neutral activity. As we
have seen, the IR holds the initiative when it comes to the topics that the IE will be questioned on. There can be no neutrality in the selection of these topics and contexts: rather
the selection will be more or less favorable (or, which is not necessarily the same, more or
less desirable) from the IEs point of view. Further, the IR can manage questioning so that
particular presuppositions are incorporated in the design of questions and at varying levels of embeddedness. These presuppositions may be more or less problematic for an IEs
position, and their degree of embeddedness may create greater or lesser difficulties for the
IE in formulating a response. Finally, the IR can manage questions so that particular audience expectations for the IEs response are mobilised: expectations that the IE may need to
resist, and where such resistance may incur an additional burden of explanation than might
otherwise be the case.
News interview questioning, then, cannot be neutral but only neutralistic. It can be
more or less pointed, more or less fair, more or less balanced in its approach to its subject
matter. Much of the evaluationby the IE and, especially, by the news audienceof these
characteristics of IR questioning is likely to be shaped by perceptions of the relevance of
particular questions. For both the IE and the news audience, the prevailing consideration in
relation to each question is why that now (Sacks, 1992). The conclusions that are drawn
by the IEs (and, just as important, the news audience) about the why that now issue will
shape how the questioners purpose is understood and, relatedly, whether a question is
judged to be appropriate or fair.15
This chapter has aimed at laying out some basic features of question design in the news
interview context, and to describe their deployment in a range of instances. Underlying
some of these observations is the suggestion that innovation in question design can be
an important element of social change in the news interview context, and broadcast journalism more generally. In particular the emergence and growth of the prefaced question
design, while initially developed and used to inform the news audience about important
contextual details, represents a formidable extension of the interviewers initiative and
power. Many of the more hostile questions discussed in this chapter simply could not be
launched in any other way.

For example, Dan Rathers questioning of George Bush was widely judged to be inappropriate
and had substantial negative consequences for Rather and, indirectly, for CBS news (Clayman and
Heritage, 2002a).

74 Studies in language and social interaction

In a nonrandom, but wide-ranging, sample of 639 questions from British and American interview data, Heritage and Roth (1995) found that nearly half of the total questions
asked were prefaced questions. In a recent study of presidential press conferences Clayman and Heritage (2002b) also found that simple questions fell from 44% of the total
during Eisenhowers first term to 21% during Reagans first term. During the same period
hostile question prefaces multiplied by a factor of 450%. Although the relative absence
of follow-up opportunities may encourage journalists to produce more complex questions
in the press conference context, these figures are nonetheless striking, and may index a parallel underlying growth in the deployment of prefaced questions in the U.S. news interview
context as well. If this is so, it is clear that journalistic initiative has expanded considerably
during the past 40 years and, in all probability, that this is directly associated with a growth
in adversarialness, which, by common consent, has also grown significantly during this
The growth of prefaced questioning may, however, have different institutional histories
in Britain and America. In Britain, legislative regulation and oversight of broadcast journalism has historically been more intense than in the United States. Moreover, until 1958
when the BBCs monopoly position in broadcasting was replaced by a duopoly, there were
no competitive pressures that might fuel a reduction in deference and a rise in adversarialness. In the United States, by contrast, FCC oversight and regulation of news program
content has been minimal, and competitive pressures have impacted broadcast journalism from the outset. It may be conjectured then that in Britain there was a more dramatic
growth in prefaced questions, beginning in the 1960s, whereas in the United States growth
was more steady and gradual and began from a higher baseline. This in turn suggests that
news interview questioning may never have been as deferential in the United States as it
was in Britain during the 1950s. Thus the Attlee example (2) with which this discussion
began may truly represent one of the more extreme cases of deferential interviewing that
one could find in the anglophone broadcasting context.
Boyd, E., & Heritage, J. (in press). Taking the patients personal history: Questioning during verbal
examination. In J. Heritage & D. Maynard (Eds.), Practising medicine: Structure and process in
primary care encounters. Cambridge, England: Cambridge University Press.
Brown, P. & Levinson, S. (1987). Politeness: Some universals in language usage. Cambridge, England: Cambridge University Press.
Chafe, W. (1986). Evidentiality in English conversation and academic writing. In W.Chafe &
J.Nichols (Eds.), Evidentiality: The linguistic coding of epistemology (pp. 261272). Norwood
NJ: Ablex.
Clayman, S. (1988). Displaying neutrality in television news interviews. Social Problems, 35(4),
Clayman, S. (1992). Footing in the achievement of neutrality: The case of news interview discourse. In P.Drew and J.Heritage (Eds.), Talk at work (pp. 163198). Cambridge, England:
Cambridge University Press,
Clayman, S. (1993). Reformulating the question: A device for answering/not answering questions
in news interviews and press conferences. Text, 13, (2), 159188.
Clayman, S. (1998, Novermber). Some uses of address terms in news interviews. Paper presented at
the annual meeting of the National Communication Association, New York.

Designing questions and setting agendas 75

Clayman, S. (2001). Answers and evasions. Language in Society 30:403442.
Clayman, S. & Heritage, J. (2002a). The news interview: Journalists and public figures on the air.
Cambridge: Cambridge University Press,
Clayman, S. & Heritage, J. (2002b). Questioning presidents: Journalistic deference and adversarialness in the press conferences of Eisenhower and Reagan. Journal of Communication 52 (4).
Clayman, S. & Whalen, J. (1988/1989). When the medium becomes the message: The case of the
Rather-Bush encounter. Research on Language and Social Interaction, 22, 241272.
Harris, S. (1986). Interviewers questions in broadcast interviews. In J. Wilson & B. Crow (Eds.),
Belfast working papers in language and linguistics (Vol. 8, pp. 5085). Jordanstown, Ireland:
University of Ulster.
Heritage, J. (1984). Garfinkel and ethnomethodology. Cambridge, England: Polity Press.
Heritage, J. (1985). Analyzing news interviews: Aspects of the production of talk for an overhearing audience. In T.A.Dijk (Ed.), Handbook of discourse analysis (Vol. 3, pp. 95119) New York:
Academic Press.
Heritage, J. (1988). Explanations as accounts: A conversation analytic perspective. In C.Antaki
(Ed.), Understanding everyday explanation: A casebook of methods (pp. 127144). Beverly
Hills, CA: Sage.
Heritage, J. (in press). The limits of questioning: Negative interrogatives and Hostile question content. Journal of Pragmatics.
Heritage, J. (2002). Oh-prefaced responses to assessments: A method of modifying agreement/disagreement. In Cecilia Ford, Barbara Fox, & Sandra Thompson (Ed.), The Language of Turn and
Sequence, (pp. 196224) Oxford: Oxford University Press.
Heritage, J., & Greatbatch, D. (1991). On the institutional character of institutional talk: The case of
news interviews. In D.Boden & D.H Zimmerman (Eds.), Talk and social structure (pp. 93137).
Berkeley: University of California Press.
Heritage, J., & Roth, A. (1995). Grammar and institution: Questions and questioning in the broadcast news interview. Research on Language and Social Interaction, 28(1), 160.
Horn, L. (1989). A natural history of negation. Chicago: University of Chicago Press.
Jefferson, G. (1984). On stepwise transition from talk about a trouble to inappropriately nextpositioned matters. In J.M.Atkinson & J. Heritage (Eds.), Structures of social action: Studies in
conversation analysis (pp. 191221). Cambridge, England: Cambridge University Press.
Maynard, D.W. (1985). How children start arguments. Language in Society,_14, 129.
Nevin, B. (1994). Quandary/abusive questions. The Linguist Discussion List, 5754.
Pomerantz, A. (1984). Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. In J.M. Atkinson & J. Heritage (Eds.), Structures of social action: Studies
in conversation analysis (pp. 57101). Cambridge, England: Cambridge University Press.
Pomerantz, A.M. (1988/9). Constructing skepticism: four devices used to engender the audiences
skepticism. Research on Language and Social Interaction, 22, 293313.
Raymond, G. (2000). The structure of responding: Conforming and nonconforming responses to
yes/no type interrogatives. Unpublished doctoral dissertation, University of California, Los
Roth, A. (1998a). Who makes news: Descriptions of television news interviewees public personae.
Media, Culture and Society, 28(1), 79107.
Roth, A. (1998b). Who makes the news: Social identity and the explanation of action in the broadcast news interview. Unpublished doctoral dissertation, University of California, Los Angeles.
Sacks, H. (1987). On the preferences for agreement and contiguity in sequences in conversation.
In G.Button & J.R. E.Lee (Eds.), Talk and social organization (pp. 5469). Clevedon, England:
Multilingual Matters.
Sacks, H. (1992). Lectures on Conversation (2 vols.) (G. Jefferson, Ed.). Oxford, England: Basil

76 Studies in language and social interaction

Schegloff, E.A. (1968). Sequencing in Conversational Openings. American Anthropologist, 70,
Schegloff, E.A. (1972). Notes on a conversational practice: Formulating place. In D.Sudnow (Ed.),
Studies in social interaction (pp. 75119). New York: Free Press.
Schegloff, E.A. (1988). On an actual virtual servo-mechanism for guessing bad news: A single case
conjecture. Social Problems, 35(4), 442457.
Schegloff, E.A. (1988/9). From interview to confrontation: Observations on the Bush/Rather
encounter. Research on Language and Social Interaction, 22, 215240.
Schudson, M. (1994). Question authority: A history of the news interview in American journalism,
1830s1930s. Media, Culture and Society 16, 565587.
Smith, D. (1978). K is mentally ill: The anatomy of a factual account. Sociology, 12,(1), 2353.

Taken-for-Granteds in (an) Intercultural
Kristine L.Fitch
University of Iowa
In a pair of articles (Hopper, 1981a, 1981b) synthesizing theory and research from diverse
areas of social science and philosophy, Robert Hopper formulated the nature and functions of taken-for-granteds (TFGs), unspoken yet ordinarily understood between-the-lines
aspects of talk. Emphasizing that TFGs were not to be equated with nonverbal messages,
Hopper noted the essentially incomplete and often telegraphic nature of much face-toface interaction. He pointed out similarities between missing premises in enthymemes,
pragmatic implications of utterances inferred from felicity conditions and conversational
maxims, and other well-studied categories of unspoken messages as the parts that when
presumed to form coherent patterns, constitute communicative frames (Bateson, 1972;
Goffman, 1974). The concept of TFGs put forth in those articles has proved a powerful
analytic tool in communication studies and related disciplines, and was recently the theme
of a Northwest Communication Association convention.
This essay illustrates one kind of TFGs in everyday talk: cultural premises, that is,
unspoken assumptions drawn from a specific communal system of symbolic resources.
This analysis draws upon a tradition within the ethnography of speaking that begins from
an assumption that peoples ways of speaking are structured by cultural codes (Philipsen,
1992) that are in turn assumed to vary across cultures. This assumption is not contradictory to the emphasis on structure and organization of talk typical of conversation analysis
(CA). It is worth noting, however, that ethnographies of speaking generally proceed under
the assumption that speakers draw upon cultural codes of meaning that are constructed
across time in order to communicate in a given conversational moment. Unlike CA, then,
there is an expectation in ethnographies of speaking that such codes will most often be
invoked implicitly, rather than being referred to explicitly, in most instances of everyday
conversation. I focus this discussion of TFGs around a conversation in which distinctive
cultural codes form the bases for contrasting proposals for action, despite speakers agreement about the objective these proposals are meant to accomplish. Although elucidating
TFGs can illuminate identification and understanding of culture in talk, which provides the
analytic vigor of the concept, I argue that making them explicit through metacommunication during interaction can be problematic. That said, the power of the TFG construct lies
in revealing the ambiguity and enigma inherent in talk, and the possibilities such incompleteness leaves open for multiple, often productive, alternative framings of interpersonal

78 Studies in language and social interaction

Family dinner table conversation has been recognized for some time as a particularly rich
setting for talk that is more obviously culturally situated than in some other settings and
activities (see, e.g., Blum-Kulka, 1997; Ochs, Smith, & Taylor, 1989). Talk within families
is a primary vehicle for socialization of children into a speech community. That talk may
be implicitly instructive, as parents model desired ways of speaking and correct childrens
deviations from them (Erica, Gabe is talking, you need to wait your turn). It may also
be quite direct, as parents give voice to cultural norms for behavior and, at times, to the
understandings that underlie those norms (Its rude to talk when someone else is talking,
it seems like what theyre saying isnt important enough for you to listen to.)
A third way in which cultural norms and premises (described by Philipsen, 1997, as
cultural codes) are discernible in dinner table talk is through examination of such talk for
TFGs that are relevant to the matter at hand in some culturally situated way. What counts
as a culturally specific or relevant TFG is discussed in more detail later. For now, I propose
that a case can be made that a particular instance of talk is consistent with, and counts as an
enactment of, a cultural system of belief. Making such a case, however persuasively, does
not constitute ruling out other explanations, particularly if it is made on the basis of a single
fragment rather than a collection of similar instances. It does, however, provide a starting
place for pursuing cultural codes through their subtle appearances in everyday talk.
The transcript that follows is the first few minutes of a family dinner table conversation
that lasted approximately 15 minutes in total. Just before the recorder was turned on, the
son asked for an increase in his allowance. The participants are the mother (M), the father
(F), the 9-year-old son (S) and the 7year-old daughter (D). Some side sequences have been
edited out for length.





but uh REASonable (.) mount (.) that Im gonna spend

What I? think is you need (.) two: (.5) containers
fer mo?ney (.) one tht you kin draw on tspend fr
things like bake sales an one that you dont touch
(.5) .hh one that you just (.) keep building up until
you have enough to go to the bank.
(4 lines deleted)
but (.) the WAY YOU have yer ((cup clinks)) finances
situated right now (.) Y YMIGHT PUT YER WHOLE
yeah but one thing<< (.5) If I did tha: t
((swallows)) I dont (.) I make LES:SSh: than (.)
ten dollarsh (.) a month (2.5) (with thllwance Im
getting) (.) right now

Taken-for-granteds 79




























ten dollars
If you gotta raise tthree dollars youd be making
twelve dollars a month
sou::nds good
Well? outta that twelve dol?lars I wancha tbe saving
(1.0) EIGHT of it
y ocho
Between six? and eight
He (.) them he could have

to spend

(right) (.5) soun?ds good
(1 line deleted)1
dan los (ahorros) a
y yo los
Give me the (savings) and Ill take
guardo (.5) Los ahorros gue van a ir para el banco=
care of them(.5)the savings that will go to the bank=
=okay (.5)


yo los man
Ill keep them
=y nadie puede tocarmelos=
=and no one can touch them=
(1 line deleted)

I TRI: (.5)

The side sequence that begins here and ends at Line 59 involves D stating her intention to go to
her room to change the page on her calendar. M vetoes the idea, asking D to finish her dinner first.

80 Studies in language and social interaction











=I said (.) will you?

(2 lines deleted)
Thats what I tried (1.0) thats what I asked you to
do once and then (.) Will you keep my allowance? for
a coupla weeksh?
Youre right? And I was not willing to (be uh)
Pero es que yo lo nico que voy a guardar es lo que
But the only thing Im gonna take care of is that
st guardao guardado (.) Lo que va a ir para el
which is saved saved (.) that which is going to the
banco (.) Lo o?tro no
bank (.) The re?st no
hm hm hm
Porque tienen que aprender a manejarlo
because they have to learn to manage it
Well? Im (.) I think if somebodys hangin onto it
for im hes not? learning ta
hm hm hm
((softly, through food)) manage it
but if it would help? im out (.5) IF IT WOULD HELP


hang onto it then (.5) whom I? tstand in the way a


Immediately obvious from this transcript is that S does not get a straight answer to his
request. In fact, when the topic shifts at the end of segment presented here, there has been
no clear indication of whether or not he will get the desired increase in his allowance.
Rather than a yes or no answer, M responds to his request with a proposal for how S should
manage his money more generally. In some contexts, between some participants, the relevance of this response could be questioned. M does not specify how the proposal is connected to the request; perhaps the increase will be contingent on S agreeing to adhere to it.
She does go on to make more explicit the basis for the proposal, noting a problem with Ss
current money management practices:



but (.) the WAY YOU have yer ((cup clinks)) finances
situated right now (.) Y YMIGHT PUT YER WHOLE

Taken-for-granteds 81
Comparing the present state of affairsthe next day you go and get all of it outwith
the preferred alternative you just keep building up until you have enough to go to the
bank (Lines 67)suggests that the habit M wishes to correct is S spending all of his
money, rather than saving some of it. Certainly this is a common theme of parental instruction to children. Of note here is Ms emphasis on the actions and choices of the child himself, signaled by her repeated use of the pronoun you: You need two containers, one that you
can draw on, one that you dont touch, one that you build up until you have enough to go
to the bank. Assuming that goodies bought at a bake sale have more allure for a 9-year-old
than a container of money waiting to be taken to the bank, adhering to this proposal will
also require (and may be intended to instill) significant self-restraint.
Perhaps calculating the degree of self-restraint that he will be expected to exercise,
S notes that if he follows Ms proposal he ends up with less actual cash in hand than he
currently has (getting $2 a week and essentially being free to spend all of it, as opposed
to getting $3 a week and having to put $2 of it into the banks container). In Line 26 M
emphasizes the vast amount of money he would be receiving, twelve dollars a month, to
counter this objection.
Although the children initially agree to their mothers stated expectation, after a lengthy
pause the father moderates itbetween six? and eight. His mitigated proposal comes in
Spanish, at the same moment that the son seems to reconsider his agreement to saving all
$8. D quickly figures what spending money would be left under this plan, whether as an
endorsement or as a demonstration of her mental math skills. Neither child misses a beat
in these responses; whatever else may be said about the fathers use of Spanish (a matter
that is explored more fully later), language choice itself does not draw a reaction of any
kind. It seems to be unmarked, typical behavior for talk at this dinner table to go on in two
languages simultaneously.
In Lines 4654, F offers a counterproposal to solve the problem of S spending all of his
money. There is a clear stylistic contrast between Ms proposal and Fs:



Me dan los (ahorros) a mi y yo los guardo

Give me the (savings) and Ill take care of
What I? think is you need (.) two: (.5) containers
fer mo?ney

M seems to have offered a suggestion that may be taken merely as an opinion: What I?
think you need Whereas Fs utterance sounds like a command: Give me Translation
is tricky here, however, and the ambiguity is not one that can be resolved from hearing the
tape. In Spanish, pronouns are often optional, particularly at the beginning of a sentence. F
may be offering a suggestion as well, with the initial pronoun/verb left implicit, as in:

(Pueden) darme los ahorros

(They can) give me the savings

82 Studies in language and social interaction

Whatever the illocutionary force of the utterance, the substance of the proposal also contrasts with the one offered earlier by M. Rather than assigning a container the job of holding onto the savings until enough money has accumulated to go to the bank, F suggests
(or declares) that he will take on that role. Ss reaction is immediate enthusiasm: This is
exactly what he wanted all along, what he tried to get M to do for him once. There is a note
of accusation in the dramatic replaying of his appeal:
58 S.

thats what I asked you to do once and then (.) Will

keep my allowance? (.) for a coupla weeksh?

When M confirms that she has previously rejected the plan F is supporting, F interjects
with clarification of his role. He will not be in charge of ALL of the childs money, only
that which is going to the bank. Although M has not voiced a reason for her refusal to
cooperate with Ss earlier attempt to instantiate this system, F anticipates that it is a parents
involvement with the childs money that was the basis for her objection. In Line 71, he
suggests a point of agreement between him and M, a commonality despite their contrasting


Porque tienen que aprender a manejarlo

because they have to learn to manage it
Well? Im (.) I think if somebodys hangin onto it for im
hes not? learning tamanage it

Ms immediate response is to disagree, not that they (both children, perhaps all children)
need to learn to manage money, but that if somebodycertainly F, since he has just
offered to do sohangs onto (part of) the money, S is not learning to manage it. In her
view, the plan F has proposed does not count as teaching S, or perhaps any child, to manage money. Her discontent with the role he has offered to play is mitigated by applying her
objection to somebodynot to him specifically, which would create an accusatory tone.
When this disagreement is met with silence (Line 79), M closes the topic with what sounds
on one level like an immediate reversal of field:



but if it would help? im out (.5) IF IT WOULD HELP
hang onto it then (.5) whom I? tstand in the way a

It is precisely this quick juxtaposition of the stated view that if somebodys hangin onto it
for him hes not learning to manage it with the (louder) opposite IF IT WOULD HELP
YOU OUT that marks the latter as sarcasm. The parting shot is an idioma prepackaged, and thus hard to object to, formulaic construction. Drew and Holt (1988) noted
that idioms frequently occur at the end of complaint sequences, serving as a figurative

Taken-for-granteds 83
summing-up of a grievance that brings the matter to a close. Although their exploration
of idioms shows a number of cases in which the function of the utterance is to bring the
speaker and recipient into some kind of alignment, M is clearly not expecting to elicit
agreement with her point of view. In speaking sarcastically, she is in fact complaining that
her proposal has not been supported by F or gotten uptake from the children. Assuming she
has not really changed her mind from one phrase to the next, she is conceding defeat (at
least in this conversational moment), but not without first commenting on the irrationality
of the proposal that has been greeted with more enthusiasm than her own.
In what sense, and to what extent, is this conversation among family members intercultural communication? What conversational features mark this (or any conversation) as an
instance of contact between members of different speech communities? The most obvious
answer seems, accurately, too easy: There are two languages used, Spanish by one of the
participants and English by the other three. Although language differences may be readily
observable boundaries between speech communities, Hymes (1972) noted that members
of a single speech community may well share two languages. The three English speakers
in this conversation plainly have no trouble understanding what is said in Spanish, so these
four could be part of a bilingual speech community in which mixing languages is in itself
unmarked behavior.
There is a noticeable disagreement in this exchange, however, in which contrasting
proposals are put forth. An attempt to state common ground that might align the two plans
is rejected. This disagreement, I suggest, reveals the existence of different assumptions
about personhood and relationships as enacted in money management practices. Those
TFG assumptions, left unstated as is most often the case, can be shown to be part of distinctive cultural codes. It is the contact between those codes that makes this intercultural
There are two questions at hand: What are the TFGs behind the distinct proposals, and
what is there to suggest that those TFGs are cultural premises? As noted earlier, Ms proposal emphasizes the childs actions and his (autonomous) responsibility for them. Dividing the allowance and keeping one part of it out of the spending loop is to be a matter
between S and two containers, physical objects that cannot praise him for compliance,
reproach him for lapses, or remind him of his promise (and thus reinforce it) at moments
of temptation. By contrast, Fs proposal that he, a human being endowed in a 9-year-olds
mind with both power and wisdom, take charge of the money allows S to draw upon the
strength of another person when his own willpower flags. Ms references in Lines 1013
suggest that S has unhappy experience with just such lapses. The remedy she suggests,
perhaps with a view toward underscoring the importance of willpower generally, relies on
increasing Ss ability to control and restrain his impulses. Fs remedy allows him to rely on
another person for help.
Left unsaid in this particular conversation are ideals of reliance on oneself as an individual versus reliance on other people, premises that are readily recognizable as a common

84 Studies in language and social interaction

contrast between cultural systems. The fact that this contrast suggests that M comes from
an individualist culture and F comes from a collectivist one (Triandis 1988; Triandis et al.
1988) does not, however, illuminate very much that is specific to either culture. In what
sense, then, is this observation any more helpful than the obvious and unremarkable one
that these people are speaking different languages?
A problem with durable dichotomies like individualism/collectivism is that, although
resonant and often useful, they are too broadly conceived to be more than blunt instruments. They hack out the most obvious differences between cultures without giving clues
about how, when, and how often such dimensions of belief actually shape peoples talk and
other actions. Based on one conversation with one family, this contrast cannot be a welldeveloped account of specific cultural themes that distinguish Ms cultural background
(U.S. middle class) from Fs (Colombian middle class). Nonetheless, this contrast is a
readily hearable TFG in this exchange. Its presence suggests that with a collection of talk,
a catalogue of specific instances that draws upon similar as well as discordant notes to
establish cultural patterns (When does the father urge self-reliance? When does the mother
offer participation as assistance? To what extent are the varied instances part of a system,
and how may that system be described?) could be the basis for a more nuanced picture of
contrasting cultural premises.
The theoretical contribution of the TFG concept is to suggest that such description would
necessarily be grounded in examination of implicit messages. The theoretical contribution
of cultural explanations such as this one is to suggest that there is a system of meaning
there to be discovered: Although implicit and subtle, cultural premises can be discerned in
everyday talk, and often become most visible when they come in contact with a different
system of premisesas is by definition the case in intercultural communication. Interlocutors, even when they are well aware that they are interacting with someone whose cultural
premises are different from their own, are highly unlikely to make those premises explicit.
I want you to use two containers so youll become an autonomous individual, which in
my cultural belief system is the only kind of self that counts as a whole and healthy one
would have been an awkward thing for M to say in this (or any) conversation, as would I
want you to depend on me so youll learn that you are incomplete on your own, that you
need other peoples help to do anything in the world, which in my cultural belief system
is the only, and so on. The reasons why this is so, in intercultural communication as in
other kinds, are a useful point with which to conclude.
It is well known, or at least widely suspected, that dissimilar TFGs are at the heart of many
misunderstandings and disagreements between members of different speech communities.
Speakers leave those elements of talk unsaid that they presume to be shared knowledge.
When they come from different systems of belief, there are quite understandably different approaches to communicative goals, divergent interpretations of action, and other
serious muddles related to language use and meaning. If the problem were not in these
implicit, between-the-lines aspects of talk, the only distinctive feature of intercultural communication would be language differences, relatively solvable (if hardly simple) through

Taken-for-granteds 85
fluency and attention to strict accuracy of expression. The notion of TFGs emphasizes how
much of culture, and how many difficulties in intercultural communication, come down
to ideologies subtly hidden (because they are never given voice to) in and around spoken
A common route to discovering such cultural differences, and working through areas
of misunderstanding, includes prescriptions to discuss TFGs in order to make explicit that
which is unsaid. The conversation examined here, and Hoppers observation in the TFG
articles that interpretation may be forever enigmatic and incomplete to some degree, make
it clear that there is a definite limit to which explication of that kind is practical in everyday talk. Alignment mechanisms are available, ranging from those that are so indirect that
they go unnoticed by participants, to those that are quite direct, clarifying TFGs by way of
making them explicit.
It seems very likely, however, that some premises are too delicate to put into words.
They must remain unspoken for a variety of reasons. One of those is the potential face
threat involved: To make explicit something that a competent hearer could be assumed to
know calls into question how competent this particular hearer actually is. Another reason
has to do with the nature of cultural premises. Making explicit a cultural premise can paradoxically call it into question. Part of the enormous weight of cultural codes to shape action
and interpretation comes from their pervasive unspokenness. Members of a speech community hear talk in well-worn grooves of inference. When human agency (for example) is
spoken of in ways that emphasize autonomy and individual selfhood (youyouyou
you) it becomes difficult to imagine other ways the (social) world could be arranged.
Besides increasing the difficulty of articulating them (a significant factor in itself), this
customary implicitness of cultural premises increases the risks involved in holding them
up to conversational daylight. For M to insist explicitly that children must develop individual self-restraint would be to open a slot for resistance, contradiction, a counterproposal
that interdependence among intimates was a more legitimate principle to instill, and other
forms of disagreement. It is far safer to argue over procedures for reaching a particular
goal, especially when the goal itself is not questioned, than it is to debate the fundamental,
sacred symbols underlying social life more generally.
A final reason TFGs must ordinarily be left implicit has to do with a further paradox, this
one related to Grices maxims (Grice, 1975). Given speakers abilities to search indirect
utterances for meaning by inference on the basis of quantity, quality, relevance, and truth,
it is not surprising that direct utterances would be subjected to a similar kind of scanning.
Suppose that M, aware after many years of conversation with F that included discovery of
cultural patterning in the disagreements between them, had explored his proposal for the
cultural premises underlying it. Is this how Colombians teach children to manage their
money? she might have inquired. However benigneven generousthe intention behind
the question, the contrast with her own position raises a face threat similar to the competence challenge just mentioned. Regardless of the phrasing, a metacomment that attributes
the meaning of an utterance to membership in a category to which the hearer belongs and
the speaker does not carries a strong suggestion that the comparison is critical or sarcastic,
dismissive of the hearers faculties to reason as an individual, along the lines of Isnt that
just like a wo/man?

86 Studies in language and social interaction

The interpretation of competing codes I have offered here rests on cultural differences.
There are certainly other readings of this conversation that might be offered. Among those,
a gender difference perspective, in which Fs more forceful proposal supersedes Ms,
undercutting the legitimacy of her position and drawing the childrens support away from
her, is undeniably plausible. There is even room to argue that the difference of opinion
here, and the ways in which participants express their competing views, is idiosyncratic.
Sorting through these possibilities, each of which entails TFGs based in distinctive kinds
of shared knowledge, would require a great deal more examination of other conversations
between this couple and in other families, perhaps with similar configurations of cultural
background. The careful excavation of everyday talk that would entail is, appropriately
enough, Robert Hoppers signature contribution to the field.
I am grateful to the members of the Discourse and Rhetoric Group in the Department of
Social Sciences at Loughborough University for their very useful discussion of the|transcript
presented in this chapter.
Bateson, G. (1972). Steps to an ecology of mind. New York: Chandler.
Blum-Kulka, S. (1997). Dinner talk: Cultural patterns of sociability and socialization in family
discourse. Mahwah, NJ: Lawrence Erlbaum Associates.
Drew, P., & Holt, E.J. (1988). Complainable matters: The use of idiomatic expressions in making
complaints. Social Problems 35, 398417.
Goffman, E. (1974). Frame analysis. New York: Harper-Colophon.
Grice, H.P. (1975). Logic in conversation. In P.Cole & J.L.Morgan (Eds.), Syntax and semantics:
Vol. III. Speech acts (pp. 4158). New York: Academic Press.
Hopper, R. (1981a). How to do things without words: The taken-for-granted as speech action. Communication Quarterly 29, 228236.
Hopper, R. (1981b). The taken-for-granted. Human Communication Research, 7(3), 195211.
Hymes, D. (1972). Models of the interaction of language and social life. In J. Gumperz & D.Hymes
(Eds.), Directions in sociolinguistics: The ethnography of communication (pp. 3571). New
York: Holt, Rinehart & Winston.
Ochs, E., Smith, R., & Taylor, C. (1989). Detective stories at dinnertime: Problem solving through
co-narration. Cultural Dynamics 2, 238257.
Philipsen, G. (1992). Speaking culturally. Albany: State University of New York Press.
Philipsen, G. (1997). A theory of speech codes. In G. Philipsen & T. Albrecht (Eds.), Developing
Theories in Communication (pp. 119156). Albany: State University of New York Press:.
Triandis, H., Bontempo, R., Villareal, J., Asai, M., & Lucca, N. (1988). Individualism and collectivism: Cross-cultural perspectives on selfingroup relationships. Journal of Personality and Social
Psychology 54, 323338.
Triandis, H.C. (1988). Collectivism vs. individualism: A reconceptualization of a basic concept in
cross-cultural psychology. In G.Verma & C.Bagley (Eds.) Cross-cultural studies of personality.
London: Macmillan.

So, What Do You Guys Think?: Think Talk and
Process in Student-Led Classroom Discussion
Robert T.Craig
University of Colorado at Boulder
Alena L.Sanusi
University of Colorado at Boulder
This study examines certain uses of think talk (expressions such as I think and What do
you think?) in student-led classroom discussions on controversial issues. Data are drawn
from recorded discussions in several undergraduate critical thinking classes at a large,
western-U.S., public university, 19961998.
Students in this course were instructed in critical thinking techniques and participated
in practical exercises, one of which involved working in a small group to prepare and lead
a full-class discussion of a current, controversial issue such as capital punishment, sex
education, or media ethics. The official purpose of the discussion was to facilitate critical thinking on the issue, not necessarily to reach consensus. 1825 students, including
the group of 46 leaders, usually participated. A graded assignment for the leaders, the
discussions were observed and recorded by the instructor, who otherwise did not officially
participate. The leaders selected and researched an issue and conducted a 40-minute class
discussion. Background readings were sometimes assigned by the leaders in advance of
the discussion.
The discussions followed variations of a standard format. The leaders usually sat
together at the front of the classroom with other participants either facing them or completing a large circle. Usually, the leaders would open with a formal presentation, based
on their research, introducing the issue and providing background information. Often they
would then break the class into small groups assigned to discuss briefly particular questions, aspects, or points of view on the issue. General discussion, sometimes structured
around reports by small groups, sometimes structured by a series of questions posed by
leaders to the class as a whole, sometimes more free-flowing or managed by the leaders in
apparently ad hoc ways, would follow the opening presentation and/or small- group discussions. When time was up, the leaders would end the discussion, sometimes abruptly, more
often with some attempt to summarize and conclude.
Previous studies of these discussions have examined the use of critical thinking terminology to mitigate the interpersonal implications of disagreement and criticism (Craig,
1997), co-construction of the issue as a metadiscursive object and its use in presenting
standpoints and managing the discussion (Craig, 1999a, 1999b), the use of animated mock
figures in the construction of arguments (Muller, 1999a, 1999b), and a problematic transition from an opening presentation to subsequent class discussion (Sanusi, 1999). Craig and
Sanusi (2000) showed how Im just saying and related discourse markers (Schiffrin, 1980,
1987) are used by participants to constitute contributions to the discussion as expressions
of continuing, consistent standpoints on the issue.

88 Studies in language and social interaction

The present study in a sense complements Craig and Sanusis (2000) analysis of continuity markers. If participants in group discussion routinely use tokens such as Im just
saying to display their contributions as expressions of unchanging viewpoints, how do
they also display the relevance of their contributions to the ongoing process of discussion?
Group discussion involves online talktalk that responds to the current state of the
discussion and occasions further such responses by others, thus moving the process along.
How is this accomplished?
Our analysis focuses on the use of I think and related expressions as markers of online
process. I think, especially in the opening stretch of a turn at talk, can be used to indicate a
particular kind of relevance to ongoing talk that characterizes the process of group discussion. In such cases, it marks the turn in progress as a presentation of the speakers response
to the currently relevant group topic, but not necessarily as a response to anything said
about the topic by other speakers. In our data, the relevant group topic is usually the discussion issue or some more immediate question or statement presented by a discussion leader.
Discussion leaders often invite these reactions by the use of expressions such as What do
you guys think? in presenting a topic to the group. I think marks the current turn as one in a
possible series of different individual reactions to the topic. As such, the current turn may
not be relevant at all to the immediately preceding turn, except by virtue of being the next
in a series of expressions of opinion by different members of the discussion group.
The following sections present data illustrating how think talk is used by discussion
participants to index their own statements of opinion as expressions of online thinking situated in the ongoing discussion, to mark transitions between canned and online discussion
and invite expressions of online thinking from other participants, and finally, to maintain a
sense of process when process seems threatened by a lack of potential for controversy on
the topic. In a concluding section, we reflect on the implications of this analysis for understanding the semantics and pragmatics of I think, and for further studies of interaction in
classroom discussions and related institutional settings.
Example 1, an excerpt condensed from a transcript of a discussion of sex education, exemplifies the density of I thinks that can occur in such discourse.
(1) (Condensed)


I think that (.) um: (.) well two things. One I
think we can rely on family structure, our family
structure, (I think) in the United States is (.)
really screwed up. ((about 13 lines deleted))
if youre ge- an I dont think that- that way
youre not attacking (.) religious or moral values
but (.) well. I- I guess in a way but at least

So, What Do You Guys Think? 89


regardless of whether or not you believe in it


still have ta (.) accept that its out there. (And

you dont like it.)

Shelley: I think just even more so going on with that I

think that (.) um people are so opposed to things

like homosexuality is because- (.) I mean- like

like out theyre just ignorant about it I mean they

think that (.) you know its all abour:t (.) just

sex or whatever an- (.) yknow but its more than

that. Its about like relationships and things like

that an- you know like every other heterosexual

relationship and I think that (.) things like that

to be taught are important for people are (.) you

know, a-homosexual but theyre too afraid to


out and this will give them (.) like a way ta- to

have better self-esteem about themselves and


better about themselves.

So (.) ()
Jennifer I was gonna say I think its of importance that

they define things like abortion and homosexuality

because we probably all knew somebody in high

school whos just totally naive: (.) about

everything. You know, and, whether we were


people or (.) we- were friends with those people

there are people in (.) high school that are

totally nave, to what those things even mean.

Really. You know, and I think its really important

that not necessarily they push one side or the

other. But that they really define them so students

know. What (.) it means? (.) What it is.

( ?)
Brooke: Na- I was just wanted to say that to say that uh-I

agree with that I think that like- things like um:

(.) maybe:: if- if someon- if you have to write a

paper or like (to have) a presentation like this.

Stuff like that in high school I think is really

90 Studies in language and social interaction



important but you cant really just teach you

you can define but- (.) w-where would you even
begin to teach about (.) abortion when theres so
many different sides? But I think it is important
to do stuff like (.) you know have an opportunity
like this ((turn continues))

In this stretch of discussion, I think is used to mark each turn as one in a series of expressions of opinion on the topic of sex education. Prior to this segment, a discussion leader
(Tad) posed the question (Wul/But) do you guys think like (.) heated topics like homosexuality:, and abortion:, should be something that (.) the government should be able to::
(.) talk about? After some elaboration by Tad (seven lines) and some brief transitional
business, Jack is called on by a leader (Line 1) and takes the floor (Line 2). Initiating his
turn with I think, Jack argues, in what can be heard as an implicit response to Tads question, that sex education in public schools is necessary because I dont think we can rely on
family structure, and that topics like homosexuality should be included because whether
or not you believe in it you have to accept that its out there.
Also (like Jack) opening with I think (Line 12), Shelley begins speaking and positions her turn as the next in a sequence by contrasting it with the preceding (I think just
even more so going on with that I think).1 Even more so going on with that indicates
broad agreement with the preceding turn and that the following remarks will extend the
preceding topic in some unspecified way. The following I think, however, shifts the
topical focus from Jacks view back to Shelleys. Her subsequent contribution, although
it continues Jacks just preceding topic of homosexuality, does not respond to his specific
points. Rather, in arguing that teaching about homosexuality is needed because the reasons
for opposing it are based on ignorance and it will help homosexual students to feel better
about themselves, Shelley can be heard as presenting another view in response to the question earlier posed by a discussion leader (Tad).
Two following speakers in succession, Jennifer and Brooke, present further views on
the topic of teaching about homosexuality in schools. Each opens her turn by marking a
temporal disjuncture (I was gonna say [Line 28]; Na-I was just wanted to say [Line
41]) between the sequential position of her turn and the unspecified recent moment in the
discussion that immediately occasioned what she is about to say. Each speaker prefaces
her contribution with I think. Each continues topical threads of previous turns (a prosex education stance, societal ignorance of homosexuality, the value of specific learning
experiences) while not responding to specific points made by previous speakers.

The transcript, based on an audio recording, omits nonverbal behaviors that were available to the
participants. It is quite possible that Shelley did not self-select as next speaker but was nonverbally
selected by Jack or a discussion leader in response to her nonverbal bid for the floor (such as a
raised hand). The omission of possibly relevant nonverbal details, especially in turn transitions,
should be kept in mind when reading these transcripts.

So, What Do You Guys Think? 91

In the linguistic literature, I think appears most regularly in discussions of modality
and hedging (e.g., Coates, 1987; Galasinski, 1996; Turnbull & Saxton, 1997). In Example
1, however the phrase does not appear to be used primarily as a hedging or downtoning
device. Note the appearance of the word important/ce in the propositions with parenthetical I think in the following excerpts:
I think that (.) things like that to be taught are important for people are (.) you know,
a-homosexual (Lines 2022).
I was gonna say I think its of importance that they define things like abortion and
homosexuality (Lines 2931).
I think its really important that not necessarily they push one side or the other.
But that they really define them so students know. What (.) it means? (.) What it is.
(Lines 3639) .
Stuff like that in high school I think is really important but you cant really just
teach you know, you can define but- (.) w-where (Lines 4547).
But I think it is important to do stuff like (Lines 4950).
It would seem unlikely that a speaker would be downtoning a proposition for which she
is claiming importance. In these cases, the I think might be understood as contributing
modality, that is, a sense of speaker commitment, in a fairly literal way, as a lexical verb.
More noticeable to us, however, is the flowing quality of the talk and how I think is used
to mark each in a series of expressions of opinion. The opinions presented in Example 1 are
all generally favorable to sex education but are otherwise diverse. Each turn links sequentially and topically to prior turns but does not primarily build on or respond to previous
speakers opinions. Each advances the discussion primarily by contributing the speakers
own view on the current topic.
Notably, there is little to suggest that speakers in this free-flowing segment of discussion are expected to build tightly on one anothers comments any more than they actually
do. Their expressions of opinion are not markedly hesitant, apologetic, or qualified; they
generally display the features of preferred rather than dispreferred turn shapes (Pomerantz, 1984). In short, there is no evidence that participants in this stretch of discussion,
in expressing whatever opinions on the topic happen to occur to them in the moment, are
doing anything other than what they apparently ought to be doing in the situation.
I think seems to function in such routine, unproblematic stretches of talk neither especially to modify illocutionary force (either to boost or downtone), nor to express politeness
or deference, but rather primarily to mark contributions as expressions of online thinking
within the discussion process. I think indicates a relevant response to the current discussion
topic but also licenses a certain topical disjunctive as the discussion jumps from one individual point of view to another. I think marks an expression of opinion on a shared topic but
from an individual point of view, distinct from succeeding and following points of view,
occasioned by the current state of the discussion. It marks process.

92 Studies in language and social interaction

In Example 2, students deal with the problem of speaking on behalf of a group rather
than on ones own behalf. Talk produced while a student is speaking as a spokesperson
is noticeable for its very lack of I thinks. The only I think in Sallys (the spokespersons)
talk is the initial I think, whose syntactic parallelism with Jacks question marks her talk as
designed as a response to his question and what follows as her sense of the discussion that
had occurred in the group in which she was participating.


I hope weve kinda outlined each one of these

(.) conclusions an hh reasoning why I asked

what yguys think (.) whether you guys are on


opposite sides or (.) bring up points and also y

know- y go specifically into the (.) examples the

Ramsey case or (.) any uh these, hhh what do you

guys think like the Globe pictures that were

published, (.) pt what do you guys think of that.

Sally: I think right now I mean theres like a little

bit of miscommunication cause our group (.)


dealt with (.) with issues [that were:-] that

Sally: =false or exaggerated or you know like whats

taken out of context and those

things like I would say you know (.) sure I- Id

like to know about things going on in the press


long as theyre true n that (.) n- n yes I agree

the public has the right to know I agree with

a lot

of those things as far as the First Amendment is

concerned .hhh but we were more focussing


werent we?
Mm hm.=
=((murmurs [of agreement))]
Sally: [focusing on that?] w(h)e(h)e

we(h)e(e)re focusing on, .hhh like taking things

So, What Do You Guys Think? 93


out of context an- an- (.) I mean- making up

stories or yknow or compensating for a lack of
facts (then,)(.) but

In Example 2, Jack asks other participants what they think as a way of transitioning from
the presentation of canned (prepared) material to open discussion. Similar uses of What do
you think? or What do you guys think? occur quite frequently in our data (also see Example
5, later). Such expressions invite reports of online thinking in reaction to some stimulus.
In this case, Jack is speaking as one of a group of students who have been assigned to lead
a discussion of media ethics. The group has just finished presenting a series of reasons for
and against increasing restrictions on the press. The presentation has been prepared by the
group members as a summary of breakout small-group discussions that they led, in order
to serve as a stimulus for a discussion that is now to occur among the entire class. Think
is used by Jack to mark this transition from canned to spontaneous discussion by inviting expressions of opinion, and then by Sally, marking her turn as a response to Jacks
Sally goes on, however, to explain that she is unable to make an acceptably relevant
response. Her small group, she says, didnt focus on the kinds of examples Jack just
referred to, so she has no thoughts to express on those issues. Sallys extended turn at talk,
including a short stretch of side talk with members of her group, is interesting in a number
of ways. Our main interest at this point in the analysis is what her response suggests about
the interactional properties of What do you think?
What do you think? invites expressions of opinion in reaction to something presented or
indicated by the speaker. It projects nothing about the contents of the reactions except that
they will be reactions to that something and that they will be reactions, that is, expressions
of online thinking from presumably differing individual points of view. Sallys account in
Example 2 displays not only her online reaction but also her awareness that the thoughts
invited by Jack are expected to be reactions specifically to the kinds of examples he has
Sallys talk in Lines 1029 also provides evidence that What do you think? invites
online rather than canned expressions of opinion. Sally indicates in a variety of ways that
she is responding to Jack not just as an individual but on behalf of her small group, but this
spokesperson role becomes rather problematic for her. If think is about in-process reactions
rather than canned presentations, then reporting on behalf of a group becomes a potential
trouble point, because the groups thoughts are either previously agreed upon (canned) or
have to be negotiated on the spot, which would require a frame shift or time out from
discussion in the larger group. Sallys orientation to this problem is reflected in her tense
shifts, her use of I think to mark her immediate thoughts and her nonuse of it in other contexts, her explicit checking with the group that she is representing their views accurately
(Lines 2122), and her laughed speech in Lines 2526.
Example 3 is a similar exchange from a discussion on health insurance reform, in which
M initially tries to speak on behalf of us guys (implying canned talk, i.e., the presentation of a view already discussed within the group) but quickly resorts to I think, marking
his talk as spontaneous.

94 Studies in language and social interaction





How do you guys feel about that.

um how do we feel about that ((several people
laughing softly)) (.) oh we we think its a
responsibility and stuff but I think that seeing
from other countries trying to do this and seeing
how it has it hasnt had any positive effects
and p
then () bring it back to our our current plan I
think that ah its just it will be more abused
than it will be used ((turn continues))

M responds to Fs question, directed to his small group, as to how they feel about the
moral responsibility to provide health care to people in need, with talk laden with signs of
trouble including false starts, hesitation, and side interaction (including soft laughter) with
members of his group. His response is first marked by we feel (arguably displaying himself
as cooperative in that he takes up Fs focus on how you guys feel; Galasinski, 1996), then
shifts (still uneasily, with a false start) to the less evaluative we think and then to I think.
In all of this, M displays the awkwardness of expressing an online group reaction as he
retreats to (and proceeds with) expressions of his own individual online reaction.
Examples 13 have illustrated some ways in which think talk is used in contexts where
the processual aspect of discussion is threatened or needs to be emphasized, including:
sequences of diverse, loosely related opinions, transitions from canned talk to open discussion, and shifts from speaking on behalf of a group to speaking as an individual. We now
turn to evidence for yet another threat to process in the context of a class discussion: lack
of controversy.



Do you guys?- Wul it seems- wul it seems to me

tha:t like most of us agree: an- you know we all
kind of we agree that (.) people need to be aware,
and- and its gonna happen anyway whatever, (.)
do you guys think that its such: (.) like a
controversial thing among parents. I mean

Example 4, from the same sex education discussion as Example 1, illustrates another environment that can constitute a threat to the discussion process, namely agreement. Agreement among all participants in a discussion can threaten the continuation of the discussion
process, because once the group has reached a point of agreement, unless the group is
following a prepared agenda, there may be nothing immediately at hand to discuss. As we

So, What Do You Guys Think? 95

can see in Example 4, a discussion leader may invite further discussion by marking points
of agreement, presenting a new question or item of information, and inviting reactions with
some variation of What do you think?
When a topic for discussion threatens to be uncontroversial, leaders may do considerable work to mark it as open to various opinions and therefore potentially controversial
when inviting opinions, as in Example 5.
(5) Condensed
1 Emily: So. But. Urn. I guess wed like to know, what you
guys think about, whether [or not sex
3 Tad:
[Before- be- before we
wanta know what you guys think
5 ?:
mm hh-huh-huh
6 Tad:
I just wanted to make sure that- Imean wer- wer
were presenting ourselves (.) in sort of a biased
standpoint because we did the whole condom exercise
and we passed out condoms an (.) we talked about
how condoms should be distributed in school, so,
may: be: showing you guy:s that that were for: sex
e:d, which we are, (.) but (.) sex ed in schools is
a different thing, bcause, I think
((turn continues, 30 lines deleted))
(.) but this kind of gives like a: just a gradual
leap into the actual talking about sex and uh what
goes on in (childrens lives)?
19 Emily: (Well?) So! What dya guys think.
2 0 Tad:
We were going to break you guys into groups but
we figured that thats been overworked a lot in
this class? so we figured we have a kinda
24 Emily: =( )=
25 Tad:
=have a open? (.) discussion? (.) and maybe just
(.) what you guys feel? because I mean my- my
views: (.)
28 F?:
I know:
30 Tad:
have changed a little bit, as far as (.) sex
education should be taught in schools because I
dont think it is up to the government to (be

96 Studies in language and social interaction



forcing) something like this, (.) I think its

more parents and more religion, (.) then) (.)
Alright![ ]So whataya guys think!

Example 5 occurs at the end of a long introductory segment in which Emily, Tad, and
other members of the group assigned to lead this class discussion on sex education have
presented a large amount of canned information about the topic. Emily at Line 1 initiates a
transition from canned presentation to open discussion with a markedly hedged expression
of interest in what you guys think. Tad interrupts and proceeds to talk at some length,
emphasizing that, even though the leaders are presenting ourselves in a sort of biased
standpoint in favor of sex education, the topic really is controversial. (Thirty lines have
been deleted from Tads long turn.) At Line 19 Emily recycles the question, more confidently than before, but Tad again interrupts to report that the leaders have decided against
using an overworked approach, that open discussion is wanted, and, again, that the topic
really does warrant discussion as evidenced by the fact that Tads own views have changed
a little bit (Line 30). What is interesting about Tads I think and I dont think is less his
attitude as speaker toward his own propositions, but the way he puts forward those markers
of thought-in-process in the service of presenting something to be reacted to in a situation
that threatens an end to the talk. Emily then repeats her question a second time, still more
emphatically than before, after which open discussion begins (finally!). In this segment,
Tad displays great concern to repair a condition that may render Emilys invitation to open
discussion unsuccessful, the condition being that the topic is uncontroversial. Think talk
again is occasioned by agreement as a threat to the discussion process.
In summary, we have described several ways in which think talk is used to index online
thinking and expressions of opinion in the discussion process. I think can be a token that
a speaker uses to bypass conditional relevance of her contribution to the immediately preceding talk in favor of the contributions relevance as one of a series of individual reactions
to a leaders question. Think talk also can be used to mark transitions between canned and
open discussion, and to maintain a sense of process when the potential for further discussion on the topic seems threatened. We conclude with a brief discussion of some possible
implications in regard to the semantics and pragmatics of I think, and further research on
classroom discussion and related forms of institutional discourse.
I Think, Modality and Politeness
Our analysis finds that think may function as meta-talk (Schiffrin, 1980) in that it focuses
attention on the status of the talk. However, it does so in ways that have little to do with
the semantics of the verb think and more to do with a need to display what kind of talk
it should be taken to be. It is only by freeing ourselves from the expectation that words
contribute semantically rather than pragmatically that we can see the way these phrases, in

So, What Do You Guys Think? 97

the examples we have presented, have been put to a metadiscursive task that is not clearly
predictable from the semantics of the words.
In the linguistic literature, as we noted earlier, I think most often appears in discussions of modality and hedging (e.g., Coates, 1987; Galasinski, 1996) or of modality as it
functions to do facework (e.g., Brown & Levinson, 1987; Turnbull & Saxton, 1997). As
Holmes (1984) pointed out, these represent the two primary reasons why a speaker would
want to modify the illocutionary force of a speech act: to convey modal meaning or the
speakers attitude to the content of the proposition, and to express affective meaning or the
speakers attitude to the addressee in the context of the utterance (p. 348). In these functions, I think seems to draw on the implications (e.g., rationality, intentionality, and their
social implications) of the semantic content of the lexical verb think, so that a speaker using
I think would be referring to and characterizing a consciously held opinion or intent.
Curiously, I think resists categorization as either a downtoner or a booster (Holmes,
1984) of speaker commitment because, like I believe, depending on its intonation pattern,
it may either hedge or intensify/highlight the speakers commitment to her utterance. Perhaps this ambiguity accounts for why I think is absent from Schiffrins (1987) discourse
marker model of discourse, although semantically, as a lexical verb, it would seem to have
a place with such discourse markers as you know and I mean, whose verbs appear to refer
to a particular way of holding a kind of cognitive entity.2
However, Schiffrin (1990) noted that although modality, or speaker commitment to a
proposition, may be marked in a variety of (perhaps redundant) ways both linguistically
and metalinguistically, there may in fact be no such marking at all. This of course raises the
question of whether the overt marking of modality through I think might have some other
discursive or communicative functions. Given that it can usually be taken that what one
says is what one thinks (as can perhaps be inferred by the fact that what is usually explicitly marked is deviation from that expectation), why bother to explicitly metalinguistically
mark the expected?
We recognize that I think no doubt often participates in meaning making through the
semantic contribution of the verb think (as Schiffrin, 1987, in fact argued to be the case for
you know and I mean), but we would like to suggest that I think may also be used in a way
that is quite independent of the semantic content of the verb. We suggest that there is evidence in our data that I think may be doing interactional work much in the fashion of tokens
like oh (Heritage, 1984, 1998) and okay (Beach, 1995), which do important interactional
work despite their scant semantic content. If we can see I think as a relatively semantically
empty token that might be doing some work besides referring to and characterizing mental
states, the question arises what that work might be. It appears to us that I think, particularly
in its (sequentially more or less distant) relationship to a discussion leaders question posed
roughly in the form, What do you guys think?, points to the speakers making this contribution as a second pair-part made relevant by that question. As such, I think also serves to
Interestingly, though, Schiffrins structuralist approach to discovering the meanings of you know
and I mean through examining their complementary functions roughly parallels our own approach:
whereas Schiffrin (1987) found that you know and I mean work to shift orientation between speaker
and hearer, we argue that Im (just) saying and I think can serve complementary functions of maintaining personal standpoint continuity and keeping the discussion going.

98 Studies in language and social interaction

disconnect the current contribution from any interpretation in terms of the immediately
preceding turn. In doing so, I think seems to be a participants resource for the performance
of discussion, where crafting ones contribution in terms of the content and form of the
previous utterance is less useful than crafting ones contribution (via I think) as just one
more answermy answerto a leaders topic-based question.
In claiming that I think is a participants resource for doing this kind of interactional
work, we are not claiming that this is the only function that I think prefacing can perform,
nor are we claiming that no other mechanisms exist for accomplishing this same interactional task. Our claim is the limited one, that this is one interactional function that I think
appears to be performing in the data we have examined.
Implications for Further Research
This research should be extended to examine the functions of other markers of process and
continuity in the conduct of classroom discussions. We wonder, for example, what might
be the distinction, if any, between I feel and I think, which alternate in interesting ways in
our data (see Example 3; Example 5, Lines 2635).
The significance of preliminary observation is that it opens up the possibility of research
that focuses on communicative problems in particular settings. In these student-led classroom discussions, one salient problem is to maintain the flow of talk as required both to
fill time and to cover ground for purposes of evaluation by the teacher. Rather different
problems may be expected to be more salient in group discussions oriented to arriving
efficiently at a consensus or decision for purposes of action. It would be interesting to see
whether I think is often used in school board meetings or jury deliberations, for example,
in ways we have noted in these classroom discussions. In these other settings, no less than
in classroom discussions, discussion process as well as continuity of standpoints in the
expression of opinions is surely important, but the process demands of the other settings
are probably quite different.
Within the classroom setting itself, the interactional problems of teacher-led discussions undoubtedly differ from those apparent in our student-led discussions. Whereas participants in our discussions are of essentially equal status, even though some perform a
differentiated role of discussion facilitator, teacher-led discussions involve a more marked
differentiation of power and authority. Huspeks (1989) study of the differential use of I
think and you know in working-class speech suggests one interesting point of comparison.
The preponderance of I think and the relative scarcity of you know markers in our data
contrasts with the overwhelming preponderance of you know sequences in the speech of
Huspeks working-class respondents. The pattern of our data is consistent with the predominantly middle-class status of our students. It may also reflect the lack of salient power differences in these student-led discussions along with the situational demand for participants
to assert their individual opinions (indexed by I think), whereas in other circumstances they
might do more to highlight common ground (indexed by you know). These possibilities can
be explored empirically in studies of different groups and settings, including teacher-led
classroom discussions.
So, what do you guys think?

So, What Do You Guys Think? 99

An earlier version of this chapter was presented at the annual convention of the International Communication Association, San Francisco, CA, May 29, 1999.
Beach, W.A. (1995). Conversation analysis: Okay as a clue for understanding consequentiality.
In S.J.Sigman (Ed.), The consequentiality of communication (pp. 121161). Hillsdale, NJ: Lawrence Erlbaum Associates.
Brown, P., & Levinson, S.C. (1987). Politeness: Some universals in language usage. Cambridge,
England: Cambridge University Press.
Coates, J. (1987). Epistemic modality and spoken discourse. Transactions of the Philological Society, pp. 110131.
Craig, R.T. (1997). Reflective discourse in a critical thinking classroom. In J.F. Klumpp (Ed.),
Argument in a time of change: Definitions, frameworks, and critiques (Proceedings of the Tenth
NCA/AFA Conference on Argumentation) (pp. 356361). Annandale, VA: National Communication Association.
Craig, R.T. (1999a). Metadiscourse, theory, and practice. Research on Language and Social Interaction, 32, 2129
Craig, R.T. (1999b, July 30). The issue as a metadiscursive device in some student-led classroom
discussions. Paper presented at the Eleventh AFA/NCA Summer Conference on Argumentation,
Alta, UT.
Craig, R.T., & Sanusi, A.L. (2000). Im just saying: Discourse markers of standpoint continuity.
Argumentation, 19, 425495.
Galasinksi, D. (1996). Pretending to cooperate: How speakers hide evasive actions. Argumentation,
10, 375388.
Heritage, J. (1984). A change-of-state token and aspects of its sequential placement. In
J.M.Atkinson and J.Heritage (Eds.), Structures of social action (pp. 299345). Cambridge, England: Cambridge University Press.
Heritage, J. (1998). Oh-prefaced responses to inquiry. Language in Society, 27, 291334.
Holmes, J. (1984). Modifying illocutionary force. Journal of Pragmatics, 8, 345365.
Huspek, M. (1989). Linguistic variability and power: An analysis of YOU KNOW/I THINK variation in working-class speech. Journal of Pragmatics, 13, 661683.
Muller, H.L. (1999a, Noverber 6). Creating expectations of appreciation by animating mock figures. Paper presented at the annual convention of the National Communication Association,
Muller, H.L. (1999b, July 30). Hypothetical examples in student arguments: Animating mock and
cited figures. Paper presented at the Eleventh AFA/NCA Summer Conference on Argumentation,
Alta, UT.
Pomerantz, A.M. (1984). Agreeing and disagreeing with assessments: Some features of preferred/
dispreferred turn shapes. In J.M.Atkinson & J.C. Heritage (Eds.), Structures of social action:
Studies in conversation analysis (pp. 57101). Cambridge, England: Cambridge University
Sanusi, A.L. (1999). Maintaining formation: An instance of frame transition. Paper presented at the
November 2000 annual convention of the National Communication Association, Seattle, WA.
Schiffrin, D. (1980). Meta-talk: Organizational and evaluative brackets in discourse. Sociological
Inquiry, 50(34), 199236.
Schiffrin, D. (1987). Discourse markers. Cambridge, England: Cambridge University Press.

100 Studies in language and social interaction

Schiffrin, D. (1990). The management of a co-operative self during argument: The role of opinions
and stories. In A.D.Grimshaw (Ed.), Conflict talk: Sociolinguistic investigations of arguments in
conversations (pp. 241259). Cambridge, UK: Cambridge University Press.
Turnbull, W., & Saxton, K.L. (1996). Modal expressions as facework in refusals to comply with
requests: I think I should say no right now. Journal of Pragmatics, 27, 145181.

Gesture and the Transparency of
Curtis D.LeBaron
Brigham Young University
Timothy Koschmann
Southern Illinois University
Most research on language and social interaction (LSI) has been decidedly action focused,
more concerned with what people do (i.e., vocal and visible behaviors) and how they do it
(e.g., through mutual orientation and coordination), less concerned with subjects possible
cognitive states (e.g., intentions, motivations, and understandings). Conversation analysis
(CA) especially has been touted as an empirically rigorous alternative to mentalistic perspectives that regard language as a way to study underlying psychological states, structures,
and competencies. Robert Hopper (1997), for example, described himself as a cognitive
agnostic. Though not denying the existence and potential importance of cognition, he
insisted that researchers should distinguish between calculated speech and most social
interactiondistinguish what actors do from what theorists may infer (p. 6). Hoppers
agnostic stance was consistent with CA as it has generally been described and applied. Heritage (1990/1991) observed that conversation analysts have sought, wherever possible, to
avoid a terminology of social action that invokes mentalistic predicates and thereby anthropomorphizes processes that may be less anthropomorphic than we conventionally believe
(pp. 328329). (See also Heritage, 1984; Hopper, 1989, 1990, 1992; Hopper, Koch, &
Mandelbaum, 1986; Hutchby & Wooffitt, 1998; Jacobs, 1988; Levinson, 1983; Pomerantz,
1990; Psathas, 1995.)
We agree with and indeed celebrate the efforts of Hopper, Heritage, and others to place
CA work on a rigorous foundation, one that does not allow ungrounded speculation with
respect to interactants hypothesized states of mind. In our own studies of classroom interaction (cf, Koschmann, Glenn, & Conlee, 2000; LeBaron, 1998; LeBaron & Koschmann,
1999; LeBaron & Streeck, 2000), however, we have been brought to examine how participants avow and ascribe mentalistic predicates to themselves and to others in the course of
their joint and ongoing learning activities. A question for us, therefore, has been how can
we as analysts document the practical methods by which these activities are accomplished
without abandoning the standards of warrantability set forth by the founders of our field?
Such questions and issues have already been raised by other researchers, who have
proposed discursive or praxiological approaches to the study of psychological matters.
Social psychologists with an interest in discourse and conversation analysis (e.g., Edwards,
1997; Edwards & Potter, 1992), for example, have considered cognitive phenomena
through detailed study of talk-in-interaction. Rather than treating cognition as prior to,
and separable from, interaction, it is treated as something that is managed in, constituted in,
and constructed in interaction (Potter, 1998, p. 35). Some conversation analysts working

Gesture and the transparency of understanding103

within the ethnomethodological tradition (e.g., Coulter, 1990; Lynch & Bogen, 1996) have
regarded cognition as largely public and observable rather than purely private and mental.
Coulter (1979) observed that members of a culture mundanely traffic in cognitive categories and predicatesand have practical ways of making subjectivity-determinations
(p. 51). He was early to propose a program of research to determine howon the basis
of what culturally available reasonings and presuppositionsdo members actually avow
and ascribe mental predicates to one another? (p. 37). An example of one such mental
predicate is the verb to understand. Following the work of ordinary language philosopher Gilbert Ryle (1949), Coulter observed that understand is not a process-verb like play,
but rather a terminusverb like win. An avowal that one understands, therefore, does not
describe a temporally-extended course of action (p. 37), but instead serves to mark out a
success-claim (p. 37). Coulter noted that Wittgenstein joined Ryle in treating understanding as other than a private, mental experience. Wittgenstein (1953/1968) wrote with regard
to an individuals claim to understanding, One might rather call it a signal; and we can
judge whether [the claim] was rightly employed by what he goes on to do (para. 180).
Coulter concluded, The criteria for understanding, for having understanding, cannot be
private, inner mental or experiential states or processes, but must be scenic (p. 39).
We use the phrase transparency of understanding to suggest that participants understandings within classrooms (and we think other settings) may be publicly performed,
sequentially organized, made available for others (and analysts) inspection, altogether
accountable1 . The scenic features whereby understandings are enacted are not restricted
to the linguistic, but include the mediation of artifacts, situated practices of inscription,
and various embodied forms of communication. Communicating bodies arguably have
primacy over talkbodies of understanding may occupy and move within social space,
appearing first and lingering long after a conversation has died. In the data presented here,
a transparency of understanding is interactively achieved through recurring hand gestures
that are coordinated with talk and other body movements in understandable or recognizable (Sacks, 1965/1992, p. 226) ways. Our approach to studying gesture and human
understanding should not be confused with earlier work of a psycholinguistic bent (e.g.,
Goldin-Meadows, Alibali, & Church, 1993; McNeill, 1985, 1992) that treated gesture as
a window into cognitive processes. Instead, we have adopted a microethnographic perspective that draws upon the traditions of CA and context analysis (cf. Kendon, 1990)
to explore how gesture contributes to shared forms of understanding as an interactional,
rather than cognitive, achievement.
Our videotaped record shows eight people involved in a problem-based learning (PEL)
exercise (Koschmann, Kelson, Feltovich, & Barrows, 1996) associated with a medical
school in the midwestern United States. The participants were divided into two groups
that communicated via a video-conferencing system. Although physically separated by
We use the term in the sense suggested by Garfinkel (1967), that is, Any setting organizes its
activities to make its properties as an organized environment of practical activities detectable,
countable, recordable, reportable, tell-a-story-aboutable, analyzablein short, accountable
(p. 33).

104 Studies in language and social interaction

approximately 100 miles, the two groups were virtually brought together as one televised
image that all participants could see and hear (see Fig. 6.1).
A faculty coach and three medical students were seen in the picture-in-picture (PIP)
window on the lower right of the screen. The four students shown in the full screen were
enrolled in a nursing program. At both locations, participants sat in a semi-circle around a
large table so that they could easily orient toward each other, toward a common workbook
(i.e., their medical case study), and toward the video-conferencing equipment (camera and
monitor) that enabled communication with the other group.

Fig. 6.1: Four nursing students, three medical students, and one faculty coach participate in a problem-based learning exercise, via a videoconferencing system.
Within this educational setting, participants were routinely called upon to display their
medical knowledge. One task facing the students was to interpret what their workbook
said. The medical students, who had not yet had any clinical experience, looked to the
nursing students to explain various clinical terms and concepts found in the workbook.
Typically, a medical student asked a question and one or more nursing students provided
an impromptu answer. Such knowledge displays2 were usually marked by hesitations,
restarts, silences, colloquial speech, self-repair, and other features typical of explainingin-themoment (Crowder, 1996). Moreover, knowledge displays were often interactive
accomplishments. That is, respondents often failed to complete their answers alone and
instead paused, shifted gaze, changed body orientation, or gestured toward another person,
Koschmann, et al. (2000) observed a recurrent structure within PBL exercises that they termed a
knowledge display segment (KDS), defined as a topic-delimited segment of instructional discourse in which participants raise a topic for discussion and one or more members elect to display
their understanding of that topic (p. 55).

Gesture and the transparency of understanding105

and thereby invited (or at least created opportunity spaces for) others to collaborate in the
knowledge display.
For instance, at one point during their discussion, the students came across the term
thrills and one of the medical students (Jack) asked a question that some of the nursing
students (e.g., Bill) elected to answer. The moment has been transcribed as follows (a
complete transcript appears at the end of this chapter):
1 Jack:
3 Bill:
5 Susan:
8 Bill:

Can you defi:ne thrills

Thrirll is what you fee:l (.) like is: (.)
ya could- (0.4)
If- if you happened to have uh huge murmur
(0.4) you could put your hand on (your)

the upbeat

Although Jacks utterance was ostensibly a closed question (which could have been
answered with yes or no), Bill treated it as a prompt to display his knowledge by
providing a definition of thrill. As Bill began speaking (Line 3), he also raised his left
hand and began gesturing (see Fig. 6.2). With his hand elevated and hence made available
for others view, he repeatedly wiggled the fingers of his left hand. By coordinating this
gesture with the lexical affiliate feel (Line 3) Bills gesture was recognizable as a tactile
representationthat is, his moving fingers were performing the behavior or experience of
feeling with the hand.

Fig. 6.2: Bill attempted to define the word thrill

106 Studies in language and social interaction

However, Bill failed to complete a coherent response alone. He did not produce an utterance that was hearably complete. As the transcription (Line 3) shows, he repeatedly paused
during his turn at talk, and he restarted his utterance to change the trajectory of his explanation. His first restart was marked by the words like is(Line 3); a second restart occurred
with the words ya could (Line 4). Moreover, each of these restarts was coordinated with
a shift in the shape of his gesturing hand. When Bill said, like is, his fingers stopped wiggling and came together in a rounded shape. When he said, you could, Bill moved his
left hand down and scratched the side of his neck, and his eye gaze simultaneously shifted
away from the monitor and down to the workbook, withdrawing from the interaction (see
Fig. 6.3). Thus, Bills knowledge display came up short: His hand gesture dissolved into a
neck scratch at the same time that his talk was suspended and his eyes dropped.

Figure 6.3: Susan performs a heart murmur gesture.

Susan (on Bills right) picked up where Bill left off. By repeating the word feel (Line 7),
Susan made her talk recognizable as a continuation of the knowledge display Bill initiatedthat is, he also used the word feel (Line 3). However, Susans utterance was more
hearably complete. The syntactic and prosodic structure of her talk indicated a transition
relevance place (Sacks, Schegloff, and Jefferson, 1974) after the words feel it (Line
7). Moreover, her utterance was coordinated with a recognizably coherent gesture. At the
beginning of her utterance (with the words if- if), she lifted her right hand to her chest,
locating it where a heartbeat might be felt. .With the words you could (Line 6) she lifted
her flattened hand a few inches from her chest (see Fig. 6.3) and then returned her hand to
her chest. Altogether, Susan performed a hand-felt heartbeat (albeit exaggerated). Notice
Bills alignment with Susans behavior. Bill collaboratively completed Susans utterance
with his words feel the upbeat (Line 8). His collaborative completion evidenced that he
heard and understood her description sufficient to complete it in overlap with her. Moreover, Bills own hand movements changed to correspond with the gesture that Susan now

Gesture and the transparency of understanding107

performed. After scratching his neck, Bill looked toward the monitor where Susans hand
was visibly flattened against her chest, at which point Bill lowered his hand toward his own
chest and spread his fingers in flattened form (see Fig. 6.4).

Fig. 6.4: Bill flattens his hand after looking toward Susans.
Bill performed a gestural shape in conjunction with Susans production. Through such
vocal and visible displays of alignment, Bill showed that Susans performance was an
appropriate continuation of the knowledge display that he had initiated.
Continuing their response to Jacks question about the term thrills, the nursing students
further coordinated their vocal and visible behaviors. After collaboratively completing
(Line 8) Susans description, Bill elected to continue:


Its like (.) flui:d thats getting caught on
somethin and its (.) twisting arou:nd the
vessel or

or whatever

14 Jean:
15 Bill:

Its tur bulence yeah

After a brief silence (Line 10), Bill added to the talk about the term thrill (Lines 11 through
13), but again failed to produce a coherent explanation that was hearably complete. Notice
the form and content of his talk: An utterance-initial hedge (its like), followed by hesitations (pauses) and nondescript words (somethin and whatever), came together in a
rather odd narrative about blood within the heart getting caught and twisting around
action words not usually associated with fluids. Nevertheless, Bills vocal behaviors were
coordinated with a hand gesture that was evidently consequential. With his index finger

108 Studies in language and social interaction

extended, he rotated his left hand in the air to iconically represent the movement of fluid
within a chamber (see Fig. 6.5).

Figure 6.5: Bill performs a turbulence gesture.

Jean (on Bills left) watched his gesture (see Fig. 6.5) before speaking the word turbulence
(Line 14). By speaking in overlap, Jean participated in Bills knowledge display. By speaking only after Bills gestural performance but before the end of his utterance, Jean showed
recognition of Bills embodied actions. Whether or not Bill was searching for the word
turbulence, Jean provided it (Line 14) and Bill then repeated it (Line 15)literally incorporating it into his description of thrills. Through repetition of Jeans word, Bill treated
Jeans interjection as collaborative.
The groups understanding of the term thrills was not a private achievement, nor was it
a hidden psychological condition inaccessible to analysts. Rather, a transparency of understanding was publicly and interactively achieved among the nursing students, through
visible and audible behaviors carefully orchestrated. Through coordination of their talk,
embodied actions, especially recognizable hand gestures, and ongoing use of material
objects and mediating tools within an organized space, the nursing students interactively
performed an understanding of the term thrills that the medical students silently observed
and thereby corroborated. The participants collaboratively completed each others utterances, repeated terms of each others talk, reproduced each others hand gestures, and in
other ways cooperated in a collective display of understanding. Eventually, the nursing students stopped talking and oriented away from the television monitors and back toward their
workbook (or toward each other, whispering quietly), thereby showing themselves to be
satisfied that an understanding of thrills had been adequately provided or accomplished.
Moments later, one of the medical students elected to speak, transcribed as follows:

Gesture and the transparency of understanding109

23 Marie: Thrill is just the: (.) youre feeling the
murmur (.) you can feel it with your

Maries description or definition of thrills came off as relatively succinct, hearably complete, perhaps polishedat least compared to Bills earlier attempts to define the term,
which involved hesitations, restarts, and eventually Susan and Jean as overlapping collaborators. Nevertheless, Maries ostensibly individual display of understanding must rightly
be regarded as a group achievement as her performance represented a composite of the
nursing students immediately prior vocal and visible behaviors. Marie used words that
had already been spoken: feel (Lines 3, 7, 8, and 17), murmur (Lines 5, 19, 21), and hand
(Line 6). Moreover, her utterance was coordinated with a gestural sequence that unmistakably resembled Susans (and Bills) prior performance: With the word thrill (Line 23),
Marie placed her flattened hand onto her chest; with the word feeling she lifted her hand
a few inches from her chest before returning it. Thus, her ostensibly individual display
of understanding was an embodied formulation (Heritage & Watson, 1979) of sorts
that summarized or performed the gist of prior interactionboth talk and gestureand
thereby displayed a certain understanding of that prior interaction, altogether advancing the
transparency of understanding within the group. Maries participation served to bridge the
telecommunications divide of the groups videoconferencing sessionthat is, she helped
to constitute the eight participants as being of one mind by registering within the PIP
window a sequence of behaviors with a recognizable pedigree of social interaction from
the larger frame.
Through microethnographic study of classroom activity, such as briefly represented here,
we have documented various forms of communication, including gesture, whereby a transparency of understanding may be interactively accomplished. Gestures may literally take
shape as new understandings publicly emerge and evolve within a group. Gestures may
be observably sharedeven repeatedly performedby those who move jointly toward a
transparency of understanding.
Among the several studies of gesture conducted within the field of LSI in recent decades
(e.g., Bavelas, 1994; C.Goodwin, 1986; C.Goodwin & M.H. Goodwin, 1986; Kendon,
1972, 1980, 1987; LeBaron & Streek, 2000; Streeck, 1993, 1994), our study seems to compare and contrast most interestingly with one: Schegloffs (1984) examination of gesture
and projection. Using conversation analytic methods to explicate empirical (transcribed)
details of talk, Schegloff found that gestures almost always occur within the same turn as
their lexical affiliates, but also tend to precede their lexical affiliates, thereby constituting a projection spacethat is, a processing period between the earliest indication of a
communicative behavior and its eventual delivery. Through study of iconic gestures, he
sought an independent estimate of the possible size of the projection space (p. 288),
which might shed light on other sorts of phenomena such as projection and conversational repair. Although Schegloff flirted with issues of cognitive processes as he focused on

110 Studies in language and social interaction

singular utterances of individual speakers3, he carefully wrote with the voice of a cognitive
agnostic: Words such as intention were displaced by terms like projection; words such as
recognized were recast as displayed recognition of; if the term preference was used, it was
redefined as a structural rather than a psychological condition; when words such as think
appeared, they were corralled by quotation marks; and so forth.
Our study of gesture involves some notable (and we think complementary) differences.
By focusing on strips of social interaction (i.e., knowledge display segments) rather than
individual utterances or even utterance pairs, we see how the life span and the meaning
of a gesture may extend across multiple turns at talk among multiple participants. We
find gestures to be strongly affiliatednot only with specific lexical itemsbut with parts
and wholes of utterances, other participants utterances, other gestures, other participants
gestures, participants embodied use of space, and so forth. Viewed from such a perspective, the ways in which gestures are employed in interaction are highly relevant to the task
of explicating how participants routinely make their understandings visible to themselves
and others. It is in this sense that we speak of the transparency of understanding, not as a
private mental event, but as an embodied, public, and, hence, analyzable achievement.




Can you defirne thrills

Thrirll is what you fee:l (.) like is: (.)
ya could- (0.4)
If- if you happened to have uh huge murmer
(0.4) you could put your hand on (your)




(?) :


the upbeat
Its like (.) fluird thats getting caught on
somethin and its (.) twisting arou:nd the







vessel or
Its tur


or whatever

See Moerman (1990), which criticized Schegloffs (1984) study because it was interested in
mental, not social mattersin cognitive processing and forms of thought rather than interactive
processes (p. 8); because it related movements to words and ideas [and did not] describe those gestures in their social context (p. 19); because it focused upon isolated or single utterances treated
more as sentences composed by individuals than as the products of interaction (p. 20); and
because it based the meaning of a gesture on its correspondance with its affiliated word (p. 40).

Gesture and the transparency of understanding111



You can feel a thri:ll or you (0.2) auscultate





a murmur



a bruit which


Or a murmur
Thrill is just the: (.) youre feeling the
murmur (.) you can feel it with your ha:nd

Bavelas, J. (1994). Gestures as part of speech: Methodological implications. Research on Language
and Social Interaction 27(3), 201221.
Coulter, J. (1979). The social construction of mind. London: Macmillan.
Coulter, J. (1990). Mind in action. Oxford, England: Polity Press.
Crowder, E. (1996). Gestures at work in sense-making science talk. Journal of the Learning Sciences, 5, 173208.
Edwards, D. (1997). Discourse and cognition. London: Sage.
Edwards, D., & Potter, J. (1992). Discursive psychology. London: Sage.
Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Goldin-Meadows, S., Alibali, M., & Church, R.B. (1993). Transitions in concept acquisition: Using
the hand to read the mind. Psychological Review, 100, 279297.
Goodwin, C. (1986). Gestures as a resource for the organization of mutual orientation. Semiotica,
62(1/2), 2949.
Goodwin, C. & Goodwin, M.H. (1986). Gesture and coparticipation in the activity of searching for
a word. Semiotica, 62(1/2), 5175.
Heritage, J. (1984). Garfinkel and ethnomethodology. Cambridge, England: Polity Press.
Heritage, J. (1990/1991). Intention, meaning, and strategy: Observations on constraints on interaction analysis. Research on Language and Social Interaction, 25, 309330.
Heritage, J., & Watson, D.R. (1979). Formulations as conversational objects. In G.Psathas (Ed.),
Everyday language: Studies in ethnomethodology (pp. 123162). New York: Irvington.
Hopper, R. (1989). Conversation analysis and social psychology as description of interpersonal
communication. In D. Roger & P. Bull, (Eds.), Conversation (pp. 4866). Avon, England: Multilingual Matters.
Hopper, R. (1990). Ethnography and conversation analysis after Talking Culture [Special section].
Research on Language and Social Interaction, 25, 161162.
Hopper, R. (1992). Telephone conversation. Bloomington: Indiana University Press.
Hopper, R. (1997, June). A cognitive agnostic in conversation analysis: When do strategies affect
spoken interaction? Paper presented at the Annual Conference of the International Communication Association, Montreal.
Hopper, R., Koch, S., & Mandelbaum, J. (1986). Conversation analysis methods. In D.Ellis &
W.Donohue (Eds.), Contemporary issues and discourse processes (pp. 169186). New York:
Hutchby, I., & Wooffitt, R. (1998). Conversation analysis: Principles, practices, and applications.
Cambridge, England: Polity Press.
Jacobs, S. (1988). Evidence and inference in conversation analysis. In J.A. Anderson, (Ed.), Communication Yearbook 11 (pp. 433443). Beverly Hills, CA: Sage.

112 Studies in language and social interaction

Kendon, A. (1972). Some relationships between body motion and speech. InA. W.Seigman &
B.Pope (Eds.), Studies in dyadic communication (pp. 177210). Elmsford, NY: Pergamon.
Kendon, A. (1980). Gesticulation and speech: Two aspects of the process of utterance. In M.R.Key
(Ed.), The relationship of verbal and nonverbal communication (pp. 207228). The Hague,
Netherlands: Mouton.
Kendon, A. (1987). On gesture: Its complementary relationship with speech. In A.W.Siegman &
S.Feldstein (Eds.), Nonverbal behavior and communication (pp. 6597). Hillsdale, NJ: Erlbaum.
Kendon, A. (1990). Conducting interaction: Patterns of behavior in focused encounters. Cambridge, England: Cambridge University Press.
Koschmann, T., Glenn, P., & Conlee, M. (2000). When is a problem-based tutorial not tutorial?
Analyzing the tutors role in the emergence of a learning issue. In C.Hmelo & D.Evensen (Eds.),
Problem-based learning: Gaining insights on learning interactions through multiple methods of
inquiry (pp. 5374). Mahwah, NJ: Erlbaum.
Koschmann, T., Kelson, A., Feltovich, P., & Barrows, H. (1996). Computersupported PBL: A principled approach. In T.Koschmann (Ed.), CSCL: Theory and practice of an emerging paradigm
(pp. 83124). Mahwah, NJ: Erlbaum.
LeBaron, C. (1998). Building communication: Architectural gestures and the embodiment of new
ideas. Doctoral dissertation, University of Texas at Austin,. University Microfilms NO. AAT98
LeBaron, C. & Koschmann, T. (1999, May). The conversation of gestures: interaction and learner
articulation. Paper presented at the 49th Annual Conference of the International Communication
Association, San Francisco, CA.
LeBaron, C., & Streeck, J. (2000). Gestures, knowledge, and the world. In D. McNeill (Ed.), Language and Gesture. Cambridge, England: Cambridge University Press.
Levinson, S. (1983). Pragmatics. Cambridge, England: Cambridge University Press.
Lynch, M., & Bogen, D. (1996). The spectacle of history: Speech, text and memory of the IranContra hearings. Durham, NC: Duke University Press.
McNeill, D. (1985). So you think gestures are nonverbal? Psychological Review, 92, 350371.
McNeill, D. (1992). Hand and mind: What gestures reveal about thought. Chicago: University of
Chicago Press.
Moerman, M. (1990). Studying gestures in social context. Senri Ethnological Studies, 27, 545.
Pomerantz, A. (1990). Conversation analytic claims. Communication Monographs, 57, 231235.
Potter, J. (1998). Cognition as context (whose cognition?). Research on Language and Social Interaction, 31, 2944.
Psathas, G. (1995). Conversation analysis: The study of talk-in-interaction. London: Sage.
Ryle, G. (1949). The concept of mind. London: Hutchinsons University Library.
Sacks, H. (1992). Lectures on conversation (2 vols., G. Jefferson, Ed.). Oxford: Blackwell.
Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking for conversation. Language, 50, 696735.
Schegloff, E. (1984). On some gestures relation to talk. In J.M.Atkinson & J. Heritage (Eds.),
Structures of social action (pp. 266296). Cambridge, England: Cambridge University Press.
Streeck, J. (1993). Gesture as communication: Its coordination with gaze and speech. Communication Monographs, 60, 275299.
Streeck, J. (1994). Gesture as communication II: The audience as co-author. Research on Language
and Social Interaction, 27(3), 239267.
Wittgenstein, L. (1968). Philosophical investigations (G. E.M.Anscombe, Trans.). Oxford: Blackwell. (Original work published 1953).

Talk in Everyday Life
It is perhaps not incidental that people have not devoted their lives to studying sentences like I had a good breakfast this morning or How are you?. There are
more or less defensible reasons for not studying such sentences. Not studying such
sentences, however, may have real consequences. The question of what language can
do, what people can do with language, what the results of an analysis of I had a good
breakfast this morning would involve, what kind of program it poses for a fieldall
these things remain absolutely open.
Sacks, 1984, p. 24
One of the hallmarks of language and social interaction research is keen interest in ordinary, commonplace interpersonal communication. Sacks, the founder of conversation
analysis (CA), observed that the mundane, trivial talk people do in living their everyday
lives risks being slighted by social scientists more concerned with finding prima facie
important topics for study. Countering this trend, LSI research has shown convincingly
that routine interaction serves as a locus for instantiations and negotiations of identity,
relationships, social structure, and culture. The LSI interest in the everyday reflects not
only a theoretical assumption about its importance but also an ideological commitment to
appreciating and even celebrating routine human communication. Robert Hopper actively
sought to open communication scholars eyes to the everyday, resisting a too narrow
concern with what he termed the great words of great people.
The articles in this section present empirical studies of casual interaction. They primarily reflect conversation analytic methods. As others (Psathas, 1995; ten Have, 1999) have
noted, the name conversation analysis proves unduly restrictive, for the methods have
proven useful to approach a variety of types of interactions beyond conversation. Nevertheless, most of the discourse examined in these articles occurred in casual (non-institutional)
situations: among acquaintances, friends, and family members.
Charlotte Jones (chap. 7) examines restarts in conversation, where a speaker begins
an utterance, abandons it, then begins again. She extends prior work by Goodwin (1980)
and Schegloff (1987), who showed how restarts can work to attract the attention or gaze
of an interlocutor. Jones finds that some restarts may direct the interlocutors attention to a
particular activity, specifically beginning a new topic, or presenting a sensitive or delicate
Charles Goodwin (chap. 8) examines how a speaker may drop the name of an assessable object in such a way that a hearer can recognize the assessable character or special
status of the referenced item. He shows how speakers produce some assessables in such a
way as to project for the recipient how they should be assessed, whereas the production of
others can constitute an assessment test for the recipient. Goodwin draws on and extends
previous CA research on assessments (e.g., Pomerantz, 1984).

114 Studies in language and social interaction

Susan Corbin (chap. 9) argues that questions containing the wording did you may
present particular interactional problems, in that the wording can indicate that the asker
expects that something should have been done. Therefore, such questions may carry an
accusatory sense that makes relevant subsequent talk that addresses the accusation. Data
for the study come from both recorded conversations and field notes. She analyzes both
logical and pragmatic presuppositions inherent in these questions, noting that the majority
of did you questions do not receive only a yes or no but an elaborated answer, often
with accounts.
The tenth chapter examines conversations about illness outside of the doctors office.
The vast majority of social scientific research on medical interaction has focused on professionals communicating and working with patients. By contrast, Wayne Beach legitimizes
the role of laypersons in issues of health and healing. He studies telephone conversations
between family members of cancer patients, who update, assimilate, and commiserate about
the diagnosis and treatment of their loved one. Whatever news or instructions patients may
receive from an expert at a medical facility, the information is necessarily understood and
given shape through the everyday relations of people communicating at home. Through
his analysis, Beach ties abstract notions such as stages of grieving and having hope
to specific social actions identifiable within transcribed data. His study of everyday talk
about cancer bears kinship with his (1996) research on bulimia, which was based on a naturally occurring conversation between a bulimic young woman and her grandmother, who
recognized and dealt with the granddaughters eating disorder.
Samuel Lawrence (chap. 11) takes up the issue of how interactants deal with unwanted
understandings. He provides a case study of one participants rejection of anothers understanding. In effect he examines an instance in which what Drew (1987) called a po-faced
receipt of a tease may have implications for the relationship between interactants. He
contrasts this with practices of third-position repair.
In chapter 12, Jenny Mandelbaum investigates ways that people accomplish interpersonal relationships through their interactions. She presents detailed analysis of two cases
to show methods through which people foreground relationship while continuing talk and
related activities. Tit for tat and conversational repair allow participants to focus attention on some prior bit of talk produced by another speaker. Relational communication is
always implicitly present; this analysis locates moments in which it becomes explicit and
thus more directly available to analysis than in the flow of relationally unmarked discourse.
It offers a bridge from LSI to relational communication research interests.
Gail Jefferson (chap. 13) describes how interactants may clarify a possible ambiguity
without explicitly doing so. On occasion when an alternative possible hearing could be
available in the talk of a speaker, that speaker continues talk in such a way as to clarify
which of the alternative hearings is meant. In that this is done without explicit self-correction (Schegloff, Jefferson & Sacks, 1977), it poses an analytical puzzle. Conversation
analysts typically rely upon interactants displayed orientations to the ongoing activities
in talk. This chapter illustrates how a researcher may deal with a phenomenon for which
these resources are elusive, raising issues about the process of analysis and offering some
suggestive findings.
Also addressing a methodological conundrum, in chapter 14 Emanuel Schegloff raises
the issue of how analysts might trace the suppression of an item and its apparent later

Part II: Talk in everyday lift 115

surfacing. The instances involve a speaker beginning a turn constructional unit in which
a next word, relatively clearly projectible, is not produced at that moment. However, the
possible word in question appears in that persons talk shortly thereafter, but used in a different sense. This analysis goes beyond prior conversation analytic work that has addressed
the issue of noticeable or bearable absences of actions (e.g., second pair parts following
first parts of adjacency pairs). Here, Schegloff accounts for the absence, and subsequent
reappearance, of particular words.
Conversation analysts have argued that in order to invoke some feature of context to
account for details of interaction, the researcher must demonstrate its relevance for participants (Schegloff, 1987). Through analysis of a single instance, in Chapter 15 Phillip Glenn
provides evidence that laughter, by features of its production and placement, may reveal
participant orientation to gender. Thus laughter, like other micro features of interaction,
shapes and renews context (Heritage, 1984, p. 242)in this case, a gendered context.
How people orient to and constitute gender in talk is one example of the larger issue of
connections between discourse and context, which continues to play a central role in much
LSI research.
Hanneke Houtkoop-Steenstra (chap. 16) uses conversation analytic methods for gathering and transcribing recordings of naturalistic interactions. She builds upon and extends
previous research by Schegloff, Hopper, herself, and others, concerning patterns of identification and recognition in telephone interaction openings. Beginning with data to support an argument for systematic differences between Dutch and American calls, she then
explores possible explanations. She develops an intriguing claim about historic change in
the ways Dutch tend to self-identify in phone openings. Further elaborating the analysis,
she then reports research suggesting that there are sex differences within the Dutch data.
Thus, her chapter suggests that variations in how people answer the phone and accomplish
identification may reflect culture, sex, and changes through time.
Beach, W. (1996). Conversational about illness: Family preoccupations with bulimia. Mahwah,
NJ: Lawrence Erlbaum Associates.
Drew, P. (1987). Po-faced receipts of teases. Linguistics, 25, 219253.
Goodwin, C. (1981). Conversational organization: Interaction between speakers and hearers. New
York: Academic Press.
Heritage, J. (1984). Garfinkel and Ethnomethodology. Cambridge: Polity Press.
Psathas, G. (1995). Conversation analysis methods: The study of talk in interaction. Thousand
Oaks: Sage
Sacks, H. (1984). Notes on Methodology. In J.M.Atkinson & J.Heritage (Eds.), Structures of social
action: Studies of conversation analysis. Cambridge: Cambridge University Press.
Schegloff, E.A. (1987). Between macro and micro: Contexts and other connections. In J.Alexander,
B. Giesen, R.Munch, & N.Smelser (Eds.), The macro-micro link (pp. 207234). Berkeley and
Los Angeles: University of California Press.
Schegloff, E.A., Jefferson, G., & Sacks, H. (1977). The preference for selfcorrection in the organization of repair for conversation. Language, 53, 361382.
ten Have, P. (1999). Doing conversation analysis: A practical guide. London: Sage

Utterance Restarts in Telephone Conversation:
Marking Topic Initiation and Reluctance
Charlotte M.Jones
Carroll College
Restarting an utterance is a common practice in natural, everyday conversation (Schegloff,
1987). Restarts (or recyclings) regularly occur at turn beginnings and serve a variety of
functions, including attention-seeking (Goodwin, 1980; Heath, 1984; Schegloff, 1987).
Schegloff (1987) argued that recycled turn beginnings function to repair the possible
impairment of overlapped talk. That is, identical repeats of turn beginningsoccur regularly when there has been an overlap of the turn beginning with the prior turn (p. 7). He
provided an example from a face-to-face encounter:

Well the uhm in fact they must have grown a culture, you know, they mustve- I mean how
long- hes been in the hospital for a few days, right? Takes a bout a week to grow a culture


I don think they grow a I don think they grow a culture to do a biopsy.

Schegloff observed that Ks recycle begins exactly at the point where her talk is no longer
being overlapped or emerges in the clear. Thus, her recycled turn beginning orients to the
end of the overlap and the coming of the listeners attention.
Goodwin (1980) discovered that certain restarts seek recipient gaze as a sign of attention. He noted that speakers have the task of constructing turns for hearers. That is, a
speaker must have a hearers attention and participation. Collaborative efforts by speaker
and hearer are fundamental. Goodwin demonstrated this idea with face-to-face data illustrating speakers use of restarts and pauses to request and gain hearers gaze before continuing their turns. Goodwin illustrated:
Tommy: You agree wi d- You agree wicher



In this instance, Pumpkin, the hearer, is not gazing at the speaker, Tommy. After the restart,
Pumpkin directs her gaze (shown by______) to Tommy. At this point, with the hearers
gaze and attention, the speaker continues his turn. In short, restarts and pauses can function
as attention-getting devices in face-to-face encounters.
But, do restarts serve different functions in a limited communicative channel such as
the telephone? Restarts in such circumstances may function differently than Goodwin and
Schegloff implicate. Do such restarts function to solicit a listeners attention? Hence, this

Utterance restarts in telephone conversation 117

project investigates restarts in telephone conversations (see Hopper, 1992), a medium in
which participants cannot rely on attracting recipient gaze.
Several points should be considered. First, as recipient gaze is not possible over the telephone, how does an interactant know that she or he has secured a listeners attention over
the telephone? Second, a number of recycled turn beginnings with overlaps were found in
a large corpus of telephone interactions. However, putting these aside, another group without overlaps still remained. Do restarts in telephone discourse serve any other functions
besides repairing overlap or seeking gaze?
This chapter attempts to answer this question by examining restarts at turn beginnings in
telephone conversations. Turn beginnings have been found to have particular implications
(Sacks, Schegloff, & Jefferson, 1974; Schegloff, 1987). Schegloff argued that turns project,
from their beginnings, aspects of their planned shape and type (p. 2). Concerning shape
projection, he stated, for instance, that a turn that begins with If may project a contingency clause of a particular length and a similarly sized consequence clause. Question
projection (e.g., wh- word turn beginning), quotation projection (e.g., He says turn
beginning), and disagreement projection (I dont think turn beginning) are all examples
of turn type projections. Thus, turn beginnings are important to turn-projection (p. 2).
Recycled turn beginnings or restarts, then, may mark or signal to the listener that there
is something in particular about to happen in the speakers remaining utterance. It is argued
in this study that certain telephone restarts can function to project a marked topic or issue
as viewed and exhibited by a speaker. Two cases of markedness identified thus far include
restarts of utterances that (a) initiate new topics and (b) indicate a reluctance to ask or
respond to particular issues (i.e., sensitive or delicate). That is, telephone utterance restarts
can serve to summon the listeners attention to a particular part of the conversationa new
topic, or a request or response concerning a perceived sensitive or delicate issue.
Speakers, then, can signal at the beginning of their turns what it is that they are interested
in doing. For example, a speaker may explicitly announce an abrupt topic shift by starting
a turn with, Not to change the subject, but. Similarly, a speaker may use a list-initiating
marker to project-as-upcoming a multi-unit turn (Schegloff, 1981). That is, beginning a
turn with, for instance, First of all, thereby projects that after the turn-unit in which the
first is done, more will follow (p. 75).
Speakers may also use less explicit methods of signaling to listeners their intent, for
instance, the intent to change the topic. During an attempt to change a topic is clearly one
point where a speaker would want a hearers attention. In fact, Schegloff (1979) argued
that if a topic-initial sentence by a speaker is not marked in some way for the listener, then
in the majority of cases the listener will initiate a repair in the next turn. He provided an


hhh A:n:d uh, (0.5) Me:h,

Oh Sibbies sistuh hadda ba: by bo: way.

118 Studies in language and social interaction

In this instance, B initiates a new topic, Sibbies sister, without including any type of repair
device within the turn such as a descriptor or modifier to key the listener. Without such
identifying information, A didnt follow the new line of talk and hence, exhibited a repair.
To successfully initiate a new topic without explicitly marking it as such, a speaker must
somehow signal to the listener this intent. Schegloff uncovered one method, topic-initial
turns that contain a self-initiated repair with a descriptor or modifier, but there may be others. Thus, in instances where a restarting telephone speaker is introducing a new topic, the
restart may function to secure the listeners attention. In the following segments, one can
observe utterance restarts being produced as speakers rather suddenly change topics. The
restarts occur precisely at the points when speakers initiate topic changes.

UTCL A10.14


see some people







he took off and said he was goin to


Hm: : : :


So- I dont know wher:e he is or what








So wha a- what are you doing tonight






In Instance 1, Rick and Jessie are talking about a friends whereabouts. In Line 11, Rick
restarts what are as he mentions Jessies plans for the evening in the form of a topic
initial elicitor (Button & Casey, 1984). Button and Casey noted that topic initial elicitors
regularly take the form of inquiries into what is new and In so doing, they provide for
new topicalizable material as dislocated from prior topical talk (p. 174). Even though
its new topically, Jessie understands his question and follows his lead as evidenced by her
answer of Nothin.
We can similarly observe a restart marking a topic change in Instance 2:
(2) UTCL A10.15


Lotta gigglin hh hhh






Hes gettin in that

Christmas spirit hh

Utterance restarts in telephone conversation 119


o:h shit


pt hh hh



.hh When u: :h


hh uh huh=



=When a you- when

are you goin home



RIC: U: : : :h the t



In this segment of conversation, Rick and Billy are discussing the behavior of a friend,
which they assess using a potentially topically terminal assessment (O:h shit) and laughter in Lines 4 and 5. At Line 6, Billy initiates the topic of going home. However, Rick
continues to laugh at Line 7, overlapping Billys turn, which he then abandons. At Line 8,
he restarts when are you twice. It appears that Ricks laughter leads to Billys first restart.
That is, his talk is now in the clear. Considering that the second restart isnt serving this
overlap function, it instead seems to be related to attention-seeking for the new topic.

And in Instance 3:

UTCL D8.12


I havent talked to my mother in a

lo::ng ti::me. (0.4) >(Tex 0 U weekend






I talked to my da:d.





hu h



Do you hav- do you have any other blo


brothers or sisters



I have a sister.

In Instance 3, Rick and Pam are discussing Pams parents potential reactions to a letter
she had written them. At Line 9, Rick brings up the topic of possible siblings of Pams. He
does this while restarting his utterance beginning do you have. Pams response, I have a
sister, shows that she understands his question and follows his lead topically.

120 Studies in language and social interaction

In all of the preceeding instances, restarts occur as a speaker initiates a topic change.
Schegloff (1979) noted that when topic-initial utterances display no hitches, repair initiations are common in the next turn. In these restart cases, the hearer is able to follow the proposed topic and continue it; no repairs occur. A restart by the speaker ensures the hearers
attention at a turning point in the conversation. Thus, I posit that the restarts are functioning
successfully to alert the listener to a new topic.
In addition to marking topic initiation, utterance restarts can function to project speaker
Participants in everyday conversations routinely make and respond to requests. However,
at times, speakers may exhibit in some fashion a reluctance to inquire about or reply to
certain issues. For instance, some people may understandably be hesitant to discuss topics such as sexual activity or personal finances, considering them to be of a sensitive or
delicate nature. Schegloff (1980) identified one way in which participants show an orientation to talk as sensitive or delicatethey first exhibit a pre-delicate. That is, a question
projection is followed by a question that is marked in some fashion as a delicate one. For
example, before asking a question that might be considered sensitive for some reason, a
participant might first say Can I ask you a question? or more explicitly, I want to ask
you a question that may seem a bit indelicate, but I have to know. Restarts may provide
speakers with another, perhaps less explicit or less marked, way to display forthcoming talk
as sensitive or delicate.
One group of telephone utterance restarts in this study involves both requests and
responses to requests. Restarts can be seen as functioning to signal or mark some type
of talk as being reluctantly produced. In the following first set of instances, the speaker
exhibits a restart as she or he is responding to a previous speakers utterance and is revealing information that she or he may consider potentially damaging, risky, or embarrassing
in this particular circumstance (e.g., personal finances, setting conditions on a friends
request). That is, speakers restarts show a reluctance to grant or respond to a certain type
of requests (i.e., delicate or sensitive).






-And he has a ra:nch for us to look art so

were gonna go just look at it just
How mu :ch
-hhhh We:ll I- I dont know I don know
h- (.)
dont wanta dis- discuss it on the

Utterance restarts in telephone conversation 121

In this segment, a mother and daughter are discussing the mothers possible purchase of
a ranch. The daughter inquires about the price of the ranch in Line 3. The mother displays a reluctance to reply. After a pause, a delay (i.e., an inbreath), and an appositional
(i.e., We:ll), the mother restarts I- and also repeats I don know. Additionally, she
restarts the word discuss as she metacommunicatively expresses that she doesnt want
to discuss the matter while on the telephone. Thus, the mother exhibits an utterance restart
(as well as other delay devices) at the point where a potentially delicate issuepersonal


(5)UTCL A10.5
Is there any way I can borrow somebodys

(16 lines omitted)

.hhhhhhhhhhh Itll probly take me twenty minutes
When he gets- when he gets back from the bank

you can u- you can borr it


Nat an

10 RIC:
Whens he leaving
12 FLA:
Oh hell proly back in like fifteen minutes and
itll proly take him fifteen twenty minutes
hell hell probly be done forty minutes and
then you use it

In Instance 5, Rick has asked to borrow Flaretys moped, which is currently being loaned
to someone else. In Line 4, Flarety agrees to let him borrow it, but exhibits the relatively
short restart when he gets while doing so. With this and his later comments, he seems to
be setting conditions for or potentially refusing the borrowing by Rick. One could argue
that potentially refusing or setting conditions on a friends request could be considered
socially risky and potentially damaging to the friendship. That is, Flarety may be reluctant
to offend his friend.
As mentioned earlier, not only can speakers reveal a reluctance to respond to particular requests, but they can also show a reluctance to make such requests. In the following segment, we can observe an utterance restart produced as the speaker asks the
recipient to reveal information about herself that is potentially damaging, socially risky, or
embarrassing (e.g., sexual activity). That is, the speakers restart shows a reluctance to ask
a certain type of question:

UTCL A10.14:4
JES: -hhhhh Uh we had it like at eight thirty

122 Studies in language and social interaction




Ye- did j- wu- did you spend the night
last night

In this segment, Rick and Jessie are discussing their workout times interwoven with Jessies anger at her dating partner (i.e., also Ricks friend). At Line 3, Rick asks Jessie if she
spent the night at her dating partners place of residence. He restarts wu- did you as well
as pausing before the utterance. Asking people to reveal where they spend their nights
(especially specifying a dating partner) is a personal and private matter. Thus, it can be
argued that Rick is showing reluctance to inquire about this delicate matter.
We have examined several instances of requesting and of responding to requests that
involve restarts. They all appear to show a reluctance or hesitancy to inquire or respond to
issues that can be considered of a sensitive or delicate nature. That is, speakers and listeners display an orientation to the talk as potentially problematic. Sometimes these displays
cluster together. In the following instances, we can observe both participants displaying
reluctance when talking about a particular issue:



UTCL A10.14:6
I told him I didnt want him to swim: Rick

Was that mean


I told him that I didnt want him to swim

Well what do ya- what do you mean

I tol- I jus- (0.4) you know: I just go::
(0.2) we were talkin about it or something
and I just go:: I dont want you to swim

In Instance 7, Jessie and Rick are discussing a prior conversation between her and the
man shes dating, Billy, about his swimming for the collegiate team. He is also a good
friend and swimming teammate of Ricks. At Line 7, Rick exhibits a short restart, what
do you, in his metacommunicative response to Jessies prior announcement or disclosure
(including the possibility of having hurt her dating partner). In addition to proffering a
query about a sensitive topic, Rick does the delicate work of not snowing alignment with
a conversational partner by displaying agreement or an agreeing assessment with Jessies
announcement. Ricks restart, then, may be serving a dual purpose, displaying a dispreferred turn shape as well as reluctance to discuss the topic.
After a short pause, Jessie then exhibits a restart involving another short pause and you
know: before the actual restart of I just. Her utterance can be seen as socially risky in
that she is revealing a serious request she made of her intimate partner, to not participate
on the university swim team. Asking a college athlete to quit his or her sport would seem

Utterance restarts in telephone conversation 123

to be a significant request. Jessie notes her orientation to her request as delicate in Line 2
when she asks Rick if he thought it was a mean thing to say. As mentioned previously,
revealing this information to a friend and teammate of Billys seems chancy in that Rick
may get upset with her for possibly hurting his friend and the team. Jessies restart may also
be displaying a sensitivity to the lack of alignment in Line 7 from her hearer. Thus, both
participants show a mutual orientation to this topic as delicate and sensitive.
An additional interesting feature about this instance is the use of ya know and a pause
before the beginning of the restart. Pomerantz (1984) found that dispreferred seconds (e.g.,
disagreements when agreements are preferred) typically include delay devices such as
pauses and tokens (e.g., uh, well). She noted that these delay devices display reluctancy or discomfort (p. 72). Thus, the delay devices evident in this segment (as well as in
Instance 4) seem to function as part of the delicate and tentative nature of the talk.
And in Instance 8:

(8) CIS 271.1

1 CAL: Okay- -hh this would be an AIDS patient

2 IS:
=> 3 CAL: I- th- is there any type

= 4 IS:
-hhh Is i- is- are

are you cailling for this patien:t um: =

6 CAL: =Yes uh huh=

In this segment, the caller to a Cancer Information Service has requested to talk with someone about nursing home placement for an AIDS patient. At Line 3, the caller restarts is
there, the beginning of a question. However, she stops and relinquishes her turn to the
Information Specialist.
In Lines 4 and 5, the Information Specialist attempts to find out if the caller is the patient
or if she is representing the patient. She displays a delay (i.e., an inbreath) and recycles
her turn beginning twice in pairs (i.e., Is i- and are- are). One might hypothesize that
she is starting to say Is it you? and then changes it as the former might be considered too
direct. Even in this semimedical situation, asking someone to reveal whether she or he has
a terminal illness such as AIDS is potentially a socially risky question. This is especially so
considering the current stigma associated with AIDS (Sontag, 1989).
This segment is interesting in that both participants restarts could be orienting to the
sensitive nature of that talk, but considering the overlap, they also could be trying to get
the floor. That is, the restarts may also be serving an attention-seeking function here. Both
participants utterances arent changing the topic (as in our first group of restarts), but are
showing orientation to different aspects of the topic. Thus, it is possible that telephone
utterance restarts may serve dual purposes simultaneously.

124 Studies in language and social interaction

This chapter describes various functions of restarts at utterance beginnings in telephone
conversations. Particular recycled turn beginnings were found to serve two attention-seeking functions. That is, a speaker initiating a new topic or showing reluctance when making
or responding to a request may signal to the other that something is up by using a restart.
These restarts occurred precisely at places where speakers were introducing new topics
or were displaying reluctance to discuss particular issues (e.g., sexual activity, personal
finances, potential refusal to lend items, dating issues, illness disclosure).
These findings expand our previous knowledge of the functions of recycled turn beginnings as attention-seeking devices. Schegloff (1987) claimed that they serve an overlap-repair function, whereas Goodwin (1980) argued for a gaze-requesting function. Considering
the absence of recipient gaze in telephone conversations, this chapter argues that restarts
may also serve different forms of attention-seeking functionsto indicate or mark the
initiation of a new topic or a reluctance to make or respond to delicate requests.
Thus, a restart may best be considered as a multifunctional conversational feature, capable of varying sequential work. Employing Mandelbaums (1990) distinction, restartsas
an interactional feature of the practices of conversationserve to accomplish multiple
practices in conversation such as gaining attention, getting the floor, shifting topics, and
marking delicacy. However, there are other conversational features of the practices of
conversation that interactants may employ to accomplish these same social or pragmatic
practices in conversation. These features may include more, or less, explicit or marked
ways of solving interactional problems and can be viewed on a continuum.
At one end of the continuum, interactants may choose fairly implicit, less marked means
to achieve a conversational action with little disturbance to the expression of an utterance
or to the conversations surface (Jefferson, 1996). For instance, a speaker may choose to
gradually change a conversational topic over the course of several turns via a stepwise
transition (Jefferson, 1984; Sacks, 1992).
At the other end of the continuum, interactants may choose fairly explicit, more marked
ways to accomplish actions such as gaining attention, shifting topics, or displaying delicacy. For example, the young woman in the following face-to-face segment employs a
rather direct way to summon her listeners attention:

Moonlight Pizza


She was sitting right here like this YA::LL look at


Rather than displaying a restart to attract gaze and attention (Goodwin, 1980), MOUSE
commands her listeners with the explicit YA::LL look at me:., using added stress, sound
stretches, and increased volume to further emphasize her demand. Moreover, as mentioned earlier, a speaker may start a turn with, Not to change the subject, but as a way
to announce an abrupt topic shift, or first query a conversational partner, Can I ask you a
question? to explicitly mark a subsequent sensitive or delicate question or request.

Utterance restarts in telephone conversation 125

However, there are potential dangers in employing such marked features. First, in some
cases, a speaker may be perceived as abrupt, demanding, or socially inept. Second, opportunities for the conversation to get momentarily or completely sidetracked or for a bid to
change the tone or mood of the conversational moment are made available. For instance,
in response to a serious Can I ask you a question?, wisecracks such as You already did
or What, another one? (Schegloff, 1980) may sidetrack and disturb the serious tone a
speaker is attempting to set. It may then take several subsequent turns to reestablish the
direction or to get out of the side sequence, so one might argue that more marked actions
are potentially less conversationally economical. Third, an interactant may be flatly refused
before the other even hears the question or request, No, I dont answer personal questions
or No, you may not, perhaps especially when in an argumentative encounter.
In comparison, restarts may pose less conversational danger than more marked actions.
First, being less explicit, restarts may be less likely to be perceived as abrupt or demanding.
Second, considering sequential implicativeness (Schegloff & Sacks, 1973), it would be
much more difficult for wisecracks or refusals to emerge with the use of restarts alone.
Regarding the marking of delicate, sensitive matters or topic shifts on the telephone,
restarts seem to be in the middle of the continuum when it comes to such activities as refusing a request, asking a personal question, or initiating a new direction in the conversation.
Thus, telephone restarts express middleground options by speakers. They illustrate how
we as interactants can produce an action to fit the specific needs of the moment-by-moment unfolding of an encounter. Moreover, restarts show us that one form can have many
Future research in the area might uncover yet other practices in conversation that
restarts serve in addition to gaining attention, getting the floor, shifting topics, and marking delicacy. Furthermore, although instances in the present study included both casual,
everyday and institutional telephone talk, a more focused study of different types of institutional interaction could reveal differences regarding the use of restarts. For example,
the use of explicit, marked forms of actions or more implicit, less marked forms may vary
in, for example, medical or therapeutic interviews versus corporate business interactions.
It would also be interesting to discover if restarts serve any of the aforementioned functions in face-to-face encounters. Moreover, investigating the occurrence or lack thereof of
restarts in particularly sensitive environments such as arguments may prove worthwhile.
For instance, might the absence of restarts display a specific stance in an argument, such as
certainty or hostility to the other?
Button, G., & Casey, N. (1984). Generating topic: The use of topic initial elicitors. In J.H. Atkinson
& J. Heritage (Eds.), Structures of social action: Studies in conversation analysis (pp. 167190).
Cambridge, England: Cambridge University Press.
Goodwin, C. (1980). Restarts, pauses, and the achievement of a state of mutual gaze at turn-beginning. Sociological Inquiry, 50, 272302.
Heath, C. (1984). Talk and recipiency: Sequential organization in speech and body movement. In
J.M.Atkinson & J.Heritage (Eds.), Structures of social action: Studies in conversation analysis
(pp. 247266). Cambridge, England: Cambridge University Press.
Hopper, R. (1992). Telephone conversation. Bloomington: Indiana University Press.

126 Studies in language and social interaction

Jefferson, G. (1984). On stepwise transition from talk about a trouble to inappropriately nextpositioned matters. In J.M. Atkinson & J. Heritage (Eds.), Structures of social action: Studies in
conversation analysis (pp. 191222). Cambridge, England: Cambridge University Press.
Jefferson, G. (1996). On the poetics of ordinary talk. Text and Performance Quarterly, 16, 161.
Mandelbaum, J. (1990). Communication phenomena as solutions to interactional problems. In J.
Anderson (Ed.), Communication Yearbook 13 (pp. 216244). Beverly, CA: Sage.
Pomerantz, A. (1984). Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. In J.H. Atkinson & J. Heritage (Eds.), Structures of social action: Studies
in conversation analysis (pp. 57101). Cambridge, England: Cambridge University Press.
Sacks, H. (1992). Lectures on conversation (Vol. II) (G. Jefferson, Ed.). Oxford, England: Blackwell.
Sacks, H., Schegloff, E.A., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking for conversation. Language, 50, 696735.
Schegloff, E.A. (1979). The relevance of repair to syntax-for-conversation. In T.Givon (Ed.), Syntax
and semantics, Vol. 12: Discourse and syntax (pp. 261288). New York: Academic Press.
Schegloff, E.A. (1980). Preliminaries to preliminaries: Can I ask you a question? Sociological
Inquiry, 50, 104152.
Schegloff, E.A. (1981). Discourse as an interactional achievement: Some uses of uh huh and
other things that come between sentences. In D. Tannen (Ed.), Analyzing discourse: Text and
talk (pp. 7193). Washington, DC: Georgetown University Press.
Schegloff, E.A. (1987). Recycled turn beginnings: A precise repair mechanism in conversations
turn-taking organization. In G.Button & J.R.E.Lee (Eds.), Talk and social organization
(pp. 7085). Clevedon, England: Multilingual Matters.
Schegloff, E.A., & Sacks, H. (1973). Opening up closings. In R.Turner (Ed.), Ethnomethodology
(pp. 233264). Baltimore: Penguin.
Sontag, S. (1989). Aids and its metaphors. New York: Doubleday.

Recognizing Assessable Names
Charles Goodwin
Robert Hoppers work has been centrally concerned with the question of how human beings
produce action in concert with each other by deploying the resources and practices used
to organize talk-in-interaction (e.g., Hopper, 1988, 1989, 1992, 1999; Hopper & Chen,
1996; Hopper & Glenn, 1994; Hopper & LeBaron, 1998). The present chapter explores
one facet of this process, focusing on the way in which culturally relevant understanding
of the names used to identify valued objects is made visible through specific interactive
What is investigated here is the ability of a hearer to spontaneously, on his own,
recognize the assessable character of an object being named (a Cord, a particular type of
car built before World War II). The name is dropped in a deadpan fashion, without alerting the hearer to its assessable status, and thus poses a recognition test for the hearer.1 Is
he a competent member of the domain of discourse indexed by the name, such that he can
recognize on his own the special status of the item that speaker has just named? Indeed, in
the data examined herein, there are two hearers, only one of whom passes this test.
This practice of producing assessable names as recognition tests must, however, be seen as
part of a larger family of practices that also includes alternative procedures used by speakers to explicitly signal their hearers that an assessable is about to be produced. As a point of
departure for the phenomenon explored in this chapter, some of these are briefly described.
In earlier work, Marjorie Goodwin and I (C.Goodwin & M.H.Goodwin, 1987) investigated
how turns at talk containing assessments can be organized as a multiparty interactive activity. Thus in the following, as the speaker pronounces an assessment adjective good, the
entity being assessedasparagus pieis formulated as a highly valued object through a
range of both talk and embodied displays by both speaker and hearer:


Jeff made en asparagus

It was s::so [: goo:d.
[I Love it.

Here, the hearer simultaneously produces a positive evaluation at the very moment that the
assessment adjective is spoken. She doesnt wait until after speaker has said good, but
in that talk about cars in this fashion is explicitly marked by the participants themselves as a distinctively gendered, male practice, I use the male pronoun to talk about an addressee of this talk.

Recognizing assessable names 129

instead starts to evaluate it before the speaker has even stated her own evaluation. What
interactive practices make such concurrent assessment possible? Before producing the talk
that constitutes the peak of the assessment, the speaker signposts its upcoming arrival
with an intonationally enhanced intensifier s::so_i. The hearer can use this prepositioned
evaluative frame to project what is about to happen, and indeed she does so by starting her
own assessment at the very end of the intensifier.
In Example 1, the projective signpost took the form of an intensifier (s::so:) and the
assessment peak occurred at the place where the speaker produced an assessment adjective. These slots can, however, be filled with other types of units. For example, one very
common type of assessment is formatted as a noun phrase within which an assessment
adjective, such as beautiful precedes a description of the object being assessed.


Tell Debbie about the dog on the golf course


((intervening talk omitted))

Noun Phrase

Eileen: An this beautiful, (.) [Irish Setter

I[rish Setter ((rev erently))

(3) Curt: This guy had, a beautiful, thirty two Olds

The assessment adjective tells the recipient that the object about to be described is being
assessed in a particular way. Moreover, though the entity being assessed may indeed be
relevant to a larger sequence of activity, such signposting is a local operation. Example 2
occurred in the midst of a story. Paul and Eileen had played golf together, and Paul asked
Eileen to tell the others present how a dog stole the speakers golf ball. Eileens pronunciation of Irish Setter, just after the assessment adjective beautiful, is overlapped by an
intonationally enhanced, appreciative version of the same name by Paul. Note how Pauls
treatment of the Irish Setter as an assessable differs markedly from the way in which he
formulates this same dog within the frame of the report being made by the larger story,
that is, as a protagonist in a laughable event (see C.Goodwin & M.H.Goodwin, 1987, for
more detailed analysis). By placing signposts before the peak of the assessment the speaker
informs the recipient of what is about to happen, with the effect that when this talk is
actually spoken, the recipient is already in a position to treat it as an assessment.
Signposting is, however, but one of many ways in which assessments can be organized
as an interactive activity. One of these alternatives is examined next. Instead of announcing
to the recipient that what is about to be said should be assessed in a particular way, speaker
produces the assessable out of the blue. In that the talk containing the assessable has not
been categorized as such (e.g., with an anticipatory signpost), the recipient is faced with the
task of discovering that an assessable has been produced on his or her own. The following
provides an example. In these data, the participants are car buffs. Curt is trying to restore a
Model T and asks Mike where he can get a rear spring for the car:

130 Studies in language and social interaction







Lemme ask a guy at work.

Hes gotta bunch a old clunkers.
Well I cant say that theyre ol: clunkers=
eez gotta Co:rd?
Two Co:rds,
[Not original,
Oh yes. Very original,
Oh:: reall[y?
[Yah. Ve(h) ry origi(h)nal.
Awhh are you shit tin m [e?
[No Im not.

In Lines 57 Mike describes a particular type of car, a Cord, without explicitly assessing it.
However his recipient, Curt, treats such a car as a very highly valued object with a series of
elaborate displays in Lines 10, 13, and 15, for example, asking Mike are you shittin me.
Once Curt uncovers the assessable character of the car, Mike joins him in displaying appreciation of it. Thus Curt initially treats what Mike said as so remarkable that it can hardly
be believed by saying not original, a proposal that if true would diminish the assessable
status of the cars being evaluated. This question provides an opportunity for Mike in Line
12 to emphasize that they are indeed original, and in so doing to display his own appreciation of the cars. Note the placement of the word Very before origi(h)nal, the enhanced
intonation with which both of these words are spoken, and the emphasis provided by placing Oh before yes at the beginning of the turn. The process of assessing the cars thus
becomes a mutual, collaborative activity.
The assessment-relevant nonvocal behavior that occurs in this sequence merits special
comment. While saying Oh yes in Line 12, Mike shakes his head from side to side.
Rather than contradicting the yes in his talk, this head shake simultaneously displays
that he is disagreeing with the assessmentdiminishing proposal just made by Curt (that the
Cords were not original) and constitutes a form of assessment activity in its own right,
an oh wow headshake. Because these phenomena have already been described in detail
elsewhere (M.H.Goodwin, 1980; Schegloff, 1987) they are not discussed further here.
In these data, Curt is able to recognize the exalted status of a Cord without being explicitly told that it is an assessable by Mike. This suggests that speakers have available to them
at least two alternatives for introducing an assessable into talk:
1. Announce to recipient that what is about to be said is an assessable. For example put
an assessment adjective like beautiful before it.
2. Produce an object without marking it as an assessable and thus place recipients in a
position where they must recognize its assessable status on their own.

Recognizing assessable names 131

For clarity, recognition of an unmarked assessable has so far been treated as something
done entirely by the recipient working alone. I now want to explore the possibility that the
process through which the recipient recognizes even an unmarked assessable can itself be
organized as an interactive activity.
Seeding the Ground for an Assessable
In Example 4, despite the speakers deadpan production and lack of explicit assessment
terms, there are in fact some features of the talk that might guide the recipient to see what
is about to be said as an assessable. Mike first describes the cars of his friend as old
clunkers, but then says that they are not old clunkers:



Mike :


Lemme ask a guy at work.

Hes gotta bunch a old clunkers.
Well I cant say that they re ol: clunkers=
eez gotta Co:rd?

The recipient is thus instructed to hear what is about to be described as something that
stands in marked contrast to old clunkers. Through the operation of such contrast organization, the assessable name in Line 5 emerges within an environment that has already been
subtly shaped by its presence; the shadow of its properties become visible before the object
itself. Though not explicitly marking the name being produced as an assessable, Mike has
nonetheless seeded the ground for its recognition.
Holding the Name Available
Despite the way in which its status has been foreshadowed, when the word Co:rd? is
actually spoken it is not treated as an assessable. Mike ends his pronunciation of the word
with a rising contour (indicated in the transcript by a question mark), an act that frequently
functions as a solicit for a response from the recipient, and leaves a space after producing
the word for the recipient to respond. However, the recipient does nothing and in Line 6 a
gap ensues.
Mike thus produces a response-relevant object that does not receive an appropriate response. He now employs a standard procedure available to speakers for pursuing a
response: rather than moving his talk forward into new material, he redisplays this object
for his recipient (Line 7):

Mike :

Well I cant say that they re ol: clunkers=

eez gotta Co:rd?

132 Studies in language and social interaction



Two Co:rds,

Indeed, in the present case Mike upgrades the assessable from a Cord in Line 5 to Two
Cords in Line 7. Continuing to hold the assessable available in this fashion both extends
the time available to recipient for producing a response2 and also subtly signals (e.g.,
through the reiteration of the assessable and its upgrade) that further response is relevant.
Mike also performs a nonvocal gesture that helps to solicit a response. To look at how
this gesture operates it is helpful to consider the actions of the third party present during
this exchange, Gary. Recall that the sequence began with Curt asking for help in finding a
high arch spring for his Model T. Right after Mike mentions his friend with the old clunkers, Gary offers the name of someone else (it is later revealed that this person builds street
roadsters and is thus a possible source for the spring):





Mike :




Lemme ask a guy at work.

Hes gotta bunch a old clunkers.
Yknow Marlon Liddle?
Well I cant say that theyre ol: clunkers=
eez gotta Co:rd?
Two Co:rds,
[Not original,

Just as Mike reveals that the cars are not old clunkers, Curt orients to the fact that Gary has
just said something by shifting his gaze noticeably away from Mike and toward Gary. He
continues to gaze away from Mike until after Line 7. Thus throughout the time that Mike
is announcing the presence of the Cords, Curt is looking away from him. As Mike says
Two Cords in Line 7, he moves his hand forward with two fingers extended in a V (i.e.,
a hand gesture for the number two) toward Curt and then back to his own face. This very
noticeable gesture occurs right at the point where Mike is upgrading his assessment and
appears to act as an additional solicit to Curt (for more detailed analysis of how gestures
can be used to attract the gaze of nongazing recipients, see C. Goodwin, 1986b). Very
shortly after this happens, Curt brings his gaze back to Mike with a movement that also
shows heightened attentiveness to what has just been said (e.g., while moving, Curt raises
his head). When this movement is completed, he begins his vocal response to the assessable in Line 10, intercepting Mikes appending And. Note that Curts head movement
See C.Goodwin (1981, chapter 3) for other analysis of how speakers add new segments to their
talk in order to coordinate the unit production of that talk with relevant actions of their recipients.

Recognizing assessable names 133

occupies the silence in Line 8 with the beginning of his response. Thus, unlike the much
shorter silence in Line 6, it is not a gap, but instead becomes a space filled with assessment
relevant activity.3
In brief, here Curt, unlike Gary, is able to display his ability to independently recognize
the exalted status of a Cord. However that independent display has in fact been made
possible through a subtle interactive process of prompting from Mike, who has worked
hard to hold the assessable name available until Curt can see its import and react appropriately to it.
More generally, here we find an instance of what seems a more general strategy of
downplaying something before its emergence, and then dropping it as a bomb, so that its
unique assessable character is highlighted by its sudden emergence within a relevant but
unlikely environment. Indeed, one can speculate that the ideal way this sequence would
have run off would have been for Curt to have asked what kind of old clunkers the guy
had, and then received a Cord in response. Note that unlike the congruent assessments
in Example 1, where both participants were enthusiastically evaluating the assessable, the
current strategy is characterized by asymmetry in participation, with each party displaying
markedly different affect. The party dropping the bomb, here Mike, talks with deadpan,
cool nonchalance. By way of contrast the recipient of the bomb displays shocked, elaborated amazement.
Recognition of assessable names, and the tasks it sets its recipients, sheds interesting light
on the organization of cultural knowledge as an interactive phenomenon. One of the central themes that has motivated research in cultural anthropology from Malinowski through
contemporary studies of cognition, is the question of how members of a society recognize
and properly interpret in a culturally meaningful way events in their phenomenal world.
Building a response to an unmarked assessable is relevant to this process in a number of
different ways.
First, in order to deal with the assessable properly recipient must recognize the object
that speaker is talking about. This is by no means a trivial matter. For example, one person
viewing these data heard the car that Mike was talking about as a (Honda) Accord, something that led her to become quite puzzled about Curts reaction to it. Being able to properly
identify items such as this is one of the things that establishes within the talk of the moment
a participants competence, and indeed membership (or non-membership) in a specific
culture. In the present data, the cultural world at issue is that of car buffs, but equivalent
recognition tests can be posed in almost any domain of discourse, for example, science,
politics, farming, sports, and so on. Frequently names are used to describe assessable
objects in talk, and a very interesting literature on the interactive organization of reference and name recognition now exists (c.f. Clark, 1996; Clark & Schaefer 1986; Clark &
Wilkes-Gibbes, 1986; Isaacs & Clark 1987; Sacks & Schegloff, 1979; Schegloff, 1972).
For other analysis of how nonvocal assessment activity can occupy silences, see M.H.Goodwin

134 Studies in language and social interaction

Second, in order to find the assessable status of what is being talked about, the recipient
must know how to rank and evaluate the object once it has been identified. A response to
an assessable can contain an alignment display of some type (e.g., Curts treatment of the
Cords as highly valued objects). Therefore, mere recognition of the name and the entity it
refers to is not sufficient to build an appropriate response to an assessable. In addition, the
recipient must be able to evaluate the recognized object and properly place it within the
larger cultural domain that it inhabits.
Third, the results of these operations can be publicly scrutinized by other participants.
The recipient is performing the tasks of recognition and evaluation in order to build an
appropriate response to the unmarked assessable. That response will display to others
whether he or she did or did not recognize the assessable and how he or she evaluated it.
Others can and do choose to disagree with a speakers assessment of a particular entity.
For example, shortly after the sequence being examined here, Curt proposed that a thirtytwo Olds should be treated as an exalted, highly valued object in much the way that the
Cord here is, but Mike refused to go along with this proposal (for detailed analysis of these
data, see C.Goodwin & M.H.Goodwin, 1987). Recognition and evaluation of a referent
are frequently conceptualized as purely internal, psychological processes. Here, however,
it becomes possible to analyze how performing these actions can be subjected to public
scrutiny, confirmation, and challenge within systematic processes of interaction.
The public, interactive practices through which a name is both recognized and evaluated are quite relevant to central issues posed in the analysis of culture. For example, they
permit empirical investigation of the process through which members of a society come to
share a culture in the sense that separate individuals form judgements about the events
they encounter that are congruent with those of their co-participants, but differ radically
from the interpretations of these same phenomena made by members of another group. By
viewing processes of categorization and evaluation within an interactive matrix, it becomes
possible to shift analysis from specific cultural categories, that is, a list of fixed, stable
entities argued to constitute the culture of the group, to the underlying social processes
through which such categories are formed, tested, used, and changed as constitutive features of the activities the participants are engaged in.
Fourth, insofar as the identifications and judgments one makes can be scrutinized by
others, and used to assess ones competence and membership in a particular culture, these
processes provide a built-in motivation for members of a group to learn the background
information, ways of speaking, and so on, necessary for appropriate participation in a specific domain of discourse. Talking about cars for these speakers is very serious business,
and indeed one of the ways in which they negotiate and establish their competence and
standing vis--vis each other. The same is true for many other domains of discourse. These
interactive processes thus provide structures for both testing and motivating acquisition of
particular bodies of knowledge.
Fifth, such considerations raise the question of how participants learn relevant information about a domain of discourse in the first place. Clearly a multiplicity of acquisition
processes are involved.4 The present data shed light on how assessments might be relevant
to such issues. Someone listening to this talk who had never heard of a Cord before could
find from the way in which it is treated by Curt and Mike (a) that a Cord is a type of car,
(b) that it is a very highly valued object in this culture, and (c) something about the criteria

Recognizing assessable names 135

used to evaluate such phenomena in this particular domain of discourse, for example, that
the status of a car as original is a most relevant attribute for judging it (i.e., this is the first
question Curt raises about the Cord in Line 10). The sequence thus provides information
about both the status of particular objects in this culture and ways of invoking these objects
and their relevant attributes within talk. Such phenomena provide a practical resource for
parties involved in the interaction. Indeed one of the men participating in this interaction,
Gary, is not able to display the competence about the world of cars that Mike and Curt
exhibit, and one can in fact see him trying to learn how to talk about them appropriately as
the conversation unfolds (see Goodwin, 1986a).
The self-explicating resources provided by assessments are available not only to participants but also to ethnographers and analysts. Such structures provide a way of getting
information about the content of a culture without querying participants. Use of methods
such as this seems especially important because membership in a culture involves not
merely recognition of content items, but also particular ways of talking about these items,
appropriate alignment displays to them, and so on.
The phenomena investigated here provide one demonstration of how fine-grained cultural knowledge is built, organized, and deployed through precise use of the practices used
to build action within talk-in-interaction.
Clark, H.H. (1996). Using language. Cambridge England: Cambridge University Press.
Clark, H.H., & Schaefer, E.F. (1986). Collaborating on contributions to conversations. Language
and Cognitive Processes 2(1), 1941.
Clark, H.H., & Wilkes-Gibbes, D. (1986). Referring as a collaborative process. Cognition, 22,
Goodwin, C. (1981). Conversational organization: Interaction between speakers and hearers. New
York: Academic Press.
Goodwin, C. (1986a). Audience diversity, participation and interpretation. Text, 6(3), 283316.
Goodwin, C. (1986b). Gesture as a resource for the organization of mutual orientation. Semiotica,
62(1/2), 2949.
Goodwin, C., & Goodwin, M.H. (1987). Concurrent operations on talk: Notes on the interactive
organization of assessments. IPrA Papers in Pragmatics, (1) 152.
Goodwin, M.H. (1980). Processes of mutual monitoring implicated in the production of description
sequences. Sociological Inquiry, 50, 303317.
Hopper, R. (1988). Speech, for instance: The exemplar in studies of conversation. Language and
Social Psychology, 7(1), 4763.
Hopper, R. (1989). Sequential ambiguity in telephone openingsWhat are you doin. Communication Monographs, 56(3), 240252.
Hopper, R. (1992). Telephone conversation. Bloomington: Indiana University Press.
Hopper, R. (1999). Going public about social interaction. Research on Language and Social Interaction, 32(12), 7784.
Hopper, R., & Chen, H. (1996). Languages, cultures, relationships: Telephone openings in Taiwan.
Research on Language and Social Interaction, 29(4), 291313.

For a very subtle example of learning within the midst of conversation, see Jefferson (1987).

136 Studies in language and social interaction

Hopper, R., & Glenn, P. (1994). Repetition and play in conversation. In B. Johnstone (Ed.), Repetition in discourse: Interdisciplinary perspectives (Vol. 2, pp. 2940). Norwood, NJ: Ablex.
Hopper, R., & LeBaron, C. (1998). How gender creeps into talk. Research on Language and Social
Interaction, 31(1), 5974.
Isaacs, E.A., & Clark, H.H. (1987). References in conversation between experts and novices. Journal of Experimental Psychology; General, 116(1), 2637.
Jefferson, G. (1987). Exposed and embedded corrections. In G. Button & J.R. E. Lee (Eds.), Talk
and social organisation (pp. 86100). Clevedon, England: Multilingual Matters.
Sacks, H., & Schegloff, E.A. (1979). Two preferences in the organization of reference to persons
and their interaction. In G. Psathas (Ed.), Everyday language: Studies in ethnomethodology
(pp. 1521). New York: Irvington.
Schegloff, E.A. (1972). Notes on a conversational practice: Formulating place. In D. Sudnow (Ed.),
Studies in social interaction (pp. 75119). New York: The Free Press.
Schegloff, E.A. (1987). Analyzing single episodes of interaction: An exercise in conversation analysis. Social Psychology Quarterly, 50 (2), 101114.

Interactional Problems With Did You Questions
and Responses
Susan D.Corbin
University of Texas at Austin
Did you questions are ubiquitous in everyday talk. The examples used in this project are
taken from recordings of actual conversations or from overheard conversations noted by
the author. A collection of did you questions and observations of their use and characteristics was made from which the examples in this chapter were drawn. In this collection,
it was noted that did you questions are used in many ways. For example, did you
questions can be used to begin a conversation upon first meeting a known other:
(1) [Corbin, FN] (Student to student)

Hi Kim, did you get that tape from the Speech

Yes, thank you so much for doing that

to continue a conversation when a previous topic has been talked out:


[UTCL A35d.l5] (Wife to husband)

HNK: pt .hhhh Did get the deal sold though
KRS : Great.=
HNK: = So (0.4)
KRS: Did you get your account straightened out

to introduce a previously unmentioned mentionable:

(3) [UTCL A35a.l2] (Daughter to mother)

KRS: .hhh Okay well you have a good day, did you have

good time over at Joyces last night?

to remind someone of an intended action:


[Corbin, FN] (Mother to teenage daughter)

Mom: Did you bring in the trash can?
Yes, I did.

Occasionally, the recipient of a did you question shows that the question is problematic:

Interactional problems With did you questions and responses 139



[Corbin, FN] (Co-workers)

Pizza worker 1:
Did you grate this cheese?
Pizza worker 2:
Whats wrong with it?
Pizza worker 1:
Well, you were supposed to put

Saran Wrap on it.

[DP 4]
Did you go in this morning?

U:h no, my back was hurtin too much

Recipient response indicates how she or he has taken the question. In example 5, an
exchange between two people working in a pizza palor, the asker does not use any vocal
intonation that might cause the question to sound as if he is accusing the recipient of
anything. However, the recipients response indicates he appears to have heard an accusation (Whats wrong with it?). In Example 6, the response shows a problem by the
dispreferred-shaped response (Pomerantz, 1984).
The next two examples show that another researcher has noticed that both hearers and
askers of did you questions may find them problematic (Tracy & Naughton, 1994). In
example 7, in an interaction between faculty and graduate students at a graduate seminar,
Beth displays that she has problems with Sams did you question:
(7) [Tracy & Naughton, 1994, pg 294, excerpt 12]


Did you, have any dilemmas of choice in terms of

experimentation here? Did you, did you sacrifice


uh external validity for control at any point?


Uh yeah the, well I, our readings, I mean when


when they read the conversations or read the


Tracy and Naughton characterized Berns disfluent answer as showing that Beth
recognize[s] a difficulty (p. 295) with the question.
Tracy and Naughton (1994) also showed an example demonstrating that askers may
indicate that they recognize the problematic nature of did you questions. They noted
that to ask a did you question of someone is to indicate that the action questioned is
something that could be expected to have been done:

140 Studies in language and social interaction


[Tracy & Naughton, 1994, p. 287, Excerpt 1]

... Did you, are you aware, I would assume that, that studies looking at self attributions and other attributions of competence generally show a pretty high correlation?
hmm mm
That, that is generally true? That, that persons own self rating of competence
correlates pretty highly with ratings of those surrounding?

Roys question concerns Sues research presentation. He starts his question as a did you
question, which inquires about the recipients actions. If one continues along his did you
line of questioning and combines it with the end of his question, one arrives at the conclusion that Roy was going to ask whether or not Sue had found other studies reporting
that self-attributions and other attributions of competence show a correlation. Tracy and
Naughton (1994) argued that to ask if they did something suggests it is an activity that
could be expected (p. 287). That is, it would be expected for Sue to find research reporting
the high correlation and perhaps untoward if she had not found this research.
However, Roy changes his did you to are you aware, which asks about the recipients state of knowledge at the time of the question. Before completing his question, he
amends his statement to I would assume that, which refocuses the knowing about the
attribution studies from the student to himself. Tracy and Naughton (1994) argued that
Roys reformulation of his did you question from did you to are you aware and
finally to I would assume that suggests he does not want to imply that she (the student)
should know what he is asking (p. 287). That is, the successive amendments move the
asker away from the did you format and softens the potential offence (or face threat) in
the question.
These four examples show that both speakers and hearers demonstrate in talk that they
recognize the problematic nature of did you questions. However, note that there is only
a potential for did you questions to be problematic. Of course, not every did you question is going to be a problem for every recipient, as seen in Examples 1 through 3. Certainly
vocal intonation and sequential location have a lot to do with the problematic potential of a
did you question. As each example is discussed, these features are noted.
This chapter discusses three characteristics of did you questions, any one of which
might induce a problematic response to a did you question. They are:
1. The use of did you at the beginning of the question indicates it is about a recipients
past action (or possible past action) and may be heard by the recipient to have problematic
linguistic logical presuppositions.
2. A did you question can be highly indexical; that is, the referent of the question is
underspecified yet the questions structure shows that the speaker believes that the recipient will understand the sense of the question. This high indexicality may lead to a recipients hearing a problematic linguistic pragmatic presupposition.
3. The did you question is grammatically packaged to elicit a yes/no response, but
usually receives an elaboration as well as the yes/no. A lack of expansion may lead to
an askers pursuit of an expansion, which can cause interactional problems. In the present
collection, no one problematic did you question contains all three of these aspects. The

Interactional problems With did you questions and responses 141

following sections include discussions of these problematic aspects of did you questions
in more depth with examples from actual conversations.
The notion of presuppositions in language has been discussed by linguists since the 1950s.
Levinson (1983) noted that there is more literature on presupposition than almost any
other topic in pragmatics (p. 167). He also observed that there is an ordinary notion of
presupposition that describes any background assumption against which an action, theory,
expression or utterance makes sense or is rational (p. 168). Contrasted with the ordinary
notion of presupposition is the linguistic notion that is restricted to certain pragmatic
inferences or assumptions that seem at least to be built into linguistic expressions and
which can be isolated using certain linguistic tests (p. 168). The most common linguistic
test for logical presuppositions is the constancy under negation test, which states that the
presuppositions of a statement remain true whether the statement is true or false. Keenan
(1971) proposes that there is also a pragmatic presupposition which is that there is a clear
relation between the statement and its context. If a statements context is not clear to a
recipient, she or he may conclude that the speaker is being ironic, silly, or stupid. Examples and discussion of problematic did you questions involving logical and pragmatic
presuppositions follow.
Logical Presupposition
According to Levinson (1983), questions will generally assume the presuppositions of
their assertive counterparts (p. 184). Consider the did you question from Example 6.

[DP 4]




Actually (0.2) I think he will, theres- (0.5) because Shawn (0.7) has been he
did the same thing, walked in, he said that was a- ((noise)) they didnt say anything to im.
When he went in this morning
Did you go in this morning?
U:h no, my back was hurtin too much

Cathy asks if Evan went in to work that morning. Her vocal emphasis indicated by a raised
tone on the word you indicates a shift of emphasis from Shawns going to work to Evans
going to work. Es answer to her question in the disperferred turn shape of a long pause and
the filler Uh indicates he has a problem with the question (Pomerantz, 1984). The assertive counterpart of Cs did you question is Evan did/did not go in to work this morning.
Cs did you question generates at least one possibly problematic presupposition: Evan
had work to go to this morning. According to the linguistic test, this statement remains
true whether or not he actually did go to work. If a recipient hears the presupposition in the
did you question as problematic, the recipients answer will probably reflect this, as in

142 Studies in language and social interaction

example 6, by giving a justifying reason for not going to work that morning. The assumptions that people make about others actions may be seen in the logical presuppositions of
their did you questions. If the recipient hears the presupposition as problematic, she or
he may answer the question in a manner indicating a problem.
Pragmatic Presupposition
The knowledge that people in relationships share is an integral part of understanding
problems with the indexical aspect of the did you question. Pragmatic presuppositions,
according to Keenan (1971), require that the question be uttered in an understandable context. For recipients to understand a did you questions pragmatic presupposition, the
recipient must understand and recall the shared knowledge of the questions topic. If the
recipient does not recall the questions indexed shared knowledge, the recipients answer
may indicate problems.
Context may be clear in at least two ways and can be shown in these examples of nonproblematic did you questions. One, as seen in example 9, is that the did you question
refers to the current topic:

[DP 4]
C: Yeah but theyre so [(tacky)
E: [Did you tell them to take

their (0.8) sandwich and sh- stick it

C: No because I had a (0.8) I had two cards (0.4)

C has no trouble understanding the context of the did you question because it does not
change the topic of conversation. She complains that the counter people at the sandwich
shop were unpleasant to her (theyre so (tacky)). Es did you question asks if she
decided to purchase a sandwich despite the unpleasantness (Did you tell them to take their
(0.8) sandwich and sh- stick it). C shows that she has no problem understanding the context of the did you question in her immediate answer and the continuation of her story.
A second way context is clear in did you question asking is by the separation of the
did you question from the previous topic with some kind of conversational boundary.
Example 3 shows a did you question during a preclosing in a mother-daughter telephone


[UTCL A35a.12]
you just pick up (.) dad and Timmy n from work and come over.
.hhh Okay well you have a good day, did you have a good time over at Joyces
last [night?
[Yeah, we did, it

was real ni:ce?

Well thats good to hear.

Interactional problems With did you questions and responses 143

Before KRS asks her mother the did you question, she and MAB have begun to close
the telephone call:
MAB: Oka:y.
= KRS: .hhh Okay well you have a good

Schegloff and Sacks (1984) described the closing of a telephone call as working in a
step wise fashion to allow the introduction of unmentioned mentionables (p. 80). With
the exchange of okays, KRS and MAB have aligned contributions toward closing the
encounter. A preclosing moves the partners to closing unless one of them thinks of something else to mention. In this instance, the preclosing separates the new topic introduced by
the did you question from the previous topic. The preclosing helps make it clear to the
recipient that the did you question is a new topic.
Pragmatic presupposition problems can occur in conversations when recipients do not
understand the reference of the did you question. Example 10 shows problems with the
pragmatic presupposition of a did you question. Two female friends are at the beginning
of a telephone call negotiating a first topic:


[UTCL A24.5]
My roommate is such a bitch
huh c(h)ause .hhh whatBET:

=No .hh- (0.3) whaa you doin

Oh. Did you find it

(Oh) did I find what

The shorts
The shorts

. hhhh O: h no: .

CARs did you question (Did you find it) is problematic on two counts. First, it is an
abrupt change of topic. She does not indicate to her interlocutor in any way that she is
changing topics from what are you doing to finding it. CAR appears to be trying to find
any topic for them to talk about other than that her roommate is a bitch. When what are
you doing does not produce a topic, she shifts immediately to did you find it:

144 Studies in language and social interaction




=No .hh- (0.3) whaa you

Oh. did you find it

The second problematic aspect of CARs did you question is the unclearly indexed it
in the question. BET indicates this is her problem in the way she asks for clarification (did
I find what) (see Schegloff, Jefferson, & Sacks, 1977). BET uses the word what to indicate
that it is where she is having problems understanding CARs did you question. Example 10 shows that did you questions can be problematic if the pragmatic presupposition
of context through topic shift and pronoun reference is not clear to the recipient.
As mentioned in the introduction, did you questions are grammatically packaged to
elicit a yes or no answer. Yet, in a collection of did you questions, the majority are
answered with more than just yes or no. Some did you questions are answered with
expansions of the yes/no answer, whereas others are answered with accounts, that is,
reasons for having done or not having done the action the question concerns. Expansions
often look like the answer MAB gave KRS in Example 3:

did you have a good time over at Joyces last

Yeah, we did, it was real ni:ce?

MAB answer with a Yeah and expands the answer with it was real ni:ce. Accounts
often look like the answer that E gave C in Example 6:
C: Did You go in this morning?

E: U:h no, my back was hurtin too much

E answers with no and an account, my back was hurtin too much. Very often, this
expansion or account addresses a problematic logical presupposition of the question, such
as the presupposition that Evan had work to go to that morning.
Not only can the did you question itself be problematic for interactants due to presuppositions, but also the pursuit of an expansion or account to the answer of a did you
question can be problematic for a recipient.
Pursuit of Expansion
In the next example, from a videotape of a couples dinnertime conversation, Tom asks
Abbie a did you question that she answers with a simple no. Tom asks for an expansion
of the no answer and is successful in getting an expansion. However, he also receives a
very marked response:

Interactional problems With did you questions and responses 145







Have y- did ya do anything today fo:r (a) (0.6) finance class

For what?
For finance class did you get anything done
Nothing at all
Nada. (0.5) I havent done anything Ive been gone, since ten oclock this morning

Toms first did you question (first arrow) concerns whether Abbie has prepared anything for the finance quiz they plan to study for later in the evening (Have y-did ya do
anything today fo:r (a) (0.6) finance class). His interutterance 0.6-second pause and
very quiet utterance finish may be what leads Abbie to ask for clarification of his question
(For what?). He starts his repeat question (second arrow) with the non-understood section of his question (For finance class did you get anything done). There is no particular
intonation in this question to indicate that he was accusing her or doing more than asking
for information. Abbie shows none of the problematic features seen in other did you
question answers, such as a hesitation or uh filler. However, she answers without an
expansion (No). Tom asks for more than her negative answer by with his next comment
(Nothing at all). The raised inflection of the word all may indicate surprise that she has
not done anything. Although Abbie does not appear to find the original did you question
problematic, she does appear to have problems with the pursuit of an expansion of her
no answer. She pauses 0.6 seconds before she answers and then reinforces her negative
answer with two more negatives (Nada and I havent done anything) before she offers
an account for not having done anything (Ive been gone, since ten oclock this morning). Her emphasis on gone shows that it has been impossible for her to do anything for
finance today. This example demonstrates that if interactional problems do not occur with
the asking of the question itself, problems may occur if the asker pursues more than the
yes/no answer offered.
In the next example, the recipient of the did you question also answers the question
of her actions without an explanation of those actions. Grammatically, she has answered
the question. However, pragmatically, her recipient appears to expect more than her no

[UTCL D9:3]
GOR: Did you: give Suzy the advice I suggested?

GOR: Are you going to?


146 Studies in language and social interaction



I dont believe you
Youre irritable

Unlike Abbies response in the previous example, Denise appears to find the did you
question problematic as seen by her 1.2-second post-question pause (see Pomerantz, 1984).
When she does answer, she gives the least amount of information that answers the question
(No). The did you question asker can choose at this point to go on to something else, to
as Garfinkel (1967) noted let it pass, or to pursue an expansion to the did you question.
Gordon chooses to pursue more (Are you going to?), perhaps in search of an explanation to the logical presupposition that he believes Denise had an opportunity to pass on his
advice to Suzy. Denise pauses very slightly and tells him No again with no expansion. At
this point, Gordon expresses disbelief:

I dont believe you



Youre irritable

Denises utterance concerning Gordons irritability shifts the conversations topic from
Denises past actions to Gordons present actions and the explanation of her no answer
to the did you question is dropped. These examples show that a did you question can
be problematic for interactants when an asker wants an expansion or an account that is not
forthcoming. The pursuit of an expansion can be as problematic as the did you question
This chapter has shown that did you question can be problematic for interactants. Not all
did you questions are problematic, but enough are that they are recognized as being problematic by recipients and speakers (Tracy & Naughton, 1994). This chapter also shows that
there are three aspects of a did you question that can foster problems for recipients. The
first is that did you questions are rich in logical presuppositions and pragmatic presuppositions. A specific did you question may not be problematic, whereas the truth of the
underlying logical presupposition may be a problem. Given the presupposition richness,
did you question may not suffice to indicate understandable context, a violation of pragmatic presuppositions. The final problematic aspect of did you questions is the pursuit of
an expansion to the did you question and the problems this may cause the recipient.
Questioning the expected past actions of another would not, on the surface, appear to
be the source of a problematic interaction. However, closer inspection of actual did you
questions reveals aspects with problematic potential.

Interactional problems With did you questions and responses 147

The author would like to extend a special thanks to Robert Hopper for reading innumerable
drafts of this chapter as both a second-year doctoral project and a comprehensive exam
Garfinkel, H. (1967). Studies in ethnomethodolgy. Cambridge, England: Polity Press.
Keenan, E.L. (1971). Two kinds of presuppositions in natural language. In C. J.Fillmore and
D.T.Langendoen (Eds.), Studies in linguistic semantics (p. 4452). New York: Holt, Rinehart &
Levinson, S.B. (1983). Pragmatics. Cambridge, England: Cambridge University Press.
Pomerantz, A. (1984). Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. In J.M.Atkinson, & J. Heritage (Eds.), Structures of social action: Studies
in conversation analysis (p. 57101). Cambridge, England: Cambridge University Press.
Schegloff, E., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the organization of repair in conversation. Language, 50, 696735.
Schegloff, E.A., & Sacks, H. (1984). Opening up closings. In J.Baugh & J. Sherzer (Eds.), Language in use (p. 6999). New York: PrenticeHall.
Tracy, K., & Naughton, J. (1994). The identity work of questioning in intellectual discussion. Communication Monographs, 61, 281302.

Managing Optimism
Wayne A.Beach
San Diego State University

Examining how family members talk through a loved ones cancer on the telephone reveals,
as a central concern, the interactional construction of hopeful and optimistic responses
to uncertain and potentially despairing cancer circumstances. I refer to such recurring
moments as managing optimism1 in talk about cancer. This chapter focuses on an initial
collection of seven excerpts wherein optimism emerges as a resource for family members
as they update, assimilate, and commiserate about cancer diagnosis and treatment. These
materials are drawn from a set of 54 recorded and transcribed phone calls comprising the
first natural history of a family talking through cancer, from Moms initial diagnosis until
her death, some 13 months later.2 Only phone calls #1 (involving Dad and Son) and #2
(Dad, Son, and Mom) of the corpus are examined, interactions drawn from a collection of
more than 100 instances where speakers engage in optimistic collaborations.

It was Robert Hopper who coined the phrase managing optimism to depict a wide range of
moments for dealing with bad and uncertain news by remaining hopeful about his health condition. This description first emerged within weeks following a diagnosis of colon cancer, during one
of a series of phone calls with me wherein his illness trajectory routinely (though not exclusively)
became an explicit topic for discussion. Following his summary of what doctors had told him about
ongoing test results, attention was given to the inherent (and often frustrating) uncertainties of
medical knowledge, including doctors being unwilling and apparently unable to lay out, in specific
terms, just what his prognosis for overcoming cancers debilitating effects might be. In the face
of more basic yet unanswered questionsHow long do I have to live? What probability for healing exists? What impacts will further treatments have?our talking about cancer diagnoses and
impacts routinely shifted to being optimistic, reassuring, at times even upbeat about the ambiguities such bad news entails. And it was in response to our being hopeful together that Robert stated
something like Managing optimism. Thats what Im calling what were doing, as a practical
Family members include the Son, Father, Mother, Daughter, Aunt, and Grandmother. The corpus
also includes an assortment of other conversations between the Son and his ex-wife, the ex-wifes
brother, representatives from various airlines (when seeking flight information and reservations), an
academic counseling office receptionist, a receptionist at an animal boarding kennel (when making
and canceling reservations for his dog during his travel), a woman the Son had begun dating, an old
friend from St. Louis, a graduate student who covered the Sons classes during travel, and a variety
of other calls involving routine daily occurrences (e.g., the payment of bills, leaving messages on
phone answering machines).

Managing optimism 149

Unique opportunities are provided when health-related family conversations are closely
inspected over an extended period of time.3 As Kubler-Ross (1969) observed years ago in
reference to different stages that people go through when they are faced with tragic news
defense mechanisms in psychiatric terms, coping mechanisms to deal with extremely difficult situations The one thing that usually persists through all these stages is hope (p.
138). In the data that follows, preliminary insights into such phenomena such as defense/
coping mechanisms and stages can be tied to specific social actions. More recently, in
his ethnographic study focusing on the social meanings of death in three hospital wards
dealing with seriously ill patients, Perkyl (1991) referred to hope work as a predominant set of practices whereby patients are getting and feeling better (curative and palliative care) or past recovery (where hope per se is dismantled). In contrast, focus here rests
not with medical staff working with their patients in institutional settings, nor attempts to
legitimate medicine by professionals, but with laypersons speaking together on the telephone within their home environments (though, as in call #2, Mom is in the hospital when
Son phones from his home).
As with Perkyl, (1991) findings, it is not necessary for hope to be explicitly named.
At times hope is invoked in situated and thus revealing ways in the data examined herein.
And though not a single instance of the word optimism has yet been identified, speakers actions are shown to display a sense of expectancy, even assurance, about a hopeful
As a preview of more complete data to follow, consider the following seven excerpts:


Dad: So .hhh n:o:: I would hope by Monday or Tu:esday




.hhh But (0.2) she did have two nice things ha:ppen today.
No theres nothin to say. >You just-< .hh Ill- Ill wait to talk to Dr.
Leedon today.=Hes the cancer man, and =
My only hope- I mean- (.) my only choice.
Well wheres our magic wand Mom.
.hh Is find a reason to keep fighting and (.) to keep being hopeful.
See, [there] theres a small battle=





Only alluded to in this chapter, research focusing on longstanding concerns with social aspects of
death and dying (e.g., see Sudnow, 1967; Kubler-Ross, 1969, 1974; Perkyl, 1991, 1993, 1995;
Holt, 1993), troubles-telling sequences (e.g., see Jefferson, 1980, 1984a,b, 1988, and chapter 13 in
this volume; see also Sacks, 1992), and interrelationships between the delivery and receipt of good
and bad news (e.g., see Maynard, 1996, 1997, in press) are more fully addressed in related and
ongoing papers (e.g., see Beach, in press b; Beach, 2000 a,b).

150 Studies in language and social interaction


[( )]
=That weve won.=

Only Excerpts 1, 4, and 6 reveal hope/hopeful as being invoked, and then in similar yet
contrasting ways: in Dads reference to medical procedures (1), a personal reflection on
Moms ill-fated circumstance (4), and her display of perseverance and tenacity (6). Yet the
other instances are also somehow related to hopeful and optimistic orientations: As Dad
lightens prior and serious discussion (2), Mom waits and relies on news from the cancer
doctor (3), Son invokes and Mom responds seriously to magic, and Sons later attempts
to edify and simply cheer Mom up (7) in response to a story she initiates.
As a whole these moments reveal managing optimism to be a practical matter for
family members, talk that is shown to be designed in alternative (at times even humorous) ways while working through troubling illness circumstances. Analysis proceeds by
giving attention to the interactionally achieved and contingent features of each successive
moment, in its natural and emergent order, to discover what might be learned about how
speakers manage various optimistic concerns.4
Hope and Uncertainty Regarding Medical Diagnosis and Procedures
We begin with the initial instance, where hope is explicitly mentioned in the midst of
talking through a family members cancer. In Excerpt 8 as follows, Dad continues by
reporting to Son a doctors description of procedures for treating Moms cancer. In Line
3, these procedures include contacting a cancer specialist and conducting this bo:ne scan
thing tomorrow.:


.hhh He said he would have somebody else look in on

her:.=He also co:ntacted this cancer specialist so

he will be in Monday. (.) .hhh And they will do this

borne scan thing tomorrow. So .hhh n:o: : I would


by Monday or Tu:esday (0.7) pt they have <pin:ned

do::wn>(0.7) the particulars of what theyre after.

This analytic exercise is part of a more encompassing project, designed to capture not just patterns
of interactional conduct co-enacted by family members facing cancer but also three interrelated
sets of activities: a time-line sense of chronology for family members undergoing cancers development; a grounded understanding of how conversations get progressively constructed from prior
interactions, as resources forming the basis for organizing here-and-now problems and their solutions (see Beach, in press b); and (as noted) an extension and elaboration of the observed tendency
for good topics to arise out of otherwise bad and troubling matters (see, e.g., Jefferson, 1984a,
Perkyl, 1991, Sacks, 1992, Maynard, 1997).

Managing optimism 151



>Now they may not have< the course of action all

figured out, but [ .hhhh]
[Umhm] =
= Theyll at least kno:w. (.) .hh And maybe this
is just simplistically in my mind >but theyll
know< .hhh what ki:nd? theyre dealing with.
That way they should know .hhh how quickly does
it spread (.) what is- (0.7) what can be done to:
to stop it >you know< .hh radiation [or chemotherapy

Following I would hope in Line 4, Dad makes reference to two basic features of cancer
treatment: when something might be known and what theyre after. Immediately next,
however, he disclaims by stating >Now they may not have< the course of action all figured out, which is quietly and briefly acknowledged by Son. Dad then proceeds by elaborating his lay understandings of what he was hopeful about, namely, bottom-line concerns
with identifying the cancer and attempting to stop it with radiation or chemotherapy.
Several features of Excerpt 8 are interesting but not unusual throughout the Malignancy phone calls. First, this excerpt represents the initial display of hopeful conduct-ininteraction. These actions follow Dads initial and extended delivery, and Sons receipt and
assimilation, of bad news regarding Moms cancer (see Beach, in press; 2000c; Maynard,
1996; 1997; in press). Second, a delicate and countervailing balance exists between hope
and uncertainty. Notice again that Dads expression of hope (Line 4) is mitigated with a
next-positioned caveat: a course of action (Line 7) replete with incomplete knowledge.
Third, Dad must inevitably rely on, and report about, what doctors have told him about their
specialized knowledge. It is clear that Dads source of hope is anchored in the involvement
of assumedly competent medical providers, professionals who are expected to do everything possible while devising a plan for halting the insidious progress of Moms cancer.
However, his attempts to describe doctors suggested treatment options to Son (e.g., this
bo:ne scan thing in Line 4, and later to simplistically in my mind in Line 11), reveal
Dads lay attempts to understand complex medical procedures and the technical expertise
comprising bone scan procedures. Qualified and simplified moments such as these, involving lay constructions of medical knowledge and procedures, are given considerable attention by family members throughout the course of Moms cancer. Inevitably, each identified
moment reveals some problems in offering medical descriptions, but also optimism about
ongoing treatment and diagnosis.
Shifting from Bad to Good News
For approximately 1 minute following Excerpt 8, Dad continues by describing to Son how
Moms original neck problem, some 35 years ago at 25 years of age, was a slow growing
lymphatic cancer. He then raises the possibility that Moms current cancer may also be
slow growing, which bone scan results will aid in determining. In Excerpt 9 which follows,
Dad summarizes what is essentially a bad news description of how Mom was doing. His

152 Studies in language and social interaction

portrayal escalates in its telling, from Moms co:nf irmation and resignation I just
hurt too b:ad to be anything else something drastic.:


A:: yeah .hhh (.) But she seemed to be doing (.) >as I

said< pt .hh at this point it was mostly (0.5)

co:nfirmation and resignation.

[Mmhmm: .]
[Cause she] said, .hhh I just hurt too b:ad to be

anything else (0.2) >ya know.< It ha: :d to be

som- (0.7) something drastic.

And she was really having some problems with pa:in

today. She had .hh one and a half (0.2) >percodans<

in her and it wasnt hardly slowin it down.

Son:Mmm wow.

Dad:.hhh But (0.2) she did have two nice things


today. She was on her way do:wn and .hhh and was
kinda, depressed or concer:ned I guess with having
>to go down< for these needle biopsies and Will?
showed up.

In Line 12, following Dads progressively distressing update, Sons Mmm wow. displays a shift from acknowledging Dads description-in-progress (i.e., with Mmhmmm:
and Mmhm) to quietly assessing it as troubling news. This response is treated by Dad
as Sons unwillingness to comment further, and not inviting Dads further elucidation of
Moms painful condition. Immediately following Sons Mmm wow., Dad initiates
transition to a new but related topic with his pre-announcement But (0.2) she did have
two nice things ha: ppen today. As an upshot of Sons closure implicative action, this
conversation restart (see Jefferson, 1984b, p. 193; see also Jefferson, 1996; Sacks, 1992)
reveals how Dads insertion of good news is on-topic, yet designed by him to ease the
burden of previously articulated grievous circumstances about which enough had been said
(at least for now).
Notice also that Dads kinda, depressed or concer:ned (Line 15) was inserted following his pre-announcement, yet before announcing the good news that Will? showed up.
Here, as with how Dad and Son collaborate on reporting bad news as a prelude to announcing good news, the close proximity of Moms reported mood, immediately prior to an old
friend showing up for a visit, reveal how everyday life is comprised of tightly interwoven
relationships among bad and good circumstances. It also illustrates how the valence of
social occasions are subject to change and alteration, literally on the cusp of interactional
time (see Maynard, 1997; in press).
The shift from bad to good news evident in Excerpt 9 is also similar to Holts (1993)
findings involving death announcements by tellers, particularly to recipients not especially
close to the deceased. In each of the 10 instances she examined, the tendency to treat the

Managing optimism 153

death of an intimate or acquaintance as bad news nevertheless eventuated in movement to
a bright side sequence revealing some positive stance toward the news (e.g., deceased
persons: worked until the time of death, died peacefully and in so doing solved problems
associated with prolonged illnesses and caregiving tasks, or had the opportunity to say
goodbye to people providing for a funeral that is less dismal). Holt observed that there
seems to be a strong link between bright side sequences and topic termination (p. 208),
not uncommonly termination of a phone call. In Excerpt 9, two exceptions can be noted.
First, Dad transitions not just to a closely related topic, but to a decidedly positive orientation to updating news. His actions reveal how the shift from bad to good news is as an
apparent resource for facilitating closure to a discussion that Son initially, and next Dad,
treated as a delicate matter. Second, in Excerpt 9 not only is good news about friends
unexpected visits elaborated, but the phone call continues for more than 15 minutes. This is
not surprising, however, because this is the first phone call between Dad and Son regarding
Moms malignant diagnosis. Perhaps even more important, however, is that a loved ones
cancer is consequential for family members. Recipients not close to the deceased neednt
be directly concerned about primary family troubles (Beach, in press). Family members
routinely (often closely) monitor the course and progression of a loved ones illness, experience anxiety regarding the future, and grieve together for the possible or probable loss of
a family member with whom extensive history is shared.
In two contrasting yet related interactional environments, Dad and Son have been shown
to collaborate in managing optimism regarding Moms cancer: In Excerpt 8, hope was
explicitly named and commented on by Dad; in Excerpt 9, talk about good news emerged
out of prior bad news descriptions. In both instances, Dad was reporting on prior incidents
involving medical staff and procedures, the latter focusing on how Mom was doing including problems with pain medication. These two instances were drawn from the first phone
A more extended instance appears in the following Excerpt, but in this case during the
second phone call, the very next day, between Son and Mom. A revealing glimpse of Moms
construction of her own cancer dilemma is evident in three ways: as she relies on medical
procedures and providers as sources of information and thus attributed (but not named)
hope, as hope gets mentioned but quickly corrected by her in favor of choice regarding
radiation and chemotherapy, and as keep fighting gives rise to being hopeful:
((Mom has just informed Son that her cancer has been diagnosed as a very fast growing adenoma type-an update from call #1, where Dad was not aware of the general cancer classification, nor whether Moms cancer was slow or fast growing. Mom has just reported that since
very few people respond well to treatment, and those who do live five years or less, Its real
And uh: >I dont know what else to tell you.<


154 Studies in language and social interaction





.hh hhh Yeah. (0.2) um- ((coughs)). Yeah, I

dont know what to say either.


No theres nothin to say. >You just-< .hh Ill

Ill wait to talk to Dr. Leedon today.= Hes the

cancer man, and =


= Um hmm.

10 Mom:

See what he has to say, and (0.4) just keep goin


forward. I mean I might be real lucky in five


years. It might just be six months.



14 Son:


15 Mom:

Who knows.

16 Son:

Phew: : .

17 Mom:


18 Son:

.hh hhh (0.4) Whadda you do: with this kind of


thing. I mean- (.)

20 Mom:

>Radiation chemotherapy.<



22 Son:

Oh bo:y.

23 Mom:




25 Mom:

My only hope- I mean- (.) my only choice.

26 Son:


27 Mom:

Its either that or just lay here and itll kill me.



29 Mom:

And thats not the human condition.

30 Son:

No. (1.0) I guess [not.]

31 Mom:

[No.] (.) So thats all I can


tell you.

It appears, at least initially, that Mom and Son collaborate in exiting from the topic of cancer. Both speakers utter I don t know (see Beach & Metzger, 1997), first in Line 1 as
Mom claims she has nothing further to tell, and next in Lines 4 and 5 as Son affirms that, as
recipient, he does not know what to say. In this sense there is indeed nowhere else to go
(Jefferson, 1984b, p. 191), and Lines 15 bring closure to further talk about the seriousness
of Moms prognosis.
Yet Lines 15 also demonstrate a transition to talking with her cancer doctor,
which Mom initiates in Line 6. As the conversation unfolds, it becomes clear that the

Managing optimism 155

insufficient knowledge they claim, and display an inability and/or unwillingness to talk further about, is tied only to Moms prior diagnosis (most notably the anguish Moms immediately prior news makes available) and not her ongoing treatment. Three features of particular
relevance to managing optimism emerge in Lines 632.
First, Moms No theres nothin to say. is one form of an extreme case formulation
(see Pomerantz, 1986), employed here to emphasize her position and to terminate her diagnostic update for Sons hearing. Next, notice that Moms I 11- Ill wait to talk to Dr.
Leedon today. = He s the cancer man, (Lines 78) implicates her having cancer without explicitly stating it. This is but one instance representing a larger collection where the
word cancer is noticeably absent and, at times, apparently and actively avoided. In this
moment, where Mom clearly has been diagnosed with cancer but fails to directly state it,
she is nevertheless left with the task of formulating herself as a sick person. One practice
for doing so, which Mom employs here, is to make reference to a provider-patient relationship in which she is involved. Thus, the professional expertise of cancer man, provides
one solution to directly stating I have cancer. And by stating See what he has to say,
(Line 10), Mom situates herself as recipient for obtaining any new information the doctor
might impart. Only the doctor has the expertise to announce any new, potentially good,
and more or less definitive news regarding her acute medical condition. A central feature
of just keep goin forward. (Lines 1011), therefore, involves waiting for the doctor and
whatever news he might disclose. As updates about Moms terminal illness evolve, this is
but one instance of how faith in your doctor is grounded in moments where waiting is
explicitly stated, whereas the possibility of hopeful news is only implied.
Of course, there is no guarantee that any update of her condition will amount to whatever good news might imply. This is revealed straightforwardly through Moms selfrepaired I mean I might be real lucky in five years. It might just be six months. (Lines
910). When 5 years is considered fortuitous, just what might constitute good news is an
altogether relative notion here. (As noted previously, Moms death occurred 13 months
following diagnosis.) Clearly, in Lines 1417, uncertainties surrounding such an illness
trajectory make it problematic for Son and Mom to do more than assimilate the quandary
they are caught up within.5
Second, in response to Sons query in Lines 1819, whadda you do: with this kind of
thing. I mean-, Mom immediately and quickly replies >Radiation chemotherapy. <. By
forwarding medical procedures as forms of treatment regimen, Mom also avoids addressing what Son may very well have been pursuing: more personal issues involving her coping (e.g., fears, anxieties, anger) with what appears to be a terminal diagnosis. Whether
Son was in fact soliciting and thus inviting Mom to talk further about her feelings remains
unclear. What is apparent is that by responding in this manner, Mom is managing optimism through steadfast reliance on medical protocol that, for now, is put forth as critical
to j ust keep goin forward. (Lines 1011).
Third, it is her resoluteness that Sons delayed and assimilating Oh bo:y. response
seems to address (Lines 21 & 22), which Mom next affirms en route to an explicit yet
Work in progress (Beach 2000) is focusing on a collection of similar moments where few words
are enough in the course of assimilating bad news (e.g., Jesus, Oh boy, Oh wow, Phew, Yuck).

156 Studies in language and social interaction

fleeing reference to hope: My only hope-I mean(.) my only choice. This is a curious self-repair, where hope and choice are at once treated by Mom as interwoven
yet distinct, an explanation for which might be gleaned from prior discussion: In light of
her 5-year prognosis as a best case scenario for life expectancy, any hope emerging from
radiation and chemotherapy is restrictive; such treatment options offer little certitude nor
assurance of healing her cancer. Thus, in this utterance, hope and the optimism it may
engender appears to give way to my only choice., which is itself clearly restrictive and
further legitimizes her decision making (see Pomerantz, 1986). It is not really a preference
but an ill-fated necessity that Mom is orienting to. Addressed in no uncertain terms in Lines
25 and 27, Mom displays an essential unwillingness to be passive while allowing the cancer to kill me., which Son aligns with here (Line 26) and following Moms elaborated
And thats not the human condition. (Lines 29 & 30). Further, it is by reference to basic
human instincts for survival that Mom expresses her willingness to be treated through
radiation and chemotherapy.
In the final utterance of Excerpt 10, Moms So thats all I can tell you. (Lines 3132)
repeats tell you from Line 1, where Mom stated >I dont know what else to tell you
.<. By so doing she exhibits her departure from this portion of an extended storytelling,
which her cancer experiences entitle her to reveal (see Sacks 1984, 1992).6 One consequence is that, through word repeat, her story ending is punctuated in a manner not providing further access to Son who, as story recipient, does not further pursue what his Whadda
you do: with this kind of thing. I mean-, (Lines 1819) may have been designed to address
(e.g., Moms personal feelings). Nor can he address the scenic particulars constructed in
Excerpt 10 by himself, as it is clearly Moms story to tell.
Invoking and Responding to Magic
What Son does do, however, is proceed with his own story, informing Mom that he is
aware of how the medication she is on can make her depressed. Mom then informs him that
her diagnosis is very serious because the cancer has metastasized. (These data are not
included here; 31 Lines were deleted between Excerpts 10 and 11, which follows). Next,
Son takes the initiative to shift orientation to cancer problems by invoking magic:


Son: Well wheres our magic wand mom.
Mom: $It- he$ (.) Beats the hell out of me.

Mom: I guess the o:nly thing: (.) I: can do: is (.)

after Im done ree:ling from this.

Mom: .hh Is find a reason to keep fighting and (.) to keep being

Schegloff s (1999) analysis of word repeats at turn endings reveal a similar resource: Tellers display their entitlement to initiate closure to stories only they are capable of narrating.

Managing optimism 157





hopeful. (0.5) You know that- thats about all you 9

can do. >Thats all a person can do.<
How can you do: that. (0.2) Thats [gotta]=
=be tough. >I mean-< I dont mean to sa:y that
sounding like a
Here comes your Papa: : .

In Line 1 Son achieves two key actions. First, through our he assumes ownership of
Moms illness predicament by making them out to be problems that can be faced together
(see Beach, 1996). This is but one relational and commiserative display of being with
(see Beach, in press-b; Goffman, 1963, 1971; Mandelbaum, 1987) that was obvious yet
implicit in prior discussion. Next, magic wand offers more than wishful thinking. It also
injects a sense of humor and brightness into a serious health scenario, one that is literally
no laughing matter, and (based on prior actions) apparently a set of dire circumstances
preventing Mom from being capable of uplifting herself.
In responding with $it- he$ (.) Beats the hell out of me. (Line 2), Mom in turn
accomplishes two key actions. First, her initial attempt at laughter ($), though quickly
aborted, nevertheless treats Son as having made an effort to invite such laughter through
his magical refraining of such critical topics. Curiously and next, however, Mom acts as
recipient of her own telling situation by producing a despairing and recognizably serious response (Jefferson, 1984a, p. 346). By so doing Mom again appears unable and/or
unwilling to take the trouble lightly and thus act in a troubles-resistant fashion (Jefferson,
1984b, 1988). Rather, and understandably so, she is totally engrossed in (and ensnared by)
her diagnostic dilemma.
But there is more here, a poetic and delicate preoccupation evident in her unwitting
and quietly tailored Beats the hell out of me. (see Beach, 1993, 1996; Hopper, 1992;
Jefferson, 1996; Sacks, 1992; see also chap. 13, this volume).7 Beginning with how the
word Beats adds valence and thus pragmatic force to Moms description, it stands in
marked contrast to how magic wands are typically employed (i.e., through a simple waving, which is sufficient to achieve magical consequences). And in unison with Beats
as a lexical choice reflecting the kind of force required to drive cancer out of her body,
so does her extended utterance precisely characterize an unintentional sensitivity to the
very troubles at hand: If a magic wand could heal an illness approaching hopelessness,
it would literally exorcise a dark and foreboding force from hell that stifles rather than
improves living.
The phrase Beats the hell out of me. may be added to the collection of idiomatic expressions as Drew & Holt (1988, 1998) have analyzed them (e.g., its gone tuh pot, down the
tubes), in that it is an utterance occurring in a sequential environment clearly involving complainable matters (i.e., a serious cancer diagnosis). However, this interactional moment is unique in
this sense: While Drew & Holt (1988) have shown that such complainable matters are routinely
directed to others treatments of them, here Moms utterance is not treating her Son as the source of
the trouble but the illness she is enduring and its varied consequences.

158 Studies in language and social interaction

Following his humorous attempt to uplift Moms condition, Son next withholds further commentary to her tepid response (Line 3). But the despair evident in her reply is
only momentary (see also discussion of Excerpt 12 in the next section). As revealed in
Moms next I guess the o:nly thing: (.) I: can do: is (Line 4), she continues by specifying that there are uncertain and limited options for coping with cancer. This utterance is
consequential in three key ways.
First, it prefaces her insertion after I m done ree:ling from this., a bewildering formulation referencing her here-and-now reaction to a malignant diagnosis (what Dad had
earlier and apparently portrayed as co:nf irmation and resignation. in Excerpt, Line 3).
Second, it also sets up Moms .hh Is find a reason to keep fighting and (.) to keep being
hopeful. (Lines 78). Framed as an ongoing and practical matter, Mom sketches out a
procedure for living with and through her cancer that exemplifies basic survival instincts
underlying the human condition (see Excerpt 10, Line 29, earlier). As she constructs it,
remaining hopeful requires motivated fighting, two interwoven yet distinct actions that
facilitate the search for reasons to live.
Third, it is interesting that a key portion of Moms I guess the o: nly thing: (.) I: can do:
is (Line 4) is repeated two more times in Lines 89: You know that- thats about all you
can do. >Thats all a person can do.<. Notice that whereas can do gets repeated, Moms
attempt to inform her Son evidences a movement from I you person. In unison
with her use of meIIm in Lines 25, Moms description becomes progressively
less my-world centered as she endeavors to manage optimism in the face of bad diagnostic
news. This stepwise shift, beginning with a revelation of her experiences yet ending with a
generic person, accomplishes three critical and interrelated actions:
1. While falling short of magic, Mom reveals herself as doing all she can within her
unique circumstances. She first discloses then normalizes her lived reality as an ordinary
feature of illness management, an orientation common for others dealing with cancer predicaments (with whom she is now indirectly yet directly associated) as well.
2. By invoking third-person characterizations, Mom distances herself by utilizing you
and person as devices for coping with the apparent inevitability of death. Through thirdperson references, her illness problems become less intimate and thus more easily managed
at a time when, clearly, coming to grips with dying is inherently problematic.
3. Mom is also designing her talk in consideration of Sons hearing, and even protection,
from having to directly confront a hopeless terminal illness. She is not disattending his
prior and attempted uplifting of the dire situation, but (as best as possible) being responsive
to it. Though her current disposition can be explained as reeling, it is only temporary:
Her confusion will give way to a more determined and hopeful condition, a fighting
perseverance that Son can himself be hopeful about.
Excerpt 11 draws to a close as Son continues by further pursuing just how Mom can
remain hopeful (Lines 10 & 12), a solicitation that is preempted with Moms announcement that Papa has just entered the room.

Managing optimism 159

A Story and its Consequences: Fighting the Battle Together
As Mom exits from talking (not shown in Excerpt 11), Papa and Son continue talking for
nearly 5 minutes about fixing cars together8 and an upcoming chili dinner Son has prepared
for when Mom returns home from the hospital. Son then requests to speak with Mom once
again and announces his dinner plans to her. It is at this juncture that Mom initiates the
following story about a sign the Son had placed in her hospital room:


>By the way< your sign Do not take me really worked.
$Did it?$
Totally confu:sed one girl. She looked, and she looked,








and she looked. Now this is a little oriental gal. =

= Mm hm.
<And ah> (1.0) [she went out]
[Do not] ah (.) take oh ( .) me
(.) [ha]
[She] went out and she brought in >ya
know< those things have liners?
Mmm hm.
She brought a li:ner of like a- of clear water in to
set it there.
Mm hm mm hm [hm]
[She] didnt- she couldnt quite figure
that whoile thing out. >But she wasnt gonna touch it. <
Mmm. (.) Good.
So that was kinda funny.
See, [there] theres a small battle=
=That weve won. =
$Right, [right,] right$

While it may appear that fixing cars together is of little relevance to understanding the interactional management of cancer predicaments, quite the contrary is the case. It is revealing to examine
just what everyday topics find their way into the midst of cancer topics, how and when they
appear and are terminated, that are seemingly not about cancer per se. For example, in this instance
of fixing cars together, is it coincidental that Dad and Son move together to talk about 1) something they are both knowledgeable about, that 2) they can thus (with some confidence) diagnose
togetherin stark contrast to technical matters of cancer diagnosis and treatment? Analysis of a
larger collection of of topic organization suggests otherwise.

160 Studies in language and social interaction




An(d) thats all ya can do is jus- just

[rack up the] sma:ll battles.
[Rirght ri:ght uh mm]
Well .hhh Well okay. Im gonna let you go:.
((Mom & Son move to phone closing.))

That Mom even initiated such a humorous story displays her attempt to lighten what had
become, prior to Son and Dads conversation, a very serious discussion of both her diagnostic condition and orientation to coping. Further, she also acknowledges Sons thoughtful effort to meet her needs, by his placement of a Do not take me sign, which stands
in contrast to her prior tepid and momentarily despairing response to his well where s our
magic wand mom. (Excerpt 11, Lines 12).
Taken together, the actions built into this shift in topic mark a contrast in Moms
demeanor: They are remedial in just the ways Moms initiation of this particular story
appears designed to invigorate her earlier and displayed unwillingness and/or inability to
take her troubles lightly, and to display appreciation for Sons ongoing concerns with her
illness predicament.
This marked shift in Moms disposition does not go unnoticed by Son. In response to
her reference to little oriental gal. (Line 4), Son collaborates by personifying the girls
scenic reaction with a stereotypic [Do not] ah (.) take oh (.) me (.) [ha] , a voiced switch
in identity (see Beach, 2000a) he treats as humorous with his final [ha]. Next, it is of particular interest that when Mom brings the story to a close (Line 19), Son relies on Moms
initiated story to revisit yet extend their earlier discussion (Excerpt 11): He retopicalizes
and reframes Moms immediately delivered story (i.e., fighting battle, our magic wand
weve won). In these ways, Son shows sensitivity to Moms keep fighting and (.) to
keep being hopeful., while simultaneously treating this as a moment for reemphasizing
that they are indeed facing the problems together.
Following Moms aligned recognition and their shared laughter (Lines 2324), Sons
An (d) thats all ya can do is jus- just [rack up the] sma:ll battles. (Lines 2526) offers
a prototypical summary that reinvokes all ya can do. Apparently, this utterance overextends an otherwise well-taken point, however, as Mom interjectively moves to close down
Sons contribution (Line 27) and end the phone conversation together (Lines 2829).
Faced with a serious and uncertain cancer diagnosis, and thus in the very midst of emergent
troubles and possible despair, family members rely on hope and optimism as resources for
dealing with and attempting to ease burdens arising from the often harsh and restrictive
impositions of such illness circumstances. Just as it has been observed that research on
the connections between hope and social psychological functioning is minimal in cancer
research, and that maintaining] a sense of control is an essential determinant of how
cancer patients cope with their illness hopefully (Bunston, Mings, Mackie, & Jones, 1995,
p. 79), so can it be noted that perhaps even less is known about what comprises hope and
control as interactionally organized moments of practical action.

Managing optimism 161

Although only calls #1 and #2 of the larger corpus were examined, managing optimism was nevertheless evident across an assortment of social actions:
Acknowledging the importance of medical personnel by steadfastly relying on medical
protocol and treatment procedures.
Lightening the discussion by shifting from bad to good topics.
Revealing how personal coping with cancer involves an inseparable relationship
between hope and restricted choices.
Offering collaboration in facing Moms illness together.
Humorously going even beyond hope by invoking magic when Mom understandably
displays an inability and deep preoccupation with not taking her troubles lightly.
Proposing fighting and being hopeful as basic survival instincts even when resistance to troubles is diminishing.
Doing all you can do to remain capable of hoping that healing might occur.
Clearly, then, such delicate instances are comprised of fine-grained subtleties through
which the process of managing optimism is being achieved. Ongoing analysis of the
larger collection of such moments (calls #3#54) will provide a useful and longitudinal
perspective for framing how the interactional activities examined herein are themselves
tied to, in fact constitutive of, key moments as Moms cancer progressed and was treated
until her death. Though yet further and critical implications require discussion, only four
can be briefly articulated here.
First, working to be hopeful together can also produce its own interactional dilemmas
in the midst of talking about other dreaded issues Perkyl, 1995). Further investigation
is needed into how the management of family relationships is itself an ongoing and often
problematic achievement, particularly when: a) doing the work of moving out of troubling
topics (e.g., Dads shift to good from bad news precipitated by Sons display that enough
had been said); b) moving talk forward even though family members express that they
do not know what to say (e.g., Mom and Son rely on few words when assimilating the
news together); c) initiating, pursuing, and responding to intimate and personal topics
(e.g., Son twice querying Mom about how she copes with her condition); d) uplifting and
compensating for responses to such edification efforts (e.g., Sons invoking magic and
Moms delayed telling of a funny story to counter her prior tepid response to his displayed
concerns); and e) in responding to Moms story Son further attempts to make the point that
small battles can be won together, which Mom interjectively initiates closure on by moving
to end the call.
Second, even a cursory inspection of these materials reveals that the query What makes
a family, a family? (e.g., see Gubrium & Holstein, 1990) is deserving of substantive,
interactionally grounded answers. Such matters as how supporting and commiserating get
interactionally managed, for example, are available to the extent they are anchored in family members practices for working as a team: when taking turns at being hopeful, injecting
humorous concerns into troubling circumstances, and working to protect one another from
fears and anxieties so often associated with death and dying. In these ways, useful contrasts might also be made with interactions among acquaintances. This chapter has shown
that bright side sequences are only one type of response available for family members

162 Studies in language and social interaction

dealing with cancer (see Holt, 1993), that the proximity and interwoven nature of good and
bad news is omnipresent, and that family members may display doing being a family by
making anothers problems their own in and through the ways they assimilate the news and
grieve together (see Beach, 1996, in press).
Third, regarding talk about troubles (see Jefferson 1980, 1984a, 1984b, 1988; see chap.
13, this volume), these family members appear remarkably sensitive to limitations on
serious topics, yet at times proceed to enact topic shifts without necessarily terminating
talking about cancer per se. How this ongoing work gets done also merits ongoing examination. Similarly, environments need to be more fully inspected when, following moments
where Moms ability to resist troubles essentially fails, she nevertheless rebounds, that
is, attempts to muster the energy required to rally her appreciation for Sons concerns and
to remain hopeful and optimistic. Further, if and when such issues as coping or defense
mechanisms are to be understood as interactionally generated and managed, as well as
stages of grieving (i.e., denial, anger, depression, bargaining, acceptance; Kubler-Ross,
1969, 1974), they must be shown to be more than psychological states wherein individuals
experiences are ultimately the units of analysis. By inspecting how family members mutually coordinate their orientations to illness predicaments and various health concerns over
time, it may become possible to describe and substantiate temporal shifts interactionally,
that is, by elucidating the social actions comprising developmental aspects of coming to
grips not just with death and dying but, even more broadly, all aspects of illness progression. A key feature of these discoveries will likely involve understanding how prior discussions, such as what the doctors told them, are employed to constantly shape and update
understandings about Moms condition (see Beach, 2001, in press). Little has been said in
this chapter about such carry over recurrences, even though the data make available such
possibilities for analysis, albeit in limited fashion (e.g., as with Dad and Moms references
to medical staff).
Finally, as described earlier (see Footnote 1), I did not invent managing optimism
as a technical term for labeling social actions of the kind examined here. But it seems an
apt description. Having been diagnosed with cancer, and just beginning to realize social
aspects of talking with others about his diagnosis and treatment, it was Robert Hopper who
observed the tendency to remain hopeful as uncertain and even bad news emerged. Given
marked contrasts between self-reporting about versus enacting social actions collaboratively in real time, it is interesting (yet perhaps not surprising) to note that the kinds of
interactional contingencies examined in this chapter extend considerably beyond those he
identified in more general terms. Similarly, the experiences and interactional involvements
of a cancer patient (with medical staff, family members, friends, and colleagues alike) are
much broader than what any single phone corpus might capture. And so it should also not
be unexpected that Robert cited other kinds of encounters central to managing optimism,
only three of which I mention here, activities involving both those undergoing cancer and
others talking with them about it: a) acting as though everything is all right when it
obviously is not, b) literally calibrating and coordinating just what and how something
might be said, if anything, yet without appearing morbid about the illness, and c) when
talk about the same cancer arises, but within different relationships comprised of varying
degrees of background and intimacy, what problems (if any) emerge as attempts to discuss

Managing optimism 163

and describe the illness and its prognosis are modified (e.g., when disclosure is solicited
and/or voluntary, withheld and/or pursued)?
Living with and through cancer, and an array of other chronic and lifethreatening illness
(e.g., see Packo, 1991), occasions diverse circumstances where managing optimism is
interactionally achieved. Only selected and comparably few instances have been introduced in this chapter. It is obvious and compelling, however, that the full social milieu of
cancer quandaries, involving what communicators do, not what scholars have validated
(Hopper 1981, p. 209), remain largely unearthed and thus taken-for-granted.
This research was made possible through funding provided by the American Cancer
Society (Grant #ROG-9817201).
Beach, W.A. (1993). The delicacy of preoccupation. Text and Performance Quarterly, 13, 299312.
Beach, W.A. (1996). Conversations about illness: Family preoccupations with bulimia. Mahwah,
NJ: Lawrence Erlbaum Asociates.
Beach, W.A. (2000a). Inviting collaborations in stories about a woman. Language in Society, 29,
Beach, W.A. (2000c). When few words are enough: Assimilating bad news about cancer. Unpublished manuscript.
Beach, W.A. (2001). Stability and ambiguity: Managing uncertain moments when updating news
about moms cancer. Text, 27, 221250.
Beach, W.A. (in press). Between dad and son: Initiating, delivering, and assimilating bad cancer
news. Health Communication.
Beach, W.A. & Metzger, T.R. (1997). Claiming insufficient knowledge. Human Communication
Research, 23, 562588.
Bunston, T., Mings, D., Mackie, A., & Jones, D. (1995). Facilitating hopefulness: The determinants
of hope. Journal of Psychosocial Oncology, 13, 79104.
Drew, P. & Holt, E. (1988). Complainable matters: The use of idiomatic expressions in making
complaints. Social Problems, 35, 398417.
Drew, P., Holt, E. (1998). Figures of speech: Figurative expressions and the management of topic
transition in conversation. Language in Society, 27, 495522.
Goffman, E. (1963). Behavior in public places. New York: The Free Press.
Goffman, E. (1971). Relations in public. New York: Basic Books.
Gubrium, J.F. & Holstein, J.A. (1990). What is family? Mountain View, CA: Mayfield.
Holt, E. (1993). The structure of death announcements: Looking on the bright side of death. Text,
13, 189212.
Hopper, R. (1981). The taken-for-granted. Human Communication Research 7, 195211.
Hopper, R. (1992). Speech errors and the poetics of conversation. Text and Performance Quarterly,
12, 113124.
Jefferson, G. (1980). End of grant report on conversations in which troubles or anxieties are
expressed (HR 4805/2) [Mimeo]. London: Social Science Research Council.
Jefferson, G. (1984a). On the organization of laughter in talk about troubles. In J.M.Atkinson & J.
Heritage (Eds.), Structures of social action: Studies in conversation analysis (pp. 3463 69).
Cambridge, England: Cambridge University Press.

164 Studies in language and social interaction

Jefferson, G. (1984b). On stepwise transition from talk about a trouble to inappropriately nextpositioned matters. In J.M. Atkinson & J. Heritage (Eds.), Structures of social action: Studies in
conversation analysis (pp. 191222). Cambridge, England: Cambridge University Press.
Jefferson, G. (1988). On the sequential organization of troubles talk in ordinaryconversation. Social
Problems, 35, 418441.
Jefferson, G. (1996). On the poetics of ordinary talk. Text and Performance Quarterly, 16, 161.
Kubler-Ross, E. (1969). On death and dying. New York: Macmillan.
Kubler-Ross, E. (1974). Questions and answers on death and dying. New York: Macmillan.
Mandelbaum, J. (1987). Couples sharing stories. Communication Quarterly, 35, 144170.
Maynard, D.W. (1996). On realization in everyday life: The forecasting of bad news as a social
relation. American Sociological Review, 61, 109131.
Maynard, D.W. (1997). The news delivery sequence: Bad news and good news in conversational
interaction. Research on Language and Social Interaction, 30, 93130.
Maynard, D.W. (in press). Bad news, good news: A benign order in conversations, clinics, and
everyday life. Chicago: University of Chicago Press.
Packo, J.E. (1991). Coping with cancer and other chronic life-threatening diseases. Camp Hill,
Pennsylvania: Christian Publications.
Perakyla, A. (1991). Hope work in the care of seriously ill patients. Qualitative Health Research,
1, 407433.
Perakyla, A. (1993). Invoking a hostile world: Discussing the patients future in AIDS counseling.
Text, 13, 302338.
Perkyl, A. (1995). AIDS counselling: Institutional interaction and clinical practice. Cambridge,
England: Cambridge University Press.
Pomerantz, A. (1986). Extreme case formulations: A way of legitimizing claims. Human Studies,
9, 219229.
Sacks, H. (1984). On doing being ordinary. In J.M.Atkinson & J.Heritage (Eds.) Structures of
social action: Studies in conversation analysis (pp. 413429). Cambridge, England: Cambridge
University Press.
Sacks, H. (1992). Lectures on conversation (Vols. 12). Cambridge, MA: Blackwell.
Sudnow, D. (1967). Passing on: The social organization of dying. Englewood Cliffs, NJ: PrenticeHall.

Rejecting Illegitimate Understandings
Samuel G.Lawrence
University of Central Florida
Conversation analytic studies have demonstrated decisively that an architecture of intersubjectivity (Heritage, 1984) provides for the recurrence and stability of understandings
in talk-in-interaction. These studies describe interactants methods for accomplishing the
routine and tacit tasks of displaying, ratifying, and updating intersubjective understandings (Heritage, 1984; Sacks, 1992a, 1992b; Sacks, Schegloff, & Jefferson, 1974). This
architecture of intersubjectivity is a systematic by-product of turn organization:
[I]t obliges its participants to display to each other, in a turns talk, their understanding of other turns talk. More generally, a turns talk will be heard as directed to a
prior turns talk, unless special techniques are used to locate some other talk to which
it is directed. (Sacks et al., 1974, p. 728)
The outcomes of interpretive operations, performed upon the prior turn in the first position,
are publicly displayed in the next turn position. The selection of some next action (e.g.,
an answer), for example, exhibits its speakers understanding that the prior turn was a corresponding first action (e.g., question). The prior speaker, in turn, inspects the adequacy
of those displayed understandings and exhibits their (inadequacy in the third turn position. The products of these inspections may contribute to or briefly impede the continued
sequential development and directionality of the talk.
According to Heritage (1984), the third position slot may be used for implementing
actions that tacitly ratify understanding displays in next turn position. Alternatively, the
speaker of the talk in the first position, upon finding evidence of misunderstanding in the
next turn position, may initiate third-position repair (Schegloff, 1987b, 1991, 1992). For
example, D understands Ms deployment of Jeff as referring to her husband who is also
named Jeff. The displayed product of this understanding is the collective pro-term we.
This pro-term refers to the speaker and her husband.

UTCL: Mother-Daughter.2.192202
M: =How are things goin with her- uh her and Jeff?
D: Fine

D: Just fine, we havent seen much of h



mean your Jeff, I mean Jeff Over

very good ((continues))

166 Studies in language and social interaction

M re-performs the operations that D had performed on Ms turn in Line 1 and displays their
products, your Jeff, in the rejection component of the third-position repair (Lines 56).
Ms actions of re-performing these operations and displaying their products treat Ds misunderstanding as the product of a methodical and legitimately alternative, though incorrect,
analysis of Jeff (Line 1). In this regard, Schegloff (1992) observed:
It is striking that misunderstandings are both orderly and accessible to the speaker of
what has been misunderstood, who might well be thought to be so committed to the
design and so-called intent of the earlier turn as to be disabled from appreciating that
(or how) it could be otherwise understood, (p. 1307)
The orderliness and accessibility of misunderstandings to speakers of talk in the first position, however, are not givens because they may misunderstand the understanding display in
the next turn position (Schegloff, 1992). Additionally, speakers may reject an understanding display as an unwarranted or illegitimate analysis of the talk in the first position. In
these cases, the understanding display may be treated as intelligible on its own; however,
the speaker of the first-positioned talk may deny the reproducibility of that understanding as the product of some methodical analysis of the prior turn. Such understandings are
rejected not as misunderstanding but as misconstruing the prior turn.
The present essay is a single case analysis (Schegloff, 1987a) of an understanding
display that is rejected as misconstruing the prior turn. Analytical resources from turn,
sequence, and topical organization are utilized to explicate: (a) how the talk in first position
is occasioned, (b) how the next speaker analyzes the prior turn, (c) how the speaker of the
first-positioned talk rejects the reproducibility of that analysis, and (d) how the speaker of
the understanding display counters the rejection and provides for the methodicity of that
The data are taken from a telephone conversation between two college students. Dee
Ann had called to check whether Skeet was willing to lend his ticket to her. After indicating
that he needed the ticket and producing a topicbounding turn, Dee Ann used a topic initial
elicitor (Button & Casey, 1984) in Line 1 to create a slot in which Skeet may formulate
newly topicalizable materials, based on his current activity. After the repair sequence in
Lines 23, he reports his current activity as, Goin ta bed.

Skeet :
Skeet :

Skeet :

What Doin=h
Wha Im doin?
Uh huh
Coin ta bed
Are you really?

Rejecting illegitimate understandings 167




Skeet :
Skeet :

Skeet :


Skeet :



Skeet :










rya si.ck?
No Im jus tired.=
Went to bed too late las night.=
=(Yep) I donno why:
huh huh

my fault. ((spoken in an exaggerated

regional dialect))
I- (.) didnt siay that

You th(h) ought it awful l(h)oud(h)

thou (h) gh huh

huh .h=

Jis okay wull- (0.2) anyway (.) thought Id

Because Skeets activity report had been solicited rather than volunteered, it exhibits downgraded newsworthiness (Button & Casey, 1984). Dee Anns topicalizing response Are you
really? upgrades the newsworthiness of that report and makes Skeets current activity
available for further topical talk. This topicalizer selects Skeet as the next speaker, but does
not specifically request an elaboration of his report although an occasion for elaboration is
provided; thus, Skeet is positioned to volunteer an elaboration. His minimal affirmation in
Line 7, however, momentarily curtails topic development, but Dee Ann pursues elaboration in Line 8 through an itemized news inquiry (Why? rya sick?). This inquiry utilizes
a correction invitation format (Sacks, 1992a) that selects a candidate account from a class
of accounts (glossed as debilitating personal states) and invites confirmation or a correction that selects an alternative account from the same class. Skeet in Line 10 opts for the

168 Studies in language and social interaction

latter by rejecting Dee Anns candidate account and attributing his early preparation for
sleep to fatigue. Skeet uses the minimizer jus to formulate his fatigue as having minimal
After the hearing check and its confirmation (Lines 1112), Dee Ann had an opportunity to self-select and pursue further topical development (Line 13); however, she did not.
Instead Skeet elected to continue speaking and volunteered an unsolicited account for his
fatigue: Went to bed too late las night (Line 14). He attributes his fatigue to his own
prior failure to get to bed on time. Important to note, the topical focus of Skeets current
unhappy state is linked up with incipient topical possibilities, namely actions and events,
within the temporal frame of las night, that preceded and possibly contributed to his
Dee Anns understandings of Skeets account are progressively displayed in two successive turn units. The initial turn unit (Yep) I donno why: (Line 15) may be viewed
as a teasing action. Evidence for this analysis may be found in its composition and sequential placement. First, Dee Ann professes ignorance of the reasons for Skeets tardiness
in getting to bed. Since Dee Ann and Skeet may share access to actions and events that
preceded and possibly contributed to his lateness in getting to bed, disavowing that shared
knowledge is in direct contrast to something they both know (Drew, 1987, p. 232). This
contrast, coupled with the stress on and the stretching of why:, contribute to the recognizability of the units ironic import. Second, Drew (1987) reports that teases occur in the
next turn position and treat prior turns as overdone in some fashion. In view of their possibly shared knowledge, Dee Anns irony may treat Skeets account in Line 14 as stating
the obvious rather than as news. Third, teases attribute deviant actions and/or categories
based on some minimally required identity (Drew, 1987). That is, Dee Anns use of why:
exploits Skeets failure to get to bed (minimally required identity) by alluding to (deviant)
actions that suggest a lack of personal discipline. In contrast, professing ignorance of these
actions may be a way for Dee Ann to take up a playful stance of innocence.
Following a beat of silence in Line 16, Dee Ann produced two bursts of laughter. In the
environment of Dee Anns ironic laugh source, the subsequent laughter proffers a laugh
invitation (Jefferson, 1979). Skeet has an opportunity to exhibit appreciation of Dee Anns
tease and playful stance of innocence by laughing together with her, thus co-implicating
himself with that stance (Jefferson, Sacks, & Schegloff, 1987). Skeet does not take up
the invitation to laugh. Instead, he displays recognition of the tease (without ratifying its
humor) through his own faint and world-weary profession of ignorance in Line 18.
In Line 17, Dee Ann does not pursue laughter. Instead, following her terminal inbreath,
Dee Ann explicitly denies culpability in regard to Skeets lateness in getting to bed. In
contrast to her prior teasing action, this denial proposes a serious version of las nights
events. Her denial is done in a kind of exaggerated countrified, regional dialect (possibly central Texas) that is compatible with her posture of innocence. Combined with her
pre-speech laughter, Dee Ann uses this speech register to distance herself form the delicate
action of treating Skeets account in Line 14 as shifting blame to her.
How does Dee Ann come to deny responsibility for Skeets failure to get to bed on
time? First, she does not solicit the account in Line 14. Instead, Skeet volunteers it. Button and Casey (1985) observed how tellers refrain from volunteering delicate tellings and
wait for recipients to solicit them. Because Skeet had volunteered the account, Dee Ann

Rejecting illegitimate understandings 169

may have understood him as making a special point of reminding her of an incident with
now unhappy consequences for him (Pomerantz, 1978). Second, Dee Ann may have
understood Skeets account as part of an unfinished telling, with more details to come. The
unsolicited production of this account (Line 14), coupled with its scanty details, may have
contributed to that understanding. If Dee Ann had participated in activities with Skeet that
preceded his failure to get to bed on time, then her disavowal of blame in Line 17 may have
anticipated and preempted forthcoming reminders of her participation that shift at least
some of the responsibility for his failure to her. Dee Ann may have anticipated descriptions
from Skeet that would have turned his failure into a consequence of her antecedent actions
(Pomerantz, 1978).1
In the third turn position (Line 21), Skeet deploys I- (.) didnt say that to reject Dee
Anns denial of culpability. The delayed onset of this rejection, coupled with the glottal cut-off of I- and the beat of silence prior to didnt, display what, for Skeet, is the
strongly unexpected character of Dee Anns denial.2 Features of Skeets rejection exhibit its
placement in the third sequential position in relation to his account (Went to bed too late
las night) in the first position and Dee Anns analysis of it (Not my fault) in the next
turn position. The pro-terms I- and say topicalize his authorship of the account, and
that ties back to, without formulating, Dee Anns denial and its concomitant attribution
of blame shifting. Like third-position repairs, this rejection treats the relationship between
the contributions in the first and next turn positions as problematic. However, the negation
of say denies that Dee Anns finding of blame shifting could have been produced from
any legitimate analysis of Went to bed too late las night. Skeet rejects her analysis, not
as misunderstanding the account, but as misconstruing it.
This method of rejecting understanding displays in the next turn position differs from
comparable practices of third-position repair. The latter treat misunderstanding displays
as viable, albeit incorrect, alternative understandings of talk in the first position. In this
instance, I- (.) didnt say that rejects Dee Anns denial of culpability as the product of an
illegitimate analysis of the account in Line 14. Additionally, third-position repairs provide
speakers of understanding displays with the resources to redo their understanding of the
first-positioned talk.3 Rather than providing these resources for Dee Ann, Skeet stands by
the import of his account as an innocuous and self-evident description of his agency in failing to get to bed on time. Skeets rejection accomplishes this action by reporting a negative
event; that is, he denies having authored talk that could be construed as shifting responsibility to Dee Ann. This negative formulation makes an implicit contrast with what he did
This line of analysis depends on the assumption that Dee Ann had been a party to the previous
nights events. Though no independent evidence is available, it is difficult to surmise otherwise how
she could have come to see herself as a candidate for blame allocation, without imputing some type
of exotic motivation to her denial of culpability.
Notice that Skeets rejection is done in reference to his talk in Line 14; the rejection does not
propose a version of the previous nights events that would treat Dee Ann as an outsider to those
events. That is, if Dee Ann could not be viewed as a party to those events, Skeet would be expected
to deliver a very different sort of rejection (e.g., Huh? You werent even there).
Referring back to example 1, after M had specified how Jeff in Line 1 was properly understood
(Line 6), D used her revised understanding of Ms question to redo her answer (Lines 78) in a
direction quite different from Lines 2 and 4.

170 Studies in language and social interaction

say. Skeet imputes a benign and self-evident intelligibility to that talk; furthermore, the
account is treated as a completed telling, as opposed to an unfinished one. Skeet invokes an
entitlement to having the account treated as having the plainfully intelligible character that
he attributes to it (Garfinkel, 1967), and thus Dee Anns misconstrual of that talk is treated
as something of a breach of that entitlement.
In line 22, Dee Ann acknowledges Skeets authorial authority over his talk. This asymmetry does not mean, however, that she is without resources to counter his rejection
(cf. Drew, 1991).


I- (.) didnt say that













You th(h) ought it awful l(h)oud(h)

thou(h)gh huh

huh .h=

Jis okay wull- (0.2) anyway (.) thought Id check

In Line 23, Skeet produces a burst of laughter (also bearable as a cough upon its occurrence) that Dee Ann joins with a pair of laughs. Out of this environment, she retrieves
the laugh source (Line 21) from which the counter You th(h)ought it awful l(h)oud(h)
thou(h)gh is produced. The pro-term it preserves the referent of that (Line 21), and
the counter half jokingly concedes that Skeets account could not have been understood as
saying she was to blame; this concession is delivered in a qualified fashion (note the use
of though in the tag position). Nonetheless Dee Anns counter preserves her finding of
blame shifting by imputing it as a thought to Skeet.
The action of attributing thoughts to an interlocutor speaks to Sacks (1992a) remarks
concerning the observability of thoughts:
And this phenomenon of seeing other peoples thoughts is really an important thing.
Exactly how its properly posed is quite tricky. First of all, its of course nonsense to
say that thoughts are things that cant be seen, unless you want to take some notion
of thoughts that Members do not employ, since they certainly do take it that one
can see what anybody is thinking. Not in every case, certainly, but you can see what
people are thinking, and there are ways of doing it. And you must learn to do it.
(p. 364)
In this particular case, Skeet has rejected Dee Anns denial of culpability and its analysis
of the account in Line 14; he denies the very possibility of construing his talk as shifting

Rejecting illegitimate understandings 171

blame to her. Dee Ann faces the problem of providing for the methodicity of her denial and
its display of understanding in Line 17. Having just conceded to Skeets authorial authority,
she is effectively prevented from using the composition of Skeets description of his own
agency as a resource in solving this problem. Furthermore, certain methodical features of
her understanding, the unsolicited production of Skeets telling and its possibly unfinished
character, may be potentially troublesome to formulate explicitly. Dee Ann provides for
the methodicity of her rejected understanding by glossing Skeets observable activity as
a thought and formulating that activity gloss as the source of her action/understanding
display. Such a practice does not involve mind reading in the sense of claiming access to
the private recesses of anothers mind. The description awful l(h)oud(h) characterizes
that thought as having a publicly conspicuous character.
The delicate nature of Dee Anns counter lies not so much in the attribution of thoughts
to Skeet but in the reattribution of the action of blaming to him. Skeets rejection is treated
as a laugh source that Dee Ann retrieves to produce a continuation of joking activity as she
distances herself from the accusatory import of her counter (Line 24). This laughter was
initiated in Line 23 by Skeet, but was sustained primarily by Dee Ann during its course
with minimal participation from Skeet (Line 26). The laughter combined with the joke-toserious No (Schegloff, 1996) in Line 27 to frame the interaction that ensued from Skeets
rejection in Line 21 as half kidding/serious. Dee Ann exits from this topical sequence
(Line 29) by returning to the previous topic and official reason for the call.
To summarize: This essay reports on a practice of rejecting illegitimate understanding
displays. Utterances such as I didnt say that refer to those displays (through the proterm that) but reject the reproducibility of such displays from a methodical analysis of
the talk in the first position. This practice may be regarded as a cousin of third-position
repair. Though both action types treat the relationship between the talk in the first position
and its display of understanding in the next turn position as problematic, the former does
not formulate a repair or solution to the problem of understanding. In the present data,
the speaker of I didnt say that reports a negative event that contrasts implicitly with
what had been said in the first position. This speaker stands by the first-positioned talk as
exhibiting a self-subsistent intelligibility. Consequently the recipient of I didnt say that
faced the problem of providing for the methodicity of her action/display of understanding
in the next turn position. The observed solution in these data involved the speaker of the
understanding display acknowledging her interlocutors authorial authority then imputing
her understanding to a thought of the interlocutor. This formulation served to gloss the
publicly noticeable activity of the interlocutor as the source of her understanding.
These data serve to suggest some possible limits to speakers tolerance for alternative
understandings of their talk. As Schegloff (1992) pointed out, speakers readily recognize
that and how their talk may be understood in ways divergent from its designed import.
However, when their talk is treated as portending some interpersonally problematic action
(such as blaming), subsequent understanding displays may be rejected as exceeding that
tolerance. One way to reject the legitimacy of an understanding display is to deny the
usability of the talk in the first position as the source of an analysis that would produce that
understanding. This finding provides a naturally occurring complement to one of Garfinkels (1967) breaching demonstrations. Next speakers were instructed to withhold displays
of understanding of the prior speakers commonplace remarks (e.g., I had a flat tire) and

172 Studies in language and social interaction

to raise problems of understanding by initiating repair (e.g., What do you mean, you had
a flat tire?) in the absence of recognizable understanding problems. The prior speakers
subsequent outrage was clearly more moral than technical. He concluded that speakers
do not merely expect to be understood but insist on an entitlement to the manifestly intelligible character of their talk. In the present data, a next speaker commits to a display of
understanding, but the prior speaker uses the third sequential position to reject the prior
action/understanding display as transgressing the self-evident intelligibility of the talk in
the first position. Unlike the explosive outrage of Garfinkels victims, the parties to the
present data drew upon the organization of laughter and used special speech registers as
ways of framing delicate actions as half joking/serious. Overall, these findings contribute
to our understanding of connections between the interactional architecture of intersubjectivity and the moral order.
Rejecting the methodicity and legitimacy of an understanding display poses certain
interactional aftershocks in which the parties orient to a possible impropriety embodied
in imputing the action of blaming to a prior speaker whose talk is excluded as a possible
source of such an understanding. So how does the speaker of the rejected understanding
display manage to re-legitimate that display? One way is to formulate conduct, other than
the talk in the first position, that would serve as an alternative source of the speakers
methodically produced understanding. Here that speaker preserves her understanding as
the product of a gloss of her interlocutors publicly conspicuous activity: You th(h)ought
it awful l(h)oud(h) thou(h)gh. The interactional uses of these glossing practices provide
both a parallel and challenge to communication models that impute messages to the private
encoding of speakers thoughts and meanings. Whereas these models treat thoughts
as residing in the private, unobservable mental storehouse of speakers, such notions of
radical subjectivity are not in use among the parties to this interaction. These observations
add credence to Sacks (1992a) remarks concerning the public observability of thoughts
and underscore the dangers of premature theorizing that glosses rather than explicates the
details of interactional practices.4
Robert Hopper was my dissertation adviser at the University of Texas in the late 1980s.
I and many others, in large measure, trace the beginnings of our intellectual commitment
to the close examination of talk to his graduate seminars in conversation analysis. Over
the years, he has continued to embody what it means to be a colleague by appreciating
our strengths and challenging us to improve our craft. He has unselfishly given of himself
during the best and worst of times. It is a distinct honor to contribute to this esteemed

Rejecting theoretical notions of radical subjectivity does not deny that people, at times, do act as
practical Solipsists. The key is finding data in which the parties to an interaction orient to such
practices instead of insisting upon their omnirelevance as many communication models do.

Rejecting illegitimate understandings 173

Button, G., & Casey, N. (1984). Generating topic: The use of topic initial elicitors. In J.M.Atkinson
& J.Heritage (Eds.), Structures of social action: Studies in conversation analysis (pp. 167190).
Cambridge, England: Cambridge University Press.
Button, G., & Casey, N. (1985). Topic nomination and topic pursuit. Human Studies, 8, 355.
Drew, P. (1991). Asymmetries of knowledge in conversational interactions. In I.Markova &
K.Foppa (Eds.), Asymmetries in dialogue (pp. 2148). Hertfordshire, England: Harvester
Garfinkel, H. (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Heritage, J. (1984). Garfinkel and ethnomethodology. Cambridge, England: Polity Press.
Jefferson, G. (1979). A technique for inviting laughter and its subsequent acceptance declination. In
G.Psathas (Ed.), Everyday language: Studies in ethnomethodology (7996). New York: Irvington.
Jefferson, G., Sacks, H., & Schegloff, E.A. (1987). Notes on laughter in the pursuit of intimacy. In
G. Button & J.R.E.Lee (Eds.), Talk and social organisation (pp. 152205). Clevedon, England:
Multilingual Matters.
Pomerantz, A. (1978). Attributions of responsibility: Blamings. Sociology, 12, 115121.
Sacks, H. (1992a). Lectures on conversation (Vol. 1, G.Jefferson, Ed.). Oxford, England: Basil
Sacks, H. (1992b). Lectures on conversation (Vol. 2, G.Jefferson, Ed). Oxford, England: Basil
Sacks, H., Schegloff, E.A., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking in conversation. Language, 50, 696735.
Schegloff, E.A. (1987a). Analyzing single episodes of interaction: An exercise in conversation
analysis. Social Psychology Quarterly, 50, 101114.
Schegloff, E.A. (1987b). Some sources of misunderstanding in talk-ininteraction. Linguistics, 25,
Schegloff, E.A. (1991). Conversation analysis and socially shared cognition. In L.Resnick,
J.Levine, & S. Behrend (Eds.), Perspectives on socially shared cognition (pp. 150171). Washington, DC: American Psychological Association.
Schegloff, E.A. (1992). Repair after next turn: The last structurally provided defense of intersubjectivity in conversation. American Journal of Sociology, 97, 12951345.
Schegloff, E.A. (1996, November). Joke-serious no. Paper presented at the Speech Communication Association Convention in San Diego.

Interactive Methods for Constructing Relationships
Jenny Mandelbaum
Rutgers University
Increasingly in the communication field, scholars are coming to recognize that the character of a relationship is built moment by moment, by interactants, in and through interaction
(Goldsmith & Baxter, 1996). Though compelling, this claim has proven difficult to document. Just how is the relationship between interlocutors constructed, and thus available,
from the particular ways in which talk is produced?
In this chapter I describe two methods whereby the interactional construction of relationships can be documented. First, in a kind of conversational tit-for-tat, one interlocutor produces a turn that could be heard to have problematic, or disconnecting
implications for the relationship. In the next turn, the other produces a similar turn that has
the result of shifting the disconnecting implications to connecting ones. In the second
method, conversational repair targets a turn that has possible problematic implications for
the relationship. The speaker of the repairables method for repairing the problem does not
take up the relationship implications, though. These two methods for taking up turns with
possible problematic implications for the relationship display the interactive process of
relationship construction.
In the vernacular, and often in scholarly work also, we take relationships to be things that
we have. That is, in the way that we talk about them, relationships are often reified, static
entities. Relationship states are often treated as independent variables, with discursive
consequences (Hopper & Chen, 1996, p. 310). This approach to relationships treats them
as social structural entities that exist outside of discourse, taking spouse or supervisor, for instance, to be social categories, from which ways of talking follow. From this
perspective, which dominates much research in communication, ways of talking could
provide an index for intimacy, and ways of talking that are characteristic of marriage,
for instance, could be discerned. In practice, an approach that sees relationships as existing external to discourse presents problems, because even within relationships that have
objective, social categorical definitions, relational states shift. Even those who might
describe themselves, and be described by others, as happily married have arguments or
difficult interactions and problematic moments.
In contrast to this view, social constructionists and others make a strong case for seeing relationships as constructed in and through interaction. Goldsmith and Baxter (1996)
emphasized the importance of this constitutive view of communication in relationships.
They drew on subjects diaries and recollections to identify a set of 29 speech events, which
they then divided into six groups that constitute everyday relationships. They pointed out
that it might prove difficult to observe all the joint enactments of talk through which an

176 Studies in language and social interaction

individuals relationships are constructed (p. 90). Therefore, they used diary studies so
that individuals could report on the events in which they engage in various relationships
(p. 90.). Conversation analysts have shown that detailed analysis of tape-recorded naturally
occurring conversations provides a method for describing particular ways interacts may
do relationships (e.g., Goodwin, 1990; Heritage & Sefi, 1992; Mandelbaum, 1987, 1989;
Pomerantz & Fehr, 1997).
Watzlawick, Beavin, and Jackson (1967) proposed that all messages have both content and relationship levels. All talk then may be taken to contain proposals regarding
the relationship between interactants. For the most part, though, these relational proposals
do not become the main business of talk, and may not be taken up at all in any discernible
or overt way. Their study often is speculative, because claims about the relational activities
that interactants may be undertaking can be hard to demonstrate.
Goffman (1971) suggested that interaction contains numerous signs whereby interactants make available to one another the current character of the relationship (p. 184).
He called these tie-signs evidence about relationships, that is, about ties between persons, whether involving objects, acts, expressions, and only excluding the literal aspects
of explicit documentary statements (p. 184.). Tie-signs may include holding hands, locking arms, using the same bottle of suntan lotion when coming to the side of the pool, and
so on. For the most part, the production and noticing of these tie-signs are not focused
involvements (Goffman, 1963) for interactants. That is, they are generally incidental to
other ongoing activities. Goffman wrote of them as a sort of social obligation, a performance that we owe others who are in the co-presence of a related couple (a pair in a
relationship). Through the performance of tie-signs, both relational partners and others are
provided with evidence of the character of a relationship being enacted.
Pomerantz and Fehr (1997) recommended as a final step in analysis that the researcher
examine the identity and relational implications of the way a particular action is packaged.
For conversation analysts, it is critical that relationship be procedurally relevant to participants (Schegloff, 1987). Like identity, although theorized to be omnirelevant, it can be
hard to document the relevance of relationship to the way talk is done. For this reason, conversation analysts often have been reluctant to address issues of relationship, using instead
such terms as alignment, and affiliation. Despite this constraint, conversation analytic
findings reveal important features of how talk may propose and/or construct relationships.
For instance, Heritage and Sefi (1992) showed how health visitors methods for questioning new mothers can propose particular alignments between participants. Goodwin
(1990) showed how the way that a directive is offered proposes a version of the relationship
between the interactants. That is, when you ask someone to do something, it formulates
who they are with respect to yousomeone over whom you can assume unquestionable control, for instance. When I say to someone Come here right now, I propose a
relationship between us in which I have some legitimate jurisdiction over that persons
actions. Some actions, then, lend themselves to fairly easy interpretation with respect to
the relationship they propose between interlocutors. The firmness of this phenomenon
is perhaps indicated by the fact that using a polite format to ask someone with whom we
have a close relationship to do something for us may be a way of a proposing (current)
distance between us. Some ways of talking to or acting with regard to others, then, have
somewhat stable relational interpretations. With respect to how we ask someone else to do

Interactive methods for constructing relationships 177

something, the extent to which we provide them with choice, or self determination over
their own actions, is a fairly tangible index of how we see ourselves relative to them. It
may indicate the kind of interpersonal power we take ourselves to be able to enact with
respect to them.
Some conversation analytic work has looked at inexplicit relational proposals that can
sometimes be disentangled in such features of conversation. For instance, the use of reporting to do such actions as blaming (Pomerantz, 1978), and inviting (Drew, 1984); the placement and nature of recipient turns in storytellings (Mandelbaum, 1989), and complaints
(Mandelbaum, 1991/1992) may enable participants to blame, invite, or complain in a collaborative rather than a unilateral fashion. However, conversation analytic work for the
most part has not turned its attention to how relationships are constructed, specifically
because this is frequently difficult to identify as the work interactants are actively undertaking. Two exceptions are the work of Jefferson, Sacks, and Schegloff (1987), and Morrison
Jefferson et al. (1987) showed that the use and uptake of obscenity may provide a way
for interactants to collaborate on constructing intimacy, and in this way make sub rosa
proposals of intimacy. Morrison (1997) demonstrated how interlocutors may use tracking
questions and answers to these questions to enact involvement. She showed how by asking a question that in effect seeks an update, relationship members talk in such a way as
to display their involvement in the life of the other.
Studying relationships involves numerous complexities for the researcher. Among them
are issues of unpredictability, privacy, and access. Scholars interested in how relationships develop note that transitions in the character of a relationship may occur at critical
moments (e.g., Baxter & Bullis, 1986). It is hard to know when critical moments of relationships will take place, and harder to have a tape recorder or video recorder present at
those critical moments in ways that will not change the character of the occurrence. Yet if
we look at interaction closely, we see that, in line with the proposals of social constructionists, relationships are constructed and negotiated moment by moment in a delicate to and
fro, some of which can be documented through close attention to the details of talk.
Both Gofftnans tie-signs and Watzlawick, Beavin, and Jacksons relationship level
of conversation may be present throughout conversation, but may not constitute a focused
activity for interactants. In this chapter, I examine places where the often overlooked relational implications of talk are taken up in some way. I discuss two methods for doing this,
tit-for-tat and repair, and contrast the apparent relational consequences of each. In both
cases, I show how both ends of the relationship (Goffman, 1971, p. 188) work together
to position themselves vis a vis one another.
During the wedding of Prince Charles and Princess Diana in 1981, while she was uttering
her vows, Princess Diana produced Prince Charles name (Charles Philip Arthur George)
incorrectly, confusing the order of the names.1 The significance of this repairable could be
interpreted in many different ways. As a unilateral, presumably unintentional action, it has

I am grateful to Paul Drew for bringing this example to my attention.

178 Studies in language and social interaction

many possible (possibly negative) implications, both regarding Princess Dianas identity
(the kind of person that she is), and regarding their relationship. It could be taken to have
implications regarding her competence or her state of mind, for instance. Psychologists
might take it to have symbolic significance regarding her feelings for Prince Charles, or
about the wedding. In his vows, though, Prince Charles produced Dianas name in a similarly incorrect way. Until Princess Dianas death it was said that this was the last nice thing
he did for her. His tit-for-tat here made available the implication, getting names wrong
during a wedding is something anyone could do. A reciprocal action of the same kind
appears to be one way to take up a problematic activity. By doing the same thing (mixing
up names, in this case), it targets the activity to which it is reciprocal. It may show that the
initial action was noticeable. Interestingly, though, by doing the same action, a possibly
problematic or disjoining action on its own is rendered benign or conjoining, because
the implications that anyone could do it or it can happen to me become available. It
becomes a common occurrence, instead of a gaping breach of etiquette, for instance.
In the following segment, a telephone conversation is begun with an apparently playful
exchange of name-calling. This tit-for-tat seems to work in a similar way to the previous
instance. Though it is clearly not its official business, the first name-calling could be
heard to set the couple apart. That is, although in the context the hearing is unlikely, it could
be heard in this way. In response, the reciprocal name-calling proposes a kind of relatedness between the callers, undoing the possible disjuncture. Kip and Cara have been put on
the phone by their roommates, who were talking together until Caras roommate reported
to Kips that Cara wanted to talk to Kip. On the tape, we hear Cara waiting for Kip. His
^ee^YEE::^ES?hh huh hih heh starts their conversation.

Romance 8
^ee^Y [EE::^E]S?hh huh hih heh=
ehh. heeYe (h) e (h) es?
hh Yih que:er w(h)at[re ya doin.]

uh ^I dunno whatre you doin you queer bait.


eh h[eh heh heh

eh hh[h eh
t(s) gon on.

After initial apparent difficulty recognizing one another (perhaps due to Kips overdone
Yes in Line 1), in Line 5 Cara calls Kip a name, Yih que:er, which could be heard as a
teasing response to his redoing, in Line 4, of his over-exaggerated Yes at the beginning
of their interaction. Though in its vernacular sense of homosexual queer has no apparent fit with Kips behavior, it could be heard as a playful version of silly or oddan
original meaning of the term queer. This is immediately followed by an inquiry regarding

Interactive methods for constructing relationships 179

what he is doing, presumably currently or immediately before he took the phone call. This
can be heard as a conventional beginning to their conversation. His response is postponed
by a post laugh inbreath. In Line 8, he gives a minimal answer to the question regarding
what he is doing, I dunno. He then asks the reciprocal question, whatre you doin, and
produces a reciprocal name-calling, you queer bait. This name-calling is reciprocal in a
special way. She has called him a queer that, if it were to be taken seriously or literally
in the current vernacular, would make her not of interest to him. Queer bait in response
to queer could be heard to be formulating her as bait for the queerthat is, bait for
Kip. It thus proposes a possible relationship between them in which she is specifically
attractive to him.
Thus a formulation of him (you queer) that taken literally (in the sense in which it is
presumably not intended) makes her of no interest to him, is recast in retrospect as making her specifically of interest to him. This is done playfully, but nonetheless might raise
a glimmer of the possibility that there could be a relationship between them that involves
a connection constituted by appropriate fit and special interest. In its aftermath, nothing is
overtly made of the reciprocal name-calling and the possible connectedness it implies. As
Kip laughs, Cara answers the inquiry that preceded the name-calling. As Kips laughter
continues, Cara makes a reciprocal busyness inquiry: (s) gon on. and talk proceeds.
In this instance, through a kind of conversational tit-for-tat interactants make available a connection between them. Immediately after talk that could be heard to indicate a
reciprocated disjunctive between them (the difficulty recognizing one another), an action
by one partner that could be heard to have possible implications for their relationship, but
could equally, and more plausibly, be heard to be directly related to prior talk (Kips playfully overdone greeting) is responded to in such a way as to constitute a reciprocation by
the other. The reciprocation takes up possible relationship implications in the first speakers
turn and provides for a proposal of connectedness between them where her turn could have
been heard to position them as disconnected. Talk simply moves on, and nothing is made of
it overtly. Like Princess Dianas flub, Caras name-calling makes available certain implications regarding participants relative positioning although these relative implications are
clearly not official business at all. In both cases, their relational partners next turn has
a similar format, yet counteracts those implications in an off-the-record fashion that
nonetheless makes the relational implications of the first turn apparent. Here then we see
a sort of advance on the tie-sign. An action that could be heard as a tie-sign with possible
disaligning relational implications, but that could also simply be ignored, is targeted, made
visible, and redressed simultaneously by a response-in-kind. Nonetheless, like Goffrnans
tie-signs, this remains an embedded action.
In the following fragment, the embedded relational implications of a turn are taken up
in a more overt way, using repair. Nonetheless, the first speakers response to repair initiation downplays the relational implications. This demonstrates interactants alertness to
problematic relational implications, and indicates the collaborative character of positioning
activities in conversation.

180 Studies in language and social interaction

Two couples, Vicki and Shawn, and Nina and Matthew, are eating dinner together. This
segment occurs after about 14 minutes of recorded conversation. Vicki reports an activity
she plans to undertake (Lines 24, 26, and 28). Shawn initiates repair in a somewhat overdone, teasing way (Lines 33, 35, and 37). Vicki completes the repair in an underdone
way (Lines 3940). The underdone character of Vickis repair is noticeable in contrast to
the overblown character of Shawns repair initiation.













[Cars ih stra:nded bout thirdy sumpn

peoplev die:d,
Becuz a that,
=Css the weather,
Wir gunnuh call [up]
[Ts in]sa[: n e .]
[Wir gn]nuh
[call up sm frjiends] =
[(sp thA:: d).]
=hih.hh[Wz e_igh] d [y degrees here
the oth]uh] =
[en say] [eighty degrees]ihh]
=[day. ih hih] [he
=[hnhh heh-hu]h-h[uh
Oh they hate tih hear that.
I kno:w. En [then hang up] eh heh u
[Well this gu]y
=^Who[was \tha[t ()
[who- [
[mn nah ah [hah
[One guy thet I
=he usually comes ^ou: t. yihknow[so
you js=
=tellm its eighdy degree:s hill get

Interactive methods for constructing relationships 181














=[n a h-ha-ha]
[heh heh heh]
=[w a i hey]woah w[oah
[ih hih heh he[h
[Wu waia
waia wu.
One: gu::y you usually ca(h)a(h)ll?
m-h [m
[No we [^c a\ : 1 1.]
[Wd is this] : : .
Oh:.Okay it wz: friend ami:net[oo.
[Oh: Shames
friend, [yeah.
[Oh thats good (thet).
Thats my[friend.
[The guy (oo) comes outn treats

In Line 24, Vicki begins a report about an unnamed guy that she wants to call. In using a
nonrecognitional reference (one that indicates she does not expect that her recipients could
recognize the person to whom she is referring [Sacks & Schegloff, 1979] the implication
is available that she does not expect any of those present to be able to recognize to whom
it is that she is referring. In Line 27, she tells what the guy referred to in Line 24 usually
does. She then reports what you need to say to produce the result of this guy coming
outtell him about the warm weather. From this recipients can draw the implication that
if she does what at the beginning she states she wants to do (call him), the result will be
that the unnamed (and unknown-to-others-present) guy will come out. In previous turns, in
Lines 1016 Shawn and Vicki together enact what they are going to do (Were gonna call
up some friends)calling people to tell them that it is 80. They synchronously report
an action that they both claim and show themselves to be going to undertake together. In
formulating the person who wants to call as I, and in contrast with their joint enactment
of calling someone to tell them that it is 80, her report of something she wants to do (one
guy that I wanna cawrll, Lines 2425), and her reference to you js tell m (Line 27
and 29) could be heard to project an action she will do by herself. Given the way in which
she refers to the person she will call, and her formulation of herself as the sole caller, it is
potentially hearable that she wants to call someone unknown to members of the present

182 Studies in language and social interaction

Immediately upon the completion of Vickis report of her future plan, Shawn stops
conversation in a very elaborate and overdone way. His wai hey woah woah Wu waia
waia could be heard to indicate some kind of trouble, but it is not available from this turn
what the trouble could be. He then produces a turn as though it were a repeat of Vickis
turn: One: gu::y you usually ca(h)a(h)ll? (Line 40). He combines elements from the
beginning of her turn in Lines 2425 (one guy that I wanna caw:ll) and the second part
of it in Line 27 (he usually comes ^ou:t) to produce a most incriminating version of
what she said: One guy you usually ca(h)a(h)ll? He slightly misrepeats her talk in such a
way as to make available as an understanding the strongest indication that there is a guy
in her life about whom he does not know, whom she calls habitually. His Wdzs (What
is this?) corroborates the impression that he is calling into question what is going on.
All of this is produced in a somewhat overdone, overblown fashion, which Drew (1987)
suggested may be characteristic of teases. It is possible to hear this turn as taking Vicki to
task in a teasing way for having produced the appearance that she is inviting out to see her
some guy that he does not know. In Goffmans terms, he displays himself to be hearing
her turn as offering a particular kind of tie-sign. Like the first turns in the tit-for-tat segments examined earlier, although it is clearly not its principal enterprise, Vickis turn could
be heard to be proposing that she has some involvement that suggests disassociation with
Shawn because of association with a guy that Shawn does not know.
In Line 42, Vicki offers a disagreement token, No, and then offers another version
of part of what he, through his reenactment, has claimed her to have said, we ca:ll can
be heard as a candidate replacement for you usually call. The repair operation involves
dropping the usually and replacing you with we. In this way, the problematic character of the activityhabitually calling an unknown guy without him knowingis removed,
because the calling is an activity that they do together.
What is anomalous about this repair is that she does not stress the repaired item. Normally in response to other-intiated repair, the item that performs the repair operation is
stressed, so as to be hearable as the repair (Schegloff, Jefferson, and Sacks, 1977). She
stresses call, which does not appear to have been targeted as the repairable. At the same
time, it is clearly the word we that has replaced the I from her turn and the you
(meaning Vicki) from his turn. In stressing call, a word that has not been repaired, it is
as though she were indicating that the activity of calling were the repairable. In this way,
she literally de-emphasizes the word that caused the troublethe one that pointed to who
was doing the calling. It was her use of I that made available the appearance or possible
hearing that she might want to, or was engaging in, some activity independent of Shawn.
Stressing the nonrepaired part could be hearable as backgrounding or playing down
the relational implications of the repair. In this way, Vicki avoids overtly taking up the
relational proposal his repair tries to make.
Shawns repair appears to be done as a teasing display of concern, yet Vicki gives a pofaced response to the tease. She treats it as though it were serious (Drew, 1987). After what
appears to be a postoverlap resolution hitch, in Line 45 Shawns change of state token,
Oh (Heritage, 1984), shows that he now has a new understanding of what Vicki meant.
His okay shows that this shift makes what she had been proposing acceptable. He then
reports a characteristic of the call-recipient that he now understands: it wz: friend amine
too. In explicitly stating that this is what makes it okay, Shawn makes available that it was

Interactive methods for constructing relationships 183

indeed the problem posed for their relationship that constituted the problem his repair initiation addresses. In calling the groups attention to it by doing a very public repair, Shawn
calls this implication into question in an overdone, teasing fashion. In so doing, he shows
that the appearance that Vickis talk could be heard to present regarding their positioning
relative to one anotherthat there is a guy whom she will call, and who will then come
out (presumably to California)is what was problematic for him. Because it is a friend of
his also, he can now rehear this as unproblematic, and make that rehearing public.
In this way, Shawn makes a public display of having the right to call into question with
whom Vicki associates without his knowledge. In her producing her repair with the stress
that she does, Vicki emphasizes the activity of calling, and not the we on which the relational implications center. In this way, she seems to focus on issues of understanding, rather
than relational concerns. There is no playing along with the tease, display of shame or
embarrassment, of having been caught red-handed, and so on. Rather, the way in which
she offers the repair has more the air of annoyance.
Drew (1987) suggested that teases are often used to produce mild social sanctions, and
that po-faced responses provide a way for the teased party to set the record straight. Here
Shawns repair initiation seems to target the problematic tie-sign, the appearance of illicit
activity that Vickis turn makes. Though Vicki could play along with the tease, she sets
the record straight in a way that seems to dismiss the tease. In playing down the relational
implications, Vicki avoids officially entering into the positioning activity that Shawns
turn takes up. Rather, her talk does relationship work by not officially taking up the implications Shawns repair indicates. For in treating it as a matter of course that it is his friend,
and showing mild annoyance at Shawns action, she displays that the concern his repair
indicates is not an issue. Here then we see an instance where the possible relational implications of a turn are taken up and made available by one participant, whereas the other
participant downplays the relational implications. Though Shawns turn makes possible
overt uptake of relational matters, Vickis shows that they are not relevant here.
These episodes demonstrate that relational implications may be taken up when they contain
problematic proposals regarding the relative positioning of interactants. The management
of these proposals is a collaborative process. In both conversational tit-for-tats, and in the
repair episode, a second turn targets possible problematic relational implications in a prior
turn. Thus we see interactants on-sight alertness to the relationship level of a conversation, and to the tie-signs that talk may contain. However, talk in third position indicates that
even where relationship implications have been targeted by one speaker in the talk of the
other, the speaker whose talk contained those implications need not take them up further.
This account suggests the subtle yet collaborative manner in which relationships are
enacted in interaction. It seems that moments where there are mild problems for relationships (or the appearance of a relationship) can prove to be fruitful sites for documenting the
interactive work of relationship construction. In this way, we can begin to see relationships
as collections of communication practices, or things that we do through communication, in
contrast to thinking of them as social structural things that we have.

184 Studies in language and social interaction

Baxter, L., & Bullis, C. (1986). Turning points in developing romantic relationships. Human Communication Research 12, 469493.
Drew, P. (1984). Speakers reportings in invitation sequences. In J.M. Atkinson & J.C.Heritage
(Eds.), Structures of social action: Studies in conversation analysis (pp. 129151). Cambridge
England: Cambridge University Press.
Drew, P. (1987). Po-faced receipts of teases. Linguistics, 25, 219253.
Goffman, E. (1963). Behavior in public places. New York: The Free Press.
Goffrnan, E. (1971). Relations in public. New York: Harper & Row.
Goldsmith, D., & Baxter, L. (1996). Constituting relationships in talk: A taxonomy of speech events
in social and personal relationships. Human Communication Research, 23, 87115.
Goodwin, M. (1990). He-said-she-said: Talk as social organization among Black children. Bloomington: Indiana University Press.
Heritage, J. (1984). A change of state token and aspects of its sequential placement. In J.M. Atkinson & J.C. Heritage (Eds.), Structures of social action: Studies in conversation analysis
(pp. 299345). Cambridge, England: Cambridge University Press.
Heritage, J., & Sefi, S. (1992). Dilemmas of advice: Aspects of the delivery and reception of advice
in interactions between health visitors and first-time mothers. In P.Drew, & J.Heritage, (Eds.)
Talk at work: Interaction in institutional settings (pp. 359417). Cambridge, England: Cambridge University Press.
Hopper, R., & Chen, C.H. (1996). Languages, cultures, relationships: Telephone openings in Taiwan. Research on Language and Social Interaction, 29, 291313.
Jefferson, G., Sacks, H., & Schegloff, E.A. (1987). Notes on laughter in pursuit of intimacy. In
G.Button & J.Lee (Eds.), Talk and social organization, (pp. 152205). Clevedon, England: Multilingual Matters.
Mandelbaum, J. (1987). Couples sharing stories. Communication Quarterly, 35, 144171.
Mandelbaum, J. (1989). Interpersonal activities in interactional storytelling. Western Journal of
Speech Communication, 53, 114126.
Mandelbaum, J. (1991/1992). Conversational non-cooperation: An exploration of disattended complaints. Research on Language and Social Interaction, 25, 97138.
Morrison, J. (1997). Enacting involvement: Some conversational practices for being in a relationship. Unpublished doctoral dissertation, Temple University, Philadelphia.
Pomerantz, A. (1978). Attributions of responsibility: Blamings. Sociology, 12, 115121.
Pomerantz, A., & Fehr, B.J. (1997). Conversation analysis: An approach to the study of social
action as sense making practices. In T.A. van Dijk (Ed.), Discourse as social interaction: Discourse studies 2A multidisciplinary introduction (pp. 6491). London: Sage.
Sacks, H., & Schegloff, E.A. (1979). Two preferences in the organization of reference to persons
in conversation and their interaction. In G. Psathas (Ed.), Everyday langauge: Studies in ethnomethdology (pp. 1521). New York: Erlbaum.
Schegloff, E.A. (1987). From micro to macro: Contexts and other connections. In J.Alexander,
B.Giesen, R.Munch, N.Smelser (Eds.) The macromicro link (pp. 207234). Berkeley: University
of California Press.
Schegloff, E.A., Jefferson, G., & Sacks, H. (1977). The preference for selfcorrection in the organization of repair in conversation. Language, 53, 361382.
Watzlawick, P., Beavin, J., & Jackson, D. (1967). Pragmatics of human communication. New York:

A Note on Resolving Ambiguity
Gail Jefferson
Just about twenty years ago, working on materials in which people talk about their troubles, I came across a possible phenomenon: Someone inadvertently produces an ambiguous utterance, then attempts to disambiguate it without speaking explicitly. Although it was
clear to me that something like that was going on, I found that I had no analytic resources
to develop a case for it. I gave a talk to some colleagues at the University of Manchester,
presenting the phenomenon as something intriguing but that my conversation analytic
resources gave me no handle on, and was told in no certain terms that my muchvaunted
conversation analytic methods had utterly failed to handle it. Theyre a lively bunch! Even
agreement turns into open warfare. At some point, someone suggested that we just go have
a drink. So ended my presentation.
Since that time Ive every now and then come across another candidate case (and
although the original instances occurred in the materials I happened to be investigating at
that time, the phenomenon is not exclusive to troubles-talk). Recently I took another shot
at itnot that I can handle the thing any better now than I could twenty years agobut just
trying to suggest that such a phenomenon might exist, and that this or that fragment of data
might comprise an instance of it.
Perhaps Robert Hoppers phrase roughing up the ground best describes what Im
up to.
Ill start out with a few fragments in which it seems to me that one participant has produced a characterizably problematic utterance, then resolves the problem, whereupon a
recipient produces an appropriate next utterance.

(1) [Goodwin:60:C:12]
( (Two women at a block party, chatting about college days and characters they have known.))
We had this one girl she wz from Flo:rida. Un

I swear tGo::d, she wannid tbe on the bes


dress list.

Ener parents apparently wereneven that

wealthy. En she wenoutn she bought tons of


clothes so she cd be on thbesdres-She even


came tcollege inna pegnoi:r se:t.


10 Lauren:
Yknow. u-mean who goes tih college inna with a=
11 Tanzi:

= [Who even o:wns] one.right?

12 Lauren:
[pegnoir set.]

A note on resolving ambiguity 187

Problematic here is that Lauren seems to be describing a young womans arrival on a
college campus wearing a negligee (in a pegnoir set, Lines 78). There may be good
grounds for Tanzi to figure that Lauren means to be saying something less drastic, that is,
that the young woman brought with her, among her tons of clothes, a pegnoir set. She
didnt arrive in one, but with one. The story structure itself may be angled toward the
less drastic alternative; a story about someone showing up on campus wearing a negligee
would probably look different from the start. On the other hand, funny things do happen at
college. So, Tanzi may be holding off taking a position.
We may be seeing Lauren discovering her error as she recycles the punchline with its
problematic in a and immediately thereafter produces the problem-resolving with a (I
mean who goes to college in a with a, Line 10).
Whereupon Tanzi produces a next utterance appropriate to the with a alternative,
addressing herself to the ostentation of having such a thing rather than, say, the brazenness
of wearing it. And this is whereupon in a strong sense. Not just somewhere afterwards,
but immediately upon the occurrence of the clarifying phrase.

who goes tih college inna


Who even o:wns one.

While the problem in the preceding fragment does have to do with alternatives, it doesnt
involve the sort of ambiguity Ill be focusing on, where a single item could mean one thing
or another.
The following two fragments do involve that sort of ambiguity. As in the preceding
fragment, immediately upon the occurrence of disambiguation, we get an appropriate next
The first of the two fragments comes out of a telephone conversation between two men
on duty at different locations during the 1964 Anchorage, Alaska, earthquake. They refer
to each other by their locations: City is the Anchorage fire department and Elmondorf
is an outlying army base. Theyve been connected by a short circuit in the telephone system, and have taken the opportunity for a chat. In this course of that chat, the following

[FD: Finger:23]
Dyou know wt-wt kinda newsere broadcastin

downn thStates et (.) presnt?

I: heard dfir:st
[xxxxxxx] rxxxxxxxxxxxxxxxxxxxxxxxxxxxrxx
[ (2.0) [The

These two fragments and my discussions of them are taken from Jefferson (1986).

188 Studies in language and social interaction



10 Edorf:
11 City:
12 Squawk:
13 Edorf:
15 Edorf:
17 Edorf:
19 Edorf:
20 Ci[ty:

firsone thet dey uh, (0.7) broadcas w z sixty

tun thr
[Yer loudn clear Muldoon Tower,
I heard d [firsbroadcasStateside,]
[(xxxxxxxxxxxxxxxxxxxxxxx]x [xxxxx)
[Justa minnit.
Ci ty,
[Ye-u- ah heard dfirsbroadcasstate det deh
wz bout sixty tthree hunnerdea:d n (0.4) city
of Anchrage is on dih grou:n

Just as City starts to answer Elmondorf s question, a squawk box on the Elmondorf side
starts up with a report from Muldoon air field (Lines 14). And we can watch Citys
work by reference to the squawk box. He initially drops out (lines 34) and then, perhaps
because he gets no indication from his coparticipant that he should maintain his silence,
he may take it that someone else on duty is handling it, and he starts up again (Lines 36).
But it turns out that his coparticipant is handling the squawk box, and interrupts him to
respond to it (Lines 58). And City drops out, remaining silent until hes invited back by
Elmondorfs Pardon?, to which he responds immediately (Lines 811).
That City hears Elmondorf s Pardon? as directed to him and not to Muldoon Tower
may be, at least in part, because Pardon? is a conversational object, in strong contrast to the instrumental Youre loud and clear with which Elmondorf responded to the
squawk box.
But again, just as City gets going the squawk box starts up, and Elmondorf, again with
a conversational object, Just a minute, indicates that City should drop out and give the
squawk box priority (Lines 1014).
Now comes what Im proposing to be the ambiguity. In his next utterance, Elmondorf uses Go ahead, which is both conversational and instrumental. This may generate
a problem for City: which of them is being told to Go ahead, he or Muldoon tower?
And it appears that Elmondorf comes to see that there is a problem and what the problem
is. After two such invitations go unanswered, he shifts to a non ambiguous item, naming
his selected coparticipant: City. Whereupon City respondsand whereupon in a very
strong sense, that is, after the first syllable of the identificatory word:2
Jefferson, (1986), the whole point of the exercise was that one cannot be certain that City starts
to talk by reference to Ci and not by reference to the prior Go ahead, his response merely incidentally occurring at a recognition point for the identificatory word. The same reservation, on an
even finer scale, holds for Fragment 3.

A note on resolving ambiguity 189





Go head.
Go head,

Ye-u- ah

In the following fragment, the whereupon feature may be really exquisite. And for this
fragment Im preserving the initial consonant and vowel of the actual names of two of the
participants, Jesse and Joan, in order to show just how delicate this business may be.
The fragment is taken from a group therapy session for teenagers. This particular session is being observed from a room behind a one-way mirror.
(3) [GTS: I:2:19: R:5]
( (Jesse is reporting a success with his parents; they have stopped interrogating him about his comings and goings.))
Nobddy sez inning yih jis keep whha:lkin.

hh yihknow






its bghuggin mhhe(h)now [ hm hm ]

[Donta] lk tih them
talk tu: S: : .
No. (.) th- (0.4) drapes er closed now I cn see
through that liddle crack et thwindow over there
Yer very consciousv thm being in the : re .
Je [sse.
[He keeps:: [talk [inthere.]
[ih [It doesn] rilly bother me,

This may be a very touchy moment. Joan having raised the issue of observers in the first
place, (Lines 56), it is possible that Davids remark (Line 11) is addressed to her. Indeed,
the appending of Jesses name by David may be directed to clearing such a possible ambiguity, similarly to Elmondorf s work in Fragment 2 with his shift from Go ahead, and
Laurens work in Fragment 1 with her shift from in a to with a.
(But whereas Laurens shift, involving as it does a mid utterance substitution, is clearly
a self-repair, Elmondorf s is less obviously a matter of repair, in that after a bit of silence
he produces a legitimate next component for a single utterance, that is, Go ahead (0.2)
City. And Davids shift is even less obviously a matter of repair, coming off as a through
produced sentenceutterance with the disambiguating name in tag position: Youre very
conscious of them being in there Jesse. Were left with some intonational details, the standard ending intonation of in the:re., which might lead us to wonder if the disambiguating
Jesse was not appended to a completed sentence-utterance specifically in order to resolve
a just discovered ambiguity.)

190 Studies in language and social interaction

And, similarly to City in Fragment 2, Joan could be monitoring for which of the two
candidate addressees (in this case, which of the two who have shown themselves to be
conscious of them being in there) is being addressed.
But the recognition work in this case would have to be a bit finer than that proposed for
City in Fragment 2, because in this case the name of the other candidate addressee starts
with the same consonant as does Joans. Involved in this case, then, would be response
upon occurrence of the crucial differentiating vowel, at which point, and no sooner, selection is achieved. And it is at just that point that Joan launches a next utterance appropriate
to Jesses being the one addressed by David:

Yer very consciousv thm being in



And that is whereupon in a very fine sense.

In the following four fragments, the circumstances become murkier. In each of them it
seems to me that someone, having produced an ambiguous utterance, then tries to achieve
disambiguation without the sort of explicitness found in the prior materials. That failing,
in three of the four we do getperhaps specifically as a last resorta disambiguating
In the first of the foura leisurely conversation between two neighbors, Reva and Jane,
in the laundry room of their apartment buildingthe talk has turned to an allergy that
Janes husband is suffering from. At some point thereafter, the following occurs:

[Gold: MS:1617]

En its annoying. cause you-jih-you-you


you nevuh had it befaw n allv a sahdn yih

getting all dih [sy:mptom [s,


[Mm:, [nYah I know.


I think it has a lo:t tdo wih tha:t.


En the fa:ct thet (.) they dont know what eez

allergic to yet.=

=Ih makes (a),


( )

En my sistuh callme today she siz to me how

evrything out the:re how is it is evry thing

unduh control?


A note on resolving ambiguity 191




Ah sid I guess it is the planes ah le(h)nding I

say I donknoh:,
uh-huh eh-heh eh-heh.

Reva presents her sisters question as a multi component utterance, How is everything out
there, how is it, is everything under control? (Lines 1416). This may be a faithful rendering of her sisters words. It may also comprise serial attempts by Reva to disambiguate
what she has come to see as a possible reference to some sort of illness-related problem
topically coherent with the prior talk, when what she intends to be referring to is a dramatic
but short-lived strike by the citys air traffic control personnel. (In the first place, out
there may be fitted to a trouble of the area in general, in contrast to, e.g., with you. And
perhaps at the subsurface, poetics level, is everything under control came to be produced
via its resonance with air traffic control.)
In this case, activities that may be attendant to a problem and its solution are embedded
in bland colloquy; Reva quoting an exchange between her and her sister consisting of a
multicomponent question and a similarly constructed answer (Lines 1419), in which one
component of the answer, the planes are la(h)anding, happens to be an explicit reference
to the topic; Jane responding, not thereupon, but after a next component, I say I dont
kno:w (which, contributing nothing substantive may work as a recompleter), with a mild
laugh (Line 20) that, although it occurs at a distance from the disambiguating component
may yet be fitted to it, given the laugh particle in la(h)anding.
So although matters in Fragment 4 are worked out in a more dilatory fashion than in the
prior three fragments, there is still some evidence of a problem and its solutionfor both
speaker and recipient.
In contrast to the foregoing where, in the first three fragments we have the recipients
whereupon responses and in the fourth, a response that, although not immediately thereupon, may yet show its relationship to the solution-bearing component, in the remaining
three fragments we lose the recipient as a resource. As far as I can tell, their responses are
completely opaque for the problem-solution issue.
The following fragment and its consideration is taken from the work I did on troublestalk and is one of the cases in which I first noticed the possible phenomenon (Jefferson &
Lee, 1980).
The situation is this: The adolescent son of divorced parents has driven down from Palo
Alto where he lives with his father, to visit his mother in Los Angeles. At some point in the
visit, his car is vandalized. Hes left the car with his mother and is flying home unbeknown
to his father who is expecting his arrival by car and has phoned the mother to find out his
sons estimated time of arrival, only to be given the news.

[MDE: MTRAC:601:2:R:12]
Hi: Sheila?
How are you.

192 Studies in language and social interaction

















I wz wondering wene left.
.t.hh Uh(d) did OH: .h Yer nod in
on wut ha:penhhnt.
No(h)o [ (wut he-)
[Hes flying.
En Nadine [Joes girlfriend] is going to meet
im:.=Becuz the TOP wz ripped o:ffv iz
car which is tih say somebddy helped thmselfs.
Stolen.=Right outn frontv my house.
Oh fer c:rying out loud En eez not gnna- eez
not gnna bring it ba:ck?
h No so its parked in: thih gihrage cz it wz
so damn co: ld. <and ez a mattuh fac snowing
on the Ridge Route.
hhh So I took him to the airpor he couldnt
buy a ticket.
B- he cd only get on sta:nby.
Uh hu: [h,
[En I left him there et abayou:t noo:n.
Uh ha:h.
Ayund uh,h
Wuts e gundo go dow:n pick it up later? er
somethn like [ey- [Bt thats AW ] fl
[hh[His friend-]
Yeh [ his friend S t e e-
[ (Boy) that really makes] me ma:d,
hhh Oh its digusti [ng iz a matteraf] a:ct.
[P o o r J o e y.]
I- I:, I told my ki:ds- who do this: . down et
the Drug Coalition ah want th TO: P ba:ckhh.

A note on resolving ambiguity 193








SEND OUT THE WO:rd. hhhkhuhh

Bu:t (.) hhghuh: his frienSte:ve en Brian er
driving up.
right after: :< (0.3) school is out. En then hill
drive dow:n here with the:m.
Oh I see
So: in the long run hhh it (.) problys gnna
Save a liddle time n: energy.

As Sheila described what happened, Monty exhibits what seems to be more concern for the
car then concern for his son; for example, his response to Sheilas initial announcement,
Hes not going to bring it back? (Lines 2122), his non response to her report of icy cold
weather in which Joe would have to be driving in a car without its convertible top (Lines
2426), and his non and minimal responses to her report of Joes troubles at the airport
(Lines 2732).
Which is to say, Montys treatment of Sheilas report raises as a possible issue that the
boy has been irresponsible, simply abandoning a problem as adolescents are wont do do .
Focusing on the arrowed series of assessments, the initial one, But thats awful.
(Line 40), may be an attempt to repair what might look like a display of more interest in
the cars return than in the boys circumstances. It occurs immediately after a statement of
concern for the cars return, prior to completion of the utterance in which that statement
is packaged: Whats he going to do, go down and pick it up later? Or something like eyBut thats awful. (Lines 3940). In that rapid juxtaposition is an echo of Fragment 1 with
Laurens shift from in a to with a. And as Lauren may there be discovering her error,
Monty, hearing himself expressing concern for the car (for the second time, cf. And hes
not going to bring it back?, Lines 2122), may be discovering the infelicitous direction
of his concern, attempting to repair that with a self-interruptive display of concern for
the boy.
However, the assessment he uses is non selective; it could apply to either concern.
And following on the heels of an expression of concern for the car as it does, it might
conceivably be heard as assessing his sons abandonment of the vandalized car.
As it happens, Montys assessment occurs in overlap with something Sheila has started
to say (Lines 4142). She, having cut off her overlapped utterance, minimally acknowledges Montys talk with Yeh and starts again, now overlapped by his next assessment,
which starts up immediately after her Yeh (Lines 4243).
The Yeh is at best no help to Monty in deciding if his initial assessment has been heard
by reference to the vandalism or to his sons irresponsibility. At worst, it may be weighted
toward the latter, hearable as Sheila, in the interests of keeping the peace, acceptingif
most minimallyhis assessment of the boys (and her own) handling of the situation.

194 Studies in language and social interaction

And conceivably it is in response to the non- or wrongly commital Yeh that Monty
makes a next attempt at disambiguation. But, as in Fragments 2 and 4, rather than producing something more selective of one or the other relevant alternatives than was his But
thats awful, he offers another item of the same sort,(Boy) that really makes me mad
(Line 43) cf. Elmondorf s repeated Go ahead and Revas added how is it, is everything under control? And it may be that the offering of a same or similar item can alert
a recipient to a problem in their response to the initial item while preserving non explicit
But in this case, whereas, for example, an expression of anger on his sons behalf such
as Boy I bet hes mad might not only have done such reoffering work but could have
fostered selection of the vandalism alternative, Montys expressing his own anger allows
for (and perhaps even promotes) selection of the irresponsible-kid alternative.
And given the persistent bivalence of the talk so far, Sheilas concurring Oh its disgusting (Line 45), which does not select for one or the other alternative but refers to whatever it is that Monty is referring to, could at least possibly be concurring with his prior
utterance as an assessment of the boys abandonment of the car and not the vandalism.
For Montys assessments and Sheilas concurrence to be unequivocally understood as
addressing the vandalism and not the boys behavior, we need to refer to and rely upon our
shared knowledge of the conventional proprietiesfor example, that a father cares more
about the welfare of his son then about a chunk of Detroit metaland assume that the
speaker and his recipient share those proper concerns.
It appears that in this case the father does not feel able to depend upon those conventional proprieties for deciding how is ex wife is hearing what hes saying, or for him to
decide what she is saying.
And what occurs next is an utterly explicit utterance that resolves any possible ambiguity, Poor Joey. (Line 46).
This utterance is positioned in just the way Pomerantz described for second assessments,
that is, with minimization of gap between its initiation and prior turns completion; in this
case, as in several of those she showed, occurring in slight overlap (Pomerantz, 1984):

Oh its distgusti
Poor Joey.

That is to say, as a sequential object Poor Joey comes off as an understanding/agreeing

response to Sheilas utterance, and not at all as some sort of repair.
Nevertheless I would argue that Poor Joey is indeed some sort of repair; this expression of pity, so unlike the sort of talk that Monty has been producing throughout the interaction, being enlisted specifically to resolve the as-yet-unresolved ambiguity.
In armchair-psychological terms, Poor Joey may have been generated out of the fact
that Monty does blame his son and is in fact angered by the boys just walking away from
the vandalized car, and thus can hear his own words and those of his recipient as at best not
clearly enough not blaming the boy. It may be that he has found himself forced to produce
something so drastically over solicitous to make himself heard through the crescendo of
blame that has only intensified with each next utterance.

A note on resolving ambiguity 195

The following fragment also involves the relational-pair categories parent-child, with
the attendant conventional proprieties. Again the disambiguation does not come off as a
solution or repair, and again the recipients responses are inscrutable.
Heres the situation. Christmas is approaching. Two young mothers, Ann and Linda,
are chatting on the telephone and talk has turned to presents for the kids. Ann has already
bought some for her own kids, and also some for Lindas kids. At one point shes remarked
that what Ive got for them theres no way youre going to be ableto get it in your car,
which sounds pretty impressive.
As the fragment begins, Linda is asking what Steven, one of Anns children, wants for
Christmas. To Anns I dont know she responds I dont know either (Lines 15); that
is, she speaks of herself as a candidate gift giver in search of the right gift (and perhaps
something pretty special) for her friends little boy.









So:: Whatd Stevn wa::nt.
hhhhhh Oh:::(m) tch I: dont kno [w, [
[tch I dont
know eether. [I
[(Bt) he keeps tell in yi.hknow
before he mentioned thet he said he wannid uh
( .) tch a tra::ctor.
[hhhh En, I donknowf they have those liddle
To: nka things? bt hes go-ot two athese (.)
grader uh not graders bt tra:ctor things
out [here.
[Yea [:h? h?
[hhhh Anthats aonly thing yihknow
he kept telling s- u- Donna one day she went with
me tin the store en she stayed in th- car with
thkids en then I hh-hh did the sa:me fer he:r,
hhhh a:n uh:m sh- t she said thet s- Steven
said he wannid the tra:ctor.=
[M m : : : : .]
=hh Oh thats ni: ce hhuh heh heh heh h [uh- u]
[heh he] h=
=He aint gettn one, [heh] heh hhhh] hh=
[Ye: ah .]
=Bu:t. I dont know I rilly(d) (0.2) phhh Hes

196 Studies in language and social interaction







so ha:rd. tuh figure out (.) what tih git

im [this year]

((ca 8 lines omitted, in the same vein))

I got im a lotta things tih jis:siddo:wn

en [:
=pk en do things. I do: nt I donknow I really
dont wannim tuh hhave a lotta stuff ....
So : : (m) tlk hhh I dont know just (0.2) ga: :me
yihknow booksn:: stu [ff he cn] do stuff hh
[Mm::.] [Yeah.

Linda as candidate gift giver in search of a gift for Steven may be what sets up the ambiguity problem here. When Ann does mention something Steven really wants, a toy tractor
(Lines 1620), one question might be whether Linda is listening to the anecdote that that
information is embedded in as a story recipient or as an information seeker. And what may
be happening in Anns series of utterances following Steven said he wanted the tractor is
an attempt to convey to Linda that shes neither to run out and buy the kid a tractor nor to
feel accountable for not doing so, without saying so in so many words. (While Ann might
have avoided the whole problem by simply not mentioning the story of Stevens telling
their friend Donna that he wanted a tractor, she might forsee Donnas mentioning it to
Linda and be trying to head off whatever problems that might entail.)
Anns initial attempt to defuse Steven said he wanted the tractor, a dont take this
seriously marker, the laughing recompleter (h)y(h)ou kno(h)w is overlapped by Lindas
simultaneous appreciative Mm:::. To the mention of the tractor (Lines 2122). (That
the next place Linda produces that sort of utterance is at the fragments end, when Ann
summarizes the foregoing talk by mentioning some things that shed like Steven to have,
So, I dont know, just game you know, books and stuff (Lines 4345) suggests that its
initial occurrence might also be produced as a response to a gift suggestion made to her
by Ann.)
Anns next attempt, her ironic self-quoted response to the storied announcement that
Steven wants a tractor, Oh thats nice huh heh heh heh is received by Linda with a little
laugh (Lines 2324).
That is, Lindas responses give no indication that she sees herself off the hook when it
comes to the toy tractor.
Anns problem here may be the reverse of Montys in Fragment 5. That is, whereas
Monty may be not at all sure that the conventional proprieties are working for him so that
hell be understood to be more concerned for the boy than for the car, Ann may be discovering that the conventional proprieties are working too well, that shes not being heard as not
wanting her child to have the toy he so much wants for Christmas.
And it is, perhaps, therefore that we get the disambiguating He aint getting one, heh
heh (Line 25).

A note on resolving ambiguity 197

I have a feeling that this utterance is as uncharacteristically callous as Montys Poor
Joey in fragment 5 is uncharacteristically solicitous. But in this case the callousness may
specifically be produced to be taken lightly, not only with the appended laughter, but with
the aint. Shes to be heard as doing talking tough to get a point across. And a bit further
on, the possible callousness of He aint getting one is shown to have been a matter of
motherly concern; that instead of toys, shed prefer him to have game books; things that
promote activity (see Lines 4144).
A quick note about aint. Ive transcribed two phone calls between these two women;
this very long one (ca 45 minutes) and another, shorter one. This is the only occurrence of
aint. All other utterances that could be done with aint are done with standard syntax.3
Not long after Id put together an earlier draft of this exercise, I began watching coverage
of the O.J. Simpson trial. Several times I heard aint used in the way Ann uses it. And
in some instances, the aint was embedded in language a cut above the ordinary. For
[TV news, caught in passing]
((Cindy Adams, New York Post columnist))
Adams: If theres a better system anywhere I aint found it yet. But theres something
inherently wrong with whats happening in this case.
[CNBC Special Report, 42495]
((Manny Medrano, commentator, asked about the feasibility of using professional jurors.))

That also (.) aint gonna happen fthe feeruh-r-f The reason thet it rilly flies in
the face of Constitutional protections,

And just recently, looking through some medical data collected in 1992,1 came across a
physician making similar use of aint.
[HospSite: PIS:82792:2122]
((Senior attending physician Slater is commenting on intern Fitchs suggestion that a patient be
scheduled for a psych consult))
It might be worth it causeit might be Yknow kind of [an unstable mo
ment where

[Mm [Mhm
hhh just getting on a waiting listn having an: (0.7) hhh (.) something happen in a couple

months just (.) aint gonna do the jo: [b.


hhh Its not that shes got a crisis its just this is the m- the right ti:me


M [hm
[period in which something ought to nappe [n.

For example (and these are all by Ann): [TCI(b):16] p. 1. Im not gnna have it done., p. 15,
Im not worryin about it., p. 57, Its not rilly like a cowboy thing, p. 60, thats not yours,
p. 79, hes not doing that. [TCI(c):12] p. 5, Were not answering., p. 13, yer not talking tuh

198 Studies in language and social interaction

(Especially nice here is that having used aint gonna do the job to make his point, Dr.
Slater returns to the standard syntax of Its not that shes got a crisis )
These sorts of materials can lead us to see Anns He aint getting one, not as an
expression of callousness, but as an idiomatic resource shes put to work to make herself
utterly clear in an environment of persistent ambiguity. And in that regard, then, it may well
be that Montys strikingly solicitous Poor Joey is a similar sort of resource being put to
similar work in a similar environment.
The final case and its consideration, like Fragment 5, comes out of the early work on
troubles-talk (see Jefferson & Lee, 1980). As in the preceding three fragments, we get
a series of ambiguous utterances. Unlike the preceding three, this one has no explicit,
last-resort component. Thingsif they are adriftremain adrift.
In this section of the troubles-talk report, the point being made is that although troublestalk seems to have the potential for progressing as an orderly sequence, it appears to be
enormously susceptible to contamination by other types of activities. One such contaminant is the negotiating of a plan, in which one participants trouble is the others obstacle.
In the following fragment, someone has phoned with a project in mind (leaving her
little boy to be looked after for a while so that she can go shopping) and discovers that the
intended coparticipant in the project (the babysitter) has a trouble that may be consequential for that project (shes got the flu). And once again, the issue of proper parental concern
for a child seems to be involved.
( (Call opening unrecorded; Lily is the caller and is now identifying herself to Cora.))

Jo:dys mothe:r?


Oh ye [h ((very hoarse, here and throughout call))


[Jo:dy Lih- tempi,


Oh: yen,


Are you si::ck,


tch ah got the flu.





aOh::::.uh [hnh [hnh ha] ha-ha-ha

[h- [hhhhh] hh-hh-hk
[Wul that ni:ps itnna bu:d, hh ah wz gonna ask
yuh if yih cd keep Jo:dy fer a c(h)ouple hours
but yih cant if yih got the flu::
tch Ah wouldnwanim arounme ho:n,t=
[hhhhhhh]hh Cause uh: ahv really ghhot it.
yo [u sure-]

A note on resolving ambiguity 199









But ahd be glad=do it if I wasnt sihhck.
-yousure sound aw:ful ul. [(hoarse.)]
[ t Oh: :] my God
ah been hhh running thhighestempihtures you
ever sa:w.
Omy go:sh well let me hang upn letchu git back
tuh be:yudh=
=eh huh [uh uhh] h h
[So:rry]I disturbed you.]
=Hayih doin hhon=
=Oh jes fi:ne.

Across the fragment, the trouble is talked about by reference to its consequences for Lilys
project; that is, will the fact that Cora has the flu stand in the way of her minding Lilys
little boy. The presence of a symptom (hoarseness) and the announcement of the flu does
not in itself terminate the possibility that the project can be carried out. This is perhaps
because a feature of the term the flu is that it gets applied to almost anything and may here
be naming something quite mild, and a feature of hoarseness is that it can be residual and
not at all debilitating. So the sheer assertion Ive got the flu (Line 8) and the presence of
hoarseness are in a range of ways unreliable indices.
And it appears that although Lily announces absolute withdrawal of the project, Well
that nips it in the bud (Line 13), she is allowing for and perhaps specifically pursuing its
being carried out. For one, several alternative courses are available to her. She might at
this point introduce the Sorry I disturbed you, which eventually closes off discussion
of Coras flu (Line 31). Less drastically, she might now initiate the diagnostic inquiry that
occurs midway into the discussion, You sure sound awful (Line 24). Instead, having
announced abandonment of the project, she goes on to describe it: I was going to ask you
if you could keep Jody for a couple of hours and her grounds for abandoning it: but you
cant if youve got the flu (Lines 1315).
And in the description of the project is at least one detail that might tend to urge for its
being taken on by Cora; that is, there is mention of the briefness of the intended period
of babysitting (a couple of hours), where, that something will take but a little while is a
routine component of such negotiations.
Then there is the proposed reason for abandoning the project, but you cant if youve
got the flu. This utterance strikes me as a proposal offered for confirmation or disconfirmation, perhaps because stating it makes it sequentially relevant; a response to it is due.
Also, the proposed grounds for abandonment of the project are specifically disattentive
to what ought to be a crucial concern if the flu is being taken seriously; that is, it ought
not to be that this sick woman cant take on the job, but that if she is sick the child ought
not to be exposed to her.
So, in this utterance that announces itself as abandoning the project, there is a minimizing not only of the task (just a couple of hours) but of the obstacle (no concern about
contagion), and the babysitter has been put into a position of confirming or disconfirming
that she cant take on the job.

200 Studies in language and social interaction

Now we come to the target series. In the utterance that confirms that the project ought to
be abandoned, it appears that Cora is addressing the seriousness of the flu by reference to
possible contagion with I wouldnt want him around me, hon (Line 16). She is in effect
hanging up a quarantine sign.
But the utterance is ambiguous. It is at least conceivable that what is being referred to
is the child as a nuisance to a sick person rather than (or as well as) the sick person as a
source of contagion for the child.
This is a very real issue, and it does show up in conversationbut interestingly, at least
in the cases Ive noticed, not as a person-to-person assertion, but as a third-party report.
So, for example, in the following fragment a woman is talking about her daughter Janets
very ill father-in-law.


Janet sd he looked (.) awflly ba:d though bcourse Fre:d ditn say e
looked so ba:d but uh: (0.4) what kinyih do:, hes ho:me en yee ah mean they
cant have
the kids aroun distur:b Yihknow

And in the following fragment a woman is talking about her daughter-in-laws mum,
who has yet to see her newborn granddaughter.



And uh (0.2) her mum rang me this morning n (0.3) they could get from Salsbry
just uh within a day but sh sez I cant go n see er Ive got bronchi:ti [s

[Oh dear what a sha:me.

Shsz I ca:nt go anywhe(h)re nea(h)r them an

she do(h)nt feel like it anyway you [know,


Fragment 7.b is especially instructive. We get both aspects specifically referred to, that is,
sick person as a source of contagion (again with the self-quarantining, stay away formulation: here, I cant go anywhere near them, in Fragment 7 I wouldnt want him around
me), and sick person as in any event unwilling.
Further, the covert character of the latter is interestingly invoked; that is, while Mattie
quotes her fellow new grandmother as saying I cant go and see her, Ive got bronchitis,
she does something else with the unwillingness aspect, not quoting but asserting and she
dont feel like it anyway. How ever she may have come to that conclusion (whether the
other woman actually said it, or some sort of common knowledge is being invoked; i.e., no
ill person would feel like it), Mattie is not ascribing those very words to her, but providing a sort of buffer by forming it up as a statement about her and not by her.
So, returning to Fragment 7, it appears that the understanding of Coras I wouldnt
want him around me, hon as an assertion of self-quarantine in the interests of protecting

A note on resolving ambiguity 201

Lilys little boy from contagion is based on a conventional public propriety. But there turns
out to be that covert aspect, that is, that behind the quarantine sign is one that reads do
not disturb.
Compounding that, is the local context, specifically, that Lily herself is exhibiting no
concern about quarantine.
Given these factors, Cora, having said I wouldnt want him around me, hon and receiving a drawn-out, sympathetic nNo::::, (Line 17), may have good grounds to suspect that
she is being heard to be invoking the do not disturb alternative.
A quick note about nNo::::,. Comparing British and American uses of No as a
response token (not an answer to a question), I found that whereas British speakers use
No for negatively framed priors, for example:

So ah dont kno::w, (.) yihknow when shes com[ing

Americans deploy Uh huh, Yeah, and so on, not only for positive but for negative
priors, for example:
(7.d) [SBL:2:2:R:1]
Allen doesnknow anything new out there eether.
Uh huh,
[TCI(b):8:23] ((re: allergy medication))
R.J. :
En I donknow where she keeps that sorta stu:ff,

reserving No for affiliation; for showing sympathy, solidarity, and so on, often where
values and morals are concerned, for example:

[JG:II(a):3:2] ((Maggie blacked out at party))

she asked me if it wz becuz Id had too much t dri:nk en I sid no
becuz et the t] i: me
[N O : : ; : :.]
[NB:II:2:R:19] ((Nancy knows that Andr lied.))
becuz Andr never stayed home all day tih call anybuddy [Y, h
hhh] hh

In any event, the nNo::::, with which Lily receives Coras I wouldnt want him around
me, hon is not unequivocally selective of either alternative (quarantine or do not disturb and, as in similar circumstances in Fragments 4, 5, and 6, another non disambiguating
item is offered, Because Ive really got it (Line 18), Cora perhaps attempting to alert her
recipient to the existence of a problem while remaining non explicit.
But, in contrast to the prior fragments with their disambiguating third items, Cora
produces yet another non explicit utterance, But Id be glad to do it if I wasnt sick

202 Studies in language and social interaction

(Line 23), and the ambiguity is left unresolved: Is she expressing concern for the child or
for herself?
It is certainly possible that she is using ambiguous talk to pursue attention to her troubles while not explicitly saying poor-me-and-the-devil-take-your-kid.
On the other hand, the ambiguity may be a by-product of an attempt to avoid being seen
as trying to instruct a mother on the proper grounds for abandoning the project; that is, that
its not that Cora cant baby-sit, but that the child should not be exposed to herand that
that ought to have been the mothers first concern.
In which case, across a series of attempts, this speaker might be characterized as invoking, while specifically declining to explicate, the proprieties in hopes that the recipient
will come to see that her prior talk exhibited a misalignment to those proprieties and now
produce talk that will exhibit correct alignment.
And whereas in each of the preceding fragments the problem can be ascribed to the one
who is producing the ambiguous talk, in Fragment 7 it may be that the trouble lies with the
In which case, whereas in each of the preceding fragments the one who produces the
ambiguous talk solves the problem with a disambiguating utterance, in Fragment 7, as the
recipient appears to remain dense to the problem, the speaker may be deciding that tactful
ambiguity is preferable to possibly confrontational disambiguation.
A closing note. One thing we can notice is that whereas in Fragment 7 disambiguation
(possibly for good reason) did not occur, in the preceding materials we did see an eventual
move to explicitness.
One question that raises is, why do we not see an immediate move to something explicit?
Why, for example, in Fragment 2, do we get Go ahead again? Why, in Fragment 5, do
we get another indexicalized complaint (Thats awful followed by That really makes
me mad)?
This may have to do with a general feature of interaction, something that might be called
understanding assumed, which involves that the way in which were talking to each other
is in principle adequate for understanding. Where, then, on any given occasion, resolving
some particular problem by explicating, explaining, and so on, could constitute a rupture
of that in-principle condition of understanding each other.
In one of his lectures, Sacks talks of how monumental in its import it is that in their
interaction people suppose that what weve been talking about all along, you know in the
way I told it to you, and I suppose that in producing any next thing I say. He goes on to
offer a rhapsodic description of a possible consequence of that assumption; that without
thinking about it, the work I do is to find for any item you sayno matter how grossly it
misunderstands what I say, how well it understands what I say (Sacks, 1992, p. 184).
The materials Ive been exploring here may involve a rather more prosaic working out
of understanding assumed on particular occasions when that assumption falters. Specifically, when an initial non explicit reference seems to be getting into difficulty, its speaker
may attempt to alert its recipient to the problem while preserving the utterances original,
non explicit character, and thereby preserving the assumption of understandingit being
only when that attempt fails that the assumption is breached and explication is brought
to bear.

A note on resolving ambiguity 203

Jefferson, G. (1986). Notes on latency in overlap onset. Human Studies, 9, 153183.
Jefferson, G. & Lee, J.R.E. (1980). On the sequential organization of troublestalk in ordinary conversation. (SSRC end-of-grant report).
Pomerantz, A. (1984). Agreeing and disagreeing with assessments: Some features of preferred/dispreferred turn shapes. In J.M.Atkinson & J. C.Heritage (Eds.), Structures of social action: Studies in conversation analysis (pp. 5964). Cambridge, England: Cambridge University Press.
Sacks, H. (1992). [Lecture 2]. In G.Jefferson. (Ed), Lectures on conversation (Vol. 2, p. 184).
Oxford, England: Basil Blackwell.

The Surfacing of the Suppressed
Emanuel A.Schegloff
I was first alerted to the phenomenon I sketch here by an incident in which I was a participant. The episode was not taped; I noticed the key occurrence when it happened in the
course of the interaction, a meeting (job interview would probably be the more accurate
term) with the Vice Chancellor of a small New England university in the early 1970s. After
the end of the interview, I wrote a note about what I had noticed onyou wont believe
thisthe back of an envelope.1 Heres the note:
Talking to Vice Chancellor; he tells about an administration report that slams some
departments and the trouble to be expected when the report becomes public. It is set
up for the shit will hit the fan, but he censors it. Still, its in his brain, as witnessed by: a few moments later, replying to a suggestion that it not be made public,
he says its already in the fan.
So there in a nutshell is a raw description of the phenomenon. If we ask what happens to the
talk that gets suppressed when an utterance gets aborted before being brought to completion, then we sometimes see the suppressed item pop up in the talk later. As I say, thats a
raw description. How can we refine it? And why, or how, is it of interest?
As an initial take, we might say it is of interest, first, because it is a recurrent occurrence
in conversation (if it turns out to be) and it is our job to describe such things. And, second,
because we may well find ourselves called upon to explore and register what has been
suppressed when talk is self-interrupted, and what prompts the suppression. If we have
grounds for looking to a particular place and knowing how to recognize what is to be found
in it, we may find evidence there to support a claim about what was suppressed. And often
enough what was suppressed is the best lead as to how come it was suppressed.
How can we refine the rough initial account? At the very least it would be nice to put
some constraints on the claim that something said later is the suppressed item, and some
constraints on later; surely it cannot be indefinitely later.
And surely we want to press such refinements not on anecdotes written on the backs
of envelopes, but on recorded data that can be inspected over and over again to give us
the best possible chance of detecting this phenomenon. And it needs detecting. As we see
herein, what happens to suppressed material often appears designed to escape noticefor
obvious reasons; if it was wanted to be kept out of the talk once, there may well be grounds
I am, it should go without saying, not recommending this way of working, especially for getting
started on a project, but one should not discard candidate phenomena only because they have come
to attention in this way.

The surfacing of the suppressed 205

for keeping it from figuring in the talk subsequently as well. In fact, I found my most recent
instance while preoccupied with some other topic, in data that I have been working on for
about 30 years, data that were in fact collected several years before my episode with the
Vice Chancellor. Thats a long time to escape detection! Here, I can examine only a few
exemplars, but I think we can at least sketch some of the key features of this phenomenon,
which I am calling the surfacing of the suppressed.2
Let me begin with an exchange that presents (at Line 29) a very simple and accessible version of some of the central features of these occurrences. (The reader is urged to examine
the transcripts with some care and not read around them; notational conventions are
explained in Appendix A. Readers are invited to access the audio of this and virtually all
the data extracts in this article, in a format suitable for most platforms, on my home page,
which can be addressed at <http://www.sscnet.ucla.edu/soc/faculty/schegloff/>, with a link
to the present paper. Should this web page cease to be available, readers should contact me
directly or search the California Digital Library at <http://cdlib.org/>. The extract is from
a telephone call in the late 1960s between two young women who grew up in the same
neighborhood and attended the same college until Bee transferred to another school; here
Bee is asking about the school that she has left and that Ava still attends.)
(1) TG, 4:345:31 (simplified)
35 Bee:
Eh-yih have anybuddy: thet uh:? (1.2) I would
know from the English deparmint there?
37 Ava:
Mm-mh. Tch! I dont think so.
38 Bee:
Oh, =<Did they geh ridda Kuhleznik yet hhh
01 Ava:
No in fact I know somebuddy who ha:s huh [now.
02 Bee:
my got hh[hhh
In Gail Jeffersons article On the Poetics of Ordinary Talk (1996), she employed the term suppression-release (at pp. 8, 18, 20 and 24) for a somewhat different, but not unrelated, phenomenon.
By that term she meant Youre being very careful not to say something, and you succeed in not
saying it, and it sneaks out in the next utterance (p. 8). However, in none of the instances that she
examined in this regard is there an overtly displayed suppression of the talk (e.g., by cutting off the
talk that would articulate the suppressed material), talk that subsequently is released. In three of
the four instances, Jefferson developed a cogent account of an ongoing suppression of some word
or theme that subsequently comes out in the talk, but that was not done as a suppressionwas not
done as a displayed suppression; in the first of the instances for which she introduced the term,
there is a displayed suppression, but it is applied prematurely, and the item hypothetically being
avoided (Blacks) is not the one subject to displayed suppression and does not in fact come out
subsequently. So although Jeffersons account of what she referred to by suppression-release is
tracking something that is thematically closely related to what I am examining here, the details of
the occurrences and their analysis are different.

206 Studies in language and social interaction










[Yeh en s he siz yihknow he remi:nds

me of d- hih-ih- tshe reminds me, hhh of you,
meaning me:.
Uh-ho thats [a- thats a s[wee:t co:mplimint]
[Kuhleznik.= [I said gee:, tha:n]ks
a lo:[t honeh,
[hhhhhhuh huh=
=.hh [Said] yih all gonna gitch mouth shuddup=
=fih you yih dont sto:p i [t.]
I think evrybuddys had her hm[hhh!
[shes the biggest] pain in the a:ss.
[ -fih something,]
.T Shes teaching uh English Lit too, no more
Ohj: :, Shes moved up in the wor[ld]
[She] must know
somebuddy because all those other teachers they
got rid of.hhhh
Yeh I bet they got rid of all the one::Well one I
had, t! hhhh in the firs term there, fer the
firsterm of English, she die::d hhuh-uhh [hhh
She died in the middle of the te:rm?mhhh! =
=Oh thats too ba:d hha ha!=

Note then that this extract begins with a topic-proffering sequence initiated by Bee to Ava,
a sequence whose second try (at Line 38) asks whether they (i.e., the authorities at the
college) got ridda Kuleznick yet, a teacher who is held in low regard by both Ava and
Bee. They work up the Kuhleznick case for a bit, and it turns out that she has not only
not been sacked, but is doing very wellteaching English Lit too, no more composition,
the import of which is registered by Bee (Line 24) as having moved up in the world, and
explained by Ava (Lines 2527) by reference to her knowing somebody because all those
other teachers they got rid of. Such a reuse of a word from a question (Line 38s Did they
get ridda Kuhleznick yet) deep into an extended answer sequence is a practice for marking
or claiming the end of the answering (Schegloff, 1998).
And here (at Line 29) it appears that Bee is aligning with this move to close the sequence
by agreeing with the claim with which Ava has proposed to end it with respect to the fate

The surfacing of the suppressed 207

of the faculty they knew in common, Yeh I bet they got rid of all the one::. Although this
is epistemically qualified to a supposition by the I bet, the turn-so-far still appears on the
way to alignment, projecting a continuation as got rid of all the one[s I had].3 But it is
aborted before getting there.
The turn is arrested in a relatively unusual waynot with a cut-off but with a sound
stretch (marked by the colons near the end of Line 29). It seems to convey, waitamminnit,
Im just thinking of something that makes what I was about to have said not quite right.
It exemplifies a suggestion made some years ago (Schegloff, 1979) that, whereas cut-offs
commonly initiate repair on the talk-already-produced, sound stretches ordinarily initiate
repair on talk as yet unsaid. And so Bee aborts the about-to-have-said-ness of it, and tells
what problematizes it: One of her former teachers could not have been gotten rid of by the
secular higher-ups (so to speak), because she died. And thereby what was on the way to
being an agreement with what Ava had said, and an alignment of their views and the closing of the sequence, is derailed. It is turned into an exception to what Ava had said, and
thereby at best a nonalignment, perhaps even a disagreement and challenge (a characterization resonant with the well that initiates the new departure, well being often deployed
as an opposition- or disagreement-marking token).
As it happens, this outcome characterizes virtually every sequence and topic in this
conversation. At one point, having secured from Ava an agreement that she is home (she
must be, because that is where Bee called her and she answered, and this is before callforwarding technology), Bee remarks in frustrated vindication (or vindicated frustration),
See? hI-Im doin somethin right tday finally, I finally said something right. (0.2) You are
home. Still, Ava finds a way to distance herself even from this inescapable truth: Yeh-1
believe so. Physically anyway. Bees backing away from the alignment we are examining,
concerning getting rid of teachers, is just one appearance of something deeper and more
pervasive going on in this interaction and in the relationship of which it is the most recent
(and possibly the last) episode.
But note how Bee starts this exception: Well one I had t! .hhhh in the firs term
there, and so forth. Note two things. First, what follows the suppression of the ending of
the turn unit that was aborted includes in its very beginning just the words that appear to
have been suppressedI had. Indeed, we almost reflexively use those words to either
reconstruct, or ground the reconstruction of, what that aborted ending was going to be.4
Second, note the break between one I had and its descriptor in the first term there. This
ends up being a single phrasal person referenceone I had in the first term there, but it
is fractured in the middle, both with the tongue click and with a substantial inbreath, an

The brackets enclose a plausibly projectable continuation of the talk that was not in fact articulated.
As we do with error correction; cf. Jefferson, 1974.

208 Studies in language and social interaction

inbreath that displays the at least transient unit-in-itsown-right status of this chunk, and
the persistence of the boundary that was projected to occur after had.5
Here is another, quite similar, instance (at Lines 38 and 41). Mark has been visiting and
schmoozing with Sherrie, Karen, and Ruthie in their dormitory room in the mid-1970s, talk
mostly dominated by Marks recounting of his recent social life. Then:
SN-4, 12:1540.
Yih know my stomach after every meal now feels

r:ea:lly weird n its been giving hh Mi:les got


Digel tablets? n stuff like tha:t?



[A:nd uh: like-(-) ts r:ea:lly weird, ( too).

hh- I find one thing .dont eat their pineapples.


They make yer stomach imme:diately after dinner


really feel lousy.<t least m.i:ne.=

=Their pineapples ca:nned.

(I ont care,) its still terrible,
hhhh HUH-HUH hhhh hh they really- just turn my

stomach. Sumpm after dinner [(ih) (-)(s)] turning


in yer stomach .hh

[hhhh hh]

But U: m:

Cest la vie, cest la vie,=


Mark: :-->


Mark: :-->

Thats about it hell I havent been doing anything
but- () s- (Well,) (0.2) going out [actu] ally.
I af tuh start studying no:w

Mark is apparently starting to complain that he has done nothing but s[tudy], which is (by the testimony of his own prior talk) the opposite of the case. When he comes to the payoff component
of this turn-constructional unit (at the start of line 39), he suppresses it, and confesses that he has
done nothing but good times. The correction from what hewas about to say to the truth is even
underscored by the actually which serves here (as it often does; Clift, 1999, 2001) as a correctionmarker. And then the suppressed studying surfaces in the turn to repentance which follows
(at line 42); one might almost hazard the conjecture that this further extension of his talk at this
juncture is designed to accommodate the surfacing of the suppressed element of the prior talk.

The surfacing of the suppressed 209

So the candidate finding I want to take away from this instance is that something that has
been suppressed in the course of producing talk in a turn may pop up in the same words in
the very next spate of talk. We add to and shape this observation as we examine additional
candidate exemplars, but, for now, we have this: what was suppressedthat is, the word
or words that were suppressed (if they appear to have been projected), may surface in the
immediately following talk. That gives us something to look for and a place to look for it,
and those two thingsposition and compositionare major parts of all sorts of practices
and phenomena in talk-in-interaction.
There are occurrences that look very much like suppressions, ones in which the suppressed item pops up in immediately following talk, which however are a quite different phenomenon. They are instances of same-turn repairs accomplishing the operation of
insertion. Thus for example:

Joyce and Stan, 4:0711

Stan: And fer the ha:t, Im lookin fer somethi:ng uh
little different. Na- uh:f: not f:: exactly funky

but not (.) a r-regular typea hhh >well yihknow

I I< have that other hat I wear, yihknow?

Joyce: Yeah,

In this telephone conversation recorded in the mid-1970s, Stan is soliciting advice from
his sister Joyce about where to purchase a hat and a pair of sandals. At Line 02 he appears
to suppress somethingwhich begins with an fwhen he says about the hat that he is
looking to buy, not f::. And when a moment later the word funky comes out, it may
look like the surfacing of the suppressed. But Stan has in effect put the utterance-so-far
on hold in order to insert somethinghere, the word exactlybefore the word he was
in the process of saying, after which he returns to the saying of it; thus not f:: exactly
funky. Funky has not been suppressed, only to surface anyway; it has been held in
momentary abeyance to insert something before it. To be sure, this practice is as deserving
of careful analysis as suppression is (because it is as much an issue for recipient as suppression is): How shall we understand a speakers disruption of the production of the talk to
insert some elementthis element in particularat this juncture? What does its insertion
do to the upshot of the turn? To what possible understandings of the talk by recipient does
a speaker show orientation by inserting this element when it was not included in the previously articulated composition of the turn? Etc. But these questions are different than the
ones mobilized by suppression. Or consider the following extract from earlier in the same
conversation. Stan has asked his sister the outcome of a traffic ticket incident in which
she was involved and she has reported deciding to pay the ticket rather than contesting it.

210 Studies in language and social interaction


Joyce and Stan, 01:2030

[I guess it would ye you figured out finally

found out itd be too much ha:ssle ta take care

of it.
Joyce:--> hh I figuired (0.4) in order: I would just haf

tig- make t.wo trips down there:,

Joyce:--> Yihknow Id hafta go down there ta pay it.,
Then make an appoi:ntment (.) ta come back

[An they wouldnt give me a date, fer a month

an a half,

Stan is offering a guess about why his sister Joyce has chosen to pay a parking/traffic ticket
rather than contesting it. Joyce appears to be suppressing something when she says (at
Lines 0405, in regularized orthography), I would just have to g-, with that something
surfacing at Line 07, Id hafta go down there ta pay it. But it is clear that here again an
insertion is being done. Joyce has temporarily put this utterance on hold while inserting
make two trips down there before the go:inserting, that is, the larger point of which
the selfinterrupted utterance is a first part.
Although this is not the place for a substantial comparative treatment of suppression
and insertion, at least this much can be said here. A speaker can show that insertion
is being done by having the previously abandoned and now repeated or returning element
be implicated in the same trajectory of utterance as was initially in progress, and this is
ordinarily implemented by employing the same grammatical form and lexiconby doing
resuming as part of the practice of doing inserting. In suppression, as we see later, the
suppressed elementwhen surfacedis often virtually unrecognizably different from
what was in progress or due next grammatically and semantically rather than resumptive
of it, and is implicated in a different trajectory of utterance.
One upshot of registering the practice of same turn insertion repair, and differentiating suppression from it, is this. It may be necessary to track the subsequent development
of the talk in order to determine exactly what practice the earlier abandonment of a
TCU-in-progress (TCU stands for turn-constructional unit) was the product ofnecessary
both for the co-participant and for the professional analyst. And, for the coparticipant,
once engaged with that subsequent talk, and with its potential sequential implicativeness for what should be said next in response, the possibility of returning to the point of
abandonmentthe point of suppressionand lingering on its import is attenuated.

The surfacing of the suppressed 211

Returning now to suppression itself, let us see what the resources developed on the first
exemplar (before the excursus), and the search that they permit, yield on another specimen. In this telephone call between two college women in the mid-1970s, Hyla has called
her good friend Nancy ostensibly to talk about the arrangements for going to the theater
that evening, but a good deal of talk about other matters gets done as well. Quite early on
in the conversation there are opportunities for each to tell anything major that happened
during the day, and it is in such a telling by Nancy that the utterance we examine occurs
(at Line 24).

HG, 2:125




[My f]:face hurts,=





Oh whatde do tih you.



GODe dis () pracly killed my dumb fa:ce,=


=Why: Ho[-ow.]







With this thing I donee I wzneven looking I


dont kno::w,





Bt e jis like orpened up,





a lo*:t* yknow(v)





the pimples I ha:ve= =








It (js) hurrt so bad Hyla I wz cry:::ing,=



=Yhher khhiddi[;ng.]



[nNo:]::He really hurt me he goes


Im sorry, hh wehh hh I khho th(h) at dznt make


i(h)t a (h) n (h) y better yihknow he wz jst (0.4)


so, e-he didnt mean to be but he wz really


hurting m[e.

212 Studies in language and social interaction

Looking at Nancys turn at Lines 2327, we can note that here too an utterance is aborted,
its ending suppressed. Nancy has reported her exchange with the doctor after crying in
reaction to the pain: He apologizes (Lines 234), she reports herself to reject the apology
(Lines 245). Then (in standard orthography), He was just (0.4) so, e-he didnt mean to
be but he was really hurting me. He was just so what? In the aftermath of pain infliction
and an apology that is treated as rejectable? He was just sowhat?
I take it that this can be not only a question for us external analysts; it can be an issue
for the parties as well, the recipients of the talk. Recall that recipients parse a speakers talk
in real time, turn-so-far by turn-so-far, projecting where it is going, what it is coming to,
what it will take for it to be possibly complete. They are projecting all the time, and using
each next bit of the speakers actual talk to confirm or modify their projection of where the
talk is goingto re-project. So Hyla is not listening in a docile manner for each next bit
of Nancys turn to fall into her lap, so to speak. She is listening proactively, in the fashion
that (as we have seen from such work as that of Sacks [1992] and Lerner [1991, 1996] on
collaboratives or anticipatory completions) can often allow such a recipient in effect to say
the projected next part of the utterance for or with the current speaker. Indeed, in just such
a place as we have arrived at, one often enough finds the recipient chiming in at the point
of the hesitation and supplying the missing item (Lerner, 1991, 1996). There are grounds
then for taking the recipient to be oriented to the possible turn completion that is being suppressed and not delivered (just as recipients can be demonstrably oriented to it when suppression is not an issue). He was so There is a virtual tension built up by the recurrent
cycle of projection (by the recipient) and delivery by the speaker of a next bit of the turnso-far, a tension deprived of resolution by the suppression. We return to this theme later.6
How about mean? He was just so mean? Look then at the immediately following
talk after the suppression, and notice: He didnt mean to be but
Now this is clearly a different mean. What is suppressed in Nancys turn, if it was
mean, was a descriptor (an adjective)was the mean of nasty, cruel, and the
like. The mean of He didnt mean to be but is a verbthe mean of intend. Still,
it is a way in which the word or words that have been suppressed find a way out, so to
speak. Sometimes they are the same lexical items used in the same senseas in one I
had; sometimes they are the same lexical items used in an entirely different sense, as with
mean. And when they come out in such a radically different usage, they are very hard
to detect. In effect, they are a form of camouflage, allowing the suppressed talk to come
out, perhaps even at some level to ground the energy or tension set up by the unfulfilled
projection of the turn completion, without actually saying the suppressed thing. It (so to
speak) grounds the energy left unspent by the nonsaying of the projected, although still
not saying the suppressed, although using its word(s).7
One sort of evidence for this line is suggested by the suppressed elements reappearing in the
immediately following talk not of the suppressing speaker but of the recipient, whose close
attention to the turn-so-far, and orientation to its projected completion, are displayed by production
of the candidate suppressed element. For discussion of several exemplars of this, see Appendix B.
Consider the blizzard of tokens of the suppressed item in the following episode of mutual accommodation in arranging to take a meal together.

The surfacing of the suppressed 213

But what is so important about not saying the suppressed? In many such instances,
what is suppressed is suppressed because in some fashion it is problematic or delicate.
Such problematicalness or delicateness also commonly figures in a speakers providing an
opportunity for anticipatory or collaborative completion by the recipient (as in the work of
Sacks and Lerner cited earlier). Getting the recipient to say the delicate item allows them to
have said it together, collaboratively; it shows the recipient to also be capable of thinking
that thought and saying it.
So what is so delicate or problematic in the episode in Extract 4? Here is another piece
of the puzzle, another ingredient of the phenomenon being described here. We want to
show not only the suppression reappearing, and reappearing in the next spate of talk (composition and position); we would like to motivate or ground the suppression interactionally,
and where it is so grounded, come to terms with the camouflaged appearance that it sometimes takes. The phenomenon can still be there without heavy interactional motivation;
but then, perhaps, it is most centrally an artifact of the speech production machinery under
interactional control and shaping.8



side 1
Bu wai- dya wanna have lunch? r dinner. Witha Big Mac.
Which dya think is best fer you.
Well I dontuh::: (1.5) I- Im- Im adjustable. I think if I know now,
yknow I cn:uh:: (1.0) adjust my time accordingly,

I take Marcia to be saying Well I dontuh:::[know] with the know suppressed. But then note
the flurry starting with if I know now yknow, none of which is the know that she|suppressed
(which was the knowing of whatis best fer you).
There is a closely related phenomenon and practice that deserves brief mention and exemplification here, without full treatment. This involves a display of orientation to public cultural norms in
the very course of transgressing them; that is, even when they do not command full assent or conformity from the speaker her or himself. This can take the form of full or partial suppression. In the
former, the speaker omits articulation of the transgressing elements. Thus, in a storytelling episode
discussed in various papers (Goodwin, 1986, 1987; Schegloff, 1987, 1988, 1992), Mike is telling
about a fight at the race track the night before. Although he later shows himself willing to articulate
far more offensive language, he begins the storytelling itself this way:


Auto Discussion, 6:234

Mike:--> Evidently Keegan musta bumped im in

And the silence at Line 24 is broken by the intervention of another party to the conversation. What
is missing here is quite clearly the word ass, which figures in similar contexts later in the story
and is articulated there. But here there is a sort of obeisance paid to the cultural impropriety of the

214 Studies in language and social interaction

So what is the problem or delicate matter here? I offer this proposed analysis, or conjecture. Nancy is a young woman, in her late teens, in the transition between adolescence
and adulthood. Under the stress of the pain and the telling about it, it appears that she is
regressing a bit, reverting to a childs grasp of painit is inflicted by those who administer it because they are mean.
usage, and it is fully suppressed. In partial suppression, the improper talk is produced in lowered
voice, sotto voce, as what I am inclined to call quiet improprieties. For example, in the following phone call recorded in the mid-1960s, a woman of some years is telling her friend about a
holiday trip to Lake Tahoe in California, and the comparative virtues and drawbacks of the venue.
This includes what could be reckoned to be prejudiced comments about various so called minority
groups. Although she has little reason to believe she can be overheard, she nonetheless lowers her
voice to register an awareness of, and orientation to, the impropriety of what she is doing.

SBL, T2: C4, 3:130 (simplified)

So you go outta California into Nevahda. All of the

motels are in California, all the ga(h)mbling

places, and the big hotels, are in Nevada.

Mm hm,
And os its- it is. -hh
iinfinitely different. And I don know, Ann, but I

thinktheyre stealing a lotta Los Vegas.

I wouldnt be surprized.
The other thing that we noticed, ((very quiet)) You

know, we didnt see any Jews, you know in Las Vegas,

you [know how you see those greasy old women an=
[Uh huh
=[men, but at
=[Uh huh,
And very few Negroes, ((voice moves to low-normal))

But we saw lots of Orientals.

[Mm hm,



[You see, I think they come in from San Francisco.

Mm hm,
((voice returns to normal)) And the Orientals, you
know, are always very well dressed,
Mm hm,
And theyre tremendous gamblers.
Mm hm,
I think thats ()
So uhm uh:: they have a grand time at the crap

The surfacing of the suppressed 215

There is evidence of such a stance elsewhere in this very conversation. Hyla, for example, a little later on, reacts to a mention of the Dear Abby advice column by launching
into a story: Oh:, she said something mea::n yesterday I didn like her, and as soon as
Nancy asks her to go on, she retracts the mean as a descriptor, Well ih wasnt mea:n bt
it wz really stupid. Mean here is a kind of generic negative. But here is another instance
of the usage of the term, this time from an adolescent boy, a hotrodder in 1960s Los
Angeles talking about the relationship between teenagers and the police, which embodies
just the usage I have suggested for Nancy.




When a cop sees a hopped up car, he doesnt care

if youre goin forty five you must be doin
somethin wrong, and if he wants to be mean, he
can bust you on a thousand things.
He doesnt have to have a reason

Here again the adult who does something painful to the kid does so because he is, or wants
to be, mean.
So here is Nancy poised on the very verge of a relapse into this childish way of seeing the world: She does not treat the doctor as hurting her incidentally, as part of doing



They really at uh- its a something to see, and Im
glad I saw it, n I had a wonderful time doin it.

Formal notice is thus taken of the cultural norms applicable here, in the very course of showing a
lack of commitment to abide by them. Finally, there are gradations between full suppression and
reduced offensiveness, in which, for example, a speaker mouths the words or parts of them without actually voicing them, or begins that way and then gradually allows some voicing to set in, as
in the following characterization (by the same Mike cited earlier in this note) of the villain in the

Auto Discussion, 9:2327

D[eWa:ld is a [big burly ( (silent))ba ( (vl)) sterd=
[Jeezuz . [
[hhhh hhehhhhhhehheh,
= [jihknow,
= [Mmhm,

Here, the first syllable of bastard is mouthed silently and its remainder is voiced very quietly
(V1 is an abbreviation for Very low). What we have in the various gradations of this practice,
then, appears to involve more than simple word production apparatus per se, and yet not some
thisinteraction-specific matter of delicateness, but one way in which culture in the anthropological sense, and an orientation to cultural prescriptions as privileged points of reference, appear in

216 Studies in language and social interaction

something for her, and when he apologizes she rejects the apology as ineffective, and characterizes him asjust as she is about to say mean, she backs away. And note, what she
backs into is precisely the adult counterpart to the childish viewits not that he means to
be hurting her, but it hurts just the same. And in the very course of articulating this newer
adult part of her, she leaks outin camouflaged formthe bit of childishness she has
almost let escape.9
A rather more public problematicity and delicateness informs the next instance, taken
from an interview on National Public Radios news program Morning Edition. President
Clinton had nominated obstetrician/gynecologist Dr. Henry Foster to be Surgeon General
of the United States, and the nomination had run into trouble in its pursuit of confirmation when Dr. Foster was reported to have performed a number of abortionsthese being
treated as immoral by one segment of the press, the Congress, and the public, and as a
medical decision by another segment. Journalist Joanne Silberner developed a story on
the attitude of obstetrician/gynecologists toward doing abortions, and one part of the story
reported on Dr. Elizabeth Garrow (Lines 14), and included her recorded response to an
inquiry during an interview (Lines 513).

NPR, Morning Edition, 2/23/95

Elizabeth Garrow does one or two abortions a

as part of her practice in suburban Virginia. She

says its one of many services she offers her

Just as if a woman comes in an says, .hh Im

pregnant=I want ta have a baby, en I- try to give

her good prenatal ca:re, .hhh or .h I dont want to
be pregnant en I g:et her on the pi.:ll, f=sh=ss I
am pregnant en I dont want ta be:, .hh thats- (.)
-helping her take care of that is just another
aspect. (0.8) of- of my jo:b. I dont see it as any:
(0.2) more a less important. Its js- its a part
of it.

At Lines 1112, it seems apparent that Dr. Garrow is on the way toward summing up how
abortion presents itself to her in her practiceas just another aspect of my job (Lines
1011)by saying I dont see it as any [moral issue]. In the context of the public controversy that prompted the story and interview in the first place, this would, of course, have
been fuel on the fire. As she approaches the problematic element of her TCU, she slows
and pauses, and suppresses moral. But note how it creeps out nonetheless. In a striking
restructuring of her TCU, the any is converted into the start of the idiom any more or
Compare the relationship of this surfacing of a suppressed item with the earlier-discussed reappearance of an item held in abeyance to allow an insertion before it, as in Extracts (2) and (3) and
the discussion of them.

The surfacing of the suppressed 217

less [important]. But her articulation of this phrase, by reducing the or to a, incorporates the suppressed moral like this: any: (0.2) [more+a+1]+ess In the very swerving
to avoid the publicly problematic moral, it occupies the turn in camouflaged form and in
the very next bit of talk.
Let me end with the instance that had escaped me all these years, and that I finally saw
while examining something quite different. This comes from the conversation drawn on
for the first extract that we examineda telephone call between two young women in late
1960s New York. Ava is telling Bee about how she came to be so tired.

TG, 02:1038
Im so:: ti:yid.I js played ba:skeball

tday since the firs time since I wz a

freshmn in hi:ghsch[ool.]
[Ba::] sk(h)et=

b(h)a(h)ll? (h) [ (Whe (h) re.)

[Yeah fuh like an hour enna

[hh] Where didju play ba:sk[etbaw.]
[(The) gy]:m.
In the gy:m? [ (hh)
[Yea:h. Like grou(h)p therapy.

Yuh know [half the grou]p thet we had la:s=
[O h : : : .]hh
=term wz there- <n we [jus playing arou:nd.
Uh-fo[oling around.
Eh-yeah so, some a the guys who were bedder

yknow wen off by themselves so it wz two

girls against this one guy en hes

ta:ll.Yknow? [hh
[Mm hm?
En, I had- I wz- I couldnt stop laughin it

wz the funniest thing bt yknow you get all

sweaty upr en evrything we didn thing we

were gonna pla:y, hh en oh Im knocked out.

Nhhkhhhh! hhhh
Ripped about four nai:ls, n okhh!
Ava: ->
=Bt it wz fun-You sound very far away

Here, it appears that the I had- at Line 33 suppresses I had [fun.] Note first that the
fun surfaces a bit later in the funniest thing (Line 34), where, however, the sense of

218 Studies in language and social interaction

fun (as having a good time) is masked by the sense of funniest (as laughter prompting) given by its following couldnt stop laughing. Then note that, just before a final
quick exit line from this topic at Line 40, Ava says, But it was fun; this is the same fun
that was suppressed earlier (as compared to the funniest as the superlative of funny
that is not, quite), but she still manages to deflect it from herself to the situation as a whole:
It was fun, rather than I had fun. Note as well that the first thing to follow the initial
suppression at I had- (Line 33) was I wz- (itself cut-off in turn), and that wz returns
in the exiting line it wz fun (Line 40). Several ties connect this exit line with the earlier
site of the suppression, then. (Note by the way that Bees otherwise odd fantastic (Line
39)odd as a response to knocked out and ripped about four nailsmay invite understanding for its resonance of fantastic with fun.) So the features that have recurred in
other instances of suppression that we have examined appear to be present here as well.
But what is going on? I would like to end with an(other) illustration of an unexpected way
in which having a sense of such a phenomenon as suppression resurfacing as a real thing
can figure in our understanding of entirely different aspects of what is going on in some
episode of interaction.
The suppression and its reappearance (or the capacity of the reappearance to warrant
that there was a suppression and what it was) throws new light on something odd in the
opening of this conversation. In the opening, Bee says a curious thing after detecting in the
sound of Avas voice and in her apparent kidding around a note that properly warrants
notice by a recipient in an opening; she says, Why whatsa matter with y- ysound happy.
Now sounding happy would not ordinarily be characterized or made accountable as
something the matter with you The allusion here, I had always taken it, was to Ava being
a sad sack type, always complaining, never being content, so that the later ysound sorta
cheeerful that follows Avas denial of being happy would, even as a reduced descriptor,
be a noticeable. But this had been mere supposition; interpretation with little in the data to
support a stronger claim of analysis.
And herein the suppression we have been examiningwe see what may be such evidence: Ava cannot bring herself to say she had funI had funeven though everything
about the telling about playing basketball conveys that. This is not quite something that
motivates the suppression, but it grounds the claim of suppression in a larger canvass of the
speakers conduct, and grounds Bees treatment of Avas sounding happy as something
the matter with her in an actual display of happiness avoidance.
This is a long way from where we started (though subsequent developments can be
brought to bear on the episode with the Vice Chancellor, even if only conjecturally for lack
of a recording of the exchange). The moral of my story is this. Taking seriously, and pursuing, an observable for the purely technical object it can be, can make available a resource
whose bearing on the warrantable analysis of what is going on in interaction is by no means
purely technical in the pejorative sense ordinarily attached to that phrase.
Perhaps the larger moral is to remove the pejorative sense attached to terms such as
technical, merely technical, purely technical, and the like altogether. If something is
correct as an account of a possible event or practice or phenomenon in talk-in-interaction,
then pursuing it in its own terms promises to deliver an analytic resource whose scope of
relevance cannot be properly imagined in advance.

The surfacing of the suppressed 219

It will not be lost on readers that my title alludes to a phrase generally associated with
psychoanalytic theorizing, and with Freud in particular, the return of the repressed. Why,
then, have I danced around this memorable phrase, and settled for something that retains
both its semantic sense and its poetic alliteration, but not its literal identity? Suppression
and repression have, to my mind, slightly different connotations. Repression is deeper,
suppression shallower; repression long-lasting; suppression, at least potentially,
shorter term and transient (a government may suppress an uprising, but we do not speak
of it as suppressive; if this is a long-term, character-revealing tendency of a regime, we
speak of it as repressive); repression fundamental, suppression, at least potentially,
relatively superficial. Still, in both of them, grounds are found by actors for affirmatively
avoiding the externalization of something assertedly (by the analyst thereof) present in the
scene and informing the conduct of participants in the scenewhether these be thought
of as regimes and bodies politic, individuals and their psyches, or participants in episodes
of interaction. Here I have been dealing with suppression; to what degree the discussion
turns out to be relevant to repression-(whatever that term may be understood to denote,
given the methodological obstacles to rigorous and clear thinking in this domain) remains
to be determined.
Dealing with suppression (and repression as well, of course) involves us in nontrivial issues of interpretation and evidence, and this in two respects. First, it involves showing
what was not saidand this implicates a host of issues bound up with making negative
observations. Second, it can involve (and does in the present case) arguing that something
that was said not only was said, but is what was specifically not said earlier, and has thus
in effect escaped.
With respect to the first of these sets of issues, it may be worth reviewing in as compact
a form as possible the problem of negative observations. Strictly speaking, an indefinitely
extendable set of things was not said at any specified point in a conversation, yet only a
very limited part of that set can relevantly be noted to have been not saidby parties
to the conversation in the conversation or by external analysts about it. As noted early on
in the conversation analytic literature, one consequence of the sequencebuilding resource
dubbed the adjacency pair (two-turn sequences such as greeting-greeting, question-answer, request-grant/reject, etc.) is that when there is no response to the first part of such a
pair, one can not only generally say who was silent, even though no one has talked; one can
say what was not said/done. After a question, then, the silence is understood as a failure to
answer or a withholding of answering. Here, formulating what was not said takes the form
of a characterization of the activity or action that was not implemented, and that line of
analysis can be grounded in the relevance rules by which a first pair part constrains, shapes,
and casts an interpretive key over the moments directly following it.
The negative observation implicated in a claim of suppression, however, can be more
detailed and specific than this. In the episodes examined in this chapter, what is claimed
is that some word(s) or phrase(s) or topically specific fragment of talksome sayable in
particularhas been specifically withheld from articulation, has been suppressed. The
relevance rules that underlie such a claim therefore have to be more fine-grained than those
underlying characterizations of missing responses to first-pair parts.

220 Studies in language and social interaction

With respect to the second set of issues, one feature of the type of understanding of
interaction (and social life more generally) sought by conversation analysis and kindred
pursuits in the social and human sciences is that analytic characterizations of actors conduct be grounded in, and warranted by, the participants own demonstrable orientations
to the setting, context, and import of what is going on. In this enterprise, one eschews
analytical claims warranted only by the theory one brings to the data, whatever the force
of the statistical or experimental or interpretive data marshaled on their behalf. Whatever
categories of action the analysts theory has generated, if we cannot show the participants
to be oriented to the conduct in its course by reference to such categories, to such an
understanding of the import of their actions, then that line of analysis is not tenable. But are
we then to argue about talk that has slipped outas is implied by the surfacing of the
suppressedthat this captures the orientation of the parties? The import of the conduct
for them? That is what is involved in arguing that something that was said not only was
said, but is what was specifically not said earlier, and has thus in effect escaped.
These are some of the more general issues mobilized by the empirical occurrences with
which this chapter engages. It would, of course, be presumptuous to claim that they have
been solved. But I hope to have indicated one way in which we can approach taking them
seriously and beginning to deal with them. Their relevance may extend past conversation
analytic work itself.
First, prepared for, and presented at, the annual meetings of the National Communication Association, Chicago, November 1997. Robert Hopper called to my attention possible
convergences with discussions in Jefferson 1996, a matter taken up in Footnote 2. The
present version of the chapter was prepared while I was the grateful beneficiary of a John
Simon Guggenheim Memorial 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.
This article is co-published in German in Volume 1, No. 4 of the new journal Psychotherapie und Sozialwissenschaft. Readers coming to the article from a background
in psychiatry or psychoanalytically oriented psychology will find in the Postscript some
reflections on the relationship between the sort of conversation-analytic work presented
here and those traditions of inquiryas reflected in the title, and may wish to consult it
first, or in due course.
Appendix A

Suppressed Elements Surface in Recipients Utterance

This appendix presents brief accounts of two episodes in which suppressed elements reappear in the immediately following talk not of the suppressing speaker but of the recipient,
whose close attention to the turn-so-far, and orientation to its projected completion, are
displayed by production of the candidate suppressed element. Consider first the following
opening of a telephone conversation.

The surfacing of the suppressed 221


D&B, 1:117
Bernie: Hllo, Dina?
Bernie: Hi, howre you.
I CAlledju las night.
Bernie: You di:d,
Bernie: Wha ti.:me.
Uh::: about seven uhclock, or was it e- tch! Oh

I- I dont remember b- but I calledju.

Bernie: Yeah.
N- nobuddy was home.
Bernie: hhhh[hhhh
[Gee I was just th- n- thats very funny.

How are you.

Bernie: Okay.
Thats good.
BerTch! hhhh I think I was home last night.

Almost certainly Dina was saying at Line 14, Gee I was just th[inking about you], something that is often accompanied by thats [very] fUnny (at the beginning of an unanticipated phone call). Here it is suppressed (perhaps because it is a further display on her part
of interest in him which may not be reciprocated or appropriate). Note then that it pops up
three turns later, in the recipients mouth (I think I was home last night.). Two observations may be made about this. First, regarding the non-immediacy of the position: This
is the first turn of Bernies following the suppression, which is not sequentially constrained
by Dinas prior turn. Second, hurdles are overcome for this utterance to be produced here.
A reciprocal howareyou question is in order, as Bernies first howaryou at Line 4 was
by-passed by reporting the effort to call him, and Dinas howaryou was marked by its
stress on the second syllable as a first inquiry of a reciprocal pair (Schegloff, 1986).
Where the reciprocal inquiry was due, Bernie does not do it. In its place he replies to the
Nobody was home of Line 12 with what is in effect a disagreement or rejection or correction, its contrariness marked by the epistemic downgrade of the I think, which was the
suppressed element of Dinas earlier turn.
The second exemplar occurs early in the conversation between Joyce and Stan examined earlier in the discussion of insertion (Extract 3), and indeed is the larger sequence
in which that insertion occurred.

Joyce and Stan, 01:0902:12

hh First of all howd that thing turn out with
the ticket. Dju: anything happen?


222 Studies in language and social interaction


Stan: -->
Joyce :
Joyce :

Joyce :


Joyce :

Joyce :

Joyce :

Joyce :
Joyce :

Joyce :

Joyce :
Stan: -->

Joyce :

Oh, I just decided ta pay it.

Decide (d) ta pay how much was it
Fifteen fifty.
Fifteen fifty?
Mm hm,
Bitch. Bitch.
I(h) kn(h)owh[h
[I guess it would ye you figured out
finally found out itd be too much ha:ssle ta
take care of it .
hh I figu:red (0.4) in order: I would just haf
tig- make two trips down there:,
Yihknow Id hafta go down there ta pay it,
Then make an appoi: ntment (.) ta come back there
[An they wouldnt give me a date, fer a month
an a half,
AnI figu:red (0.9) the case [just wu
[(Plus) ya gotta
yih gotta put down the money. ahead a time.
Yeah t [hey give it back to you. l:ater.=
[ (Yeah the)
= [ (see an)
= [The way I beat mine it was a pa:rking ticket.
yihknow, so I was able ta go to ta night court .
(wu) then beat the ten dollar ticket.
hh Yihknow just the principle a thing that
bugged me.
=U: :m (1.4) tch! (.) So wudja do pay it through
the auto club

The surfacing of the suppressed 223

On the theme that the suppressed item may show up in the immediately following talk of
recipient, note that Stan surely appears to suppress something at Line 02: Djuianything
happen? He is starting to ask an agentive question: Did you: [pay it] The sound stretch
on the you shows him thinking the better of it, and he shifts to a non-agentive form of
the inquiry, one that does not introduce the relevance of any particular action on Joyces
part (which she might have to report having failed to do, e.g.). Then note that the suppressed item shows up in the next turn by the recipient, Oh, I just decided ta pay it. and
is then repeated by Stan (Line 05) as a form of registering the response (Schegloff, 1997).
Once out in the open, Stan uses it again (at Line 41), as he brings the the topic/sequence
to a close. (For further discussion related to this general topic, see also Jefferson, 1974;
Schegloff, 1979)
Clift, R. (1999). Grammar in interaction: The case ofactually. Essex Research Reports in Linguistics 26, University of Essex.
Clift, R. (2001). Meaning in interaction: The case of actually. Language 77, 24591.
Goodwin, C. (1986). Audience diversity, participation and interpretation. Text, 6, 283316.
Goodwin, C. (1987). Unilateral departure. In G. Button & J.R.E.Lee (Eds.), Talk and social organisation (pp. 206216). Clevedon, England: Multilingual Matters.
Jefferson, G. (1974). Error correction as an interactional resource. Language in Society, 2, 181199.
Jefferson, G. (1996). On the poetics of ordinary talk. Text and Performance Quarterly, 16, 161.
Lerner, G.H. (1991). On the syntax of sentences-in-progress. Language in Society, 20, 441458.
Lerner, G.H. (1996). On the semi-permeable character of grammatical units in conversation: Conditional entry into the turn space of another speaker. In E. Ochs, E.A. Schegloff, &
S.A.Thompson (Eds.), Interaction and Grammar (pp. 238276). Cambridge, England: Cambridge University Press.
Ochs, E., Schegloff, E.A., & Thompson, S. (1996). Interaction and Grammar. Cambridge, England:
Cambridge University Press.
Sacks, H. (1992). Lectures on conversation. (2 vols.). G.Jefferson, Ed., (with introductions by
E.A.Schegloff). Oxford, England: Blackwell.
Schegloff, E.A. (1979). The relevance of repair for syntax-for-conversation. In T.Givon (Ed.), Syntax and semantics 12: Discourse and syntax (pp. 261288). New York: Academic Press.
Schegloff, E.A. (1986). The routine as achievement. Human Studies, 9, 111151.
Schegloff, E.A. (1987). Analyzing single episodes of interaction: An exercise in conversation analysis. Social Psychology Quarterly, 50, 101114.
Schegloff, E.A. (1988). Description in the social sciences I: Talk-in-interaction. IPRA Papers in
Pragmatics, 2, 124.
Schegloff, E.A. (1992). In another context. In A.Duranti & C.Goodwin (Eds.), Rethinking context:
Language as an interactive phenomenon (pp. 193227). Cambridge, England: Cambridge University Press.
Schegloff, E.A. (1997). Practices and actions: Boundary cases of other-initiated repair. Discourse
Processes, 23, 499545.
Schegloff, E.A. (1998). Word repeat as a practice for ending. Paper presented at the 84th Annual
Meeting of the National Communication Association, New York, NY, November, 1998.

Sex, Laughter, and Audiotape: On Invoking Features
of Context to Explain Laughter in Interaction
Phillip J.Glenn
Emerson College
Laypersons and analysts sometimes invoke gender as an explanatory variable that, it is
presumed, shapes or even determines some feature of interaction. In its simplest formulation, this variable shows up in studies devoted to identifying differences in how women and
men talk, move, listen, and so on. Underlying such studies is an assumption that particular
features of speech or interaction reflect and constitute gender differences. For example,
there are claims that women use more tag questions, disclaimers, and hedges; and that
men interrupt women more than women do men. Tannen (1990) claimed that women give
more audible and visible feedback when listening than men do. Wood (1996) summarized
research findings indicating a tendency for women to do more conversational maintenance work, including behaviors to signal interest and involvement (p. 157). The differences then are found in particular features but also in clusters of these features adding up
to activities, such as maintenance, affiliation, or support.
Researchers offer various conceptual explanations for such differences. Some argue
that these may not reflect behavioral differences as much perceptual differences: that people perceive women and men as speaking differently. Initially researchers were willing to
explain such differences in terms of lesser confidence or competence on the part of women.
Recent studies tend to treat such claims as problematic (see West, 1995), and suggest that
differences may in fact show women as being highly competent, perhaps more so than
men. Another explanation lies in asserted power differences: that speech features reflect
varying degrees of relative power, and that differences between men and women reflect
more fundamentally different power currencies. Others account for variations as reflecting
different primary styles of communication. Pushed to an extreme, style difference arguments pose women and men as coming from different cultures or even different planets
(e.g., Gray, 1992; Tannen, 1990).
Laughter may be one feature of discourse that reflects and constitutes gender differences.
There are shared cultural assumptions (perhaps based in stereotypes) that men produce
more laughable, humorous behavior, and that women do more laughing in response to men,
than the converse. Laughter can do such conversational work as displaying involvement
or interest and achieving maintenance; to the extent that such work is more common for
women, this may suggest women do more laughing in the presence of, and responsive to,
men. In an observational study Robert Provine found that most instances of conversational
laughter between two persons occurred when men were talking and women were listening,
and the least took place when women were talking and men were listening (reported in
Kluger, 1994).

Sex, laughter, and audiotape 225

Two recent studies make use of naturalistic data to investigate gender differences in
conversational laughter. Jefferson (1994) explored the possibility that in male-female
interaction, if the male laughed, the female would join in laughing; if the female laughed,
the male would not join in laughing (p. 1). From analysis of a collection of instances of
laughter in interactions of women with men, she found tentative support for some gender
difference trends. However, her claim is not a straightforward one that women laugh more
than do men. Rather, it is that laughing (or not laughing) may, depending on sequential
environment, display receptiveness or resistance to what the other speaker is doing.
Her gender difference argument is that men more often display resistance and women tend
more to display receptiveness. Whether laughing or withholding laughter in any particular
instance displays resistance or receptiveness is shaped in part by the immediate sequential
environment. Thus the organization of laughter seems subsumed under the organization
of a more fundamental set of activities, displaying receptiveness or resistance. However,
Jefferson cautioned against making too much of these tentative claims; emphasizing the
cartoonish nature of the crude female-male binary split, she referred to participants in her
data as Tarzans and Janes.
Glenn, Hoffman, and Hopper (1996) set out to test Jeffersons preliminary claims in a
larger corpus of laugh instances. In general, their counts did not match the trends Jefferson
identified. When they separated data into two kinds of interactions, courtship-relevant
and noncourtship, some numerical gender difference trends emerged. However, the
increasing number of cells made for such small sample size that results remain inconclusive. Outside of courtship situations, men more often showed appreciation for womens
laughables-with-laugh-invitations by laughing along than women did for men. This contradicts Jeffersons receptiveness-resistance theory. Within courtship, however, instances of
laughter produced responsive to anothers laugh more closely supported patterns described
by Jefferson. When one speaker offered a positive laughable without laughing and the
other showed appreciation for it by laughing, a female speaker of the laughable would
provide second laugh, but a male speaker usually would not. In courtship-relevant interactions, women were much more likely than men to produce negative laughables at their
own expense and offer first laugh. This suggests another way in which laughter may mark
gender differences: that women may be more likely than men to laugh as an accompaniment to self-deprecation.
Research questions driving such studies begin with the presumption that communicative differences do exist, or at least may exist, and that the binary, biologically based categorization scheme of women and men is an appropriate way to conceptualize this
variable. Claims of gender difference notice trends across numbers of cases. Empirical
findings reflect this in proquantifier terms like more often or less likely. However, we
do not live our communicative lives in the aggregate. We live them one moment at a time,
or, in researchers terms, one instance at a time. If people communicate differently from
each other, and if they do so systematically in some way linked to biological sex or gender
role, then our task as analysts is to examine the means by which people accomplish such
differences in single instances.
Increasingly, scholars are calling for more context-sensitive treatments of gender as
socially constituted (see Wodak, 1997). Garfinkel (1967) noted the omnirelevance
(p. 118) of sexual status in everyday life in that humans continually display features

226 Studies in language and social interaction

readable as gendered. However, this does not mean that people orient to gender equally
at all times. Many individual attributes or features of context are potentially available as
participant resources in the ongoing tasks of organizing and making sense of conduct.
For analysts too, gender is but one of many features available to draw on for explanations
of communicative phenomena. How can we develop and support a claim for gendered
communication being part of a particular communicative moment?
This may be understood as a question of context (see Tracy, 1998). Making a distinction
between text and context helps us examine words, actions, utterances, sequences, and so
on, somewhat apart from features of the individuals, setting, surrounding talk, relationship,
culture, and so forth, that shape and help explain features of the text. In the present study,
I treat context as emergent, fluid, and locally occasioned by participants in interaction.
Consistent with this perspective (one advocated by Schegloff, 1987, among others), we
may make the strongest empirical claims about the relevance of some feature of context
(such as gender) in explaining communicative phenomena when evidence exists in the data
that participants themselves orient to that feature as relevant. This intrinsic-to-messages
approach (Hopper, 1992) helps avoid the danger of the researcher imposing a priori theories that may unduly limit or mislead analysis.
The sense of contextual features being located in the moment is different in an intrinsicto-messages approach. If a man interrupts a woman, or a woman uses a tag question and
a man does not, there is not an a priori assumption that such differences arise because the
actor is a woman or a man. Rather, the initial interest is less in individual behavior than in
joint construction of actions, and less in imposing external explanatory variables than in
trying to characterize the procedures by which people do whatever it is that they do. Thus,
the moment can be investigated, not as a site for the inevitable realization of gender or
some other feature(s) of context, but as a site for creativity, change, and constitution. The
analytic focus, then, is on details of talk and action as patterned ways of accomplishing
activities in interaction. There is a suspension of theoretical explanations in order to retain
as long as possible analytic focus on what is being done and how it is being done.
The following analysis is aimed at investigating the possibility that people orient to gender in the organization of conversational laughter. For this purpose I selected an instance of
talk in which gender (and sex) clearly become relevant for participants, in close proximity
to laughter. In other words, I begin with a hunch that something gendered is happening
with laughter here. I argue, with evidence from this instance, that acoustic and sequential
features of laughter can display participant orientation to gender. Thus, this analysis stands
as an example of how to demonstrate empirically the relevance of gender as a feature of
The following instance comes from the radio program Car Talk, broadcast live on National
Public Radio affiliate stations. In the show, brothers Tom and Ray Magliozzi, who run an
automobile repair shop in the Boston area, dispense advice to people calling in with carrelated problems. In addition to giving advice, the brothers joke and play, often punctuating
the talk with laughter. The show combines face-to-face interaction between Tom and Ray,
telephone interaction with the caller, and broadcasting to an overhearing radio audience.

Sex, laughter, and audiotape 227

The interactions with callers typically reflect a structure common to other advice-based talk
shows: opening, problem formulating, advising, and closing (Crow, 1986).
The fragment under consideration is shown in its entirety as follows. It comes from the
beginning of a phone call, the second one broadcast on this particular day:
Car Talk, National Public Radio, 30 March 1997 Tom and Ray Magliozzi and Caller
One eight hundred (.) three three two (.)

nine two eight seven=Hello youre on Car


Hi this is Chandler? Im calling from

From Denver=
10 Chand:
11 Tom:
12 Chand:
13 Ray:
Thats an unusual (.) first name?
14 Chand:
Well (.) I know Im not supposed to tell you
15 Ray:
for- for a woman
16 Chand:
my last name my last names even wilder.
18 Chand:
19 Tom:
Even wilder
20 Chand:
21 Tom:
=Ooh! Chandlers even wilder than the last
22 Ray:
23 Chand:
24 Tom:
26 Tom:
HAH HAH HAH HAH HAH (.) hh huh huh
27 Chand:
:-huh huh u h h h h [We:ll.
28 Ray:
Theres a
29 Ray:
theres a hyphen in there?
30 Chand:
31 Tom:
No its just a sentence
32 Chand:
Its just a sentence? Thats right
33 Tom:
[Ha ha ha ha ha ha ha ha hh
34 Ray:
35 Chand:
Anyway, I: have (.) I have this problem.
I have a Ford Escort (.) wagon (.)

Of particular interest for this article are Lines 2127. In Lines 2122, Tom playfully assesses
the caller as even wilder in contrast to the last girl he went out with. As is shown later,

228 Studies in language and social interaction

in this utterance he treats the caller as female, as someone he might go out with, and as
wild with possibly sexual implications. All three participants laugh, although her laughter displays a different, less affiliative stance toward the laughable than those of the brothers. This appears to be a moment of highly gendered, laughable, and laugh-inducing talk.
Are the laughs themselves contributing to gender marking? Do they display orientation to
gender? Before addressing this question, lets back up and trace how the participants get
to this moment.
The caller identifies herself as Chandler from Denver. This use of first-name-only plus
location for self-identification is standard practice on the show.

Chand: Hi this is Chandler? Im calling



The name gets immediate and marked attention. Ray repeats it with increased melody and
emphasis; Tom does the same, shifting the pronunciation of the initial affricate ch to sh and
stretching it.




Hi, this is Chandler?, Im calling


Repeats can function as next-turn repair initiators (Schegloff, Jefferson, & Sacks, 1977). At
the least, they retrieve some prior item and make it available for further talk or action (they
also divert, at least momentarily, from moving toward the purpose of the call; in Car Talk
such playful diversions are common). Consistent with the structure of repair sequences,
the repeat returns the floor to other to confirm or amend the repeated item. In overlap with
Toms second repeat, Chandler confirms that this is her name.
Ray now repeats the second half of Chandlers self-identification, from Denver. This
repeat has a marked melody paralleling that which he used in repeating her name a moment
earlier. It is a poetic moment: the melody echo emphasizes the rhyming of Chandler and
Denver. This too fitting the structure of a next-turn repair initiator, it returns the floor to
her, and she confirms Denver as correct.


From Denver=

Tom repeats the name again (line 11), once more with marked, melodic intonation. She
again confirms it.

Sex, laughter, and audiotape 229




That its been repeated multiple times and already confirmed provides evidence that this
is not a problem of hearing or understanding on their part. Rather, the repetitions open up
possibilities for topicalizing her name as something to talk about, and/or for keying a playful treatment of it (on repetitions role in keying play, see Hopper & Glenn, 1994).
Ray assesses the name as unusual (Line 13). This assessment, with questioning intonation, would make relevant further talk about her first name, perhaps including an account
for it. Chandler begins to speak. In overlap, Ray (Line 15) produces a delayed completion
(Lerner, 1989) of his prior turn:




-Thats an unusual (.) first name?

Well (.) I know Im not supposed to tell
for- for a woman
my last name my last names even wilder.

This added prepositional phrase modifies his assessment such that the name Chandler is
unusual, not for all people, but for a woman. By this he introduces gender explicitly into
the talk, for the first time in this call.
Her unfolding turn does not attend explicitly to the delayed completion. Instead, Chandler shifts to discussing her last name. She does not actually produce it, but states the programs rule prohibiting use of last names. She assesses this name comparatively as even
wilder than her first. Through this turn she continues the pattern of playful assessments of
her name yet shifts attention from her first name to her last.
After she says her last name is even wilder, there is a pause. Several possibilities are
relevant here. They could talk more about her first name, although she now has shifted
focus to her last name. They could talk about her wild but unstated last name, although
such talk might be limited because Tom and Ray do not have the name itself as a present
resource. They could go on with the business of the call. Two of these three possibilities
get pursued almost simultaneously. Chandler speaks, and her Anyway displays willingness to close this section of talk and move on. In overlap, Tom repeats her preceding phrase
even wilder.




my last name my last names even

Even wilder

Toms repeat/repair initiator picks up on and furthers the topical shift she had made from
her first name to her surname. In contrast to her Anyway, his repeat displays willingness
to delay proceeding to the business of the call. She confirms his repeat.

230 Studies in language and social interaction

Now comes Toms joke. Its prefaced by an exclamation of delight or excitement.
21 Tom:

=Ooh! Chandlers even wilder than the

girl I went out with


He repeats the assessment even wilder but applies it to her, not to her last name as she
had done. To retain the contrastive form of the adverb-adjective assessing pair, Tom must
provide something or someone against which to compare Chandler. He does so by inventing the last girl he went out with. The jibe is clever: he uses her words to assess her playfully by invoking a nonexistent dating/romantic relationship between them and implying
that within it she is wild.
In this utterance, it is not just gender that creeps into talk (Hopper & LeBaron, 1998);
it is also sexthe act, not the biological category. Toms use of girl in the jest about
her being even wilder suggests a younger orientation and perhaps playfulness on his
part (contrast to Rays prior use of the term woman). It seems fitted as category to the
activity go out with (see Sacks, 1992, Vol. 1, pp. 515516, 594597, on category-bound
activities). More specific than simply the broad categories female and male, the talk
now invokes, albeit jokingly, participant identities as heterosexual woman and man who
represent, for each other, potentially dateable partners. For such persons, the assessments
wild or even wilder may carry sexual meanings.
Now comes the laughter. Ray begins to laugh immediately after the words last girl,
displaying recognition of the joke in progress. He produces a lengthy and mirthful stream
of laughter.



=Ooh! Chandlers even wilder than the last

girl I went out with
Hu hu hu ha ha ha ha ha ha

Chandler starts laughing at completion of Toms utterance and following several syllables of Rays laugh. She produces two initial closed-mouth syllables then six open-mouth



=Oooh! Chandlers even wilder than the last

girl I went out with
Hu hu hu ha ha ha ha ha ha
Hhhh hhh huh huh huh huh
huh huh uhhhh

Toms is the biggest laugh of all, loud and hearty (Lines 2526).



=Ooh! Chandlers even wilder than the last

girl I went out with
Hu hu hu ha ha ha ha ha ha
Hhhh hhh huh huh huh huh

Sex, laughter, and audiotape 231



HAH HAH HAH HAH HAH (.) hh huh huh
= [huh huh u h h h h [We: ll.

Rays laughter ceases. Chandler produces an audible inbreath (Line 27), and Tom pauses
briefly, produces an inbreath, then laughs a bit more (Line 26). More laugh particles
following inbreath may show willingness to keep laughing and constitute an invitation to renew and extend shared laughter. At that moment, however, Chandler resumes
nonlaughing talk, and Tom ceases laughing (ends of lines 2627).


huh huh

The word well is spoken with a tone of mock indignation. Placed here, following Toms
jest about her wildness plus shared laughter, it shows some degree of resistance (albeit
playful) to what has just gone on. Perhaps sensitive to this, the brothers abandon the laughable plus shared laughter to resume speaking. Ray suggests an implicit pun, for his reference to hyphen invites a hearing that Even-Wilder literally is her last name.


huh huh
Theres a
theres a hyphen in there?

Rays grammatical jest provides a way for them to continue playing with her name without
continuing the explicitly gendered, sexual talk (although gender still may be remotely relevant, in that hyphenating surnames is a practice more often characteristic of women than
of men, and may invoke marital status). Tom laughs, but neither of the other participants
does. Chandler then moves on to the business of the call:



Theres a- theres a hyphen in there?

No its just a sentence
Its just a sentence? Thats right
Ha ha ha ha ha ha ha ha hh
Anyway, I: have (.) I have this problem.
I have a Ford Escort (.) wagon (.)

In this passage, the callers name serves as a resource for play. Tom uses Chandlers name
and her own words to construct a sexual jest about her. The brothers laughs align with
each other and appreciate the jest, which is done (however innocuously) at her expense.
By laughing at the sexual jest, Chandler displays some willingness to play along (cf. Drew,
1987, concerning the range of responses to teases). By laughing less enthusiastically and

232 Studies in language and social interaction

responding with Well, she displays some resistance to the jest. By resuming talk, she
helps move them away from the sexual reference. At the first sign of lack of enthusiasm
from Chandler, Tom and Ray immediately move away from sexual innuendo. Talk continues on topics for which gender/sex seem not to be foregrounded: a hyphen in the sentence,
a Ford Escort wagon, and more.
Participants mark the relevance of sex categories and sexuality as features of context.
They do so in the service of word play and shared laughter. The laughs themselves reflect
and constitute different orientations to this invoking of context. Laughs orient to context
through their acoustic features, length, and sequential placement, all of which contribute to
marking laughters footing in relation to the laughable, the participants, and the situation.
The instance here turns out to be consistent with Jeffersons (1994) preliminary claim that,
in laughing along, Janes interacting with Tarzans exhibit receptiveness (p. 17). That is,
Chandlers laughing shows her to be receptive to what the brothers are about. This Jane
may not be thrilled about what happens, but she is willing to laugh along while at the same
timethrough features of her laughterdistancing herself somewhat from the stance of
the two Tarzans.
Participants sometimes foreground gender issues explicitly as topic of talk. Perhaps
more subtly, they sometimes orient to gender through features of the sequential organization of interactions. The choice to laugh or not to laugh provides partial clues for hearers
and analysts concerning the work that laughter may be doing. Placement and production
features of laughs help show laughter to be affiliating, disaffiliating, or partially affiliating
with some evident resistance. To the extent that these displays are about gendered issues,
they allow participants to orient to gender and thereby, allow analysts access to the social
constitution of gender in discourse. This analysis, then, offers a method for demonstrating
empirically the relevance of gender to interaction. It is intended as an alternative to beginning with a priori assumptions that gender is always equally relevant for participants. It is
also intended as an alternative to assuming that the study of gender equates to the study of
difference. The laughing that women and men do may not always differ from each other,
but laughter stands as one of a host of phenomena through which people engender sexual
identities. Finally, all pragmatics researchers must deal with how and under what circumstances to invoke features of context to explain discourse. This argument shows one way
to locate context in talk. Providing evidence in details of interaction that participants are
orienting to some feature of context (such as gender) provides an empirical warrant for
invoking that feature in an explanatory fashion.
Crow, B.K. (1986). Conversational pragmatics in television talk: The discourse of Good Sex.
Media, Culture, and Society, 8, 457484.
Drew, P. (1987). Po-faced receipts of teases. Linguistics, 25, 219253.
Garfinkel, H (1967). Studies in ethnomethodology. Englewood Cliffs, NJ: Prentice-Hall.
Glenn, P.J., Hoffman, E., & Hopper, R. (1996, March). Woman, laughter, man: Gender and the
sequential organization of laughter. Paper presented at the American Association of Applied
Linguistics Convention, Chicago.
Gray, J. (1992). Men are from Mars, Women are from Venus: A practical guide for improving communication and getting what you want in relationships. New York: HarperCollins.

Sex, laughter, and audiotape 233

Hopper, R. (1992). Telephone conversation. Bloomington: Indiana University Press.
Hopper, R., & Glenn, P. (1994). Repetition and play in conversation. In B. Johnstone (Ed.), Repetition in discourse: Interdisciplinary perspectives (Vol. II,) (pp. 2940). Norwood, NJ: Ablex.
Hopper, R., & LeBaron, C. (1998). How gender creeps into talk. Research on Language and Social
Interaction, 31, 5974.
Jefferson, G. (1994). A note on laughter in Male-Female interaction. Unpublished manuscript.
Kluger, J. (1994, January). Survival of the funniest. Discover, pp. 1620.
Lerner, G. (1989). Notes on overlap management in conversation: The case of delayed completion.
Western Journal of Speech Communication, 53, 167177.
Sacks, H. (1992). Lectures on conversation (Vol. 1) (G. Jefferson, Ed.). Cambridge: Blackwell.
Schegloff, E., Jefferson, G., Sacks, H. (1977). The preference for self-repair in the organization of
repair in conversation. Language, 53, 361382.
Schegloff, E.A. (1987). Between macro and micro: Contexts and other connections. In
J.C.Alexander, B.Gieson, R.Munch, & N.J.Smelser (Eds.), The macro-micro link. Berkeley and
Los Angeles: University of California Press.
Tannen, D. (1990). You just dont understand: Women and men in conversation. New York: William
Tracy, K. (1998). Analyzing context: Framing the discussion. Research on Language and Social
Interaction, 31, 128.
West, C. (1995). Womens competence in conversation. Discourse & Society, 6, 107131.
Wodak, R. (1997). Introduction: Some important issues in the research of gender and discourse. In
R. Wodak (Ed.), Gender and discourse (pp 120). London: Sage.
Wood, J.T. (1996). She says/he says: Communication, caring, and conflict in heterosexual relationships. In J.T.Wood (Ed.), Gendered relationships (pp. 149162). Mountain View, CA: Mayfield.

Gender Differences in Telephone Conversations
Hanneke Houtkoop-Steenstra
Utrecht Institute of Linguistics
When we read Harvey Sacks very first Lecture Notes of 1964, we see that conversation analysis (CA) has developed from Sacks observation of how North Americans open
their telephone conversations. Phone openings have been studied ever since, especially by
Emanuel Schegloff and Robert Hopper.
Schegloff (1979) studied North American telephone openings and found a pattern of
four canonical sequences: (a) summons/answer sequence, (b) identification/recognition
sequence, (c) greeting sequence, (d) how are you sequence. For example:

Hopper et al. 1990/91:370

a. 1

b. 3
Hello Ida?
c. 5
Hi.=This is Carla
Hi Carla.
d. 7
How are you.
10 R.
=How about you.
11 C.
Fine. Don wants to know

Schegloff showed how the conversationalists establish the participants identification.

When somebody hears the ringing of the telephone, he or she will answer this summons
by providing a voice sample (hi or hello) to be recognized by the caller. If the caller
recognized the answerer from the voice sample in the answering turn, then the caller should
show (or claim) such recognition in the next turn, the second in the call. Subsequent to the
callers recognition of the answerer, the answerer displays recognition of the caller. For in
example, in Fragment 2:
(2) Schegloff 1986:127

turn answer:
A. (Hell)o,
turn recogniC. Hello Missiz Feldman,
turn recogniA. Hi Bonnie.

Gender differences in telephone conversations 235

This is different from how people in the Netherlands deal with the tasks of identification
and recognition. Rather than doing other-recognition, Dutch telephone conversationalists
self-identify. First the answerer mentions his or her name, and then the caller does. This is
shown in Fragments 3 and 4:
(3) Houtkoop-Steenstra 1991

Met Mies Habots=

With Mies Habots=
=Da:g, met Anneke de Groot.

=Hi:, with Anneke de Groot.
(4) Houtkoop-Steenstra 1998

Goedenvond. Met Francien de Veer.

Good Evening. With Francien de Veer.
Goeienavond. U spreekt met Annette

Bos van Marktonderzoeksbureau

((NAAM)) uit Amsterdam

Good evening. Youre speaking with

Annette Bos from Market Research

((NAME)) in Amsterdam.

Before Dutch answerers mention their names, they may provide a greeting token, usually
Good Morning/Evening as in Fragment 4. This greeting token then tends to be returned
by the caller in his or her next turn. A more common way of answering the telephone is
to say (Hello), with So-and -So, as happens in Fragment 3. Note that the with is the
remains of youre speaking with
In 1991 I reported a study of 87 Dutch phone openings (HoutkoopSteenstra, 1991) that
were recorded in the later 1980s by Paul ten Have and myself. In 78 cases the answerers
provided a self-identification, and in 5 cases they provided a voice sample. The rest of the
4 cases were referred to as variant cases, for example, answerer picking up the phone by
saying just a second please. In Table 16.1, I refer to this set of data as late 1980s data.
TABLE 16.1: Late 1980s Data

Answerer provides self-identification

Answerer provides voice sample
Variant cases, e.g., switch calls





236 Studies in language and social interaction

It was on the basis of these data that I came to the conclusion that in the Netherlands we
find a strong preference for answerers explicit selfidentification, whereas in the United
States, we typically find other-recognition.1
Such a difference then suggests that there are differences between speech communities
with respect to how people answer the phone. Robert Hopper and his students, as well as
various other authors, have shown differences with respect to how members of different
speech communities routinely answer the telephone in domestic contexts. In the literature
we find three variations: providing a voice sample, explicit self-identification, and mentioning the households telephone number.
Answering the phone by using some form of voice sample was found in Taiwan (Hopper
& Chen, 1996), in Lebanon2 and England (Hopper & Doany, 1989), in Northern Mexico,
Spain, and Paraguay (Hopper, Doany, & Drummond, 1990/1991), and in Greece (Sifianou, 1989). English answerers say either Hello or give their telephone number (Sifianou
1989). Based on her own intuitions as a member of French society, Godard (1977) claimed
the French use voice sample too. However, Hopper & Doany (1989) did not find evidence
for this in a follow-up study.
Lindstrm (1994) showed that, in Sweden, explicit self-identification is the most common answer to a summons, followed by the phone number. Although Hello is used in
Lindstrms data, it is as infrequent and marked as in the Dutch data.3 Adler (1993) reported
a similar procedure for Germany.4
The example that follows shows how two speakers from different cultures, here North America
and The Netherlands, may stick to their own opening procedure. The opening has been transcribed
from memory immediately after the call took place.


Met Hanneke Houtkoop
Hanneke. ((with an American
Its Doug.
oh. Hi: Doug.

In fact, Hopper and Doany (1989) spoke of Arabic openings. This suggests that their findings
also apply to other Arabic-speaking countries, such as Morocco, Egypt, and the like. Hopper et al.
(1990/1991) mentioned the possible effect of language on opening sequences. It seems plausible
to expect possible differences to occur in speech communities rather than in languages. It is possible that the ways in which members of speech communities answer the phone may have been
influenced by a countrys colonial history. As Hopper and Doany pointed out for the former French
colony Lebanon: the use of allo as a response type in Arabic calls in Lebanon is the result of
linguistic borrowing of the term (pp. 165166).
Hello was found in 5 of the 100 transcribed openings. One of these appeared in a call that
seemed to involve some technical problem. Three of the four remaining instances were produced by
the same person.
As telephone technology is changing, it should be stressed that all studies mentioned apply to calls
to pre-modern telephone sets, that is, no cell phones and no telephone sets that have the provision
of displaying the callers telephone number. Especially in countries in which answerers selfidentification is the norm, it is possible that such technical devices may change the way in which people

Gender differences in telephone conversations 237

There is an ongoing debate in CA on the question as to whether or not these differences
in how members of certain speech communities routinely answer the phone reflect a cultural difference. More generally, how universal is Schegloffs description of the four canonical sequences? Hopper et al. (1990/1991) seemed to suggest that the difference between
answerers providing a voice sample versus explicitly self-identifying falls within the scope
of withincultural variance in the details of telephone openings as Schegloff (1979, 1986)
and Hopper (1989) found in North America. They considered the systematic practice of
answerers and callers self-identifying as fitting within Schegloffs model: In fact, Schegloffs (1979) discussion of identification and recognition includes virtually every format
that have been argued as being unique to Greece, France or Hollandand all from North
American data! (p. 378). If we read Schegloffs work closely, especially his unpublished
dissertation (Schegloff 1967), it is clear that he saw the voice sample Hello as the typical answer to the summons. He wrote: Hello is the unmarked form of answer to the
telephone; whereas yeah or Hi may type a prospective conversation as expected, and
a self-identification form of answer, such as Police Desk may type it as business (p.
43). For the Dutch situation, we may state that selfidentification is the typical, unmarked
form, and all other forms of answering the phone (e.g., yes, hello, and hello?) are
marked forms.
Schegloff (1967) furthermore stated that it is up to the caller rather than to the answerer
to start the identification work. In discussing self-identifications by North American
answerers (e.g. Police Desk), Schegloff wrote:
The work of identification [is] the initiators work, in the case of telephone conversation the callers work, for it is his entitlement to [start] the conversation that may be
at issue. Answering the telephone with a self-identification is pre-emptive because
it does the work of identification before the turn-taking organization has provided
caller his first opportunity for doing so. (pp. 4445)
This is fundamentally different from the Dutch situation, in which the answerer begins the
work of identification.
So, there are two clear differences between the North American and the Dutch situation,
(a) Hello versus self-identification being the typical answer to the summons, and (b) the
caller versus the answer beginning the work of identification.
Whether or not these are cultural differences depends on how we define the concept
of culture. I see them as cultural differences, because I see culture as a set of typical behavanswer the phone. Cell phones provide for the possibility to be used in public spaces such as streets,
shops, and public transportation, where strangers can listen in to the conversation. This may have
an effect on the way people answer these calls. A second feature of cell phones is that they usually
are not shared with other members of the household, and callers know this. Answering the call by
mentioning ones name is thus a redundant action, because a voice sample suffices as a selfidentification (cf. Sanders, 1998). In the situation in which answerers can read the incoming telephone
number, they may know who is calling before the telephone has been picked up. Theoretically
speaking, this provides for the possibility to answer the call by saying, for example, Hi Mom.
(Compare Hopper et al. 1990/1991 on the possible effects of technology on how people answer the

238 Studies in language and social interaction

iors, norms and values that are largely shared and oriented to by the members of a (speech)
Apart from the issue of whether or not these are cultural differences, it seems safe to say
that the Dutch practice does not quite fit Schegloffs description of the first two sequences,
the summons/answer sequence and the identification/recognition sequenc: especially not
because the party who starts the identification sequence in the United States is the caller (in
Turn 1), whereas in the Netherlands it is the answerer (in Turn 2). Schematically, it breaks
down as in Table 16.2.
TABLE 16.2



The Netherlands

A. Voice sample

A. Self-identification
C. Self-identification

C. Other-identification


My Dutch colleague Leo Lentz (Lentz, 1997) introduced me to the thought that conversational practices (like answering the phone) may change over time within a culture. Lentz
asked himself: Did the Dutch always self-identify, or may they have started out in a different way? The problem with studying the history of telephone conversation is that we
do not have recordings of calls that were done before the last few decades. However, there
is an indirect way to approach the issue. Lentz analyzed theater plays and novels written
between 1920 and 1940 with respect to the use of the telephone. He also studied early telephone directories that not only provided telephone numbers but also instructed the Dutch
people how to use the phone. What Lentz found is the following.
The very first Official Guide of the Dutch Bell Telephone Company, No. 1 of 1881
instructs telephone conversationalists as follows: If the member is called by the telephone
bell, one takes the telephone off the hook, pushes it against the ear, makes clear he is present, and listens. The Namelist for the Interlocal Telephone Service of 1925 says: In
case of a call, one says his name and does not shout Hallo. The call should be answered
immediately. Ten years later it was said: One does not answer the telephone with Hallo,
but mentions name or telephone number. This time the directory also gave accounts for
the advice: this in order to prevent loss of time, and in case of a wrongly dialled number,
to give the caller the opportunity to put back the receiver and disconnect.
Not only directories, but also books on etiquette would instruct the Dutch how to behave
in case of a call. In 1945 it was written that The etiquette requires that the one who is
being called self-identifies immediately. A book from 1960 not only explained how to
do it: In case you are being called, mention your name, but also how not to do it: Do
not say just Hallo because this does not inform the caller. A few years later, in 1964,
another etiquette book is even more precise: We do not begin our conversation with the
silly Hallo, whos there? but mention shortly and consisely ones name or give ones tele-

Gender differences in telephone conversations 239

phone number. As we see later on, providing ones telephone number never made it as a
practice in the Netherlands.
Lentz came to the conclusion that in the early days of telephone communication, the
Dutch must have started out their answering practice by saying hello. Only after World
War II and after the Dutch phone company had kept telling their costumers to mention their
names, rather than saying Hello, the Dutch gradually developed from providing a voice
sample into selfidentifying.
An interesting point in Lentzs work is the idea that some new piece of technology
requires a conversational practice that does not yet have a precedent that can simply be
followed by newcomers in the conversational arena.
A second point of interest is that people can gradually change a conversational practice
for whatever reasons. As a member of Dutch society, I had the impression that the Dutch
way of answering the phone was slightly changing over the last several years. This impression was based on two mundane observations. First, Dutch people, especialy women,
sometimes state that they say Hello when picking up the phone, because, as they say:
You dont know whos calling, after all, right? They seem to see this as a safe practice
that protects their privacy. This may well be in line with the fact that more and more Dutch
people have unlisted phone numbers nowadays.
The second mundane observation derives from my research on interaction in telephone
survey interviews, that I have done since 1991. I collected hundreds of recorded survey
interviews that are carried out from Dutch survey research centers. These interviewers randomly phone to Dutch citizens homes. Listening to these recordings I got the impression
that, compared to the late1980s data, more people who answered the phone would say only
Hello or Yes, either with a rising or a falling intonation contour. Moreover, it seemed
as if these were the people who angrily inquired how the interviewer got hold of their
phone number, as it was an unlisted number. So, I wondered if the Dutch might be moving
up a little toward the American system.
When one day, my student Titia Houwing asked me for an idea what to study for her
thesis, I proposed she might look at my interview data and compare the openings with
the late 1980s data that were reported in HoutkoopSteenstra (1991). Titia transcribed the
first 142 opening sequences of these interview tapes, leaving out the cases that would fall
into my 1991 category of variant cases. Table 16.3 shows what she found. Note that the
four variant cases of the late eighties data reported in Houtkoop-Steenstra (1991) are left
out here.
TABLE 16.3


X2=0.70, DF=1, P=0.40

LATE 1980s








240 Studies in language and social interaction

Let me first make clear why I use the term non-self-identification rather than voice sample
in this table. Anita Pomerantz pointed out (personal communication) that if my Dutch
informants claim they say hello in order not to be recognized by creepy callers, they can
not be seen as providing a voice sample. Remember that a voice sample is meant to be
recognized by caller (Schegloff 1972, p. 353; 1986, p. 123).5
Table 16.3 shows that there is no significant increase in the percentage of answerers who
withhold self-identification. Maybe the people who claim they answer the phone by saying
hello nowadays, do not actually do say hello once they are being called. Perhaps it is less
easy to say goodbye to a conversational routine than one might wish.
The next step in this study was to look for possible gender differences in the way Dutch
people answered the phone in 1995. Could it be that the 9% of the answerers who withheld
self-identification were mainly women?6 As was said earlier, it were especially women
who claimed they answered the phone by saying Hello in order not to be identified. After
going through the transcripts again in order to find out whether the answerer was male or
female, we did not find significant gender differences, as Table 16.4 shows.

TABLE 16.4

% of Men


X2=0.60, DF=1,P=0.439

% of Women 1995







As we reanalyzed the transcribed openings, we came to realize that selfidentification is

a broad category that comprises different ways of selfidentification. When answerers perform the activity of self-identification, they also choose a certain formulation with which
they self-identify. There were four ways in which the Dutch self-identified in these 1995

Pomerantz remark is a challenging one, which, however, can only be confirmed if we would
know what action the Dutch answerers intend to perform when saying Hallo. And it may well be
that some mean to withhold self-identification and/or to invite the caller to self-identify, whereas
others mean to indeed provide a voice sample to be recognized by the caller. As we have no clear
means to decide on participants intentions, we need a less interpretive term than voice sample for
the Dutch situation. Therefore I use the more descriptive term non-self-identification.
Conversation analysts are very reluctant to engage in quantitative and distributional studies of
conversation for reasons that were laid out by Schegloff(1993); see also Hopper (n.d.) and Schegloff (1987). I wish to point out that the problems that Schegloff discussed appear not to apply to
the study at hand, that is, the study of the response to yet unknown callers summons.

Gender differences in telephone conversations 241

data: (a) mention first name: (With) Hanneke; (b) mention both names: (With) Hanneke Houtkoop; (c) mention last name: (With) Houtkoop; (d) Title+last name: (With)
Misses Houtkoop. Theoretically speaking, one could also provide ones telephone number, but nobody did so.
After we did a statistical analysis of our data, we found some striking gender differences
in ways of self-identifying (see Table 16.5).
TABLE 16.5



First Name
15 19
First+Last Name
33 24
Last Name
52 16
Title+Last Name
0 16

54 100% 75
X221.77, DF=3, P=0.000



The most striking finding is that whereas 21.5% of the women in this sample say Mrs. Last
Name, not one man says Mr. Last Name. The background of this difference is unclear.
Another result is that only when it comes to the percentage of persons using First+Last
Name men and women act the same. The genders score very different on the other two
ways of self-identifying. The women provide First Name almost twice as often as the men
do (25% vs. 15%), whereas the men provide Last Name more than twice as often as the
women do (52% vs. 21%).
What do answerers do when they identify themselves as First Name or as Last Name?
Providing a self-identification as such may well be a cultural specific routine, but making
the choice for one form of self-identification over another, is a different issue. Do people
present a certain aspect of themselves, when choosing for one or the other form? Do people
project informality when they present themselves by First Name, and do they project formality when they present themselves by Last Name?
There is a yet unmentioned aspect of these calls to domestic homes that may be relevant here. In two thirds of the cases, the phone was answered by women.7 If we leave
out all women and men who live on their own, and concentrate on households, this may
mean that answering the phone is primarily the business of the woman in the house. So for
women, the telephone may be part of the domestic and private world of relatives, friends
and aquaintances. And in answering the phone by providing first name only they recipient
design their answer and are doing being intimate (cf. Lindstrm 1994). Dutch men, on
the other hand, may consider the telephone as belonging, in the first place, to the public
domain, where more formal ways of speaking are being used. So, one might suggest that
Ton Boves (personal communication), a Dutch survey researcher confirmed that in The Netherlands calls from survey research centers are answered far more by women than by men.

242 Studies in language and social interaction

the different ways in which a large proportion of Dutch men and women answer the phone,
reflect their different orientations to the category of people whom they expect to call.8 One
could object to this line of reasoning by saying that Dutch women, just like Dutch men, are
being called by potential strangers in their workplaces. However, if we look at the statistics
of the Dutch labor market (NRC Handelsblad 1998), it turns out that in 1969, only 30 years
ago, no more than 30% of the Dutch women had a paid job. For the men, this was 98%
(See Table 16.6).
TABLE 16.6
Dutch Labour participation (2064 years old)




If we also consider the fact that a large percentage of these womens jobs were, and still
are, part-time jobs, it seems reasonable to say that for Dutch women the telephone used
to be primarily part of their domestic lives. And for Dutch men, the phone used to belong
to their public lives. These then are the different settings in which the genders may have
come to develop their gender-related ways of answering the phone. For the time being I
think that the gender-specific way of self-identifying is, in the first place, a reflection of
the traditionally and still existing very unequal labor division in the Netherlands. If this
suggestion holds true indeed, and considering the growing number of working women in
the Netherlands, we may expect the gender-related differences in answering the phone to
gradually decrease in the future.
From a conversation analytical point of view, one might say that in the way the Dutch
men and women in our data answered the phone they displayed an orientation to a different class of potential callers, and that they recipient designed their answering utterances.
Although this may well have been the case in specific cases, I strongly believe that the
way in which individuals answer the phone is a case of socialization and routine in the first
place. Dutch children are explicitly taught to answer the phone by mentioning their names.
There is no research on how Dutch children develop their phone answering practices, but
one may expect the following: They start out answering the phone by saying Hello? as
my collection of telephone openings suggests. Soon their parents instruct them to mention their name when answering the phone, which they take as mentioning their first name
only. And in hearing how adults answer the phone, they will gradually come to see that
adult women provide First Name or First 4- Last Name, whereas the adult men provide
First+Last Name or Last Name only. At some point in their lives a large proportion of the
Dutch children will adopt this gender-specific way of answering the phone. Had these
children been raised in the United States, they would have learned to answer the domestic
phone by saying hello. As Hopper (1992) says about this American routine, it was established in the early years of telephone use and has remained somewhat stable.
I owe this perspective to Gitte Rasmussen, with whom I discussed these gender-related differences
in self-identifying.

Gender differences in telephone conversations 243

The way in which people answer the phone is not only a matter of socialization, but also
of routine behavior.9 Each Dutch person has his or her own idiosyncratic routine; they not
only differ in the form of self-identification they use but also in whether or not they begin
their self-identification with Hello and/or with, and in the intonation contour of the
answering utterance and their speech rate.
When discussing how telephone conversationalists proceed in establishing the parties
identities, it was already suggested that cultural differences exist with regard to how to
carry out the interactional task of mutual identification. In some cultures, answerers typically provide vocal recognition cues; in other cultures, they typically self-identify. This
study shows that there may also be genderrelated differences within one and the same culture when it comes to how to answer the phone. These differences may be seen as stylistic
differences (cf. Hopper et al. 1990/1991). In Dutch society, the genders do not differ in
whether or not they self-identify, but in how they self-identify.
I thank Huub van den Bergh, Paul Drew, Paul ten Have, Henk Lammers, Leo Lentz, Joost
Schilperoord and two anonymous reviewers for their helpful comments on an earlier
version of this paper.
Adler, J. (1993). Telephoning in Germany. Telecommunications Policy, 281296.
Godard, D. (1977). Same setting, different norms: Phone call beginnings in France and the United
States. Language in Society, 6, 209219.
Hopper, R. (n.d.). Quantity envy. Unpublished manuscript, University of Texas at Austin.
Hopper, R. (1992). Telephone conversation. Bloomington, Indiana University Press.
Hopper, R., Doany N., Johnson, M., & Drummond, K. (1990/1991). Universals and particulars in
telephone openings. Research on Language and Social Interaction, 24, 369387.
Hopper, R., & Doany, N. (1989). Telephone openings and conversational universals: A study in
three languages. In S.Ting-Toomey & F. Korzenny (Eds.), Language, communication and culture (pp. 157179). Newbury Park, CA: Sage.
Hopper, R., & Chen, C.H. (1996). Languages, cultures, relationships: Telephone openings in Taiwan. Research on Language and Social Interaction 29, 291313.
Houtkoop-Steenstra, H. (1991). Opening sequences in Dutch telephone conversations. In D.Boden
& D.H.Zimmerman (Eds.), Talk and social structure (pp. 231252). Cambridge, England: Polity
Lentz, L. (1997). The history of opening sequences in Dutch telephone conversations. In L.Lentz &
H.Pander Maat (Eds.), Discourse analysis and evaluation: Functional approaches (pp. 87111).
Amsterdam/Atlanta: Rodopi.
It is striking that in Lindstrms Swedish data it was one individual who was responsible for 3 of
the 5 Hello-answers.

244 Studies in language and social interaction

NRC-Handelsblad. Vrouwendeelname groeit. [Womens participation grows]. (1998, July 2). p. 4.
Lindstrm, A. (1994). Identification and recognition in Swedish telephone conversation openings.
Language in Society, 23, 321352.
Placencia, M.E. (1998, July 19). Telephone conversation openings in Ecuadorian Spanish and British English. Paper presented at the 6th IPrA conference, Reims, France.
Sanders, E. (1998, October 7). Ik zeg: Hallo. [I say: Hello.]. NRC Handelsblad.
Schegloff, E.A. (1967). The first five seconds. Unpublished doctoral dissertation. University of
California at Berkeley.
Schegloff, E.A. (1972). Sequencing in conversational openings. In J.J. Gumperz & D.Hymes
(Eds.), Directions in sociolinguistics (pp. 346380). New York: Holt, Rinehart & Winston.
Schegloff, E.A. (1979). Identification and recognition in telephone conversation openings. In G.
Psathas (Ed.), Everyday language: Studies in ethnomethodology (pp. 2378). New York:
Schegloff, E.A. (1986). The routine as achievement. Human Studies, 9, 111151
Schegloff, E.A. (1993). Reflections on quantification in the study of conversation. Research on
Language and Social Interaction, 26, 99128.
Sifianou, M. (1989). On the telephone again! Differences in telephone behaviour. England versus
Greece. Language in Society, 18, 527544.

Talk in Institutional Settings
The importance of social institutions is indicated by the extensive attention devoted to
them in scholarly work (e.g., Drew & Heritage, 1992; Labov & Fanshel, 1977; Morris &
Chenail, 1995). Drew and Heritage pointed out that an occasions institutionality is not
derived simply from its setting. Rather, interaction is institutional insofar as participants
institutional or professional identities are somehow made relevant to the work activities in
which they are engaged (p. 4). Thus interaction is central to the constitution of institutional settings. As Heritage (1984, p. 242) pointed out, interaction is both context shaped
and context renewing. Work in language and social interaction (LSI) has examined institutional settings from a number of different perspectives. In this section authors focus on
a range of institutions from several different perspectives, showing both how institutions
impinge on interaction, and how interaction is constitutive of institutions.
A formal distinction between casual and institutional talk (see Heritage & Drew, 1992;
Sacks, Schegloff, & Jefferson, 1974) assumes that casual conversation occupies one end of
a continuum of speech-exchange systems, the other end of which is marked by increasing
restriction on turn taking. Other than distinguishing various institutional events by their
turn-taking features (such as meetings, interviews, and debates), little work has attempted
systematically to explore variations in talk in different types of institutions. In chapter 17,
Paul Drew explores the possibility that formulations, which are utterances providing a
summary or gist of preceding talk, might vary across four different institutional contexts.
Drew addresses this question while reflexively considering analytic issues raised in the
Robert Sanders (chap. 18) explores how methods and findings designed for studying
face-to-face and telephone talk might apply to interactions taking place over Marine VHP
radio between people on different boats. This medium carries particular constraints on
interaction due to the limitations of being unable to use the same channel for both listening
and speaking. Sanders shows how participants manage coherent interactions despite these
limitations. In particular, he demonstrates how laughter gets accomplished between speakers who cannot hear each other laugh in overlap.
Next, the article by Jennifer Molloy and Howard Giles (chap. 19) exemplifies work on
intergroup communication, taking up the important but understudied area of communication between civilians and law enforcement officers. This chapter shows how sociolinguistic research can have real-life applications that offer hope for improving communication
between groups. It pays tribute to an interest area of Robert Hopper, who coauthored with
his former student Dennis Gunderson a textbook for law enforcement officers on communication (Gundersen & Hopper, 1984).
The next three chapters examine interaction in a therapeutic setting. This has been a
popular site for research on LSI. Harvey Sacks, a founder of conversation analysis, made
some of its earliest applications in the study of a therapy group for troubled teenagers
(e.g., 1992, pp. 281299). Since the publication of Labov and Fanshels (1977) classic,

Part III: Talk in institutional settings 247

Therapeutic Discourse, analysis of clinical discourse has flourished. Through close observation and analysis of therapy recordings (e.g., Morris & Chenail, 1995), researchers
have shown how therapeutic discourse may be structured in ways that ordinary talk
is not, which has practical import for the discourses of healing that clinicians and clients
interactively bring about.
First, G.H. Bud Morris (chap. 20) examines preventatives, that is, utterances that orient
to and forestall the possibility of interactional trouble. In this study he builds on previous
research on disclaimers and accounts, grounded in the study of alignment as a fundamental interpersonal activity. He briefly introduces seven types of preventatives and offers
an instance of each type, arguing for both an ordering of them in terms of seriousness
and a time sequencing of them, such that speakers may start with the mildest and build
toward the strongest. He suggests that preventatives serve an important role in minimizing interactional problems that could deepen; he also argues that a rule of the earlier, the
better guides the doing of preventatives, as people seek ways to keep interactions going
Next, Duff Wrobbel (chap. 21) examines a recording of a family therapy session, focusing on several minutes of interaction leading up to an aha moment in which the wife
experiences (or at least displays) a sudden flash of insight or self-revelation. The author
identifies various external antecedents associated with the wifes internal experience,
including subtle communicative moves on the part of her therapist.
Taking seriously the social constructionist view that individual selves and psychological states are largely products of social interaction, Kurt Bruder (chap. 22) promotes
a discourse analytic approach to therapeutic intervention. The author argues that therapists can (perhaps should) analyze (in real time) the moment-by-moment and turn-by-turn
unfolding of therapy sessions, noticing and calling clients attention to the inevitable display and enactment of identityconstituting talk. Not only would the therapist gain insight
into a clients discursively generated psychosocial experiences, the argument goes, a therapist could share these insights with the client, who might thereby be acculturated into
processes of self-healing.
The last three chapters in this section examine interaction in the medical setting. Anita
Pomerantzs article (chap. 23) on modeling as a teaching strategy is part of an ongoing
research project concerning medical precepting, the process through which supervising
physicians train and oversee medical students working with patients in clinical settings.
She argues that modeling provides a solution to the complexities of needing to ensure
proper patient care, instruct interns, and yet avoid compromising the interns professional
role in front of patients. The chapter examines not only interactional phenomena, but also
participants perceptions of the effectiveness of a particular pedagogical strategy, as determined through surveys and interviews, which are standard ethnographic methods.
Douglas Maynard and Richard Frankel (chap. 24) examine a sequence of conversations
between a doctor and a female patient whose mammogram results were mixed, warranting additional tests (e.g., ultrasound) that also turned out to be indeterminate. The authors
focus on diagnostic news as an interactive and emergent accomplishment: The patient in
this case happens to also be a registered nurse, able to interpret test results and ready to
resist the doctors conclusion that the results constitute good news. By attending to the
details of this particular case, the authors show how diagnostic negotiations are delicately

248 Studies in language and social interaction

woven into conversations between health care professionals, who sometimes joke (in a
self-conscious or self-reflexive way) about the medical practice in which they simultaneously participate.
In the final chapter of this section, Daniel Modaff (chap. 25) investigates coordination of
talk and subtle body movements during doctorpatient interviews. Specifically, he examines
transitional moments interactively brought about: Doctors sometimes turn away from their
patient and toward some object in the room (e.g., a stool), indicating a shift in the immediate focus of attention, giving the patient an opportunity to align with the transition possibly
being cued. Through such small and subtle forms of interaction, large social institutions
(such as a medical community) are sustained day by day, mostly taken for granted.
Drew, P., & Heritage, J. (1992). Talk at work. Cambridge: Cambridge University Press.
Gunderson, D. & Hopper, R. (1984). Communication and law enforcement. New York: Harper &
Heritage, J. (1984). Garfinkel and ethnomethodology. Cambridge: Polity Press.
Labov, W., & Fanshel, D. (1977). Therapeutic discourse: Psychotherapy as conversation. New
York: Academic Press.
Morris, G., & Chenail, R. (Eds.). (1995). The talk of the clinic: Explorations in the analysis of
medical and therapeutic discourse. Hillsdale, NJ: Lawrence Erlbaum Associates.
Sacks, H. (1992). An Introduction Sequence. In Lectures on Conversation (2 vols.) (G.Jefferson,
Ed.) (pp. 281299). Oxford: Blackwell.
Sacks, H., Schegloff, E.A., & Jefferson, G. (1974). A simplest systematics for the organization of
turn-taking in conversation. Language, 50, 696735.

Comparative Analysis of Talk-in-Interaction in
Different Institutional Settings: A Sketch1
Paul Drew
University of York
There has in recent years been some discussion and debate (e.g. Hak, 1995; Hopper, 1995;
and Schegloff, 1992) concerning the study of talk that takes place in institutional settings. Much of this debate is about how (and whether) institutional interactions are to be
distinguished from those that are not institutional: for instance, if the family is an institution, why then are telephone calls between members of a family not institutional? But
the question about what is special or different about institutional interactions shades into
others, including whether, since the practices and organizations of talk are generic to talkin-interaction, and are not specific to talk in any given setting, it is appropriate to separate
the study of talk in one setting (for instance in medical consultations, courts, or in news
interviews) from others? There is a tendency, it is argued, to treat the conduct of talk and
interaction in a particular institutional setting as unique to that setting. Because researchers generally focus on one specific institutional setting, they commonly assume that any
patterns or practices that are observed in that setting can be attributed to the particular
organizational features and exigencies associated with that setting.
The analytic connections between the very identification and delimitation of institutional interactions, and the readiness to attribute to features of talk in a given setting a
certain kind of uniqueness, is summarized succinctly by Hopper in a trenchant commentary
about whether the study of institutional settings might, as he puts it, blunt the cutting
edge of conversation analysis?:
A problem with analyses of institutional talk is embedded in describing it as institutional talk. This terminology carries the traditional setting divisions of communication study. Given a characterization of a strip of talk as the opening of a medical
interview, or given a title of an essay as Host Talk on X TV Show it becomes
difficult to resist offering an institutional setting explanation as the explanation for
whatever we find in these materials. (Hopper, 1995, p. 374)
This paper is based on a talk which I first gave at a meeting of Nordic sociolinguistics projects,
held at the Swedish School of Social Sciences, Helsinki University, in May 1992. A previous version was published in H.Lehti-Eklund ed., 1998. In revising this for publication in this volume, I
have benefited from the particularly thoughtful comments of two anonymous reviewers: although I
have not accepted all their suggestions, I have borrowed from these at certain points without further

250 Studies in language and social interaction

Hopper develops these arguments in a number of directions, some of which have also been
articulated by other voices in this debate. I would like to take up just one of these directions
hereone which is more or less implicit in his commentary, but which is quite explicit at
some points in his own research (Hopper and Drummond, 1992; Hopper and Chen, 1996).
That is, that comparative analysis may be required in order to assess how far a certain pattern, device or practice is generic to talk-in-interaction, and therefore not restricted to any
one type or setting; or whether, perhaps, there are systematic variations in the occurrence,
scope, properties and form of certain practicesvariations associated with the specific
settings in which they occur and the activities in which participants are engaged in those
Although work on institutional interactions often implies or explicitly claims a comparative justification for attributing a pattern or device (or the salience and import of that
pattern or device) to a given setting, nevertheless it is true that those claims are generally
not supported by comparative research. Hopper is correct when he points out that Most
essays about talk within institutions have treated just one setting, which foregrounds setting-based explanations for things happening as they do (Hopper, 1995, p. 373). Typically,
researchers (and I include myself here) investigate interaction in the particular setting they
are studying, perhaps with only an indistinct comparative perspective in minda general
awareness that what they are finding in their data/setting is unlike patterns or features which
(probably) obtain in other settings, but without exploring that suspected comparative difference at all systematically. And there is something further which is worth highlighting in
a remark which Hopper makes about such comparisons, Analyses of talk in institutional
settings frequently proceed by posing comparisons between practices used in that settings
and those in mundane conversationpractices that seem relatively context-free (Hopper,
1995, p. 372). I take this to mean, in part at least, that we can claim about a practice that it
has some relatively specialized use or consequences in a given settingeven though the
practice itself is not restricted to that setting (just as oh is not restricted to mundane conversation) and is therefore relatively context-free, and despite our not having investigated its
various uses or properties in other settings (hence the tendency to attribute to that practice
in that setting some unique properties, or to explain its occurrence in terms of the special
properties of that setting).
As a way to begin to address some of these issues of comparative analysis, to sketch
what such an analysis might involve and what kinds of properties of a practice we might
investigate, it occurred to me to bring together some findings about a particular conversational practice, that of formulating what another speaker is saying or has said. Plainly
the practice is in some respect context free; it is not restricted to any particular context,
whether mundane or institutional. However, I wondered whether the practice may exhibit
some systematic variations associated with the settings in which it is used. What follows,
This is a slight re-statement and amplification of the proposal which Heritage and I made, that
The basic forms of mundane talk constitute a kind of benchmark against which other more formal or institutional types of interaction are recognized and experienced institutional forms of
interaction will show systematic variations and restrictions on activities in their design relative to
ordinary conversation (Drew and Heritage, 1992, p. 19).

Comparative analysis of talk-in-interaction 251

then, is a sketch, the objective of which is to enquire whether, if a practice appears to be
context free, we should let it rest there, assigning this to one of those generic practices of
talk-in-interactionor rather, whether the practice is molded into distinctive shapes by
participants when they engage in the specific interactional work associated with certain
institutional settings.
The sense or meaning of a conversation, part of a conversation, or a turn in a conversation,
is not unambiguous. The meaning of what someone said or what we have been talking about
can be describedor formulatedin different ways. Of course most of the time, participants in a conversation take it that they have understood the others meaning sufficiently
to be able to produce a relevant response, without first having to check their interpretation
of what the other meant. But from time-to-time, participants May treat some part of the
conversation as an occasion to describe that conversation, to explain it, or characterize it, or
explicate, or translate, or summarize, or furnish the gist of it, or take note of its accordance
with rules, or remark on its departure from rules. That is to say, a member may use some
part of the conversation as an occasion to formulate the conversation (Garfinkel and
Sacks 1970, p. 350). Thus formulations are a means through which participants may make
explicit their sense of what we are talking about, or what has just said: they are a means
for constructing an explicit sense of the gist of the talk thus far.
In their seminal paper on formulations, Heritage and Watson (1979) (following Garfinkel and Sacks) identify and describe a range of types of formulations. I shall focus here
on those in which a speaker offers his or her interpretation of what the other meantan
activity which generally takes the form (So) what you mean/are saying is, or something resembling that. These are familiar to linguists as metacommunicative acts, expressions through which participants comment on the nature of the discourse in which they are
engaged, or are about to be engaged. One reservation I have with the term wetacommunication is the implication that such expressions stand above or outside the talk. Heritage
and Watson (1979) argue cogently that formulations are themselves events or moves within
the talk, and as such may be geared primarily to participants ongoing, specific practical
interactional tasks. In this respect, they are as much part of the talk as any other kind of turn
or discourse practice. Indeed we can see that formulations are produced in very specific
interactional environments or circumstances in various kinds of institutional discourse, and
that they serve to perform specific interactional tasks which vary according to the setting.
But more of that in a moment: for the present, I want to make an initial observation about
the claim I made earlier that they are contextfree.
In their title Formulations as conversational objects, Heritage and Watson can be taken
to imply that formulations are the realizations of a generic practice in talk-in-interaction
(mundane conversation as well as other forms of talk). Just parenthetically, I take it that
formulating is the practice, and that a formulation is the object or device through which
the practice is mobilized by participants in a given interaction. At any rate, I supposed

252 Studies in language and social interaction

from the title that Heritage and Watson were describing a practice/device that had its home
base in conversation.3 I was wrong. Re-reading the article, I discovered that none of the
instances they show (at least, none resembling the form I outlined above) was taken from
ordinary conversation: instead, they were from a variety of institutional contexts, mostly
news interviews.
This led me to make a search of the recordings of mundane (mostly telephone) conversations we have (much of this data obtained in the years since Heritage and Watson wrote
their article) in order to check whether their data were skewed by their happening to have
been working at that time on news interviews. I was surprised to find almost no instances
of formulations which in any way resembled (So) what you re saying is. This is one of
the only two clear cases I found during a not-quite-exhaustive search.
(1) [HG: 45] (Talking about Nancys skin problems, and the medication she has been given) the
Nan: So e gay me these pills tih ta:ke=
Hyl: =What.Tetracykuhleen?

Nan: .PT NO: cuz I usetuh take that an it didn

he:lp so e gay me something e:lse.=

Hyl: =Hm: .

Nan: He sai:d- yihknow, (0.2) sometimes Tetracyklene

jus doesn he:lp.


Nan: Also he sid that (0.3) .t what you ea:t, (0.2)

end how you wash yer face has nothing tih do


with it,

Hyl: Yer kiddin[g.
Nan: [nNo:,

Nan: He says ts all inside you its n emotional

thingn, .hhh e[:n,

Hyl: [Yeah buh whatchu ea:t if you

eat greasy foo:d=

I think that the use by Heritage and Watson of conversational in their title was owed in part to
how enquiries in conversation analysis were cast, at the time they wrote (1979). They were describing the properties of what they took to be a general (if not quite generic) practice: they did not then
have the more accurate nomenclature talk-in-interaction (which as far as I know was introduced
by Schegloff in the early 1990s) with which to refer to its scope. What they might have meant by
conversational was general: however that misled me, at least, into assuming that it was primarily in
ordinary conversation in which participants employed this practice.

Comparative analysis of talk-in-interaction 253







=We:h he said its no:t the fact thet youve

eaten the greasy food its a fact thet you
worry about it. En that makes you
[break ou[;t.
[.Teh.k.h[hhhhhh Ymean I cd sit here en eat
french fries n ez longz Im not worrying
about it I [wont break ou]hhhthh
[I g z a : :]:ctly,
.hh[hh] tsa [buncha [h:::::::[horse:]:
[I] belie [ve im [too hes[rilly-]
[e.-hes rilly a [smart,]
[(isk-skih-) f: [father]s, .huhh [.hn

Here, in Lines 2628 Hyla seems to offer an interpretation of what the other has said (note
that in #1 Nancy reports in Lines 8, 11, 18 and 22 what the doctor said to her, as he said
such-and-such: these are instances of indirect reported speech, and not therefore the practice on which I am focusing here). However, without being very technical about this, it is
reasonably clear that Hyla is being tendentious in her interpretation (remembering that
Nancy is reporting what the doctor said to her; but she is doing so in a fashion which makes
it evident that she is aligning with or accepting what he said). Moreover in producing this
version of what Nancy means, Hyla is making a move that is a preliminary to expressing
her scepticism with the doctors advice (Lines 30 and 31, its a bunch of horse feathers).
So this practice of offering an interpretation of what the other meant is employed in
mundane conversationbut apparently only very infrequently. This is in contrast to various institutional settings in which, as I knew from some previous research,4 such formulations are very frequent indeed. This then is a practice/device that might be generic, though
not much found in conversation. However, to regard this as a generic practice, we need to
explore whether it has properties which are context-free in so far as they are exhibited in,
and underpin, its use in talk in any setting in which it occurs (this to paraphrase Hopper,
1995, p. 372). This led me to considering how this practice was employed in other settings,
and whether its properties, its form or linguistic features, were identical; or whether instead
the design features associated with formulating in various settings differed systematically
according to the kind of interactional work which formulating is done to manage in particular settings (a kind of correspondence between form and function)in which case,
whilst the practice of formulating may be context-free, we cannot discern a generic device
through which the practice is implemented.
In order to pursue this question, I will briefly describe instances of formulations in
psychotherapeutic consultations, call-in radio programmes, news interviews and industrial
I am drawing here particularly on work of two of my previous graduate students, Ian Hutchby
and Esther Walker. Their research into, respectively, radio call-in programmes and negotiations
between management and unions in an industrial setting is cited in the bibliography. I would like to
acknowledge my indebtedness to their work, from which many of the data extracts are taken.

254 Studies in language and social interaction

negotiations. Space allows me to show only a single example in each setting: but this will
perhaps be sufficient to sketch a comparisonone which will suggest that the precise
linguistic forms that such formulations take may differ, and do so in ways which seems to
relate to the interactional task (function) which the formulation serves in each setting.
An instance of a formulation in psychotherapeutic sessions, which illustrates what
appear to be some of the characteristic features of such expressions in this setting, is the

[Therapy: PB:53172:7]
Well hhm I ve been ah:m,hh .k better:

with her. (.) u-lately, then I had been (.) in

a long ti.-ime. (0.2) .p.hh (0.9) e-Oh: Go:d but

that couldnt I mean if that ever created a

problem like Im having no:w.

May not create a problem: it might make it

possible for a problem to come ou:t

You mean she couldve always felt like this.


.pl.hhhhh (0.6) mYou know Sams been very

about this. N he: (0.4) s-aid that I shouldnt

have sent her to school when I did. (1.5) And

thats probably what caused it.

The patient (Brenda) has been telling about some aspect of a problem she is having with
her very young daughter: in her first turn in this fragment she appears to notice a paradox
between an improvement in her conduct (Ive been better with her lately) and the worsening of her childs problems (a problem like Im having now). The therapist responds
(Lines 78) by commenting on, or making an observation about, the patients account, and
seems to offer an alternative association between Brendas improvement and her daughters apparently increasing problem (in other instances, such commenting may be done in
an interrogative form, as in Think she might be trying to tell you something about you?).
The patients formulation (Line 10) is an expression through which she offers her interpretation of the characteristically implicit, allusive or indirect message which she discerns in
Laurals remark. The patient is constructing a sense of what the therapist might be alluding
to in her comment/observation, putting that implicit message into so many words, for the
therapist to confirm. The sense of the therapists intending to be allusive, to hint at but not
make explicit some point to be found in the patients telling, is perhaps to be found also in

Comparative analysis of talk-in-interaction 255

Laurels minimal and unelaborated confirmation of Brendas formulation/understanding
check (see Line 12). Moreover, in this kind of therapy at least, there is evidently an orientation to a strategy whereby the therapist guides the patient towards finding for herself
what might be the true nature of her problem. But at any rate, it is evident that the patient
treats the therapists comment or observation in Lines 78 as implying or alluding to something about the problem, which goes beyond what she (the therapist) has said explicitly.
Brendas formulation is an attempt to put into words that implied message, and thereby
constitutes an action that is part of her finding, and showing that she is finding, the direction
in which the therapist is pointing her.5 In this way, her formulation embodies an orientation to the reciprocal role of therapist and patient, and the behaviors expected of each (on
formulations in psychotherapeutic settings, see also Davis, 1986).
In his study of radio call-in programmes (specifically, a program broadcast by a London
radio station, in which listeners called in to air and discuss with the program host their
views about any matters of current interest, or which concerned them), Hutchby (1996)
focused on the ways in which the host constructed controversy. Whatever topic a caller
had called in about, and whatever position he or she held, the host invariably managed to
challenge their point of view and contested their argumentso that often the most unexceptional views were turned into the subject of a controversy between host and caller.
Among the moves which the host made in seeking to defeat the callers view was to
formulate the callers argument, to summarize the gist of what he or she was saying.
(3) [BH:2/2/89:12:12] (from Hutchby 1996, pp. 7071) (The caller has phoned in to
recommend a product which will prevent dogs fouling the footpath outside ones home)
U: sually when a dog fouls:, .hh e::r it, it

lea:ves-=the scent that is left behind even if


you:, clean up with boiling wa:ter an


disinfectant, .hhh is a mar:ker. .h An when e


comes on is e::r, (w-) wa:lk the next da:y,


when e gets tuh that ma:rk, he does the same


thing again.


Er you s-seem tuh be suggesting that they go

A simple way to put this is that Brenda is checking her understanding of what Laurel has said.
However, she may be doing so in the service of another activity, namely showing that she is seriously considering the implications of the alternative association which is implied in Laurals remark
(that the daughters problem hasnt been caused by Brendas recent conduct, but is the result of
pre-existing circumstances or events which have only now come to the surface). So that checking
her understanding can be a way to show that she is considering this possibility.

256 Studies in language and social interaction



tthe same place evry ti:me. Becuz theyve

been there buhfore,
Ooh yes,=quite often ye:s.=
=Yeah but er(h)n(h) then:, .h e:r,=
=An[d other [dogs will: also.
[This- [This mea:ns that they never go in
a diffrent pla:ce,=doesnt it.

In Lines 1012 the host formulates the callers account (opening turn) as amounting to
an argument that when dogs poop on the pavement they go tthe same place evry ti:me.
Becuz theyve been there buhfore,. It is readily apparent from this extract that that formulation is the first part in an argument sequence: after the caller confirms this formulation,
the host subsequently constructs an upshot of the callers position, an upshot which reveals
the absurdity of that position (here, the absurdity of holding that a dog always poops in
the same place). Three features of this formulation are worth drawing attention to at this
stage. First, it is tendentious; it is constructed to serve the hosts purpose to challenge and
undermine the callers position (there are several features which are associated with that
tendentiousness, including extreme case constructions such as same place every time).
Second, the callers attempt to qualify his confirmation of that formulation, in quite often
(Line 13), is possibly evidence that he has recognized the hosts strategy and is trying to
head off an anticipated line of argument. So the hosts formulation is likely to have been
analyzed by the caller as a move which has the aim of setting him up. And third, the
formulation is the initial move in a sequence designed to challenge and defeat the callers
position, the third turn in that sequence being the hosts rebuttal in Lines 1617 (an attempt
having been made in Line 14 to go straight to that third turn rebuttalan attempt which
collides with the caller continuing to support or defend his position, perhaps as further
evidence that he understands the hosts strategy, and is trying to deflect it).
Heritage reports that in news interviews, interviewers do not respond with news marks
(particularly, oh) to answers which interviewees give to their questions (Heritage, 1985).
In order to avoid being seen to align with the IE, or in other ways to treat his or herself
as the primary recipient of the talk, IRs regularly use formulations of the gist of the IEs
prior answerformulations which do not exhibit any empathy or alignment with the IEs
position, but which topicalize or highlight an implication of what the IE has said in answer
to a prior question (Heritage, 1985). All that such formulations do, officially, is to make
explicit something in the prior answer, for the IE to confirm or disconfirm. But by highlighting some particular aspect of what the IE has just said, the IR manages to give the IE
the opportunity to comment further, or elaborate, or defend, his or her position.
(4) [News interview: TVN:Tea] (from Heritage, 1985, pp. 108109)
IE: What in fact happened was that in the course of

last year, .hh the price went up really very


sharply, .hhh and-uh the blenders did take

Comparative analysis of talk-in-interaction 257




advantage of this:-uh to obviously to raise

their prices to retailers. (0.7) .hhh They
havent been so quick in reducing prices when
the world market prices come down. (0.3) .hh
And so this means that prd.ce in the sh- the
prices in the shops have stayed up .hh really
rather higher than wed like to see them.
So you-youre really accusing them of
.hhh No theyre in business to make money
thats perfectly sensible.=Were also saying
that-uh: its not a trade which is competitive
as we would like it.=Thre four (0.2) blenders
which have together eighty-five percent of the
market .hhh and-uh were not saying they (.)
move in concert or anything like that but wed
like the trade to be a bit more competitive.

The IRs formulation in #4 is an instance of the kind of formulation that Heritage describes
as an inferentially elaborative probe (1985, pp. 108112): it invites the IE to assent to
a rather strong or dramatic version of what he has said in his previous answer. The IR
characterizes the IEs stance as being particular critical of the tea blenders: this is perhaps
designed to commit the interviewee to a stronger (and more newsworthy) version of his
position (in relation to the blenders) than he was initially prepared to adoptthe point
being to test how far the IE is prepared to go in criticizing the blenders (Heritage, 1985, p.
110). In formulating the IEs position in such dramatic, controversial or conflictual terms,
the IR might, of course, expect the IE to deny such a strong version of his position. The IR
may do so in the interests of making the item more newsworthy or controversial (if only by
getting the IE to go on record as denying something).
The final setting that I want to consider as part of this comparative exercise is that of industrial negotiations. In her study of negotiations in the workplace between management and
trades union representatives, Walker (1994) reports that formulations are used at particularly critical junctures in negotiations. Following periods of extensive discussion on a matter under dispute, one or other side (i.e., either management or union) may formulate the
position each is taking, summarizing where they now stand, in an effort to explore whether
they can reach an agreed settlementa compromise, in which well agree to x if you agree
to y. The following is a particularly transparent instance.

258 Studies in language and social interaction

(5) [PORT: WGE:2:A:314] (From a wage negotiation between management, here Andy, and
the work staffs union, represented here by Pete. Management are offering a flat rate pay
award, with no additional deals/inducements. The union is seeking a package, to include discussions about a shorter working week) (from Walker, 1994)
Er:m (1.4) er so (1.0) youre (.) com- on the

basis of feedback youre getting from (.) from


people (.) you (.) started off giving me the


impression that we were (.) still hundreds of


miles apart (.) we now seem to have come down


to a position where (.) in essence what youre


asking us to consider is the six percent on


basic which weve already offered you (.) but


you would like in addition to that for us to


consider the possibility (.) of: an increase


(.) on the (.) bonus rate (.) and to include in


any agreement we reach (.) a paragraph


indicating the willingness to (.) have dialogue


on the subject of a thirty seven hour week


(1.2) during the period of this agreement.


N:o (.) thats not whaI said.

I says in six months time to have a look at it

(.) again
You want to be specific an say six months do


I think you have to () bu I mean if you:

(.) talked about it for six months as well

Andys extended formulation, which begins with in essence what youre asking us to
consider is (Lines 615), is an attempt to summarize where they have got to in their
discussion, and to construct a package in which the union will recommend a pay rise of
6% (they had been asking for more), in return for management agreeing to enter into talks,
during the next twelve months (i.e., during the period of this agreement.), about reducing
the working week. It may be noted that although the union representative (Pete, in Line
17) objects to this formulation of what his (union) team has been demanding, his objection
is only to one aspect of it. In effect he is correcting only that part of Andys formulation
of his (Petes) position which concerns the period in which the union are seeking to have
discussions begin on the issue of a shorter working week (Lines 19, 20). Andys enquiry
(Line 21)to which Pete responds with an interpretation of the six months stipulation
which would make it more acceptable to Andy (ie. starting in six months, and talking
for six months, would bring it into line with the managements preferred timetable)is
a preagreement move towards his accepting the compromise settlement adumbrated

Comparative analysis of talk-in-interaction 259

in his formulation of what the union is now asking for. So his formulation of the other
sides position played a key role in achieving an agreement on the matter of the wage rise.
Through that formulation he was proposing a compromise that struck a balance between the
interests of the two sides, in a (successful) attempt to reach agreement.
In summary, then, formulations in these negotiations occur after there has been discussion about some issue of contention; and they are constructed to articulate what each side
may be willing to offer by way of a compromise package. That is, although ostensibly
formulating only what the other side is saying, these formulations are constructed in a turn
package which conveys what the proposer (i.e., the one doing the formulation) is willing
to agree to. Hence formulations are the objects through which a settlement is proposed.
Because of their strategic characterone side may be trying to slip in to the wording of
the formulation something in line with their preferred outcome, and something which the
other side may wish not to acceptthe other side may be cautious in confirming such formulations: for instance in #5 the formulation was rejected, and an alternative one proposed
(Lines 2125), whilst elsewhere the other side may avoid explicitly accepting or rejecting
the formulation but instead give a very qualified version of what they are saying (a version
which avoids commitment to the principle which the other sides formulation attempted to
build into the settlement).
The single instances I have shown of formulations in each of the four settings are taken
from collections of such objects in these settings. Although these are likely to be representative of such collections, one cannot yet draw firm conclusions on the basis of this preliminary review. So in comparing formulating in the different settings considered above, I am
not claiming that these are anything like definitive findings. Recalling that this is a sketch
or an exercise, all I mean to indicate is that these are the kinds of comparisons that can be
made, and the kinds of conclusions that emerge, at least on the basis of these limited data.
So the following points sketch the dimensions or properties in terms of which formulating in different settings can be compared, if we are to consider whether formulations are a
generic device of talk-in-interactionand I think it would follow from these comparisons
(if the observations on which they are based hold for large-scale data sets) that whilst
formulating is a generic practice in talk-in-interaction, the forms through which it is realized are not. These forms (objects or devices) are not unique to particular settings, so they
are not setting-specific; rather there are clusters of similarities which relate to the kinds of
activities which are managed through formulating.
It appears that formulations have different interactional functions in the different settings reviewedwhere by interactional function I mean that participants manage different activities through formulating, which is therefore associated with different kinds of
activity sequences. In psycho therapeutic sessions, the patient formulated a version of the
therapists prior comment, by way of checking her understanding of the therapists implicit
meaning (this being associated with the therapists strategy of making a comment, or asking a question, which leaves it to the patient to find for herself what the problem is, what

260 Studies in language and social interaction

should be done etc.). In the radio call-in programmes, the host formulated a (tendentious)
version of the callers argument, as an initial move in an argument sequence (confirmation
by the caller of the formulation, leading to a reductio ad absurdum by the host). It is worth
noticing the similarity between this and the formulation shown from ordinary conversation, in extract #1in which Hyla constructed a tendentious version of what Nancy was
saying, before expressing scepticism with the latters argument (or rather, with the doctors argument, with which she was aligning). News interviewers offered formulations of
interviewees prior answers, as a means to invite or encourage them to elaborate on some
particular aspect of that answer (often as a means of dramatizing the issue, and making it
more conflictual and newsworthy). And finally, in industrial negotiations one side offers a
formulation of what the other is saying/proposing, in an effort to construct a compromise
which will settle the matter under negotiationthe formulation being designed strategically to hold on to one sides preferred conditions, whilst characterizing this in a package
designed to be acceptable to the other side.
Each of these activities is central to the tasks in which participants are engaged in these
settings. They are not peripheral, epiphenomenal activities. Constructing controversy and
undermining the others argument, getting the interviewee to elaborate, figuring out the
implicit meaning in a therapists comment, and trying to arrive at a compromise settlement
with which both sides in a negotiation can agree, are each core activities in these settings.
So that formulations are associated with activity sequences which are especially characteristic of certain forms of talk-in-interaction (psychotherapeutic discourse, negotiating, etc.).
We begin to see, now, why formulations of this kind might be so rarely employed in mundane conversation: formulations are the means of conducting these activitiesand though
these activities are not unique to these settings, they are relatively restricted, in so far as
we do not, generally speaking, need to arrive at compromises after long negotiations in
mundane conversation, nor do we need to be allusive, set someone up in order to challenge
their argument, or present what theyve said in a more dramatic or newsworthy way. We
may engage in these activities in conversation from time to time,6 but they are not the kinds
of routine, organizationally salient activities that they are for the settings discussed here.
Moreover, it appears that small but significant differences in the linguistic realization of
formulations in these settings may be associated with their different activity environments.
In psychotherapy (at least, of the kind represented here), formulations are done in interrogatives, in the form of You mean. In the call-in program in which the host challenges
callers arguments, he used formulations such as What youre saying is and You seem to
be suggesting. News interviewers formulate the upshot of what an interviewee has just
answered, in So. .; and uses a wider range of verb forms, including for instance accusing, as
well as suggesting, in constructions like So youre really accusing them. And in industrial
negotiations, formulations seem restricted to forms such as (What) youre saying is, and
Youre asking us. The principal difference between these is the lexicalisation of the verb
describing the kind of saying attributed to the other. Mean occurs in psychotherapy, but
not in the other settings; saying is used in each of the others but stronger forms, such as

On being allusive in conversation, see Schegloff, 1996.

Comparative analysis of talk-in-interaction 261

accusing, predominate in news interviews and are not used in the others; and suggesting is
used in news interviews and in the radio call-in program, but not in negotiations. But there
are differences also in other features of the turn design package, So yourebeing used in
news interviews, but not in the others; and you seem to beoccurs in the radio call-in program, but apparently not in the others. I think also that there may be prosodic differences
between otherwise identical lexical verb forms, so that saying in radio call-in programmes
has different prosodic features from saying in negotiations (on prosodic aspects of the
realization of the same lexical token, see Couper-Kuhlen and Selting, 1996).
These varying patterns of lexicalising the verb with which a formulation is proposed
(i.e., the verb of saying) are associated with the different activities in each of these settings
in the following way. In psychotherapy, the patient is involved in a search for the meaning
to be found in the therapists allusive remarks or questions: the patient is endeavoring to
interpret and show that she understands what the therapist is meaning to sayhence the
lexical selection you mean with which the patient formulates a sense of the therapists prior
remark. In the other settings, most notably in industrial negotiations, there might be good
reason to avoid any suggestion that one is having to interpret what the other is saying. A
speakers purpose in formulating what the other said is to claim a certain transparency to
what they said, whereas interpretation is associated with speech that is opaque in its meaning. So in an industrial negotiation, a speaker is aiming to be able to pin on the other side
this transparent sense of what they are saying (rather than having to resort to an interpretation). The more dramatic verbs to be found in news interviews, such as accusing., are
associated with attempts by interviewers to inject something controversial or newsworthy
into the interview: of course such a verb would be alien to psychotherapy and to industrial
negotiationsin the latter case, were one side to claim that the other is accusing, this might
lead not to resolution and compromise, but rather to outright breakdown.
So in a very exploratory fashion, I have tried to show that we can track a particular
linguistic phenomenon through its use in a range of different (institutional) settings,here
the phenomenon of formulating what the other is sayingand find that the same object is
associated with different core activities in each setting. Hence the object or phenomenon is
employed in different activity sequences. Furthermore, associated with the different contexts in which it occursand by context now, I mean the different activity sequences in
which it is to be foundare patterns of different linguistic realizations of the object: for
instance, the lexicalisation of the verb with which what the other is saying is formulating
is different in the different settings/activity contexts. Hence if formulating is a generic
practice, the devices or objects through which it is realized are shaped by the activities, and
thus the settings, in which they are employed.
Couper-Kuhlen, E. and Selting, M. (1996). Prosody in conversation: Interactional studies. Cambridge: Cambridge University Press.
Davis, K. (1986). The process of problem (re)formulation in psychotherapy. Sociology of Health
and Illness, 8, 4474.
Drew, P. (1998). Comparative analysis of institutional discourse: The case of formulations. in H.
Lehti-Eklund (Ed.), Samtalsstudier: A Festschrift for Anne-Marie Londen (pp. 2942). Helsinki:

262 Studies in language and social interaction

Drew, P., & Heritage, J. (1992). (Eds.), Talk at work: Interaction in institutional settings. Cambridge: Cambridge University Press.
Garfinkel, H. and Sacks, H. (1970). On the formal structures of practical actions. In J.D. McKinney
and E.A. Tiryakan (Eds.), Theoretical sociology (pp. 337366). New York: Appleton-Century
Hak, T. (1995). Ethnomethodology and the institutional order. Human Studies, 18, 109137.
Heritage, J. (1985). Analyzing news interviews: Aspects of the production of talk for an overhearing audience. In T. Van Dijk (Ed.), Handbook of discourse analysis (Vol. 3) (pp. 95117). London: Academic Press.
Heritage, J. and Watson, R. (1979). Formulations as conversational objects. In G. Psathas (Ed.),
Everyday language: Studies in ethnomethodology (pp. 123162). New York: Irvington.
Hopper, R. (1995). Studying conversational interaction in institutions, Communication Yearbook,
18, 371380.
Hutchby, I. (1996). Confrontation talk: Arguments, asymmetries and power on talk radio. Mahwah,
NJ: Erlbaum.
Schegloff, E.A. (1997). Confirming allusions: Toward an empirical account of action. American
Journal of Sociology, 102, 161216.
Walker, E. (1994). Negotiating work. Unpublished doctoral thesis, University of York.

Conversational Socializing on Marine VHF Radio:
Adapting Laughter and Other Practices to the
Technology in Use1
Robert E.Sanders
University at Albany, SUNY
Marine VHF radios are the primary medium for communication between vessels in coastal
waters and between vessels and shore facilities.2 They are standard equipment on commercial vessels, and widely but not universally installed on recreational vessels. Unlike
CB radio, marine VHF radio is not intended as a folk medium. It is used for official communications by law enforcement (the Coast Guard, marine police and harbormasters), in
search and safety operations, by towing/salvage services, and commercial operations in
coastal waters (drawbridge operators, port operations and traffic control). There are prescribed protocols and languagedrawn from long-standing procedures for signaling at
seafor hailing other stations, repeating information, acknowledging transmissions and
ending them, requesting priority on a channel, prefacing messages to index their urgency,
pronouncing some words and numbers (see-lonce for silence and niner for nine) and
pronouncing letters when spelling (Alpha, BravoYankee Zulu), and so on.3 And
the Federal Communications Commission (FCC) has designated for whom, or for what
purpose, each of a marine VHF radios approximately 55 talk-receive channels is reserved
(including two channels for hailing, one for search and safety, one channel for digitized
emergency broadcasts, six channels for the Coast Guard, eleven channels for port operations and traffic control, ten channels for commercial users, six channels for marinas and
recreational boaters, and nine channels for connecting to a landside telephone line).
I cant think of a more fitting place for this study than in a volume in honor of Robert Hopper. Its
not so much because it resonates with his interest in telephone conversation, but more his broader
interest in the adaptation of conversational particulars to the technological environment. I have in
mind at least one project of his I know of, some relatively early work on the way persons playing
pinball adjusted their turn constructional practices to accommodate the interruptions to be expected
from the noises and activity of playing the game.
A growing number of boaters are also using cell phones for communication in coastal waters,
and some recreational boaters use a cell phone exclusively. However, marine VHF radios are not
becoming obsolete. The broadcast capabilities of radio make it essential for vessels in distress to
call for help from anyone in the vicinity, for coordinating search and rescue operations, for making
known the location of obstructions, or movements of vessels, in channels and harbors, and so on.
Perhaps in conjunction with the FCC having stopped requiring recreational users to have radio
station licenses, a number of recreational boaters do not observe these restrictions and protocols of
use (at least in the waters of Long Island Sound, and probably around the United States generally).
Many have imported CB-radio jargon and protocols, and, as on CB radio, use any clear channel for
transmission rather than just the ones reserved for them.

264 Studies in language and social interaction

However, in practice marine radios are not consistently used as the FCC intends, that
is, for carrying out the business of safely operating vessels at sea and providing marine
services. They are also commonly used for matters of logistics and convenience, especially by recreational boaters (contacting marinas to arrange for overnight dockage; calling
water taxis for transportation ashore when moored; contacting fellow boaters about mooring together at days end; contacting others who are fishing to exchange information about
where the fish are; and so on). And sometimes, furthest from FCC intent, these radios are
not used in service of marine operations at all, but as a folk medium for conversational
socializing when there is no particular business at hand.
Because the phenomena of interest here come mainly from this genre I call conversational
socializing, and because the genre itself is of interest, it needs a brief exposition. When
persons engage in conversational socializing, they talk about matters that are entertaining,
or present each other with news items, commentary, and gossip about subjects or persons
of mutual interest. It is definitive of such talk that it not be material to transacting business on any practical matter, nor for bringing about any particular resultexcept to have
spent time together entertainingly. Use of marine VHF radios for conversational socializing is most widespread among recreational boaters, but occasionally tugboat captains
engage in it while in transit between harbors or while docked waiting for a barge to be
loaded or unloaded.
It is while auditing conversational socializing on marine VHP radio that I observed
the two practices of interest here, involving gaps and conversational discontinuities, and
laughter and other affiliative responses. Although the practices I observed in that regard
might occur in other media, other genres of conversation, other cultures, what is of
interest here is how those phenomena reflect and have been adapted to the operational
contingencies of the medium of two-way radio.
There are two prominent technological differences between the telephone and two-way
radios in general that seem responsible for the phenomena I examine below. First, the radio
technology makes it physically impossible for more than one person at a time to occupy
the floor. Anyone transmitting cannot hear (i.e., receive) others who are transmitting at the
same time. Further, if two persons in a conversation speak (transmit) at the same time, they
cannot tell they are doing so as long as they continue speaking. And if a person in a conversation and a third party outside that conversation transmit at the same time, only the one
with the strongest or closest signal can be heard by the other(s) in the conversation. Hence,
third persons outside a conversation can inadvertently step on (block) the transmission of
someone with whom one is speaking, with neither of the speaking persons aware of it and
the person(s) listening unable to intervene.
Second, unlike other aural media, to make oneself heard one has to do more than just
vocalize. One has to take the prior, physical steps of bringing the microphone up near ones
mouth, and pressing and holding down the microphones transmit key. This alone delays

Conversational socializing on marine VHP radio 265

response and reduces its spontaneity. The spontaneity of responses is further reduced by
being unable to make oneself heard until the other person stops transmitting. And there is
a potential for further delay in responding if something occurs just then that is material to
operating ones vessel, or to fishing, so that one may temporarily not have a free hand to
operate the microphone even if one could otherwise have continued talking while taking
action at the same time.
Because conversational socializing comprises a minority of the transmissions being made
at any time on marine radio, and they can occur on any of about 40 talk-receive channels on
which boaters contact each other,4 locating and recording such conversations is somewhat
happenstance (barring the use of 40 receivers each on a separate channel and 40 recorders).
I relied on a scanning radio: My radio completes a scan of all talk-receive channels roughly
once every 3 seconds unless it comes to a channel on which someone is transmitting. When
transmission on a channel is detected, scanning is suspended for 4 seconds so that the
transmission can be listened to, and then scanning resumes unless it is manually stopped.
Conversations thus got found in that way and recorded, usually after they had already
begun. Further, in recording a conversation, other conversations that may have been taking
place at the same time on other channels necessarily went undetected. Sometimes when I
resumed scanning after a conversation had been recorded, I came upon a final fragment of
another conversation that exhibited a practice in which I had become interested, too late
to record it. I took notes on some of these. Hence, the methods I employed do not make it
possible to estimate the relative frequency of the practices of interest here, but they occur
more often than I was able to record.
I should note (given Hoppers interest in gender and communication) that the great
majority of speakers on marine VHP radio are men. This is probably an artifact of the extent
to which men dominate boating (though women have the option: Two boats of approximately 46 on my dock are owned and operated by women). On those relatively few occasions when women use the radio, it is often as a stand-in while the captain is engaged
in operation of their vessel. Hence, with the exception of one example in my notes, the
conversations in which the phenomena of interest were exhibited were all between men. It
remains to be seen whether this is incidental to the corpus or reflects a gender bias in the
medium itself, and whether women would adopt different conversational practices.
The transcripts of these social conversations use notation conventions developed by Jefferson (e.g., in Atkinson and Heritage, 1984), but with two slight modifications. First, gaps
of less than 0.5 seconds are not hearable as delays in response in this medium and were not
Although there are a total of 55 talk-receive channels on marine VHP radios, some channels
are not potential sources of conversation between boaters. A few of these are now in use by civil
authorities ashore and are avoided by boaters. On the nine channels dedicated to connecting with
the land-based telephone system, only the land-based side of the conversation is hearable; the
marine telephone company that provides this service sends out a masking signal that prevents the
boaters transmission from being heard on other VHP radios.

266 Studies in language and social interaction

noted: It takes at least that long for the next speaker to register that the prior transmission
has ended, and to then press his or her transmit key and begin speaking. Second, the symbol
# denotes the electronic click sound made at the end of transmissions when the current speakers transmit key is released: Notating this serves as a reminder that the ends of
transmissions are audible, and allows notating any occurrences of dead air between the
speakers last utterance and release of the transmit key. Finally, laughter was transcribed
with symbols intended to more closely reflect its actual phonetic qualities, and it appears
in boldface to set it apart visually from the surrounding talk, not to indicate any relatively
greater loudness.
It is not uncommon in the conversations I recorded for there to be gaps between conversational turns of 56 seconds and longer. Based on my own experience in the region and
subcultures of the Northeastern United States, these gaps are much longer than what these
same speakers would generally find tolerable in face-to-face or telephone conversations
ashore. Yet on marine radio, these prolonged gaps are almost always tolerated and not oriented to as breaches, something I attribute to the operating requirements of this medium,
as explained below. Persons waiting for a response often do not prompt the other at all,
and if they do, it is not as quickly or directly as they might in a different medium. Moreover, the party waiting for a response is careful generally to avoid making the other person
accountable for the gap. At the furthest extreme, these gaps are sometimes terminal: The
conversation just ends for lack of anyone taking a next turn, without any closing.
In the following examples, gaps of 9 seconds and longer went unremarked, and ended
when the next speaker finally did respond:
M2 :

M2 :
M3 :



M6 :

Hey:, the guys on the boat here, they invited us to go out to Hooters tonight
theyre so happy. #
That right? Theyre gonna go hoot n holler. #
Yeah:, they want Ja:ck n Gary myself, they all (want) us to go up to Hooters. #
Read todays Newsday? #
Okay, then you must be in sight o:f me. #
I got a visual on Penfield reef now:.#
Cant Ginny talk % im into: (.) goin out to eat? #
Ah, I wish she would, itd be so much nicer. #
Well take im. (.) Well pay for im, well do anything. #
Thats right, lets- lets take up a collection,= well seh-h-nd im to Alaska h- #
heh-heh-heh-heh #
Howd doctor Mike do today? #

Conversational socializing on marine VHP radio 267

The likely reason for the occurrence of such notable gaps and the evident respect they are
given is that, as noted, the technology requires the speaker to have one free hand to hold
the microphone and operate the transmit key. Having a hand free is not something one can
count on from anyone currently operating a vessel or fishing. This is evident even in the
one instance in my corpus when the prior speaker did directly prompt the next speaker
after a gap of 6.2 seconds (Example 4 below: Diyou copy that, i.e., did you receive my
transmission). The next speaker replied by citing his current attention to fishing as a reason
not only for the gap but for thereupon ending the conversation. This was accepted without
protest and the conversation ended exactly then:


Ill tell ya, you gotta use these ci.rcle hooks. (0.5) Theyre great. (0.5)
You stick the rod in the rod holder, ya put the brake on: (0.5) -hh uh: ya
hook (.) ninety nine percent of all yr fish in the (lip). (0.7) Ah:: theyre
a son of a gun t get ou:t, (0.5) but- ya dont gut hook many fish at all, ya
get a really () fish. (0.5) Works pretty nice. (0.2) over. (.) over. #


Diyou copy that, (Dan)? (0.5) #

Ah:, I gotta leave now, like we gotta fish on. (0.7) Talk tya later. (0.2)
over. #
Awright. (0.2) Coin back to nine ((nine is the standby/hail ing channel; this
declaration is equivalent to hanging up a telephone)). #

Long gaps were not always ended by the next speaker eventually taking a turn, as in Examples 13. Although it is atypical as in Example 4 to directly prompt the next speaker, same
speaker sometimes resumed after a gap to prompt the other indirectly for a response. The
person seeking a response after a gap begins hailing the silent vessel, as one would do in
making initial contact. This implicates that it is not a matter of a response delayed too long,
but of having lost contact, an accidental happenstance that warrants an effort to reestablish

M3 :

M4 :


Okay, then you must be in sight o:f me. #

I got a visual on Penfield reef now:. #
Alone Again, Alone Again, Happy Days. # ((a hail to the vessel Alone Again to
answer the vessel Happy Days))


M4: Alone Again, Alone Again, ya got Happy Day#


M3 :
How farre you from Penfield? #

Nine miles n hour. #


268 Studies in language and social interaction







(Whatre) you doin? #

(it) like Jo:hn. #
(uh) I got eight point o:ne. #
((possibly his speed))
(Any ideas?) # ((possibly a transmission from a third party in another conversation))
Bout fifteen hunderd. # ((possible reference to RPM))
(Glitterbox), whe:re are ya?

Finally, as was noted, gaps sometimes were not closed at all; the conversation just stopped
continuing. There are several examples of this in my corpus. A case from my notes in
which the conversation does continue after a gap reveals an orientation by both parties
to the potential for discontinuation after such a gap. In arranging for their boats to tie up
together at anchor that night, M11 expresses the joking concern to F1 that F1s husband
might snore too loudly:



Yeah, but I dunno. Do you think well get any sleep? You know,
Franks snoring and all.
((710 second gap))
Switching back to nine, ((nine is the standby/hail ing channel))
No, sorry Tom, we were just working on a comeback.

Mils announcement after the gap that he was switching to the standby/hailing channel is the
equivalent of simultaneously saying goodbye and physically hanging up the telephoneit
is not the same as opening up a closing where one then waits for the other to respond. That
M11 did this after a notable gap displays an orientation to gaps as possibly terminal. However, Fl responded anyway (probably aware that while persons often do switch channels
right when they make such announcements, they are sometimes slow to do so and may still
be listening). What is important about F1s response is that she apologizespresumably
for producing a gap that it would be warranted for M11 to infer was terminal. She continues
by giving an account for the gap as interactionally produced, thus canceling the implicature
of termination.
There is thus a relatively greater tolerance for gaps during conversations on marine VHF
radio, and the potential for discontinuations, than one is likely to find in conversations
ashore among these same speakers. However, note that similar tolerances and potentials
have been observed among Native Americans, and attributed to the communal value they
give to privacy and autonomy (Basso, 1979; Scollon and Scollon, 1981). With two different accounts of the same practiceattributed to the practicalities of boating on one hand,
and to Native American communal values on the otherit would be parsimonious to find a
common denominator. As it happens, there is one. It seems that in both cases, if talk occurs

Conversational socializing on marine VHP radio 269

at the same time speakers are engaged in an activity that the community of speakers gives
precedence, the talk will be suspended whenever it interferes with that activity, and the
resulting gaps and discontinuities tolerated. For boaters, there are certain practicalities that
are given precedence over talk. Perhaps what has been observed among Native Americans
arises from their giving most or perhaps all other concurrent activities priority over conversation, whether these are practical/material activities, or spiritual or cognitive ones. The
cultural aspect of the tolerance for gaps and discontinuities, then, is not about values placed
on privacy or autonomy, for example, or marine exigencies. Rather, it is about the priority
that the community gives to conversation relative to specific other activities that persons
can be engaged in concurrently.
In conversation generally, whether on a two-way radio or not, the presumption is that ones
substantive reply to, or follow-up on, what the current speaker is saying in the moment
will be withheld until it is ones turn to speak (Sacks, Schegloff, and Jefferson, 1974). A
notable exception is affiliative responsesfor example, acknowledgment tokens, newsmarks, laughterspontaneous responses to what is just then being said that are ordinarily produced by the listener while the current speaker still has the floor. However, in
conversation on two-way radios, not only substantive but affiliative responses have to be
saved up until one has clear air in which to transmit. The opportunity for and the spontaneity of affiliative responses is thus greatly reduced during conversational socializing on
marine VHP radio.
This has the apparent effect of pruning out some, and simplifying other, affiliative
responses in radio conversations. In my corpus, there are few or no newsmarks, acknowledgment tokens, supportive assessments, and so on. Speakers did sometimes produce linguistically elaborated acknowledgments at the next opportunity (in this corpus primarily,
I gotcha, I hear ya, I copy that and Yeah::), but they lack a functional complexity on
two-way radio that they can acquire in conversation ordinarily. In other media of aural
conversation, such elaborated acknowledgment tokens not only have an affiliative function
but a turn coordination function. If a person produces simpler back channel acknowledgment tokens during the current speakers turn, and then produces an elaborated form, it
often implicates the speakers readiness to take the floor just then and produce a full turn
at speaking (Drummond and Hopper, 1993; Jefferson, 1993). But on two-way radio it is
impossible to bearably make simpler back channel responses while the others turn is in
progress, so that producing an elaborated acknowledgment token cannot display a change
of state. Besides that, if one has clear air to transmit the elaborated token and does so,
one already has the floor and the issue of speakership is moot anyway. Accordingly, such
acknowledgment tokens can only serve an affiliative function on two-way radio.
The one affiliative response that is not pruned out or functionally simplified on marine
VHP radio is laughter. At times, persons who are conversationally socializing via marine
radio press down the transmit key, they transmit laughter, then end that transmission. In
that case, they took the special steps needed to transmit just to make laughter heard. This
in itself is evidence that speakers are capable of being knowing and deliberate, even

270 Studies in language and social interaction

calculating, about the social functionality of laughter, more so than previous analysis has
revealed. Further evidence of this is presented below.
Ordinarily the functionality of laughter as an affiliative response depends on its being, or
being made to seem, an immediate and spontaneous response to what occasions it (Glenn,
1989, 1991/1992; Jefferson, 1979, 1984). But this is impossible to display on a two-way
radio. One has to wait for clear air, then transmit, so at minimum there is an unavoidable
micro-delay before laughter is heard, and a marked deliberateness about making it hearable. Ordinarily this would make laughter seem artificial.5 But laughter is produced on twoway radio anyway, moreover with the apparent presumption that it is genuine unless there
is reason to think otherwise. The evidence for this is that persons laughing on marine radio
sometimes take special measures to register their laughter as artificial, and conversely,
sometimes take special measures to establish it as genuine when there is a circumstantial
reason to doubt it. It is these phenomena that are of particular interest here.
Let us posit that a laugh response on a two-way radio is presumed genuine the extent to
which it has the requisite vocal qualities of genuine laughter, and is transmitted immediately (though not spontaneously)that is, at the first opportunityafter it is occasioned.
It is not any more difficult to produce laughter with the requisite vocal qualities on two-way
radio than any other medium, especially if it actually is genuine. The relative immediacy
of a laugh response will be enhanced the extent to which the current speaker ends his
transmission just when laughter is occasioned, and this is common. In example 8 (from
the same excerpt as example 3), there is actually a gap of a few l0ths of a second between
M6s occasioning remark and laugh particles, and M5s laugh response, but my own experience is that such gaps do not register as a delay in response when one is accustomed to
the mechanics of two-way radio:




Cant Ginny talk im i.nto: (.) goin out to eat? #

Ah, I wish she would, itd be so much nicer. #
Well take im. (.) Well pay for im, well do anything. #
Thats right, lets- lets take up a collection,= well seh-h-nd im to Alaska
h- #
heh-heh-heh-heh #

However, it was not unusual that when the current speaker ended transmission as soon as
he or she occasioned laughter, there was a marked delay before the laugh response was
transmitted, sometimes several seconds in duration. In itself, in any other medium, this
The one exception would be if laughter were delayed because the person did not immediately
get it. But when this happens, there is usually a marked display of getting it when the delayed
laughter begins, even verbalizations such as Oh:: I:::: get it, that function to cancel any implicature that the delayed laughter is artificial. Although there is no obvious reason why delays for that
reason would not occur (or be feigned) over marine radio as in other aural media, and be marked in
that way, this did not occur in my corpus.

Conversational socializing on marine VHP radio 271

would mark the laughter as artificial. Although extra effort could be made to establish
laughter as being genuine anyway, as we see in examples 12 and 13 below, no such effort
was made in the following examples, 9 and 10, despite a notable gap between the occasioning utterance and the laughter in response. Of course it is possible that in these instances
the persons responding did not care whether their laughter seemed genuine or not. But then
why take the trouble to transmit it? The alternative is to suppose that delays in transmission
on marine radio are accepted as potentially unavoidable, canceling the implicature that
delayed laughter is artificial.



M12 2







(in) Anthony. #
Come in Anthony. #
Chuck- pa:l, how are ya?
That voice. #
I missed that voice. #
hah-hah-hah-hah #

It doesnt work that way. (0.2) Set the anchor, (.) two guys on top, (.) pull
against the anchor, it works.#
heh-heh-heh #

In contrast, there were instances when active steps were taken to mark transmitted laughter
as artificial. In general, laughter can be made to seem artificial (not genuine, insincere) by
positioning it so that it is bearably delayed or withheld, and/or by giving it vocal qualities
that are not natural. Of course, delaying or withholding laughter is not distinguishable on
two-way radio from being unable to immediately transmit it, and so persons have to rely on
vocal quality alone to register laughter as artificial. In example 8, M17 produced laughter
that was too loud and intense a response to what occasioned it, his final laugh particle was
artificially elongated, and he gave the laughter a guttural quality reminiscent of the villains
laugh in an old movie:

M16 :




Hey (A1), whaddya suppose hes doin over there? (0.5) #

Probably (checkin on our) maneuvers. #
Yeah::. Thats okay:, we know how ttake care o that, right pal?
*YEAH: : : : .HEH-HEH-HEH:::::::::::::*#

272 Studies in language and social interaction

Conversely, one does occasionally find that speakers take steps to establish their laughter
as genuine. In general, the laugher was marked as genuine by means of transmitting it
twice, separated by an interval. In the instances when this happened, there was reason to
doubt that the laughing person would have been genuinely amused by what occasioned the
laughter. In example 12 M18s laughter could potentially be regarded as insincere because
it came in the context of a mild disagreement, where it expressed M18s affiliation with
M19 on something that Ml8 had been disputing. In their conversation, Ml8 and Ml9, apparently commercial fishermen or lobstermen, disagreed whether a supplier of theirs treated
customers badly because of the business pressures involved (as Ml8 contended) or because
he was a hateful person in his own right (as M19 contended). After Ml9 finds a pithy way
to make his point that it was this persons intrinsic qualities that made him hateful, not the
business context, M18 responds with laughter even though he presumably disagrees. Note
that he transmits laughter twice in two contiguous transmissions separated by a gap of 4.2
seconds, even though nothing new (interactively) occurred in that interval to occasion the
second transmission.







Well, dont forget, too, I mean, uh:::, ya know -hhh n always sitting
there trying to collect money from thirty different guys, n uh::: you
take any thirty guys Zs gonna be: : hh a certain
amount of em thata al:ways pay their bill on time=n theres gonna be
a certain amoun:t v em yalways gonna have t chase down:: n look
for. hhh Ya know, n I think thats ((mic noise)) where we had problems. #
Was always the same guys. #
W ha-ha yeah:: h-h- ya know, I:: -hhh I:: Im not- pickin out any names
or anything like that, but I mean thats just the rule of thumb, ya know:
hhh ya know ya (never) gonna have thirty people all make their payments on the exact same time or be prompt.= Theres always ((noise))
fusing, ya know, hhh n I think thats where a lot of problems used to
stem from, -hhh N Im sure Pete had to do that with some of us too,
but- ya know, (.) now he doesnt have to worry about as many. #
(), when he- (was/noise) at Bayshore we hated him. (0.2) I mean(.) we
havent changed. We still hate him at (Jethreys). #
hu-hu-hu yeah-h-h- gotchah-h- #
eh: heh-heh- #

By transmitting his laughter twice, Ml8 gives the appearance of finding M19s quip so
funny that he actually sustained laughter during the interval between his two transmissions
(or at least the appearance that on reflection he had found M19s quip funny again and

Conversational socializing on marine VHP radio 273

resumed laughing). He thereby marks his first laugh response as genuine by transmitting
the second one.
In the instance that follows, M20 and M21 are tugboat captains who evidently have
known each other for a time, but have been out of touch for several years. In the course
of catching up on personal news, M21 reports that he has a young daughter now. He then
complains that continuing to have children would leave him without a seat at the dinner
table, and M20 comments that he should therefore not have more children. At that juncture,
a potentially delicate matter is introduced. I take M21s response as an indirect disclosure
that he had a vasectomy ((I) better not [have more children]. (0.2) Ill be after that doctor
with a baseball ba:t.). M20 responds with laughter and then moves to closing. A second laugh particle is transmitted after they have closed, though I could not identify who
produced it.








Ho::ly smoke, I havent seen you in awhile. #

Yeah, I got two and a half, (almost) three year old daughter,=shell be three
in uh: : (0.5) (just) before Christmas. (0.5) (little Emma). (#)
((1.5/garbled utterance, either a continued transmission by M21 or a transmission from some third party, ending with rising inflection)).
Su: re! #
(0.5/open mic) ( ), just screw yourself right out of a seat at the table ya keep
goin. #
(Then) dont have no mo:re. #
(I) better not. (0.2) Ill be after that doctor with a baseball ba:t. #
WA- ha-ha- -hhh I:: gotcha. (Alright there, Rod), (.) hh you have a good
trip back in there. #
Yah, okay Steve, well be talkin to ya.=((smiley voice)) Take ca::re, (keep
your sanity with the cattle). ((Tugboat captains sometimes refer to pleasure
boats or boaters as cattle, perhaps because they dot the landscape and are
slow to move out of the way)) (0.5) Well talk to ya. #
heh-heh (0.2) #

Of interest here is the second laughter token in the transcripts last line, transmitted after
they closed. No matter which of them transmitted it, that token has a similar functionality. The only evident laughable is M21s allusion to his vasectomy, where M20 did laugh.
Hence, given that that second laugh particle was transmitted long after it was occasioned,
it displays sustained amusement, as in the prior example. If it was M20 who produced the
laughter, then like M18 8 in example 12, he affirmed the genuineness of his laughter about
a matter he might have not found amusing. However, if it was M21 who transmitted that
laughter, it could not mark previous laughter as genuine because he had not previously
laughed (aside from a smiley voice in closing). But it would affirm that he had alluded
to his vasectomy as a joking and not a delicate matter, and underscore his own residual

274 Studies in language and social inter