You are on page 1of 290

New Adventures in Language and Interaction

Pragmatics & Beyond New Series (P&BNS)

Pragmatics & Beyond New Series is a continuation of Pragmatics & Beyond and
its Companion Series. The New Series offers a selection of high quality work
covering the full richness of Pragmatics as an interdisciplinary field, within
language sciences.

Editor Associate Editor

Anita Fetzer Andreas H. Jucker
University of Würzburg University of Zurich

Founding Editors
Jacob L. Mey Herman Parret Jef Verschueren
University of Southern Belgian National Science Belgian National Science
Denmark Foundation, Universities of Foundation,
Louvain and Antwerp University of Antwerp

Editorial Board
Robyn Carston Sachiko Ide Deborah Schiffrin
University College London Japan Women’s University Georgetown University
Thorstein Fretheim Kuniyoshi Kataoka Paul Osamu Takahara
University of Trondheim Aichi University Kobe City University of
Miriam A. Locher Foreign Studies
John C. Heritage
University of California at Los Universität Basel Sandra A. Thompson
Angeles Sophia S.A. Marmaridou University of California at
University of Athens Santa Barbara
Susan C. Herring
Indiana University Srikant Sarangi Teun A. van Dijk
Cardiff University Universitat Pompeu Fabra,
Masako K. Hiraga
St. Paul’s (Rikkyo) University Marina Sbisà
University of Trieste Yunxia Zhu
The University of Queensland

Volume 196
New Adventures in Language and Interaction
Edited by Jürgen Streeck
New Adventures in Language
and Interaction

Edited by

Jürgen Streeck
The University of Texas at Austin

John Benjamins Publishing Company

The paper used in this publication meets the minimum requirements of

American National Standard for Information Sciences – Permanence of

Paper for Printed Library Materials, ansi z39.48-1984.

Library of Congress Cataloging-in-Publication Data

New adventures in language and interaction / edited by Jürgen Streeck.

p. cm. (Pragmatics & Beyond New Series, issn 0922-842X ; v. 196)
Includes bibliographical references and index.
1. Sociolinguistics. 2. Social interaction. 3. Conversation analysis. I. Streeck, Jürgen.
P40.N425â•…â•… 2010
306.44--dc22 2010010772
isbn 978 90 272 5600 3 (Hb ; alk. paper)
isbn 978 90 272 8814 1 (Eb)

© 2010 – John Benjamins B.V.

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

New adventures in language and interaction 1

Jürgen Streeck

Interlocutory logic: A unified framework for studying

conversational interaction 9
Alain Trognon and Martine Batt

Beyond symbols: Interaction and the enslavement principle 47

Stephen J. Cowley

The case for an eclectic approach to discourse-in-interaction 71

Catherine Kerbrat-Orecchioni

Grammar: A neglected resource in interaction analysis? 99

Peter Muntigl and Eija Ventola

Researching intercultural communication: Discourse tactics

in non-egalitarian contexts 125
Angel Lin

Studying interaction in order to cultivate communicative practices:

Action-implicative discourse analysis 145
Karen Tracy and Robert T. Craig

Healthcare interaction as an expert communicative system:

An activity analysis perspective 167
Srikant Sarangi

Interacting with difficulty: The case of aphasia 199

Elizabeth Armstrong and Alison Ferguson

Ecologies of gesture 223

Jürgen Streeck
 New Adventures in Language and Interaction

The neglected listener: Issues of theory and practice

in transcription from video in interaction analysis 243
Frederick Erickson €

Dialogical dynamics: Inside the moment of speaking 257

John Shotter

Author index 273

Subject index 275
New adventures in language and interaction

Jürgen Streeck
The University of Texas at Austin

This book has developed from an idea by Carlo Prevignano and Paul Thibault to
have a diverse group of well-known researchers who are engaged in the empirical
study of language and interaction discuss the start of the art of the field. All of the
contributors are either influenced by or develop their own line of inquiry in criti-
cal dialogue with conversation analysis and/or interactional sociolinguistics, as
conceived by John Gumperz. This book thus complements the prior discussions
with Emanuel Schegloff (Prevignano & Thibault 2003) and John Gumperz
(Eerdmans, Prevignano & Thibault 2003) that these scholars have published. The
resulting volume gives an impression of how diverse the field has become.
While all of the contributors see their work as grounded in what we may now
call the interactionist canon – G.H. Mead, the late Wittgenstein, perhaps Vygostky
and Bakhtin, Bateson, maybe Garfinkel, but above all Goffman (see, for example,
the contributions by Trognon and Batt, Streeck, and Shotter) – they have branched
out from this shared ground in a number of different directions:
– by subjecting these dominant paradigms in language and interaction research
– conversation analysis (CA) and interactional sociolinguistics (IS) – to a crit-
ical review and formulating their own distinct theoretical edifice, vision, or
methodology in contradistinction to them, sometimes by giving these a
name that sets them off from those methodologies, e.g.€ interlocutory logic
(Trognon and Batt) or action-implicative discourse analysis (Tracy and Craig);
Kerbrat-Orecchioni pleads for an eclectic methodology, while Shotter articu-
lates a philosophy of dialogical dynamics as it has emerged from his research
as well as from the interplay of utterances by Vygostky and Wittgenstein that
has unfolded in his work; or
– by taking these methodologies, notably conversation analysis, and integrating
them with separate, differently focused but philosophically compatible meth-
odologies, such as systemic-functional linguistics as proposed by M.A.K.
Halliday (Muntigl and Ventola); or
 Jürgen Streeck

– by incorporating previously understudied components such as prosody

(Cowley) listener action and temporal coordination (Erickson) and gesture
(Streeck), and using findings from their research to revise received notions in
interactionism (e.g.€ context) and introducing new ones (such as resonance)
that emphasize embodied aspects of intersubjectivity in interaction.
Other contributors are committed to elucidating interactional difficulties that re-
sult from impairments (Armstrong and Ferguson) and cultural conflict (Lin), or to
give guidance or provide representations of dialogue for the critical reflection of
practitioners in such fields of communicative practice as health-care (Sarangi),
therapy (Cowley, Muntigl and Ventola, Shotter), and politics (Tracy and Craig).
The adventures that await those who enter the microcosm of human interac-
tion thus offer challenges of multiple kinds: the need to revise one’s guiding as-
sumptions and find new bearings in a changing, interdisciplinary landscape, but
also to make sense of little-known fields of practice which are also undergoing
change, as well as making practitioners feel a greater need to reflect on what is go-
ing on. All of the contributions combine reflections of method with an assessment
of some aspect of the state of the art and illustrate their reflections with limited
presentations and analyses of cases.
Specifically, Alain Trognon and Martine Batt exemplify one discernible trend
in contemporary work on language and interaction, which is to seek to integrate
some of the different existing strands of research as well as their conceptual foun-
dations into a single integrated theoretical and analytic framework. Central to
their endeavor are Goffman’s notion of the interaction order (Goffman 1983) and
Schegloff ’s dictum that interaction (or, rather, talk in interaction) is the “natural
habitat” for the deployment and development of language (Schegloff 1991). The
interaction order, whose “backbone” is sequentiality, underwrites mutual under-
standing (intersubjectivity) via language, but traditional conceptions that deprive
language use from its scaffolding by institutional orders of interaction unfailingly
overestimate the contributions of linguistic structures to human understanding.
Trognon and Batt propose interlocutory logic as a unified framework for studying
cognition in interaction. It comprises formal procedures to (a) represent and mod-
el the sociolinguistic organization of interlocutions (i.e. dialogues consisting of
social action sequences), and (b) generate hypotheses about “subjective processes
at the cognitive level” and study “the conversational emergence of both declarative
and procedural knowledge”. They exemplify this framework with an analysis of the
talk during brief “hand-overs” between work-shifts in a paper factory.
Stephen Cowley’s contribution breaks with more orthodoxies than most oth-
ers. His point of departure is, like Trognon and Batt’s, “the exaggerated importance
of linguistic forms” in traditional research and theory, but in his rejection of this
New adventures in language and interaction 

tradition he is particularly keen on re-somaticizing language and communication

and to demonstrate the “importance of patterns beyond symbols”. He contends
that traditional conceptions – including conversation analysis and, to a more lim-
ited extend, Gumperz with his notion that contextualization cues steer the inter-
pretation of situated uses of symbols – misconstrue our participation in embodied
events by construing it in analogy to how we read pictures: as recognition of pat-
terns whose meanings can be looked up in some social or mental lexicon. What
Cowley offers is a dynamic model of self-organizing bodies that continuously res-
onate with unpredictable events and dynamic changes in co-acting bodies: a body’s
“pico-level resonance” constantly responds to and, in turn, impacts another body.
Cowley offers a case-analysis of dynamic prosodic processes in an Italian conver-
sation; he models meaning-making as an embodied phenomenon, in line with a
perspective that is grounded in the “bio-mechanics” and “bio-semiotics” of inter-
action, in light of which traditional units such as words and utterances, even inter-
action itself, appear as folk-constructs.
Catherine Kerbrat-Orecchioni makes the case for methodological eclecticism,
presenting ideas towards an approach that incorporates concepts from various
versions of discourse analysis, pragmatics, politeness theory, speech act theory,
and conversation analysis. She develops these ideas mainly through a critique of
core concepts in conversation analysis. Thus, she takes issue with the notion of
adjacency pair, pointing out that the acts that are connected in sequences may of-
ten be far apart (cf.€Schegloff, 2007, which also addresses this issue), emphasizing
that the units that make up sequences are not turns, but actions (or speech acts),
and arguing that the notion of pair betrays a CA bias toward two-party conversa-
tion. Kerbrat-Orecchioni also separates the dimensions of ordering and conditional
relevance in sequence organization and proposes a reinterpretation in light of po-
liteness theory (Brown & Levinson 1987) of various preferences structures that con-
versation analysts have uncovered (e.g.€the preference for agreement; Sacks 1987).
She is critical of what we may call the “maxim of contextual parsimony”, advanced
in particular by Schegloff (1987), according to which the situated relevance of
some category of context or identity (gender, place, social class, etc.) should not be
presumed as relevant unless this relevance has in fact been made apparent by the
parties: many dimensions of context are assumed by the parties even if they are not
made apparent in the discourse. Interaction analysis, therefore, is always interpre-
tive, and inevitably the parties will draw upon “external” parameters of context,
e.g.€their respective social backgrounds. Kerbrat-Orecchioni exemplifies her eclec-
tic approach, which “draws on descriptive resources from different fields and puts
them together”, in an analysis of a strategically misplaced greeting during a tele-
vised interchange between the French politicians Sarkozy and Le Pen in 2003.
 Jürgen Streeck

Peter Muntigl and Eija Ventona take a very different turn than, for example,
Cowley and Trognon and Batt: instead of reducing or questioning the relevance of
grammar in the achievement of intersubjectivity in interaction, they exemplify its
relevance – the relevance of grammatical choices – for the development of social-
interactional processes and interpersonal life. Their approach, which draws on
Halliday’s systemic-functional linguistics is compatible with recent work by conver-
sation analysts and interactional linguists on positionally sensitive grammar, in-
volving a type of analysis of grammatical structures and devices that pays attention
to the positions within emergent turns, sequences, and activities where these struc-
tures and devices can be deployed. Both systemic-functional linguistics and con-
versation analysis, they point out, “insist on examining speakers’ meaning-making
resources”, specifically the constructional alternatives that are available to them at
a given point in an interaction and which include such choices as that between ac-
tive and passive voice or action-, achievement-, and relational verbs. Aligning with
the “behaviorist” view of the relationship between linguistic resources and socio-
cognitive behavior espoused by Ryle (1949) and Wittgenstein (1953; see Coulter
1989), they show that grammatical resources are not simply resources for making
clauses, but that choosing one type of clause over another means enacting a differ-
ent action, construing an experience in different ways, and offering different types
of relationship.€This conception is developed through the analysis of a brief excerpt
from a couple’s therapy in which the therapist successively alters the grammatical
frame in terms of which the husband’s habits are construed.
Angel Lin’s chapter reconciles the perspective of Gumperz’ interactional socio-
linguistics with the methods and procedures of conversation analysis. She rejects
Gumperz’ criticism that conversation analysis treats social groups and categories as
static and refers to Sacks’ account, in one of his lectures (cf.€Sacks 1992:€288), of how
groups are locally constituted through specific linguistic practices. Lin argues for
the integration of micro- and macro-perspectives and draws on Giddens’ structura-
tion theory for a concept of structuration practices and a framework that recognizes
the constraining influence of social macro-structures and institutions and can guide
an investigation of how these structures and institutions emerge from situated so-
cial action, interaction, and work. Lin is particularly interested in what happens in
the “borderlands” of inter-group and inter-positional interaction in the context of
an increasingly globalized world. Her focus is on the structuring activities that take
place during inter-cultural communication in non-egalitarian encounters. Follow-
ing Davies and Harré (1990), she describes how in such encounters participants
position one another according to conflicting, but consistent storylines and assign
each other parts in these stories. While high-status participants may seek to use
their power to secure the enforcement of their storyline, participants in less power-
ful positions use tactics (de Certeau 1984) to counteract non-egalitarian interactional
New adventures in language and interaction 

structures. Lin’s contribution illustrates the need for language and interaction re-
search that focuses on settings and engagements in which horizontal relations of
affiliation and engagement and vertical relations of power and control over linguis-
tic and communicative norms are negotiated and contested.
Karen Tracy and Robert Craig position their approach to the study of lan-
guage and interaction – which they call action-implicative discourse analysis
(AIDA) – in contradistinction to interactional sociolinguists and conversation
analysis. They argue that all approaches always reflect and have to reckon with the
traditions, concerns, and thematic regimes of their respective home disciplines: in
the case of their own approach, communication studies; linguistics and anthropol-
ogy in the case of interactional sociolinguistics; and sociology in the case of CA.
According to Tracy and Craig, CA takes on different flavors and is less firmly
bounded when it is transported into other fields. For communication scholars,
Tracy and Craig suggest, the practical dimension of research has always been a
main concern, reaching back to the beginnings of the rhetorical tradition in an-
cient Greece and Rome. Communication studies and, a fortiori, action-implicative
discourse analysis, are fundamentally interested in cultivating communication, that
is, in providing, on the basis of research on actual cases, guidance for improved
practice. The authors illustrate this perspective with an analysis of a school-board
meeting. Approaching the setting with an ethnographically grounded version of
discourse analysis (cf.€Schiffrin 1994), they aim to reconstruct the “situated ideals”
of different categories of participants, in order to explicate their normative stand-
ards, which can then guide the cultivation of communicative practice.
Srikant Sarangi is similarly motivated by the task of professional communica-
tion analysts to guide professionals, for example in health-care, in reflections of
their own communicative practice. Health-care professionals, he argues, are likely
“to apply such insights about interaction selectively, in the same way they deal with
theories and models of scientific and technical knowledge”. In contrast to other
contributions, Sarangi does not begin with a generic, bottom-up analysis of inter-
action sequences and their various embeddings within specific macro-contexts
and fields of social practice. Rather, he comes to his research with an understand-
ing of current policy-induced changes in the professional roles of health-care pro-
viders and the impact that these changes have on their interactional statuses and
roles as knowledge providers. To give an example, software-based expert systems
play an increasing role in health-care, but while they potentially lower the status of
physicians as human experts, they are also apt to empower nurses, which are ena-
bled to perform minor surgery and dispense advice independently of physicians.
Drawing upon Levinson’s study of “activity types and language” (Levinson
1979/1992), Sarangi conducts activity analysis, complemented by theme-oriented
discourse-analysis, and illustrates his approach here with analytic observations of
 Jürgen Streeck

an example of genetic counseling. He shows that the interaction is characterized

by frequent frame-shifts (from history-taking to diagnosis, etc.), as well as by re-
peated patterns of rhetorical escalation and de-escalation of the risk that the med-
ical condition poses to the patient.
Elizabeth Armstrong and Alison Ferguson discuss the contribution that inter-
action research can make to clinical research and intervention involving people
with aphasia; like Muntigl and Ventona, their research brings together conversa-
tion analysis and systemic-functional linguistics. The authors point out that apha-
sia is usually characterized by a breakdown of lexicogrammatical order in the
presence of relatively intact pragmatic skills. Microanalytic research on interac-
tions with people affected by aphasia has shown how these are able to draw on
“normal” conversational mechanisms (e.g.€ repair), but also on grammatical re-
sources such as cleft-constructions to compensate for difficulties in making them-
selves understood. However, this type of research has also revealed interactional
and linguistic accommodations that conversation partners of aphasics make
(e.g.€in prosody) and which sometimes amount to over-accommodation and yield
the risk that the speaker is perceived as talking down to the person who has the
speech impediment. Armstrong and Ferguson offer a detailed discussion of both
the history and the possible futures of research on language and interaction in-
volving people with aphasia and argue that many of the communication difficul-
ties resulting from aphasia are not categorically different from problems of mis-
communication in ordinary conversations; rather, they highlight and expose the
“generic” underpinnings of conversational collaboration.
Jürgen Streeck revisits some core ideas in the work of G.H. Mead, Bateson,
Goffman, and conversation analysis (sequentiality; the relationship between ges-
ture and action; an ecological view of context) and shows how in a series of brief
moments of workplace interaction hand-gestures are understood and function by
reference not only to concurrent speech, but also in relation to their positions
within turns and sequences of action, ongoing physical activities, and the material
environment – the landscape or setting – at hand. He thus pleads for an ecological
approach to the analysis of modalities (including language and gesture) in social
interaction. He also suggests that an ecological conception of embodied and mul-
timodal interaction will benefit from greater engagement with contemporary work
on embodied cognition in phenomenological philosophy and cognitive science.
Attention to the embeddedness of gesture, language, and interaction in the mate-
rial world raises the issue of the historical specificity and determination of that
world. Streeck concludes that, while the methodological issues of studying talk
and gesture in interaction have been settled, the real challenge now is to integrate
the interactionist conception of human competence into a broader conception
that encompasses the encultured quality not only of the mind, but also the body,
New adventures in language and interaction 

while reckoning with the historical constitution of the resources, contexts, and
practices that we study.
Frederick Erickson addresses the issue, first forcefully raised by Ochs (1979),
that transcription methods incorporate theories of the (linguistic, interactional)
realities that are meant to be represented by them, and he criticizes the conversa-
tion-analytic system devised by Gail Jefferson, which arranges turns at talk as in a
playscript, for its inherent neglect of the hearer and its logocentrism. Interaction,
Erickson points out, predates language. The transcripts that inform our research
and theorizing should accordingly show the embeddedness of talk in interaction,
which includes not only sequential phenomena (i.e. phenomena ordered by suc-
cession), but also the phenomena of mutuality that are embodied in participants’
concurrent spatial orientations to one another. Visible behaviors of co-participants
may shape both what they say and how they say it while they are saying it. These,
too, are absent from Jeffersonian transcripts (although they are recorded in detail
in such variations as the system of gaze notation developed by Goodwin 1981).
Erickson proposes and illustrates a transcription system that is derived from musi-
cal scores. It organizes the vocal and visible behaviors of all participants relative to
a common time-line and displays units of time in analogy to measure in music
notation. Importantly, musical notation can also record, as Erickson demonstrates
in an analysis of a “collective complaint sequence” during a family mealtime con-
versation, the timing of instrumental acts such as those of food distribution. It
thus allows the display and analysis of the temporal articulation of different strata
of activities and interactions. Erickson concludes that “the time is ripe for a re-
newed effort toward the study of space, time, and visual phenomena in social in-
teraction. The prospect seems promising for nonverbal and temporal aspects to
receive more attention in relation to speech than in the recent past.”
John Shotter’s final, visionary chapter, Dialogical dynamics: Inside the moment
of speaking, replaces, as Cowley’s does, the Cartesian view of understanding as a
congruence of representations with a view of intersubjectivity grounded in bodily
resonance. “All communication begins in, and continues with, our living, sponta-
neous, expressive-responsive (gestural), bodily activities that occur in the meetings
between ourselves and the others and othernesses around us.” Our intellectual
lives, Shotter writes, are based in “’inner’, dialogically-structured movements, in a
dialogical dynamics giving rise to unfolding movements which shift this way and
that in a distinctive fashion, movements whose ‘shape’ can be ‘felt’ or ‘sensed’ but
not pictured, or known at all in a propositional form”. Thought is not separate from
feeling; it orchestrates heterogeneous influences and provides us with senses of how
to go on in a situation. Shotter, who avows to be influenced by “specific utterances
or expressions” in the writings of Wittgenstein, Vygotsky, Bakhtin, Garfinkel, and
Merleau-Ponty, envisions a therapeutic form of analysis that will allow us to
 Jürgen Streeck

improve our practices from within our practices. Words and ways of talking are
psychological tools that enable us “to relate ourselves to our circumstances in a dif-
ferent way”. Therapy can engage us in Bakhtinian modes of thinking in the voices of
others and teach us new ways across the totality of our shared language games.
All of the contributors address themselves, implicitly or explicitly, to some of
the questions that Prevignano and Thibault initially posed to them and that related
to the authors’ conceptions of interaction and interaction analysis, the relative use-
fulness of analytic procedures, the categories of analysis, and the obstacles and
future that the analysis of interaction and language currently faces. The answer is
a multivocal Baktinian polylogue in which the authors act like ventriloquists and
the voices of the founders of our fields resonate in concert with the submerged
voices of Prevignano and Thibault. Cries of excitement continue to be heard on
our adventurous journeys into the microcosm of human encounters.


Coulter, Jeff 1989. Mind in Action. Cambridge: Polity Press.

de Certeau, Michel 1984. The Practice of Everyday Life. Berkeley: University of California Press.
Eerdmans, Susan L., Prevignano, Carlo L., and Thibault, Paul. J. (eds.) 2003. Language and Inter-
action: Discussions with John J. Gumperz. Amsterdam: Benjamins.
Goffman, E. 1983. “The interaction order.” American Sociological Review 48: 1–17.
Levinson, S. C. 1992. “Activity types and langsuage.” In Talk at Work. Interactions in Institu-
tional Settings. Paul Drew and John Heritage (eds.), 66–100. Cambridge: Cambridge
University Press.
Ochs, Elinor 1979. “Transcription as theory.” In Developmental Pragmatics, Elinor Ochs and
Bambi Schieffelin (eds.), 43–72. New York: Academic Press.
Prevignano, Carlo L. and Thibault, Paul J. (eds.). 2003. Discussing Conversation Analysis: The
Work of Emanuel A. Schegloff. Amsterdam: Benjamins.
Ryle, Gilbert. 1949. The Concept of Mind. London.
Sacks, Harvey. 1992. Lectures on Conversation. Vol.s I and II. Edited by Gail Jefferson. Cambridge:
Schegloff, Emanuel A. 1987. “Between micro and macro: Contexts and other connections.” In
The Micro-Macro Link. Jeffrey C. Alexander, Bernhard Giesen, Richard Münch & Neil J.
Smelser (eds.), 207–233. Berkeley: University of California Press.
Schegloff, Emanuel A. 1991. “Conversation analysis and socially shared cognition.” In Perspec-
tives on Socially Shared Cognition, Lousie Resnick, John Levine and Stephanie Behrend
(eds.), 150–171. Washington D.C.: American Psychological Association.
Schegloff, Emanuel A. 1992. “Repair after next turn: The last structurally provided defense of
intersubjectivity in conversation.” American Journal of Sociology, 975: 1295–1345.
Schiffrin, Deborah. 1994. Approaches to Discourse. Oxford, UK; Cambridge: Blackwell.
Wittgenstein, Ludwig. 1953. Philosophical Investigations. Cambridge: Blackwell.
Interlocutory logic
A unified framework for studying
conversational interaction

Alain Trognon and Martine Batt

Université Nancy

After exhibiting the historical and epistemic context of the discovery of the
interaction order, the authors develop a global theory of the cognitive-affective-
social organization of talk-in-interaction: «Interlocutory Logic». On the basis
of insights provided by the pragmatics of natural languages and the theory and
methods of contemporary logic, this theory deals with elementary illocutionary
acts and higher-order units (turns, exchanges) of conversation with the help of
methods of natural deduction and dialogical logic. The authors present a model
constructed within the empirical domain of functional dialogues during «hand-
overs» between work-shifts in a factory to demonstrate the descriptive and
explicative values of «Interlocutory Logic».


The word «interaction», which according to the French dictionary Larousse de la

Langue Française (1975) means «reciprocal action of two or more phenomena», did
not enter French dictionaries until recently. The 1892 Dictionnaire des dictionnaires
and the 1877 supplement of the 19th century French dictionary Littré both date its
appearance back to 1876. However, the short history of the word has not prevented
it from spreading and becoming widely used within the last century, in nearly eve-
ry knowledge domain. One finds it in mathematics, and in natural sciences like
physics, where all known phenomena are classified into four types of interaction,
and biology, where «interaction» refers to relationships between two or more or-
ganisms that coexist within a biotope. And of course, one finds it in the human and
social sciences, where, based on Goffman’s (1971) definition, it is generally under-
stood to mean the reciprocal influence that partners exert upon their respective
actions when they are in the immediate physical presence of each other.
 Alain Trognon and Martine Batt

What are the main success and failures of studying interaction in this last do-
main? And what tasks are needed? Answering these two questions in a first para-
graph where we will see that «accounting for interaction, and especially for verbal
interaction (...) still constitutes an essential challenge for cognitive psychology
(Caron 1997:€234)» (Trognon et al. 2008:€624) will bring us to propose a unified
framework, namely Interlocutory Logic. We shall present this theory in a second
paragraph: its goals, its formal organization, and a model which will illustrate its
semantics. The corresponding domain, a real one, of this model will be a func-
tional dialogue during a shift changeover situation.

1. Success and failures of the “interactionist paradigm”

1.1 Origins of the «interactionist paradigm»

In the human and social sciences, the first substantial work on interaction as a
concept began in the 1930’s, with Bakhtin (1929) in linguistics, and Mead (1934),
Vygotski (1934), and even «young» Piaget (1928, 1932) in psychology. At that time,
the theses set forth by these authors were purely speculative. But they proved ex-
tremely prolific forty years later in the 1970’s, when «this subject became a more or
less autonomous field of research [research that] has now reached maturity» (Ker-
brat-Orecchioni 1997:€1). This is when a full-fledged scientific program began to
emerge, one which, according to a synthesis of the field proposed by Kerbrat-Orec-
chioni (1989, 1990, 1997, 1998), could be called the «interactionist paradigm».
It should be noted that technological innovations had reached a level that per-
mitted the empirical study of interaction. Indeed, at the same time as this field of
research was developing, was offering increasingly sophisticated techniques for
recording sound and images (tape recorders and video cameras) and reproducing
observed events (digitization, and more generally artificial intelligence). With in-
teraction having become a fully accessible entity for observation and modelling on
the computer (Geniffey and Trognon 1986), research had a new scientific object at
its disposal. In fact, conversational analysis was to grow out of this very setting, as
Sacks noted in his famous Lectures (Sacks 1984:€26).
In the thirty years that followed, the research proliferated at the intersection
between sociology, psychology, linguistics, the philosophy of language, and artifi-
cial intelligence, each of these disciplines founding in interactional analysis mate-
rial to fuel its research and renewing its issues.
Interlocutory logic 

1.2 Fundamental theses of the interactionist paradigm

Although studies on interaction «include a heterogeneous series of investigations

[...], i.e., more a ‘sphere of influence’ than a coherent and unified set of theoretical
proposals» (Kerbrat-Orecchioni 1989:€7, our translation), they all are based on the
premise that «there is interaction». This thesis has been asserted with variable de-
grees of forcefulness. At one extreme, Goffman interprets it more or less methodo-
logically in writing for example, «I am convinced that if you want to study some-
thing, you have to start by approaching it as a system of its own, at its own level
[...]. It was this premise that led me in my thesis to treat interaction as a specific
domain and to attempt to rescue the term ‘interaction’ from the state in which
social psychology and those who claim to be its disciples had left it» (1981:€306,
our translation). Conversely, Schegloff (1991) granted it a more ontological status,
considering interaction to be a constituent phenomenon of human societies. Sche-
gloff states indeed that, for species living in a society, direct interaction is the pri-
mordial scene of social life, and that for human beings, this kind of direct interac-
tion is typically organized into «talk-in-interaction», with conversation acting as
its basic structure. But in either case, the interaction «plane», «level», «stage», or
even «order», as Goffman (1983) termed it, is a specific phenomenon and thereby
justifies a relatively autonomous approach or grain of analysis (Quéré 1989).
So, studies conducted in the interactionist paradigm have a number of com-
mon features or orientations, three of which are worthy of our attention. Although
they are interdependent, we will present them separately here. The first is the focus
on «natural» data, meaning data that has not been generated for the purposes of
analysis as is experimental or survey data. Taking this orientation, which is a re-
versal of the traditional ways, interactional linguists or psychologists place priority
on oral rather than written language, the dialogical to the monological, the actual
productions of people during real social activities (carried out in the here and
now) to the introspective accounts of the ideal speaker-listener heard by the schol-
arly speaker-listener (Boutet 1989; Kerbrat-Orecchioni 1989; Mondada 1995).
This first feature is not rooted in a «methodological stance» chosen for humanistic
reasons, with the ethnomethodologist acting as a sort of worker-priest of sociolo-
gy. It draws its legitimacy from the direct inscrutability of the object of observa-
tion, as Mondada so rightly stated: «Conversational data is ‘discoverable’ but not
‘imaginable’ – which makes any simulation problematic» (1995:€ 3, our transla-
tion). And as Gumperz noted:
As to «regularities» of communication practice, I believe that these should ulti-
mately be derived from or related to in-depth analyses of situated encounters in a
variety of settings. Because of the way interaction works, the usual questionnaire
studies are not satisfactory here (Gumperz in Prévignano and Thibaut 2002:€151).
 Alain Trognon and Martine Batt

All data, whatever their nature, whatever channel they tap, for they are multimo-
dal (Kerbrat-Orecchioni 1990), exhibit this first characteristic. An interpretation,
for example, is no more ineffable than a behaviour and can be grasped in the same
way. Both Gumperz (1982, 2002) and Garfinkel (1967, 1996), each in his own
way, offered us definitive proof of this, and in doing so, provided the groundwork
for the emergence of an original method of investigation and knowledge acquisi-
tion in the human and social sciences. This method (Levinson 1983; Heritage
1990) espouses the processes and devices by means of which people give inter-
subjective meaning to the everyday activities they accomplish in the here and
now, devices and processes which support the ever-changing, immanent, contex-
tual anchoring of social relations (Trognon 1994). The method thus helps us dis-
cover the repertoires which – along with different types of indexicals (Levinson
2002; Thibaut 2002) like the contextualization cues discovered by Gumperz (1982,
2002) – «index» interpersonal relationships in social relations. With such meth-
ods, the human and social sciences are in a position to go beyond the traditional
survey-based dichotomy between externality (e.g., sociology surveys) and im-
mersion (e.g., ethnology surveys conducted in the field, participation surveys).
The second feature, which follows from the first, involves the mixing of theo-
retical and epistemological references – interaction studies are multidisciplinary
(Boutet 1989; Kerbrat-Orecchioni 1989; Thibaut 2002) – and this is where re-
searchers who venture into this area run the risk of stigmatization in their com-
munity of origin. Investigators like Garfinkel, the first to call himself an eth-
nomethodologist before ethnomethodology was instituted, have been subjected to
numerous attacks and insults.
The third orientation, which in some sense overarches the other two, is the
desire to edify a natural science rather than a formal discipline (Goffman 1981;
especially Sacks 1984) and thus to adopt an empirical rather than hypothetico-
deductive approach (Kerbrat-Orecchioni 1989).

1.3 Properties of the «interaction order»

The «interaction order» is the interindividual relationship that embodies social

institutions through the use of language. In other words, the relationship is the
basis of the interaction order:
One cannot conceive of communicative interaction as an extrinsic coordination
process that harmonizes the individual activities of actors as if behind their backs
(see the law of supply and demand in economic individualism); nor as the in-
fluence of one speaker on another, which presupposes the heterogeneity of the
interactants (see classical rhetoric); nor as the secondary coordination of projects
thanks to the conscious effort required of the actors to reach a consensus [...]. In
Interlocutory logic 

fact, it is the relationship that explains reciprocal expectations and not vice versa:
permissible perturbations are ones that cannot be exceeded without breaking the
relational couple or destroying the self-organized system it engendered. (Jacques
1988:€52–53, our translation).

1.3.1 Dual opening of the “interaction order”

As the basis of the interaction order, the relationship must not be seen as a self-en-
closed phenomenon. The interaction order (Trognon and Bromberg 2007) opens
onto two horizons. Upwards, it opens up onto social institutions, and onto a par-
ticular institution that is specific to the human species. This prominent part of all
social institutions (Searle 1995) is what Saussure (1969) called the «speaking mass»,
namely, natural language. On this first horizon, interactions can be seen as consti-
tuting a space where institutions are accomplished. Since the early 1970’s, eth-
nomethodology and conversational analysis on the one hand (see, for instance,
Maynard and Zimmerman 1984; Watson and Sharrock 1990; Schegloff 1991; and so
on), and Gumperzian interactional sociolinguistics (Gumperz 1982, 2002) on the
other, have brilliantly illustrated this conception. Now concerning the institution of
language, we need to refer to interactionist linguistics as a whole, for language is just
like any other institution, and the constituents of language are just like any other
institutional state. Accordingly and analogously to what happened to ethnomethod-
ology with respect to academic sociology, very many studies conducted in interac-
tionist linguistics have shown how the categories referred to as «Platonian» in tradi-
tional linguistics are reproduced through interaction (Trognon, Saint-Dizier and
Grossen: 1999). What is more, numerous elements of language are overdetermined,
partially fashioned in return by the «interaction order». As Schegloff again wrote:
If the basic natural environment for sentences is in turns at talk in conversation,
we should take seriously the possibility that aspects of their structure (e.g.€their
grammatical structure) are to be understood as adaptations to that environment.
(Schegloff 1991:€154–155)

Downward, the interaction order is the womb of the self and of the mind (Stern
2004; Trevarthen 1993, 2001, 2004).

1.3.2 Relative autonomy of the “interaction order”

The interaction order nevertheless exhibits a «relative autonomy». It is «an inter-
mediate and in many ways analytically distinct level of organization» (Gumperz
in Previgano and di Luzio 2002:€8). This relative autonomy derives from the fact
that the interaction order is structured. The interaction order is thus the opposite
of a neutral or shapeless milieu. As Schegloff (1991) wrote, it is not an astructured
medium that would simply transmit messages, knowledge, and information. The
 Alain Trognon and Martine Batt

backbone of this structure is sequentiality, and this is why it has been placed at
the crux of conversational analysis: «CA is concerned with the study of the se-
quential organization of interaction and of the reasoning that is inherently em-
bedded within it» (Heritage 1990:€27). For example, as Schegloff (1991) wrote,
conversational sequentiality [t1,€ t2,...,€ ti,€ tj,...,€ tn] quasi-automatically creates a
special type of repair mechanism, which he called third position repair.
Third position repair may be thought of as the last systematically provided oppor-
tunity to catch (among other problems) divergent understandings that embody
breakdowns of intersubjectivity, that is, trouble in socially shared cognition of the
talk and conduct in the interaction (Schegloff 1991:€158).

Schegloff does not give a detailed analysis of the third position repair mechanism.
For our part, we have proposed an interlocutory analysis of it (Trognon and Brassac
1993; Trognon and Saint Dizier 1999; Trognon 2002), founded on Interlocutory
Logic (cf.€ supra), which shows quite clearly how the interlocution produces the
shared world or intersubjectivity evoked by Schegloff. The resulting model, named
“The Conversational Mechanism of Mutual Understanding” (CM2U), bridges the
gap between the speaker meaning and the meaning intersubjectively taken to be
the speaker meaning (Clark 1996; Trognon 2002).
So, as Schegloff (1991) wrote it, repairs, especially third position repairs, sup-
port the intersubjectivity of the participants in the interaction, i.e., «the mainte-
nance of a world (including the developing course of the interaction itself) mutu-
ally understood by the participants as some same world» (Schegloff 1991:€ 151).
The «discovery» of third position repair thus brings empirical content to the pro-
cedural solution that Garfinkel, by inventing ethnomethodology, suggested for
academic sociology’s unsolvable problem, that of the acquisition, confirmation,
and revision of «common» sense (Bernicot et al. 1997; Trognon 2002)€ in such a
way that even if interaction were not the source of intersubjectivity, the latter being
somehow innate (as believed by Trevarthen (1993, 2001, 2004; for example) or
even embodied (as the discovery of mirror neurons would lead one to believe;
cf.€di Pellegrino et al. 1992), it would still constitute a medium, a receptacle, a con-
tainer necessary for the psychological survival of the human individual.

1.3.3 Human beings communication and the “interaction order”

It is very important to emphasize along with Schegloff that the third position re-
pair mechanism (or CM2U) has two main consequences on Human Beings’ Com-
munication: on the individuals and on the human species.
Firstly, the interaction order clearly represents an advantage for the human
race in natural selection. As one of us wrote earlier (Trognon 1991:€403–404; Tro-
gnon 2003), a communication system made up solely of a language, i.e., a
Interlocutory logic 

haphazardly-designed code which, besides, relies on an inferential process, is cer-

tainly not an excellent communication system. Its reliability is totally relative and
its productions uncertain. It is hard to imagine, for example, how it could effec-
tively coordinate the individual actions required to carry out a collective project.
From this angle, a far more efficient communication system is the encoding-de-
coding system used by bees. But imagine for a moment that in this communica-
tion system, the haphazard code has attached to it a tuning subroutine that serves
to perfect its operations. This would give us the most efficient communication
system of all. Virtually any mishap or failure could be recovered by the tuning
subroutine. Even better, the system’s representational capabilities would be supe-
rior to those of a code. Indeed, to represent a new piece of information, a code
must necessarily have a new unit available, otherwise it loses the very property
that makes it useful: the one-to-one correspondence between the set of elements
to encode and the set of encoded elements. Consequently, it is costly to integrate
new information into a code. A [natural language, tuning subroutine] system does
not suffer from this type of limitation. And this is what makes language, plus the
interaction order, an advantage in natural selection. As Edda Weigand writes it:
It is precisely this risk of misunderstanding which makes communication so effi-
cient since it permits us to cope with an unlimited, ever changing world (Weigand
Communication functions so well because it allows the risk of misunderstand-
ing and trusts in the fact that it will be corrected by the ongoing dialogue itself
(Weigand 1999:€783).

Schegloff adopted an analogous point of view when he wrote:

Given that hearers have resources available for addressing problems in understand-
ing, should they arise, the resources of natural language need not, for example, be
unambiguous. They need not have invariant mappings of signs or symbols and
their signifieds. They need not have a syntax that assigns only a single interpreta-
tion to a given expression. They need not be limited to literal usage, but may be
used in idiomatic, metaphoric, and other non literal tropes (Schegloff 1991:€155).

In short:
The kinds of language components from which it is fashioned – sounds, words,
and sentences – have the character they do and are formed the way they are in part
because they are designed to inhabit an environment in which the apparatus of
repair is available and in which, accordingly, flexible arrangements can be permit-
ted (Schegloff 1991:€155).

Finally, phylogenetically speaking, it would seem as if Mother Nature (borrowing

Dennett’s expression 1987) invented the interaction order to make up for the
 Alain Trognon and Martine Batt

imperfections of communicating with natural language. By finding a procedural

solution to the undecidability of natural-language communication which over-
comes that undecidability without eliminating it, the human race has created an
optimal, robust, flexible, and perfectly adaptive system of communication.
In any case, for the human species as a species, for one must indeed admit that
human evolution has a psychological cost. In effect, it seems that «when the hu-
man species ate of the fruit of the Tree of Knowledge, it discovered that automatic
signs could be turned into signals and emitted with conscious or unconscious pur-
pose. With that discovery, of course, also came the possibility of deceit and all sorts
of other possibilities» (Bateson 1956:€ 157–158; Lacan 1981:€ 47). Evolution thus
demands a very peculiar kind of competence: «knowing what to do with» uncer-
tainty, for «misunderstanding is a human condition» (Prevignano 2002:€17).
It is not merely a question of what something «means». Ultimately, agreement on a
specific interpretation presupposes the ability negotiate repairs and re-negotiate mis-
understandings, agree on how parts of an argument cohere, follow thematic shifts and
shifts in presuppositions, that is, share indexical conventions (Gumperz 2002:€18).

Fundamentally, this competence is a competence whose object is the procedural.

It could be considered an important component of the «interactional competence»
to which Prevignano, Thibaut, and Gumperz refer in their discussion (2002). It is
even thought to be located somewhere in the frontal lobes (Bernicot and Trognon:
2002; Dubois: 2002). But regardless of where one stands on the issue of this com-
petence, it is clear that there are human beings who have been deprived of it, be it
accidentally due to a brain injury for example (Peter-Favre 2002) or otherwise as
in schizophrenia and autism (Trognon 1987, 1992; Trognon and Collet 1993). And
in this case,€ «it is not only that you don’t survive, you don’t learn» (Gumperz
2002:€ 19). As Prevignano (2002:€ 19) stated: «In psychiatric cases, it’s just these
metapragmatic procedures that fail.» Jakobson (1955) already knew this: brain
pathologies and mental illness are directly related to the communication system
used by human beings.

1.4 Advocacy for extending the interactionist paradigm to individual cognition

1.4.1 Epistemological consequences of the interactionist revolution

As could be expected, traditional views of cognition, which take it to be a solipsis-
tic, acontextual psychological event, have been blemished by the success of the
interactionist paradigm.
In Western tradition, it is the single, embodied, minded individual who constitutes
the autonomous reality. Organized aggregations – whether of persons or of activi-
ties – tend to be treated as derivative, transient, and contingent. They are something
Interlocutory logic 

to be added on, after basic understanding is anchored in individual-based reality.

It has accordingly seemed appropriate in the cognitive sciences to study cognition
in the splendid isolation of the individual mind or brain, and to reserve the social
aspect for later supplementary consideration (Schegloff 1991:€168).

But after the interactionist breakaway had extended the pragmatic reorientation
(Bernicot et al. 2002), it became clear that:
Knowledge and understanding (in both the cognitive and linguistic senses) do not
result from formal operations on mental representations of an objectively exist-
ing world [but] arise from the individual’s committed participation in mutually
oriented patterns of behaviour that are embedded in a socially shared background
of concerns, actions, and beliefs (Winograd et al. 1989:€78).

Thus, cognition is no longer defined as an abstract mental event but as a situated

phenomenon (Suchman 1987) that depends on available resources, granted, but
also and especially on the here and now of the situation: situated actions corre-
spond to «the view that every course of action depends in essential ways upon its
material and social circumstances» (Suchman 1987:€50). Cognition, then, is no
longer just an individual event; it is also a distributed phenomenon embodied in
the transactions that cognitive agents carry out with each other and with the ar-
tefacts that occupy their environments, transactions which, moreover, cannot be
described as the sum of the resources utilized by the partners, but as joint, origi-
nal constructions that emerge from the interactive dynamic (Hutchins 1995). As
Lave (1993) also wrote, cognition is constituted by dialectic relationships be-
tween the actions of the persons in interaction, the contexts of their activity, and
the activity itself.
In sum, rationalism, the dominating epistemology of the human and social
sciences in the 1960’s, came out deeply disrupted by the accumulation of knowl-
edge about the «interaction order». However, the disruption is unequally distrib-
uted across the various established disciplines. It showed up clearly in sociology
and even more so in linguistics, where, according to Auroux: «With the introduc-
tion of interaction into the analysis of linguistic phenomena, we are witnessing a
revolution in the representation of these phenomena which is not very compatible
with the type of philosophical rationalism defined by Chomskians» (1989:€ 221,
our translation). For this same author again, while the philosophical interpreta-
tion of this revolution is unclear, one can surely uncover in it «the confirmation of
the social conception of language [with] a shift of the transcendental sphere, which
would cease being linked to solipsistic subjectivity and start to depend on a found-
ing dialogical intersubjectivity» (Auroux 1989:€221, our translation). The collec-
tive representation of a linguistic event – given that it is no longer identical to the
representation of that event in each of the interactants and hence is no longer a
 Alain Trognon and Martine Batt

shared representation – must, in order to be founded, also be related to what hap-

pens among the interactants.

1.4.2 And what about psychology?

It is clear now that «methodological solipsism» is erroneous:
Such a stance may be deeply misconceived, because our understanding of the
world and of one another is posed as a problem and resolved as an achievement,
in an inescapably social and interactional context – both with tools forged in the
workshops of interaction and in settings in which we are answerable to our fel-
lows. Interaction and talk-in-interaction are structured environments of action
and cognition, and they shape both the constitution of the actions and utterances
needing to be «cognized» and the contingencies for solving them. To bring the
study of cognition explicitly into the arena of the social is to bring it home again
(Schegloff 1991:€168).

«The cognitive subject is not a monad: he interacts in an ongoing way with his
environment, and in particular, with other subjects» (Caron 1997:€233, our trans-
lation). «What this interaction brings into play, above all, is language» (ibidem).
So, «accounting for this interaction, and notably verbal interaction, which is un-
doubtedly its most elaborate form, poses an essential challenge to the cognitive
psychology [...] upon which it must be based»(ibidem). But psychology could ben-
efit a great deal from the study of interaction and talk-in-interaction as “structured
environments of action and cognition”, as it may lead to less speculative, more
parsimonious and descriptively more adequate modes of cognition (Good 1990).
If psychology is starting to acknowledge, especially in developmental psychology,
that cognitions (or at least their expressions) are overdetermined by «the logic» of the
interaction order. For example, «the young child’s non-conserving responses could
reflect, not so much the inability to understand the effects of the transformation (i.e.,
failure to grasp conservation) than the inability to understand the experimenter’s in-
tentions. The confusion is not so much conservational as it is conversational» (Light
et al. 1989:€103; our translation; see also Siegal 1991). If, more generally, we are willing
to agree in establishing cognitive psychology that «a reasoning experiment is more
like a conversational exchange where the subject, the listener, infers certain conclu-
sions from what the experimenter, the speaker, says» (Van der Henst 2002:€291, our
translation; the italics are the author’s) and that «taking the conversational context
into account allows us to distinguish that which pertains to reasoning per se from that
which pertains to an interpretation» (ibid, 301; our translation), it is still too often
proposed in the petitio principii mode. Indeed, studying the production of cognitions
during «talk-in-interaction» on the one hand, and studying its appropriation by in-
teractants in situ on the other – even if they could end up providing empirical answers
Interlocutory logic 

to the questions raised nearly a century ago by Mead (1934), Vygotski (1934) and
even Piaget (1928) are no longer a top priority in cognitive psychology.
It seems to us that the absence of a conceptual framework capable of describ-
ing the occurrence and the outcome of cognitions during talk-in-interaction does
not facilitate an aggiornimento in psychology. Interlocutory logic was designed to
provide some elements of such a framework.

2. Towards a unified framework for studying talk-in-interaction

2.1 What theoretical requirements should the

desired theoretical apparatus satisfy?

Any theory that attempts to satisfy the requirements mentioned above must have
at least two properties. Firstly, it must be able to formally describe interlocutory
events. Secondly and more generally, it must be able to build a «grammar» of the
types of dialogue in which we engage and their felicity conditions. The phenome-
nal properties that must be rendered are:
– Illocutionarity: an interlocution is made up of acts of discourse.
– Successiveness: an interlocution is a concatenation.
– Dialogicity: the contributions to an interlocution come from several sources.
– Recursiveness: the hierarchical elements are organized into different text levels.
These empirical properties determine the organized set of logical methods that
enable one to apply a calculus to interlocutory events, and thus to treat interlocu-
tion as a rationally accessible phenomenon.
The first principle (illocutionarity) defines the alphabet of interlocutory logic.
It encompasses the acts of discourse and the «relaters» of natural language, which
include not only connectives, but also marks used to structure the conversation,
interjections, adverbs, etc. The other principles define the empirical forms that
articulate these components: for example, the principle that they succeed each
other linearly and hierarchically.
The logic methods chosen are not necessarily independent of each other. For
example, if interlocutory logic adopts natural deduction as one of its logic meth-
ods, it is because natural deduction enables one to operationalize the principle of
successiveness. But recourse to dialogical logics à la Barth and Krabbe enables one
to operationalize the principle of successiveness and the principle of dialogicity at
the same time, since this type of logic shows how to go from one to the other. Now
if interlocutory logicians are so enthusiastic about Hintikka’s theories, it is
because referring to these theories allows one to simultaneously operationalize the
 Alain Trognon and Martine Batt

Table 1.╇

Interlocutory Properties Logic Methods Used

Illocutionarity General Semantics (Searle & Vanderveken, 1985;

(with the other constraints) Trognon, 2002; Vanderveken, 1990)
Successiveness Natural Deduction (Barth & Krabbe, 1982; Gentzen, 1935)
(with the other constraints)
Dialogicity Dialogical Logics
(with the other constraints) – Dialogue Proposition Evaluation Theory (Barth &
Krabbe, 1982; Lorenzen, 1967)
– Search and Discovery Dialogue Theory (Carlson, 1983;
Hintikka, 1983)
Recursiveness Hierarchical Conversation Structure Theories (Moeschler,
(with the other constraints) 1985; Roulet et coll., 1985)

principles of dialogicity and recursiveness. More specifically, the main merit of

Hintikka’s theories is having shown that the logic required to study the logical
facet of natural language is precisely a theory that is dialogical in nature. By obey-
ing the principle of dialogicity, one also obeys the principle of recursiveness. More
generally, contrary to what some authors thought they could write or think, inter-
locutory logic is not a logic «toolbox».
Table 1 presents these different constraints and how they are operationalized.

2.2 Interlocutory logic: A system that satisfies the above requirements

Two aims can be assigned to interlocutory logic. One is quite modest: describe the
sociocognitive organization of interlocutions by devising a system of formal pro-
cedures that adheres as closely as possible to the «phenomenology» of interlocu-
tion. Many studies published since 1995 illustrate this first approach (Trognon
1999; Trognon and Kostuski 1999). The other aim is much more ambitious: it con-
sists in using the above system to make hypotheses about the subjective processes
that take place at the cognitive level and lead logically to the products of a conver-
sation as they are expressed formally at the end of the descriptive phase of the
analysis. Accordingly, interlocutory logic proposes an analysis of the cognitive
production of a conversation, i.e., the study of the conversational emergence of
both declarative and procedural knowledge. Some examples of situations where
the emergence of knowledge has already been analyzed in this framework include
learning to move the cursor in a word-processing tutorial (Trognon and Saint-
Dizier 1999), learning to handle a pneumatic drill in an apprenticeship setting
(Sannino et al. 2003), volume conservation tasks (Marro et al. 1999), school
Interlocutory logic 

learning of division (Trognon, Saint-Dizier et al. 1999; Trognon 2007) and propor-
tionality (Schwarz et al. 2008; Trognon, Batt et al. 2006), hypothetico-deductive
reasoning applied to an empirical problem (Trognon and Batt 2003, 2005) and a
logical problem (Batt et al. 2009; Laux et al. 2008; Trognon and Batt 2004; Trog-
non, Batt et al. 2007), and making diagnoses (Brixhe et al. 2000). Interlocutory
logic has also been used to formalize the co-resolution of the Tower of Hanoi
problem by children (Trognon et al. 2008).
Past studies have already shown, then, that interlocutory logic, as a theoretical
and technical system that abides by and reconstructs the phenomenal properties
of interlocution, is a useful method of analysis for describing interlocutory events
and uncovering their underlying processes, notably by way of a computational
analysis that is as natural as possible (Trognon 1999, 2003; Trognon et al. 2008).
The system of coordinate logics that constitutes interlocutory logic is suited to
natural language. It is true that they form a limited class of logics because of the
constraints they must satisfy, but they are nevertheless of general relevance to any
discourse expressed in natural language. More fundamentally, because of the in-
tertraductibility of model-theoretical, dialectical and derivational methods of
modern logic (see for example Barth et al. 1982:€306, Theorem 29), proved along
1980’s years, interlocutory logic is able to find out strategies by which intersubjec-
tivity passes into intrasubjectivity and vice versa. According to a psycholinguistics
point of view, by virtue of Theorem 29, what an individual can cognate, a dyad
(and moreover a small group) can cognate it, and often better (Trognon, Batt et al.
2007; Laughlin et al. 2006). Reciprocally, what a dyad (and moreover a small
group) can cognate, an individual can cognate it. It is why a talk-in-interaction
context generally favours acquisition. Moreover interlocutory logic might explain
why individuals take off cognitive benefits from talk in interaction, as we have
showed it in Laux et al. (2008) and Trognon, Batt et al. (2007) for individuals who
resolved the Wason’s task, in Trognon et al. (2008) for children who resolved the
Tour of Hanoï’s problem, and in Schwarz et al. (2008) and Trognon, Batt et al.
(2006) for the acquisition of proportionality.

2.3 Conducting an analysis in interlocutory logic

An interlocutory logic analysis is conducted in two phases. First, the interlocution

is assigned a formal representation using a table called an «interlocutory analysis
table». This purely descriptive phase is followed by an analysis aimed at demon-
strating the existence of an underlying subjective and intersubjective dynamic,
whose utterances written in the interlocutory analysis table constitute potential
expansion points. An interlocutory analysis table looks something like this.
 Alain Trognon and Martine Batt

Table 2.╇ Interlocutory Analysis Table

Transaction Structure Sequential Conversational

Illocutionary Cognitive

Type Token
F(p) C
1 2 3 4 5 6

The rows in the table are filled in gradually as the elementary components of the
participants’ contributions to the dialogue are made. The middle column of the
table (column 3) is filled in first with the component’s position in the sequence.
This step is essential since it breaks the sequence down into its final elements,
which it ranks by order of occurrence. It is also a preliminary task, in that sequen-
tiality is what lays the empirical grounds for all interlocutory events, as conversa-
tional analysis specialists have so clearly demonstrated (Trognon 2002). The or-
dered elements are then assigned their illocutionary interpretations in columns 4
and 5 (type, for example, if it is an assertion; token, for example, if that assertion
defends a point of view), and their cognitive interpretations in column 6, which is
itself split into as many columns as there are participants in the conversation. The
illocutionary interpretations correspond to the forces of the speech acts accom-
plished during a given contribution. We are indeed referring to speech act theory
here, for no matter what has been said about speech acts, no one has yet been able
to replace them as the minimal units of linguistic communication which best ac-
count for the fact that communication via language is fundamentally a sociocogni-
tive event. The illocutionary force is the pragmatic function accomplished when an
utterance is produced; it corresponds to what uttering the speech act amounts to
doing during a conversation. Five major types of actions can be accomplished by
uttering a speech act: assertive, commissive, directive, declarative, and expressive.
The actions are qualified by a number of properties: the goal (and its direction of
adjustment), powerfulness, mode of accomplishment, propositional content con-
ditions, preparatory conditions, and sincerity conditions. The cognitive interpreta-
tions of the tokens located in the middle column are the propositional contents of
the speech acts accomplished by the interacting partners as the conversation un-
folds. Propositional content is the representation with respect to which an illocu-
tionary force is applied in the world. Corresponding to the cognitive-representa-
tional function of a speech act, it is a proposition that represents the state of affairs
targeted by the utterance, with the force being a sort of operator of that state of af-
fairs. Here again, the illocutionary and cognitive interpretations are assigned in
order of occurrence, but – and this is a very important point – via a process that is
Interlocutory logic 

both prospective and retrospective. The illocutionary interpretation «type» often

corresponds to what in pragmatics is called the «utterance meaning», and the illo-
cutionary interpretation «token» corresponds to what is called the «speaker mean-
ing», although not totally since the speaker meaning is stabilized conversationally
(cf.€infra). In regards to conversation, Clark recently wrote (see also Grice 1982):
The notion ‘what the speaker means’ is replaced by ‘what the speaker is to be taken
to mean’. The change is small, but radical. The idea is that speakers and addressees
try to create a joint construal of what the speaker is to be taken to mean. Such a
construal represents not what the speaker means per se – which can change in
the very process of communicating – but what the participants mutually take the
speaker as meaning, what they deem the speaker to mean (Clark 1996:€213).

In Interlocutory Logic, this process is described with the CM2U model. It consists of
at least two speakers (S1 and S2), the first and second speakers, respectively) and three
speech turns (T1, T2, and T3) successively distributed in the following manner:
T1: S1
T2: S2
T3: S1
The mechanism is composed of the two relationships organized in the diagram below.
(T1, T2) forms an interpretation relationship.€Its second element enacts S2’s
interpretation of the action performed by S1 in T1. S2 thus makes this interpreta-
tion mutually obvious (Sperber and Wilson, 1986). Explaining why (T1, T2) con-
stitutes an interpretation relationship is equivalent to explaining why T2 «inher-
ently embodies and displays its producer’s interpretation of the prior actions in the
sequence». T2 can be defined in this manner because T2 is an action that creates a
state of affairs available to each interactant, and because that state of affairs appears
«after» T1 and thus, according to the rules of communication (cooperation or rel-
evance) in response to T1. In general semantics, T2 can be defined as an action


T1 T2 T3


Figure 1.╇
 Alain Trognon and Martine Batt

that fulfils the satisfaction conditions of an illocutionary interpretation of T1.

More formally again, i.e., in illocutionary terms, the second turn (T2) will be in-
terpreted as the action which, by default, satisfies (Trognon 2002) the satisfaction
conditions of a certain illocutionary interpretation of the first turn (T1). ((T1, T2),
T3) forms an evaluation relationship.€Given that S2’s interpretation of T1 is avail-
able to S1 in T2, he can compare it with his own interpretation and enact a ratifica-
tion if the two interpretations correspond, that is to say, if S1’s interpretation of
this initial utterance is equivalent to that enacted by S2 in T2. He can also reformu-
late T1 if the two interpretations diverge, that is to say, if S1’s interpretation of his
initial utterance is not equivalent to that enacted by S2 in T2. The main conse-
quence of ((T1, T2), T3) is the mutual knowledge of S2’s interpretation of T1. In
T2, this interpretation is obvious to S1. In T3, it is obvious to S2 that the interpre-
tation is obvious to S1. Here again, the reasoning that leads to this latter knowl-
edge involves laws of general semantics. There are two possibilities, depending on
whether T3 confirms or invalidates S2’s interpretation of T1. If it is confirmed,
what S2 understood becomes the speaker meaning for the participants of the con-
versation. If it is invalidated, the invalidation signals a misunderstanding. But in
either case, intercomprehension has progressed.
The interactants’ belief, progressively acquired through this confirmatory ac-
tivity, can be represented in the following manner, where B is a modal belief indi-
cator, a and b are the speakers, and i is the illocutionary force of the initial utter-
ance as it is elaborated during the interaction.
Based on the hierarchical theory of conversations, the left side of the table
represents the interlocution’s discursive organization into acts, exchanges, and
moves, and via combinations of these different elements, into structures or even
More technically, to analyze an interlocution fragment in interlocutory logic
amounts to decomposing this fragment into a series of utterances. Each utterance
is represented by an expression ф of the system: <Mi, {Mi-k},{Mi-k} ˇ Mi, RD, DG>.
Mi is the conversational move accomplished by the utterance under examination.
{Mi-k} is the set of all the conversational moves that precede the move Mi and from

Table 3.╇ Legend: B = Belief

Levels a b

a1 1
b1 2 BaBb(i)
a2 3 BbBaBb(i)
b2 4 BaBbBaBb(i)
... ... ...
Interlocutory logic 

which Mi follows. Mi can then be conceived as a conclusion that results from

premises {Mi-k}. The reasoning that leads from {Mi-k} to Mi, and that is represented
by the schema {Mi-k} ˇ Mi, is called, in logic, a sequent.1
Let’s specify this notion more precisely by adapting an analysis proposed by
Carlson (1983).2 Suppose that A goes to the stadium to attend a pole-vaulting final
between Jack and Bob. Delayed, he only arrives at the stadium once the competi-
tion has finished. When arriving A entertains the following «ideas»: If someone
won it is Jack or Bob and someone has won. Moreover, he asks himself “Who has
won?” Catching the sight of his friend B, he engages in the following dialogue:
1A : Has Jack won? (1)
2B : No (2)
3A : Then Bob won (3)
4B : No (4)
5A : But then nobody won! (5)
(2) is a response to (1). The rule (or rules) (RD) of dialogue (DG) allowing to de-
rive (2) from (1) is represented by the sequent {1A} ˇ 2B. The rules that lead from
the premise to the conclusion are the questioned rules of semantics exposed in the
research of Hintikka (1976, 1981, etc.) of whom Carlson is a student. (3) comes
from the «thoughts» entertained by A3 and from the information 2B worked out
together by the logic rule of disjunctive4 dilemma. (5) is again deduced from the
thoughts entertained by A and from 4B in using reductio ad absurdum.
In a relatively informal manner for understanding the meaning of our ap-
proach, the previous short dialogue has just been reproduced as the product of a
set of dialogue rules formulated as sequents. Some of these rules belong to (dia-
logical forms of) standard logic. This is the case with the disjunctive dilemma or
the classic reductio ad absurdum rule which would be used to demonstrate 5A.
Other rules concern, instead, the semantics and pragmatics of natural language.
We call sequents of dialogue the setting of the relationship in a set of dialogical
events (Trognon, Batt et al. 2006; Trognon and Batt 2007a, b; Trognon et al. 2008).
This relationship is an inference composed on the one hand from a set of premises

1. “a sequent” is a pair (note Γ ˇ F) where€:

– Γ is a finite set of formulas. Γ represents the hypotheses that one can use. This set is also
called the sequent context
– F is a formula. It is the formula that one wants to demonstrate. This formula is said to be the
conclusion of the “sequent” (David, Nour and Raffali 2003:€24).
2. Even in his recent publications (1996, for example) Carlson does not refer to the logic the-
ory of sequents. We think nevertheless that his theory calls for this extension.
3. This is the cognitive environment of A defined by Sperber and Wilson (1986).
4. p€v q, or ~p, so q.
 Alain Trognon and Martine Batt

given in the dialogue and, on the other hand, of events which are deduced from
these premises. The rules intervening in this inference are rules of a dialogue game.
For example, the dialogue sequent {1A}ˇ 2B rests on a rule belonging to the game
theory of the question-response dialogue. This example shows that interlocutory
logic seems to generalize the sequent notion to all the illocutory acts used in an

2.4 An example: The interlocutory logic of a shift-changeover dialogue

The dialogue recorded here took place during a shift changeover in a production
shop of a paper company (Grusenmeyer et al. 1995, 1996; Trognon et al. 1997).
Shift changeovers result in specific kinds of talk-exchanges. What is peculiar to
them is that the common activity is time shared; the operators contribute succes-
sively in carrying out the collective intentionality. Therefore, as the succession of
operators is governed by an abstract rule, it is very difficult for each operator to es-
timate in real time his own contribution to the collective intentionality. The opera-
tor leaving the work station (outgoing operator) cannot directly evaluate all the ef-
fects of his actions, since it is the operator taking over (incoming operator) who is
affected by them. If the outgoing and incoming operators alternate (the second suc-
ceeds the first, the first the second, and so on, as in the present dialogue), the outgo-
ing one can only receive feedback regarding this evaluation during his next shift,
when he asks his predecessor and/or the latter inform him. The information «trans-
mitted» by the outgoing operator is therefore useful to his incoming colleague. On
the one hand, it allows the latter to construct a global representation of the process.
He is then likely to receive feedback on his own prior actions, thereby enabling him
to evaluate the effectiveness of these actions. On the other hand, he can get to know
the reasons for the action he has to undertake. This information is «transmitted»
during shift changeover when the outgoing operator meets the incoming one. Use-
ful in «normal running», this information becomes necessary and even crucial in
problematic or unusual situations, such as a malfunction or an incident.
The operator who was leaving work (the outgoing worker) will be called A, and
the operator who was taking over for A at that same workstation (the incoming
worker) will be called B. These two workers operate a machine that produces sheets
of paper. A sheet of paper arrives on the partially damp machine. It is trimmed on
both sides by two edgers located at the front and back of the damp part of the ma-
chine. The edgers produce two very fine sprays of water that mark off the edges of
the sheet (hereafter called «pistols»). It sometimes happens that projected paper
pulp accumulates in the edgers and causes tearing. The problem under debate here
is how to assess (and explain) the present working of the machine.
Interlocutory logic 

A1: (...) and the pistols seem to be working well.

B1: And the back one, it still lifts the sheet a little, if you noticed.
A2: Well, yeah, maybe.... But I didn’t have any pulp compared to yesterday;
I didn’t have any pulp, hm.
B2: I had some.
A3: You had some? Me, I didn’t have any, hm.
B3: And I had decreased it a little more because I felt it was moving the sheet
away a little and that made it uh... squirt out.
A4: Oh, right, me, what I did, I reopened it, it’s the front one this morning,
a tiny bit, because, well, you saw today I tore it (the outgoing worker
shows the notebook, morning, right-hand side), OK, I pulled the end,
uh... three times.
B4: Yes
A5: But the strip, it wasn’t cut. I didn’t clean the pistols hm. I didn’t even take
off the pulp by hand, nothing, and there isn’t any, just a few fibers, that’s
all hm.
B5: Because they’re set right.
A6: It’s just as good, huh?
In this interlocution, we find two sequences pertaining to two different interplays
of speech: (i) a declared difference of opinion (Barth et al., op.€cit.), from A1 to A3,
and (ii) the cooperative resolution of the conflict, from B3 to A6. The dialogue
ends with a shared solution to the problem that is understood by both interacting
For each of these sequences, let us now describe the discursive architecture, the
dialogical format, and finally, the reasoning in which the interlocutors engaged and
what they learned from it. The entire process will then be assembled to form a model
representing how a third party (e.g., an instructor) might perceive the interlocution.

2.4.1 The declared difference of opinion Discursive architecture of the difference of opinion A reports a property of the world
In A1, A states a property of the world, more specifically, a property of the pistols:
«and the pistols seem to be working well». A makes an assertion whose truth con-
ditions are found in «what seems to be». In this way, A notices a state of the world
and reports it. Formalized in the language of first-order predicate calculus, A1 is
written: Ass A1-B((∀p)€Wp), where Ass stands for assert, A for speaker A, 1 be-
cause it is his first speaking turn, and B because B is the addressee. As we have seen
 Alain Trognon and Martine Batt

above, the propositional content of this assertion is formalized5 as (∀p) Wp, or

more simply Wp.€Along with A’s interlocutor, B, we know that when A asserts the
proposition «the pistols seem to be working well», he commits (Searle et al. 1985;
Vanderveken 1988, Trognon et al. 1992) to supporting the analytic proposition
that the back pistol works well and that the front pistol also works well, since there
are only two pistols, one in front and one in back. This proposition is itself the
conjunction of two primitive propositions: (∀p)€ Wp€ ≡€ Wpd€ ∧€ Wpf (where pf
stands for front pistol and pb for back pistol, these two arguments falling under the
unary predicate W). "Working well" at time tn is a state that is posterior to the state
at time t', where the pistols "were not working well", i.e., they did not make the
formula (∀p) Wp true, which, as we have seen, is the propositional content of A1.
The thesis stated by A in A1 can thus be paraphrased by the compound statement
ϕ: ϕ: (∀p)Wp€ ≡€ Wpd€ ∧€ Wpf. The implicated proposition "the pistols were not
working well", which A seems to have set in the past, at time t’, can be paraphrased
by the compound statement ϕ’: ϕ': (∀p) ¬Wp ≡ ¬(Wpb€∧€Wpf).
To express the anteriority relationship of ϕ'with respect to ϕ, we use the pred-
icate U, and we write: Ut’ϕ’tnϕ, which reads "event ϕ' that happened at time t'
precedes event ϕ which is happening at time tn" (or "event ϕ which is happening
at time tn succeeds event ϕ' that happened at time t'"). Statement ϕ is said to be
valid if and only if the universal statement (∀p) Wp is true when the compound
statement (Wpb ∧ Wpf) is true. Thus, if A’s listener considers assertion A1 to be
true, i.e., that it corresponds to the state of the world in the utterance context tn of
A1 (that the pistols are working well) then he also considers the statement (Wpb ∧
Wpf) to be true in that same utterance context, i.e., that the back pistol and the
front pistol are both working well at tn. The proposition is true if each term of the
conjunction is true, i.e., if the statement Wpb is true and if the statement Wpf is
also true. Now let us look at the interpretation of A’s assertion in the discourse of
his interlocutor, B, in B1. Objection!
The reaction, B1, seems to raise an objection. B points out that one of the pistols, the
back one, «still lifts the sheet a little». B thus implies (Grice 1979; Sperber et al.
op.€cit.) or implicitly signifies that the pistols still are not working correctly, and thus
that what A asserted is false, since one of the necessary conditions of this state
(working well) is still not met in the utterance context of A1 and B1, at time tn.
Clearly, if a pistol lifts, it can cause paper pulp to accumulate, and that can cause a
malfunction. This knowledge is shared by A and B. In B1, B assumes (sarcastically?),

5. ¬: negation; ∨: disjonction€; ∧: conjonction€; ≡€: equivalent€; ⊃, implication€; ∃: existential

quantifier€; ∀: universal quantifier.
Interlocutory logic 

«if you noticed», that A also realized that the back pistol was still lifting. So B seems
to be contradicting A. This contradiction is put up for discussion in the subsequent
conversation. Using the same formalization (presented above), let Lpb be the prop-
ositional content expressed explicitly by B in B1 and, in the same manner, let Lpb€⊃
¬Wpb symbolize the propositional content implicated in B1. A can only think that,
for B, at least one pistol still is not working right at time tn, and from there, deduce
that for B, it is false to say, for all pistols, that «the pistols seem to be working well»
at time tn. B1 thus conveys the following propositions, translated into the language
of first-order predicate calculus: (Lpb ⊃ ¬Wpb) ⊃ (∀p)¬Wp. Italics denote what is
implicated or implied. At this point in the conversation, we do not know how A will
interpret B1. Let us examine the possible inferences for A. Note, however, that the
expression «the pistols aren’t working well», which will be understood as a contra-
diction of A1, will be taken to communicate information shared by all speakers and
readers of this dialogical sequence: «¬(Wpf€∧€Wpb) or the equivalent expression
(¬Wpf€∨€¬Wpb)». On a linear representation of time, this gives Figure 2.
By saying «the back one still lifts the sheet a little», B gives the value true (1) at
instant tn of the utterance to the proposition «the back pistol lifts» (Lpb); by utter-
ing the adverb «still», he seems to also indicate the truth of this proposition at
some instant set in the past. Without any other precise indication about what mo-
ment in the past B is referring to, the proposition «the back pistol lifts» takes on the
value true (1) at all times prior to utterance time tn. Let us draw up the truth table
of this proposition (Lpb) using a Prior matrix (following Gardies, 1975). Let us
also put in the value false (0) that B claims to assign to the proposition «the pistols
are working well» (Wp) at times tn and t’.

t’ tn
φ’ ¬φ
Wpb Wpb
¬Wp ¬Wp
¬Wpf ∨ ¬ Wpb ¬Wpf ∨ ¬ Wpb

Figure 2.╇

B’s shift A’s shift

t’ tn

Lpb 1 1
Wp 0 0

Figure 3.╇ Representation of B’s opinion (in B1)

 Alain Trognon and Martine Batt B’s argument and A’s interpretation of it

For A, B1 is not very forceful and does not seem to be truly informative. By saying
«yeah, maybe», A is apparently insinuating what B can understand: that his utter-
ance does not convey more information than he already possessed, and that that
information is not relevant to the current topic because it does not question his
first assertion. A more or less accepts B’s assertion: the back pistol lifts. A adds a
restriction, introduced by the conjunction «but»: «But I didn’t have any pulp com-
pared to yesterday; I didn’t have any pulp, hm. What A says, in A2, is that it is not
because the back pistol lifts that the pistols (plural) accumulate paper pulp, i.e.,
that it is not because the back pistol lifts that the pistols are not working well.
Thus, A says that during his work shift (which corresponds to time period j),
he noticed that, granted, the back pistol lifted (Lpb), maybe, but the pistols were
working well (Wp) anyway. On a Prior matrix (1957), A’s discourse in A2 can be
formalized as.
As a comparison of the two Prior matrices shows (Figures 3 and 4), the B1-A2
pair thus constitutes the exchange that «materializes» the interlocutors’ divergence
regarding B’s implied thought. Indeed, until A2, there is no debate about B’s ex-
plicit discourse but about what B implies in B1. From B2 on, the implicit will be
formulated explicitly, which gives the following conversational structure (Roulet
et al.,€op.€cit.)€(Figure 5).

B’s shift A’s shift


Lpb 1
Wp 1

Figure 4.╇ Representation of A’s opinion (in A2)

I Assertion A1: the pistols (...) working well

I Implicitly opposed assertion B1: the back one (...)
I Assertion validating the explicit A2a: well, yeah, maybe
E I Assertion canceling the implicated utterance heard
A2b: but I didn't have any pulp
I Explicitly opposed assertion B2: I had some
E I Question A3a: you had some?
I Assertion A3b: me, I didn't have any

Figure 5.╇ (E = exchange ; I = move)

Interlocutory logic 

This diagram shows that A1 generates an exchange where the interlocutors re-
spond explicitly as well as implicitly. B1 consists of the implicit negation of A1. The
explicit constituent of B1, the back pistol lifts (Lpb), is an utterance that gets vali-
dated by the dialogue in A2a. By contrast, its implicit constituent is already being
debated in A2b. For A, B1 implies an assertion that does not have a truth condition
in utterance context tn. For A, the implication relation suggested in B1 is false, i.e.,
for A, ¬(Lpb€⊃€¬Wp) is true. For A, the back pistol lifts, granted, but the pistols are
not working improperly when B1 is uttered: Lpb is true, and, a minima, ¬Wpb is
false (¬¬Wpb). Dialogical format of the debate

Let us describe the dialogical format of the debate that took place in the first part
of the conversation. This will point out the strategy used by each of the interacting
individuals as they attempted to impose their respective points of view. The format
of this dialogue is described by Barth et al. (op.€cit.), and some of the dialogical
rules described by these two philosophers are applied by the conversers. The dif-
ference between this debate and the conflicts of opinion presented by Barth and
Krabbe is that here, it is an implicated implication relation that is being debated, so
the formula under debate is not completely formulated. The dialogical analysis is
presented in Table€4. Reciprocal insemination of each partner’s propositions

To analyze the thought processes carried out by the two agents during this conver-
sation, we must apply the methods of natural deduction. We have seen that A and
B share knowledge about the way the paper machine functions and about the «the-
oretical» implications of a potential breakdown; this mutual knowledge thus be-
longs to the set of conversation premises. Let us now prove that if A reasons by
taking assertion B1 as a hypothesis, and if he «combines» this assumption with
proposition A1 he himself uttered earlier, then he can deduce the opposite of what
B implicated: even if the back pistol lifts, the pistols are working well, which we
shall symbolize as Lpb€⊃€Wp.€A demonstration using this same method proves
that, on the basis of assertion B2 assumed to be true and logically linked to A’s own
discourse, A can legitimately conclude: the back pistol «is working well», Wpb.
This gives Table€5.
 Alain Trognon and Martine Batt

Table 4.╇ Dialogical Table of the Debate


Wp A1
A communicates an assertion to B: initial thesis
Lpb B1
B hears A’s assertion, B makes an assertion
himself, and tacitly defies A about Wp; he seems
to not accept Wp (doubt, disbelief, disagree-
ment?). B1 is thus considered to be the source of
the debate. B thereby becomes A’s opponent. B’s
opposition of A1 has to do with the consequent of
an implicated implication relation, which is
written: Lpb ⊃ ¬Wp. So, Lpb⊃ ¬Wp expresses an
attack of A1 via B1: B1 = aA1
( ?)Wpb
A’s response is a direct (structural) defense:
apparently, for the purposes of this debate, A was
inclined to accept Lpb, but was also ready to
defend Wp. A’s response is a challenging
declarative sentence (Barth and Krabbe, op. cit.):
A2 = dA1
Wp A2
B doesn’t question A’s moderate acceptance of Lpb ∧ Wp A2
Lpb ¬(Lpb ⊃ ¬Wp) A2
¬Wpb B2
B doesn’t withdraw his tacit challenge but rather
accepts its criticism. For B, at the very least, it is
Wpb that is not true, which is what B says he be-
B2 = caA2
¬Wpb? A3a
A’s counter-attack of the argument ¬Wpb: A3a =
caB2. By means of this question, A invites B to re-
spond to a sentence of the form: “How do you de-
fend ¬Wpb?”. The interrogative sentence uttered
by A functions as a counter-attack of B2.
Wpd A3b
For B, a verbal attack of the sentence Wp is no A does not withdraw his assertion; he seems to in-
longer possible because B used up the only ar- tend to unconditionally defend Wpb. The conflict
gument he had: Lpb. A’s strategy is a structural is over: A won the debate initiated by the criticism
counter-attack of B1. B is unable to falsify the (B1). He used dialogical rules that are authorized
conditional relation Lpb ⊃ Wpb. B1 thus consti- and strategically recommended for winning, i.e.,
tutes what Barth and Krabbe call an attack or systematically defend Wp against B’s criticism.
critical remark of the first kind. If he wishes to A3b = dA1
pursue the discussion, B can do so, but he has
lost this particular declared conflict of opinion.
Interlocutory logic 

Table 5.╇ Joint reasoning process to explain the solution to the pulp-projection
problem (first part)


Shared premises: â•⁄╆Wp = Wpf ∧ Wpb

¬Wp = ¬(Wpf ∧ Wpb)
¬Wp = ¬Wpf ∨ ¬Wpb
Lpb B1 Explicit discourse Wp A1

Conclusion of an accessible A’s deduction which integrates B’s

line of reasoning discourse B1 into his own
discourse: Lpb ⊃ Wp
Explicit discourse Lpb ∧ Wp A2
¬(Lpb ⊃ ¬Wp) A2
¬Wpb B2
Conclusion of an accessible A’s deduction which integrates B’s
line of reasoning discourse B2 into his own
discourse: Wpb
Explicit discourse Wpb ∨ ¬Wpb A3a
Wpb A3b

2.4.2 Solving the problem and building intersubjective knowledge Discursive architecture B’s experimentation, related in B3
Starting in B3, B explains himself. He reconstructs all of the cognitive reasoning
steps he took before being able to assert, at time tn in B2, «I had some (pulp)».
During the time period representing his shift (period i), at time t’’, he identified a
problem, that it was squirting out  and made a diagnosis, that  it (the back pis-
tol) was responsible for the squirting (that made it squirt out) and it was moving
the sheet away a little . This is why (because) at intermediate time t’, he decreased
it a little more  (the back pistol). Thus, in B3, B states what he implicated in B1,
and what A indeed realized that he was implying: by saying «it lifts» he meant «it
isn’t working right». He notes that even though he had adjusted the back pistol
 Alain Trognon and Martine Batt

(note the use of the past perfect tense in B3a), there is projected pulp.€This line of
reasoning, as B tells it to his interlocutor, can be described as:
B2: I had some 

B3a: And I had decreased it a little,

B3b: because I felt it was moving the sheet away a little and

B3c: that made it uh... squirt out
In B’s discourse, then, we find, firstly, the temporal modalities that define (i) the
time period of his work shift, period i (A’s period being period j), (ii) the utterance
time, tn, and (iii) the times that preceded it, t’’ and t’. Secondly, we find the tempo-
ral modalities that reconstruct the order of the search for a solution to the problem
(diagnosis, hypothesis, action) expressed via different verb conjugations. To these
temporal modalities, one can add a modality assigned to the predicate «lift». In-
deed, B says he decreased the opening of the back pistol «a little». B’s action, which
involved preventing the back pistol from lifting, was thus incomplete. So when he
claims that the back pistol still lifts, A can understand that the back pistol still lifts
a little even after the lifting has been decreased a little. To account for this modality,
we thus assign the value 0.5 to the proposition Lpb. At utterance time tn, B seems
to interpret the modality «a little» in such a way that he attributes the semantic
value true to the predicate «lift» (i.e., the back pistol lifts) and in the universe situ-
ated in the past, at time t’, the value false (by decreasing it a little more, the pistol
doesn’t lift). Put into a Prior matrix, B’s discourse is represented on Figure 6. B becomes aware of his knowledge

The conclusion to which B is thereby committed is that no matter how the back
pistol is set, whether it lifts or does not lift, there is pulp (). However, at the time
of the conversation, tn, B again entertains the theory that the back pistol is the cause
of the malfunction, and this is what grounds his implicit theory. He will not be held
accountable for the proposition to which he committed. Since there are pistols
only at the front and back of the machine, B commits to the proposition that

B’s shift A’s shift

t” t’ tn

Lpb 1 0.5 = 0 0.5 = 1

Wp 0 0 0

Figure 6.╇ Representation of B’s opinion (from B3 on)

Interlocutory logic 

the cause of the malfunction is the set€«front pistol and back pistol», irrespective of
the fact that the back pistol lifts. He does not seem to be «aware» of this proposi-
tion, and it is the conversation that will make him understand the logical deduction
that follows from what he does. The hypothesis that the front pistol is responsible
for the problem should arise in his thoughts, but it is A who formulates it in A4-
A5: «(...) I reopened it, it’s the front one (...)» and if there is accumulated pulp, then
it is the responsible one. At time t, which came before tn and after t’, A changed the
settings of the front pistol (Lpf) and for him, there is no pulp.€This can be formal-
ized using the anteriority predicate «U»: Ut’¬Lpf tLpf tnWp, which gives Figure 7.
B can only agree with A, and in B5, he will finally be held accountable for the
proposition to which he committed. It all appears as if he were becoming aware of
his knowledge. The two men agree: «because they’re set right». This give Figure 8.
Being prompted by a collective or joint intentionality, namely the desire to
make the paper machine function properly, the workers are satisfied. It is A who
expresses this in A6: «It’s just as good, huh?» Dialogical format of the collaboration

We are going to show here what discursive format the pair (A, B) used to overcome
the conflict and manage to co-produce the solution upon which they agreed. The
«instrument» that permitted this was a subtle dialogical interplay based on con-
cessions by both parties. Briefly, A knew how to remain quiet before trying to
make his thesis acceptable by B, and B knew how to listen. While it is his own

B's shift: i A's shift: j Changeover

t'' t' t tn
¬Wp Wp Wp
Lpb Lpb
¬Lpf Lpf Lpf

Figure 7.╇

B’s shift A’s shift

t” t’ t tn

Lpb 1 0.5 0.5 0.5

Lpf 0 0 1 1
Wp 0 0 1 1

Figure 8.╇ Representation of A’s and B’s opinions (from A4 on)

 Alain Trognon and Martine Batt

discourse that threatens to make B lose face (Goffmann 1959), it is A who, through
his silence, neutralizes this potential threat. After B3, A could point out to B that
his utterance consisted quite simply in reinforcing the opposite thesis to the one he
was claiming to defend: through his argumentation, B in fact «fuels» A’s point of
view. A does not so this; he does not cross over into that territory, but sets forth an
intermediate thesis that finally convinces B. This move by A constitutes a sociocog-
nitive process, insofar as it is both formally «correct» and socially protective of the
interlocutor. Shared thoughts resulting from idea confrontation

Our analysis based on natural deduction shows that as early as B3, B can under-
stand that what he says means «it is not true that the slight lifting of the back pistol
implies that the back pistol is not working properly», because the logical conse-
quence of his argumentation is just the opposite, that the back pistol is set right
when it lifts a little. B’s maneuver thus contradicts his own conception.
Indeed, B seems to believe that the fact (and we shall see below, solely that
fact) that the back pistol lifts the sheet a little is what is responsible for the pro-
jected pulp.€Let Lpb be the cause of ¬Wp, i.e., «Cause (Lpb, ¬Wp)». We suggest
expressing this causal relation logically using an equivalency (Lpb€≡€¬Wp), in such
a way that Lpb is a necessary and sufficient condition of ¬Wp.€Let (Lpb€≡€¬Wp)€⊃
((¬Wp€ ⊃€ Lpb) ∧ (Lpb€ ⊃€ ¬Wp)), or – given that by counter-positioning
(¬Wp€⊃€Lpb)€≡ (¬Lpb€⊃€Wp) – ¬Wp€≡€((¬Lpb€⊃€Wp) ∧ (Lpb€⊃€¬Wp)). The expla-
nation proposed by B in B3 adds this last expression to his set of premises, and it
adds in particular the left-hand expression of this conjunction, i.e. (¬Lpb€⊃€Wp).
But with the observation he relates in B2-B3, B should realize that he is stating
¬(¬Lpb ⊃ Wp) and thus, that his set of premises is inconsistent. As for A, while he
bases his reasoning on what B says, and he makes the logical connection between
these statements and his earlier conclusions, he notes the same logical conse-
quences as those that should be obvious to B. In sum, after B3, if each interlocutor
were to state the implications of what he was saying and ascertaining from what was
said, he would formulate the proposition that «whether the back pistol lifts or
doesn’t lift, it works properly».
In his relationship with «nature», B learns nothing; it is by way of his partner
that he becomes able to learn. If he had expressed the result of his manipulation in
a proposition, he would have produced an inconsistent discourse. Because he does
not express the conclusions of his experiment, he does not produce an inconsistent
discourse. However, he is committed to inconsistency because he formulates all of
the elements which, logically «worked out», lead to a contradiction.
It is A who expresses a new proposition: Lpf, the front pistol was reopened. This
proposition, which B does not contradict, logically commits A to concluding that the
Interlocutory logic 

pistols are set right; it is B who states this. This all seems to suggest that B assumed that
A4 was true, and that he incorporated it into his own thinking before saying «they’re
set right». The cognitive moves in this conversation are represented in Table€6.
More schematically, three moves show up in the interlocution: (i) two separate
points of view are asserted in succession, (ii) the interlocutors each carry out their
own lines of thinking irrespective of the other, and (iii) the interlocutors each car-
ry out their lines of thinking by integrating certain propositions of their partner,
in what one might call a subliminal manner or by formulating, in the conversa-
tional space, the conclusions they came to draw in the process.
It is the (iii) process which opens the possibility for the interlocutors to con-
verge on Wp.6 We have named it a «learning in interaction» (Trognon et al. 2003,
2006; Schwarz et al. 2008). A Learning in interaction is a process by which

Table 6.╇ Joint reasoning process to explain the solution to the pulp-projection problem
(second part)

¬Lpb B3a Explicit discourse

¬Wpb ∧ ¬Lpb B3b
Lpb ⊃ ¬Wpb B3c
(¬Lpb ⊃ Wpb) B3c
B’s commitment Conclusions of an accessible A’s deduction
based on his discourse: line of reasoning which integrates B’s
(Lpb ⊃ ¬Wpb) ∧ (¬Lpb ⊃ Wpb) discourse
Wpb into his own discourse:
(Lpb ∨ ¬Lpb) ⊃ Wpb Wpb
¬Lpb ⊃ Wpb
(Lpb ∨ ¬Lpb) ⊃ Wpb
Explicit discourse Lpf A4a
¬Wpf A4b
(¬Lpf ⊃ ¬Wpf) ∧ A4b
(¬Wpf ⊃ ¬Lpf)
B’s deduction Conclusions of an accessible A’s commitment
which integrates A’s discourse line of reasoning based on his own discourse:
into his own discourse: (Lpf ∧ Lpb) ⊃ Wpb
(Lpf ∧ Lpb) ⊃ Wpb
Wp B5 Explicit discourse Wp A6
Shared conclusion: Wp

6. We demonstrate the possibility for the intervention to converge on Wp thanks to learning

in interaction formula:
 Alain Trognon and Martine Batt

an interlocutor takes an inference that he/she constructed using a «thesis» of his/

her interlocutor and integrates it into his/her own set of propositions as a hypoth-
esis. This operation, which integrates «the intersubjective into the intrasubjective»
is theorized as a “discharge of assumptions” during a natural deduction process.
“Discharging assumptions” during natural deduction amounts to «replacing a
propositional form p€that was asserted as an assumption, by a propositional form
in conditional form but asserted unconditionally» (Gochet et al. 1990:€136). More
intuitively, discharging an assomption in natural deduction is «moving» a sub-
reasoning proposition conceived on the basis of that hypothesis into the reasoning
that contains it. The following natural deduction illustrates this process:
Rank Principal Auxiliary Derivation
Reasoning Reasoning Rules
1 p Premise
2 q Hypothesis
3 Assumption
4 q⊃r r Discharge

In natural deduction theory, one is not concerned with the origin of the hypotheses
that trigger a sub-reasoning process, for «(...) we can, at any stage of a derivation,
introduce the auxiliary hypothesis of our choice, provided, of course, we begin a
new sub-derivation». Our contribution to this theory therefore consists in adding
to a converser’s auxiliary hypotheses certain propositions of his/her partner.

2.4.3 The cognitive layers of an interlocution

On the surface, this discursive organization is only the emerging part of a deeper
cognitive framework that «supports» the conversation and is distributed normally
between the participants.
But one can also consider this framework as a whole, from the standpoint of a
third party such as an instructor, a team leader, or a participant striving to grasp
the «group mind» that comes out of an interlocution. Its cognitive dimension in
this case is represented as a succession of joint, interwoven activities that produce
an object – the conversation – that is relatively independent of the individual ac-
tivity of the agents.


It has been fifteen years that Jean Caron, a French psycholinguist, feared that the
success of the interactionist paradigm might lead to a regression in linguistics. He
Interlocutory logic 

wrote: “In the name of the «linguistics of talk», we could lose sight of language. (...)
The skeleton in a living body cannot be seen; and of course, the flesh is so much
more interesting! But if we take the skeleton away, only a shapeless, limp mass re-
mains. Won’t taking away language, its structures and its constraints, have us prac-
ticing a spineless linguistics?” (Caron, 1989a: 138, our translation). Twenty years
later, we can reassure him. The «interaction order» did not invade the «institution
order» nor the «language order». Granted, the language institution is partly im-
mersed in the interaction order (see Figure 9), so the formal structures of language
are available to interactions, which select the ones that suit the ever-changing state
of the interaction.
This explains why a sentence is sometimes recognized on the basis of syntax
and sometimes on the basis of semantics, depending on the requirements of the
situation (Caron, 1989b). But it is only partially that the formal structures of lan-
guage depend upon the interaction order, regardless of Thibaut’s statement that
«the notion of a ‘language system’ does not refer to some reified entity ‘out there’
which exists independently of the social meaning-making practices of a given so-
cial group and which has an independent causal status» (2002:€136). Besides, nei-
ther Garfinkel and the conversational analysts on one side, nor Gumperz and in-
teractionist sociolinguists on the other, ever claimed to have discovered new social
categories: it is their embodiment in social relations that interests these authors.
Moreover, Gumperz’s definition of communicative practices, in line with Saus-
sure’s, assumes that language is a system (Gumperz in Prevignano and Thibaut,

Social Institution

The transition
from “the x x Speaking Mass
intersubjective x x
to the x x
Interaction (the
“Interaction Order”
= communicative
practices (Gumperz))

Persons in interaction

Figure 9.╇ The system of interaction

 Alain Trognon and Martine Batt

Lewis wrote in 1972: “I distinguish two topics: first, the description of possible
languages or grammars as abstract semantic systems whereby symbols are associ-
ated with aspects of the world; and second, the description of the psychological
and sociological facts whereby a particular one of these abstract semantic systems
is the one used by a person or population. Only confusion comes of mixing these
two topics” (Lewis, 1972:€170). In the end, then, after thirty years of existence, the
interactionist paradigm will have recognized this distinction, on inventing an in-
termediate area between the two terms of the dichotomy. And it is what Bakhtin
want in 1929 (Gardin, 1989; Thibaut, 2002).
But there are many unresolved difficulties. The main ones are epistemologi-
cal. Firstly, what sort of object is the “interaction”, and in particular “interlocu-
tion”? Here, more than a taxonomy, we need a sort of generative theory of inter-
locution, even rough, capable of generating dialogue types from ordinary
conversation, in much the same way as a generative grammar produces sentence
types from a core sentence (Trognon and Bromberg, 2007; Trognon et al. in
press.). With such a theory, one could work under a thesis now acknowledged
both in conversational analysis (see Heritage, 1990) and in the cognitive psy-
cholinguistics of dialogue (see Clark, 1996), in fact, shared by all researchers in
the domain, which stipulates that ordinary conversation is the matrix of all oth-
er forms of conversation. Secondly, how to combine formal and observational
researches on “interaction” (Searle: 1992 vs Duncan, Fiske et al.: 1985). Con-
cerning this second sort of researches what becomes of statistical proofs when it
comes to demonstrating structures? For instance, non-preferential actions are
not always correlated with the discourse pauses, delays, and repetitions assumed
to indicate – both to the interactants and to the analyst – that those actions are
not the preferred ones. Does this mean that it is invalid to establish a corre-
spondence€between them? The answer is no, if we agree with Schegloff (1993)
that the informal quantification used in conversational analysis and statistical
quantification are two fundamentally different enterprises, that the latter cannot
be substituted for the analysis of isolated cases and groups of sequences, and
that, in sum, statistical quantification is not an educated form of informal quan-
tification. Conversely, it is not because we recognize the merits of informal
quantification that statistical tools should be avoided. But what is true is that the
ambivalence (exhibited by Sacks, 1987, for example) and sometimes even casu-
alness of certain researchers with respect to statistical quantification (Duncan,
Fiske et al.: 1985) does not help in conducting a discussion that deserves, fol-
lowing Gumperz’s (2002) example, an unbiased, dispassionate approach.
Finally, there is still much ground to cover to explore the «interaction order».
Interlocutory logic 


Auroux, Sylvain. 1989. ‘‘L’interaction et les limites du rationalisme.’’ Buscila 221–232.

Bakhtin, Mikhail. 1929. [V. N. Volochinov]. Le marxisme et la philosophie du langage. Essai
d’application de la méthode sociologique en linguistique. Paris: Minuit.
Barth, Else M. and Krabbe, Erik C. W. 1982. From axiom to dialogue. Berlin, N. Y.: Walter de
Bateson, Gregory. 1956. “The message ‘This is Play’.’’ In Group Processes: Transactions of the Second
Conference, Bertram H. Schaffner (ed.), 145–242. New York: Josiah Macy Jr. Foundation.
Batt, Martine and Trognon, Alain. 2009. “Ergonomie cognitive d’un conseil génétique pour le
test présymptomatique de maladie de Huntington.’’ Psychologie du Travail et des
Organisations, 15 (1): 21–41.
Bernicot, Josie and Trognon, Alain. 1997. “Dimensions de la conversation.’’ In Conversation,
interaction et fonctionnement cognitif, Josie Bernicot, Alain Trognon and J Caron-Pargue
(eds.), 15–41. Nancy: Presses Universitaires de Nancy.
Bernicot, Josie and Trognon, Alain. 2002. “Le tournant pragmatique en psychologie.’’ In La prag-
matique en psychologie, Josie Bernicot, Alain Trognon, Michel Musiol and Michèle Gui-
detti (eds.), 13–32. Nancy: Presses Universitaires de Nancy.
Boutet, Jeanne. 1989. “Construction sociale du sens et interaction.” Buscila 196–204.
Brixhe, Daniel, Saint-Dizier, Valérie and Trognon, Alain. 2000. “Résolution interlocutoire d’un
diagnostic. Etudes d’explications dans un corpus de dialogues finalisés.’’ Psychologie de
l’interaction 9–10: 211–237.
Carlson, Lauri.1983. Dialogue games. An approach to discourse Analysis. Dordrecht: Reidel.
Caron, Jean.1989a. In “interventions à la table ronde.” Buscila 139.
Caron, Jean. 1989b. “Que peut apporter la psycho-linguistique expérimentale à l’analyse de
l’interaction?” Buscila 76–94.
Caron, Jean. 1997. “Psychologie cognitive et interactions conversationnelles.” In Conversation,
interaction et fonctionnement cognitif, Joise Bernicot, Alain Trognon and Josiane
Caron-Pargue (eds), 221–238. Nancy: Presses Universitaires de Nancy.
Clark, Herbert H. 1996. Using language. Cambridge: Cambridge University Press.
David, René, Nour, Karim and Raffalli, Christophe. 2001. Introduction à la logique. Paris: Dunod.
Dennett, Daniel C. 1987. The Intentional Stance. Cambridge: The MIT Press.
Di Pellegrino, L., Fadiga, L., Fogassi, V., Gallese and G. Rizzolatti. 1992. “Understanding motor
events: a Neurophysiological study.” Experimental Brain Research 91: 176–180.
Doise, Wilhem, Mugny, Gabriel and Perret-Clermont, Anne Nelly. 1975. “Social interaction and
the development of cognitive operations.” European Journal of Social Psychology 5: 367–383.
Dubois, Bruno. 2002. “Lobe frontal et régulation du comportement.” In La pragmatique en
psychologie, Josie Bernicot, Alain Trognon, Michel Musiol and Michèle Guidetti (eds.),
269–279. Nancy: Presses Universitaires de Nancy.
Duncan, Starkey D., Fiske, Donald W., Denny, Rita, Kanki, Barbara G., and Mokros, Hartmut B.
1985. Interaction structure and strategy. New York: Cambridge University Press.
Gardies, Jean Louis. 1975. La logique du temps. Paris: PUF.
Gardin, Bernard. 1989. “L’interaction en sociolinguistique.” Buscila 143–146.
Garfinkel, Harold. 1967. Studies in Ethnomethodology. Englewood Cliffs NJ: Prentice Hall.
 Alain Trognon and Martine Batt

Garfinkel, Harold. 1990. “The Curious Seriousness of Professional Sociology.’’ In Les formes de
la conversation, Bernard Conein, Michel de Fornel and Louis Quéré (eds.), 1, 145–161.
Paris: CNET.
Garfinkel, Harold 1996. “Ethnomethodology’s program.” Social Psychology Quaterly 59(1): 5–21.
Geniffey, Yves and Trognon, Alain. 1986. “La chasse aux étoiles, sur les formalismes statistiques
en psycholinguistique.” In Pratiques des sciences sociales, 67–73. Nancy: Presses Universitaires
de Nancy.
Gentzen, Gerhard. 1935. “Untersuchungen über das logische Schliessen”. Mathematische
Zeitschrift 39: 176–210 and 405–431.
Gochet, Paul and Gribomont, Paul. 1990. Logique. Méthodes pour l’informatique fondamentale.
Paris: Hermès.
Goffman, Erving. 1959. The presentation of self in everyday life. NY: Doubleday.
Goffman, Erving. 1971. Relations in Public: Microstudies of the Public Order. New York: Harper
and Row.
Goffman, Erving. 1981. Forms of Talk. University of Pensylvania Press.
Goffman, Erving. 1983. “The interaction order.” American Sociological Review 48: 1–17.
Good, David. 1990. “Repair and cooperation in conversation.’’ In Computers and conversation,
Paul Luff, David, M. Frolich, Michel G. Gilbert (eds), 133–150. New York: Academic Press.
Grice, Herbert Paul. 1979. “Logique et conversation.” Communications, 30, 57–72.
Grice, Herbert Paul. 1982. “Meaning Revisited.’’ In Mutual Knowledge, N. V. Smith (ed), 223–243.
New York: Academic Press.
Grusenmeyer, Corinne and Trognon, Alain. 1995. “L’analyse interactive des échanges verbaux en
situation de travail coopératif: L’exemple de la relève de poste.” Connexions 65 (1): 43–62.
Grusenmeyer, Corinne and Trognon, Alain. 1996. “Structures of natural reasoning within func-
tional dialogues.” Pragmatics and cognition 4 (2) 305–346.
Gumperz, John. 1982. Discourse Strategies. Cambridge: Cambridge University Press.
Gumperz, John. J. 2002. “Response essay.” In Language and Interaction: Discussions with John J.
Gumperz, Susan, L. Eerdmans, Carlo, L. Prevignano and Paul J. Thibault (eds.), 105–125.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
Heritage, John. 1990. “Interactional Accountability: A Conversation Analytic Perspective.” In Les
formes de la conversation, Bernard Conein, Michel de Fornel and Louis Quéré (eds.), 1, 23–49.
Paris: CNET.
Hintikka, Jaakko. 1976. The semantics of questions and the questions of semantics. Amsterdam:
North-Holland Publishing Company.
Hintikka, Jaakko and Kulas, Jack. 1983. The game of language. Studies in Game Theoretical
Semantics and its applications. Dordrecht: D. Reidel.
Hutchins, Edwin. 1995. Cognition in the wild. Cambridge CA: Bradford Books, MIT Press.
Jacques, Francis. 1988. “Trois stratégies conversationnelles: conversation, négociation, dialogue.”
In Echanges sur la conversation, Jacques Cosnier, Nadine Gelas and Catherine Kerbrat-
Orecchioni (eds.), 45–68. Paris: Editions du CNRS.
Jakobson, Roman. 1955. Aphasia as a linguistic problem. In On expressive language, Heinz
Werner (ed.), Worcester: Clark University Press.
Kerbrat-Orecchioni, Catherine. 1989a. “L’approche interactionnelle en linguistique.” Buscila 7–25.
Kerbrat-Orecchioni, Catherine.1989b. L’implicite. Paris: Armand Colin.
Kerbrat-Orecchioni, Catherine.1990. Les interaction verbales, Tome 1. Paris: Armand Colin.
Kerbrat-Orecchioni, Catherine. 1997. “A multilevel approach in the study of talk-in-interaction.”
Pragmatics 7 (1): 1–20.
Interlocutory logic 

Kerbrat-Orecchioni, Catherine. 1998. “La notion d’interaction en linguistique: Origines, apports,

bilan.” Langue Française 117: 51–67.
Kerbrat-Orecchioni, Catherine. 2005. Le discours en interaction. Paris: Armand Colin.
Lacan, Jacques. 1981. Le séminaire III, les psychoses. Paris: Seuil.
Laughlin, Patrick. R., Hatch, Ervin C., Silver, Jonathan S., Boh, Lee. 2006. “Groups perform bet-
ter than the best individuals on letters-to-numbers problems: Effect of group size.” J. Pers.
Soc. Psychol. 90(4): 644–651.
Laux, Jennifer, Trognon, Alain and Batt, Martine. 2008. “Une approche expérimentale et micro
génétique de la tâche de sélection de Wason réalisée par des dyades pratiquant un dialogue
coopératif.” Psychologie Française 53: 375–397.
Lave, Jean. 1993. Cognition in practice. New York: Cambridge University Press.
Levinson, Stephen C. 2002. “Contextualizing “contextualization cues.” In Language and Interac-
tion: Discussions with John J. Gumperz, Susan L. Eerdmans, Carlo L. Prevignano and Paul J.
Thibault (eds.), 31–39. Amsterdam/Philadelphia: John Benjamins Publishing Company.
Levinson, Stephen. 1983. Pragmatics. Cambridge University Press: Cambridge.
Lewis, David, K. 1972. “General Semantics” In Semantics of Natural Language, Donald Davidson
and Gilbert H. Harman (eds.), 169–218. Dordrecht: D. Reidel Publ. Co.
Light, Paul and Perret-Clermont, Anne-Nelly. 1989. “Social context effects in learning and test-
ing.” In Cognition and social worlds, John A. Sloboda (eds.), 99–112. Oxford: Oxford Science
Publications, University Press.
Lorenzen, Peter. 1967. Métamathématique. Paris: Gauthiers-Villars.
Marro-Clément, Pascale, Trognon, Alain and Perret-Clermont, Anne-Nelly. 1999. “Processus
interlocutoires dans une tâche de conservation des liquides”. In Apprendre dans l’interaction,
Michel Gilly, Jean Paul Roux and Alain Trognon (eds.), 163–180. Nancy: Presses
Universitaires de Nancy.
Maynard, Douglas. W. and Zimmerman, Don H. 1984. “Topical Talk, ritual and the Social Or-
ganization of Relationships.” Social Psychology Quaterly 47: 301–316.
Mead, Georg, H. 1934. Mind, Self and Society. Chicago: University of Chicago Press.
Moeschler, Jacques. 1985. Argumentation et conversation. Eléments pour une analyse pragma-
tique du discours. Paris: Hatier.
Mondada, Lorenza. 1995. “Introduction: pour une approche des formes linguistiques dans les
dynamiques interactionnelles.” In Formes linguistiques et dynamiques interactionnelles,
Lorenza Mondada (ed.). Cahiers de l’ILSL 7: 1–18.
Peter-Favre, Claire. 2002. “Discours et conversation chez les traumatisés cranio-cérébraux”. In
La pragmatique en psychologie, Josie Bernicot, Alain Trognon, Michel Musiol and Michèle
Guidetti (eds.), 281–304. Nancy: Presses Universitaires de Nancy.
Piaget, Jean. 1928/1976. “Logique génétique et sociologie.” In Les sciences sociales avant et après,
Giovani Busino (ed.), 44–80. Genève: Droz.
Piaget, Jean. 1932. Le jugement moral chez l’enfant. Paris: Alcan.
Prevignano, Carlo L. 2002. “On Gumperz and the minims of interaction.” In Language and In-
teraction: Susan, L. Eerdmans, Carlo, L. Prevignano, Paul J. Thibault (eds.), 63–77. Amster-
dam/Philadelphia: John Benjamins Publishing Company.
Prevignano, Carlo L. and di Luzio, Aldo. 2002. “A discussion with John J. Gumperz.” In Language
and Interaction: Susan, L. Eerdmans, Carlo, L. Prevignano, and Paul J. Thibault (eds.), 7–29.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
 Alain Trognon and Martine Batt

Prevignano, Carlo L. and Thibaut, Paul J. 2002. “Continuing the discussion with John J. Gumperz.”
In Language and Interaction: Susan, L. Eerdmans, Carlo, L. Prevignano, and Paul J. Thibault
(eds.), 149–161. Amsterdam/Philadelphia: John Benjamins Publishing Company.
Prior, Arthur Norman. 1957. Time and Modality. Oxford: Clarendon Press.
Quéré, Louis. 1989. ““La vie sociale est une scène” (Goffman revu et corrigé par Garfinkel).” In
Le parler frais d’Erving Goffman, 47–82. Paris: Minuit.
Roulet, Eddy, Auchlin, Antoine, Moeschler, Jacques, Rubattel, Christian and Schelling, M. 1985.
L’articulation du discours en français contemporain. Berne: Peter Lang.
Sacks, Harvey 1984. “Notes on Methodology.” In John Maxwell Atkinson and John Heritage
(eds.), Structures of Social Action: Studies in Conversation Analysis, 21–27. Cambridge:
Cambridge University Press.
Sacks, Harvey. 1987. “On the Preferences of Agreement and Contiguity in Sequences in Conver-
sation.” In Talk and Social Organization, Graham Button and John R. Lee (eds.), 54–69.
Clevedon: Multilingual Matters.
Sannino, Annalisa, Trognon, Alain and Dessagne, Lara. 2003. “A model for analyzing knowl-
edge content and processes of learning a trade within alternance vocational training.” In
Between school and work: new perspectives on transfer and boundary-crossing, Terrtu Tuomi-
Groh and Yrjö Engeström (eds.), 271–289. Amsterdam: Pergamon.
Saussure, Ferdinand de. 1969. Cours de linguistique générale. Paris: Payot.
Schegloff, Emmanuel A. 1991. “Conversation Analysis and Socially Shared Cognition.” In Per-
spectives on socially shared cognition, Lauren B. Resnick, John M. Levine and Steph-
anie D. Teasley (eds.), 150–170. Washington: American Psychological Association, DC.
Schegloff, Emmanuel A. 1993. “To Searle on Conversation: a note in return.” In On Searle on
Conversation, Herman Parret and Jeff Verschueren (eds.), 113–128. Amsterdam: John
Benjamins Publishing Company.
Schwarz, Baruch, Perret-Clermont, Anne Nelly, Trognon, Alain and Marro, Pascale. 2008.
“Emergent learning in successive activities: Learning in interaction in a laboratory context.”
Pragmatics and Cognition 16(1): 57–87.
Searle, John R. 1990. “Collective intentionality and action.” In Intentions in Communications, Philip
R. Cohen, Jerry Morgan and Martha E. Pollack (eds.), 401–415. Cambridge, Mass.: MIT Press.
Searle, John R. 1992. “Conversation.” In (On) Searle on Conversation, Herman Parret and Jeff
Verschueren (eds.), 7–30. Amsterdam: John Benjamins Publishing Company.
Searle, John R. 1995. The Construction of Social Reality. New York: Free Press.
Searle, John R. and Vanderveken, Daniel. 1985. Foundations of illocutionary logic. Cambridge:
Cambridge University Press.
Siegal, Michael. 1991. “A clash of conversational worlds: interpreting cognitive development through
communication.” In Perspectives on socially shared cognition, Lauren B. Resnick, John M. Levine
and Stephanie D. Teasley (eds), 23–41. Washington, DC: American Psychological Association.
Sperber, Dan and Wilson, Deirdre.1986. Relevance: Communication and Cognition. Basil:
Stern, Daniel N. 2004. Un système de motivation primaire et fondamental. Psychiatrie Française
25 (1): 8–20.
Suchman, Lucy. 1987. Plans and situated actions: The problems of human-machine communica-
tion. New-York: Cambridge University Press.
Thibaut, Paul. J. 2002. “Contextualization and social meaning-making practices.” In Language
and Interaction: Susan, L. Eerdmans, Carlo, L. Prevignano, Paul J. Thibault (eds.), 41–61.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
Interlocutory logic 

Thibaut, Paul. J. 2002. “Body dynamics, social meaning-making, and scale heterogeneity. Re-
considering contextualization cues and language as mixed-mode semiosis.” In Language and
Interaction: Susan, L. Eerdmans, Carlo, L. Prevignano, and Paul J. Thibault (eds.), 149–161.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
Trevarthen, Colwyn. 1993. “The self born in intersubjectivity: an infant commumnicating.” In
The Perceived self: ecological and interpersonal sources of self-knowledge, Ulric Neisser (ed.),
121–173. New York: Cambridge University Press.
Trevarthen, Colwyn. 2001. “Intrinsic motives for companionship in understanding: Their origin,
development and significance for infant mental health.” International Journal of Infant
Mental Health 22 (1–2): 95–131.
Trevarthen, Colwyn. 2004. “Learning about ourselves, from children: why a growing human
brain needs interesting companions.” Research and Clinical Centre for Child Development,
Annual Report 2002–2003, 9–44. Hokkaido University: Graduate School of Education.
Trognon, Alain. 1987. “Débrayages conversationnels.” DRLAV 36–37: 105–122.
Trognon, Alain 1991a. “Fonctions de la conversation.” In Le dialogue, George Maurand (ed.),
375–408. Toulouse: C.A.L.S.
Trognon, Alain. 1991b. “L’interaction en général: sujets, groupes, cognitions, représentations
sociales.” Connexions 57: 9–27.
Trognon, Alain. 1992. “L’approche pragmatique en psychopathologie cognitive.” Psychologie
Française 37 (3/4): 191–203.
Trognon, Alain. 1994. “Sur la théorie de la construction interactive du quotidien.” In La con-
struction interactive du quotidien, Alain Trognon, Ulrich Dausendschon-Gay, Ulrich Krafft,
Christiane Riboni (eds.), 7–52. Nancy: Presses Universitaires de Nancy.
Trognon, Alain. 1998. “Interactional Pragmatics of hypnotic Induction.” Pragmatics 8(2): 255–270.
Trognon, Alain. 1999. “Eléments d’analyse interlocutoire.” In Apprendre dans l’interaction,
Michel Gilly, Jean Paul Roux and Alain Trognon (eds.), 69–94. Nancy: Presses Universitaires
de Nancy; Aix-En-Provence: Publications de l’université de Provence.
Trognon, Alain. 2002. “Speech Acts and the Logic of Mutual Understanding.” In Essays in Speech
Acts Theory, Daniel Vanderveken Susumo Kubo (eds), 121–133. Amsterdam: John Ben-
jamins and sons.
Trognon, Alain. 2003. “La logique interlocutoire. Un programme pour l’étude empirique des
jeux de dialogue.” Questions de Communication 4: 411–425.
Trognon, Alain. 2007. “€L’antilogicisme de Gérard Vergnaud.€” In Activités humaines et conceptu-
alisation: Questions à Gérard Vergnaud, Maryvon Merri (eds.), 195–204. Toulouse: PUM.
Trognon, Alain and Batt, Martine. 2003. “Comment représenter le passage de l’Intersubjectif à
l’Intrasubjectif? Essai de Logique Interlocutoire.” L’Orientation Scolaire et Professionnelle 32
(3): 399–436.
Trognon, Alain and Batt, Martine. 2004. “Logique interlocutoire d’un biais de raisonnement
dans un jeu de dialogue.” In Psychologie Sociale de la Communication, Marcel Bromberg
and Alain Trognon (eds.), 135–156. Paris: Dunod.
Trognon, Alain and Batt, Martine. 2007a. “Comment conduire l’examen d’un fragment
d’interlocution au moyen de la logique interlocutoire?” In Enseigner dans l’interaction, An-
tonietta Specogna (ed.), 13–33. Nancy: Presses Universitaires de Nancy.
Trognon, Alain and Batt, Martine. 2007b. “Quelles méthodes logiques pour l’étude de l’interaction
en psychologie?” In Interactions communicatives et psychologie, Claude Chabrol and Isa-
belle Orly-Louis (eds.), 53–65. Paris: Presses Sorbonne Nouvelle.
 Alain Trognon and Martine Batt

Trognon, Alain and Batt, Martine. In press. “Discharging assumptions: a symbolism to construe
learning in interaction.” In Logic and dialogue, J. Caelen, D. Vanderveken and D. Vernant
(eds). Netherlands, Dordrecht: Kluwer.
Trognon, Alain, Batt, Martine and Laux, Jennifer. 2007. “Psychologie Sociale et raisonnement.”
In Psychologies du Raisonnement, Sandrine Rossi and Jean-Baptiste Van der Henst (eds.),
139–168. Bruxelles: De Boeck.
Trognon, Alain, Batt, Martine, Schwartz, Baruch, Perret-Clermont, Anne Nelly and Marro, Pas-
cale. 2006. “Logique interlocutoire de la résolution d’un problème d’arithmétique élémen-
taire au sein d’une dyade.” Psychologie Française 51 (2):€171–187.
Trognon, Alain, Batt, Martine and Saint Dizier, Valérie. In press. “Dialogue games of argumen-
tation.” In Dialogue’s logical properties, Alain Trognon, Martine Batt, Jean Caelen and Denis
Vernant. Nancy: Presses Universitaires de Nancy.
Trognon, Alain and Brassac, Christian. 1993. “L’enchaînement conversationnel.” Cahiers de
linguistique française 13: 76–108.
Trognon, Alain and Bromberg, Marcel. 2007. “L’interaction sociale.” In Psychologie sociale et res-
sources humaines, Alain Trognon et Marcel Bromberg (eds.), 63–94. Paris: PUF.
Trognon, Alain and Collet, Véronique. 1993. “L’autisme et les mystères de l’intentionnalité.”
Revue internationale de psychopathologie 10: 225–257.
Trognon, Alain and Grusenmeyer, Corinne. 1997. “To resolve a technical problem through con-
versation.” In Discourse, Tools and Reasoning: essays on Situated Cognition, Lauren B. Res-
nick, Rogé Säljö, Clotilde Pontecorvo and Barbara Burge (eds), 87–110. New York: Springer.
Trognon, Alain and Kostulski, Katia. 1999. “Eléments d’une théorie socio-cognitive de
l’interaction conversationnelle.” Psychologie Française 44 (4): 307–318.
Trognon, Alain and Saint-Dizier, Valérie. 1999. “L’organisation conversationnelle des malenten-
dus: le cas d’un dialogue tutoriel.” Journal of Pragmatics 31: 787–815.
Trognon, Alain, Saint-Dizier, Valérie and Grossen, Michèle. 1999. “Résolution conjointe d’un
problème arithmétique.” In Apprendre dans l’interaction, Michel Gilly, Jean Paul Roux and
Alain Trognon (eds.), 121–141. Nancy: Presses Universitaires de Nancy; Aix-En-Provence:
Publications de l’université de Provence.
Trognon, Alain, Sorsana, Christine, Batt, Martine and Longin, Dominique. 2008. “Peer interac-
tion and problem solving. One example of a logical-discursive analysis of a process of joint
decision making.” European Journal of Developmental Psychology 5 (5): 623–643.
Van der Henst, Jean-Baptiste. 2002. “Contexte et raisonnement.” In Le raisonnement humain
Guy Politzer (ed.), 271–305. Paris: Hermès.
Vanderveken, Daniel. 1990. Meaning and Speech Acts (1): Principles of Language Use (2):
Formal Semantics of Success and satisfaction. Cambridge: Cambridge University Press.
Vanderveken, Daniel. 1988. Les actes de discours. Bruxelles: Mardaga.
Vygotski, Lev S. 1934. Myshlenie i rech’: Psikhologicheskie issledovaniya. Moscow and Lenin-
grad: Gosudarstvennoe Sotsial’no-Ekonomicheskoe Isdatel’stvo.
Watson, Rod and Sharrock, Wesley W. 1990. “Conversational Actions and Organizational
Actions.” In Les formes de la conversation, Bernard Conein, Michel de Fornel and Louis
Quéré (eds.), vol 2, 21–35. Paris: CNET.
Weigand, Edda. 1999. “Misunderstanding: the standard case.” Journal of Pragmatics 31 (6): 787–815.
Winograd, Terry and Florès Fernando. 1986. Understanding Computers and Cognition. Norwood
(N.J.) Ablex Publishing Corporation.
Beyond symbols
Interaction and the enslavement principle

Stephen J. Cowley
University of Hertfordshire, UK and University of KwaZulu-Natal, South Africa

Humans often contextualize without using cues. While Gumperz showed that
analysis is not sufficient to explain interaction, his view of what lay beyond
symbols was based in cognitive internalism. Opposing this, prosody can be
shown to contribute directly to conversational sense-making. Humans use self-
organizing dynamics in ways that resemble what happens in gas-lasers. Voices
attract each other and, at times, set off laser like synergies. Using these effects,
we modulate our actions in situation-transcending events that give sense to the
dynamics. Conversation is distributed cognition during which prosodic sense-
making links the world with brains and bodies. Far from being based in word-
forms, interaction and language are dynamical first and symbolic afterwards.

1. Beyond symbols

The linguistic turn of the 20th century produced, among other things, the compu-
tational theory of mind and social constructionism. Breaking with post-Saussuri-
an assumptions, this paper presents another perspective. Rejecting models where
language systems are separated from their use, weight is given to conversing. By
linking the work of Gumperz (1982), Maturana (1978), Hutchins (1995), Love
(2004; 2007) and others, 20th century tradition is shown to exaggerate the impor-
tance of linguistic forms. This appears, above all, in the curious opposition of ex-
ternalism to mentalism. Instead of acknowledging that conversations are founda-
tional to minded behaviour, theorists retreated into two camps. On the one hand,
externalists tended to identify what can be known with symbolic patterns or dis-
course; on the other, cognitive internalists reified symbols by appeal to informa-
tion processing. Historically, of course, Saussure’s (1916) vision of synchronic lin-
guistics can be used to legitimize both sets of views. By opposing system (langue)
to use (parole), he gave us a picture of a cognitive world where context reduces to
no more than the perceived external setting. Having separated people from the
 Stephen J. Cowley

world, agency is ascribed to language. Bizarrely, brains and/or bodies are regarded
as producing and processing social activity (including language). Theorists who
appeal to programmed brains thus make the same error as those who reduce peo-
ple to rule-followers. On both sides, language and interaction are separated from
the full-bodied activity which shapes the experience of human subjects.
One great merit of Gumperz’s opus (e.g. 1982; 2003) was that he showed that
making sense of language depends on events that lie beyond symbols. While
adopting the cognitive internalism of the mid-twentieth century, he saw the “se-
mantic importance of context” (2003: 9). Accordingly, his work took on a practical
relevance that is all too rare in the language sciences. This arose in focusing or
miscommunication which, in his terms, exploited contextualization cues. While
these constructs are discussed below, the move establishes two de facto principles:
– Recognition of interindividual communication
– The importance of vocal patterns beyond symbols
Gumperz’s models are less convincing than his practices. Thus, Levinson notes
lack of theoretical cleanliness (2003: 32), Prevignano doubts that interpretations
often coincide (2003:17) and Thibault (2003a) stresses that indexicality is all per-
vasive. Rather than contrast the ‘symbolic’ with the ‘indexical’, the paper offers an
alternative to contextual semantics. Denying that interaction is an ‘analytical
prime’ (Gumperz, 2003: 106), conversing or first-order language is traced to how
people control dynamics. Invoking Haken’s (1993) enslavement principle, parallels
and contrasts are drawn between gas lasers and conversation. In stressing real-
time contextualizing, humans are shown to orient to patterns in a pico-scale: we
not only use phonological regularities but also voice dynamics that operate in time
domains around the threshold of consciousness (200–300msec). While drawing
on something like the enslavement principle, we monitor the sound of speaking to
set up counter-effects. For this reason interaction analysis needs to be supplement-
ed by other modes of investigation. Language and interaction can be rethought as
distributed cognition (Hutchins, 1995; Clark, 1997; Wheeler, 2004; Cowley &
Spurrett, 2003; Cowley, 2007d, 2009b). Using circular causation, we hear social
norms that enable us to think as we engage in interaction.

2. Bewitched by von Neumann machines

The spell of von Neumann machines led to a 50 year fixation with symbols and rep-
resentations. At the end of the 20th century, indeed, many assumed that both lan-
guage and interaction depend on inferencing that links text with context. It was
widely touted that, whatever else we do, symbolic forms lie at the heart of human
Beyond symbols 

interaction. Regardless of whether internal or external, symbols become the je ne sais

quoi that link putative language-systems, the words actually spoken, and ‘context’.
With Grice (1987), Austin (1962) Searle (1969) and others utterances are supposed
to function because of illocutionary force. Indeed, even Gumperz treats human com-
munication as ‘intentional’ and ‘based on inferences’ (2003:11). Conversational
events are taken to depend, in the first instance, on symbol processing.
Gumperz’s work has been criticized for reliance on information theoretic
models (Levinson, 2003: 33). Regarding indexical signs as modifying symbols
masks the full-bodied nature of semiosis. Appeal to pure contextualizing cues is
mistaken, Thibault argues (2003a), because all interaction is indexical. Given its
situated nature, it is unlikely that cues can ‘add’ to symbolic output. Further while
robots use such architectures, these differ from what is found in nature. As Clark
(1997) argues, intelligent behaviour by animals and people is not controlled by
the central processes of physical symbol systems. Unlike von Neumann ma-
chines, brains do not reduce ‘content’ to physical analogues of linguistic forms.
While ‘Cartesian materialism’ (Dennett, 1991) has many philosophical problems,1
this paper uses dynamic observations to challenge both symbolic models and
contextual semantics. Symbol processing is too slow to explain many human re-
actions. Human interaction is fast because, as in Tetris (Blair & Cowley, 2003),
we rely on embodied coupling. As in playing the computer game (Kirsh & Magl-
io, 1994) events cannot be clarified by appeal to algorithmic models. For that
same reason, interaction analysis needs to be supplemented by other kinds of
While some view discourse as programmed, conversation analysts picture talk
in terms of sequential events. Word-forms are treated as constituting social action
based on rule-following. Rather than view people as living subjects, they become
agents whose subjectivity is (largely) illusory. Eschewing appeal to programs, such
theories posit choice making based on conformity to social rules. Although inter-
nalists and externalists appeal to inferences and norms respectively, both groups
reduce conversations to sequences of units. Real-time dynamics are systematically
excluded from account on purely methodological grounds. It is as if social action
were managed by a computer in the sky. This effect depends on:
– treating transcriptions as formalizations of ‘data’;
– emphasising recurrent patterns;

1. Attacking the ghost in the machine Ryle (1949) pointed out the error of confusing brains
with minds. To invoke a central processing mechanism is metaphorical because brains are dis-
tributed self-organizing systems that integrate events with different evolutionary and develop-
mental histories (Deacon, 1997). Computationally, central processing is too rigid for simulating
adaptive behaviour (Clark, 1997). In addition, many deny ‘psychological reality’ to language and
belief (e.g. Matthews, 1979; Linell, 1979; Harris, 1981; Dennett, 1987).
 Stephen J. Cowley

– analysing discourse in terms of determinate units;

– using methods that pick out program-like regularities
In line with representationalism, events reduces to talk-in-interaction. All context
becomes local. As Linell (2009) argues, no room is left for either trans-situational
regularities or how we use dialogue to transcend current circumstances. Taking
language-processing (whatever that is) for granted, talk is desomatized. By focus-
ing on patterns, analysts overlook circumstances, experience, bodies and relation-
ships. First, singularities are removed by analysis into word-forms. Then, using
transcripts (or, recordings) some detail is restored in invoking features determi-
nate units: interaction consists in sequential arrangements of word-forms. In il-
lustration I sketch an event that is later discussed in detail. Echoing Gumperz,
transcription uses semi-standard spellings and signs represented on typewriter
keys (viz. #, = and […]).

English version Italian original

1.B y’know when we’re starting # haha the 1.B sai quando iniziamo noi # haha il dieci
tenth of October ottobre
2.P of next year 2.P di un altr’anno
3.B of no of this year but 3.B di no di quest’anno ma
4.P. yeah but that’ll be 4.P si` ma passa tutto=
5.B [mm well] 5.B [be’ mm]
6.P almost next year 6.P l’anno praticamente
7.B [well yeah] 7.B. [be` si`]
8.B oh he’s got no time he says 8.B ah non c’e` tempo dice
9.P [in Oct]ober 9.P [a ot]tobre
10.B the beginning of October before the 10.B i prima di ottobre prima dei primi

Transcriptions prompt us to compare sense-making with reading. An analyst

can view text-in-interaction as if it resulted from a program.2 The approach
generates explanada such as sharing bad experience or recognizing complaints.
By focusing on ‘members’, it can seem that, in principle, interaction can be
understood by analysis. Disagreeing, Cowley (1993) shows the value of spec-
trographic display for investigating what people do. Close attention to voice
dynamics can be used in understanding the singularity of events. This is based
on linking routines with interindividual sensitivities (as described below).
Thus, on the basis of tens of listenings, he shows that, in 3, Brunella displays

2. The analyst claims to draw on shared knowledge to reconstruct the presuppositions that, in
context, give rise to a determinate illocutionary act.
Beyond symbols 

that she has heard Patrizia make a faux pas (in 2). In 4–6, Patrizia tries to rec-
tify this and, in 9, makes up for her error. How can this be explained? For a
context semanticist, the gaffe, recognition, and making-up can only depend on
social or mental programs. In what follows, distributed cognition is used to
sketch an alternative.

3. Complexity in social behaviour

Contextual semantics uses transcripts to bring our parallels between embodied

events and acts of reading. Conversation is made to resemble a product of social
knowledge and/or inner programs. In using ‘contextualization cues’ to save this
view, Gumperz stresses indexical events. Turning to inner mechanisms, he looks
beyond symbols without recognizing the power of vocal and visible dynamics.
Instead, he posits that presuppositions encode social experience. Challenging this,
mental life can be examined ‘out in the open’. Transcription traces full-bodied ac-
tivity which, in Lemke’s (2000) terms, exploits diverse space-time scales.
Unless humans actually depend on formal units, transcriptions do not record
linguistic output. Rather, they privilege the words actually spoken over the con-
tinuous fluctuations of prosodic and non-vocal activity. While moving towards
recognition of full-bodied interaction, Gumperz made non-symbolic dynamics
subordinate to symbols. Elsewhere, Cowley (2006) emphasizes that appeal to re-
current patterns sets up theoretical issues that include the following:
– Brains are not mere symbol processors
– Neither minds nor machines depend entirely on encoding/decoding
– Cognitive powers evolved to use physical features of the world
– Adaptive, flexible behaviour – and language– integrates activities that criss-
cross between body and world
Whatever their neural basis, conversations are part of social behaviour. Interaction
is situated and embodied because, as Harris (1995, 1998) sees, we integrate activi-
ties in real-time. As we do so, feeling, acting bodies orient to norms that evoke
thoughts. By using local dynamics, we can the current situation. This is possible
because, while having a routine aspect, utterance-activity arises as bodies concert
(Spurrett and Cowley, 2004) As we co-ordinate, we hear and sense what happens.
As a result, vocal and other movements “self-organize into patterned behaviour”
(Thibault, 2003b: 137). We use anticipative dynamics (see, Christensen, 2000;
Cowley, 2004) to influence what people feel, act and say.
 Stephen J. Cowley

4. Interaction is dynamical

Text-context theorists focus on pairings and sequences of utterance-sized units.

Given their methodological concerns, they tend to reduce sense-making to the use
of (putative) conversational units. On a dynamical view, this is like the armchair
philosopher’s error of describing physics exclusively around everyday middle-
sized dry objects. Instead of pursuing the lay-person’s perspective, interaction-
analysis focuses on how talk can be described. It thus obscures how singular events
contribute to human lives. This is inevitable as analysis relies on a metalanguage
that maps symbolic types onto repeatable contexts. Explanations of social life thus
model of how ‘minds’, not living beings, deal with language.3
In biology, living systems use dynamics in flexible adaptive behaviour. Where-
as simple systems may use separable time-frames (e.g. immediate stimuli and evo-
lutionary dispositions), animals exploit both proximate and ultimate causation
(Tinbergen, 1952; Hinde, 1987). As they become more complex, they use (or create)
multiply framed temporal organization (Lemke, 2000). While bacteria draw on
two time-frames, even invertebrates link evolution and development with learn-
ing-based experience. Animals rely on brains and bodies that evolved in using the
world as a cognitive resource (Clark, 1997). Bees, for example, learn about the col-
our of flowers (Cowley, 2004). When rats are in a maze, objects become landmarks
in future action. Robots, as Matarić (1997) shows, can also use landmarks to con-
trol movement. Finally, in Tetris, human experts use what Kirsh and Maglio (1994)
call epistemic action. In fitting pieces into a pattern, the screen serves as a cognitive
resource. Instead of imagining rotations, players carry out micro-movements
whereby ‘useless’ moves set them up for what comes next. They draw on skills,
experience and the setting to transcend current circumstances or, in other terms,
off-load information in ways likely to serve mid-term goals. Similar strategies are
required in performing long division and assembling jig-saw puzzles where, in-
stead of thinking, strategic action couples with real-time perception. Building on
Blair and Cowley (2003), talk is traced to real-time events. As Clark (1997) sug-
gests, humans make the world smart so that we can act dumb in peace.
How can speaking persons serve each other as cognitive resources? While
analysis can track use of regular patterns, it cannot show how acoustic events affect
feeling, thinking and acting. To understand the sense we ascribe to talk, contex-
tual semantics can be supplemented with pico-scale investigation (Cowley, 2007c;

3. Recognising that lay views are essential to language, it is often erroneously thought that
theory must be based on these views. With Dennett (1987), this is like arguing that since behav-
iour can be described from an intentional stance, it must be explained intentionally. In fact, he
argues the contrary: accounts from physical and design stances are necessary even if, ultimately,
they must specify how we come to believe in the abstracta posited from the intentional stance.
Beyond symbols 

2009a). In such models the sequential dimension of conversation is treated as re-

sulting from how events are integrated across time-scales. What is routine
(or situated) is thus connected with trans-situational or situation-transcending as-
pects. Utterance-activity can be scrutinised from afar, verbally, and with respect to
resonating bodies. Before pursuing this, a dynamical system is used to show how,
without thinking, self-organized systems generate complexity.

5. The enslavement principle and the dynamics of gas lasers

Physicists are impressed when systems show design that gives rise to ‘higher-level’
organization. Such events depend, not on push-pull causality, but on how systems
self-organize. When circular causation (Haken, 1993) applies, coupling imple-
ments the ‘enslavement principle’. This occurs, for example, in gas lasers. In simple
cases, these consist in a sealed tube containing a light emitting gas. To precipitate a
laser effect, an electric current of a certain voltage is passed through the tube. The
light remains disorganized until, at a certain threshold, gas molecules self-organize
as a laser-beam. Strikingly, this appears instantaneous. Indeed, while possible to
calculate the threshold at which circular causation arises, the underlying micro-
processes rely on neither the chemistry of the gas nor the electric charge. Thanks
to the current, high-level interaction between molecules prompts competition and,
remarkably, this leads to self-organization. The enslavement principle ensures that,
even in principle, outcomes are unpredictable. This is typical of self-organizing
systems which depend on principles more complex than rhythmic entrainment.
Next, therefore, I use the model in thinking about the real-time coupling of human
voices. While finding similar patterns of attraction, it is stressed that people moni-
tor the results to co-ordinate in ways that give talk much of its singularity.

6. Enslavement and interaction: Parallels and contrasts

The enslavement principle presupposes a physicist’s perspective. Interaction can

be shown to feature similar processes when one views cultural circumstances as a
locus for self-organizing. The necessary condition for enslavement is thus affec-
tive-cognitive activity between living human subjects. Like an electric current, the
flow of pico-scale events prompts each body to set off self-organizing activity that
evokes kinds of resonance. Movements, and especially vocalizations, resemble gas
molecules in that minute fluctuations of speaking (and moving) are interpenetrat-
ed by what another person says and does. True to the enslavement principle,
change arises when, acting jointly, voices and bodies attract each other. By analogy,
 Stephen J. Cowley

extrinsic motivation generates higher-level organization. Each party orients to sa-

lient effects that arise as speaking is co-ordinated. By analogy to the laser-beam,
orienting to a joint outcome can set off change in the quality of vocal events
People anticipate and, by so doing, prompt response. Not only does other-
orientation shape dynamic coupling but, of course, it elicits verbal response.
As in Gumperz’s (1982) classic case, ‘gravy’ can sound neutral or unfriendly
(Cowley, 2006; 2009). In an instant, vocal and visible dynamics produce laser-like
effects. Word-forms are inseparable from co-ordinated pico-scale activity. Con-
versing is thus both dynamic and symbolic. Drawing on something like the en-
slavement principle, sound-patterns can resonate with memories. Unlike gas-la-
sers, however, we can monitor the results and, by so doing, act to modulate each
other’s actions. Thus, living subjects use resonances to integrate speaking with
emergent, repressed and modified verbal activity. As sounding, moving bodies, we
link how things feel, experience and ‘what is actually said’. Talk yokes verbal pat-
terns to dynamics based in, above all, emotional affiliation, jarring, and conflict.
Since aspects of talk are laser like, salient patterns can spark adroit response or,
indeed, discomfort. As Gumperz saw, communicational effects arise from a flow of
events whose dynamics are, at once, affective, cultural and prosodic.
To pursue the laser analogy, I return to events that intrigued me a decade ago.
During the events shown in transcription, Patrizia and Brunella show exquisite
sensitivity to voice dynamics. In pursuing this, my primary concern is with how
prosodic patterns shape co-action. They rely on how the words are spoken in en-
acting a singular event that serves in maintaining their friendship.

7. Patrizia and Brunella’s dynamics

A transcript can show how interpersonal relationships play out in time. It hints at
how, in the pico-scale, brains jointly control coupled microdynamics. In the talk
described, Brunella is seeking empathy from Patrizia. This is done because, as
close friends, she is likely to understand the difficulties with getting builders. Such
empathy, however, does not depend on social routine. In these circumstances, in
fact, she “inadvertently indicates that she is not paying close attention to what she
is hearing” (Cowley, 1993: 146). This is a social gaffe: it is unacceptable to misin-
terpret ‘the 10th of October’ (il dieci ottobre). Since Brunella show signs (sic) of
hearing her misconduct, her friend tries to cover up. Seconds later (in 9), Patrizia
reiterates Brunella’s ‘in October’ (a ottobre). By doing so appropriately, Patrizia
‘makes good’. For convenience, key utterances are printed in italics.
Beyond symbols 

English version Italian original

1.B y’know when we’re starting # haha the 1.B sai quando iniziamo noi # haha il dieci
tenth of October ottobre
2.P of next year 2.P di un altr’anno
3.B of no of this year but 3.B di no di quest’anno ma
4.P. yeah but that’ll be 4.P si` ma passa tutto=
5.B [mm well] 5.B [be’ mm]
6.P almost next year 6.P l’anno praticamente
7.B [well yeah] 7.B. [be` si`]
8.B oh he’s got no time he says 8.B ah non c’e` tempo dice
9.P [in Oct]ober 9.P [a ot]tobre
10.B the beginning of October before the 10.B i primi di ottobre prima dei primi

Before turning to acoustic evidence, the transcript can be used to justify this as-
cription. It needs to be demonstrated, first, that saying di un altr’anno (of next
year) is a faux pas. Second, it needs to be shown that social equilibrium is restored
by Patrizia’s a ottobre (in October). This matters because, often, saying ‘of next year’
would be a sensible response to ‘ha ha the tenth of October’. Accordingly, I con-
sider why, in these circumstances, it inappropriately ‘goes beyond the situation’. It
is a social gaffe which, as things turn out, is put right by a blatant interruption (9)
that restores social equilibrium.
Spoken on a fall-rise tone di un altr’anno (of next year) looks unexceptional. It
might be said, for example, at the end of the year or, indeed, where normal to wait
five months for builders.4 This talk, however, occurs in May and Patrizia knows the
local builders. She thus fails to show adequate understanding or, indeed, to give
her friend support. The problem, then, is not one of inference. Rather, saying ‘in
October next year’ is not licensed by Brunella’s (unmarked) tone. Just as one would
expect in Italian and other languages (including English) Brunella’s fall-rise an-
ticipates an empathetic echo. However, rather than repeating what her friend has
said, she offers:

4. Any such interpretation would be false. First for biographical reasons (Patrizia has heard
talk about building in the village – a major topic of conversation – all her life.) Second, even if
she forgets it is May, her wording shows that she sees the new year as far away (she says not
‘dell’anno prossimo’ but ‘di un altr’anno’). Third, if the mistake were based on such ‘reasons’, her
gaffe would not be social.
 Stephen J. Cowley

Di un’ altr’ a n n o

— ✓
Patrizia speaks as if five months would be a short wait; she replies as if Brunella’s
tone suggested a longer delay. While transcription shows no outward sign of hurt
(‘no di quest’anno ma’), she hears the faux pas. In terms of the next utterance proof
of CA, Brunella’s ‘no of this year but’ displays that she expected something different
(especially, the final ‘but’). Then, Patrizia responds by offering a lame explanation
of her error. Waiting from May to October is, she implies, ‘like’ waiting 17 months
(in 4 and 6). This cover up is meant as a socially acceptable way of justifying sloppy
response. As such, it shows that Patrizia is monitoring the events. With ‘yes, but
practically the whole year passes’, however, she gets into more trouble. She implies
not only that the wait must feel long but, by so doing, shows that she knows it is
about a year away. By neither covering up her ignorance nor making a joke, she
shows that she ‘got the wrong end of the stick’.5 It is this, indeed, which is lame. In
treating the wait as if it were more than 3 times as long, she violates social norms.
She breaks the rule that, with friends, you ‘listen’. Thus, in 2, Patrizia is guilty of
lack of due attention.6 While such analysis clarifies what happens, it gives no in-
sight into how effects are achieved. Another mode of investigation is needed if we
are to understand how, in a few hundred milliseconds, a gaffe becomes salient,
prompts a lame utterance, and changes the flow of the talk. First, however, I sketch
how the disruption is overcome and equilibrium restored.
When Brunella is 4 syllables into her story, Patrizia interrupts with ‘a ottobre?’
How can use of known information be “verbally and prosodically right” (Cowley,
1993: 147)? Why does Patrizia use a fall-rise to repeat the fact and thus make up
for her previous gaffe?

a╅ o t╇ t o b╇ r╇ e

First, in the lived present, three seconds is not long enough for reasoning based on
the words that are actually spoken. The parties depend on intricate voice dynamics

5. Since it is May the period is actually 17 months. Neither party notices this (and, in 1993,
nor did I- in spite of having listened to the tape tens of times).
6. During conversations, people who do listen (in a normative sense) break the social rules.
Patrizia establishes that, in this sense, she is not guilty (She only misinterpreted what was said).
Beyond symbols 

that are much too rapid to rely on presuppositions. Not only are these beyond sym-
bols but, as argued elsewhere (Cowley, 2006), the vocalizations act directly. While
humans can be compared to turn-taking, inference-using symbol processors, we
are also self-conscious affective beings. We orient to each other as we integrate
bodily dynamics with manifest beliefs. Social life happens as we draw on affec-
tively charged ‘words’: we use pico-scale sound patterns:
– Persons use real-time sounding as they engage in co-action.
– Each picks up on subtleties in the other’s utterance-activity.
– Co-action is not simply entrained; persons are sensitive to affect and motives
as well as pico-scale events.
To capture the function of real-time dynamics, acoustic events can be repeatedly
frozen (see Cowley, 1993, 1994, 1998; in press). Pico-scale investigation shows the
delicacy with which, together, we link action and perception. Sound patterns shape
what, given repeated listening, an observer hears ‘in’ the words actually spoken. As
dynamic resources, vocalizations can regulate co-temporal activity. Events consist,
in part, of utterance-activity like that of 8–9 where Patrizia and Brunella couple their
talk. Not only is this a laser-like effect but its audible result makes up for the gaffe.
The women talk within a narrow pitch-range, mesh their timing, and echo the pitch
cadence of concurrent syllables (tempo and a ottobre). Prosodic closeness enables
the women to use pitch-matching which, while often noted (e.g. Brazil et al., 1980;
Gumperz, 1982), has many crucial functions (Cowley, 1998). Below, this is illustrated
with measures of fundamental frequency (on voiced segments) shown iconically.
(Fundamental frequency is in Hz and bold segments represent overlap).7

8. B╛╛ah non╛╛c’e` tem po╛╛dice

225 — — 210
— 207
190 —

9. P a╇ o t╅ t o b╇ r e

╇╇ 202 190 — 210
185 160

The observations show parallels between one person’s perceiving and the other’s
acting. As if based on circular causation, the voice dynamics show:

7. All measures were made on a Kay DSP Sona-graph 5500. In relation to fundamental fre-
quency, attention was given to vowels (and some voiced consonants) and, especially local pitch
peaks, troughs and moments of onset and offset. Given the settings of the machine, measures are
likely to be accurate within a range of approximately 4.5 Hz. For details, see Cowley (1993; 1998).
 Stephen J. Cowley

– Convergence in the pitch range of the voices;

– By timing ‘a ot’(tobre) to coincide with Brunella’s (te)‘mpo’, Patrizia allows
‘dice’ to be spoken during the silence of a geminate (o)tt(obre);
– Brunella responds, auditorily, by saying ‘dice’ in little more than a whisper;
– The pitch cadence on ‘a ot’ mirrors the rising cadence of ‘tem’ (Patrizia’s rises
over 0.6 ST, her friend’s falls over the same distance);
– The initial F0 measure on Patrizia’s pre-head (202 Hz) is within 5Hz of the end
point of Brunella’s tonic (207Hz);
– The final pitch of Patrizia’s tail is within 5 Hz of the initial measure on Brunel-
la’s rise (190Hz).
Pico-scale coupling is characteristic of intimate talk (Cowley, 1994; 1995; 1998;
2009). In conversations in Italian families, its complexity is comparable with that
of verbal patterns (Cowley, 1993). While challenging sequential models of interac-
tion (Cowley, 1998), it is striking that such events are often described colloquially.
While on the edge of awareness (Cowley, 2007c), this laser-like coupling is salient.
Thus, Patrizia speaks as if the information has ‘just sunk in’. Did we not hear the
synergy, we would not understand the ‘sinking’ metaphor.8 Rather as bodies reso-
nate, judgement uses dynamics. As if detecting a laser-beam, the women use pico-
scale fluctuations to give life to their speech.
The closeness of the coupling of the second pair of utterances is usefully con-
trasted with the event of three seconds earlier. In her faux pas too, Patrizia showed
empathetic surprise. On that occasion, however, her speech did not mesh closely
with Brunella’s. Rather, it had ‘a hollow ring’.

1. B. il die ci╇ ot╇ t o╇ bre 2. P. di un’ altr’ a n n o


270 302 —

✓ 320 205
— —
215 â•… —
— 280 220 — —

These utterances show contrast in both pitch-range and other musical features.
This arises as, launching into her story Brunella emanates tension while using fal-
setto to make demands of empathetic display. With respect to 9, Patrizia shows a
pitch range (+10 Hz) far from Brunella’s expansion (+80 Hz). Equally, the ‘hollow
ring’ depends on looser timing and lack of overlap. In contrast with events three
seconds later, Patrizia is unresponsive to her friend’s voice. The events, however,
are neither neutral speech nor, indeed, ‘noise’. Far from being rule-governed or
prosodically determined, the gaffe is – in one sense – a reasonable response. In this

8. In parallel to Wittgenstein’s (1958) remarks on colour, the phrase makes no use of inner samples.
Beyond symbols 

informal talk, Brunella’s voice dynamics prompt a mistaken ‘conclusion’. Misinter-

preting its literal sense, she picks up on the exaggerated display. Technically, how-
ever, this is a social gaffe where a party does not ‘listen’.
Patrizia’s speech of 9 synergizes Brunella’s by showing that, in one sense, she
was listening. In real-time, a socially significant pattern emerges as the friends
engage affectively. While aspects of speaking are laser-like, living subjects can also
regulate co-action. While partly entrained, each prompts the other to modulate
how they act and feel. Within relationships, appropriate emotions, attitudes and
motives animate utterance-activity:
– Variations in sounding help persons achieve specifiable goals
– Unexpected activity can dam the behavioural flow
Patrizia is less laser-like when (in 4, 6 and 9) she senses something is wrong. As a
person, she attempts to make up for her gaffe by using verbal response. Hearing 2 as
striking a false note thus contributes to the coupling of 8 with 9. Indeed, Brunella’s
perturbance shows in how she lowers her tone, drops into Patrizia’s pitch range and
makes it easy for her friend to ‘make up’.9 Adjusting to this, Patrizia integrates proso-
dy with syllabic and verbal patterns. Relying on sensitivity to pico-scale events, both
parties prod and probe by bring out empathetic and conflictual patterns. Brunella is
thrown off track by how the sound is integrated with the wording ‘di un altr’anno’.
Although showing no disturbance, the unexpected ‘content’ slows the flow of Brunel-
la’s speech in ‘polite’ underturns of 5 and 7. Not only are these uncharacteristic of
intimate talk but, strikingly, they seem to be a common spur to attentive listening.
Like gas lasers, human voices use mutual attraction to produce vocal synergies
while coupling in pico time scales.10 Since such events can occur in +/- 200msec,
they must depend on action-perception cycles (see Preston & de Waal, 2002). Yet,
outcomes are not ‘triggered’: the parties re-establish relationships by using affective
experience. Elsewhere, Cowley (2007c; 2009) suggests that its basis is biosemiotic.
We hear laser-like effects that arise as voices to act as mutual attractors. Not only
does this strike observers, but it shapes interactional experience. Humans feel how
others orient to pitch levels, cadences, rhythmicality, rapidity, tempo and voice-
quality (see Cowley, 1993). We monitor how members of relationships or groups
weigh our words. Humans use sounding and moving (e.g. Auer, 1992; Cowley,
2004) to gauge ‘what this is like’ and, thus, form attitudes and person-impressions.

9. In 1993, I failed to notice that, in compensating for the gaffe, Brunella adjusts to make
things easier for Patrizia. While in line with circular causation presented, this challenges my
earlier Wittgenstein-inspired emphasis on reacting-responding bodies.
10. It takes about 200msec. to speak a stressed syllable. The argument parallels Kirsh and
Maglio’s (1994) demonstration that Tetris playing is better served by epistemic action than
(inner) mental life.
 Stephen J. Cowley

As confirmed by both the example and experience, moreover, this can influence
what is said. Especially if talk is spontaneous, vocal coupling prompts feelings and
‘ideas’. Such mechanisms are most powerful, of course, where people know each
other well and are unconcerned with specific tasks or ideas. While talk is accom-
plished by co-ordinating bodies it is also experienced by living human subjects.

8. Rethinking text-context relations

Emphasising the biomechanics of interaction is consistent with what Hutchins

(1995) calls (culturally) distributed cognition. While some cognitive and communi-
cative events depend on output from ‘internal’ processes, much depends on causal
loops between bodies and the world. As in other species (Cowley, 1997), human
meaning-making is dynamic and embodied: real-time vocalizing is integrated
with non-vocal expression. More uniquely, humans co-ordinate against a back-
ground of salient events that share cultural and normative worlds. Accordingly,
indexical activity can function without (verbal) interpretation. Even if, at times,
participants sense how joint activity is heard, much is beyond conscious control.
Events are embodied even if, with different degrees of skill, analysts also ascribe
sense to how utterances (and their wordings) are heard to sound. Next, therefore,
I contrast this view of dynamics with models of semantic context.
Until the 1990s, brains were often seen as symbol-processing systems. This view
faded with parallel-processing (Rumelhart et al. 1986; Clark, 1997), discovery that
brains meet the needs of bodies (Edelman, 1992; Deacon, 1997; Johnston, 2004) and
recognition that human cognition is encultured (Hutchins, 1995). In parallel, intel-
ligent action was examined independently of representations (Brooks, 1991; Matarić,
1997) and traced to perception/action loops. In support, functionalist thinking was
enriched by use of Vygotsky (1962; 1978), Wittgenstein (1958), Merleau-Ponty
(1968), Gibson (1967; 1979), Maturana (1978) and others. As shown in Dennett
(1987; 1991; 1995), Hutchins (1995) and Clark (1997), cognition draws on causal
chains that link bodies and world. Whereas we exploit ‘cognition in the wild’, repre-
sentational models describe “a sociocultural system from which the human actor
has been removed” (Hutchins, 1995: 63). This echoes in, say, integrational linguis-
tics (Harris, 1981; 1995; 1998), Harré and Gillet’s (1992) discursive psychology, and
Bickhard’s (1995) rejection of cognitivism. More recently, a growing group has come
to emphasise the distributed nature of human language (see, Cowley, 2007; Cowley,
2009). On this view, real-time activity resembles, not output from von Neumann
machines, but how we use an environment to co-ordinate with robots. Given how
investigation clarifies intelligent activity, dynamical approaches (e.g. van Gelder,
1998) can be used to challenge the view that brains are von Neumannesque structures
Beyond symbols 

(Dennett, 1991) that use symbolic representations (Clark, 1997). Program-like

models explain little unless they map onto neurobehavioural processes. Going be-
yond appeal to entrainment, for example, Christensen and Hooker (2000) show that
bumble-bees, leopards and humans use brains in anticipative modelling. Expecta-
tion-based systems influence utterance-activity in mother-infant interactions (Cow-
ley, 2004). Far from relying on ‘language use’, conversing uses space, time and bod-
ies. For Hodges (2007), these anticipatory dynamics realize values.
On externalist models, by contrast, the focus typically falls on sequences that
allow interaction to be modeled in terms of rules. Thus, it is often supposed that
just as text constitutes context, context shapes text (e.g. Goodwin & Duranti, 1992).
On the distributed view, while using cultural constructs (e.g. texts and social
norms), language is grounded in full-bodied activity. From this perspective, Gestalt
analogies are limited. Words can act as foreground as ‘indexicals’ reconfigure as we
experience pico-scale events. While shaping some inferences, decision-making
often uses salient dynamics. The Gumperz model invokes presuppositions rather
than how bodies shape interaction. In talk, action-perception cycles integrate in-
formation across sensory modalities. To explain this, Cartesian models are replaced
by appeal to dynamic coupling. Vocal patterning reduces neither what is meant nor
computed but is, above and beyond this, a matter of using speaking bodies as cog-
nitive resources. Just as in other animals, human interaction is based in full-bodied
coupling. In Linell’s (2009) terms, talk uses other-orientation to enact events while,
at the same time, permitting generalizations that transcend the situation.

9. The distributed view

Much can be meant without being internally processed. Acts, thoughts and feel-
ings can be correctly ascribed by making judgements about circumstances. In il-
lustration, consider an example from Hutchins:
One evening a marine commander on board the Palau, Major Rock, telephoned
the Charthouse. Quartermaster Smith answered the phone. Major Rock asked
Smith what phase the moon would be that night. Smith asked Chief Richards who
was sitting nearby. Richards immediately replied, “Gibbous waning”. Smith re-
layed the answer to Rock. Rock apparently did not understand the answer and he
and Smith talked past each other for several conversational turns. Finally, Smith
put his hand over the mouthpiece and said, “Chief, he says it’s got to be one of,
‘new’, ‘first’, ‘full’, and ‘last’.” Chief Richards said, “It’s last.” Smith told Rock and
Rock hung up…. After Smith hung up the phone, Chief Richards said: “Rock is a
great big guy with a brain about this big (making a circle with the tip of his index
finger touching the first joint of his thumb). He must never have taken an amphib
 Stephen J. Cowley

mission onto a beach at night. He might get by with a crescent moon, but on a
gibbous moon he’ll be dead. (Hutchins, 1995; 230–231).

Events depend neither on what Richards (or Smith) meant by ‘gibbous waning’ nor
how Rock inwardly understood ‘it’s last’. Rather, they show Rock’s poor use of the
words actually spoken. Given limited vocabulary and naval experience he fails to in-
tegrate what he needs to know with hearing ‘gibbous waning’. As shown by the gaffe
described, human interaction often features such failings. Yet, as Richards remarks,
familiarity with the moon can, at times, make the difference between killing and being
killed. As we co-act we use publicly available routines to link with past experience
while anticipating what may happen. Though interaction is situated, we use experi-
ence in projecting possible outcomes. While analysis highlights the words that are
actually spoken, much depends on pico-scale events. Thus Patrizia and Brunella inte-
grate these with verbal patterns to show exquisite sensitivity to each other’s feelings. As
they do so, they shape each other’s thinking and the flow of events. In historical time,
we have developed a meshwork of constructs, artifacts and modes of action (Love,
2004). People hear linguistic forms in the flow of talk. In utterance-activity, of course,
we also draw on ‘words’.11 Patrizia and Brunella connect what they hear, see, say and
feel with current expectations. Like the naval officers, the women monitor available
information by attending to more than is actually said. By hearing her gaffe, Patrizia
uses the sound of her speaking as a prompt to make good. In (9), she synergises af-
fectively mediated dynamics with Brunella’s story telling. Just as for the officers, events
originating in one brain influence the other person. We move beyond the current situ-
ation. While this may be reducible to information (in Shannon’s probabilistic sense),
we also draw on text-based inferences, affect and past experience. On a distributed
view, much ‘thinking’ is co-action elicited by barely noted sound patterns.

10. Contextualizing and the enslavement principle

Gumperz might applaud a situation-transcending view of interaction and see the

above as examples of contextualizing.12 Stricto sensu, however, Patrizia and Brunel-
la use patterns that emerge and spread. While metaphorically, these are ‘cues’, they
need no shared meanings, presuppositions, or inner intentions. From a distributed

11. In Love’s (2004) terms, conversing is first-order language that can be heard in terms of
second-order cultural constructs (including ‘words’).
12. In an influential definition this is: “speakers’ and listeners’ use of verbal and nonverbal signs
to relate what is said at any one time and in any one place to knowledge acquired through past
experience, in order to retrieve the presuppositions they must rely on to maintain conversa-
tional involvement and assess what is intended” (Gumperz, 1992: 231).
Beyond symbols 

perspective, vocal synergies prompt lay judgements. Brain-side events are insepa-
rable from affect and, connecting people in time, set off social events. While some
expression has an internal basis, many ‘intentional’ properties are extrinsically de-
rived. In hearing “gibbous waning” anything that Richards and Smith ‘mean’ mat-
ters less than Rock’s failure to grasp the expression. In similar vein, instead of ago-
nizing over the ‘meaning’ of the dynamics of 2 and 9, what matters is how the talk
is integrated. There is a disjunction between human sense-making and what inter-
action analysis sdescribe. This is because, while analysis demands a shared meta-
language, human experience is body-based. As a result, we use dynamics in mak-
ing judgements and ascriptions of meaning. Indeed, once we consider how
pico-scale events self-organize, we can rethink how Brunella and Patrizia syner-
gise affect, social practice and their relationship. Contextualization uses utterance-
activity that orients to both this now and what, on average, is likely to bring future
benefit (see, Cowley, 2004). Conversing is flexible behaviour that occurs alongside
explicit intentions and inferences. During utterance-activity, moreover, contextu-
alizing can often make direct use of what is felt, done, and thought.
Given trust in phonetic intuitions and nose for social meaning, Gumperz re-
jected text-context duality. He avoided reading transcriptions as evidence of
program-like units and, therefore, over-inflating symbols. Using an ear trained to
structural analysis, he identified interactional constituents by means of careful lis-
tening. In line with classic cognitive science, he assumed that these functioned
‘between the ears’. As a cognitive internalist, he treated the cues as ‘real’ and as-
similated them to sui generis models of language. Emphasizing the primacy of in-
teraction, he separated languages from the speech of determinate communities. By
extension, sense-making is activity and, for Gumperz, it is mistaken to think that
“meaning resides in language” (2003:121). While rejecting his view of cues, the
distributed view uses this insight by tracing meaning-making to prosodic syner-
gies. Talk is co-constructed as feelings, presuppositions and inferences spread.
Pico-scale events also permit complex ascriptions. Given this post hoc quality, we
can use voice dynamics to predict and, indeed, shape events. Expanding on work
like that of Malinowski (1927), Abercrombie (1967) and Bolinger (1986), Gumperz
shows that language-in-time cannot be reduced to folk constructs. As complex
social behaviour, utterance-activity is based in human biomechanics.13
Since contextualizing integrates activity between persons, there is no a priori
limit on its functions. Above all, while synergetics shape interindividual aspects of

13. Malinowski (1927) saw that primitive language must be grounded in action. Abercrombie
(1967) recognised that what is now called ‘prosody’ must inevitably deal with what he called
‘voice dynamics’: Bolinger (1986) saw that the forms and cues of prosodic analysis are no more
that cultivated versions of ‘wild’ ways of using the voice.
 Stephen J. Cowley

talk, interaction also uses symbolic and probabilistic processes. The distributed
view thus rejects appeal to a priori language ‘systems’. Rather, meaning-making is
traced to human capacities for co-ordination. Continuous, embodied utterance-
activity prompts us to re-member experience as circular causation sets off judge-
ments. Given the slow pace of experience, we can (partly) transcend our own bio-
mechanics. While mistaken to trace interaction to the units of lay analysis – words,
utterances and intentions– talking demands belief in such fictions. In develop-
mental time, contextualizing must prompt a baby to believe in rules and units
(see, Cowley et al. 2004; Spurrett & Cowley, 2004). In learning to converse, sym-
bols must be grounded into brains, normative activity and first-person experience.
As children come to hear words, they take a language stance (Cowley, 2007a; in
press): they begin to hear vocalizations as verbal patterns. Saussure’s error, there-
fore, lay in reifying a mere ‘point of view’. By emphasizing forms, he reduced lan-
guage to models of system and use. For many linguists, therefore, people are agents
who link texts to contexts (and vice versa). By contrast, on a distributed view,
people are inseparable from language: human agency arises as bodies act together
in a cultural world.

11. Interaction & language revisited

To treat interaction as an ‘analytical prime’ is to reduce persons to ‘users’ of word-

based input-output. As Thibault (2003a) suggests, social formations should thus
be seen as neither logically nor ontogenetically prior. While interaction plainly
predates human ontogenesis, the vicissitudes of human activity depend on talking
bodies. Developing this argument, I claim that interaction and its (putative) com-
ponents – words, minds, intentional states etc.– are valuable folk constructs. They
rely on social norms favored, amongst other things, by a tendency to believe in
their reality. Human bodies use the self-organizing powers of brains to discover
biomechanical ways of orienting to social abstracta.
Challenging the primacy of interaction, ‘construing experience’ is partly
embodied. This, of course, is consistent with Gumperz’s view that ‘meaning’
pertains to persons and, for this reason, little explanatory power accrues to
theories that posit a priori systems. Equally, while transcription has descrip-
tive value, analysis cannot ‘explain’ human sense-making. In the investigation
reported here, Patrizia and Brunella show that conversing draws on how we
perceive vocal modulation. In development, of course, such contextualizing is
primary. Even if interaction is historically prior, experience and laser-like syn-
ergies influence talk. As the friends show, self-organizing bodies use affect
that, moments later, shapes what happens. Ontogenetically, such co-ordination
Beyond symbols 

grounds language (Cowley et al. 2004; Cowley, 2007a). Indeed, without per-
ceiving and monitoring affect, there can be no attention sharing and, as a con-
sequence, babies would lack the perceptual skills used to hear utterances in a
verbal aspect. In adults, reciprocal monitoring allows talk to proceed without
need for either Thibault’s (2000) downward causation or Gumperz’s (2003)
contextual semantics. Internal and external motive systems co-develop as we
orient to a shared cognitive world. Without thinking, Patrizia commits and
makes up for a social gaffe while, without words, Brunella accepts an unspo-
ken apology.
In rejecting the primacy of social formations, weight falls on human reso-
nance. It is only by emphasizing the causal loops of affective expression that we
can move beyond post-Saussurian symbols. In tracing pico-scale coupling, real-
time events are shown to be constituted by dynamic information that sets off
effects. During interaction, the epistemic functions of affective activity are suf-
ficient to prompt (explicit) ‘thoughts’. In lay terms, attitudes spark ‘what to say
next’ and, for an analyst, suggest construals of events. More technically, we rely
on full-bodied events that animate what we do together. Rather than view inter-
action in terms of formal units, we turn to how, using a pico-scale, bodies use a
world of social norms to display manifest beliefs. Experience with bodily
prompts and probes gives us neuro-behavioural resources based in affect, ex-
pectations in human relationships. While we can describe contextualization
‘cues’ (and conventional gestures), social events are also directly embodied. Giv-
en shared history, affective patterns often render redundant any appeal to cues
and conventions.
Presenting lay analysis as invoking abstracta runs against post-Saussurian
emphasis on linguistic (or semiotic) autonomy. Instead of ascribing independent
reality to language, interaction is traced to real-time dynamics. Far from being
based in word-forms, language bottoms out in pico-scale co-ordination. While
material processes (Thibault, 2003b) are synergetic, active subjects manage
events in larger scales. Monitoring word-forms and ‘talking without thinking’
shape relationships. What the friends show is important but limited. First, some-
thing like the enslavement principle allows verbal functions to be integrated with
affect. Second, an analyst can use third person ascriptions based on how the
sound of voices is integrated with wordings. However, such processes are mar-
ginal in erudite talk. Patrizia and Brunella show only that talk can proceed with-
out inner thoughts, presuppositions, and inferences. In close relationships,
movements and vocalizations often ‘direct’ activity. Drawing on a shared past,
events are assessed and managed under dual control. In humans too, much de-
pends on bodily resonance. As social vertebrates (Cowley, 1997), we often con-
textualize without cues.
 Stephen J. Cowley


Abercrombie, David. 1967. Elements of General Phonetics. Edinburgh: Edinburgh University Press.
Auer, Peter and di Luzio. Aldo (eds.). 1994. The Contextualization of Language. Amsterdam:
John Benjamins.
Austin, John. 1962. How to Do Things with Words. Oxford: Clarendon.
Bickhard, Mark H., 1995. “Intrinsic constraints on language: grammar and hermeneutics”. Jour-
nal of Pragmatics, 23, 541–554.
Blair, Grant and Cowley, Stephen J. 2003. “Language in iterating activity: Microcognition re-
membered”. Alternation, 10(1): 132–162.
Bolinger, Dwight. 1986. Intonation and its Parts: Melody in Spoken English. Stanford University
Press: Stanford, CA.
Brazil, David, Coulthard, Malcolm and Johns, Tim. 1982. Discourse Intonation and Language
Teaching. London: Longmans.
Brooks, Rodney A. 1991. “Intelligence without representation”. Artificial Intelligence. 47: 139–159.
Christensen, Wayne D. and Hooker, Cliff. 2000. “An interactivist-constructivist approach to in-
telligence: self-directed anticipative learning”. Philosophical Psychology, 13(1): 5–45.
Clark, Andy. 1997. Being there: Putting Brain, Body and World Together Again. MIT Press:
Cambridge, MA.
Condon, William S. 1979. An analysis of behavioural organization. In Nonverbal Communica-
tion: Readings with Commentary, Shirley Weitz, (ed.) 149–167. New York: Oxford University
Press. (Originally published, 1976, Sign Language Studies, 13: 285–316).
Cowley, Stephen J. 1993. The Place of Prosody in Italian Conversations. Unpublished PhD dis-
sertation, University of Cambridge.
Cowley, Stephen J. 1994. “The role of rhythm in conversations: A behavioural perspective”. Lan-
guage and Communication, 14: 353–376.
Cowley, Stephen J. 1997. “Conversation, co-ordination and vertebrate communication”. Semiot-
ica, 115 (1): 27–52.
Cowley, Stephen J. 1998. “Of turn-taking, timing and conversations”. Journal of Psycholinguistic
Research, 27 (5): 541–571.
Cowley, Stephen J. 2004. “Contextualizing bodies: Human infants and distributed cognition”.
Language Sciences, 26 (6): 565–591
Cowley, Stephen J. 2006. “Language and biosemiosis: A necessary unity?” Semiotica, 162: 417–444.
Cowley, Stephen J. 2007a. “How human infants deal with symbol grounding”. Interaction Stud-
ies, 8(1): 81–104.
Cowley, Stephen J. 2007b. Distributed language: biomechanics, functions and the origins of talk.
In The Emergence and Evolution of Linguistic Communication, Caroline Lyon, Chrystopher
Nehaniv, and Angelo Cangelosi (eds.) 105–127. London: Springer Verlag.
Cowley, Stephen J. 2007c. The codes of language: turtles all the way up? In The Codes of Life,
Marcello Barbieri (ed.) 319–345. Berlin: Springer.
Cowley, Stephen J. 2007d. Cognitive dynamics and distributed language. Language Sciences, 29
(5): 575–583.
Cowley, Stephen J. 2009a. “Language flow: Opening the subject”. Cognitive Semiotics, 4: 64–92..
Cowley, Stephen J. 2009b. “Distributed language and dynamics”. Pragmatics & Cognition, 17/3:
Cowley, Stephen J. In press. “Taking a language stance”. To appear, Ecological Psychology, 2010.
Beyond symbols 

Cowley, Stephen J. and Spurrett, David. 2003. “Putting apes, (body and language) together
again”. Language Sciences, 25: 289–318.
Cowley, Stephen .J. Moodley, Sheshni and Fiori-Cowley, Agnese, 2004. “Grounding signs of
culture: primary intersubjectivity in social semiosis”. Mind, Culture and Activity, 11/2:
Deacon, Terrence. 1997. The Symbolic Species: The Co-evolution of Language and the Human
Brain. Harmondsworth: Penguin.
Dennett Daniel. 1987. The Intentional Stance. Cambridge MA: MIT Press.
Dennett, Daniel. 1991a. Consciousness Explained. Boston: Little Brown and Co.
Dennett, Daniel. 1991b. “Real patterns”. Journal of Philosophy, 88, 27â•‚51.
Dennett, Daniel. 1995. Darwin’s Dangerous Idea. Harmondsworth: Penguin.
Duranti, Allesandro and Goodwin, Chris (eds.). 1992. Rethinking Context: Language as an Inter-
active Phenomenon. Cambridge: Cambridge University Press.
Edelman, Gerald M. 1992. Bright Air, Brilliant Fire: On the Matter of the Mind. Harmondsworth:
Eerdmans, Susan, Prevignano, Carlo and Thibault, Paul (eds.). 2003. Language and Interaction.
Amsterdam: John Benjamins.
Gibson, James J. 1966. The Senses Considered as Perceptual Systems. Boston MA: Houghton
Gibson, James J. 1979. The Ecological Approach to Visual Perception. Boston MA: Houghton
Giles, Howard, Coupland, Justine and Coupland, Nikolas (eds.). 1991. Contexts of Accommoda-
tion: Developments in Applied Sociolinguistics. Cambridge: Cambridge University Press.
Goodwin, Charles and Duranti, Allesandro. 1992. Rethinking context: an introduction. In Re-
thinking Context: Language as an Interactive Phenomenon, Allesandro. Duranti, A. and
Charles Goodwin, (eds.) 1–42. Cambridge: Cambridge University Press.
Grice, Paul. 1987. Studies in the Way of Words. Cambridge MA: Harvard University Press.
Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.
Gumperz, John J. 1992. Contextualization and understanding. In Rethinking Context: Language
as an Interactive Phenomenon, Allesandro Duranti, and Chris Goodwin, (eds.) 229–252.
Cambridge: Cambridge University Press.
Gumperz, John J. 2003. Response essay. In Language and Interaction, Susan Eerdmans, Carlo
Prevignano, and Paul Thibault, (eds.) 105–126. Amsterdam: John Benjamins.
Haken, Hermann. 1993. Are synergistic systems (including brains) machines? In The Machine
as Metaphor and Tool. Hermann Haken, Anders Karlqvist and Uno Svevin (eds.) 123–138.
London: Springer-Verlag.
Harré, Rom and Gillett, Grant. 1994. The Discursive Mind. Los Angeles: Sage.
Harris, Roy. 1981. The Language Myth. London: Duckworth.
Harris, Roy. 1995. Language, Signs and Communication. London: Routledge.
Harris, Roy. 1998. An Introduction to Integrational Linguistics. Oxford: Pergamon.
Hatfield, Elaine, John T. Cacioppo, and Richard L. Rapson (eds.) 1994. Emotional Contagion.
Cambridge: Cambridge University Press.
Hinde, Robert. 1987. Individuals, Relationships, Culture. Cambridge: Cambridge Unversity Press.
Hodges, Bert. 2007. “Good prospects: Ecological and social perspectives on conforming, creat-
ing and caring in conversation”. Language Sciences, 29 (5): 584–604.
Hutchins, Edwin. 1995. Cognition in the Wild. Cambridge MA: MIT Press.
Johnston, Mark. 2004. Developmental Cognitive Neuroscience. Oxford: Blackwell.
 Stephen J. Cowley

Kirsh, David and Maglio, Paul. 1994. “On distinguishing epistemic from pragmatic action”. Cog-
nitive Science 18, 513–549.
Lemke, Jay. 2000. “Across the scales of time: Artifacts, activities, and meanings in ecosocial sys-
tems”. Mind, Culture and Activity 7 (4), 273–290.
Levinson, Stephen. 2003. Contextualizing ‘contextualization cues’. In Language and Interaction,
Susan Eerdmans, Carlo Prevignano, and Paul Thibault, (eds.) 31–40. Amsterdam: John
Linell, Per. 1979. Psychological Reality in Phonology. A Theoretical Study. Cambridge: Cambridge
University Press.
Linell, Per. 2009. Rethinking Language, Mind and World Dialogically. Charlotte NC: Information
Age Publishing.
Love, Nigel. 2004. “Cognition and the language myth”. Language Sciences 26, 525–544.
Love, Nigel 2007. “Is language a digital code?” Language Sciences, 29 (5), 690–709.
Malinowski, Bronisław. 1927. The problem of meaning in primitive languages. In The Meaning
of Meaning, Charles K. Ogden and Ivor A. Richards (eds.) 296–236. London: Kegan Paul,
Tench, Trubner and Co Ltd.
Matarić, Maja. 1997. “Behavior-based control: Examples from navigation, learning and group
behaviour”. Journal of Experimental and Theoretical Artificial Intelligence, 9(2–3): 323–336.
Matthews, Peter H. 1979. Generative Grammar and Linguistic Competence. London: Routledge.
Maturana, Humberto R. 1978. Biology of language: The epistemology of reality. In Psychology
and Biology of Language and Thought: Essays in Honor of Eric Lenneberg, George A. Miller,
and Elizabeth Lenneberg. (eds.), 27–63. New York: Academic Press.
Merlau-Ponty, Maurice. 1995. The Phenomenology of Perception. London: Routledge. Originally
published (1945) as Phénoménologie de la Perception, Maurice Merleau-Ponty. Paris:
Preston, Stephanie and de Waal, Frans. 2002. Empathy and its proximal bases. Behavioral and
Brain Sciences 25: 1–72.
Prevignano, Carlo and di Luzio, Aldo. 2003. A discussion with John Gumperz. In Language and
Interaction Susan Eerdmans, Carlo Prevignano, and Paul Thibault (eds.) 7–30. Amsterdam:
John Benjamins.
Rumelhart, David E. Smolensky, Paul, McClelland, James L. and Hinton, Geoffrey E. 1986. Sche-
mata and sequential thought processes in PDP models. In Parallel distributed processing:
explorations in the microstructure of cognition, Vol. 2. Psychological and biological models,
David E. Rumelhart, and James L. McClelland (eds.) 7–57. Cambridge MA: MIT Press.
Ryle, Gilbert. 1949. The Concept of Mind. Chicago: University of Chicago Press.
Saussure, Ferdinand de. 1916. Cours de Linguistique Générale. [English translation by R. Harris.
London: Duckworth, 1983.]
Searle, John. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge
University Press.
Spurrett, David and Cowley, Stephen J. 2004. “How to do things without words”. Language Sci-
ences, 26(5): 443–466.
Thibault, Paul J. 2000. “The dialogical integration of the brain in social semiosis: Edelman and
the case for downward causation”. Mind, Culture and Activity, 7(4): 291–311.
Thibault, Paul J. 2003. Contextualization and social meaning-making practices. In Language and
Interaction, Susan Eerdmans, Carlo Prevignano, and Paul Thibault (eds.) 41–62. Amsterdam:
John Benjamins.
Beyond symbols 

Thibault, Paul J. 2003. Body dynamics, social meaning-making, and scale heterogeneity: re-
considering contextualization cues as mixed-mode semiosis. In Language and Interaction,
Susan Eerdmans, Carlo Prevignano, and Paul Thibault (eds.) 127–147. Amsterdam: John
Tinbergen, Niko. 1952. “Derived activities: Their causation, biological significance, origin and
emancipation during evolution”. Quarterly Review of Biology, 27: 1–32.
Trevarthen, Colwyn. 1998. The concept and foundations of infant intersubjectivity. In Intersub-
jective Communication in Early Ontogeny, Stein Bråten. (ed.) 15–46. Cambridge: Cambridge
University Press.
van Gelder, Timothy J. 1998. “The dynamical hypothesis in cognitive science”. Behavioral and
Brain Sciences, 21: 1–14.
Vygotsky, Lev S. 1978. Mind in Society: The Development of the Higher Psychological Processes.
(Edited by Michael Cole, Vera John-Steiner, Sylvia Scribner and Ellen Souberman). Cam-
bridge MA: Harvard University Press.
Vygotsky, Lev S. 1986. Thought and language. (Alex Kozulin, Trans. and Ed.). MIT Press:
Cambridge MA.
Wheeler, Mike. 2004. “Is language the ultimate artifact?” Language Sciences, 26 (6): 693–715.
Wittgenstein, Ludwig W. 1958. Philosophical Investigations (2nd edition). Oxford: Blackwell.
The case for an eclectic approach
to discourse-in-interaction

Catherine Kerbrat-Orecchioni
Université Lumière Lyon€2

This paper advocates an eclectic approach to discourse-in-interaction analysis,

not only because adopting a single point of view on such a complex object is
too restrictive, but also because it is impossible to account for fundamental
aspects of the ways it operates without having recourse to notions coming from
different theoretical paradigms. For this we shall consider first the question
of units (particularly speech acts and adjacency pairs) then the question of
“preference organization”, a notion which can be dealt with more adequately by
resorting to face-work considerations. This investigation will lead us to revisit
two problems which are central to discourse analysis (whether in interaction or
not): what place is to be allocated to context in description and what the analyst’s
interpretation consists in.

1. Introduction

In the wake of Levinson (1983:€286–294), Discourse Analysis and Conversation

Analysis are usually opposed, due to their different “styles of analysis”. However, if
we consider that any discipline is defined essentially by its object of investigation
rather than by its particular approach, then we have to admit that analyzing con-
versations is part of discourse analysis, since conversations are a specific kind of
discourse. It follows that Discourse-in-Interaction Analysis would appear to be the
best way of labelling the field of research in which I am engaged,1 which consists of
examining different aspects of various kinds of discourse produced in an interac-
tive context, that is co-produced by different participants involved at the same time
in a linguistic exchange. “Interactional linguistics” or “verbal interaction analysis”

1. See Le discours en interaction (2005), where the different principles exposed here are devel-
oped and illustrated.
 Catherine Kerbrat-Orecchioni

would also be suitable terms here as they refer to the study of different types of
interactions (reciprocal action systems) performed by mainly linguistic means.2
The boundaries of this field are fuzzy, for discourses can display different de-
grees of interactivity: an informal conversation is clearly more interactive than a
lecture, a television debate more interactive than the news, etc. (certain types of
oral discourse are therefore barely interactive, whereas certain types of written
discourse such as Internet Relay Chats are, to some extent, interactive).
This kind of investigation requires natural data, which are highly complex. In
order to account for this complexity in a satisfactory way, it seems preferable to put
together one’s own comprehensive toolbox. Mine includes, besides classical lin-
guistic tools, borrowings from different trends in “discourse analysis” (especially
the so-called Birmingham and Geneva schools), pragmatics (Ducrot, Grice, speech
act theory), and, of course, interactional linguistics (conversation analysis – hence-
forth CA – and also Gumperz, Goffman, Brown and Levinson’s linguistic polite-
ness theory, etc.). In other words, this kind of approach is founded on methodo-
logical eclecticism (that is the controlled use of tools coming from different
paradigms). As this term sometimes carries a pejorative connotation, it is worth
noting that it has been employed in a positive way by authors as different as Aston
(1988:€13); Wetherel (1998:€388); Vicher and Sankoff (1989), who speak of “meth-
odological hybridization”; Eggins and Slade (1997:€273); House (2000:€146), who
advocates “an eclectic model comprehensive and powerful enough to handle di-
verse cases of misunderstandings”; Gumperz (see Eerdmans, Prevignano and
Thibault 2003:€32, 50, 71); or Heritage (1995:€397):
Many CA insights and observations are profoundly compatible with the view-
points developed in connection with, for example, Gricean implicature or polite-
ness theory.

The purpose of this article is to present the case for an eclectic approach towards
discourse-in-interaction. I will begin by reviewing some general principles on
which such an approach is based.

2. Principles

Looking at my materials, these long collections of talk, and trying to get an abstract
rule that would generate, not the particular things that are said, but let’s say the
sequences […]. (Sacks 1992, Vol.€I: 49; emphasis added)

2. “Interaction” refers first and foremost to a process (that is a series of actions and reactions)
and secondly, by metonymy, to the very event in which this process occurs.
The case for an eclectic approach to discourse-in-interaction 

The way of working outlined here by Sacks is not unfamiliar to linguists: the con-
stant toing and froing between the observation of materials and the search for
underlying abstract rules. However, what is new compared to the usual practices
of linguists (especially considering the time when Sacks delivers his first lectures)
is the nature of these materials and the relationship to data that he recommends.

2.1 The data

Any generalization should be the result of scrupulous and detailed analysis of “ac-
tual episodes of interactions of one sort or another”. The data must be plentiful
(“long collections of talk”) and for the most part, naturally occurring.3

2.2 The rules

The goal of analysis is to go beyond the description of particular occurrences in

order to extract regularities and to discover reproducible phenomena (such is the
condition of any scientific undertaking). The very first of Sacks’ Lectures is called
“Rules of conversational sequences”, and the most famous article in conversational
literature is entitled “A simplest systematics for the organization of turn-taking in
conversation” (Sacks, Schegloff and Jefferson 1974). The constant use of the terms
“rules” and “regularities”, “procedures”, “methods”, “norms”, “conventions”, “rou-
tines”, “organizing principles”, “order/orderly”, etc., appears to be a real leitmotiv in
this literature. Indeed, the objective is primarily to do away with the idea that con-
versations are chaotic (and therefore unsuitable for scientific investigation) and to
prove the existence of some kind of order in the midst of this chaos, this “apparent
disorderliness of natural speech” (Goodwin 1981:€55). Contrary to popular belief,
ordinary talk is “systematically and strongly organized”. Consequently the analyst’s
task is to bring to light this kind of organization which is both strong and flexible
(“structured sets of alternative courses or directions which the talk and the inter-
action can take”, Schegloff 1986:€114). Moreover, conversational rules can be con-
sidered as procedures, as they are directed towards practical use. Lastly, most con-
versational rules deal with sequential organization: order prevails in interactions,
in every sense of the word.

3. Sacks does use made-up examples from time to time, however; and for some phenomena,
it can be interesting to take advantage of literary examples.
 Catherine Kerbrat-Orecchioni

2.3 The question of units

We need some rules of sequencing, and then some objects that will be handled by
the rules of sequencing. (Sacks 1992, Vol. I: 95).

We will now look at these objects which are handled by the rules of sequencing, that
is “the embarrassing question of units” (Goffman 1981:€22) – they are in fact very
diverse in nature, including strictly linguistic units (phonemes, words, different types
of grammatical units) but also pragmatic, discursive and conversational units.
On a “superficial” level, conversations and other types of verbal interactions
present themselves as a succession of “turns” made up of “turn-constructional
units”. We will not go back over these units here as they have been deeply investi-
gated in conversation analysis, but rather prefer to make some comments about
two types of units which, in our opinion, are of a totally different nature and be-
long to another level of analysis, namely adjacency pairs and speech acts.

2.4 Adjacency pairs

The adjacency pair (henceforth AP) is widely accepted as being the prototypical
smallest unit of sequential organization, yet it poses many theoretical and descrip-
tive problems. First, the minimal units are not always “in pairs”, nor “adjacent”, as
we can see considering the most common examples dealt with in the literature:
greetings only come in pairs in two-party conversations;4 questions often initiate a
ternary sequence as the answer should generally be followed by an acknowledg-
ment; and invitations frequently open an extended sequence. Moreover, these
types of sequences often contain more complex configurations (insertion sequenc-
es, embedded and side-sequences, interlocking organizations, etc.), and this great-
ly limits the adjacency property, despite the “preference for contiguity” principle
(Sacks 1987). All these structural particularities lead us to the conclusion that ad-
jacency pairs are only one specific type of inclusive units, generally labelled “se-
quences” in CA where the meaning of this term is never made fully explicit.
Another problem with APs is knowing what they are made up of. Sacks talks
about “utterances”, but “turn” is also often employed in the literature. However, it
should be made clear that pair parts are not turns, primarily because their bound-
aries do not always coincide: pair parts are simply “housed” inside turns (“you
have a turn and in it a first pair part”, Sacks 1987:€56). Above all, turns are units
which “pertain to the surface structure of conversation” (Roulet 1992:€92), where-
as pair parts are units of a pragmatic nature: they correspond to “actions” that

4. Incidentally, the importance given to this notion of “pair” is an index of the primacy given
to a dyadic conception of talk-in-interaction (Kerbrat-Orecchioni 2004a).
The case for an eclectic approach to discourse-in-interaction 

“epiphenomena” such as turns serve to convey (Selting 2000:€511). APs are made
up of speech acts, or more precisely of segments revolving around a “head act”
which may be accompanied by one or more “subordinate acts”. These monological
units which form minimal dialogical units are called moves in discourse analysis
(or by Goffman) and the dialogical units they form are labelled exchanges or inter-
changes (the term is ambiguous, but its technical definition is relatively clear: any
set of moves depending on one and the same initiative move). These kinds of or-
ganization have been well described for example by Sinclair and Coulthard or the
Geneva school, whose objective is to show how a conversation is built by combin-
ing units of different hierarchical “ranks”.5
They raise a certain number of issues (where one exchange begins and another
one ends, the occurrence of “Janus elements” which are oriented both towards the
previous and the next move, etc.), but the important thing is to recognize that
moves and exchanges pertain to another level of analysis than the turn.6 Take this
example of an interaction in a bakery as a brief illustration of the sequential or-
ganization of these pragmatic units:
1 B madame bonjour/
2 Cl bonjour (.) je voudrais une baguette s’il vous plaît
3 B bien cuite [ou
4 Cl [bien cuite oui
5 B (tendant la baguette) alors voilà (.) et avec ça/
6 Cl ça sera tout merci […]
1 B good morning madam/
2 C good morning (.) I’d like a baguette please
3 B well-done [or
4 C [yes well-done
5 B (giving the baguette) here you are then (.) anything else/
6 C that’s all thanks […]
The excerpt is made up of six turns, five exchanges and ten moves. The first exchange
stretches over T1 and T2: it is an exchange of greetings (“good morning/good morn-
ing”) which is primarily built on lexical material. However, T1 also works as a ques-
tion, due to the rising intonation which means “What would you like?” (two speech
acts are amalgamated here in one segment because of its multimodality). This
initiative move opens up a second exchange which is interwoven with the first:

5. See Kerbrat-Orecchioni 1990: chap.€4 on the “hierarchical functional” model.

6. As Schegloff reminds us too (1992a: 124): “Although it is true that the organization of turn-
taking and the organization of sequences (or speech acts) are not independant […], they are
largely distinct and only partially intersecting”.
 Catherine Kerbrat-Orecchioni

question-answer exchange, with the answer coming in T2, after the reactive greet-
ing. As for T2: on the one hand, the assertion “I’d like a baguette please” assumes a
request value, by virtue of the illocutionary derivation rule “any assertion of some
need, addressed to someone who can satisfy this need, must be interpreted conven-
tionally as an indirect request to satisfy this need” –€it should be noted that some
contextual elements should be incorporated into the formulation of pragmatic rules:
in the commercial context, the assertion is addressed to somebody who not only can
but must satisfy the customer’s need (unless the requested product is not available),
the indirect speech act is all the more obvious here. On the other hand, this “request-
ing answer” is not clear enough for the baker (problem of applying the Gricean
“maxim of quantity”), who therefore asks the customer for more detail by introduc-
ing in T3 an embedded question-answer type exchange. Once B has obtained the
necessary detail, she can go ahead and fulfill the request, as evidenced in T5 where
the reactive move is made up of a head act which is, in fact, a non-verbal action (ac-
companied by an utterance to be considered as a subordinate act: “here you are
The end of T5 initiates a new question-answer type exchange (the answer is
delivered in T6).
Hence, this sequence consists of the following exchanges:
(1) E1 (T1€-€T2’s first segment): greetings;
(2) E2 (T1€-€T2’s second segment): question-answer; amalgamation of E1 and
E2’s initiative moves, the reactive moves being distinct and delivered suc-
cessively in the same turn;
(3) E3 (T3€-€T4): question (request for details) followed by answer; E3 is em-
bedded in E4;
(4) E4 (T2’s second segment and T4€-€T5’s first segment): request-request ful-
filled; the request is grafted onto the answer and delivered in two parts
(“I’d like a baguette please” and “well-done”);
(5) E5 (T5’s second segment€-€T6): question-answer (the answer is accompa-
nied by a subordinate act of thanks).
Such a banal example as this shows the limits of the notion of “adjacency pair”: the
boundaries of exchanges coincide only exceptionally with the boundaries of turns,
insofar as one same turn can be made up of several moves which are either successive,
or amalgamated by virtue of indirect speech acts, and sometimes, of multimodality.

2.5 Speech acts

The trickiest problem with the notion of the adjacency pair is this: if two adjacent
turns can sometimes be considered as forming an AP, but at other times not, which
The case for an eclectic approach to discourse-in-interaction 

criteria should be used to decide whether or not one is dealing with an AP? It is based
on the feeling that both items have a “conditional relevance” relationship (“given the
first, the second is expectable”, Schegloff 1972:€363). But what is this feeling based on?
To answer this question we have to turn to the notion of the speech act.
Despite what is often said following a famous controversy,7 it is not obvious
that CA and speech act theory are incompatible. On the contrary even, CA con-
stantly refers to an implicit theory of speech acts with its persistent use of notions
like “question”, “request”, “greeting”, “offer”, “complaint”, etc., to which we can add
more interactive kinds of pragmatic units such as “challenging”, “repairing”, etc.
These units are usually referred to as “actions” in CA, but as they are very specific
actions realized by verbal8 means, it would seem that the expression “speech acts”
is more appropriate –€ the use of this term does not imply the adoption of all
Austin’s, Searle’s or Venderveken’s theoretical postulates: today the notion of speech
act has come into the public domain; it is part of the basic vocabulary for those
dealing with discourse, in interaction or not.
However we label this class of object, the questions, requests or greetings are
defined by the “job” they do (something very similar to Searle’s illocutionary force)
which dictates their sequential properties (if a question is generally followed by an
answer, surely it is because its purpose, by definition, is to obtain some sort of in-
formation, supposed to be unknown to the person asking the question). “Sequenc-
es” are not interpretable on the sole basis of their sequentiality (the fact that an
utterance immediately follows a question is neither necessary nor sufficient to say
that it is the answer to the question);9 and the application of conditional relevance
and sequential implicativeness principles rests above all on the content of the utter-
ances, which creates some specific expectations about the nature of what follows.
Without going back over the criticism of speech act theory, I would say that in
my opinion, it does not fundamentally challenge this theory. We could also show
that Schegloff ’s relevant analysis of “For whom” or “Do you know who’s going to
that meeting?” (1984 and 1988) does not in any way contradict standard speech
acts theory.10 But there is no doubt that work carried out in the framework of CA
has helped to make this theory more operative by examining how speech acts re-
ally function in an interactive context, the role of the utterance’s position in the
sequence for determining its pragmatic value, how this value can be negotiated by

7. See Searle et al.€(eds.) 1992.

8. On the fundamental difference between verbal and non verbal actions, see Kerbrat-Orec-
chioni 2004b.
9. As Mey says (1994:€241): “The mere fact of utterances following each other is no guarantee
of coherence”.
10. See for example Cooren 2004.
 Catherine Kerbrat-Orecchioni

the participants, and also the possible transformations a speech act can undergo
throughout the exchange; for example, how a greeting question such as “how
are you?” can progressively become a real question (Kerbrat-Orecchioni
2001:€118–120), or how an offer can be turned into an order, as in the “stubborn
old man and the herring” sequence examined by Sacks (1992, Vol. II: 327–331).
Sacks’ analysis brilliantly shows the successive transformations that the speech act
undergoes (somewhat like the music of Steve Reich, where the theme changes so
surreptitiously that the listener suddenly finds himself in the middle of a new mo-
tif with no idea as to when exactly the change took place); it shows also that this
metamorphosis is already potentially contained in the offer’s very first utterance.
This kind of analysis has greatly refined the description of speech acts, but it is
founded upon Austin and Searle’s theory.11

2.6 The sequential organization of interaction

2.6.1 Preference organization

According to the grammar of speech acts, an initiative act accomplished by A will
open up a paradigm of possible reactive acts to be accomplished by B. However, all
these reactions are not equally probable, some are “preferred” and others are “dis-
preferred”, not only in context, but already at the level of the system. This “prefer-
ence organization” is based firstly on considerations of frequency: the “preferred”
reaction is the reaction which is chosen “ordinarily” in a paradigm of alternative
forms, which also best conforms to the participants’ normative expectations and
which has correlatively fewer interactional effects. However, the observation that
the preferred reactions are generally expressed in a more immediate and econom-
ical fashion12 has led to formal considerations being given precedence over no-
tions of frequency. Unfortunately these two criteria do not always coincide, which
makes it possible, according to Lerner (1996), to find such strange things as dis-
preferred sequences within a preferred format or vice-versa. Similarly, when ana-
lyzing compliment responses Pomerantz notes that “actual performances are often
discrepant from ideal or preferred performances” and that “a large proportion of
compliment responses deviate from the model response of accepting compli-
ments” (1978:€ 80–81). However, one could also reach the conclusion that this
“ideal” and this “model” should be questioned and thought about differently, call-
ing on politeness theory (seen in terms of face-work). This theory, which is the

11. In other terms, I don’t think that “a gap between the theory [of SA] and the reality of dis-
course” exists (Streeck 1980:€133).
12. See for example the different articles in the book edited by Atkinson and Heritage,
1984: Part II.
The case for an eclectic approach to discourse-in-interaction 

result of a kind of hybridization between Goffman and Searle’s work (as it looks at
speech acts from the point of view of the effects they can have on participants’
faces) was drawn up by Brown and Levinson (1978 and 1987), and several modi-
fications of it have been suggested.13
Whichever version of this model we choose to adopt can render the descrip-
tion of discourse-in-interaction considerable service. In particular, it helps to
make the problem of preference organization clearer, as Lerner notes (1996:€304):
It has been widely observed […] that matters of face, on the one hand, and preference
organization in conversational interaction, on the other, are intimately connected.

If we accept that preference is “intimately connected to matters of face”, we can

understand why the preferred sequence is often the most costly one. As face-work
generally involves some overcost in utterance formulation (which is compensated
for by the benefit obtained regarding the interpersonal relationship), it is not sur-
prising that in many cases, a relatively costly solution is chosen, for instance after
a compliment (request for confirmation, downgraded agreement, softened disa-
greement, etc.), or following an offer: if a brutal refusal is “less preferred” than
immediate agreement, a certain preference for temporary refusal has also been
noted, which brings about a more or less developed cyclical exchange, made up of
repetitions and refusals of the offer before its final acceptance (Kerbrat-Orecchioni
2005:€220–226). For both offers and compliments, the elaborate nature of the se-
quence is due to the fact that they are particularly complex acts as far as face-work
is concerned. They comprise both a “threatening” and “flattering” component for
the recipient’s face, so that he finds himself in a “double bind” situation which can
only be resolved by relatively sophisticated strategies.
Pomerantz says that sequences are subject to multiple and sometimes contra-
dictory constraints. It should be added that these constraints do not all belong to
the same level of analysis. First, the reactive move should comply with the seman-
tic-pragmatic rules of coherence which depend on the nature of the initiative move
(as well as on the utterance’s specific content). Moreover, sequencing is subject to
additional types of constraints: these are ritualistic face-work constraints, which
may come into conflict with one another€– in the case of a compliment: the gen-
eral principle of preference for agreement (which is “flattering” for the face) ori-
ents towards a positive reaction, but the principles of protecting one’s own terri-
tory, and of self-praise avoidance (“modesty rule”), join together to favour a

13. Personally I have suggested introducing the notion of Face Flattering Act (FFA) into the
model, alongside the notion of “Face Threatening Act” (FTA) –€ in a similar way others talk
about “face enhancing acts” or “face boosting acts” –, as politeness consists not only of softening
up threatening acts (negative politeness) but also of producing valorizing acts (positive polite-
ness). See Kerbrat-Orecchioni 1992 and 2005€: chap.€3.
 Catherine Kerbrat-Orecchioni

negative reaction. It is a double bind situation, therefore, and the only exit is via
compromise, delay or avoidance strategies: the complexity of this exchange with
regard to the face-work system can explain why responses to compliments can
often be seen as excessively fussy and affected.14
On the contrary, if after most assertions agreement is preferred to disagree-
ment (that is the least expensive reaction rather than the most expensive one), it is
because agreement is more satisfactory for face: whatever the initiative act, the
preferred reaction (that is the most expected and frequent one) is not the least
expensive one but the one which best satisfies mutual face preservation.

2.7 Dialogical interpretation principle

The principle of preference organization operates on the side of the production of

discourse by participants. If we now look at interpretation mechanisms (which
concern both participants and analyst alike), we meet another principle which is
sometimes labelled the “dialogical interpretation principle” and can be summa-
rised like this: given that utterances are often pragmatically ambiguous (as speech
act markers are as polysemous as lexical items), how can we interpret them? The
answer: by observing the T2 reaction to the problematic turn T1, which will “pub-
licly demonstrate” how T1 has been understood, that is the real meaning of T1.
This idea has been largely developed in conversational literature,15 for example by
Heritage comparing these two adjacency pairs (1984:€255):
(1) A: Why don’t you come and see me sometimes
B: I’m sorry. I’ve been terribly tied up lately
(2) A: Why don’t you come and see me sometimes
B: I would like to
A’s utterance is pragmatically ambiguous: it could correspond to several different
“actions”. However, B’s sequence will give us the key to its meaning: in (1) it is a
complaint, as B responds with an excuse, and in (2) it is an invitation as B’s reac-
tion is to accept it.

14. Regarding the different possible reactions to a compliment, see Pomerantz 1978 for English
and Kerbrat-Orecchioni 1994: chap.€5 for French€; and on how to apply the “double bind” notion
to description of interactions, Kerbrat-Orecchioni 1992:€279–289 and 1994:€273–275.
15. This question is at the centre of the debate between Schegloff and Wetherell in Discourse &
Society, 8(2), 1997 and 9(3), 1998.
The case for an eclectic approach to discourse-in-interaction 

A first objection to this “sequential analysis” as it is dealt with in CA arises im-

mediately: perhaps B misunderstood A’s utterance. However, if that is the case, in
the third turn A will rectify B’s mistaken interpretation.16 This means that:
– A knows how B interpreted his turn T1 thanks to the reaction T2 provided by B;
– B knows if his interpretation is correct or not thanks to A’s response (third turn).
At the same time, the meaning of the whole exchange is made available to the ana-
lyst thanks to what is also called the “theorem of mutually recognizable actions”
(Heritage) or the “intersubjective construction of meaning principle” (Schegloff).
Unfortunately, things are not always so simple. Let us look now at Sacks’ well
known example (1992, Vol. I: 689):17
A: Can you fix this needle?
B: I’m busy.
A: I just wanted to know if you can fix it.
In the third turn A claims than the first one had to be interpreted as a mere ques-
tion and not as a request (as B did), but he is more likely committing here a kind
of denial of the indirect value of his previous utterance (for face saving reasons, the
request having failed he pretends he has not intended such a speech act). This ex-
ample shows that it is not always the case that the third turn “publicly demon-
strates” the real meaning to be accorded to the first turn.
We cannot say either that the second turn necessarily shows the meaning that
B accords to the first turn. Let us look again at the case of compliment responses:
One of the most frequent in French consists of treating the compliment as a ques-
tion about the nature or origin of the praised item (“I love your perfume – Kenzo”;
“These cups are great! – I got them in Brittany”, etc.), even in cases where there is
nothing to suggest that this treatment corresponds to the compliment payer’s in-
tention. Here, Marandin (1987:€86) maintains that the compliment “does not oc-
cur in the interlocution”, which is more than debatable: following our linguistic
conventions, the utterance “These cups are great” cannot not be considered ana-
lytically as a compliment, even if it is not treated as such, and the recipient prefers
to “avoid” it. Certainly, no one is fooled by the avoidance strategy: the author of the
compliment suspects that his compliment has been recognized (no need to repair
with something like “It was a compliment, not a question!”), and the recipient
suspects that the author suspects that his compliment has been recognized. This is
a sort of routinized misunderstanding, which is accepted easily by both participants
as it is most suitable to the mutual preservation of their faces.

16. On “third position repair”, see Schegloff 1987.

17. In this example A is supposed to be a very young child, but similar cases of denial could also
be found from adult speakers.
 Catherine Kerbrat-Orecchioni

So it is a little simplistic to assimilate “signification of X produced by A” to

“interpretation of X made by B”, and “interpretation of X by B” to “interactional
treatment of X by B”.
Besides, this treatment, that is B’s reaction, must itself be interpreted by the
analyst. This can be just as problematic, as reactive moves are often formally less
characteristic than initiative moves (compliments are easier to identify than com-
pliment responses, questions are easier to identify than answers, etc.). The follow-
ing turn may help the analyst with his interpretation, but it is unrealistic to expect
the participants to do the work for him:
Although the value of sequential analysis in getting at the participants’ under-
standings of the talk has been stressed in this section, it should not be concluded
that the way in which a speaker responds to a prior utterance can, in every case,
be treated as criterial in determining how the utterance should be viewed analyti-
cally. (Atkinson and Heritage 1984:€11)

2.8 The role of sequential placement

All linguists admit that the location of an item in the syntagmatic chain deter-
mines in part its meaning. However the relative importance accorded to the con-
ventional meaning (pre-existing) and to the location (occasional) varies according
to authors and theories. As for the role of sequential placement of an utterance for
determining its value, all depends on the type of utterance we are dealing with: if
it is true that we identify an “answer” because it comes after a question (and also
because of its semantic content),18 then we must accept that questions present
specific formal characteristics. Consider as well the often mentioned example of
greetings: Sacks (1992, Vol. I: 94) states that “we need to distinguish between a
‘greeting item’ and a ‘greeting place’”, and quite rightly adds that “if some other
item occurs in a greeting place it’s not a greeting”. More questionable is his asser-
tion that “if a greeting item occurs elsewhere it’s not a greeting”: a “Good morn-
ing!” occurring in the middle of a conversation will be a greeting as well (a “mis-
placed” greeting), which proves the robustness of the intrinsic meaning of most
linguistic items.
The effects of these misplacements can be very diverse: polemical (an example
of which we will see shortly), playful, or even pathological (as we find in the “ab-
surd” dialogues in Ionesco’s or Tardieu’s plays), or metaphorical as in the example

18. Cf. Levinson (1983:€193): “Answerhood is a complex property composed of sequential loca-
tion and topical coherence across two utterances amongst other things; significantly there is no
proposed illocutionary force of answering.”
The case for an eclectic approach to discourse-in-interaction 

of an exchange “between two adults after making love” quoted by Streeck

A: Hi! (smiling)
B: Hi! (smiling)
Although “misplaced” to some extent, “Hi!” retains the value of a greeting. Given that
this speech act is submitted, as Searle put it, to the preparatory condition “the speaker
must have just encountered the hearer”, the mere utterance of this greeting means that
the speakers are acting as though they are meeting again after being separated (far from
challenging the existence of this condition of felicity, Streeck’s example proves its exist-
ence, which is responsible for the effect of this little game between the two lovers).
The temptation to overestimate the role of placement can be easily understood:
it is because this property has an “objective” nature –€ whether Y follows X or
whether T2 comes after T1 is “obvious”, and this obviousness is the same for both
participating members in the interaction and for the analyst. Unfortunately, how-
ever, sequentiality does not provide the key to meaning... The problem is that the
term “sequentiality” is used to refer to two types of relationship of a very different
nature: firstly a simple relationship of order, which is effectively objective; but
sometimes also a logical-semantic relationship of “conditional relevance” (a ques-
tion requires an answer, and the answer comes to complete the question to form
with it an “adjacency pair”), and this relationship is of an interpretative nature.19
Placement is simply one of many resources used to interpret utterances. For
example, the actual meaning of “How are you?” and other “greeting questions”
depends on a series of factors which work simultaneously: the position of the ut-
terance in the exchange (the earlier it occurs, the more it acts like a greeting), the
degree of elaborateness of the wording, the accompanying prosody, face expres-
sion and gestures, the participants’ mutual knowledge about the current situation
of the addressee and other contextual factors.

2.9 The role of context

Context shapes utterances and utterances shape context in turn: this conception,
which reconciles determinism and constructionism, is widely accepted today. Dis-
course is both an activity which is conditioned by the context and which transforms
this very context: pre-existing to interaction, context is continuously renewed

19. The notions of “sequence”, “sequencing” and “sequentiality” are used in a number of theo-
retical frameworks. For an overview of the different approaches towards these notions from
both formal and functional perspectives, see Fetzer and Meierkord (eds.) 2002.
 Catherine Kerbrat-Orecchioni

throughout. The renewable nature of context is partly due to the fact that the term
refers not only to the “external” or “situational” context (this “frame” surrounding
the interaction, itself comprised of different components which can be classed in
several ways)20 but also to the “sequential” or “intra-interactional” context21
(Schegloff 1992b), which continuously reframes conversational activities (for ex-
ample, the introduction of a question creates “a new arena for subsequent action”,
Duranti and Goodwin 1992:€29). In addition, context should be considered as a set
of cognitive representations (shared at least partially by the different participants)
which are constantly enriched and modified – even if it is obvious that all is not
“renewable” in the context and that many components remain constant through-
out the interaction.
It is also widely agreed that context strongly determines the production and
interpretation mechanisms which work in an “indexical” or “setting-related” way.
Opinions diverge, however, regarding to what extent the analyst can and should
use information which is not actually contained in the interaction’s text.22 Gump-
erz (in Eerdmans, Prevignano and Thibault 2002:€22) claims that the analyst “al-
ways needs a prior analysis of context”, and should therefore collect as much “eth-
nographic” information as possible on the setting being studied. However, most
conversation analysts consider that we should do without “external” interpretative
resources, which may be “misleading” (Heritage 1984:€282): considering that con-
text is an infinite set, all the elements are not interactionally relevant and those
which are will be revealed as such by the participants. At any moment of the inter-
actional process, they can bring to light some specific aspect of context and make
it significant; so the relevant contextual facts are “internalized” to some extent in
the form of “indicators” which allow the analyst to do without external informa-
tion. In this light, context is entirely considered as
something endogenously generated within the talk of the participants and, in-
deed, as something created in and through that talk. (Heritage 1984:€283)

Let us note first of all that even when analysis wants to be purely “internal”, the
analyst always draws on some external information. Take the example of a small
shop: from what point is the mention of context legitimate? As soon as the shop-
keeper “exhibits” his status as shopkeeper, by saying something like “Yes madam?”
or “May I help you?” Yet the question cannot be correctly interpreted without

20. See among others Hymes’ (1972) SPEAKING model or Brown and Fraser’s (1979) model.
21. In textual linguistics, the internal surroundings of a given segment are sometimes called cotext.
22. On “controversy in CA about the proper attitude toward contextual knowledge”, see
Bilmes 1996:€184. Note that this debate reminds us somewhat of the raging textual linguistics
debate in the 70’s.
The case for an eclectic approach to discourse-in-interaction 

knowing that it is occurring in a small shop... Is it necessary to wait until the cus-
tomer says something like “I’d like a loaf of bread” to discover that the scene is
taking place in a bakery (which is an interpretation made possible by a lexical and
“encyclopedic” knowledge that cannot be considered as purely internal)? Such an
attitude of refusing external information can seem rather artificial – all the more
so as the constitution of a corpus, at least when dealing with a “collection” (interac-
tions in small shops, telephone calls, work meetings, etc.), entirely depends on
external criteria... Let us note too that the distinction between endogenous and
exogenous context is entirely dependent on the way the corpus is made up
(for example, for the analyst who only has an audio recording the characteristics
of the setting constitute external information, whereas in a video recording they
form entirely part of the semiotic material which makes up the interaction), and
the way it is delimitated (if we widen the corpus, what is “external” may become
“internal”). The analyst’s first gesture is to isolate a more or less long segment of
talk-in-interaction, but for the members this segment only takes on meaning with
regard to its environment, immediate or not. This initial gesture therefore consti-
tutes a “bias” for the analyst who cannot therefore claim to take on the members’
point of view, especially if he chooses to look no further than the boundaries set by
the sequence on which the analysis is focussed.
Whether in a bakery, or in another setting, participants know where they are
well before “showing” it. And this is the main objection we can make against this
descriptive attitude: it contradicts the principle by which description should be
made as far as possible from “the members’ point of view”; for when they enter a
shop or a classroom, or when they take part in a television show, the members al-
ready have an idea regarding what kind of event they are involved in, and this
mental picture will be called up constantly throughout the interaction.
If taking context into account can be “misleading”, refusal to do so can be even
worse. In a classroom, Heritage reminds us, we can have a non-didactic interac-
tion and the opposite. But it is not because the teacher and pupils (or the doctor
and patient) can from time to time in the classroom (or in the doctors’ office) pro-
duce an exchange which is apparently a conversation between peers that it will
really be one. For example, in the co-constructed narrative taking place in a school
analyzed by Mondada (1998:€252–3), the first ten turns give no explicit indication
as to the status of the teacher; this does not mean, however, that it is a true “peer
interaction”: the “members” are well aware of their respective status (which can be
here considered as “omnirelevant”, cf.€Sacks 1992:€594–596), and this knowledge
cannot help but influence their interpretation of conversational events. Moreover,
when these markers occur, they have to be interpreted, which can only be achieved
on the basis of prior conventionalization. I can understand that a speaker is “doing
being teacher” or “doing being doctor” thanks to the kind of questions he asks; but
 Catherine Kerbrat-Orecchioni

this understanding is based on my knowledge of some pre-existing correlations

between a type of interactional behavior and a type of status or role. Lastly, the
interactional effect produced by this “doing” depends entirely on whether or not
the speaker really possesses teacher or doctor status,23 or to mention Sacks’ origi-
nal example regarding this, the fact of “doing being ordinary” will not have the
same interactional meaning if the “doing” is being done by a genuinely “ordinary”
citizen or by the French President: the interactional meaning of any behavior is to
be found somewhere at the crossroads between being and doing being, that is, at
the interface between external and internal information.
So an affirmation such as “context is not predefined” (Mondada 1998:€ 258)
cannot be taken literally, it is useful above all for underlining that context is per-
manently recreated as the interaction goes along: the context is pre-existing to the
interaction and constructed throughout the interaction at the same time. The ana-
lytic attitude which consists of pretending to discover the main characteristics of
context only from some explicit markers occurring during the interaction is not
only artificial but often misleading. For instance, when dealing with media-type
interactions the audience needs absolutely to be taken into account, even though
it is generally not explicitly referred to. It seems preferable to begin with the most
relevant points of the context to which the members have access (“frame” of the
interaction or its “schema”: type of setting, participation framework, goal of the
exchange, etc.). Then we must see how these elements are activated in the dis-
course and how they may be changed and negotiated among the participants.24 In
other words, we have to reconcile two ways of dealing with discursive data, la-
belled top-down and bottom-up by Aston (1988:€26):
The schema provides initial presuppositions and expectations, but through the
discursive process its instantiation may be modified and renegotiated on a bot-
tom-up basis.

3. Analyzing an example: Good evening Mr Le Pen

In order to illustrate the fact that when describing an interactional sequence, we

have to resort to tools from various origins, I will take the example of this small

23. See Garfinkel for “breaching” experiments, and Goffman for “breaking frame” and “fraud-
ulent identity”. We can also think of the fascinating “phoney” character, defined as someone
whose “doings” are not in harmony with his “being”.
24. The notion of “conversational negotiation” appears best suited to deal with the reciprocal
action between what exists before and what emerges during the interaction; on this notion see
for example Roulet 1992 and Kerbrat-Orecchioni 2005: Chap.€2.
The case for an eclectic approach to discourse-in-interaction 

excerpt from a television debate which took place in 2003 between Nicolas Sarkozy
(who was Interior Minister at the time) and Jean-Marie Le Pen (President of the
National Front party)25, and which will give me the chance to go back over the case
of greetings.
Sarkozy has already been on the show for a while and has been confronted
with several members of the political arena. Le Pen makes his entrance, gives a
general greeting, and makes himself comfortable as invited to do so by the host of
the show Olivier Mazerolle (OM):
1 OM: monsieur Sarkozy alors euh Jean-Marie Le Pen président du Front
National est avec nous euh vous allez débattre ensemble bonsoir mon-
sieur Le Pen
2 LP: bonsoir/
3 OM: voilà (.) prenez place (.) monsieur Le Pen […]
1 OM: so Mr Sarkozy euh Jean-Marie Le Pen the president of the National
Front party is here with us euh and you are going to talk together good
evening Mr Le Pen
2 LP: good evening/
3 OM: right then (.) take a seat (.) Mr Le Pen […]
No sooner is he sitting down than he launches into a diatribe against the political-
media world for treating him as an “outcast”. Sarkozy lets him get on with his act
for more than a minute. At the very moment when Le Pen after this general pre-
amble turns towards Sarkozy, getting ready to attack, here is what happens:
4 LP: […] ASP26 monsieu:r/ le ministre de l’Intérieu:r/ vous me donnez
5 NS: [bonsoir/] monsieur Le Pen
6 LP: bonsoir/ bonsoir monsieu::r eh j’ai dit bonsoir en arrivant/ ASP mais
euh vous étiez inclus collectiv- dans mon bonsoir collectif\
4 LP: […] ASP Mr Minister/ you seem to me::/
5 NS: [good evening/] Mr Le Pen
6 LP: good evening/ good evening si::r eh I did say good evening when I arrived/
ASP but uh you were included collectiv- in my collective greeting\

25. This is the program 100 minutes pour convaincre [100 minutes to convince], France 2,
20th Nov. 2003.
26. ASP signals an audible inspiration.
 Catherine Kerbrat-Orecchioni

I will describe the greeting good evening Mr Le Pen, which appears identically
twice (turns 1OM and 5NS), but with considerably different values due to the dif-
ference in sequential placement in the following way:
As soon as we speak of a “greeting” it is necessary to turn to speech acts theory,
since a greeting is a specific speech act, where the speaker “courteously indicates
that he has recognized the listener”: such is the illocutionary value of so-called
“greeting” utterances, according to Searle (1972:€107)27 who adds that this type of
utterance is subject to the preliminary condition that “speaker and listener have just
met”. This implies that the greeting should normally come at the very beginning of
the interaction: this is in fact the case for the first occurrence of the greeting. This
occurrence is also perfectly orthodox as far as sequencing is concerned: one greet-
ing calls for another one in return (the exchange is “symmetrical”). Let us note,
however, in the reactive move (or if preferred, in the second part of the adjacency
pair), the absence of all term of address and the fact that it is collective (Le Pen ut-
ters this greeting as he walks towards his seat looking at no one in particular).
The greeting delivered by Sarkozy (in 5) is more unusual: on the one hand it
brutally interrupts Le Pen’s speech, on the other hand it arises when Sarkozy and
Le Pen have been engaged in conversation for a while, it can therefore be consid-
ered as “misplaced” – but in part only for as Le Pen had until this point not been
addressing Sarkozy exclusively, we can accept that “Mr Interior Minister” marks
the beginning of a new interaction, embedded in the previous one (a “dilogue”
finds itself embedded in a “polylogue”).
The question is whether in a case like this a new exchange of greetings is nec-
essary. Nothing is less sure: our ritual system is uncertain here, and the greeting is
far from expected (in any case for Le Pen, whose norms apparently diverge from
Sarkozy’s on this point, the greeting is totally unexpected). Nevertheless, due to its
very specific placement, Sarkozy’s “good evening”, without ceasing to be a greeting,
also serves as an indirect act of reproach. This value is the result of implicit reason-
ing like this: in engaging in an exchange with me you should have greeted me first,
but you did not, so you lack good manners. It is also reinforced by intonation
(markedly more rising than in turn 1OM), and also by the expression of triumph
used by Sarkozy to welcome Le Pen’s reactive “good evening” (nodding his head up
and down with a little smile), which is a sort of retroactive clue as to the indirect
meaning of the reproach.
Sarkozy’s utterance therefore receives a double illocutionary force: the force of
a greeting conventionally attached to the “good evening”, and the force of a re-
proach which emerges in this particular context. It calls for a double reaction,

27. In fact, the “essential condition” of the greeting is rather that “the speaker indicates to the
addressee that he has recognized him and/or intends to engage in a verbal exchange with him”.
The case for an eclectic approach to discourse-in-interaction 

which indeed comes: obliged to return the greeting (which he repeats not without
some annoyance), Le Pen also feels that he must justify his behavior (“I did say
good evening when I arrived but you were included in my collective greeting”: his
reaction to the reproach). Other interactional values (“perlocutionary” for speech
acts theory) come along on top of these two illocutionary values, for example: The
unexpected irruption of the greeting will upset the exchange and unsettle the op-
ponent, as we can see in 7LP –€cut off in mid sentence, Le Pen is clearly thrown
off balance, for at the end of his turn he is victim of a failure followed by a repair
(“you were included in my collectiv- in my collective greeting”), which is not like
him. In addition, this greeting will invalidate what has come before: since a greet-
ing should normally appear at the very beginning of the exchange, what came be-
fore will to some extent become null and void; in this way Sarkozy suggests that
the general preamble should not have taken place, and that Le Pen should have
addressed him personally right from the beginning (Sarkozy’s reproach also con-
cerns this point).
In order to describe what happens at this moment of the interaction we can
also call on face-work theory: A greeting is in principle an act of “courtesy” (see
Searle’s definition above), in that it constitutes a “face flattering€act” (FFA), which
is obviously the case with the host’s “good evening Mr Le Pen” (the politeness of
the greeting is reinforced here by the term of address). But the FFA is seriously
harmed by the way in which the greeting is used by Sarkozy afterwards, as a “face
threatening act” (FTA) interfering with the greeting, namely the reproach aggra-
vated by the interruption: Le Pen is severely reprimanded by Sarkozy and is obliged
to justify his behavior like a naughty child (he is clearly placed in a low position by
the reproach which is addressed to him). In such an example the threatening com-
ponent is stronger than the flattering component and even cancels it out (here we
have a sample of this “courteous delegitimization” which has been identified as one
of Sarkozy’s favorite debating strategies).
Finally we could call on Goffman for his work on “presentation of self ” and
correlative construction of self-image and of that of others throughout the interac-
tion: in this way Sarkozy, killing two birds with one stone, uses the greeting to
show that he is a polite debater (even if this politeness is rather suspect here), and
to label his opponent as having an unmannerly ethos. We can call upon Goffman
again for the notions he has introduced in order to describe the “participation
framework” –€in this example the setting is complex as we are dealing with a media
interaction, the exchange taking place on the TV set is in fact a show intended for
another party: the audience, who for the most part, we can imagine are overjoyed
at the trick Sarkozy has played on Le Pen, who has the reputation of being a very
skilled debater.
 Catherine Kerbrat-Orecchioni

Obviously when necessary other tools can be called upon depending on the
needs of the description. But this short example allows us to get some idea as to
what an eclectic, and correlatively syncretic approach can mean, as it draws on de-
scriptive resources taken from diverse fields of analysis and puts them together.
This eclecticism is justified by the fact that discourse, and more especially dis-
course-in-interaction, functions on different levels which are both distinct and
articulated (management of turns-at-talk, semantic-pragmatic coherence of the
dialogue, global as well as local organization, face-work and ritualistic constraints),
one same item working simultaneously on several levels, for example:
– In “Can you close the door please?”, please acts as both a marker of illocution-
ary derivation (confirming the question’s indirect request value which is al-
ready suggested by the conventionalization of the structure and by the extral-
inguistic context), and as a softener of the face threatening act that the request
– In “I’d just like a baguette”, the adverb just firstly serves to prospectively organ-
ize the interaction by projecting that the transaction will merely consist of one
purchase and that the shop assistant will have no need to ask “anything else?”
afterwards. “Just” also has a ritualistic purpose: it is a kind of apology (the
customer is sorry for issuing such a banal request); this adverb tends to appear
systematically, in this context of small shops, when the request is for a product
whose value is inferior to an agreed norm, defined according to either the av-
erage of the other transactions carried out in this setting or more locally to the
prior transaction.

4. Concluding remarks

CA has established the golden rule of absolute respect of data, which should be
observed “carefully, closely, seriously, open-mindedly” (Schegloff 1999:€581). For
us, “open-mindedly” means that we can make use of all available analytical re-
sources, as long as they are compatible and adapted to both the object and the
objective of the description. Thus, I have tried to show that in order to account
effectively for what happens in interactions, it is often desirable –€and indeed in
some cases even essential€– to include theoretical propositions from different par-
adigms (for example, we cannot do without the notion of speech act to describe
the sequential organization of a conversation, or the notion of face-work to de-
scribe preference organization).
In their introduction to (On) Searle on conversation (1992:€ 5), Parret and
Verschueren evoke “the classical debate concerning the complementarity or the
exclusivity of different orientations within pragmatics”. Personally, I am inclined to
The case for an eclectic approach to discourse-in-interaction 

favor complementarity, which makes description much richer and more exciting,
albeit less “pure” –€if any approach can ever be totally pure: no model is made up
of exclusively endogenous notions, for concepts spread out, migrate and cross
school and even disciplinary boundaries (as the notion of “interaction” itself il-
lustrates). This commitment in favor of complementarity (that is “eclectism”) is
not without risks, as regards the global coherence of the description. The problem
of how to connect the different layers of analysis, and how to put together a kind of
“integrated model” which would take into account the different constraints that
condition the construction of conversations (“technical” constraints such as turn-
taking, linguistic constraints, ritualistic and cognitive constraints too...) also arises.
This point can perhaps be considered premature.28 To conclude, I would rather
come back to the question of interpretation, which is central to discourse analysis.
As we consider like Gumperz (in Eerdmans, Prevignano and Thibault 2002:€150)
that the analysis of discourse-in-interaction’s main objective is to try to understand
“how speakers understand each other”, we must accept the leitmotiv of conversa-
tion analysis whereby the analysis should “take on the members’ point of view”. But
what exactly should we understand by that? Not that the analyst’s job is the same as
that of the participants: the conditions in which these two categories of interpreters
find themselves are different in all respects. With regard to the “members”, the ana-
lyst is both at a disadvantage (he generally does not have all the relevant contextual
information at hand)29 and at an advantage as he has time (to review the recording
at leisure, and continue to discover new relevant details), and a full set of theoretical
tools and descriptive categories, which are never “natural” ones.30 They are catego-
ries which have been constructed in the context of a specific theory – TCUs, adja-
cency pairs, preliminaries, repairs, etc. are certainly not “membership categories”;
and even apparently “commonsense” categories such as “turn” or “interruption”
must be redefined and refined in order to become operative.
Clearly this does not mean either that the analyst leaves full responsibility of
interpretative work to the members: even if the conversation sometimes carries
“metaconversational” elements, it goes without saying that the presence in the cor-
pus of a term like “request” or “interruption” is not in any way a necessary condi-
tion (nor even in some cases a sufficient one) for identifying a request or an inter-
ruption. To adopt the members’ point of view can only mean that the analyst’s

28. See, however, Roulet et al.’s “modular” model (Roulet, Fillietaz and Grobet 2001) which
proposes an integrated representation of the different constitutive dimensions of all kinds of
discourse, be it monological or dialogical.€
29. Unless he himself also took part in the interaction, which in turn creates other problems.
30. See Segerdahl 2003:€95€; and Schmale 2008 on folk conceptions of conversation (the study
shows how far they are from scientists’ conceptions).
 Catherine Kerbrat-Orecchioni

interpretation must be seen as a reconstruction of the participants’ interpretation,

which is achieved by means that are only partially the same, and explicitly ex-
pressed in a suitable metalanguage.
Whether it concerns the participants or the analyst, the interpretative activity
always consists of extracting meanings from meaningful material delivered by the
speakers, by calling up interpretative resources of a very various nature. This activ-
ity is always a real “job”, despite what we could conclude from some affirmations
like this one:
Mutual understanding is thus displayed, to use Garfinkel’s term, “incarnately” in
the sequentially organized details of conversational interaction. Moreover, be-
cause these understandings are publicly produced, they are available as a resource
for social scientific analysis. (Heritage 1984:€259; emphasis added)

This conception, which represents the participants as working towards “displaying

publicly” the interpretation that has to be attributed to their utterances, is very
comfortable for the analyst, who has just to examine these utterances in detail in
order to extract their meaning. We can never emphasize enough the importance
that should be accorded to even the slightest details of the utterances to be ana-
lyzed; but it is not enough to look, even very closely, at the data to miraculously
have access to their meaning. Indeed, the principles of availability and accountabil-
ity reformulate in their own way the semiotic principle: what is “displayed” are “sig-
nifiers” (in French signifiants –€“cues”, “markers”, “indicators” or “features”), which
orient us towards specific meanings (in French signfiés). But these signifiers are no
more “transparent”31 in a dialogue than in a monologue: describing always means
interpreting, with some degree of subjectivity and risk inherent to this process.
Interpreting dialogical discourse is, in fact, even more complex than interpreting
monological discourse, as it involves reconstructing the interpretations made by
the different participants turn after turn.32 The conversation analyst is an “arch-
interpreter”: he has to hypothesize step by step about the interpretative hypotheses
of conversationalists who are involved in the dynamic process of collectively and
sometimes conflictually constructing interactions. The higher the number of par-
ticipants and the more complex the participation framework, the more difficult

31. For an example of this confidence in the “transparency of understanding”, see LeBaron and
Koschman 2003 (where they talk above all about gestures which accompany verbal material
–€gestures which are more or less iconic, therefore effectively more or less “transparent”, but the
authors do not even discuss this very fundamental point). This text also illustrates the fact that
what appears today to be a sort of denial of the semiotic process goes hand in hand with the dread
of “mentalism” and the obsession of “objective” analysis (like in the good old days of distribu-
tionalism and behaviorism).
32. The linear nature of interpretative work does not rule out the possibility of “retro-interpretation”.
The case for an eclectic approach to discourse-in-interaction 

the analyst’s task will be. For example, when dealing with a TV talk show he has to
take into account not only the different participants in the studio but also the audi-
ence, which is made up of “members” whose understanding is not made “available”
during the communicative event, and which is most heterogeneous. How these
different participants interpret the interaction depends on their point of view (in
every sense of the word) on the interactional scene, as well as on their background
information, which cannot be completely reconstructed by the analyst.
To finish, let us remember that the process of collective discourse construction
is first made possible by the linguistic and more generally communicative compe-
tence of the participants, that means their knowledge of all kinds of conventions
which are pre-existent to discourse (these rules that “generate the sequences”, to
come back to Sacks quoted at the beginning of this article) and which are continu-
ously activated and negotiated throughout – but no negotiation would be possible
without a prior set of rules (which can themselves be subject to negotiation). If the
progression of the exchange develops in such a tentative way, by incessant failures
and repairs, this relentless search for the correct word, the appropriate expression
and the right construction proves very clearly the existence of a system of rules
that have been internalized by the participants in the interaction, and to which
they do their best to conform. The analysis constantly refers (consciously or not,
explicitly or not) to these rules – grammatical rules, pragmatic rules such as those
which regulate speech acts, or lexical rules such as those on which activities of
“categorization” are founded: for example, a speaker can categorize his emotional
state via the term “angry” and its conventional meaning; if he says later that he is
“furious”, we will be led to conclude that he “recategorizes” his state, even in the
absence of any marker such as “I am even furious”, because we know that in Eng-
lish, the terms “angry” and “furious” are not synonymous. In other words: dis-
course analysts constantly make use of their linguistic knowledge; study of “lan-
guage system” and study of “language uses and practices” are to be seen as
definitely complementary.
We cannot seriously claim that language reinvents itself in each moment of
discourse. What is problematic, however, is knowing to what extent the linguistic
system is affected by interaction and more particularly with regard to turn-taking
constraints (given that language use is far from being limited to interactive con-
texts). This question is today under debate within CA (see for example Ochs,
Schegloff and Thompson 1996). Schegloff is cautious about this point (saying in
Prevignano and Thibault eds. 2003, 168: “it seems to me too early to offer ‘answers’
to these questions”), whereas others go so far as to claim that “language is interac-
tionally structured”, and that “linguistic resources […] are shaped by interactional
principles” (Mondada 2000:€24). This point needs to be clarified, therefore, just as
the question as to what we should understand by “emergent” grammar (Hopper
 Catherine Kerbrat-Orecchioni

1988, Streeck 1996) –€we can wonder if this term really means anything different
than the traditional idea that the language system is nothing but a sedimentation
of language use,33 as Saussure or Benveniste already said:
C’est dans le discours […] que la langue se forme et se configure. On pourrait dire,
calquant une formule classique: nihil est in lingua quod non prius fuerit in oratione.
(Benveniste 1966:€131)34

These sedimentation phenomena can only be dealt with from a diachronic point
of view. In one particular instance of discourse-in-interaction, what we can ob-
serve is the collective and tentative construction of malleable utterances according
to a set of flexible rules:
Conversation is like playing tennis with a ball made of Krazy Putty that keeps
coming back over the net in a different shape. (David Lodge, Small World, Penguin
Books, 1985:€25)


Aston, Guy. 1988. “What’s a public service encounter anyway?” In Negotiating Service, Guy
Aston (ed.), 25–43. Bologna: CLUEB.
Atkinson, J. Maxwell and Heritage, John (eds.). 1984. Structures of Social Action. Studies in Con-
versation analysis. Cambridge: CUP.
Benveniste, Emile. 1966. Problèmes de linguistique générale. Paris: Gallimard.
Billig, Michael. 1999. “Critical Discourse Analysis and Conversation Analysis: an exchange be-
tween Michael Billig and Emanuel A. Schegloff ”. Discourse & Society 10 (4): 543–582.
Bilmes, Jack. 1996. “Problems and resources in analyzing Northern Thai conversation for Eng-
lish language readers”. Journal of Pragmatics 26: 171–188.
Brown, Penelope and Fraser, Colin. 1979. “Speech as a marker of situation”. In Social Markers in
Speech, Klaus R. Scherer and Howard Giles (eds.), 33–62. Cambridge: CUP.
Brown, Penelope and Levinson, Stephen. 1978. “Universals in language use: Politeness phenom-
ena”. In Questions and Politeness. Strategies in Social Interaction, Esther Goody (ed.), 56–289.
Cambridge: CUP.

33. Concerning the question as to whether grammar precedes discourse or whether “grammar
is secondary to discourse” (as Hopper asserts, 1988:€121), it reminds us of the sophistic question
about which came first, the chicken or the egg; and regarding the drastic opposition he makes
between two conceptions of grammar, “a priori” vs. “emergent” (supposedly belonging respec-
tively to a “structuralist” vs. “hermeneutic” ideology), this is made possible by caricaturing the
first conception as being “a deterministic view of grammar as a static, prioristic, complete sys-
tem” (ibid.: 132), which no one has ever advocated.
34. “It is in discourse […] that the language is formed and configured. […]. One could say, to
borrow a classical formula: nothing exists in language that was not before in discourse”.
The case for an eclectic approach to discourse-in-interaction 

Brown, Penelope and Levinson, Stephen. 1987. Politeness: Some Universals in Language Use.
Cambridge: CUP.
Cooren, François. 2004. “The Contribution of Speech Act Theory to the Analysis of Conversa-
tion: How Presequences Work”. In Handbook of Language and Social Interaction, Kristine L.
Fitch and Robert E. Sanders (eds.), 21–40. Mahwah, NJ: Lawrence Erlbaum.
Duranti, Alessandro and Goodwin, Charles (eds.). 1992. Rethinking Context. Cambridge: CUP.
Eerdmans, Susan L., Previgano, Carlo L. and Thibault, Paul J. (eds.). 2002. Language and Interac-
tion. Discussions with John Gumperz. Amsterdam/Philadelphia: John Benjamins.
Eggins, Susanne and Slade, Diana. 1997. Analysing Casual Conversation. London/New York:
Fetzer, Anita and Meierkord, Christiane (eds.). 2002. Rethinking Sequentiality. Amsterdam/
Philadelphia: John Benjamins.
Goffman, Erving. 1959. The Presentation of Self in Everyday Life. New York: Anchor Books.
Goffman, Erving. 1981: Forms of Talk. Oxford: Blackwell.
Goodwin, Charles. 1981. Conversational organization: Interaction between Speakers and Hearers.
New York: Academic Press.
Gumperz, John. 2002. “Response essay”. In Eerdmans et al. (eds.) 2002, 105–126.
Heritage, John. 1984. Garfinkel and Ethnomethodology. Cambridge: Polity Press.
Heritage, John. 1995. “Conversation Analysis. Methodological aspects”. In Aspects of Oral Com-
munication, Uta M. Quasthoff (ed.), 391–418. New York: Walter de Gruyter.
Hopper, Paul J. 1988. “Emergent Grammar and the A Priori Grammar Postulate”. In Linguistics
in Context: Connecting Observation and Understanding, Deborah Tannen (ed.), 117–134.
Norwood: Ablex.
House, Juliane. 2000. “Understanding misunderstanding: A pragmatic-discourse approach to
analysing mismanaged rapport in talk across cultures”. In Culturally Speaking, Helen Spen-
cer-Oatey (ed.), 145–164. London/New York: Continuum.
Hymes, Dell. 1972. “Models of the interaction of language and social life”. In Directions in Socio-
linguistics: The Ethnography of Communication, John Gumperz and Dell Hymes (eds.),
35–71. New York: Holt, Rinehart and Wilson.
Kerbrat-Orecchioni, Catherine. 1990–1992–1994. Les interactions verbales (3 vol.). Paris: A.€Colin.
Kerbrat-Orecchioni, Catherine. 2001. Les actes de langage dans le discours. Paris: Nathan.
Kerbrat-Orecchioni, Catherine. 2004a. “Introducing Polylogue”. Journal of Pragmatics 36: 1–24.
Kerbrat-Orecchioni, Catherine. 2004b. “Que peut-on faire avec du dire?”. Cahiers de Linguis-
tique Française 26 (“Les modèles du discours face au concept d’action”): 27–43.
Kerbrat-Orecchioni, Catherine. 2005. Le discours en interaction. Paris: Nathan.
LeBaron, Curtis D. and Koschmann, Timothy. 2003. “Gesture and the transparency of Under-
standing”. In Studies in Language and Social Interaction. In Honor of Robert Hopper, Phillip
J. Glenn et al. (eds),119–132. Mahwah, NJ: Lawrence Erlbaum.
Lerner, Gene H. 1996. “Finding ‘Face’ in the Preference Structures of Talk-in-Interaction”. Social
Psychology Quaterly 59 (4): 303–321.
Levinson, Stephen. 1983. Pragmatics. Cambridge: CUP.
Marandin, Jean-Marie. 1987. “Des mots et des actions: ‘compliment’, ‘complimenter’ et l’action
de complimenter”. Lexique 5: 65–99.
Mey, Jacob. 1994. Pragmatics: An Introduction. London: Blackwell.
Mondada, Lorenza. 1998. “Variations sur le contexte en linguistique”. Cahiers de l’ILSL 11: 243–266.
 Catherine Kerbrat-Orecchioni

Mondada, Lorenza. 2000. “Analyse conversationnelle et ‘grammaire-pour-l’interaction’”. In

Modèles du discours en confrontation, A.-C. Berthoud and L. Mondada (eds.), 23–42. Berne:
Peter Lang.
Ochs, Elinor, Schegloff, Emmanuel A, and Thompson, Sandra A. 1996. Interaction and Gram-
mar. Cambridge: CUP.
Parret, Herman and Verschueren, Jef. 1992. “(On) Searle on conversation: An introduction”. In
Searle et al. 1992, 1–5.
Pomerantz, Anita. 1978. “Compliment Responses”. In Studies in the Organization of Conversa-
tional Interaction, J. Schenkein (ed.), 79–112. New York: Academic Press.
Prevignano, Carlo L. and Thibault, Paul J. (eds.). 2003. Discussing Conversation Analysis. The
work of Emanuel A. Schegloff. Amsterdam/Philadelphia: John Benjamins.
Roulet, Eddy.1992. “On the Structure of Conversation as Negotiation”. In Searle et al. 1992, 91–99.
Roulet, Eddy, Fillietaz, Laurent and Grobet, Anne. 2001. Un modèle et instrument d’analyse de
l’organization du discours. Berne: Peter Lang.
Sacks, Harvey. 1987. “On the preferences for agreement and contiguity in sequences in conver-
sation”. In Talk and Social organization, Graham Button and John R.E. Lee (eds.), 54–69.
Clevedon/Philadelphia: Multilingual Matters LDT.
Sacks, Harvey. 1992. Lectures on Conversation (2 vol.). Oxford (UK)/Cambridge (USA): Blackwell.
Sacks, Harvey, Schegloff, Emanuel A. and Jefferson, Gail. 1974. “A simplest systematics for the
organization of turn-taking for conversation”. Language 50: 696–735.
Schegloff, Emanuel A. 1972. “Sequencing in conversational opening”. In Directions in Sociolin-
guistics: the Ethnography of Communication, John Gumperz and Dell Hymes (eds), 346–380.
New York: Holt, Rinehart and Winston.
Schegloff, Emanuel A. 1984. “On some questions and ambiguities in conversation”. In Atkinson
and Heritage (eds.) 1984, 28–52.
Schegloff, Emanuel A. 1986. “The routine as achievement”. Human Studies 9: 111–151.
Schegloff, Emanuel A. 1987. “Entre Micro et Macro: Contextes et Relations”. Sociétés 14: 17–22.
Schegloff, Emanuel A.1988. “Presequences and indirection. Applying speech act theory to ordi-
nary conversation”. Journal of Pragmatics 12: 55–62.
Schegloff, Emanuel A. 1992a. “To Searle on Conversation: A Note in Return”. In Searle et al. 1992,
Schegloff, Emanuel A. 1992b. “In another context”. In Duranti and Goodwin (eds.) 1992,
Schegloff, Emanuel A. 1999. “Critical Discourse Analysis and Conversation Analysis: an exchange
between Michael Billig and Emanuel A. Schegloff ”. Discourse & Society 10 (4): 543–582.
Schmale, Günter. 2008. “Conceptions populaires de la conversation”. Pratiques 139–140: 5–80.
Searle, John R. 1972. Les actes de langage. Paris: Hermann [Translation of Speech acts, Cambridge,
CUP, 1969].
Searle, John R. et al. 1992. (On) Searle on Conversation. Amsterdam/Philadelphia: John Benjamins.
Segerdahl, Pär. 2003. “Conversation analysis as a rigourous science”. In Prevignano and Thibault
(eds.) 2003, 91–108.
Selting, Margret. 2000. “The construction of units in conversational talk”. Language in Society 29
(4): 477–517.
Sinclair, John Mac H. and Coulthard, R. Malcolm. 1975. Towards an Analysis of Discourse: The
English used by Teachers and Pupils. London: Oxford University Press.
Streeck, Jürgen. 1980. “Speech Acts in Interaction: A Critique of Searle”. Discourse Processes 3:
The case for an eclectic approach to discourse-in-interaction 

Streeck, Jürgen. 1996. “A little Ilokano grammar as it appears in interaction”. Journal of Pragmat-
ics 26: 189–213.
Thibault, Paul J. 2002. “Contextualization and social meaning-making practices”. In Eerdmans et
al. (eds.) 2002, 41–61.
Vicher Anna, and Sankoff, David. 1989: “The emergent syntax of pre-sentential turn openings”.
Journal of Pragmatics 13: 81–97.
Wetherell, Margaret. 1998. “Positioning and interpretative repertoires: conversation analysis
and post-structuralism in dialogue”. Discourse & Society 9 (3): 387–412.
A neglected resource in interaction analysis?

Peter Muntigl and Eija Ventola

Simon Fraser University and University of Helsinki

Our aim in this paper is to explore the variety of ways in which grammar can
be used as a resource for interaction. We propose that a grammatical analysis
of social interaction needs to take two perspectives into account. The first
involves showing how the meaning of a grammatical unit depends on its
context of use (Schegloff 1996). The second involves identifying speakers’
specific choices of grammatical categories (e.g., verbs, nouns, etc.) and showing
how these categories combine to create specific inference-rich meanings
(Halliday 1978, 1994). Drawing from couples therapy data, we show that by
adopting grammatical categories into our analysis, a unique perspective on the
moment-to-moment unfolding of therapy can be provided.

1. Introduction

There seems to be a growing trend in conversation analytic research these days in

which increasing attention is being given to the relationship between language and
social interaction. On the one hand, the emergent interest in language stems from
the observation that ‘words’ play a key role in constructing our social lives and re-
lationships. As Lerner (2004a: 93) has argued in a recent special issue on Practices
of Turn Construction in Conversation, “The words people use in conversation with
others are not only building blocks for sentences. Words are issued as part of the
unfolding lives of their speakers in almost all social encounters with others – and
most of these encounters are organized by talk-in-interaction.” On the other hand,
this renewed interest has been fueled by a different kind of research agenda, one
that we feel is even more significant; this is the recognition that grammar is a central
 Peter Muntigl and Eija Ventola

resource used in social interaction1 (see Ochs, Schegloff and Thompson 1996). One
major line of research in this regard focuses on grammatical projectability in turn
taking. In this work, grammar is seen as an interactional resource that allows speak-
ers to project (i.e., infer) actions in progress and possible turn completions (Good-
win 1986; Sacks, Schegloff and Jefferson’s 1974; Schegloff 1996). Moreover, since
different languages may use different clause formats, a speaker’s ability to project
turn completions may depend on the language being spoken (see Thompson and
Couper-Kuhlen (2005) for a comparison of English and Japanese).
To those linguists who adopt a functional approach to language – who have an
interest in language in use – this endorsement of grammar from conversation ana-
lysts is certainly welcome. But for all the renewed interest in words and the gram-
matical formats of turns and turn constructional units, conversation analysts still
seem to be somewhat wary about the extent to which grammar, and linguistic
categories/concepts in particular, may play a role in explaining social interaction.
There are, presumably, many reasons for why grammar has an equivocal status.
One reason involves the conversation analytic tenet that any analysis of social in-
teraction must be sensitive to participants’ orientations to meanings. This means
that if categories are imported into an analysis, the analyst must demonstrate that
conversational participants are in some way orienting to these categories. As a re-
sult, linguistic categories such as Agent, Subject, Modality, Theme can only form
part of an analysis if these categories can be shown to be interactionally relevant to
the conversational participants and to the trajectory of the interaction.2 A second
reason stems from Schegloff ’s (1984) observation that there is no one-to-one rela-
tionship between grammatical structure and social action. For example, a turn
containing a wh-interrogative (“why don’t you come and see me sometimes”) does

1. It should be noted that in systemic linguistics, words (i.e., lexis) and grammar are consid-
ered to be part of the same phenomenon, with lexis being the most delicate realization of the
grammar (see Hasan 1987).
2. An even stronger constraint on analyst categories comes from the conversation analytic
view on social context and sociological variables. Schegloff takes a very strong position on this
issue, arguing that all interpretations of the data must be endogenously grounded in the details
of talk-in-interation; for, as Schegloff (1997:€165) very forcefully argues in his abstract, “…this is
a useful constraint on analysis in disciplining work to the indigenous preoccupations of the
everyday world being grasped, and serving as a buffer against the potential for academic and
theoretical imperialism which imposes intellectual’s preoccupations on a world without respect
to their indigenous resonance.” To put it briefly, Schegloff is arguing against “critical discursive”
research that assumes certain sociological variables such as gender or class to be relevant to the
text, without first having looked at the text. Schegloff ’s position is to examine the details of the
text first, and then see if any meaningful links can be made to such sociological variables. From
this we can assume that any links postulated between grammar and social interaction will be put
to the same test.
Grammar 

not necessarily function as a question. Rather, it may be doing different things

such as inviting (Schegloff 1984:€31). A third reason may be Schegloff ’s (1996) in-
sistence that grammar is only one resource among many others that speakers draw
from in instances of collaborative meaning making; laughter, silences, aspirations,
cut-offs, sound stretches, gestures, etc. may all influence the structure of a turn
constructional unit in progress (Schegloff 1996:€103–104). All in all, conversation
analysts seem to be fairly cautious when asked about the relationship between
grammar and other levels of language or social organization. When Schegloff was
asked in an interview if he could “… clarify the bi-directional relationship between
the levels of grammar, or linguistic forms, and turns in conversation and their
organization?”3, part of his response was “It seems to me too early to offer “an-
swers” to these questions yet. In fact, it is probably too early to know yet whether
these are the right questions, or ones formulated in the most productive way”
(Prevignano and Thibault 2003:€167–168).
Faced with these caveats concerning which categories can or should be in-
cluded in an analysis of social interaction, linguists may feel that they must tread
carefully for fear of focusing too much on grammatical structure and not enough
on interactive meaning making. We would argue, however, that by taking a func-
tional linguistic approach to social interaction, linguistic categories can be used in
an analysis without compromising the concerns raised by conversation analysts.
An approach that we feel is particularly well suited for this endeavour is systemic
functional linguistics (Halliday 1994). In this chapter we aim to demonstrate some
important ways in which speakers’ grammatical structures realize (and are real-
ized in) social interaction. Drawing from a systemic functional linguistic
(hereafter SFL) approach, we argue that a grammatical analysis does not necessar-
ily detract from how other meaning-making resources are being deployed in inter-
action; especially if the analytic focus includes meaning or function in addition to
structure. Nor does a grammatical analysis using grammatical categories necessar-
ily imply that an analyst’s as opposed to a participant’s perspective on social inter-
action is being taken. For example, showing that speakers are meaningfully link-
ing their grammatical structures to other places within the interaction provides a
strong warrant for saying that speakers are drawing from (i.e., orienting to) vari-
ous grammatical structures and meanings.

3. The complete question was: “From your perspective, could you clarify the bi-directional
relationship between the levels of grammar, or linguistic forms, and turns in conversation and
their organization? In other words, how do you see constraints emanating from the level of
grammar operating in relation to turn-taking in conversation and the construction of turns,
and, secondly, how does turn-taking impinge on grammar and its specific contribution?”
(Prevignano and Thibault 2003:€167)
 Peter Muntigl and Eija Ventola

In the remainder of this paper, we attempt to explore the concept of social in-
teraction from an SFL perspective, by showing how speakers draw on grammatical
resources in situations of meaning making. In particular, we focus on three differ-
ent modes of the grammar that are commonly referred to as interpersonal, idea-
tional and textual (see Halliday 1994). We aim also to make links with studies in
conversation analysis (hereafter CA) and CA-related approaches that have an in-
terest in the meaning making potential of grammar. Our goal, therefore, is not
simply to illustrate that the grammar is functionally-oriented, but also to show
that, in Schegloff ’s (1996) sense, that the grammar is “positionally sensitive” to the
interactional context in which it occurs.

2. Grammar in conversation analytic(-related) work

The importance of grammar in CA (and CA-related) work can be succinctly sum-

marized as follows: “Grammar is not only a resource for interaction and not only
an outcome of interaction, it is part of the essence of interaction itself. Or, to put it
another way, grammar is inherently interactional” (Schegloff, Ochs and Thompson
1996:€28). This clearly takes any notion of grammar away from the more formal
linguistic position that grammars are autonomous systems. What is highlighted
instead is the grammar’s ‘role’ in discourse and its positional or context sensitivity;
grammars are resources for doing things, but they also derive their meaning from
the surrounding talk. A wonderful example of this is how a seemingly ‘simple’
word such as “oh” can mean a variety of things depending on how it is uttered and
where it occurs within a sequence of turns (Heritage 1998).
Every CA-type analysis will, of course, involve examining words and/or gram-
matical structures to some degree. Sacks’ “Lectures on Conversation” are filled
with examples in which he shows how a speaker’s choice of words and grammati-
cal elements form a constitutive part of social interaction. Most CA studies, how-
ever, do not take grammar as a point of departure for their investigations; that is,
conversation analysts are not interested in grammar per se, but in how language
more generally is used to make meanings and construct a social world in com-
mon. But there are studies in CA that place special emphasis on words and gram-
mar. We mention but a few of these studies that have focused specifically on
English. Turning first to the study of ‘words in use’, much work has been done on
membership categories (see especially Sacks 1992 and Antaki and Widdicombe
1998), conjunctions (Heritage and Sorjonen 1994; Raymond 2004; Streeck 2002;
Turk 2004), evidentials such as ‘seem’ and ‘think’ (Roth 2002; Heritage and
Raymond 2005) and response tokens such as ‘uh huh’, ‘mm’, ‘yeah’, ‘oh’ (Beach 1993;
Gardner 1997, 2001; Heritage 1984b, 1998; Jefferson 1984; Schegloff 1982). Turning
Grammar 

now to grammatical structures, studies have focused on mood such as interroga-

tives (Heritage and Roth 1995; Raymond 2003), reported speech (Holt and
Clift 2007), sentence grammar (Goodwin 1981), sentence/turn completions and
extensions (Lerner 1991, 2004b), turn-constructional-units (Sacks, Schegloff and
Jefferson 1974; Schegloff 1996) and footing (Clayman 1992; Antaki et al. 1996).
Other approaches to language that draw heavily from CA methodology have
also contributed substantially to our understanding of how grammars are used in
discursive contexts. One approach that especially foregrounds speakers’ uses of
linguistic (i.e., syntactic and prosodic) resources is known as interactional linguis-
tics (for recent overviews see Barth-Weingarten 2008 and Selting 2008). Another
approach, known as discursive psychology, aims to reconceptualize ‘psychological’
phenomena such as ‘memory’, ‘affect’ and ‘perception’ in discursive and rhetorical
terms (Edwards 1997; Edwards and Potter 1992; 2005; Potter 1996, 1998). In par-
ticular, their work has contributed to our awareness of how ‘cognitive’ terms such
as ‘thinking’, ‘wanting’, ‘knowing’ and the like are positionally relevant and may do
a variety of discursive and rhetorical work. This theme of ‘mental predicates’ and
how they operate as interactional resources has also been taken up by ethnometh-
odologists. Most notable is their examination of the interactional operations of
memory and how expressions involving ‘remember’, ‘recall’ and ‘forget’ can do ac-
counting work by avoiding blame (Coulter 1985; Lynch and Bogen 2005).
In CA, the interest in words and grammatical structures has largely been on
their ‘positional sensitivity’ within a sequence of talk (Schegloff 1996). This cor-
responds to a ‘view from above’ in which language is examined with respect to the
social and interactional context in which it is embedded (see Halliday 1978).
Linguists, however, are often prone to take what Halliday calls a ‘view from within’;
that is, grammar is examined with respect to its component grammatical struc-
tures (e.g., prepositional phrases, nominal groups, verbal groups). For instance,
some grammatical approaches are interested in examining the ‘grammatical’ com-
pany that a certain verb takes. The verb ‘walk’, for example, takes a noun phrase
(someone who does who walking) and may, but need not, take a prepositional
phrase indicating temporal duration or the target location of the walk (e.g., ‘for an
hour’, ‘to the store’). Examining a clause in this way can provide us with informa-
tion about how combinations of words are construing experience. What this means
is that the grammar of a word, e.g., walk, can tell us something about its meaning.
Some ethnomethodologists such as Coulter (1985, 1989, 1999) have been strong
proponents of this form of grammatical analysis. In his work on ‘forgetting’ and
‘remembering’, Coulter (1985) argues that an understanding of what ‘forgetting’
means requires grammatical inquiry along the lines of Wittgenstein (1958) and
Ryle (1949). For example, the term forgetting is not an activity-type verb such as
walking or playing. We do not generally ask “when did you forget?” (i.e., at what
 Peter Muntigl and Eija Ventola

specific hour/minute/second did you forget?) or “how long did your forgetting
take?”4 Through various forms of grammatical tests we can tease out the behav-
iour of verbs and determine whether they construe states (know, believe), activi-
ties (run, walk), accomplishments (paint a picture, make a table) or achievements
(win a race, realize).5,
Functional linguists would also call attention to the kind of situation or event
that the clause and/or verb is bringing into existence (e.g., Dik 1997; Halliday 1994;
van Valin and Lapolla 1997). But, as someone more interested in discourse (and
less in clause grammar) may ask, how would knowing that a verb construes a state
or activity tell us anything about a social interaction? Wouldn’t this just merely be
a linguist’s concern in sorting out verb typologies? We would argue that it cer-
tainly can be a participant’s concern. Consider example (1), taken from a couples
therapy session. The transcript notation used for all examples is shown in Table€1.

Table 1.╇ Transcript Notation

(.) untimed short pause (less than .5 seconds)

(1.0) timed pause
[] overlapping speech, e.g., A: how was the [movie ]
B: [great ]
= contiguos utterance e.g., A: how was the movie=
B: =great
: extended sound, e.g., we:::ll
underline emphasis
CAPS greater emphasis
.hh in breath
hh outbreath e.g., t(hhh)ake
() transcriber’s guess at speaker utterance
(?) unidentifiable speaker
((laughter)) Non-speech vocalizations are placed in double parentheses

4. Coulter was principally arguing against cognitive science views of the ‘mental’, in which
remembering and forgetting are viewed as processes that involve storage, retrieval, etc. Coulter’s
point is that our grammar reveals that we do not understand these terms as being activities.
5. This typology of verb-types is taken from Vendler (1967). Ryle (1949) proposed similar
categories such as ‘act-verb’, ‘achievement-verb’ and ‘success-verb’.
Grammar 

(1) 1 Ther: a::nd ((directs gaze at S)) what was the occasion of your,
2 [(0.5) ]
3 Ther: [((shifts gaze towards M)) ]
4 Ther: split. (in February)
→ 5 Mark: domestic violence.
6 Mark: ((gazes at T, slightly raises eyebrows when saying “domestic”))
The therapist asks Mark what caused (i.e., occasioned) Mark and Stacey’s (Mark’s
spouse) separation, to which Mark replies “domestic violence.” Now, although
Mark does not supply an actual ‘verb’, we may infer an activity-type event realized
in the nominalized form of “violence” in which a certain kind of (violent) action
has taken place. Furthermore, there is the implication that there are two ‘partici-
pants’ involved in the action: this would be someone doing the action and someone
(or something) to whom the action is done. From the classifier “domestic”, we may
also infer that the violence is done to someone, and in this particular case it is more
specifically an act of spousal violence. But, the way it stands, we do not know from
Mark’s response who was violent to whom. This information is left unspecified.
In order to show that it is not just us analysts but more importantly it is the
therapist who is also orienting to these implied meanings, let us examine from
example (2) what immediately follows.
(2) 7 Ther: o:kay. ((makes ‘body’ nod, nodding head and leg))
8 (1.8) ((maintains gaze with M))
9 Ther: on uh,
10 (1.0)
→ 11 Ther: on yer part?
What the therapist does in line 11 is prompt Mark to expand on his response
(see Muntigl and Hadic Zabala 2008 on therapist prompts), and the Therapist does
so by designing his prompt so as to grammatically latch onto Mark’s noun phrase:
“domestic violence … on yer part?” The prompt targets the ‘who to whom’ infor-
mation and one could well imagine that this sort of information may be important
for a therapist in sorting out the degree of violence in the relationship.€What this
brief example shows is that speakers are attuned to how experience gets construed
and they will, therefore, pay close attention to the grammar of these construals.
The therapist worked to unpack the unspecified yet inferable information deriving
from the nominalized term “domestic violence”.6 Other relevant information

6. Now, of course, conversation analysts are aware that terms and categories are inference-rich
(see Sacks 1992). In CA terms, the production of membership categories such as ‘boyfriend’ or
‘girlfriend’ will immediately make relevant specific meanings. In a similar vein, Potter (1996)
argues that certain terms carry with them certain category entitlements that can map onto certain
 Peter Muntigl and Eija Ventola

associated with the expression “domestic violence” was also elicited by the thera-
pist later on in the sequence. This included ‘what form the violence took’ (pushing
Stacey through the door backwards), ‘how the violence came about’ (initiated by
drinking) and ‘the result of the violent episode’ (physical injury).
To sum up, although we whole-heartedly agree with the general claim made
within CA that sequence and turn-taking organization provides the architecture
for negotiating meaning and intersubjectivity (Heritage 1984a; Schegloff 2007), we
would also add that the grammar, and what we will be later referring to as interper-
sonal, ideational and textual grammar, can influence how a sequence of social ac-
tions gets played out. For the remainder of this paper, we show how Hallidayan
functional linguistics offers some powerful conceptual tools with which one can
identify ‘grammars in use’.

3. A brief view of grammar & interaction in systemic functional linguistics

Within an SFL approach, interaction can be generally defined as persons engaged

in a collaborative meaning making activity that occurs in a socially relevant con-
text; as writers/readers if we are dealing with written text, or speakers/hearers if we
are dealing with spoken text. What makes the SFL approach to interaction distinc-
tive is the special focus given to language and social context in the meaning mak-
ing process (for extensive overviews of SFL theory see Halliday 1978, 1994, 2003,
2005; Halliday and Hasan 1985; Halliday and Matthiessen 1999). We will address
language first. According to Halliday (1975), language has developed in response
to three kinds of social-functional ‘needs’. The first is to be able to construe experi-
ence in terms of what is going on around us and inside us. The second is to interact
with the social world by negotiating social roles and attitudes. The third and final
need is to be able to create messages with which we can package our meanings in
terms of what is New or Given, and in terms of what the starting point for our
message is, commonly referred to as the Theme. Halliday (1978) calls these

social roles or epistemic rights. Potter (1996:€ 133) provides an example of the differences in
categorizing a ‘newspaper reporter’ as ‘journalist’ or ‘hack’. Whereas the former category can
evoke meanings of ‘objectivity’ and ‘truth’, the latter may instead call to mind contrastive mean-
ings such as ‘subjectivity’ and ‘corruption’. Returning to the therapy example, the term ‘domestic
violence’ is also a category that generates certain kinds of inferences. Our point is that some of
these inferences are bound up with the grammar of violence, as in ‘X does Y to Z’ where ‘does Y’
is a violent act, X is the doer and Y is the one to whom the violent act is done. Further, violent
acts can be construed as having causes and results. A nominalization, however, leaves a range of
meanings unspecified and in this way opens the door to their future specification and negotia-
tion in the unfolding interaction.
Grammar 

language functions metafunctions, and refers to them as ideational, interpersonal

and textual respectively.
Halliday’s point is that any piece of language calls into play all three metafunc-
tions simultaneously. Let us take the clause in lines 07–08 from Ex. (3) as an exam-
ple, which is taken from a different couples therapy session.
07 Wen: he likes to lecture? (1.2) on any: any subject
08 that he feels even mildly uh uh y’know animated abou::t

Ideationally, the speaker is construing experience in terms of how she views her
husband’s behaviour. This construal is realized in three parts. He is given the par-
ticipant role of ‘behaver’, likes to lecture realizes a ‘behavioural process’ and on any:
any subject that… realizes the ‘aboutness’ of the husband’s behaviour. Interperson-
ally, the clause is a highly evaluative statement about the husband. Evaluation is
accomplished by linking the husband’s behaviour with meanings such as desire
and strong inclination. Through these links the husband’s behaviour may gain the
interpretation of being overly persistent and extreme or not quite ‘normal’. Textu-
ally, the clause places the husband in thematic position, with the new information
realizing the scope of the husband’s lecturing. In other words, the clause is ‘about’
the husband and the new information involves the husband’s ‘extreme’ or overly
persistent behaviour.
Although, as we mentioned before, all three metafunctions are always ‘in op-
eration’ for any given discourse, the interpersonal dimension does seem to play a
prominent role in social interaction. This is because the interpersonal mode pro-
vides such resources as epistemic modal expressions, social actions and the or-
ganization of actions in sequence. This is very much in keeping with the CA con-
cept of sequence organization as “…the vehicle for getting some activity
accomplished” (Schegloff 2007:€2). The interpersonal mode can thus be seen as a
frame in which the other metafunctions are realized; our ways of construing expe-
rience (ideational) and organizing our message (textual) is framed by our unfold-
ing collaborative actions.
The above illustration served to briefly sketch out the relationship between the
different language functions and the various clause units that helped to realize those
functions. Function, therefore, is closely tied up with the clause and the clause’s
units. Changing the content of a clause (i.e., modifying a clause’s grammatical struc-
ture by inserting a different verbal group or prepositional phrase) will certainly
have an impact on its function. Related to the notion that language is functional is
the claim that language always occurs in some relevant social context or ‘context of
use’. For, according to Halliday (1978:€ 28), “we do not experience language in
isolation – if we did we would not experience it as language – but always in relation
 Peter Muntigl and Eija Ventola

to a scenario, some background of persons and actions and events from which the
things which are said derive their meaning.” The term ‘social context’ refers to
something very different from aspects of the material environment in which people
interact. Instead, social context is an abstract concept that corresponds to “first,
what is actually taking place; secondly, who is taking part; and thirdly, what part the
language is playing” (Halliday 1978:€31). The first aspect of social context, what is
actually taking place, is termed ‘field’ and refers to the subject matter and social ac-
tivity that people are co-producing. For example, W’s clause above may be said to be
realizing the social activity of complaining about her husband (see all of example
(3) to convince yourself of this). The second aspect, who is taking part, refers to the
social roles of all participants, which include status differences and the degree of
social distance and affect found between participants. The third and final aspect,
what part the language is playing, refers to the ‘mode’ of the interaction. Relevant
aspects of mode include the medium (does the interaction involve writing or speak-
ing?) and channel (are both visual and aural channels open to all participants?). It is
important to emphasize that social context is a dynamic concept; that is, social roles,
social activity, and mode of interaction are not static and nor do they in some way
come before language. Rather, social context and language unfold together. Just as
certain language selections can construct and alter the social activity or the social
roles of the speakers, so do the already co-constructed social activities and social
roles influence how language will be used. The aim in any analysis of text is to begin
at both ends and not assume that the social context is somehow ‘given’ and will ex-
ercise its influence over the unfolding of the whole text.
In order to take stock of how people collaboratively make meaning, or inter-
act, we need to take how speakers use linguistic resources into account; that is, we
need to be able to describe how speakers are using grammatical resources with
respect to ideational, interpersonal and textual meanings. In the next sections, we
explicate the role these resources have in the meaning-making process.

4. Grammar as an interactive resource

The basic grammatical unit of analysis in examining interaction is the clause. A

central assumption is that meanings – whether they are ideational, interpersonal
or textual – are organized with respect to clausal units.7 Researchers in conversational
analysis and interactional linguistics have also argued for the primacy of the clause

7. This does not mean that meanings are only realized in major clauses. Minor clauses such as
exclamations, calls, greetings and alarms also realize metafunctional meanings (Halliday
1994:€95). These minor clauses, however, tend to be mono-functional and not tri-functional.
Grammar 

with respect to organizing interaction (see Thompson and Couper-Kuhlen 2005).

In this work, clauses are viewed as a speaker’s resource for projecting next turn
onsets, joint utterance completions and turn unit extensions. We would argue that,
in addition to the clause being an essential resource for anticipating an interlocu-
tor’s current or next actions, the clause also provides insight into how speakers
construct multifunctional utterances and how a speaker’s meanings are cohesively
and coherently tied to her/his own construction of the message and to another’s
previous message. Paying close attention to clause and clause unit production also
helps to explain how speakers co-construct an architecture of intersubjectivity (Her-
itage 1984a: 254); that is, it is partly through speakers’ production of meaningful
ideational, interpersonal and textual ties within and between turns that enables
speakers to display an understanding of each other’s social actions.
In order to demonstrate how the grammar works as an interactional resource
through which speakers construct multiple meanings that are cohesively and co-
herently tied to other meanings within and across turns, a short conversational
segment taken from a couples therapy session will be analyzed (see Ex. 3, parts of
which derive from Muntigl 2004a: 189–190; 2004b: 118–120).
(3) 01 Wen: .hh hhh well I think
02 that uh (2.0) Fred just gives up now hh heh.hh
03 I think
04 that secretly he still wants to win the argument.hh
05 he wants to prolo::ng
06 Fred is a: uh
07 he likes to lecture? (1.2) on any: any subject
08 that he feels even mildly uh uh y’know animated abou::t
09 he likes to lecture
10 and and go on and on and on and on about it.hh
11 and I- there wuz one this morning or yesterday or something
12 that.hh that I thought
13 well its deci:ded
14 but Fred still had to::.hh really make sure
15 that I knew.hh what wuz going on uh
16 that that uh he had he had pressed his point
17 [he has] done that all uh all my my years with him
→ 18 Ther: [what I’m]
→ 19 Ther: .hh what I’m starting tuh see here is a pattern uh um.hh
→ 20 as a couples therapist um I’m always looking for patterns?.hh
→ 21 that people get into that they get stuck in.hh
→ 22 and I’m I need your agreement
 Peter Muntigl and Eija Ventola

→ 23 as tuh whether or not what we’re seeing here is this particular
→ 24 .hh umm which is leading tuh the kinda communication
→ 25 that (both) you’re talking about.hh
→ 26 your experience Wendy Sue.hh is of uh living with a man
→ 27 who has a lecturing style
28 Wen: yes
→ 29 Ther: who talks to you cons-
30 excuse me if I I’m relating back in her ways.hh ahh
→ 31 who who lectures you constantly.hh
→ 32 who goes on about things bove and beyond the point
→ 33 of being [resolved ] when they’re already resolved
34 Wen: [yes ]
35 Wen: yes
36 Ther: [(what perhaps would you create tuh) ]
37 Wen: [I would agree I I would agree with that ]
38 Fred: [that’s uh ( ) ]
39 Wen: I jus I just want tuh give up an and just say yes
40 Ther: okay I wanna I wanna talk about that
41 I wanna talk in the little bit of the time that we have left about um
42 about that particular pattern.hh
43 um m: maybe since you’re speaking about I’ll ask you um.hh
44 what are y- impact or effect.hh does this lecturing style that you
45 experience from Fred
46 what does- what impact does that have on you.hh as a person
47 Wen: .hh oh well I it makes me feel like uh a child?.hh
48 to a certain degree.hh it makes me feel
49 like I’m (1.0) I I’m ignorant
50 that I can’t grasp it quickly.hh umm
One key component of Ex. (3) involves a therapist (re)formulation of a client’s turn
(lines 18–33). The term formulation is taken from Heritage and Watson (1979:€141)
and refers to “the provision of candidate readings for the sense established in pre-
ceding stretches of talk”. These candidate readings generally do the work of provid-
ing the gist or an upshot of what a previous speaker had said (for examinations of
formulations in therapy see Antaki 2008; Antaki et al. 2005; Buttny 1996; Davis
1986; Hak and de Boer 1996; Muntigl 2004a, 2007). We feel that formulations are
Grammar 

especially illustrative of the grammar at work in interaction, because they reflect

the therapist’s attempt to, on the one hand, tie her/his meanings to the client’s
meanings and, on the other hand, subtly modify the client’s meanings. Thus, by
paying close attention to the therapist’s use of grammar, we will be able to detect
similarities and differences in how the client has construed experience and how
the therapist is now construing (the client’s) experience.
In the following subsections, the therapist formulation will be examined
with respect to the ideational, interpersonal and textual work that the therapist
is performing.

4.1 Interpersonal meaning

An interpersonal analysis requires that the social actions of exchanges, or se-

quences of actions, be identified along with the speakers’ appraisals or assessments
of people and events (Martin 1992, 2000; Muntigl 2009; Thompson and Muntigl
2008; Ventola 1984, 1987). We will be restricting our analysis to the first 35 lines of
example (3). The first ‘move’ involves an extended complaint in which Wendy pro-
vides an elaborate critique of her husband Fred’s lecturing (lines 1–17). This is
followed by a therapist formulation in lines 18–33, which then ends with Wendy
voicing agreement (lines 34–35).
Wendy’s extended complaint bears the features of what Edwards (1994, 1995) re-
fers to as a script formulation. From his work on couples therapy, Edwards (1995:€319)
argues that they appear in the form of relationship scripts and are “descriptions of ac-
tions and events that characterize them as having a recurring, predictable, sequential
pattern.” In Wendy’s case, she is making the claim that Fred’s lecturing is not only
frequent and predictable, but that he has a strong inclination and desire to lecture her.
These meanings of frequency and inclination/desire are spread throughout the vari-
ous grammatical structures of Wendy’s clauses. Frequency, for example, is realized
through iterative present tense (“likes to lecture”) and within a temporal circumstance
(“he has done that all uh all my my years with him”). Inclination, on the other hand, is
realized in the desiderative component of a verbal group (“likes to lecture”), through
the repetition of the verbal group (“go on and on and on and on”) and metaphorically
within the predicator and complement (“still had to::.hh really make sure”; “pressed his
point”). Fred’s strong inclination to lecture is also expressed through the trope ‘I
thought X, but then Y happened’ (Sacks 1992; Jefferson 2004). This trope sets up a
contrast between what Wendy was expecting to occur (i.e., the end of a discussion and
the cessation of Fred’s lecturing) and what actually occurred (i.e., Fred’s continued and
persistent lecturing); that is, even in situations where there is no justifiable need to
lecture (“I thought well its deci:ded”), Fred is ‘driven’ to perform his behaviour (“but
Fred still had to:: …”). By citing this example and by realizing meanings of usuality and
 Peter Muntigl and Eija Ventola

inclination mentioned above, Wendy is able to negatively appraise Fred’s behaviour as

being overly persistent and highly frequent (and, by implication, abnormal).
Seen within the larger context of this therapy session, Wendy’s script formula-
tion may be performing other kinds of interactional and rhetorical business above
and beyond merely demonstrating how obsessive Fred’s lecturing is. For example,
Edwards (1995) argues that script formulations may serve to counter a spouse’s
previous version of events. Prior to Wendy’s extended complaint, Fred had also
been complaining that Wendy is ‘not enthusiastic about things’ and tends to with-
draw. In this way, Wendy’s lack of enthusiasm and lack of interest can be re-inter-
preted as a reaction to Fred’s excessive lecturing. As Wendy herself puts it later on
in line 39, “I jus I just want tuh give up an and just say yes.” But, a script formula-
tion may have additional interactional and rhetorical relevance. It may also index
a ‘relationship issue’ that is worthy of subsequent therapeutic work (Muntigl 2004a,
2004b); that is, it may sequentially implicate a therapist response, such as a thera-
pist formulation, that in some way addresses the client’s complaint.
Now the therapist’s next turn does indeed contain a formulation, but if we
examine his turn more closely, it should be apparent that there is more interac-
tional work going on than a ‘bare’ formulation might indicate. For instance, the
therapist does not immediately launch into a restatement of what Wendy had said.
Instead, his formulation is prefaced by an account or justification. For this reason,
lines 19–25 have been analyzed as a Prefacing move that is separate from the core
formulation found in lines 26–33. The breakdown of moves contained within the
therapist’s turn is shown in example (4).
(4) Pre-formulation
Noticing T: .hh what I’m starting tuh see here is a pattern uh um.hh
Account as a couples therapist um I’m always looking for pat-
terns?.hh that people get into that they get stuck in.hh
Solicit and I’m I need your agreement
as tuh whether or not what we’re seeing here is this par-
ticular pattern.hh umm which is leading tuh the kinda
communication that (both) you’re talking about.hh
Formulation your experience Wendy Sue.hh is of uh living with a
man who has a lecturing style
Agree W: yes
Formulation T: who talks to you cons-
excuse me if I I’m relating back in her ways.hh ahh
who who lectures you constantly.hh
who goes on about things bove and beyond the point
Grammar 

of being [resolved ] when they’re already resolved

Agree W: [yes ]
Agree W: yes
We identified three acts within the Pre-formulation. These included noticing, ac-
count and solicit. The term ‘noticing’ derives from Schegloff (1988). One of the
main features of noticings is that they tend to create an explanation slot
(see Antaki 1994). For instance, noticing a stain on someone’s shirt may implicitly
call for an explanation on how the stain got there. So, noticing something about an
addressee may also put the onus on the addressee to provide an account. However,
the burden of explanation may also be put on the speaker. Returning to the thera-
pist, by noticing (i.e., seeing) a pattern, it would seem that the expectation is placed
on the therapist, not the client, to explain the relevance of this noticing. We suggest
two reasons for this. The first is that the therapist’s stance to what he is noticing is
explicitly subjective; that is, he formulates it as “I’m starting tuh see…”, rather than
beginning the clause with “you” as in “you seem to be saying that…,” which would
shift the ‘responsibility’ of what was said more on Wendy. The second, and more
powerful, reason is the therapist’s mentioning of a pattern. In her work on pre-
announcement sequences, Terasaki (2004:€199) argued that ‘thing-like’ nominals
such as “something” or “thing” have a projective quality (e.g., “I got sumpn thet’s
wi::ld”, “I got two good things”). They sequentially implicate a step-wise delivery of
news such that the “something” will become specified later in the interaction. In a
similar way, by claiming to have noticed a pattern, the therapist projects further
talk that will not only explain the pattern, but will also serve as an account for hav-
ing noticed the pattern in the first place.
Immediately after the noticing, the therapist does two things. First, he ac-
counts for having noticed the pattern by linking his institutional role of being a
couples therapist with activities that therapists ‘always’ perform (i.e., couples ther-
apists look for patterns). Second, he expresses the need to solicit agreement from
the couple that his upcoming interpretation of the pattern will be correct. Note
that towards the end of the pre-formulation, the therapist begins to shift from an
“I” to a “We” perspective (i.e., “whether or not what we’re seeing here…”). Through
this subtle shift of reference, the therapist is able to emphasize the importance of
achieving consensus as a prerequisite for any future therapeutic work. In other
words, it is not enough for the therapist to have noticed the pattern. The couple
must also ‘share’ his vision.
In the final and ‘core’ part of the therapist’s turn, a candidate reading is given
to Wendy’s extended complaint, which appears in the form of a gist or summary.
Although the initial part of the formulation tends to downplay Wendy’s extreme
 Peter Muntigl and Eija Ventola

assessments by equating Fred’s behaviour with a lecturing style, meanings of ex-

treme persistence and frequency do get reintroduced later when the therapist em-
phasizes the manner in which Fred lectures (e.g., “lectures you constantly” and
“goes on about things bove and beyond the point…”). The selections and modifica-
tions that the therapist makes, however, pertain largely to the area of ideational
meaning and it is to this metafunction to which we now turn.

4.2 Ideational meaning

With this metafunction, an SFL analysis begins to part company with CA. In an
ideational analysis, specific attention is given to what is called the transitivity struc-
ture of each of the clauses that comprise a turn and the conjunctive meanings that
link clauses together (Halliday 1994, 2005[1967, 1968]). This essentially means
identifying the conjunctive, verbal, nominal, adverbial and prepositional elements
of each clause. For analyzing social interaction, however, the point of doing this
kind of analysis is not simply to identify how often a certain verb, noun, preposi-
tion, etc. occurs. Instead, these grammatical structures are analyzed for their func-
tions and for their ‘positional sensitivity’ or ‘function within the exchange’.
Within ideational grammar, formulations are interpreted in a way that bears simi-
larity to CA. They are rhetorical/conjunctive relations that function to summarize
and/or clarify what a speaker had formulated in a previous turn (Muntigl 2004a, 2007).
If we turn now to the specifics of the therapist’s formulation, two linguistic resources
play an especially important function in summarizing Wendy’s meanings. The first
corresponds to relational verbs (e.g., ‘to be’, ‘exemplify’, ‘represent’, ‘equals’) in which
two entities are placed in an identifying relationship (‘x’ = ‘y’). The second is T’s use of
grammatical metaphor (Halliday 1987, 1994, 1998), in which the client’s construal of
experience becomes grammatically more abstract (e.g., turning a Process into a Par-
ticipant; that is, a verb becoming a noun or adjective).
One significant difference between the speakers’ use of ideational resources
involves process (i.e., verb) type. Whereas Wendy construes experience in terms of
action or behavioural verbs (e.g., to lecture, go on and on, pressed his point), the
therapist construes experience in terms of relational verbs (e.g., “is”). The rela-
tional verb allows him to equate two ‘thing’-like phenomena. For example, the
therapist equates his experience with “a pattern” (line 19), everyone’s experience
with “this particular pattern” (line 23) and Wendy’s experience with a number of
Fred’s behaviours (lines 26–33). From the ‘x’ side of the relational clause equation,
the therapist moves from ‘my experience’ to ‘our experience’ to ‘Wendy’s experi-
ence’. On the ‘y’ side, Wendy’s prior formulation moves from ‘a pattern’ to ‘this
particular pattern’ to ‘the features of the pattern’. The summarizing component of
the Therapist’s formulation consists of his narrowing down what Wendy had
Grammar 

meant/experienced; first, her recount of Fred’s lecturing is identified as a particu-

lar pattern and, second, only selected elements of what Wendy had said are in-
cluded in the pattern.

what I’m starting tuh see here is a pattern

what we’re seeing here is this particular pattern
your experience Wendy Sue is of uh living with a man who has a lecturing style who
talks to you cons-
who who lectures you constantly.hh
who goes on about things bove and beyond the point of
being resolved when they’re already resolved

‘X’ = ‘Y’

Another important aspect of the formulation includes how the client’s talk be-
comes modified. In particular, the therapist does not simply repeat what Wendy
had said; instead, he reworks her wordings by placing some wordings in a new
grammatical context. For example, “he likes to lecture?...” in line 07 (repeated in
line 09) is transformed into “…a man who has a lecturing style.” Grammatically,
Wendy’s construal of Fred as doing lecturing is changed into ‘lecturing’ as being an
attribute of Fred. This type of transformation is commonly referred to as gram-
matical metaphor (Halliday 1994; 1998). Note, however, that the therapist does
unpack lecturing style later on in his turn, reinstating lecturing to its original dy-
namic interpretation of “lectures you constantly” and “goes on about things.” But,
as will be discussed shortly, ‘lecturing style’ in its metaphorical attributive form
retains a high level of interactional relevance and has significant implications for
the ensuing conversation.
To sum up, the therapist’s use of ideational grammar serves a number of inter-
active functions. First, the relational verb structures provide a recurring frame for
much of the summarizing work performed by the therapist. Second, transitivity
resources such as grammatical metaphor are used to modify Wendy’s original de-
scription of Fred (i.e., lecturing → lecturing style). Third, the unique clausal con-
structions help to organize the therapist’s noticing and subsequent formulation as
a newsworthy event. This last function, however, forms part of the textual
metafunction and will be discussed in more detail below.

4.3 Textual meaning

Examining interaction from the perspective of the textual metafunction involves

identifying thematic progressions and the location of new information within and
across different discourse units. In the English language a theme occupies the ‘first
 Peter Muntigl and Eija Ventola

position’ of the clause and the new information in speech tends to realized via in-
tonation (see Halliday 1994:€296). Thematic progressions can be examined for any
given conversational sequence if we identify the themes within speaker turns. The
topical themes (i.e., the ideational component of the theme) contained within
Wendy’s script formulation and the therapist’s formulation are shown in Table€2
(lines 1–33).
By examining the topical themes of Wendy’s clauses, it immediately becomes
apparent that her husband is most often realized in theme position. In the major-
ity of cases, Wendy’s husband serves as the point of departure for her assessments.
For Wendy, thematic development does not consist of a change in her ‘point of
departure’, but of consistently maintaining the same point of departure throughout
her turn. In this way, she is able to build up a gradually increasing negative assess-
ment of her husband.
The topical themes of the Therapist’s formulation mark a different kind of pro-
gression. No longer is Fred the point of departure of the message. Instead, ‘seeing/
experiencing’ or ‘the therapist’s perspective’ is commonly made the theme. In oth-
er words, through these new themes the therapist is able to re-direct the trajectory
of the talk such that what the clients and the therapist ‘see’ or ‘experience’ is placed
under negotiation. Relational clauses, especially thematic equatives (or pseudo-cleft
sentences) of the form “what I’m starting tuh see here is…”, are often used by
speakers for thematic purposes (Halliday 1994:€42).
We now turn to the aspect of new information or, more specifically, the way in
which the therapist orients to the general issue of what is newsworthy. First of all,
we should point out that much of Wendy’s script orients to the ‘typicality’ or ‘non-
news’ character of Fred’s behaviour; that is, Fred has ‘always’ lectured and, from
Wendy’s perspective, this is not ‘front page headline’ material. But, what is deemed
unremarkable for the speaker may, by contrast, be highly newsworthy for the ad-
dressee. First, and perhaps most obvious, is the therapist’s mention of a pattern. In

Table 2.╇ Wendy’s and Therapist’s Topical Themes

Wendy’s Topical Themes Therapist’s Topical Themes

Fred, he, Fred, he, he, he (ellipted), what I’m starting tuh see here, as a couples
there, I, it, Fred, I, he, he therapist, I, what we’re seeing here, which, your
experience, I

this way, the therapist announces that he has identified something therapeutically
relevant; that is, his noticing constitutes a new way of looking at Fred’s lecturing,
not merely as ‘typical’ behaviour but something that is deserving of more attention
within a therapeutic context. Further, the therapist’s noticing of a pattern also
Grammar 

projects future talk that specifies the news, in the similar way that pre-announce-
ment sequences do (Terasaki 2004). The pattern, therefore, will need to be ‘un-
packed’ and elaborated upon. Newsworthiness is also realized in the verbal group
“starting to see”; that is, by explicitly stating that he is “starting to…”, the therapist
implies that he has just, at that moment, come to see/notice/realize something that
was not available to him before. The newsworthiness, therefore, comes from his
sudden ability to make an important, therapeutically relevant, inference.

5. Grammar influencing the trajectory of the interaction

Our grammatical analysis has selectively focused on how interpersonal, ideational

and textual meanings develop and become negotiated over the course of a ‘script
formulation’-(re)formulation exchange. Interpersonally, the negotiation of moves
consisted of the therapist restating Wendy’s extended complaint and Wendy agree-
ing with the restatement. Although the restatement did maintain Wendy’s initial
evaluative meanings to a degree, it was noted that the therapist initially formulated
Fred’s behaviour as a lecturing style. With this formulation the original meanings
of usuality and inclination become lost, leaving only the attributive meaning of
‘lecturing’. Ideationally, construals of experience moved from Wendy using ‘doing’
processes/verbs to construe Fred’s lecturing behaviour to the Therapist’s formula-
tions in which he used relational verbs to equate what he or the clients ‘see’ or
‘experience’ with ‘a pattern’, ‘a lecturing style’ or features of the lecturing style. And
textually, the consistent thematic focus on Fred within Wendy’s turn changed to a
focus on ‘seeing’ or ‘experiencing’ in the Therapist’s formulation. New information
was also indexed by the therapist as a noticing, which also signalled the projection
of upcoming therapeutic work that will elaborate on the clients’ ‘pattern’. In sum,
by focusing on the grammatical resources used by each speaker on a clause-by-
clause basis, we were able to map out the trajectory of the exchange and the subtle
shifts in interpersonal, ideational and textual meanings as the reformulation se-
quence unfolded.
There are, however, certain grammatical shifts that seem to play a decisive role
in influencing the trajectory of the conversation. From Schegloff ’s (1992:€111) per-
spective this would mean that certain grammatical resources have specific rele-
vance or are procedurally consequential for the ensuing talk. A powerful example
of a salient or relevant grammatical transformation involves the therapist’s
reformulation of Wendy’s “he likes to lecture” into “a lecturing style”. The relevance
of this example does not merely lie in the observation that lecture has changed
from a behavioural process to an attribute. If we look further down in the example
on lines 44–46, we notice that the therapist now uses “this lecturing style” as the
 Peter Muntigl and Eija Ventola

nominal group (i.e., participant) of a clause (“what are y- impact or effect.hh does
this lecturing style that you experience from Fred what does what impact does that
have on you.hh as a person”). As a result “this lecturing style” (also realized by
“that” on line 46) becomes reconstrued as an Agent that impacts on/affects Wendy.
Reformulating “lecture” into “a lecturing style”, therefore, creates certain interac-
tional possibilities for both the therapist and Wendy. For this therapist nominal-
izing “lecture” allowed him to construe lecturing as an agentive participant that
influences Wendy’s life. This unfolding activity of reformulating problems for the
purpose of exploring their causal influences has been well documented in Muntigl
(2004a, 2004b). Furthermore, this specific therapist activity relates to what is
known in narrative therapy as externalizing the problem in order to unpack nega-
tive identity conclusions (for overviews of narrative therapy see White and Epston
1990; White 2001). If we view this process grammatically, it transforms Wendy’s
formulation of ‘Fred lectures’ into simply ‘lecturing’, leaving the doer of the lectur-
ing unspecified – note that the therapist does not say “Fred’s lecturing style”, which
would merely have been a nominalized version of what Wendy had said before. So,
by ‘de-thematizing’ Fred, the therapist is able to uncouple the strong association
that Wendy makes between Fred and his lecturing behaviour. The behaviour
(i.e., the lecturing) rather than the person is placed into the foreground.
The practices of nominalizing, construing experience through ‘causal’
grammatical structures and using relational verbs to connect up different ‘par-
ticipants’ also plays a significant part in linking different domains of the cli-
ents’ experience (see also Peräkylä 2004). These practices also helped to pro-
duce a substantially expanded ‘grammar of lecturing’. In order to show how
domains of experience and grammatical realization of lecturing are mutually
constitutive of each other, let us first consider the way in which Wendy’s initial
script formulation was entirely focussed on Fred. It should be noted, however,
that although Wendy clearly was the target of Fred’s lecturing, this was never
expressed explicitly as such. This can be seen from if we take a sample of
Wendy’s clauses:
a. he likes to lecture? (1.2) on any: any subject
b. he likes to lecture and and go on and on and on and on about it
Notice that there is no ‘complement’ to the verb ‘lecture’; that is, Wendy does not say
“he likes to lecture me…”. Fred’s lecturing, therefore, is formulated as a general ac-
tivity, done to various people (including Wendy) in various situations. So, although
we can infer that Wendy construes Fred as a lecturer, the link to herself as the
repeated target of the lecturing is often not made explicit (with the exception of the
‘I thought X, but Y happened’ trope she provides towards the end of her script).
Grammar 

One of the things that the therapist does in the subsequent turn is tighten up
this link between Wendy’s experience and Fred’s behaviour. Consider the follow-
ing formulations made by the therapist:
a. your experience Wendy Sue.hh is of uh living with a man who has a lecturing
b. who who lectures you constantly
In clause ‘a’, an identifying relationship is made between Wendy’s experience and
the lecturing, and in clause ‘b’, the complement of lecturing is specified (“lectures
you…”). By placing the focus on Wendy’s experience and by specifying Wendy as
the target of Fred’s lecturing, the therapist personalizes Wendy’s domain of experi-
ence. This means altering descriptions of Fred as a general lecturer, to one who
lectures Wendy specifically.
As a last step, the therapist provides causal connections between ‘the lecturing’
and Wendy’s personal domains of experience on which the lecturing was having a
negative impact; in this way, the meaning and relevance of the lecturing becomes
expanded in that it now also speaks to her feelings of self-worth or self-esteem
(e.g., “it makes me feel like uh a child”; “it makes me feel like I’m ignorant”).

6. Conclusions

One of the guiding principles that both SFL and CA share in common is the insist-
ence that any analysis of social interaction include an examination of speakers’
meaning making resources used to (co-)construct the interaction. In this article,
we have attempted to draw attention to some grammatical resources involved in
meaning making and to show how links can be made between speakers’ gram-
matical selections and the kinds of interactional business these selections can ac-
complish. By taking stock of grammatical selections that are drawn on by a thera-
pist and a client, we were able to develop a ‘grammatical profile’ of the therapeutic
activity and to ascertain the therapist’s professional stock of interactional knowledge
(see Peräkylä and Vehviläinen 2003). This did not simply mean identifying differ-
ent nouns or verbs or counting grammatical structures to see how frequently they
occur. Instead, it meant locating grammatical structures within sequences and ac-
counting for their positional sensitivity. It meant identifying what interactional
business the grammatical selections perform and what kinds of meanings, actions
and activities they projected in the unfolding conversation.
In the course of this paper, we tried to foreground different aspects of the
grammar (interpersonal, ideational and textual) in our analysis. This was done in
 Peter Muntigl and Eija Ventola

the tradition of SFL research by examining the functions of grammatical struc-

tures. Halliday (1994: xix) summarizes this position as follows:
Since the relation of grammar to semantics is in this sense natural, not arbitrary,
and since both are purely abstract systems of coding, how do we know where the
one ends and the other begins? The answer is we don’t: there is no clear line be-
tween semantics and grammar, and a functional grammar is one that is pushed in
the direction of the semantics.

We might say that we tried to go even further than Halliday by pushing a gram-
matical analysis in the direction of a sequential analysis. Proceeding along the
lines of Schegloff (1996), we were also interested in discovering the grammar’s
positional sensitivity within specific sequential contexts and, more specifically,
what kinds of interactional work the grammar is capable of doing. But in order to
do this, we still need to maintain our focus on the functional grammatical units
that speakers deploy. The challenge for interaction analysts, as we see it, is to oper-
ate at both ends (i.e., grammar & interaction) and to make the appropriate links,
where relevant.


Antaki, Charles. 1994. Explaining and Arguing. London: Sage.

Antaki, Charles. 2008. “Formulations in psychotherapy.” In Conversation Analysis and Psycho-
therapy, Anssi Peräkylä, Charles Antaki, Sanna Vehviläinen and Ivan Leudar (eds), 26–42.
Cambridge: Cambridge University Press.
Antaki, Charles, Barnes, Rebecca and Leudar, Ivan. 2005. “Diagnostic formulations in psycho-
therapy.” Discourse Studies 7 (6): 627–647.
Antaki, Charles, Diaz, Félix and Collins, Alan. 1996. “Keeping your footing: Conversational
completion in three-part sequences.” Journal of Pragmatics 25: 151–171.
Antaki, Charles and Widdicombe, Sue (eds.). 1998. Identities in Talk. London: Sage.
Barth-Weingarten, Dagmar. 2008. “Interactional linguistics.” In Handbook of Interpersonal Com-
munication [Volume 2 of Handbook of Applied Linguistics], Gerd Antos, Eija Ventola and
Tilo Weber (eds), 53–76. Berlin: Mouton de Gruyter.
Beach, Wayne. 1993. “Transitional regularities for ‘casual’ “okay” usages.” Journal of Pragmatics
19: 325–352.
Buttny, Richard. 1996. “Clients’ and therapists joint construction of the clients’ problem.” Re-
search on Language and Social Interaction 29 (2): 125–53.
Clayman, Steven. 1992. “Footing in the achievement of neutrality: The case of news-interview
discourse.” In Talk at Work, Paul Drew and John Heritage (eds), 163–198. Cambridge:
Cambridge University Press.
Coulter, Jeff. 1985. “Two concepts of the mental.” In The Social Construction of the Person,
Kenneth J. Gergen and Keith E. Davis (eds), 129–144. New York: Springer Verlag.
Coulter, Jeff. 1989. Mind in Action. Atlantic Highlands, NJ: Humanities Press International.
Coulter, Jeff. 1999. “Discourse and mind.” Human Studies 22: 163–181.
Grammar 

Davis, Kathy. 1986. “The process of problem (re)formulation in psychotherapy.” Sociology of

Health and Illness 8: 44–74.
Dik, Simon. 1997. The theory of functional grammar, Part 1. The structure of the clause, 2nd Edn.
Ed. Kees Hengeveld. Berlin: Mouton de Gruyter.
Edwards, Derek. 1994. “Script formulations: A study of event descriptions in conversation.”
Journal of Language and Social Psychology 13: 211–247.
Edwards, Derek. 1995. “Two to tango: Script formulations, dispositions, and rhetorical symmetry
in relationship troubles talk.” Research on Language and Social Interaction 28(4): 319–350.
Edwards, Derek. 1997. Discourse and Cognition. London: Sage.
Edwards, Derek and Potter, Jonathan. 1992. Discursive Psychology. London: Sage.
Edwards, Derek and Potter, Jonathan. 2005. “Discursive psychology, mental states and descrip-
tion.” In Conversation and Cognition, Hedwig te Molder and Jonathan Potter (eds), 241–
259. Cambridge: Cambridge University Press.
Gardner, Rod. 1997. “The conversation object mm: A weak and variable acknowledging token.”
Research on Language and Social Interaction, 30(2): 131–156.
Gardner, Rod. 2001. When Listeners Talk. Amsterdam: John Benjamins.
Goodwin, Charles. 1981. Conversational Organization: Interaction between Speakers and Hear-
ers. New York: Academic Press.
Goodwin, Charles. 1986. “Between and within: Alternative sequential treatments of continuers
and assessments.” Human Studies 9: 205–217.
Hak, Tony and de Boer, Fijgje. 1996. “Formulations in first encounters.” Journal of Pragmatics 25:
Halliday, M. A. K. 1975. Learning How to Mean. London: Edward Arnold.
Halliday, M. A. K. 1978. Language as a Social Semiotic. London. Edward Arnold.
Halliday, M. A. K. 1994. An Introduction to Functional Grammar (2nd ed.). London: Edward
Halliday, M. A. K. 1998. “Things and relations: Regrammaticising experience as technical
knowledge.” In Reading Science: Critical and Functional Perspectives on Discourses of Science,
James R. Martin and Robert Veel (eds), 185–235. London: Routledge.
Halliday, M. A. K. 2002 [1987]. “Spoken and written modes of meaning.” In On Grammar [Vol. 1
in the Collected Works of M.A.K. Halliday], Jonathan Webster (ed), 323–351. London:
Halliday, M. A. K. 2003. On Language and Linguistics [Vol. 3 in the Collected Works of M.A.K.
Halliday], Jonathan Webster (ed). London: Continuum.
Halliday, M. A. K. 2005 [1967–68]. “Notes on transitivity and theme in English – Parts 1–3.” In
Studies in English Language [Vol. 7 in the Collected Works of M.A.K. Halliday], Jonathan
Webster (ed). London: Continuum.
Halliday, M. A. K. and Hasan, Ruqaiya. 1985. Language, Context and Text: Aspects of Language
in a Social Semiotic Perspective. Geelong, Vic: Deakin University Press.
Halliday, M. A. K. and Matthiessen, Christian M. I. M. 1999. Construing Experience through
Meaning: A Language-Based Approach to Cognition. London: Cassell.
Hasan, Ruqaiya. 1987. “The grammarian’s dream: Lexis as most delicate grammar.” In New
Developments in Systemic Linguistics: Theory and Description, M. A. K. Halliday and Robin
Fawcett (eds), 184–212. London: Pinter.
Heritage, John. 1984a. Garfinkel and Ethnomethodology. Cambridge: Polity Press.
 Peter Muntigl and Eija Ventola

Heritage, John. 1984b. “A change-of-state token and aspects of its sequential placement.” In
Structures of Social Action, J. Maxwell Atkinson and John Heritage (eds), 299–345. Cam-
bridge: Cambridge University Press.
Heritage, John. 1998. “Oh-prefaced responses to inquiry.” Language in Society 27: 291–334.
Heritage, John and Raymond, Geoffrey. 2005. “The terms of agreement: Indexing epistemic au-
thority and subordination in talk-in-interaction.” Social Psychology Quarterly 68 (1): 15–38.
Heritage, John and Roth, Andrew. 1995. “Grammar and institution: Questions and questioning
in the broadcast news interview.” Research on Language and Social Interaction 28(1): 1–60.
Heritage, John and Sorjonen, Marja-Leena. 1994. “Constituting and maintaining activities across
sequences: And-prefacing as a feature of question design.” Language in Society 23: 1–29.
Heritage, John and Watson, Rod 1979. “Formulations as conversational objects.” In Everyday
Language: Studies in Ethnomethodology, George Psathas (ed.), 123–162. New York: John
Wiley & Sons.
Holt, Elizabeth and Clift, Rebecca (eds.). 2007. Reporting Talk: Reported Speech in Interaction.
Cambridge: Cambridge University Press.
Jefferson, Gail. 1984. “Notes on a systematic deployment of the acknowledgement tokens ‘yeah’
and ‘mm hm’.” Papers in Linguistics 17(1–4): 197–216.
Jefferson, Gail. 2004. “At first I thought.” In Conversation Analysis: Studies from the First Genera-
tion, Gene Lerner (ed), 131–167. Amsterdam: John Benjamins.
Lerner, Gene. 1991. “On the syntax of sentences in progress.” Language in Society 20: 441–458.
Lerner, Gene. 2004a. “Introduction.” Research on Language and Social Interaction 37 (2): 93–94.
Lerner. Gene. 2004b. “On the place of linguistic resources in the organization of talk-in-interac-
tion: Grammar as action in prompting a speaker to elaborate.” Research on Language and
Social Interaction 37 (2): 151–184.
Lynch, Michael and Bogen, David. 2005. “‘My memory has been shredded’: A non-cognitivist
investigation of ‘mental’ phenomena.” In Conversation and Cognition, Hedwig te Molder
and Jonathan Potter (eds), 226–240. Cambridge: Cambridge University Press.
Martin, James R. 1992. English Text. Amsterdam: John Benjamins.
Martin, James R. 2000. “Beyond exchange: Appraisal systems in English.” In Evaluation in Text:
Authorial Stance and the Construction of Discourse, Susan Hunston and Geoff Thompson
(eds), 142–175. Oxford: Oxford University Press.
Muntigl, Peter. 2004a. Narrative Counselling: Social and Linguistic Processes of Change. Amster-
dam: John Benjamins.
Muntigl, Peter. 2004b. “Ontogenesis in narrative therapy: A linguistic-semiotic examination of
client change.” Family Process 43 (1): 105–124.
Muntigl, Peter. 2007. “A metapragmatic examination of therapist reformulations.” In Metapragmat-
ics in Use, Wolfram Bublitz and Axel Hübler (eds), 235–262. Amsterdam: John Benjamins.
Muntigl, Peter. 2009. “Knowledge Moves in Conversational Exchanges: Revisiting the Concept
of Primary vs. Secondary Knowers.” Functions of Language 16 (2): 225–263.
Muntigl, Peter and Hadic Zabala, Loreley. 2008. “Expandable answers: How clients get prompted to
say more during psychotherapy.” Research on Language and Social Interaction 41 (2): 187–226.
Ochs, Elinor, Schegloff, Emanuel A. and Thompson, Sandra (eds). 1996. Interaction and Gram-
mar. Cambridge: Cambridge University Press.
Peräkylä, Anssi. 2004. “Making links in psychoanalytic interpretations: A conversation analytic
perspective.” Psychotherapy Research 14 (3): 289–307.
Peräkylä, Annsi and Vehviläinen, Sanna. 2003. “Conversation analysis and the professional
stocks of interactional knowledge.” Discourse and Society 14 (6): 727–750.
Grammar 

Potter, Jonathan. 1996. Representing Reality. London: Sage.

Potter, Jonathan. 1998. “Cognition as context (whose cognition?).” Research on Language and
Social Interaction 31: 29–44.
Prevignano, Carlo and Thibault, Paul. 2003b. “Continuing the interview with Emanuel A.
Schegloff.” In Discussing Conversation Analysis: The Work of Emanuel Schegloff, Carlo Pre-
vignano and Paul Thibault (eds), 165–171. Amsterdam: John Benjamins.
Raymond, Geoffrey. 2003. “Grammar and social organization: Yes/no interrogatives and the
structure of responding.” American Sociological Review 68 (6): 939–967.
Raymond, Geoffrey. 2004. “Prompted action: The stand-alone “so” in ordinary conversation.”
Research on Language and Social Interaction 37 (2): 185–218.
Roth, Andrew L. 2002. “Social Epistemology in Broadcast News Interviews.” Language in Society
31 (3): 355–381.
Ryle, Gilbert. 1949. The Concept of Mind. London: Hutchinson’s University Library.
Sacks, Harvey. 1992. Lectures on Conversation (2 Vols.). Oxford, England: Basil Blackwell.
Sacks, Harvey, Schegloff, Emanuel A. and Jefferson, Gail. 1974. “A simplest systematics for the
organization of turn-taking for conversation.” Language 50 (4): 696–735.
Schegloff, Emanuel A. 1982. “Discourse as an interactional achievement: Some uses of ‘uh huh’
and other things that come between sentences.” In Analyzing Discourse: Text and Talk.
Georgetown University Roundtable on Languages and Linguistics, Deborah Tannen (ed),
71–93. Washington, DC: Georgetown University Press.
Schegloff, Emanuel A. 1984. “On some questions and ambiguities in conversation.” In Structure
of Social Action, J. Maxwell Atkinson and John Heritage (eds), 28–56. Cambridge:
Cambridge University Press.
Schegloff, Emanuel A. 1988. “Goffman and the analysis of conversation.” In Exploring the Inter-
action Order, Paul Drew and Anthony Wooton (eds), 89–135. Boston: Northeastern
University Press.
Schegloff, Emanuel A. 1992. “On talk and its institutional occasions.” In Talk at Work: Interac-
tion in Institutional Settings, Paul Drew and John Heritage (eds), 101–134. Cambridge:
Cambridge University Press.
Schegloff, Emanuel A. 1996. “Turn organization: One intersection of grammar and interaction.”
In Interaction and Grammar, Elinor Ochs, Emanuel A. Schegloff and Sandra Thompson,
(eds), 52–133. Cambridge: Cambridge University Press.
Schegloff, Emanuel A. 1997. “Who’s Text? Whose Context?” Discourse & Society 8 (2): 165–87.
Schegloff, Emanuel A. 2004. “On Dispensability.” Research on Language and Social Interaction
37 (2): 95–149.
Schegloff, Emanuel. A. 2007. Sequence Organization in Interaction: A Primer in Conversation
Analysis, Volume 1. Cambridge: Cambridge University Press.
Selting, Margret. 2008. “Linguistic resources for the management of interaction.” In Handbook
of Interpersonal Communication [Volume 2 of Handbook of Applied Linguistics], Gerd
Antos, Eija Ventola and Tilo Weber (eds), 217–253. Berlin: Mouton de Gruyter.
Streeck, Jürgen. 2002. “Grammars, words, and embodied meanings: On the uses and evolution
of so and like.” Journal of Communication 52: 581–596.
Terasaki, Alene. 2004. “Pre-announcement sequences in conversation.” In Conversation Analysis:
Studies from the First Generation, Gene Lerner (ed), 171–223. Amsterdam: John Benjamins.
Thompson, Geoff and Muntigl, Peter. 2008. “Systemic functional linguistics: An interpersonal
perspective.” In Handbook of Interpersonal Communication [Volume 2 of Handbook of
 Peter Muntigl and Eija Ventola

Applied Linguistics], Gerd Antos, Eija Ventola and Tilo Weber (eds), 107–132. Berlin:
Mouton de Gruyter.
Thompson, Sandra and Couper-Kuhlen, Elizabeth. 2005. “The clause as a locus of grammar and
interaction.” Discourse Studies 7 (4–5): 481–505.
Turk, Monica. 2004. “Using and in conversational interaction.” Research on Language and Social
Interaction 37 (2): 219–250.
van Valin, Robert and LaPolla, Randy. 1997. Syntax: Structure, meaning and function. Cambridge:
Cambridge University Press.
Vendler, Zeno. 1967. Linguistics in Philosophy. Ithaca, NY: Cornell University Press.
Ventola, Eija 1984. “Orientation to Social Semiotics in Foreign Language Teaching.” Applied
Linguistics 5 (3): 275–286.
Ventola, Eija. 1987. The Structure of Social Interaction: A Systemic Approach to the Semiotics of
Service Encounters. London: Pinter.
White, Michael. 2001. “Narrative practice and the unpacking of identity conclusions.” Gecko: A
journal of deconstruction and narrative practice 1: 1–17.
White, Michael and Epston, David. 1990. Narrative Means to Therapeutic Ends. New York:
Wittgenstein, Ludwig. 1958. Philosophical Investigations (2nd ed.). Oxford: Blackwell.
Researching intercultural communication
Discourse tactics in non-egalitarian contexts

Angel Lin
Faculty of Education, The University of Hong Kong

In this chapter key sociological traditions forming the theoretical backdrop of

current discourse-based approaches to intercultural communication research will
be discussed and John Gumperz’s contribution to highlighting the interactional
nature of everyday communication and language use will be outlined. Then I
shall introduce the central thesis of this chapter: that discourse-based approaches
to intercultural communication provide helpful frameworks for understanding
how power is fluid and mediated through discourse and meaning-making, and
how different social actors located in differential, hierarchical social positions,
and coming from different cultural backgrounds, can negotiate through
discourse for more advantageous positions for themselves. This thesis will
then be delineated through drawing on positioning theory, (Davies and Harré,
1990; Harré and Langenhove, 1999), a discourse-based social identity theory, to
analyse two examples of intercultural/inter-group communication.

1. Interaction analysis and discourse-based approaches

to intercultural communication

What is our conception of interaction analysis and on what notion of interaction

can it be based? It seems that the notion of interaction cannot be essentially de-
fined. A broad range of phenomena or activities can be seen as interaction by dif-
ferent people engaged in different forms of life (Wittgenstein’s notion, see Sluga
and Stern, 1996). At one extreme of the continuum, any human (some would also
argue machine) meaning-making activity can be seen as a form of interaction.
One can conceive someone finding a fruitful way of seeing “reading” as a form of
interaction (the reader making meaning of/interacting with the text and indirectly
interacting with the invisible/non-physically present author; or to stretch the ar-
gument a bit, the computer “reading” some input and making certain responses to
the input) (similar arguments can be made for watching TV, movies or watching
 Angel Lin

an exhibition). However, stretching the notion to that far end will not be too useful
for the practical linguistic anthropologist interested in everyday human interac-
tion. I shall therefore focus my analysis on the range of activities that involve some
form of bi- or multi-party, face-to-face meaning-making, which is embedded in
some shared forms of life or ways of living engaged in by the interactants. And
“face-to-face” is to be understood broadly, i.e., can be mediated via some form of
technology, e.g., phone talk, net talk, e-mail talk, etc.
Interaction analysis thus has as its aim the uncovering of the kinds and nature
of the meaning-making, interpretive processes involved and the semiotic resourc-
es drawn upon to enable the achievement of some mutual sense of inter-subjectiv-
ity (i.e., the perception on both/all parties that they achieve the sharing of certain
perspectives with each other/one another). How is this sense of inter-subjectivity
achieved? What is happening when this is not achieved (e.g., in cases of perceived
communication barriers or breakdowns)? What is it that can bring about the over-
coming of the communicative barriers or breakdowns?
Under an interactional conception of language, language should not be seen as
a reified object of study by linguists and language as a bounded concept is an ideo-
logical, theoretical and social construct – born of the activities of armchair lin-
guists and/or political, national unifying/segregating agendas. The analytical focus
should be on how languages as (continuously changing) systems of semiotic re-
sources (among other semiotic systems of resources) are recruited and utilized for,
and at the same time also transformed, during interaction.
While the above brief summary will be familiar to those working in the inter-
pretive traditions of discourse analysis, scholars working in the broader field of
communication and/or intercultural communication might, however, need a brief
introduction to discourse-based approaches. In the next section, key sociological
traditions forming the theoretical backdrop of current discourse-based approach-
es to intercultural communication research will be discussed and John Gumperz’s
contribution to highlighting the interactional nature of everyday communication
and language use will be outlined. Then I shall introduce the central thesis of this
chapter: that discourse-based approaches to intercultural communication provide
helpful frameworks for understanding how power is fluid and mediated through
discourse and meaning-making, and how different social actors located in differ-
ential, hierarchical social positions, and coming from different cultural back-
grounds, can negotiate through discourse for more advantageous positions for
themselves. This thesis will then be delineated through drawing on positioning
theory, (Davies and Harré, 1990; Harré and Langenhove, 1999), a discourse-based
social identity theory, to analyse two examples of intercultural/inter-group com-
Researching intercultural communication 

2. Symbolic interactionism (SI), structuration

theory, and discourse-based approaches

Symbolic interactionism (SI) or that branch of sociology that focuses on human

meaning-making and interpretive processes evolving around the use of symbols,
or semiotic resources, has its roots in the pragmatist philosophers such as John
Dewey, Charles Horton Cooley, and George Herbert Mead. The SI perspective
puts an emphasis on human interaction and communication via the use of sym-
bols for meaning-making, and human interpretive processes which are central to
interaction and communication. SI studies the interaction order of everyday life
and focuses on the social, interactional, and discursive construction of self and
other. Concepts such as power, social relations, contexts, self, and identities are
seen as fluid, always open to negotiation and re-negotiation, and interactively co-
constructed via discourse and other semiotic resources. In sum, the SI perspective
emphasizes human interaction and communication as mediated by the use of
symbols, by interpretation, or by ascertaining the meaning of one another’s actions
(Blumer, 1986).
While Anthony Giddens seems to have developed structuration theory
(Giddens, 1984) quite independently of SI, structuration theory and SI are com-
patible with each other. The SI perspective sees people as active social agents, quite
different from the solitary, rational, Cartesian individual (or subject). People are
seen as social actors – constantly actively adjusting, interpreting, and organizing
and re-organizing their ways of speaking and their ways of being (e.g., ways of
dressing, looking, thinking, viewing, feeling, interpreting, hearing, etc.) to adjust
to others in social interactions. The self is created through such on-going social
interactions, and it is a self that is fluid, and constantly negotiating with and adjust-
ing to others. The SI concern is with how the social order (macro forces and struc-
tures) is constantly being created, reproduced, or contested and transformed.
Similarly, structuration theory provides a solution to overcome the sociologi-
cal macro-micro, structure-agency theoretical divide by seeing the macro and mi-
cro, social structures and agency, as mutually constitutive and shaping. Giddens
(1984) thus attempts to provide an overall theoretical framework to deal with two
major sociological issues: (i) the division between the conscious subject and social
structures, and (ii) agency or praxis and collective forms of social life (i.e., the
agency/structure problem). Giddens (1984) sees social action and interaction as
tacitly enacted social practices and discusses how they become institutions or rou-
tines and reproduce familiar forms of social life:
The basic domain of study of the social sciences, according to the theory of struc-
turation, is neither the experience of the individual actor, nor the existence of any
 Angel Lin

form of social totality, but social practices ordered across space and time. Human
social activities, like some self-reproducing items in nature, are recursive. That is
to say, they are not brought into being by social actors but continually recreated
by them via the very means whereby they express themselves as actors. In and
through their activities agents reproduce the conditions that make these activities
possible. (Giddens, 1984, p.€2).

With structuration theory, Giddens attempts to integrate human social action

with the larger systems, structures, and institutions of which we are a part. It is
the continual repetition of social action and interaction in more or less routines
or repeated practices that constitute what may appear to be the larger social forms
or systems. Under structuration theory, structure is not outside of and imposed
on social action, but is both constituted/structured by and shaping/structuring
social action.
Structuration theory thus seems to attempt to overcome the structuralist de-
terminism that is sometimes attributed to social theorists who emphasize too
much the reproduction tendency of social structures. Under structuration theory,
precisely because structures and social actions are mutually constitutive and shap-
ing/structuring, there is the possibility of transformation of larger social structures
through situated social actions, which often involve discursive practices. This per-
spective is especially important to the central argument of this chapter: that in
interactional contexts where power relations figure predominantly, social actors
can draw on discourse tactics to attempt to transform the larger social forces
(more on this later).
Both SI and structuration theory thus seem to have formed the sociological
backdrop of discourse-based approaches to research on intercultural communica-
tion (e.g., Scollon and Scollon, 1995; Carbaugh, 2005). Recent readers in intercul-
tural communication (e.g., Gudykunst, 2005; Holliday, Hyde and Kullman, 2004;
Kiesling and Paulston, 2005) also provide entries on discourse-based approaches.
While Giddens (1984) does not focus on discussing language and discursive prac-
tices, many current discourse analysis frameworks have in one way or another
drawn on Giddens’ structuration theory in seeing social actions as predominantly
mediated through language and other semiotic resources (e.g., Gee, 1999). The
ethnography of communication started by Dell Hymes and John Gumperz
(Gumpez and Hymes, 1986) also appeared in around the same period of time as SI
and structuration theory. Gumperz’s work on intercultural/inter-group interac-
tion is important as not many intercultural communication studies focus on non-
egalitarian situations. Below I shall discuss John Gumperz’s contribution to
theorizing about intercultural or inter-group communication, especially in non-
egalitarian contexts.
Researching intercultural communication 

3. Gumperz’s contribution to an interactional

conception and analysis of language

Gumperz’s research has made great contribution to de-centering language as a

research and analytical focus in his privileging of communicative practice, or the
everyday communicative event embedded in mundane everyday activities as the
central analytical focus, and in his constantly stressing the importance of situating
the communicative event (the interaction) in its larger sociocultural and institu-
tional context including the larger context of power relations. One might see him
as a pioneer in critical sociolinguistics (although he might not like to attach such
labels to himself and his work). His rich work in developing theories of intercul-
tural, inter-dialectal, inter-group communication is also a major contribution
which few will dispute about.
Although Gumperz seems to hold reservations about the methods and proce-
dures of conversation analysis (CA), the methods and procedures developed in
conversation analysis, though considered to be clinical by some, do seem to offer
some useful empirically grounded and practical research tools to interaction ana-
lysts, especially if they are used flexibly and not subscribed to religiously. Gumperz
seemingly asserted that conversation analysts have apriori, static assumptions
about groups, communities or group membership.€ An examination of Harvey
Sacks’ early lectures on conversation as well as subsequent work in CA, however,
does not warrant such an assertion. Nevertheless, it is important to take Gump-
erz’s warning about not taking “community” or “membership” as static, given cat-
egories but as something negotiated, constantly evolving in interactions. CA
methods can and need to be more applied to the analysis of interactions at “bor-
derland places”, i.e., cross-cultural, cross-group, cross-community (if you want),
and cross-position interaction (more on this below), and one can see such ana-
lytical projects a bit on the minority side in mainstream CA studies – projects that
will take as its central analytic goal to uncover and describe how sense-making
“methods” and “procedures” come into sharing by participants (e.g., cross-cultur-
al, cross-generation, cross-gender, etc.), who might generally be seen as not shar-
ing much in common. To borrow a metaphor from developmental psychology,
one can say that we need to develop CA analytical projects that are more “develop-
mental” or “longitudinal” than “cross-sectional”.
What are the analytical categories that an interactional analyst should take as
the most relevant ones at the present time, and what should be the short-term and
the long-term objectives of an interactional analysis approach? It seems that a
good unit of analysis is a speech event that is ordinarily recognizable as such by
interaction participants. One should also take what is recognizable as communica-
tion barrier and communication breakdown as a focus for analysis. Gumperz
 Angel Lin

compares this approach with the grammarian’s approach: while grammarians ana-
lyse grammatical and ungrammatical sentences and compare them to yield gram-
matical insights, an interactional analyst should analyse both successful commu-
nication events and instances of communication barriers and breakdowns. The
long-term objective is to uncover the methods and procedures that people
(e.g., coming from very different backgrounds or with very different member-
ships) can possibly use to co-construct common methods/procedures of sense-
making, of achieving some perceived (provisional) sense of inter-subjectivity. This
is a theoretical project with important implications for a number of disciplines and
for practical challenges facing us now in an increasingly globalized world of in-
commensurable discourses (with both processes of homogenization and fragmen-
tation taking place). Communication after 911 takes on different meanings – is
communication or sharing some form of consensus possible only among “mem-
bers”? How do “non-members” (e.g., coming from radically different positions,
backgrounds, be it linguistic, racial/ethnic, religious, social, gender, sexuality, gen-
erational) become recognizable to one another as “fellow members” (of shared
humanity) – i.e., recognizable to one another as sharing some common methods
and procedures of meaning-making and co-inhabiting some shared forms of life
(including methods and procedures for resolving conflicts of interests and cul-
tures), no matter how provisional it is?

Issues in contemporary studies of interaction

A major issue in contemporary studies of interaction seems to be related to the

tendency of researchers to hold a dichotomous micro-macro view of human inter-
actions, especially in conducting interaction analysis. There seems to be a tradi-
tional dualistic view in sociolinguistics (e.g., in terms of micro-interactional socio-
linguistics vs. macro-sociolinguistics) which Gumperz readily speaks against and
shows in his work how unhelpful such a perspective is. This dualistic division re-
flects a lack of ability on the part of the analyst to overcome Cartesian dualism in
theorizing human phenomena. This is also reflected in some general criticism
sometimes directed towards CA: e.g., the accusation that CA is too “micro” ori-
ented. In this regard I want to quote Harvey Sacks in his “micro” analysis of an
introduction sentence in a group therapy session (Spring 1966, Lecture 04a – An
introduction sequence, collected in Gail Jefferson (Ed.), 1992):
One thing we can come to see is that producing the introductions in the form of
a sentence might specifically be done to make available to Jim that ‘a group’ is be-
ing presented. That is, we want to differentiate between Jim being introduced to
‘three people’ in close order, and Jim being introduced to ‘the group,’ one by one.
In that regard, it seems to follow that sentence-making is to be conceived as a kind
Researching intercultural communication 

of social institution in perfectly conventional ways. I suppose we don’t ordinarily

think of the use of grammar as a social institution for demonstrating organization.
Courses in social organization don’t have, I suppose, sections on the way you can
build sentences to present a group, where you use the resources of the grammar to
do that. But it might not be a bad idea. It isn’t, then, that we have sentence-making
on the one hand and social structures on the other, and one can study their rela-
tionship by, e.g., studying dialectics. …. (Sacks, 1966/1992, p.€288)

Thus, the micro-macro analytical division is unproductive (as the discussion of

structuration theory above shows) and tends to divide theoretical and research
work into the work of critical social theorists and the work of micro-(socio)lin-
guists. Whereas a more productive analytical stance would be to see, as both
Gumperz and Sacks do, the marco (e.g., social structures) as being enacted, main-
tained, reproduced, taking shape, or being contested, being transformed…etc.,
through and through in the micro interactional event (e.g., sentence making in
introducing someone to a group). This seems to be the most challenging and yet
most interesting task for the interaction analyst; i.e., not to leave social theory to
the social theorists, as argued by Gumperz himself (2003).
Another issue, which will also form the thesis of this chapter, is the relative
lack of theoretical and methodological attention to analysis of intercultural-com-
munication-situated-in-non-egalitarian-contexts. While there does not seem to
be any dearth of research findings on intercultural communication, they tend to
fall into the trap of linguistic and cultural essentialism; e.g., making claims like:
people from the background of Language A and Culture A make and respond to
compliments in these ways while people from the background of Language B and
Culture B make and respond to compliments in those ways, etc. Also, there is a
need for more study on conflicts or oppositional practices in intercultural com-
munication located in non-egalitarian contexts. Gumperz’s famous studies of the
job interview and the research student ‘pleading’ to the professor (2003) are among
the few classical and pioneering studies in cross-group communication marked by
hierarchical power relationships. And we must also note that Gumperz under-
stands intercultural communication in a broad sense as inter-group (or in what I
would call: inter-location or inter-position) communication; i.e., communication
between people coming from different languages, cultures, dialects, social net-
works or classes, etc. Stretched to the extreme it can be said that all communica-
tion shares in some features (albeit to hugely varying degrees) of some inter-group
or inter-position communication. This is also what Bakhtin means by saying that
we need to acknowledge the “otherness” inherent in any dialogic encounter
(Gardiner, 2004).
In this chapter, I would like to continue in the tradition started by Gumperz in
focusing on analysis of inter-group communication situated in larger sociopolitical,
 Angel Lin

non-egalitarian power matrixes, and in understanding the tactics used by non-

powerful participants to make the best out of a bad situation. Michel de Certeau
discussed and described the everyday tactics used by non-powerful people and
pointed out that tactics are ‘weapons of the poor’ (1984). Understanding the dis-
course tactics used by the non-powerful in inter-group communicative events will
contribute to understanding the discourse strategies that Gumperz has devoted
much attention to studying. In the next section, I shall draw on the analytical tools
of positioning theory and storyline analysis (Davies and Harré, 1990; Harré and
Langenhove, 1999) to analyse interactional examples from two case studies, each
marked by a different configuration of power relations among the interactants.

Drawing on positioning theory and storyline analysis to understand

discursive tactics in inter-group communication in non-egalitarian contexts

In this section I shall draw on the analytical resources of positioning theory (Davies
and Harré, 1990; Harré and Langenhove, 1999) to analyse discursive tactics in two
examples presented. In typical colonial encounters, the colonizer discursively po-
sitioned the colonized as a cultural, ethnic and linguistic ‘other’, establishing bi-
nary separation of the colonizer and the colonized and asserting the naturalness
and primacy of the former (Ashcroft, Griffiths and Tiffin, 1998). In both our daily
conversations as well as public discourses such discursive construction of self and
other and of different subject positions for self and other routinely occurs. Posi-
tioning theory (Davies and Harré, 1990) proposes that such subject positions are
linked to our discursively constructed storylines which are constantly being nego-
tiated by different parties:
One speaker can position others by adopting a story line which incorporates a
particular interpretation of cultural stereotypes to which they are ‘invited’ to con-
form, indeed are required to conform if they are to continue to converse with the
first speaker in such a way as to contribute to that person’s story line. Of course,
they may not wish to do so for all sorts of reasons. Sometimes they may not con-
tribute because they do not understand what the story line is meant to be, or
they may pursue their own story line, quite blind to the story line implicit in the
first speaker’s utterance, or as an attempt to resist. Or they may conform because
they do not define themselves as having choice, but feel angry or oppressed or af-
fronted or some combination of these. (Davies and Harré, 1990, p.€7)

The construction of storyline is central to the establishment and articulation of

collective and personal identities, which involve assigning different subject posi-
tions (or ‘characters’) to different people in a certain context according to a sto-
ryline projected by one’s discourse. By ‘giving people parts in a story’, a speaker
Researching intercultural communication 

makes available ‘a subject position which the other speaker in the normal course
of events would take up.’ (Davies and Harré, 1990, p.€5). Below we shall quote Dav-
ies and Harré (1990) to delineate the key concepts of positioning theory for ana-
lyzing discursive tactics through analyzing the kinds of subject positions and sto-
rylines being both enabled and contested in discourse by different parties:
We shall argue that the constitutive force of each discursive practice lies in its
provision of subject positions. A subject position incorporates both a conceptual
repertoire and a location for persons within the structure of rights for those that
use that repertoire. Once having taken up a particular position as one’s own, a
person inevitably sees the world from the vantage point of that position and in
terms of the particular images, metaphors, story lines and concepts which are
made relevant within the particular discursive practice in which they are posi-
tioned. (Davies and Harré, 1990, p.€3)

In projecting storylines, people routinely draw on culturally available stereotypes

(or recurring storylines) as resources to position themselves and others. In addi-
tion, different storylines are linked to different moral orders, with different sets of
norms about what counts as right, legitimate and appropriate to do (Davies and
Harré, 1990).
It is in light of the conceptual framework and analytical tools offered by
positioning theory that we shall understand “non-egalitarian contexts”. By
“non-egalitarian contexts”, I do not mean a static, fixed, essentialist context out
there. Instead I want to describe the larger power structures in which the inter-
actants are located and the ways in which the different interactants draw on
these structural resources bring into shape, to reproduce (e.g., by the relatively
“more powerful” party – powerful as defined by her/his location in larger socio-
political structures) or to contest and subvert such a move to create a non-egal-
itarian context (e.g., by the relatively “weaker” party – weaker as defined by her/
his location in larger sociopolitical structures). Thus, while interactants are dif-
ferentially located in larger social structures and occupy differential positions in
the larger power matrixes, the local context is being discursively constructed
(reproduced or contested, negotiated and subverted and so on) in situ by the
interactants (e.g., through the different storylines and subject positions being
projected by different parties in conversation). For instance, while one party
starts off by trying to shape the context as a non-egalitarian one (by putting
him/herself in a more powerful position through a particular storyline being
projected), the other party might contest and subvert the effect of such a move
by using discourse tactics (e.g., through negotiating a different storyline and
thus invoking different subject positions and a different moral order).
 Angel Lin

In the following case study of Carman Lee (pseudo-name), a Hong Kong busi-
ness executive and her US client on the phone, we seem to be witnessing such ten-
sions in the negotiation of a less or more egalitarian context. Then in Case Two, we
shall look at the discourse tactics of ‘Long Hair’, a grass-root, leftist, democracy fight-
er in Hong Kong, and how he negotiates a more egalitarian discourse context when
interacting with powerful middle class politicians and party leaders in public.

Case One: A business executive in Hong Kong

Carman Lee works in a medium-size gift and premium company in Hong Kong.
Her company manufactures and trades gifts and premiums, plastic products, both
generic and tailor-made. They have a factory with 200 workers in China where the
manufacturing takes place, and a marketing and sales office in Hong Kong where
designing of products and negotiations with clients take place. Their clients come
from the Middle East, Europe and their biggest clients are from the US. However,
these US clients seldom come to Hong Kong and they communicate with them
mainly through email. The clients she comes into face-to-face contact with are
mostly from the Middle East (e.g., Dubai), and these are diasporic ethnic Indian
and Pakistani business executives who are very hardworking and very willing to
travel. Some clients are from Europe (e.g., Italy), and when they come she will
speak a few words such as Italian and they will be very happy to hear them; they
will also learn a few phrases in Chinese, such as “Ni hau ma?” (How are you?).
English to her is easier to learn than Mandarin Chinese although she is ethnic
Chinese, because to her Mandarin Chinese comes in a more formal style than
Cantonese, which is her mother tongue (e.g., “go haak hou yiu-kauh” in Canton-
ese; in Mandarin Chinese, one should say: “go haak yiu-kauh hou yimh-gaak” – a
more formal, elaborate style needs to be used).
Her job responsibility lies mainly in sales and marketing; solving the problems
of clients, e.g., helping them to do promotion; e.g., a big pharmaceutical company
wants to use their company logo to design a stationery holder plus a clock; her job
focuses on communication with them; e.g., explain the design, negotiate the price,
and the schedule, etc.
Now with e-mail in very common use, she mostly uses e-mail to communicate
with overseas clients. Thus, more written English than spoken English is used, es-
pecially when they are in the same time zone. On socializing with clients: she
mainly needs to socialize with long-term clients who have become personal
friends; when they come she will take them to lunch; these clients are frequent
visitors to Hong Kong and have visited HK for over 10 years; so, they are very
familiar with the places in Hong Kong; and in dinners with them they will talk
about things such as different education systems in different places.
Researching intercultural communication 

The following are excerpts from the interview exchanges conducted in June
2004 between the author and Carman on intercultural communication experi-
ences (the interview was conducted in both Cantonese and English and both par-
ties code-switch naturally in the interview; the following is an English translation
of the exchanges):
Carman: Yes, we have a Hong Kong accent. I care about it a little bit; I feel that it’s
not nice to hear; I’ll learn by imitation; e.g., paying attention to the
English on TV; sometimes when I hear some Hong Kong people speak-
ing English on TV with a distinctive Hong Kong accent I would feel a
bit uncomfortable;… Anson Chan’s (the former high official in Hong
Kong, an ethnic Chinese educated in the University of Hong Kong)
English is okay; and Uncle Tung’s (the former Chief Executive of Hong
Kong) English is not bad either. But I don’t have any problems com-
municating with my clients.
Lin: If you have children, which accents of English do you want them to learn?
Carman: Well, I don’t care much about that; as long as they can communicate, it’s
okay. Because, even within the same country, people have different accents
and you cannot say which ones are the best or more superior. My former
colleague in the bank, when she spoke English we can tell she’s a Hong
Kong person but she is someone who’s speaking rather good English. ….
I think I can handle them (English-speaking foreigners) in my job do-
mains; so I can speak English in certain domains only, e.g., some jargon
related to their culture, which I’m not familiar with; sometimes we
guess each other’s meanings but we can communicate alright. ….
Lin: Have you ever come across any communication difficulties with your
Carman: In particular I have an Italian client, and he’s very happy when I speak
Italian to him, but my Italian is limited (to several sentences) and English
is not his mother tongue and sometimes he’d say, sorry, my poor Eng-
lish; but we can understand each other alright; speaking is more diffi-
cult, because of loss of meaning or misunderstanding; so before each
meeting his secretary will e-mail the agenda to me first and then after
the meeting he or his secretary will give us the minutes, or we’ll e-mail
him to confirm what has been discussed, to do this, just “for sure”.
I have an Engineering colleague who writes very well in English, and
he’s very good in using simplified English to express technical details
and people can understand his writing clearly.
One incident of difficulty in communication: one time in a very noisy
environment, a long-distance call from a US client, and the topic is
 Angel Lin

rather complicated and I experienced some difficulties in communica-

tion with him. And I used a strategy: I said to him: it’s very noisy here;
please let me go somewhere and talk to you again; actually it’s a strategy
to get him to say the things again.
Americans do not “jauh neih” (Yale transcription of Cantonese words
meaning “accommodate you”), i.e., accommodate you; they just speak
as if you speak English as your mother tongue, and they will have
sounds omitted and so on; those who speak English as a mother tongue
will not articulate every sound, e.g., they will not say “I will”, but will
shorten it; whereas second language speakers of English (e.g., those cli-
ents from the Middle East) will articulate every sound clearly and so it’s
easier to understand their English.
Some clients are very arrogant and will not speak to you if you speak
slowly; e.g., some clients on the phone will start with the sentence: Is there
anyone who speaks English. I’d answer him or her: where are you from?
What can I help you with anyway? I’ll answer them directly in English.
On the whole, in the industry or in HK, our colleagues might not be
confident to speak English and especially when the clients speak fast
and our colleagues will become diffident and hesitate to speak English
even further. But I won’t be like that, e.g., I’ll ask them to spell their
names, e.g., spell it please, and then I’ll say it’s strange, as it is not a com-
mon name.
We can see from Carman’s remarks that she is a very confident speaker of English
and she uses English in intercultural communication with other second and for-
eign language speakers of English, such as ethnic Indian and Pakistani clients from
the Middle East or clients from Europe. With these clients she communicates
comfortably in English – a language that does not belong to them as a mother
tongue but a useful communication tool that has forged their business relation-
ships and sometimes personal friendships. Such intercultural communication is
characterized by egalitarian mutual respect. For instance, an Italian client would
admit to her, “sorry, my poor English,” but she accommodates him by using her
limited Italian with him, while he will also use some Chinese phrases to show his
good will. In this sense, both parties show willingness to use the other’s language,
if only as a symbol of respect and interest in the other’s language and culture.
The storyline being co-produced in conversation between Carman and her
Middle-East and European clients is thus one that projects subject positions which
are more horizontally related rather than vertically related to each other; e.g., no
speaker claims her/himself to be occupying a subject position higher than the
other. Also, English comes in not as the superior communication tool, but just as
Researching intercultural communication 

a useful tool for intercultural communication between egalitarian, mutually re-

spectful parties occupying near-peer subject positions. No one claims the subject
position of an expert speaker of English. Also, their creative use of multiple com-
munication strategies to ensure communication of important business informa-
tion (e.g., e-mailing agendas in advance and written records of meetings after-
wards) proves that successful communication does not depend on only one
channel and second/foreign language speakers of English can use English fruit-
fully for intercultural communication without invoking the notion of the need for
native-speaker-defined “good” English. In the storyline co-produced in their in-
tercultural communication, it is a world and a moral order under which both con-
versation participants bear equal responsibilities to make oneself intelligible to
each other and to try one’s best to appreciate each other’s efforts in communicating
across cultural and linguistic boundaries without expecting any one party to lop-
sidedly make all the efforts for making oneself intelligible to the other party.
The mutuality and egalitarian atmosphere that characterize Carman’s interac-
tions with clients from non-English countries (e.g., Middle-East, Europe) stand in
sharp contrast with the kind of attitudes shown by some of her US clients. For in-
stance, some US clients will start a phone conversation by saying: ‘Is there anyone
who speaks English?’ The storyline being projected by this US client’s question
presupposes a world and a moral order that has at least two inter-related ideologi-
cal underpinnings: that it is entirely the responsibility of the other party to accom-
modate the US client linguistically, and that the burden of successful intercultural
communication rests entirely with those who need to mater English to communi-
cate with those who already speak English as a first language.
In this utterance we see the reincarnation of the storyline in imperialist litera-
ture, e.g., Robinson Crusoe, the legacy of imperialist and colonial mentalities
(see analysis by de Certeau, 1984). This brings us to the witnessing of another
practice of Carman, which can be seen as subversive. Instead of answering this
client in a subservient way, she asked in her own variety of English: ‘Where are you
from? What can I help you with anyway?’ By responding to a question not with an
answer but with a question, she turned the tables and showed her agency and con-
fidence in answering back to the voice posing as a colonial master. In her defiant
act, she answered with a voice that belongs to a self-respecting, empowered agent
who does not subscribe to the master-primitive imperialist storyline and resists
the first party’s attempt to define the context as a hierarchical one. She is projecting
a totally different storyline in her reply. In this storyline the subject positions are
reversed: she is someone who demands to know the background of the caller;
i.e., she is the one who has the right to demand information from the caller in the
first place. In the storyline that she counter-projects with her reply, she is an equal
partner in this business and professional relationship with her US clients, not a
 Angel Lin

linguistic or cultural inferior. For instance, she will handle them quite confidently;
e.g., by asking them to spell their names when the pronunciation is not clear. In
doing this, she indicates to the other party that the burden of intercultural com-
munication rests with both parties, and not only on her side and she successfully
uses her discourse tactics to negotiate a more egalitarian intercultural communi-
cation context through projecting a different storyline with more egalitarian sub-
ject positions linked to a moral order under which both parties share equal re-
sponsibilities for making the communication work rather than expecting one
party to lopsidedly accommodate the linguistic demands of the other party. In the
next section we shall use positioning theory and storyline analysis to analyse dis-
course tactics used by people of the marginalized.

Case Two: ‘Long Hair’: A defiant, outspoken, grass-

root, democracy fighter in Hong Kong

‘Long Hair’ is the nickname of Leung Kwok-hung, a leftist, outspoken, grassroot,

political activist in Hong Kong for many years. He was elected a Legislative Coun-
cillor in Hong Kong on 12€September 2004. His winning of the election was main-
ly due to the support of young voters, mainly disenfranchised youths in Hong
Kong who are discontent with the education system, high unemployment rates
and the increasingly stratified society along social class lines. Many university stu-
dent associations also invited him to give talks right after his successful election.
My discourse data consists of his public televised debates with powerful right-
wing business leaders who are also powerful party (e.g., Liberal Party) leaders in
Hong Kong before and after the election. Due to limited space here, I shall quote
only one excerpt from one such debate in a public forum shortly after the election
(20 Septermber 2004, City Forum, televised live by Radio Television Hong Kong;
the event was recorded by the researcher for analysis). I shall briefuly describe the
context of the excerpt and then present the excerpt of the exchanges between
James Tien (a powerful business leader and also the Chairman of the Liberal Party
in Hong Kong) and Leung Kwok-hung (Long Hair).
When James Tien Pei-Chun, Chairman of the Liberal Party then, is debating
with Andrew Cheng of the Democratic Party, Leung Kwok-hung (Long Hair) in-
terrupts and speaks to James Tien in an assertive tone (The original Cantonese
utterances are transcribed in Chinese characters, with English glosses tabulated
next to them in the table below):
Researching intercultural communication 

Cantonese utterances English translation

1. 梁: 田少,我唔會再要你對我道 Leung: Young Master Tien, I will not request you to

歉,你唔使驚。 apologize to me again, you don’t have to be afraid.
2. 田:你對我咁友善,我點會驚。 Tien: You are so friendly to me, I will not be afraid.
3. 梁:不過我段報紙、有段新聞比 Leung: But see the news report, a news report for
大家睇,就係田北俊公司被爆欠 all of us to see, is the reporting of the incident of
薪(說時拿出剪報給現場觀眾 wages owed by James Tien’s company (Leung
看)。你為打工仔著想,你竟然 pulls out a newspaper cutting and shows to the
間搞成咁!我初時以為你係清潔 audience). You are considerate towards the work-
先生,你有咩野講?欠左人百幾 ers, you did something like this! I at first believe
萬!個個人都好老下架喇,做左 you are Mr Clean, you have anything to say? Ow-
咁耐幫你。 ing people a million dollars or so! That person is
quite old, and have served you for so long.
4. 田:依個係我一個合資公司= Tien: This is one of my joint ventures…=
5. =梁:你係咪口蜜腹劍? =Leung: Are you poison in the honey? (literal
translation: honey-mouth and sword-stomach)
6. 田:依個係合資公司,響國內既 Tien: This is a joint venture, it’s a case in Main-
情形,宜家仲響度打緊官司= land, now it is engaging in a lawsuit=
7. =梁:即係你唔知?= =Leung: That means you don’t know?=
8. =田:我唔係好詳細了解,但係我 =Tien: I don’t quite understand the details, but I
會負責既。 will be responsible.
9. 梁:你到宜家都唔知呀?你有無 Leung: You still don’t know now? Have you ever
關心過個個人呀? cared about that person?
10. 田一臉為難,想開口時梁又搶著 Tien looks embarrassed, Leung again interrupts
說:宜家好簡單,我就唔會再攞d before Tien can speak: Now it is simple, I won’t take
野過黎比你睇架喇!我有首詩送 any more thing out and give it to you to see! I have
比你,你有無、你識唔識水滸 a poem as a present for you; have you, do you know
傳,水滸傳度有首詩。 Water Margins, there is a poem in Water Margins.
11. 田笑說:你好似武松吖! Tien jokes: You are just like Wu Song.
12. 梁不理他,續說:叫苦熱歌,苦 Leung ignores him, and continues: It’s called Bit-
熱歌。赤日炎炎似火燒,野田禾 ter-Hot Song, Bitter Song. Hot red Sun is burning
稻半枯焦,農夫心內如湯煮,皇 like fire, crops are half-withered. Farmers’ hearts
孫公仔把扇搖。(說時真的拿出 are like boiling soup, the royals are fanning.
扇子在搖)嗱!你,依把扇就送 (Leung takes out a real fan and fans himself with
比你,第日你搖下搖下,睇下香 it.) See! You, this fan is for you; on the other day
港咁多人失業,三十萬人失業, you fan and fan, seeing how many people in Hong
你、你都話你自己有功既!七年 Kong lose jobs, 300,000 people are unemployed,
以黎,董建華禍港央民,斗零救 you, you still say you have your contribution!
窮人,二千億港市,你係咪搖住 Seven years from now, Tung Chee-wah caused
扇響度睇?你響行政會議... disasters to the country and its people, 50¢ to save
the poor, $200 billion for the Hong Kong market,
are you fanning and watching? In the Executive
Council, you…
 Angel Lin

Cantonese utterances English translation

13. 田笑著反問:今日係咪好過舊年 Tien smiles and asks back: “Isn’t today much bet-
呀已經?係咪你把扇多少撥倒d比 ter than last year? Is that your fan fanning stuff for
窮人呢宜家? the poor?
14. 梁:你係咪搖左扇?你話你係聽 Leung: Have you fanned the fan? You said you are
依個居民既心聲,你落區,你知 listening to the people’s voice at heart, you visited
唔知翠華餐廳一杯齋啡加一個菠 the community, do you know how much a black
蘿油係幾錢?你淨係識紅酒既價 coffee and a pineapple-bun-with-butter cost at
格,你主張紅酒減稅! Tsui Wah Restaurant? You only know the price of
red wine, you proposed red wine tax-reduction.
15. 田:以前我係唔識,依排我都知 Tien: I didn’t know in the past, but now I know
道係有ABC餐。 there are set-meals A, B and C.
16. 梁:幾多錢呀? Leung: How much then?
17. 田:十五蚊倒啫。 Tien: Like around 15 bucks.
18. 梁:十五蚊?! Leung: 15 bucks?!
19. 田:ABC餐喎! Tien: Set-meals A, B C!
20. 梁:我話比你聽,係二十蚊,齋 Leung: Let me tell you, it’s 20 bucks, a black coffee
啡加菠蘿油,翠華係全港最多人 with a pineapple-bin-with-butter. Tsui Wah is the
食既餐廳。 most popular restaurant in Hong Kong.
21. 田:咁你食得貴我好多喇長毛! Tien: So you dined more expensively than I did,
Long Hair!
22. 梁:嗱!紅酒減稅,你地就講到 Leung: See! Red wine tax-reduction, you guys
好優惠喇自由黨,當係政績咁 claiming a good offer (by) the Liberal Party, say-
講。你有無諗過D人,連食過十五 ing it like a contribution. Have you ever thought
蚊既餐都無,你講得啱喇! of the people, not even having eaten a 15-buck-
meal; you have said so right!

In this exchange, we can see that Long Hair is very skillful in using quick, witty,
discursive tactics to position his interlocutor, his debating opponent, James Tien,
as a rich family’s son not knowing much about the living conditions and suffering
of grassroot people. James Tien, being well-known in Hong Kong society as com-
ing from a rich family, is often addressed to as ‘Tien-siu’ in public media (literally:
Young Master Tien). In the Chinese language, ‘personal name + siu’ is an address
term reserved for young masters, usually used by servants to address their young
masters (‘siu’ being a word to attach to the name of the young master; ‘siu’ means
‘young master’). In public media in Hong Kong, sons of wealthy families are often
referred to as X-siu (X is the name of the person). Long hair (Leung), by using this
membership category term (Jayyusi, 1984; Hester and Eglin, 1997) right from the
beginning of the exchange, is positioning Tien as someone coming from the rich
upper classes, and as someone who does not share the lifeworld of the majority of
people in Hong Kong.
Researching intercultural communication 

Then Leung pulled out a newspaper clipping to show that one of Tien’s em-
ployees was treated unfairly (with wages unpaid to him). By showing concrete
evidence and by cornering Tien about his ignorance of the plight of his own em-
ployee, and then juxtaposing/equating Tien’s ignorance with his lack of concern
(Turn 9), Leung is launching a powerful accusation against Tien in Turns 3–9. Be-
ing caught unexpectedly by Leung on this incident, Tien (apparently without any
assistant beside him to brief him on this incident) acts in a role that Leung seems
to have both expected and positioned him to act in the storyline projected in Le-
ung’s discourse: That Tien-siu (Young Master Tien) is uncaring and unkind even
to his own employee (or servants who have served him – his company – for so
long; see Turns 3 and 9).
Having cornered Tien with this concrete incident showing Tien’s lack of con-
cern and care for his own employees, Leung immediately recited a Chinese ancient
poem (‘as a present’ to Tien) which talks about the plight of poor people under a
cruel government in the Sung Dynasty. The poem was taken from the famous
Chinese classical novel, Water Margins, which depicted the story of a group of
disenfranchised people who were forced to rebel against an oppressive, uncaring,
corrupt government which let the rich and the powerful bully poor, powerless,
ordinary people in the Sung Dynasty of China. It must be pointed out here that
while Leung is from the grassroots, he is widely-read in the Chinese classics and
can recite Chinese classical poetry and essays at ease. Compared with Leung, Tien
is shown to be not only an uncaring rich son (due to family wealth), but also some-
one who is unfamiliar with Chinese classics. Leung’s fluent recitation of this an-
cient Chinese poem in one of the most famous Chinese classical novels, has again,
given Leung an upper hand. By reciting this poem from Water Margins, Leung is
also evoking the collective memory of the storyline of Water Margins: how decent,
honest people were forced to become anti-government rebels to fight for justice.
After travelling on the time line from the present (Tien’s apparent unfair and
unkind treatment of his employee) to the ancient (reciting the poem from Water
Margins to evoke the storyline of an unfair and unjust ruling elite), Leung again
takes Tien back to the present by interrogating him about his knowledge of the
living conditions of the grassroot people in Hong Kong (Turns 14–20): asking Tien
how much it costs to have a common meal in Hong Kong). Again, Tien’s knowl-
edge is shown to be inadequate, and Tien is further positioned as a typical member
of the rich not knowing the plight of the poor.
Leung’s discursive tactics are systematic, almost like well-planned, and he has
cleverly drawn on popular cultural and discursive resources: news reports, ancient
Chinese classical stories, Chinese poem depicting the plight of poor people, and
everyday streetwise knowledge (of the living conditions) of grassroot people.
 Angel Lin

When reciting the poem, Leung fans a traditional Chinese paper fan, which
serves as a hook to anchor the audience’s imagination (those watching this debate
in front of the television) in Leung’s storytelling – his projecting of a storyline not
too dissimilar to that of Water Margins.
Tien is thus put on the defensive, but given his lack of Chinese cultural and
discursive resources (Tien was Western and English-educated, not familiar with
Chinese classics), his rebuttal seems so ineffective in front of Leung’s consecutive
attacks, the last of which being the accusation of Tien as only knowing and caring
about the reduction of red wine tax (Turn 22). Again, the middle class symbol of
red wine (in Hong Kong, red wine consumption is associated with a middle and
upper class life style) is invoked by Leung to position Tien as a bona fide middle
class person, neither cognizant of, nor caring about, the life conditions of the
grassroot people in Hong Kong.
Long Hair has always been well-known for his eloquent, outspoken, defiant
discourse style and this is precisely why some young people and many working
class people like him. They like his upfront, straightforward, no-nonsense dis-
course style and his consistent voicing out of the economic difficulties of the grass-
roots and his direct attacks on the non-democratic political structure of Hong
Kong. When a well-known rich guy, James Tien, who was also Chair of the Lib-
eral Party representing business interests, was in the debating show, Long Hair
deployed his discursive tactics skillfully to position Tien in a negative light: as
someone who does not know about, and cannot, and will not care about grassroot
people in Hong Kong.
Has Leung been unfair to Tien in cornering him with his superior Chinese
cultural and Hong Kong streetwise knowledge and linguistic resources? Has he
been not interacting in a rational way? Recent critiques of Habermas’s ideal com-
municative situation, where interactants interact in a constraint-free, egalitarian
context, have pointed out how unrealistic it is when the interactions are between
people located in different power relationships (e.g., Crossley and Roberts, 2004).
Gardiner (2004) has even pointed out that subscribing to such rationality norms
will bring more damage to the already marginalized in such a context. In the above
analysis, I attempt to show how Leung (relatively powerless in terms of wealth and
in the existing governing structure of Hong Kong) skillfully deploys his other
kinds of cultural and linguistic capital (e.g., his familiarity of Chinese classical
stories and street knowledge of Hong Kong) to position an otherwise much more
powerful person (Tien) in a negative light. Tien is shown to be of a lesser statue
given the moral order projected by Leung’s storyline. Such a (re)presentation of
the world (and the moral order and accompanying rights and obligations sets
linked to it) gives Long Hair the moral high ground.
Researching intercultural communication 


Having looked at the two examples above, it seems to us that intercultural or inter-
group communication is more likely to be (at least provisionally) successful if both
parties are willing to make the effort to overcome communication barriers, to mu-
tually respect each other’s language and culture (e.g., Carman and her European
and Middle-East clients), and to mutually share the burden of intercultural com-
munication. In their conversation both parties co-produce a storyline which offers
relatively more egalitarian subject positions for both parties. However, in non-
egalitarian contexts (which are in fact not static and are open to negotiation and
re-negotiation through discourse), intercultural communication does not always
resemble the well-intentioned, civil, good-mannered interactive styles of interact-
ants in other intercultural communication contexts, and ‘weaker’ parties might
draw on discourse strategies or tactics; e.g., returning an arrogant question with a
question, turning the tables, and counter-projecting a different storyline with a
more empowered subject position for self (as in Carman’s example when interact-
ing with an arrogant U.S. client) to subvert the power relations and to negotiate for,
and reconstitute the context into a more egalitarian context for interaction. Such
discourse tactics often do not subscribe to rationality, appropriateness or polite-
ness norms as these discourse tactics (or strategies, in Gumperz’s terms) are ‘weap-
ons of the poor’ (de Certeau, 1984). The use of positioning theory and storyline
analysis seems to be a promising direction to help intercultural communication
researchers understand how different social and cultural groups located in differ-
ent positions in the larger social structures, nevertheless, attempt to project a dif-
ferent social and moral order under which they can mitigate their structural dis-
advantage and create a discursive context where more egalitarian subject positions
are discursively made possible, if only momentarily, thus, attempting to change the
context and larger social forms, norms and structures through in situ social ac-
tions and discourse tactics (see earlier discussion of structuration theory). This
paper represents a preliminary attempt to analyse two examples of such inter-
group communication in non-egalitarian contexts and it is hoped that further re-
search in this area will help us understand the different discursive resources (and
constraints) leading to both the challenge and the degree of (im)possibility of
achieving intersubjectivity in inter-group/intercultural communication in adver-
sarial situations.
 Angel Lin


The author is indebted to the anonymous reviewers for their useful comments and
suggestions for revision on an earlier draft of this paper.


Ashcroft, Bill, Griffiths, Gareth and Tiffin, Helen. 1998. Key Concepts in Post-colonial Studies.
London and New York: Routledge.
Blumer, Herbert 1986. Symbolic Interactionism. Berkeley: University of California Press.
Carbaugh, Donal 2005. Cultures in Conversation. Mahwah, New Jersey: Lawrence Erlbaum.
City Forum. Televised live on 20€September 2004 by Radio Television Hong Kong.
Crossley, Nick and Roberts, John M. 2004. After Habermas: New Perspectives on the Public
Sphere. Oxford: Blackwell.
Davies, Bronwyn and Harré, Rom. 1990. Positioning: “The discursive production of selves”.
Retrieved from
De Certeau, Michel. 1984. The Practice of Everyday Life. Berkeley: University of California Press.
Eerdmans, Susan L., Prevignano, Carlo L. and Thibault, Paul (eds). 2003. Language and Interac-
tion: Discussions with John J. Gumperz. Amsterdam, The Netherlands: John Benjamins.
Gardiner, Michael E. 2004. “Wild publics and grotesque symposiums: Habermas and Bakhtin
on dialogue: Everyday life and the public sphere”. In After Habermas: New Perspectives on
the Public Sphere. Nick Crossley and John M. Roberts (eds), 29–48. Oxford: Blackwell.
Gee, James P. 1999. An Introduction to Discourse Analysis: Theory and Method. London:
Giddens, Anthony. 1984. The Constitution of Society: Outline of the Theory of Structuration.
Cambridge: Polity Press.
Gumperz, John J. 2003. “Response essay”. In Language and Interaction: Discussions with John J.
Gumperz. Susan L. Eerdmans, Carlo L. Prevignano and Paul Thibault (eds), 105–126.
Amsterdam: John Benjamins.
Gumperz, John J. and Hymes, Dell (eds). 1986. Directions in Sociolinguistics. Oxford: Blackwell.
Gundykunst, William B (ed.). 2005. Theorizing about Intercultural Communication. Thousand
Oaks: Sage.
Harré, Rom and Langenhove, Luk V. 1999. “The dynamics of social episodes”. In Positioning
Theory. Rom Harré and Luk V. Langenhove (eds), 1–13. Oxford: Blackwell.
Hester, Stephen and Eglin, Peter (eds.). 1997. Culture in Action: Studies in Membership Categori-
zation Analysis. Washington, D.C.: International Institute for Ethnomethodology.
Sacks, Harvey. 1966/1992. “Lecture 04a – An introduction sequence”. In Lectures on Conversa-
tion. Gail Jefferson (ed), 288. Oxford: Blackwell.
Holliday, Adrian, Hyde, Martin, and Kullman, John. 2004. Intercultural Communication.
London: Routledge.
Jayyusi, Lena. 1984. Categorization and the Moral Order. Boston: Routledge.
Kiesling, Scott F. and Paulston, Christina B (eds.). 2005. Intercultural Discourse and Communica-
tion. Oxford: Blackwell.
Sluga, Hans, and Stern, David G (eds.). 1996. The Cambridge Companion to Wittgenstein.
Cambridge: Cambridge University Press.
Scollon, Ron and Scollon, Suzanne Wong. 1995. Intercultural Communication. Oxford: Blackwell.
Studying interaction in order to cultivate
communicative practices
Action-implicative discourse analysis

Karen Tracy and Robert T. Craig

Communication Department, University of Colorado,
Boulder, CO 80309, USA

Action-implicative discourse analysis (AIDA) is an ethnographically informed

discourse-analytic approach that works to provide normative understandings
of situated communicative practices that are action-implicative for social life.
Extending the logic of grounded practical theory (Craig and Tracy 1995), AIDA
develops reconstructed accounts of the communicative problems, interaction
strategies, and normative ideals of a practice. We introduce AIDA and illustrate
the approach with an example from recent research on school board meetings
in an American local community. We compare AIDA with other approaches
to language and social interaction, focusing on interactional sociolinguistics
(Gumperz) and conversation analysis (Schegloff). We argue that to understand
the distinctive character of these approaches requires recognizing each one’s
orientation to the discursive context of a particular academic discipline.


Interview Comments:
(1) John Gumperz, Interactional Sociolinguistics
As to “regularities” of communicative practice, I believe that these should
ultimately be derived from or related to in-depth analyses of situated en-
counters in a variety of settings (Prevignano and Thibault 2003a: 151).
(2) Emanuel Schegloff, Conversation Analysis
If one is committed to understanding actual actions (by which I mean
ones which actually occurred in real time), it is virtually impossible to
detach them from their context for isolated analysis with a straight face
(Cmejrkova and Prevignano 2003:€39).
 Karen Tracy and Robert T. Craig

We begin this chapter by echoing words that Schegloff and Gumperz uttered in
interviews in which each was questioned about his approach to studying interac-
tion. Conversation analysis (CA) and interactional sociolinguistics (IS) differ from
each other in many significant ways, as does action-implicative discourse analysis
(AIDA), our own approach. As a starting point, however, all three approaches hold
this in common: to understand social action, interaction, or communicative prac-
tices – whatever this stuff is to be called – requires looking at it in the context in
which it occurred.
Our chapter is divided into two sections. In the first section we overview
AIDA, providing an example to show how we analyze interaction. For the example
we draw upon some recent work studying community-level school board meet-
ings. The second section of the chapter gives focal attention to the similarities and
differences of AIDA with two alternative approaches, CA and IS. We argue that to
understand the distinctive character of these three approaches requires recogniz-
ing each approach’s orientation to the context of a particular academic discipline.
These disciplinary contexts have shaped what each approach takes for granted or
treats as contested about language and social life. That CA originated in sociology,
IS in linguistics, and AIDA in communication is crucial to understanding why
each approach poses the questions about interaction that it does.

Action-implicative discourse analysis

AIDA is centrally interested in describing the problems, interactional strategies,

and ideals-in-use within existing communicative practices. It is an approach that
melds the analytic moves of discourse analysis – attending to situated talk and
texts – with the goal of developing an understanding that will be action-implica-
tive for practical life. AIDA works to provide a reconstructed account of the com-
municative problems, interaction strategies, and normative ideals of a practice so
that participants will be able to reflect in more sophisticated ways about how to
act. AIDA takes a rhetorical point of view, presuming that people can make more
or less reflective decisions about how to communicate in order to act well and
achieve or avoid certain outcomes. It is a normative approach: potential usefulness
for being able to think and act wisely is a key criterion for assessing the contribu-
tion of particular studies. In what follows, we describe intellectual traditions that
shaped AIDA and say a bit about its focal unit and aims, methodological profile,
and rhetorical-normative stance. A more elaborated description of AIDA can be
found in Tracy (2005).
Action-implicative discourse analysis 

AIDA’s intellectual heritage

AIDA is best described as the coming together of two traditions: practical theory,
an approach developed in the field of communication, and discourse analysis as it
is practiced in the multidisciplinary community. Consider each tradition.
Craig (1989, 1996, 1999, 2008b; Craig and Tracy 1995) has argued that com-
munication studies should be conceived as a practical discipline rather than an
empirical science. Rather than assuming that the ultimate goal of inquiry should
be to produce descriptions and theoretical explanations of empirical phenomena,
as is the case when a discipline is conceived as a science, a practical discipline takes
its ultimate goal to be the cultivation of practice. This difference in goals has impli-
cations for the role of theory, because the cultivation of practice requires attention
to normative as well as empirical questions. Whereas explanatory scientific theory
lends itself to the cultivation of an instrumental (means-ends) orientation to prac-
tice, practical normative theory is “centrally concerned with what ought to be; it
seeks to articulate normative ideals by which to guide the conduct and criticism of
practice” (Craig and Tracy 1995:€249). How exactly to integrate the technical-pro-
ductive (techne) side of communication with its moral-political (praxis) aspects is
a major challenge for communication studies conceived as a practical discipline.
Practical theory seeks to reconstruct communicative practices and provides
methodological guidance for doing so (Craig and Tracy 1995). To reconstruct a
practice means to conceptualize an idealized, normative model that is grounded in
close observation as well as critical reflection. Researchers can reconstruct commu-
nicative practices at three levels. First and most crucial is the problem level: identi-
fying the problems that occur for different categories of participants in particular
social practices. Second, reconstruction can describe the specific conversational
techniques and strategies that are employed to manage focal problems (the technical
level). Finally, reconstruction can formulate the abstract ideals and principles that
account for the selection of techniques for addressing particular kinds of problems
(the philosophical level). Of note, the philosophical level must be grounded in situ-
ated ideals, the beliefs about good conduct that can be inferred from patterns of
praise and blame made by participants in actual situations of practice.
AIDA adopts the goals of practical theory and pursues them through the
method of discourse analysis. Discourse is a term that gets used in quite different
ways (e.g., Cameron 2001; van Dijk 1997a, 1997b). Our usage is similar to that
found in linguistics (e.g., Schiffrin 1994), where “discourse” is paired with the term,
“analysis” and treated as an umbrella term to refer to a variety of approaches to the
study of talk or text. At its simplest, discourse analysis involves careful study of
recorded and transcribed talk or text, where excerpts are used to make scholarly
arguments. A second and different meaning of the term “discourse” is informed by
 Karen Tracy and Robert T. Craig

the work of Michél Foucault (1972) – what Gee (1999) refers to as big-D discourse
in contrast with little-d discourse. Big-D discourse, often mentioned in the plural
(discourses), refers to complex social practices such as education or business. Some
forms of discourse analysis, for example critical discourse approaches (Fairclough
2001), are interested in both big-D and little-d discourse, but many discourse ana-
lysts are not. For this reason it is important to keep the two meanings distinct.
As an approach that analyzes interaction, AIDA has been influenced by CA,
anthropologically-influenced speech act traditions, discursive psychology, and
critical discourse analysis (CDA). From CA, AIDA takes the commitment to study
everyday interaction and the practice of repeatedly listening to exchanges that re-
searchers have transcribed while attending to many particulars, including intona-
tion, abrupt word or phrase cut-offs, and repetition and vocalized sounds
(uh, um, eh). Moreover, although not accepting the CA principle that an interpre-
tation should only use what is visibly displayed in a next turn at talk (Schegloff
1992, 1998), AIDA does share the CA view that how an interactional partner re-
sponds is an important resource for anchoring proposals about participant mean-
ing. From anthropologically-influenced speech act traditions (Blum-Kulka, House
and Kasper 1989; Brown and Levinson 1987; Gumperz 1982b), AIDA assumes the
importance of seeing assessments about conversational actions as culturally-in-
flected judgments. Discursive psychology contributes to AIDA through its notion
of dilemma (Billig et al. 1988) and in its development of a rhetorical stance toward
discourse. Finally, critical discourse approaches argue that small-d discourse
should be connected with big-D discourses (Fairclough and Wodak 1997). As
AIDA is committed to cultivating the communicative practices that are studied,
CDA offers one model of how that linkage might be made. But, let us consider
what AIDA studies of interaction look like in their own terms.

Distinctive features of AIDA

AIDA focuses on communicative practices in institutional sites, with an analytic

aim of reconstructing the web of actor problems, conversational moves and strate-
gies, and situated ideals involved in those practices. An obvious question becomes,
then, what is a communicative practice? Practice as a term has some useful ambigui-
ties; at its core, though, it can be thought of as a way of referring to activities that
occur in specific places among specific kinds of people; practice is another way to
refer to a speech event (Hymes 1974) or what participants take to be a situation’s
frame (Goffman 1974; Tannen 1993). Ordinary names given to practices often call
up a constellation of site-people-purposes connections. “School board meetings,”
“departmental colloquia,” and “classroom discussions” are examples of easily
recognized practices related to educational settings. Practice is a way of unitizing
Action-implicative discourse analysis 

the social world to enable analysis. Since institutional practices involve multiple
categories of people who are positioned differently within the practice, the problems
of a practice will differ with a participant’s position. Getting a handle on the interac-
tional problems from the points of view of the main categories of participants is one
aim of AIDA, although often this aim is pursued across multiple studies.
Having identified an important communication practice, a next question be-
comes how to study it. AIDA is a type of discourse analysis that is also ethno-
graphic. To reconstruct a communication practice well demands that a researcher
have extensive knowledge about the routine actions and variation in the practice.
This requires the analyst to do sustained observation of the practice. It also re-
quires analysts to develop an understanding of both how participants talk with
each other in the practice (the focal discourse) and how they talk about their prac-
tice (meta-discourse). What exactly will be the necessary ethnographic compo-
nents will depend on the practice being studied.
In the analysis of school board meetings, soon to be illustrated, the focal dis-
course data were 200 hours of one community’s school board meetings recorded
from a local cable broadcast and collected over a several-year time span. In addi-
tion to the focal discourse, only a small proportion of which was transcribed, were
the following kinds of data: notes taken from viewing the televised meetings; sev-
eral observations of the meetings on site; agenda, minutes, and other documents
related to particular policy discussions; local newspaper articles and editorials
about Board activities; and interviews with a variety of participants. Moreover,
since all of these materials came from one community, the final activity involved
observing meetings in other communities. Thus, a first step in AIDA is to develop
extended knowledge of a focal practice. This can be accomplished by taping
(or getting access to tapes of) a good number of hours of the central discourse
activity, and by building up a portrait of the scene, the people, and the practice
drawing on whatever additional materials are relevant and accessible.
A next step for AIDA is to identify the segments of a focal practice for tran-
scription and analysis. At the selection and transcription stage, AIDA differs from
CA in two ways. First, AIDA would never begin with discourse moments that
before analysis, as Harvey Sacks would advocate, seem to be “utterly uninteresting
data” (1992:€293). While there is no dispute that such analyses can be valuable, for
AIDA, not all moments of interaction are equally promising places to start. In
AIDA, selecting stretches of discourse to be transcribed is a theoretically shaped
activity. Since one goal is to understand the problems of a practice, moments in
which participants seem to be experiencing discomfort, tension, or conflict are
especially promising targets to focus on. Since another goal is to understand the
situated ideals of a practice, instances where participants express evaluation of
other people’s actions are a second type of talk likely to be selected. Finally,
 Karen Tracy and Robert T. Craig

segments of interaction that seem at odds with how an institution describes its
aims and practices are also potentially of interest.
Second, AIDA studies typically work with relatively long segments of interac-
tion and give limited attention to timing and prosody. The reason for this choice
flows from the AIDA commitment to develop ideas that contribute to participants’
reflection about a practice. For this reason, AIDA gives primary attention to the
aspects of communication about which people are most able to reflect: choices
about wording, speech acts, arguments, and speech or story organizations.
In its normative orientation and its interest in both big-D and little-d discourse,
AIDA resembles CDA. The normative principle that guides AIDA differs, however,
from that of CDA. Whereas CDA is centrally committed to a negative critique that
exposes invisible practices of power and domination rooted in macrosocial inequi-
ties, AIDA is centrally committed to addressing normative problems that arise with-
in particular, situated social practices. AIDA, unlike CDA, aims toward a positive
reconstruction that conceptualizes how particular communicative practices should
be conducted. From an AIDA point of view, power and status differences are an una-
voidable, and often desirable, aspect of institutional life. Practices cannot be judged
without attending closely to their particular contexts. AIDA draws upon the Aristo-
telian idea of phronesis – good judgment, prudence, practical wisdom, sound and
thoughtful deliberation, reasonableness – as a basis for the critique of practices. Ph-
ronesis is “not a simple process of applying principles or rules to cases that leaves the
principles or rules unchanged; in prudential practice, there is a negotiation between
the case and the principle that allows both to gain in clarity” (Jasinski, 2001:€463).
Within AIDA, the central starting point for development of normative propos-
als is to identify the practice’s situated ideal(s). Situated ideals are participants’ be-
liefs about good conduct that can be reconstructed from discursive moments in
which they praise and criticize. Situated ideals capture the complex prioritizing of
competing concerns and values that not only will, but also arguably should, be
operative in actual practices. Situated ideals may be reconstructed from analysis of
participant interviews (Tracy 1997) or from study of interactive moments in con-
junction with institutional documents or other segments of interaction (e.g., Agne
2007). In the school board meeting project, the school district was developing its
policy position toward students and staff who were gay (Tracy and Ashcraft 2001).
In this deliberative body, the group’s espoused principle of communicative conduct
was to avoid arguing over words. Yet, in reflective moments and in its actual prac-
tices, participants treated word arguments positively, framing them as serving
valuable functions. Arguments over document language were used to manage a
dilemma. To make a decision, the group sought to advance the value to which the
majority of the group was committed – in this case, advocating acceptance of gays.
At the same time, the group majority wanted to maintain good relations with group
Action-implicative discourse analysis 

members committed to a contrary value. Because the school board wanted to avoid
being dismissive and sought to show that it was treating all views seriously, it need-
ed to spend a significant amount of time talking about the wording options rather
than moving ahead merely because they had the needed number of votes. Arguing
over words was how the group attended to these competing commitments.

AIDA example: School board meetings

To illustrate how AIDA analyzes interaction, we focus on one exchange that oc-
curred at a school board meeting. As background, it is important to note that, in
the United States, local governance committees, commonly referred to as school
boards, are influential in shaping educational policy. These boards, usually ranging
between 5 and 11 members, are elected by their local communities and make a
host of decisions about policies, resource allocation, and to a certain degree, cur-
riculum. Of all the decisions that school boards make, none is quite as important
as the task of selecting the person to fill the role of superintendent. It is the district
superintendent who interprets and implements the board’s policies and directs the
day-to-day operation of the school district. This person is enormously influential.
A school board meeting typically involves the elected officials, the superin-
tendent and selected school staff, and varying numbers of citizens from the com-
munity. Meetings are public, often broadcast over community-sponsored radio or
television stations, and include times for citizen commentary and for discussion
among the board members about issues on which they will soon be voting. Among
some boards there is little disagreement and almost all votes are unanimous
(Newman and Brown 1992). At other times, though, boards become sites for the
playing out of serious disagreements that exist in the community. The exchange
that is analyzed below comes from a board meeting in which there was a history of
votes routinely splitting into majority and minority positions.
The exchange occurred among one of the board members who took the mi-
nority position and was usually outvoted (Shoemaker), two of the board officers
who were part of the majority coalition, (Hult and Shonkwiler) and a consultant
(Ceruli) who had been hired to assist with the district’s search for the next
superintendent. On the meeting’s agenda, the item of discussion was described as
“Approval and Acceptance of the Superintendent Search Committee.”1
Meeting Excerpt: Minority Member Shoemaker’s No Vote

1. This analysis is a shortened version of one that appears in more detail, with more spe-
cifics of the school board meetings and other segments of meeting interaction in Tracy and
Standerfer (2003).
 Karen Tracy and Robert T. Craig

(H=Hult, the president; Sh=Shoemaker, board member in minority position;

S=Shonkwiler, the vice president; C=Ceruli, paid consultant to the Board)
1 H: Susan
2 Sh: I have some questions, um. I guess clarification first. I assume that these names are
3 added to, um (.) what we are going to be voting on here. We’re voting on the process,
4 the budget, and these names, is that correct?
5 H: We’re voting on the process and the budget. Search process and budget.
6 Sh: Will we=
7 H: = I guess the names are an inherent uh element of that (.)
8 Sh: So we are voting on the names orâ•‚
9 H: It’s the whole thing we are voting on
10 Sh: not
11 H: Yeah.
12 Sh: Okay. So we -are voting on the names
13 H: Yeah. Yeah. I think so (pause) in effect. I mean there’s no separate category for it but
14 S: It’s part, it’s part of the whole package
15 Sh: Well it just
16 H: Part of the whole thing
17 S: Part part of the package
18 Sh: It seems to me that if we vote on the members of DAC [District Advisory Council] and
19 we vote on the members of our real estate task force, we certainly should be voting on
20 our superintendent search committee.
21 H: Wanna do it name by name or d’you wanna do it as a lump sum? Lump group.
22 Sh: Lump sum is fine.
23 H: Okay.
24 S: hh Move that we appoint the listed members to the task force that was approved by the
25 board at the last meeting.
26 H: I just hâ•‚ guess that would be just a friendly amendment (.) to the motion.
27 S: I stand corrected. That would be an amendment to the motion.
28 H: ºOkayº Great.
29 Sh: ºOkayº Now. Some questions for Mr. Ceruli, please? Um. how many searches for
30 superintendents have you conducted in the past?
31 C: Uh, wâ•‚
32 Sh: Approximately
33 C: Our firm is a, ah research and facilitation firm. So we have not conducted a ah
34 superintendent search. We’ve bâ•‚ been in involved with um um one particular search in
35 Denver. Uh but what we do are public process and research and so what were, ah what
36 we’ve offered to do here. And what we’ve done thus far is um put together the parts of
37 the public process that would uh uh accompany this um and all of the research and that
38 it whâ•‚ thâ•‚ it which essentially a search is. It is aâ•‚ it’s an effort to um aâ•‚ acquire the
39 information you need from these individuals. And so that’s what we would conduct.
40 Sh: So the answer is that you haven’t. ((laughs))
41 ((Audience laughter))
42 C: We, uh. That’s right. That is the answer.
Action-implicative discourse analysis 

43 Sh: Is that correct? ((laugh))

44 C: That is the answer. Correct.
45 Sh: That you have not uh supervised or organized or whatever you are doing for us a
46 superintendent’s
47 C: Right.
48 Sh: search committee ever before.
49 C: Uh we’ve conducted searches for the um the scientific and cultural facilities district uh
50 executive director that’s been with them for eight years. We uh assisted on the
51 conducting of the Great Outdoor Colorado search for their executive director, so we
52 have done executive director searches before. We haven’t done a search for a
53 superintendent specifically.
54 Sh: Uhâ•‚huh, I just wanted to clarify that. Um, I guess am going to have to vote against this.

For analytic purposes, we will divide the exchange into two sections: lines 1–28
and 29–54. If we were to interpret Shoemaker’s actions through the focal decision
– approval of the search process – we would likely “see” evidence of hidden agen-
das and the irrationality of much of the talk that goes on in decision-making
groups. However, if we assume that people and their talk are reasonable, attending
to legitimate problems, then what becomes visible?
In lines 1–28, Shoemaker questions the meaning of voting to approve the su-
perintendent search procedures. Although the most straightforward function of
questions is to seek information, questions frequently challenge and criticize
(Tracy 1997). In lines 2–4, where Shoemaker questions whether committee mem-
bers’ names are to be included in the vote, it seems possible that she is merely
seeking information. However, when she twice repeats the upshot of Hult’s answer
(“so we’re voting on the names,” lines 8 and 12) and then explicitly states why she
regards it as unreasonable not to specify the committee make-up, it becomes clear
that the “question” is a challenge. Hult’s response (line 21), in fact, acknowledges
Shoemaker’s criticism and offers a solution. Yet the choice she offers Shoemaker–
“name by name or lump group” – frames Shoemaker as unreasonable. In light of
the shared view that school board meetings were already too long, a proposal to
turn the approval process into a yes-no vote on 11 citizens, as well as all the other
pieces of the process, implicated Shoemaker negatively. Stated differently, Hult’s
comment humors and therein seeks to silence a difficult member. This humoring
is underscored by Shonkwiler’s proposal (lines 24–25) when he states, “Move that
we appoint the listed members to the task force that was approved by the Board at
the last meeting.” In essence, the President’s and Vice President’s comments frame
Shoemaker as haggling over something that has already been decided, and there-
fore implicitly wasting time and being unreasonable.
Shoemaker’s response, “lump sum is fine” (line 22), is interesting because it is
at odds with an implication established through her prior questioning – that there
 Karen Tracy and Robert T. Craig

was something troubling about the search committee’s make-up.€Allowing approv-

al of the committee to be bundled into the “search process” decision would seem to
be just the issue to which Shoemaker had earlier been objecting. Yet, at this juncture
in the meeting, she pursues the issue no further, shifting attention to other con-
cerns. How, then, might it be possible to see Shoemaker’s talk as reasonable?
Models of group interaction often assert competing notions of good member
behavior. On the one hand, members are encouraged to be vigilant and not to go
along with the majority to avoid conflict (Janis and Mann 1977). On the other,
they are expected to avoid actions that contribute to the negative reputation that
meetings have come to have in Western society: as ineffectual, a waste of time, te-
dious, and so forth (Schwartzman 1989). In groups that use majority rule rather
than consensus, problematizing the direction the group is going, but then
permitting the group to continue, is a reasonable strategy for a person whose posi-
tion is in the minority. Such a move allows the member to establish his or her
reservations and yet to avoid being cast as the group “problem.”
In the second half of the exchange, Shoemaker challenges consultant Ceruli’s
competence to be organizing the superintendent search process. In asking Ceruli
how many superintendent searches he had previously conducted (lines 29–30), and
then tacking on that it would be acceptable for Ceruli to offer an approximate
number (line 32), Shoemaker implies the reasonableness of expecting Ceruli’s firm
to have done a number of searches. In adding “approximately” to her initial ques-
tion formulation, Shoemaker’s question offers a “candidate answer” (Pomerantz 1988).
Approximation of a number makes sense if one is dealing with relatively large num-
bers, at least, say, 10 or 15. But, if it is expected that a person has done only one or
two searches, there is no need to ask for an approximate number. This is even more
the case if a questioner expects that a firm may have done no searches.
Ceruli’s nonfluent and rambling answer makes visible his awareness of the
implications of this question as well his own discomfort with those implications.
Although Ceruli tries to reframe the experience he does have, Shoemaker does not
accept his reframing. In summarizing the gist of Ceruli’s comment (line 40) as “the
answer is you haven’t [any experience]” she offers an unfriendly reading. Not only
does Shoemaker respond unsympathetically, but she also underscores it with her
follow-up questions and, thereby, forces Ceruli to acknowledge publicly and re-
peatedly that he has no experience conducting a superintendent search. From
Ceruli’s point of view, it is hard to imagine that he did not see Shoemaker as delib-
erately working to undermine him in a situation where the group (i.e., the board
majority) had already hired him.
If we raise the question concerning what purpose Shoemaker’s talk serves, a
function does become apparent. Shoemaker’s interrogation draws attention to the
fact that some persons, but not she, hired a consulting firm with questionable
Action-implicative discourse analysis 

competence. Furthermore, her pursuit of this issue strongly implies that she was
not part of that decision; either the decision occurred behind her back (because
the majority favored it and there was no need to get her input), or it was made
despite concerns she may have raised. Shoemaker’s comments, thus, construct a
version of recent events that make visible for citizens in the community (i.e., voters
in the upcoming election) that the board majority led by the president acted in a
high-handed and/or questionable manner.
Shoemaker went on to vote against the search committee composition, a posi-
tion that was decisively outvoted by other board members. But, although in this
immediate decision, Shoemaker lost – her arguments did not lead the group to
change direction – a negative assessment of her talk is not warranted. When we
look at this deliberation process in a larger frame, her talk on this occasion
functioned to shape longer-term outcomes. In the subsequent election a key issue
became the reasonableness of the incumbents’ conduct in board meetings, both
with each other and in dealing with members of the public (Craig and Tracy 2005;
Tracy 1999; Tracy and Muller 2001). Were members of the board majority acting
democratically with each other and the larger public? Were they exercising good
judgment in the decisions with which they were entrusted? This interactive seg-
ment, as well as others like it, helped create a community impression that the board
leaders were acting “undemocratically.” In the election that followed, the president
and the two other majority coalition members running for election were voted out
of office, and Shoemaker became president.
Arriving at a reconstruction of the problems, conversational techniques, and
situated ideals of a practice, such as school board meetings, requires observing and
reflecting on multiple instances and kinds of interaction from the viewpoints of
various categories of participants. A developed reconstruction needs to attend to
the larger interactional scene. Based on this single analysis, we would highlight the
following. First, using AIDA makes visible a problem. When elected officials in
community groups know their opinions are in the minority, they face a difficulty.
As elected officials, brought to power though a process of voting, they are expected
to show respect for democratic decision-making. They are also expected to exert
influence and shape policies and decisions in a direction consistent with the views
they advocate. How to do this is a major challenge when members know they will
be outvoted. Shoemaker’s moves, analyzed above, point to some of the conversa-
tional techniques that persons in this position can and do use. In essence,
Shoemaker’s way of posing questions and reformulating others’ answers func-
tioned to challenge the good judgment and fairness of the board majority while
displaying her own commitment to democratic process (i.e., her willingness to be
outvoted). Simply put, when elected officials cannot affect the immediate decision,
their talk can be employed to shape the larger decision-making context.
 Karen Tracy and Robert T. Craig

Finally, analysis of a single interaction, such as we have done here, is inade-

quate for developing a situated ideal of school board meeting conduct. Construc-
tion of situated ideals, an important aim within AIDA, necessitates looking at
multiple instances of a kind of interaction along with collecting and studying par-
ticipant interviews and institutional documents. In the contextual crevices – the
spaces between what people actually do and how they evaluate their own and oth-
ers’ actions within the practice itself, in interviews, and in institutional documents
– are to be found the raw materials for reconstructing a situated ideal. Based on
study of materials from the larger project (Craig and Tracy 2005; Tracy 1999, 2008,
2010; Tracy and Craig 2003; Tracy and Muller 2001), the exchange offers a glim-
mer of the ideal that participants seem to hold. A belief in the goodness of “democ-
racy/democratic process,” seems to represent an ideal for school board meetings.
Participants’ situated ideals, however, differ from those that philosophers and po-
litical theorists stake out in conceptual essays. The ideal for school board meetings
in this study is one that recognizes the value of extended talking and consensus-
decision-making, and, at the same time, voting and majority rule to settle differ-
ences (Mansbridge 1980). It is an ideal that assumes the desirability of elected of-
ficials exercising their judgment and at the same time assumes that elected officials
should represent their constituents (Schudson 1998). In addition, the situated ide-
al is one that sees formal rules as the cornerstone of fairness but also seems to
recognize that rules, at least on particular occasions, can be impediments to “real
democracy.” Stated a bit differently, the situated ideal for school board meeting
conduct, reconstructed from what participants say, espouse, and criticize, is a di-
lemmatic ideal. It is not philosophically coherent, but it is pragmatically useful and
defensible. The situated ideal for these American school board meetings identifies
competing criteria for assessing conduct and leaves the selection of applicable cri-
teria for participant argument in the interaction moment. The situated ideal shapes
and constrains conduct and at the same time is a resource for justification and cri-

Studying interaction: A disciplinary conversation

Disciplinary discourses2

Academic disciplines are discourse communities. A discipline is “a conversational

community with a tradition of argumentation” (Shotter 1997:€42) that participates
in broader communities of disciplines with their own traditions of argumentation.

2. Portions of this section have been adapted from Craig (2008b).

Action-implicative discourse analysis 

Modern disciplines do not represent eternally fixed categories of knowledge; rath-

er, they are institutional objects that emerge and evolve within this ongoing “con-
versation of disciplines” (Craig 2008b). Each discipline is constituted in discourse
in its own particular way, in part by being routinely contrasted against neighbor-
ing disciplines. As Godzich (1986: x) commented, “the mutual relation of the dis-
ciplines is never one of autonomy or of heteronomy, but some sort of complicated
set of textual relations that needs to be unraveled in each instance.” This process
can be illustrated briefly in the cases of three disciplines central to our present
discussion: sociology, linguistics, and communication.
The “sociological perspective” of sociology can be defined only against a back-
ground that includes traditions of argumentation about sociology’s differences
from history, psychology, anthropology, economics, and other disciplines. Classic
writings in sociology assert the uniqueness and importance of a sociological per-
spective with compelling intellectual force, but sociologists have always disagreed
among themselves about the meaning and value of such a perspective. The socio-
logical tradition can be read as a series of arguments about how much and in what
ways sociology differs from other disciplines. Perspectives within sociology can be
described as economic, cultural, historical, political, psychological, and so forth. If
the idea of a sociological perspective were no longer felt to be worth discussing,
even among sociologists, then the conversation would break up or turn to other
topics and sociology would cease to exist as a meaningful discipline. However un-
likely this scenario may seem, sociologists have sometimes expressed the fear that
something like it may be happening (Brewer 2007; Halliday and Janowitz 1992;
Osborne and Rose 1997; Turner and Turner 1990).3
The disciplinary identity of linguistics is equally agonistic (Badger 2006; Harris
1993) and negotiable with reference to other fields such as anthropology, cognitive
science, communication studies, sociology, and rhetoric (e.g., Bucholtz and Hall
2007; Leith 1994; Ochs, Schegloff and Thompson 1996; Tusting and Maybin 2007).
Of particular importance, the discourse community of linguistics has been shaped
by a debate over “autonomous linguistics,” the idea that language is an autonomous
structure independent of social behavior generally (Newmeyer 1986). Although
many linguists, and especially those who identify as sociolinguists or anthropo-
logical linguists, would not agree with this position, it is a position that cannot be
ignored. For linguists, the position that language is not autonomous is controver-
sial and must be argumentatively defended. In sociology and communication, in
contrast, the interpenetration of language and the social is generally taken for

3. On the history and disciplinary identity of sociology, see also: Collins (1985), Lepenies
(1986), Levine (1995), Mazlish (1989), and Ross (1991).
 Karen Tracy and Robert T. Craig

granted. Few sociologists or communication scholars would dispute such a claim.

For them, the arguable issues concern how language and society are connected.
Our own discipline of communication, although notoriously heterogeneous
and not yet fully institutionalized, is not without its own traditions of argumentation
and identity negotiations vis-à-vis other fields (Craig 1989, 1999, 2008a, 2008b; see
also Buzzanell and Carbaugh 2009; Donsbach 2006). Tracy (2001) described sev-
eral features as characterizing a communicative approach to interaction.4 A first
feature is the prominence of strategy and audience as key terms in the analysis of
interactional moments. Second is the attention given to problematic interaction,
including persuasive and conflict situations, whether the conflicts are between
people or are among an actor’s multiple situated aims. Third, seeing talk as a form
of practical and moral action has deep roots in the communication field (going
back to the ancient verbal arts of rhetoric and dialectic), even while it is important
note that it is by no means the dominant tradition at the present time. Finally,
communication research about interaction has tended to use a more argumenta-
tive writing style than is typically used in other social science disciplines
(Tracy 1988). This greater amount of argumentative discourse accomplishes many
things, but one important one is to discursively enact communication as a hetero-
geneous discipline in which a wide range of assumptions cannot be taken for
granted but must be argumentatively defended in each publication.

Interactional sociolinguistics, conversation analysis,

and AIDA in Their disciplinary contexts

IS originated in linguistics, CA in sociology, and AIDA in communication. Each

of the three disciplines is a distinct discourse community with a particular intel-
lectual-institutional history that forms a background for judging claims that can
be straightforwardly asserted or assumed in research versus claims that are contro-
versial and must be justified by explicit arguments appropriate to that discipline.
Until the last few decades, as Schegloff (Cmejrkova and Prevignano 2003a) has
noted, social scientists did not regard ordinary interaction as deserving systematic
study. To merit study, interaction needed to be either defective (e.g., mental retarda-
tion or schizophrenia) or seen as directly related to profit-making (e.g., salesmanship,
negotiation). This lack of interest in ordinary talk no longer persists. All kinds of
informal and institutional interaction have been or are being studied by discourse

4. The essay, addressing linguists, included 5 features that distinguish a communicative kind
of discourse analysis. The first, which applies to discourse analysis but not to the study of inter-
action, was a preference for discourse that is interactive (i.e., talk) rather than written texts.
Action-implicative discourse analysis 

scholars. The interactional-conversational-linguistic turn of so many social science

disciplines has been impelled by diverse forces, but certainly two important ones
have been the discipline-challenging moves of Gumperz and Hymes (1972) in lin-
guistics and anthropology, and of Sacks (1992) in sociology.
In the early 1970s, Gumperz and Hymes, working at the intersection of lin-
guistics and anthropology, developed the ethnography of speaking, an approach
that challenged the dominant traditions of both disciplines. In anthropology at
that time, little attention was given to speaking and language. How, Gumperz and
Hymes asked, can culture be understood if attention is not given to how people
speak to others in the events that compose their lives? IS, the second stream that
flowed from the ethnography of speaking tradition, attended more to the field of
linguistics, and especially the subfield of sociolinguistics. To understand verbal
exchanges, Gumperz (1982b: 1) argued, requires “knowledge and abilities which
go considerably beyond the grammatical competence we need to decode short
isolated messages.” In framing the proposal this way, interactional sociolinguistics
can be seen as centrally arguing with fellow linguists. The proposal takes for grant-
ed a central goal of linguistics – to explicate knowledge underlying “language” –
but disagrees with many linguists as to where that knowledge is to be grounded.
Not in the grammatical or semantic properties of the code (linguistics, proper),
nor in social and language variables detached from interaction (sociolinguistics,
e.g., Labov 1966); instead, Gumperz claimed, the most interesting component of
language knowledge is to be found in social interaction.
The idea of contextualization cue, perhaps Gumperz’s most important idea,
necessitated attending to linguistically peripheral information (e.g., prosody, dis-
course particles) to develop a good picture of what situated meanings were being
made and how interactional problems could arise (Levinson 2003). Not only did
Gumperz’s discipline affect his argument but his area of specialization also did.
That Gumperz was an anthropological linguist influenced the interaction scenes he
selected for study. Although a variety of interaction genres have been studied
(e.g., Gumperz 1982a), they have almost entirely involved persons of different
speech communities. In pursuit of understanding this kind of complex interac-
tion, interactional sociolinguistic studies have drawn upon interviews, analysis of
text genres such as African American preaching (Gumperz 1982c), and, on occa-
sion, simulations (Akinnaso and Ajirotutu 1982). In reflecting on the intellectual
contributions of Gumperz, Levinson (2003:€ 32) noted, “Gumperz’s analyses of
conversation have nothing of the theoretical cleanliness to be found e.g.€in conver-
sational analysis. His tools are eclectic and the toolbox cluttered.” Gumperz’s stu-
dents (e.g., Tannen 1986) and grand-students (e.g.€Yamada 1992) have continued
the tradition of using multiple means to understand situated sense-making. It was
within the disciplinary context of linguistics, particularly in the American academic
 Karen Tracy and Robert T. Craig

scene, that Gumperz and students’ focus on interaction, and moreover, a func-
tional approach to it, was radical.
In communication, taking a functional approach is mainstream. For communica-
tion scholars, function and its close relative, strategy, are taken-for-granted key con-
cepts to use in studying social life (Craig and Tracy 1983). In contrast to that of CA,
Gumperz’s influence in the field of communication has been relatively limited. An un-
sympathetic reading of his work could frame him as asserting no more than a discipli-
nary commonplace in communication. That communicative functions are important is
an unquestioned assumption in the discourse community of communication studies.
Although IS and Gumperz’s work are not synonymous, for many purposes
they can be treated as alternative forms of reference. This is not the case with CA
and Schegloff ’s work. CA is a broad enterprise. Many scholars internationally and
across disciplines currently would define themselves as doing CA or being strongly
influenced by “it.” Yet, as CA has been taken up in locations outside the US and in
disciplines outside sociology, it has to some degree been refashioned. In each case,
CA has merged with other impulses that are specific to the academic tradition
(US, European) and the particular discipline.
“CA” in communication (e.g.€Beach 1996; Glenn, LeBaron and Mandelbaum
2003), linguistics (Ochs, Schegloff and Thompson 1996), or feminist psychology
(Kitzinger and Frith 1999; Speer 2002) – to identify only three of the most obvious
alternatives – each has a distinctly different flavor from the kind of CA that
Schegloff does. Moreover, in contrast to what is stated in the discussion about CA
with Schegloff that occurred in the Prevignano and Thibault (2003b) volume, in
many intellectual corners (e.g., Hutchby and Wooffitt 1999), Harvey Sacks is treat-
ed as the originator of CA. This way of framing CA is especially visible in work that
builds on Sacks’ analyses of membership terms (Fitzgerald and Housely 2002;
Hester and Eglin 1997). Sometimes this work is treated as a kind of CA; at other
times it is treated as something entirely different and labeled “membership catego-
rization analysis,” an approach to be contrasted with CA.
What is to be treated as inside or outside of CA is by no means obvious. When,
for instance, does a study become CA-influenced rather than a piece of CA schol-
arship proper? Is any study that goes beyond claims that can be grounded in the
recipient’s uptake not a CA study? Is all the work done by visible conversation
analysts actually CA? For instance, would the quantitative coding study of ques-
tions in US presidential press conferences conducted by CA scholars Clayman and
Heritage (2002), be considered CA? How much ethnographic work can a CA
scholar do and how can it be used in interpretation of an interactional scene before
the work’s CA status is called into question? Are studies that pursue issues such as
gender inequality through a close look at conversations that have been transcribed
using the Jeffersonian transcription system CA research?
Action-implicative discourse analysis 

We do not have answers to these questions. The point we wish to highlight is

that as CA has become widely influential, its boundaries have become less clear. In
this more intellectually diffuse landscape, Schegloff can be seen as anchoring a
position that emphasizes a structural view. A leap from interaction structure to
language structure is a small one. That this is so, we believe, accounts for the
spreading attractiveness of Schegloff ’s version of CA among linguists. His view of
CA meshes with assumptions about structure and function familiar to the dis-
course community of linguistics. Interestingly, other CA scholars (e.g., Drew 1992,
1998; Pomerantz 1989/90), who build on Sacks’ less structural ideas, seem to have
been somewhat less influential in linguistics but more influential in functionally-
focused disciplines like psychology and communication.
At its inception, CA both sought to address a key sociological issue and chal-
lenge the position most sociologists were taking toward it. CA developed a way to
understand social structure and offered a radical critique of the macro, “top-down”
kind of answers that were and continue to be dominant in sociology. Initially,
studies of interaction in CA focused on conversations among family and friends,
often on the telephone, or among juvenile delinquents in treatment, suicide hot-
lines (Sacks 1992), or exchanges with the police (Zimmerman 1984), all interac-
tion sites traditionally connected to sociology. Today, CA and CA-influenced stud-
ies can be found of all kinds of interaction. Much more than interactional
sociolinguistics, CA has succeeded in tearing loose from its disciplinary mooring.
As it has done so, though, its character has become fuzzier.

The distinct contribution of AIDA

AIDA shares with IS a concern with problematic interaction. But, the kinds of
problems to which AIDA and IS give attention are different. Operating within a
linguistic tradition, Gumperz has built an analytic frame on the opposition be-
tween central and peripheral linguistic information (Levinson 2003). Lexical and
syntactic kinds of information are treated as focal, whereas prosody, the use of
discourse particles, and several other features are seen as background language
information. Gumperz ‘s research has highlighted the problems that occur within
language processes (e.g., vocal intonation patterns) that are largely out of aware-
ness. IS, as is true of culture-attentive discourse approaches generally, can help
people recognize that moments of interactional trouble arise from reasonable but
culturally-specific meaning-cueing practices. In contrast, AIDA is primarily inter-
ested in institutional problems that arise among nationally and ethnically similar
persons. Rather than cultivating better understanding of subtle out-of-awareness
practices, AIDA seeks to make visible discourse strategies that can be named, re-
flected upon, and adopted by participants to make their practice work better.
 Karen Tracy and Robert T. Craig

A second distinctive feature of AIDA is its metatheoretical stance. Instead of

pursuing the building of a descriptive science of interaction as Schegloff espouses
for CA, or exposing ideology and social inequity as CDA aims to do, AIDA’s
approach to the study of interaction is guided by its practical theory view of re-
search. AIDA aims to develop practically useful and morally defensible reconstruc-
tions of interactional problems, conversation techniques, and situated ideals of a
variety of communicative practices. With an end goal of enabling people to better
manage the very particular communication practice that they care about, AIDA, as
its name suggests, is a discourse approach that aims to be action implicative.


Agne, Robert. 2007. “Reframing practices in moral conflict: interaction problems in the nego-
tiation standoff at Waco.” Discourse & Society 18: 549–578.
Akinnaso, F. Niyi and Ajirotutu, Cheryl. S. 1982. “Performance and ethnic style in job inter-
views.” In Language and Social Identity, John J. Gumperz (ed), 119–144. Cambridge:
Cambridge University Press.
Badger, Richard G. 2006. “Investigating agonism in linguistics.” Journal of Pragmatics 38: 1442–1456.
Beach, Wayne A. 1996. Conversations about Illness: Family Preoccupations with Bulimia. Mahwah
NJ: Lawrence Erlbaum.
Billig, Michael, Condor, Susan, Edwards, Derek, Gane, Mike, Middleton, David and Radley,
Alan. 1988. Ideological Dilemmas. London: Sage.
Blum-Kulka, Shoshona, House, J., & Kasper, G. 1989. Cross-Cultural Pragmatics: Requests and
Apologies. Norwood, NJ: Ablex.
Brewer, John D. (ed.). 2007. “Sociology and its strange ‘others’” [special issue]. History of the Hu-
man Sciences 20 (2): 1–175.
Brown, Penelope and Levinson, Stephen C. 1987. Universals in Language Usage: Politeness Phe-
nomena. Cambridge: Cambridge University Press.
Bucholtz, Mary and Hall, Kira. 2008. “All of the above: New coalitions in sociocultural linguis-
tics.” Journal of Sociolinguistics 12: 401–431.
Buzzanell, Patrice and Carbaugh, Donal. (eds.). (2009). Distinctive Qualities of Communication
Research. New York: Routledge.
Cameron, Deborah. 2001. Working with Spoken Discourse. London: Sage.
Cmejrkova, Svetla and Prevignano, Carlo L. 2003. “On conversation analysis: An interview with
Emanuel A. Schegloff.” In Discussing Conversation Analysis: The Work of Emanuel A. Sche-
gloff, Carlo L. Prevignano and Paul J. Thibault (eds), 11–56. Amsterdam: John Benjamins.
Collins, Randall. 1985. Three Sociological Traditions. New York: Oxford University Press.
Craig, Robert T. 1989. “Communication as a practical discipline.” In Rethinking Communication;
Volume 1; Paradigm Issues, Brenda Dervin, Lawrence Grossberg, Barbara J. O’Keefe and
Ellen Wartella (eds), 97–122. Newbury Park CA: Sage.
Craig, Robert T. 1996. “Practical-theoretical argumentation.” Argumentation 10: 461–474.
Craig, Robert T. 1999. “Communication theory as a field.” Communication Theory 9: 119–161.
Action-implicative discourse analysis 

Craig, Robert T. 2008a. “Communication as a field and discipline.” In International Encyclopedia

of Communication, Wolfgang Donsbach (ed.), Vol. II, pp.€ 675–688. Oxford and Malden,
MA: Blackwell.
Craig, R. T. 2008b. “Communication in the conversation of disciplines.” Russian Journal of Com-
munication 1: 7–23.
Craig, Robert T. and Tracy, Karen (eds.). 1983. Conversational Coherence: Form, Structure and
Strategy. Beverly Hills, CA: Sage.
Craig, Robert T. and Tracy, Karen. 1995. “Grounded practical theory: The case of intellectual
discussion.” Communication Theory 5: 248–272.
Craig, Robert T. and Tracy, Karen. 2005. “‘The issue’ in argumentation practice and theory.” In
The Practice of Argumentation, Frans H. van Eemeren & Peter Houtlosser (eds), 11–28.
Amsterdam & Philadelphia: John Benjamins.
Drew, Paul. 1992. “Contested evidence in courtroom cross-examination: The case of a trial for
rape.” In Talk at Work, Paul Drew and John Heritage (eds), 470–520. Cambridge: Cambridge
University Press.
Drew, Paul. 1998. “Complaints about transgressions and misconduct.” Research on Language
and Social Interaction 31: 295–325.
Fairclough, Norman. 2001. “The discourse of new Labour: Critical discourse analysis.” In Dis-
course as Data, Margaret Wetherell, S. Taylor and S. J. Yates (eds), 229–266. London: Sage.
Fairclough, Norman and Wodak, Ruth. 1997. “Critical discourse analysis.” In Discourse as Social
Interaction, Teun A. van Dijk (ed), 258–284. London: Sage.
Fitzgerald, R. and Housley, W. 2002. “Identity, categorization and sequential organization: The se-
quential and categorial flow of identity in a radio phone-in.” Discourse & Society 13: 579–602.
Foucault, Michel. 1972. The Archeology of Knowledge. New York: Pantheon Books.
Gee, James P. 1999. An Introduction to Discourse Analysis: Theory and Method. London:
Glenn, Phillip, LeBaron, Curtis D. and Mandelbaum, Jenny. 2003. Studies in Language and Social
Interaction. Mahwah, NJ: Lawrence Erlbaum.
Godzich, Wlad. 1986. “Forward: The further possibility of knowledge.” In Michel de Certeau, Het-
erologies: Discourse on the Other, vii–xxi. Minneapolis, MN: University of Minnesota Press.
Goffman, Erving. 1974. Frame Analysis: An Essay on the Organization of Experience. Cambridge,
MA: Harvard University Press.
Gumperz, John J. and Hymes, Dell. 1972. Directions in Sociolinguistics: The Ethnography of Com-
munication. New York: Holt, Rinehart and Winston.
Gumperz, John J. (ed.) 1982a. Language and Social Identity. Cambridge: Cambridge University Press.
Gumperz, John J. 1982b. Discourse Strategies. Cambridge: Cambridge University Press.
Gumperz, John J. 1982c. “Fact and Inference in Courtroom Testimony.” In Language and Social
Identity, John J. Gumperz (ed), 163–194. Cambridge: Cambridge University Press.
Gumperz, John J. 2003. “Response essay.” In Language and interaction: Discussions with John J.
Gumperz, Susan L. Eerdmans, Carlo L. Prevignano and Paul J. Thibault (eds), 105–126.
Amsterdam: John Benjamins.
Halliday, Terence C. and Janowitz, Morris. (eds.). 1992. Sociology and Its Publics: The Forms and
Fates of Disciplinary Organization. Chicago and London: University of Chicago Press.
Harris, Roy A. 1993. The Linguistics Wars. Oxford and New York: Oxford University Press.
Hester, Stephen and Eglin, Peter (eds.). 1997. Culture in Action: Studies in Membership Categori-
zation Analysis. Lanham, MD: University Press of America.
 Karen Tracy and Robert T. Craig

Hutchby, Ian and Wooffitt, Robin. 1998. Conversation Analysis: Principles, Practices and Applica-
tions. Malden, MA: Polity Press.
Hymes, Dell. 1974. Foundations in Sociolinguistics: An Ethnographic Approach. Philadelphia:
University of Pennsylvania Press.
Janis, Irving L. 1982. Groupthink. Dallas, TX: Houghton Mifflin.
Jasinski, James. 2001. Sourcebook on Rhetoric. Thousand Oaks, CA: Sage.
Kitzinger, Celia and Frith, Hannah. 1999. “Just say no? The use of conversation analysis in devel-
oping a feminist perspective on sexual refusal.” Discourse & Society 10: 293–316.
Labov, William. 1966. The Social Stratification of English in New York City. Washington, DC:
Center for Applied Linguistics.
Leith, Dick. 1994. “Linguistics: A rhetor’s guide.” Rhetorica 12: 211–226.
Lepenies, Wolf. 1988. Between Literature and Science: The Rise of Sociology (R. J. Hollingdale,
trans.). Cambridge: Cambridge University Press.
Levine, Donald N. 1995. Visions of the Sociological Tradition. Chicago: University of Chicago Press.
Levinson, Stephen C. 2003. “Contextualizing ‘contextualization cues’.” In Language and Interac-
tion: Discussions with John J. Gumperz, Susan L. Eerdmans, Carlo L. Prevignano and Paul J.
Thibault (eds), 31–40. Amsterdam: John Benjamins.
Mansbridge, Jane. 1980. Beyond Adversary Democracy. New York: Basic Books.
Mazlish, Bruce. 1989. A New Science: The Breakdown of Connections and the Birth of Sociology.
New York: Oxford University Press.
Newman, D. L., & Brown, R. D. 1992. “Patterns of school board decision making: Variations in
behavior and perceptions.” Journal of Research and Development in Education 26: 1–6.
Newmeyer, Frederick J. 1986. The Politics of Linguistics. Chicago: University of Chicago Press.
Ochs, Elinor, Schegloff, Emanuel A. and Thompson, Sandra A. (eds). 1996. Interaction and
Grammar [Studies in Interactional Sociolinguistics 13]. Cambridge and New York:
Cambridge University Press.
Osborne, Thomas and Rose, Nikolas. 1997. “In the name of society, or three theses on the his-
tory of social thought.” History of the Human Sciences 10: 87–121.
Pomerantz, Anita. 1988. “Offering a candidate answer: An information seeking strategy.” Com-
munication Monographs 55: 360–273.
Pomerantz, Anita. 1989/90. “Constructing skepticism: Four devices used to engender the audi-
ence’s skepticism.” Research on Language and Social Interaction 22: 293–313.
Prevignano, Carlo L. and Thibault, Paul J. 2003a. “Continuing the discussion with John J. Gump-
erz.” In Language and Interaction: Discussions with John J. Gumperz, Susan L. Eerdmans,
Carlo L. Prevignano and Paul J. Thibault (eds), 149–162. Amsterdam: John Benjamins.
Prevignano, Carlo L. and Thibault, Paul J. (eds.). 2003b. Discussing Conversation Analysis: The
Work of Emanuel A. Schegloff. Amsterdam: John Benjamins.
Ross, Dorothy. 1991. The Origins of American Social Science. Cambridge and New York:
Cambridge University Press.
Sacks, Harvey. 1992. Lectures on Conversation. Cambridge, MA: Blackwell.
Schegloff, Emanuel A. 1992. “In another context.” In Rethinking Context: Language as an Interac-
tive Phenomenon, Alessandro Duranti and Charles Goodwin (eds), 191–227. Cambridge
and New York: Cambridge University Press.
Schegloff, Emanuel. A. 1998. “Positioning and interpretative repertoires: Conversation analysis
and post-structuralism in dialogue – Reply to Wetherell.” Discourse & Society 9: 413–416.
Schiffrin, Deborah. 1994. Approaches to Discourse. Oxford: Blackwell.
Schudson, Michael. 1998. The Good Citizen. New York: The Free Press.
Action-implicative discourse analysis 

Schwartzman, Helen B. 1989. The Meeting: Gatherings in Organizations and Communities. New
York: Plenum Press.
Shotter, John. 1997. “Textual violence in academe: On writing with respect for one’s others.” In
Transgressing Discourses: Communication and the Voice of the Other, Michael Huspek and
Gary P. Radford (eds), 17–46. Albany: SUNY Press.
Speer, Susan A. 2002. “What can conversation analysis contribute to feminist methodology?
Putting reflexivity into practice.” Discourse & Society 13: 783–803.
Tannen, Deborah. 1984. Conversational Style: Analyzing Talk among Friends. Norwood, NJ: Ablex.
Tannen, Deborah. 1993. Framing in Discourse. New York: Oxford University Press.
Tracy, Karen. 1988. “A discourse analysis of four discourse studies.” Discourse Processes 11: 243–259.
Tracy, Karen. 1997. Colloquium: Dilemmas of Academic Discourse. Norwood, NJ: Ablex.
Tracy, Karen. 1999. “The usefulness of platitudes in arguments about conduct.” In Proceedings of
the Fourth International Conference of the International Society for the Study of Argumentation,
Frans H. van Eemeren, Rob Grootendorst, J. Anthony Blair and Charles A. Willard (eds.),
799–803. Amsterdam: SicSat.
Tracy, Karen. 2001. “Discourse analysis in communication.” In Handbook of Discourse Analysis,
Deborah Schiffrin, Deborah Tannen and Heidi Hamilton (eds), 725–749. Oxford: Blackwell.
Tracy, Karen. 2005. “Reconstructing communicative practices: Action-implicative discourse
analysis.” In Handbook of Language and Social Interaction, Kristine Fitch and Robert
Sanders (eds.), 301–319. Mahwah, N.J.: Lawrence Erlbaum.
Tracy, Karen. 2008. “‘Reasonable hostility’: Situation – appropriate face attack.” Journal of Polite-
ness Research: Language, Behaviour, Culture 4: 169–191.
Tracy, Karen. 2010. Challenges of Ordinary Democracy: A Case Study of Deliberation and Dissent.
University Park, PA: Pennsylvania State University Press.
Tracy, Karen & Ashcraft, Catherine. 2001. “Crafting policies about controversial values: How
wording disputes manage a group dilemma.” Journal of Applied Communication Research
29: 297–316.
Tracy, Karen and Craig, Robert T. 2003. Communicative Ideals within Democracy: Conduct Ar-
guments at School Board Meetings. Paper presented at 8th International Pragmatics Confer-
ence, Toronto.
Tracy, Karen and Muller, Heidi. 2001. “Diagnosing a school board’s interactional trouble: Theo-
rizing problem formulating.” Communication Theory 11: 84–104.
Tracy, Karen and Standerfer, Christina. 2003. “Selecting a school superintendent: Sensitivities in
group deliberation.” In Group Communication in Context: Studies of Natural Groups,
Lawrence Frey (ed.), 109–134. Mahwah, NJ: Lawrence Erlbaum.
Tusting, Karin and Maybin, Janet. 2007. “Linguistic ethnography and interdisciplinarity: Open-
ing the discussion.” Journal of Sociolinguistics 11: 575–583.
Turner, Sephen Park and Turner, Jonathan H. 1990. The Impossible Science: An Institutional
Analysis of American Sociology. Newbury Park: Sage.
van Dijk, Teun A. (ed.). 1997a. Discourse as Social Interaction. London: Sage.
van Dijk, Teun A. (ed.). 1997b. Discourse as Structure and Process. London: Sage.
Yamada, Haru. 1992. American and Japanese Business Discourse: A Comparison of Interactional
Styles. Norwood, NJ: Ablex.
Zimmerman, Don H. 1984. “Talk and its occasion: The case of calling the police.” In Meaning,
Form and Use in Context, Deborah Schiffrin (ed.), 210–228. Georgetown: Georgetown
University Press.
Healthcare interaction as an expert
communicative system
An activity analysis perspective

Srikant Sarangi
Health Communication Research Centre, Cardiff University

In this paper I argue that interaction – as a communicative system – is central

not only to forms of everyday social encounters but also to professional-
client relationships in institutional settings. The role language plays in these
interactional trajectories has to go beyond the dichotomous language as system
and language as behaviour divide, and focus on the phenomenon of interaction
itself, but not necessarily reducing interaction to language practice. This means
that as interaction analysts we can utilise linguistic insights when interpreting
professional-client encounters, but not be limited by them. Healthcare interaction,
as an institutional and professional site, can be seen as an expert communicative
system, with complex variations – along different modalities – reflecting
different specialities and participant frameworks. I focus here on the linguistic
dimension in the counselling domain, where communicative expertise is to be
conceptualised in terms of hybrid interactional competencies for the management
of different initiation-response frames, including aspects of uncertainty, risk, self-
and other-initiated diagnostic and prognostic scenarios. Interaction analysis –
what I refer to as activity analysis – should take as its starting point the structural,
interactional and thematic maps of whole encounters, while aligning with the
agenda of the professionals and the clients in a given setting in order to make any
findings uptake-oriented in practically relevant ways.

1. Introduction: Expertise as mediated knowledge-in-interaction

Expertise, for many, equates with ownership of knowledge, or declarative knowl-

edge in a content-intensive sense. According to Stehr (1994), we live in ‘knowledge
societies’, where experts exert knowledge-based power in all aspects of our social
lives. But exactly what counts as expert knowledge and what relationship
professionals as experts establish with available knowledge systems is open to
 Srikant Sarangi

debate. Over the years, bureaucracy as a rule-governed expert system has given
way to knowledge-based expertise with the attendant growth in professionalism,
although professional power and expertise are routinely monitored through state
institutions (Johnson 1972, 1993; Larson 1977; Freidson 1994). Freidson (1994:€137)
believes that ‘bureaucratic organisation is assumed to be antithetical to the free-
dom of activity traditionally imputed to the professional’. However, in many do-
mains, e.g., education, healthcare, social welfare, the professional groups are close-
ly embedded and governed within an institutional/organisational frame, thus
risking their independence and credibility. Policy-level changes at the institutional
and organisational level can potentially transform the everyday practices of pro-
fessionals, including their interactional trajectories with clients. In the healthcare
setting, for instance, new regulatory practices such as clinical governance, evi-
dence-based practice, patient-centredness, shared decision-making will no doubt
have epistemological and ontological ramifications regarding the status of expert
knowledge and power, with potential communicative consequences at the interac-
tional level for a given professional-client encounter.
Shils (1968) draws a distinction between experts and intellectuals: while the
latter are preoccupied with general knowledge, experts deal with specialised
knowledge. However, Merton (1957:€209) captures this difference differently: ‘in-
tellectuals devote themselves to cultivating and formulating knowledge’ whereas
experts are more interested in transmitting and applying that knowledge. Moving
away from such narrow dichotomies, Stehr (1994) characterises professional ex-
perts as both ‘knowledge-bearing’ and ‘knowledge-disseminating’ agents. He ar-
gues that in the process of disseminating knowledge, professionals as experts affect
the very knowledge base they mediate. In stressing that their function is not a pas-
sive one, Stehr (1994:€186) writes:
The knowledge these occupations employ is not, under most circumstances, di-
rectly of their creation. That is, these occupations serve as mediator between the
knowledge producers and the knowledge users, between those who create a ca-
pacity for action and those whose job it is to take action.

Stehr’s observation above underlines the significance of situated interaction which

provides a platform for both knowledge-dissemination and knowledge-bearing.
In our contemporary society, the intellectual-expert dichotomy in terms of
knowledge orientations becomes further complicated with the rise of technology-
based expert systems.1 Many of the professional-client encounters – in healthcare,
in education, in other public and private sectors such as inland revenue, banks,

1. In the healthcare context, expert systems may include technologies such as X-ray proce-
dure, laboratory-based tests, software-assisted risk assessments as well as patients’ case records,
official forms and certificates.
Healthcare interaction as an expert communicative system 

insurance firms – are now managed through technological know-how character-

ised by rule-based calculations and algorithms. The mediator role of the profes-
sional expert is potentially under threat when his/her activities are constrained by
the kind of knowledge generated through available expert systems. In a sense, the
very existence of an expert system undermines the expertise of the profession,
individually and collectively. To quote Illich (1977:€ 33), ‘as techniques multiply
and become more specific, their use often requires less complex judgements and
skills’. Generally, expert systems are targeted at the so-called non-experts requiring
explicit rule-following, which is rather different from what professional experts
are expected to do when dealing with clients. Professional experts work with a set
of rules, but these rules are often mediated through their experience and the de-
mands of the particularities of the case in hand. While advances in science and
technology (as expert systems) have direct consequences for what constitutes
knowledge as well as the authority of the professional experts, the tensions are
once again manifest at the communicative level. The communicative relationship
between clients, experts and expert systems is thus in a constant flux.
New experts and new expert systems are continually emerging, especially in
the area of healthcare delivery. Consider the case of nurse-led telephone-mediated
information/advice service in the UK (NHS Direct) for people with minor con-
cerns who may not need to attend a clinic physically to consult their General
Practitioner (GP). At a political, institutional level, this service is a means of cut-
ting waiting time on the part of an organisation that is under considerable stress
– financially and in terms of human resources. However, as far as the nursing pro-
fession is concerned, it appears to be a transfer of expertise from General
Practitioners to nurses. Under this new system, nurses can interact with callers/
patients directly without being overshadowed by the presence of a GP. In actual
fact, it turns out to be a case of trading places: the expertise of the GP is replaced
by an impersonal, institutional expert system. When dealing with patients, the
nurse professionals are obliged to routinely and systematically follow algorithms
in conducting their people-processing activity as they lack the institutional power
to diagnose and prescribe. In interactional terms, this is a complex expert ma-
noeuvre as nurses have to manage a three-party encounter, especially when the
caller is not the actual patient needing intervention. This is further complicated at
a procedural level as the nurse professional has to decide which aspects of the in-
formation exchange do or do not fit within the dictates of the software-based ex-
pert system. What we have here is a situation where the interaction/participation
framework is constrained by derivative knowledge, which does not allow nurse
professionals to draw on their own experience and expertise as and when they see
relevant. Remotely accessible expert systems can thus marginalise the expert-in-flesh
 Srikant Sarangi

and discount the complex nature of knowledge that professionals socialise into
and have access to in their everyday communicative practices.

1.1 Access to expertise and the ‘lay expert’

‘The client comes to the professional because he has met a problem which he can-
not himself handle’ (Hughes 1958:€141). As Agar (1985) suggests, institutional dis-
course, which inevitably involves professional experts, is constituted in three key
stages: diagnosis of the client, directives targeted at the client’s problems, and re-
ports in the form of case notes which encompass both diagnoses and the direc-
tives. This underscores the fact that professionals are privy to scientific knowledge
as well as organisational knowledge that clients do not have direct access to. Ac-
cording to Rueschemeyer (1986:€166), experts
define the situation for the untutored, they suggest priorities, they shape people’s
outlook on their life and world, and they establish standards of judgement in the
different areas of expertise – in matters of health and illness, order and justice, the
design and deployment of technology, the organisation of production.

This then allows for a lay-expert distinction, which is a long standing one. Schutz
(1964) contrasts expert knowledge and lay knowledge as follows:
The expert’s knowledge is restricted to a limited field but therein it is clear and
distinct. His opinions are based on warranted assertions: his judgements are
not mere guesswork or loose suppositions. The man on the street has a working
knowledge of many fields which are not necessarily coherent with one another.
His knowledge of recipes indicating how to bring forth in typical situations typical
results by typical means. (Schutz 1964:€122).

Expertise, according to the above stipulation, implies an in-depth mastery of a

field of knowledge. ‘Warranted assertions’ can only be made within a ‘limited field’.
Lay knowledge, by contrast, is not distinctly specific: it is rather ‘typical’, although
over time clients can acquire such know-how and become ‘professional clients’
(Sarangi and Slembrouck 1996).
In its contemporary form, expert knowledge is accessible to many of us, espe-
cially with the wider availability of web- and internet-based information systems. In
the healthcare context, the notion of ‘lay expert’ (Tuckett et al 1985, Moore at al
2001, Sarangi 2001, Prior 2003) is central to our understanding of expert knowledge
as something which is accessible to patients and their carers and that it is not some-
thing which only the professional expert embodies, although professionals supple-
ment their scientific knowledge with highly valuable experiential knowledge. In the
case of chronic illnesses, we can safely assume that patients do have access to and
experience of more specialised knowledge about their conditions (DoH 2001). The
Healthcare interaction as an expert communicative system 

same holds for healthcare concerning children. As Strong and Davis (1978:€63) point
out, parents have expert and detailed knowledge about their children, so doctor’s
scientific and generalised expertise is bound to be contingent on parents’ expertise:
‘The dependency of doctors’ expertise on parents’ expertise exists where the medical
problem can only be resolved by recourse to knowledge that only the parents pos-
sess’. The same is true of elderly patients or patients with limited mental abilities
where carers take on the ‘expert’ role of managing their illness and lifeworld.

1.2 The interactional basis of professional expertise

With regard to the medical profession, Freidson (1970) suggests that professional
expertise is constituted in a combination of scientific/technical knowledge and
clinical/experiential knowledge. As I see it, both these knowledge systems are in-
teractive, cumulative and systematic and do give rise to an array of expert interac-
tion systems in functional specific ways. Another inevitable component of this
‘expertise mix’, as I have pointed out earlier, is the institutional/organisational
ethos in which professional activities are carried out. Much of the expert knowl-
edge (scientific, clinical and organisational) are discernible in the interactional
level in terms of systematic history taking, diagnostic reasoning, use of evidence,
offer of causal explanations etc. More than mere rule-following, the contingent
character of interaction assumes significance. As Freidson (1970:€90) points out:
People [in the medical profession] are constantly responding to the organised
pressures of the situations they are in at any particular time, that what they are is
not completely but more their present than their past, and what they do is more
an outcome of the pressures of the situation they are in than of what they have
earlier internalised.

In Foucauldian terms (Foucault 1970), appropriation of discourses is an expert

knowledge activity, constituted in both the what of knowledge and the how of
knowledge (Ryle 1949), although these knowledge-claims may not always be made
explicit (Polanyi 1958, Heath 1979, Sarangi 2005a). In interactional terms, knowl-
edge management becomes a communicative activity, which professional experts
have to embody in their everyday lives when dealing with clients and fellow pro-
fessionals. Professionals’ acquisition of new technical knowledge and familiarisa-
tion with the changing organisational/institutional ethos as well as clients’ access
to expert knowledge more widely contribute towards transforming the nature of
situated interactional trajectories.
In light of the discussion above, interaction is constitutive of expert knowl-
edge alongside scientific, experiential and organisational dimensions, and that
healthcare professionals have explicit and tacit levels of knowledge about
 Srikant Sarangi

interaction in their specific professional, institutional settings. I proceed as fol-

lows. In Section€ 2, I briefly outline the interactionist turn in social and human
sciences, followed by, in Section€ 3, an overview of communicative practices in
healthcare encounters. In Section€4, I offer a general framework of activity analysis
for analysing interaction in professional and institutional settings. This leads me to
an illustration of a case study dealing with familial breast cancer in the context of
genetic counselling (Section 5). In conclusion (Section 6), I acknowledge the ec-
lectic nature of the activity analysis framework, but underscore its relevance as an
uptake-oriented analytic endeavour.

2. The interactionist turn in social and human sciences

Let us take a cursory look at the debate concerning the primacy of interaction
within the social and human sciences (for a recent overview see Atkinson and
Housley 2003). Interactionsim, in a broad sense, is a perspective which allows situ-
ated human agency to mediate social structure. The nature of this mediation, how-
ever, is not necessarily shared among various perspectives. A starting point, for
our purposes, is Blumer’s (1969) model of symbolic interaction, which was prima-
rily a reaction against a deterministic view of the social world which relied heavily
on causal explanations. According to Blumer (1969:€11–12):
The position of symbolic interactionsim is that the ‘worlds’ that exist for human
beings and for their groups are composed of ‘objects’ and that these objects are the
product of symbolic interaction… The nature of an object – of any and every ob-
ject – consists of the meaning that it has for the person for whom it is an object.

Glassner (1980) considers Blumer’s position as being idealistic, which is conceptu-

alised as an alternative to materialism by denying the existence of phenomena in
their own right. Other interactionists such as Mead (1934) and Simmel (1950)
would distance themselves from such extreme symbolism and recognise the exist-
ence of social phenomena in their own right. This has led Glassner (1980:€22–23) to
propose what he calls ‘essential interactionism’ consisting of events, states, phenom-
ena and processes: ‘Interactions may be described as processes made up of phenom-
ena within various events, which at each point make up states amid other states’.
Bakhtin’s project in dialogicism can also be seen as an exercise in interaction-
ism. The basic unit – utterance – is not reducible to an objective meaning outside
of their communicative environment. For Bakhtin (1986:€99):
Language as a system has an immense supply of purely linguistic means for ex-
pressing formal addresses … But they acquire addressivity only in the whole of a
concrete utterance.
Healthcare interaction as an expert communicative system 

The notion of addressivity is never exhausted, as ‘anticipated responsive reactions’

are considered part and parcel of one’s individual style. In a similar vein, Voloshinov
(1987:€99) suggests that the context of the utterance must consist of three factors:
(1) the common spatial purview of the interlocutors (the unity of the visible…);
(2) the interlocutors’ common knowledge and understanding of the situation, and
(3) their common evaluation of that situation.

Wilson (1971:€ 60) characterises this trend of interactionism as a shift from the
normative paradigm in which ‘interaction is viewed as rule-governed in the sense
that an observed pattern of action is rendered intelligible and is explained by refer-
ring to rules in the forms of dispositions and expectations to which actors are
subject’. As Voysey (1975:€24) puts it:
The major problem with this [normative paradigm], however, is that if rules are
to account for observed or imputed similarities in action in different situations, or
over time, there must be an assumption of ‘substantial cognitive consensus’. Ac-
tors must agree not only that a situation is one in which a particular rule should be
followed, but on what counts as evidence that it is being followed.

This raises particular questions about the positioning of the analyst in looking for
patterns of similarities and differences across a given interactional trajectory. For
Wilson (1971:€67), within the ‘interpretive paradigm’, unlike the normative one,
‘interaction is an essentially interpretative process in which meanings evolve and
change over the course of the interaction’. This echoes the Weberian position:
What distinguishes an interpretive explanation is that it involves explaining be-
haviour by reference to the agent’s conceptions of what he is doing, as opposed to
explaining it by causal laws. Interpretive explanation takes into account the fact
that an agent’s knowledge of his actions differs in important ways from that which
an observer can have of those actions. (Levison 1974:€101)

Such a perspective on social action/interaction is not only at the heart of phe-

nomenology (Schutz 1964), ethnomethodology (Garfinkel 1967) and cognitive
sociology (Cicourel 1974), but also what characterises Goffman’s call for the study
of interaction in its own right. In the preface to Relations in Public, Goffman
(1971:€13) writes:
The realm of activity that is generated by face-to-face interaction and organised by
norms and co-mingling – a domain containing weddings, family meals, chaired
meetings, forced marches, service encounters, queues, crowds, and couples – has
never been sufficiently treated as a subject matter in its own right. In fact, a con-
venience has often been made of it. Whenever a concrete illustration has been
needed of how it is with a social establishment, or a bit of social structure, or even
a society, interaction vignettes have been fetched in to provide vivid evidence and,
 Srikant Sarangi

incidentally, a little obeisance to the fact that there are people out there moving
about. Thus interaction practices have been used to illuminate other things, but
themselves are treated as though they did not need to be defined or were not
worth defining. Yet the nicest use for these events is the explication of their own
generic character.

Following from this conviction, Goffman (1983) formulates his notion of the ‘in-
teraction order’:
My concern over the years has been to promote acceptance of this face-to-face
domain as an analytically viable one – a domain which may be titled, for want of
any happy name, as the interaction order. (Goffman 1983:€2)

In making a strong case for the study of ‘the neglected situation’, Goffman offers a
distinction between interaction order to mean interactional practices and the tradi-
tionally conceptualised ‘elements of social organisation’ in the sense of social struc-
tures – or what Wilson (1971) refers to as the ‘normative paradigm’. He goes on to
capture the linkage between these two domains as ‘loose coupling’ (Goffman 1983).
In this sense, the ‘interaction order’ for Goffman goes beyond face-to-face encoun-
ters, and by extension, talk-in-interaction within the conversation analytic tradi-
tion. This is clearly reflected in what Goffman treats as data in his writings and what
knowledge of context he invokes to aid the interpretive procedure.
One striking observation is that Goffman issues a challenge to linguistics to
explicate systematically the role of language in interaction. As he puts it with rela-
tion to his notion of footing: ‘linguistics provides us with the cues and markers
through which such footings become manifest, helping us to find our way to a
structural basis for analyzing them’ (Goffman 1981:€157). As Tannen (1993) points
out, Gumperz’s (1982) theory of conversational inference is one such response. In
addition to contextualisation cues functioning as a signalling mechanism for ne-
gotiation and shifts in frames and footings, other pragmatic notions such as pre-
supposition, implicature, coherence, indexicality are intricately embedded in
Goffman’s (1974) frame analysis. I shall return to these analytic resources in my
proposal for activity analysis in Section€4.

3. Communicative practices in healthcare encounters

Over the years, communicative practices in healthcare consultations have attract-

ed the attention of scholars from within conversation analysis, sociolinguistics and
discourse studies (e.g., Mishler 1984, Fisher and Todd 1986, Heath 1986, Drew
and Heritage 1992, Sarangi and Roberts 1999, Candlin and Candlin 2003, Sarangi
2004, Sarangi 2005a, Heritage and Maynard 2006). It is worth acknowledging that
Healthcare interaction as an expert communicative system 

a vast number of healthcare interaction studies have prioritised the doctor-patient

encounter, especially in the primary care setting, which has been described as a
genre in structural and sequential terms (Ten Have 1989). Several insights have
been gained at the interactional level by scholars working within the conversation
analytic tradition: e.g., Maynard’s (1991) perspective display series in delivery of
bad news; Heritage and Sefi’s (1992) step-wise organisation of advice by health
visitors; Silverman’s (1997) information-as-advice formats in HIV counselling.
Identification of such interaction patterns is testimony to the power of micro-anal-
ysis, but this also shows that healthcare professionals expertly bring off certain
preferred interactional trajectories when communicating with their clients.
Peräkylä and Vehviläinen (2003) have recently drawn our attention to what
they call professional ‘stocks of interactional knowledge’ (SIK). By SIK, they refer
to the normative models and theories found in communication textbooks and
manuals. Quite rightly they challenge the rather simplistic conceptualisation of
interaction and, in the Goffmanian spirit, call for the need to systematically exam-
ine the interaction process itself. One needs, however, to keep the textbook char-
acterisation of interactional knowledge separate from how professional practition-
ers conceptualise and operationalise interaction in their everyday practice. When
we interview professionals or become involved in long-term ethnographic field-
work, we realise the complex nature of interactional knowledge shared among
professionals. Following Polanyi (1958), my concern here is with the embodied,
tacit knowledge of interaction with which healthcare professionals conduct their
communicative practices.
Consider Byrne and Long’s (1976) identification of interactional differences
between a medical consultation and an employment interview. Focusing on a ‘re-
peat prescription’ visit, they suggest that ‘all of the patient’s replies to questions
have been absorbed by the doctor who has never used any of the information given
to develop further responses’ (p.13). Although they readily acknowledge this as
untypical of consultation, they do claim that in 75% of their recorded consulta-
tions, ‘the doctor provides all of the causes and the patient all of the effects’ (p.13).2
By contrast, in the employment interview situation, the interviewer ‘initiates the
discussion and is thus causative, but all of his subsequent questions are the result
of something the interviewee has previously said. Thus all of his interventions are
really effects rather than causes’ (p.12). At the risk of premature generalisation, this
finding underscores how professional dominance is routinely manifest as

2. Interestingly, this lack of attention to the patient’s contributions is used as a rationale for
Byrne and Long to focus their analysis only on the doctor’s consulting style, broadly based on
the interaction category system developed by Bales (1970). The complexity of the communica-
tion process, however, has been acknowledged by many researchers from within this tradition
(see, for example, Davis [1982] who draws upon key insights from Goffman).
 Srikant Sarangi

interactional asymmetry and can have potential implications for patient-centred

healthcare delivery. It would follow that doctors will need to develop a more in-
volving style of participation and allow more interactional space to patients for a
fuller articulation of ‘the voice of the lifeworld’ (Mishler 1984, Mishler et al 1989).
Within the spectrum of medical consultations, there are interactional varia-
tions to reckon with. One would expect different interactional patterns for repeat
visits as opposed to first visits, in primary care settings as opposed to tertiary clin-
ic settings, for consultations dealing with simple curable conditions as opposed to
those involving chronic illnesses. Providing information and communicating in-
tervention in the context of antibiotics prescription vs. chemotherapy intervention
will necessitate different interactional trajectories and outcomes. Davis (1982),
dealing with children’s clinics, observes that when children are called in for rou-
tine check-ups (receptive stance) as opposed to when they come with presenting
problems (initiative stance), different patterns of interaction management emerge.
In his seminal study of Down’s Syndrome children, Silverman (1987) shows how
doctors de-medicalise the child’s condition as part of a non-interventionist agen-
da, and thus shift responsibilities about care to parents, which has to be accom-
plished interactionally.
Let us consider in detail the communication of uncertainty in the cancer set-
ting. In an early study, McIntosh (1978) shows how doctors use routine responses,
including doses of euphemism, to deal with uncertain diagnosis and prognosis. In
addition to clinical uncertainty, doctors have to cope with two other forms of un-
certainty: (i) uncertainty about patients’ genuine desire to know a bad diagnosis
and prognosis; and (ii) uncertainty about how patients might react to bad news,
whether diagnostic or prognostic. Such an awareness then has interactional con-
sequences. McIntosh found that doctors routinely used typification strategies,
i.e., allocate patients first to illness categories and then ‘profess’ or ‘imply’ certainty
and uncertainty. This demonstrates that healthcare professionals operate with dif-
ferent models of communication management depending on the severity of the
condition as well as certainties and uncertainties surrounding clinical and external
factors. The data examples McIntosh provides may come across as being explicitly
directive and paternalistic, with no indication of shared decision making proto-
cols. In the current climate of patient-centred healthcare in the UK, such maxims
will have undergone transformation, although the same clinical outcomes/inter-
ventions may probably be the result. On the surface of interaction, we would ex-
pect the deployment of patient-oriented styles such as perspective display series
(Maynard 1991) to negotiate treatment options. This suggests that expert interac-
tion management systems are subject to general socio-political, organisational and
technological changes in healthcare delivery.
Healthcare interaction as an expert communicative system 

The counselling encounter, which is the focus of my analysis in this paper

(see Section€ 5 below), is different from mainstream medical consultations. The
nondirective ethos of counselling and therapy is often manifest in a series of inter-
actional features: e.g., minimal back-channelling on the part of the professional
and use of repetition to signal confirmation or request elaboration surrounding
interpretive summaries (Ferrara 1994); withholding of answers to clients’ ques-
tions (Turner 1972); packaging of information-as-advice (Silverman 1997); initia-
tion of reflective frames to elicit clients’ levels of understanding and coping mech-
anisms (Sarangi et al 2004, 2005), and use of hypothetical questions to announce
future states of affairs, upgrade the conditionality of such assertions and prepare
the client for worst scenarios (Peräkylä 1995).
In the context of therapeutic encounters, Ferrara (1994) draws an important
distinction between the rhetorical force and the speech act force underlying par-
ticipants’ interpretive practices. In focusing on the interactional significance of
repetition in therapeutic settings, she identifies two types of repetition: echoing
and mirroring. Echoing involves ‘the contiguous repetition of another’s utterance
or statement using the same downward intonation in an adjacency pair’, which is
usually done by the client, sometimes allowing for pause. Echoing signals emphat-
ic agreement more than explicitly formulated confirming utterances such as ‘yes’, ‘I
know’, ‘you’re right’. What the therapist proposes as a candidate for echoing is an
‘interpretive summary about the client’s experience’; so the repetition by the client
signals agreement of assessment proffered by the therapist. Mirroring, on the oth-
er hand, involves ‘partial repetition by the therapist of a client’s statement’ using
the same downward intonation. This is meant to be heard by the client as a request
for elaboration. Both therapists and clients share these interactional norms and
can be seen as aligned to each other’s communicative practices.
With regard to group therapy encounters, Turner (1972) discusses participa-
tion structure in relation to question-answer sequences. When clients ask ques-
tions such as ‘why are we here’, ‘is therapy working’ as ‘first action’, the therapist is
not supposed to occupy the ‘second action’ position and offer an immediate re-
sponse. This withholding of the ‘second action’ position is not to be regarded as
indifference on the part of the therapist; indeed it constitutes a display of situated
interactional expertise. In a similar vein, Scheff (1968:€12) characterises the psy-
chotherapeutic interview as ‘a series of offers and responses that continue until an
offer (a definition of situation) is reached that is acceptable to both parties’. Ac-
cording to him, the success of a therapeutic session relies on ‘patients who accept,
or can be led to accept, the problems as internal, as part of their personality, rather
than seeing them as caused by external conditions’ (Scheff 1968:€13). This perspec-
tive underlines how questions are framed and how the therapist controls the inter-
action by shifting topics, while rejecting the clients’ offers. At the interactional
 Srikant Sarangi

level there seems to be a relationship between the staged nature of interaction

management and the therapeutic outcome. In some clinical contexts, Peräkylä,
Ruusuvuori and Vehviläinen (2005) point to a potentially strong association be-
tween interaction theory and treatment theory. Different healthcare sites will pri-
oritise different interactional features based upon their diagnostic and
treatment regimes, while also aligning with institutional/organisational realities
(Sarangi 2005b). Shifts at the organisational policy level, such as movements to-
wards patient-centred, evidence-based healthcare, have interactional consequenc-
es that are not necessarily anticipated. Likewise, developments in scientific knowl-
edge and technology no doubt impact upon interactional trajectories. In other
words, interaction as an expert communicative system constitutes an essential
component of healthcare delivery.

4. Activity analysis and interaction types

Generally speaking, institutional encounters are a structured activity (Agar 1985),

organised around tasks (Drew and Heritage 1992), with an interplay of institu-
tional, professional and personal-experiential modes of talk (Sarangi and Roberts
1999). Levinson (1979) proposes the notion of activity types (characterisation of
settings such as a medical consultation, a courtroom cross-examination) and the
inferential schemata that accompanies them as the basis for analysing institutional
encounters. Healthcare encounters, whether involving clients or fellow profession-
als, can be analysed in activity-specific terms, while allowing for variations within
and across specialties and sites. A further notion of discourse type or interaction
type (characterisation of forms of talk such as history taking, cross-examining,
troubles telling, advice giving) is useful in our understanding of the dynamic, hy-
brid nature of a given activity type.
What I propose here as ‘activity analysis’ (Sarangi 2000, Sarangi 2005c, Sarangi
[in press]) does build on the seminal work of Levinson (1979), especially his no-
tion of activity type. A theme-oriented discourse analysis underpins such an exer-
cise (Roberts and Sarangi, 2005). Within this orientation, variations within and
across healthcare encounters – in terms of focal and analytic themes – are legiti-
mately warranted. More importantly, the analytic task is to be based on an overall
mapping of structural, interactional and thematic trajectories of a given encounter
as a way of identifying activity-specific coherence and incoherence as well as criti-
cal moments for further detailed analysis.
Activity analysis pays attention to the flexible nature of the relationship be-
tween form and content of a given encounter. Gumperz and Cook-Gumperz
(1982:€15) point out that ‘communicative flexibility enables us to tell, by looking
Healthcare interaction as an expert communicative system 

only at actual performance features and without knowing the content, whether two
speakers are actively communicating’. While this is true, and Goffman’s notions of
alignment, frame and footing would attest such a position, there is an imperative
for the analyst to engage with the content of the interaction order. This is particu-
larly so in the professional/institutional context, where interactional patterns can-
not be disentangled from the treatment of content or theme (Sarangi 2007).
The key concepts for activity analysis derive from Goffman – the concepts of
frame, footing, face work and alignment. It is worth noting that although Goffman’s
notion of interaction order and the attendant methodology of ‘frame analysis’
(Goffman 1974) have had a lasting impact on many forms of interaction analysis
undertaken within sociolinguistics and discourse studies, Goffman hardly en-
gaged himself with analysing recorded social encounters in a systematic way. What
is not always clear is how one goes about defining and working with notions such
as frame and footing. There has been a tendency to define such concepts in a con-
tingent manner to suit the analyst’s current purposes. Like the term discourse,
frame has been used at different layers of meaning. For instance, Tannen and
Wallat (1993) use frame to draw distinctions between ‘consultation frame’, ‘exami-
nation frame’ and ‘reporting frame’ in the context of medical examination/inter-
view. Also, they go on to suggest that there are tensions in the way such frames are
manifest, although one could regard this as part of hybrid competencies of both
the medical professional and the client.
Another key notion to activity analysis is alignment. The notion of alignment
is based on a view of interaction as jointly produced. In a seminal paper, Stokes and
Hewitt (1976) suggest that the notion of ‘aligning actions’ encompasses two mean-
ings: (i) how individual conduct accords with that of co-participants in the creation
of social acts; and (ii) how problematic situations involve discrepancies ‘between
what is actually taking place in a given situation and what is thought to be typical,
normatively expected, probable, desirable or, in other respects, more in accord with
what is culturally normal’ (1976:€843). Alignment, from an activity analytic per-
spective, involves both well-synchronised turn-taking and a display of shared un-
derstanding of what is talked about and what participant roles are expected at a
particular point in time. As interaction analysts, therefore, we need a reasonable
understanding of the topics of counselling and therapy in order to be able to inter-
pret in a meaningful way the interactional trajectories of a given encounter.
Activity analysis therefore has to be grounded in what I would call ‘thick par-
ticipation’ (Sarangi 2007) in the professional/institutional events, and ‘thick de-
scription’ has to draw upon structural, interactional and thematic mapping of
whole encounters (Roberts and Sarangi 2002, Sarangi [in press]). It is worth point-
ing out that structural and interactional maps have been particularly useful in
educational settings (Green and Wallat 1981, Mehan 1979) and in narrative
 Srikant Sarangi

analysis (Labov and Waletsky 1967, Gee 1997). Interactional mapping can be real-
ised both in terms of distribution and volume of turn taking and such maps can
tell us something about the interaction system and the positioning of the partici-
pants within it. This is not to suggest that the more one talks, i.e., holds the floor,
the more powerful s/he is. Interactional asymmetry should not be confused with
knowledge asymmetry. Equal access/right to turn taking does not necessarily
equate with symmetrical knowledge relations.
To summarise, the following analytic features are constitutive of activity analysis:
– Mapping of entire encounters at structural, interactional and thematic levels
– Communicative flexibility in terms of activity types and discourse/interaction
– Integration of discoursal and rhetorical devices
– Goffman’s notions of frame, footing and face-work
– Gumperz’s notions of contextualisation cues and conversational inference
– Alignment: sequential and normative
– Social and discourse role-relations
– Thick participation and thick description
Within a framework of activity analysis, interactions are seen as a narrative un-
folding of events and characters, organised temporally and spatially. In addition to
the sequential order, rhetorical moves are also central to how events and charac-
ters are portrayed and managed in interaction (see, for instance, Goodwin’s [1994]
general proposals about nature of salience, backgrounding and foregrounding of
information; see also use of devices such contrast, reported speech, question, rep-
etition, metaphor). In this sense, activity analysis as an enterprise in interpretive
understanding can be situated between sequential description (as is the case with
conversation analysis) and extra-situational explanation (as is the case with critical
discourse analysis).

5. An extended case study from genetic counselling

In this section I adopt the activity-analytic perspective for the purposes of analysing
interactional patterns in counselling encounters, especially genetic counselling
which can be characterised as a hybrid activity type (Sarangi 2000). The choice is
between taking the route of corpus-based analysis of genetic counselling to demon-
strate variations in interaction types (both patterns of differences and similarities)
or focusing on a single case in order to understand the staged dynamism character-
istic of the overall interactional trajectory. Here I choose the latter option and focus
on a single genetic counselling session involving familial breast cancer. My focal
Healthcare interaction as an expert communicative system 

theme is risk and uncertainty and the analytic themes will centre around the notions
of frame, alignment and the rhetorical devices of escalation and de-escalation.
Management of risk and uncertainty is typical of genetic counselling, espe-
cially involving predictive testing. The professional expertise of the counsellor is
evident in giving clear explanations about genes, patterns of inheritance, popula-
tion risk etc. It also constitutes a space for discouraging unwarranted decisions,
whether about testing or acting on positive test results. The genetic counselling
protocol in South Wales, from where the following case of breast cancer is taken,
includes at least two or more sessions before a gene test is offered to the client.
The extended extracts below constitute the very first session to ascertain the
at-risk status of the client (AF), who is in her early twenties.3 The genetic counsel-
lor (G2) opens the session as follows:
the purpose of this session is is really to discuss breast cancer – how it can run in
families and (.) just wonder what your own sort of worries and concerns were eh
eh in the matter.

Such a definition of the situation (McHugh 1968) is integral to activity analysis, as

this not only provides a frame of reference for the participants in the interaction in
question, but also for the interaction analyst.
Given the space constraints, I shall not offer the structural, interactional and
thematic maps for this clinic session (for illustrations, see Sarangi [in press]). Let
me first summarise the first 12 turns. AF reports that she has always been aware of
breast cancer running in her family. AF has already had a mammogram, following
which she has now been referred to the genetics clinic. In a routine fashion and
using the typification strategy (see McIntosh 1978 above), G2 explains at length
the general basis of breast cancer: that it is a common condition affecting one in
ten/twelve women in the general population, independent of hereditary factors;
that only a small number of women inherit it genetically, in which case they will
develop it at a younger age than normal, and that they often develop cancer in both
breasts. G2 then goes on to explain the working of genes as follows (see appendix
for transcription conventions):
Extract 1
01 G2: these genes when they work normally they protect against breast
cancer (.) and they also protect against ovarian cancer as well
02 AF: mhm

3. The names of participants have been anonymised. The case study is taken from The Well-
come Trust funded project on Communicative Frames in Counselling for Predictive Genetic
Testing (2000–2004).
 Srikant Sarangi

03 G2: but if they become altered in some way or or damaged then (.) they
stop working and that protection is lost so if someone if a woman
inherited one of these altered copies of the gene
04 AF: yeah
05 G2: she has a higher risk of developing breast cancer and also it’s like
higher risk of developing ovarian cancer as well (1.5) and that’s (.5)
those are the sort of (.) women that (.) those families are the ones
where you you see a lot of women [affected] (1.5)
06 AF: [yes]
07 G2: (.5) now- these families aren’t that common but what we do find
quite often is that because breast cancer in itself is so common we
find families where there are several people affected just by chance
08 AF: mhm
09 G2: so what we need to do is to try and decide whether and when we’re
looking at someone’s family tree try and decide whether [it’s chance]
10 AF: [it’s chance]
11 G2: or whether we think there might be a gene involved basically
12 AF: mhm
Here G2 dominates the interactional space, much of which has been taken up by
explanations about normal vs. altered genes, which will become the frame of refer-
ence for the rest of the encounter. Unlike psychotherapeutic encounters where the
therapist, through use of minimal backchannelling cues, allows the client to ex-
press his/her concerns (Ferrara 1994), genetic counselling often involves a fair
amount of explanation talk, given that not everybody who comes to clinic will be
familiar with the complexities of how genes work. Through consistent use of min-
imal cues, AF remains interactionally aligned to the explanation routine. The con-
cept of risk is introduced within a framework of uncertainty – in a mixture of hy-
pothetical language (‘if ’), and by using the generic third person referents (‘a
woman’, ‘someone’, ‘those’/’these’ families) – which may be seen as an instance of
typification strategy to manage uncertainty of routine explanation (McIntosh
1978). AF is allocated to an already existent category of at-risk condition, rather
than singling out her personal circumstances and addressing them specifically
(see McIntosh’s distinction between patient’s condition and patient’s circumstanc-
es). There is a significant pause of 1.5 seconds in turn 05, which, as part of delivery
of routine explanation, can be taken as a signal for continuation of G2’s turn (see
Boden’s [1994] discussion of long pauses in committee meetings). The contribu-
tions made by AF are minimal, but taken together they indicate her awareness of
basic genetic knowledge (also evident from her opening remarks, not shown here).
In turns 07 and 09, G2 prepares the ground for diagnostic work, as he draws a
Healthcare interaction as an expert communicative system 

distinction between chance mutation and patterns of inheritance, as the latter will
have a significant bearing on whether AF will develop breast and ovarian cancer at
a younger age.
Extract 2 (continues from extract above)
01 G2: so one of the ways we do that is by drawing a family tree like the one
that we’ve drawn for you
02 AF: mhm
03 G2: the one that (nurse) drew for your own family and we’ve to see if we
find any patterns of of breast cancer in in (.) in the family
04 AF: mhm mhm
05 G2: so if we look at your family tree ((both shift their gaze to family tree
diagram on paper)) the circles are women squares are men and the
ones that are coloured in have had breast cancer (.5) so you’re down
here and that’s your mother had breast cancer [at]
06 AF: [yeah]
07 G2: the age of forty two
08 AF: yeah
09 G2: and she’s how is she now?
10 AF: she’s fine
11 G2: she’s fine [yeah okay] alright so she is (.) now sixty-five
12 AF: [yeah she’s fine]
13 G2: and her mother had breast cancer first at the age of forty-three
14 AF: erm
15 G2: and then again at the age of sixty-eight?
16 AF: erm (.5)
17 G2 and also he:::r sister-
18 AF: yeah
19 G2: had breast cancer at the age of sixty and she’s now seventy-eight
20 AF: yeah
21 G2: and you think there might be someone else as well [on that side of
the family]
22 AF: [yes yeah there
is ] (.) there is yeah
23 G2: right also on your father’s side of the family there’s some more dis-
tant relatives [who] who have breast cancer as well but that’s quite
24 AF: mhm
25 G2 [they’ve got-]
26 AF: [my father’s] aunt and cousin
27 G2: =yeah so they’re- they’re quite distantly related already to you
 Srikant Sarangi

28 AF: mhm
29 G2: so I think this side of the family is definitely the more important side
What we see here is a retrospective history taking routine, which is typical of ge-
netic counselling. It is striking that G2 presents the genetic facts concerning AF’s
family members in the format of A/B-events4 (Labov and Fanshel 1977), which
only requires AF to confirm G2’s statements. Such a history taking routine is mark-
edly different from those undertaken in many other tertiary consultation settings.
The latter are characterised by embedded and chained questions (A-events) asked
by the doctor to which patients provide information (B-events). In this instance,
G2’s knowledge about AF’s family history is derived from the family pedigree chart
alluded to in turn 05 (note that family pedigree is routinely charted by specialist
nurses prior to the first clinic session). Starting with turn 05, G2 focuses on the
maternal side and foregrounds the age of onset in each instance. It is thus a display
of expert genetic knowledge, which requires attributing discriminatory significance
to patterns of inheritance across AF’s maternal and paternal lines. Affected family
members’ age at onset is crucial information for determining if another close rela-
tive will develop breast cancer at a younger age. The particularities of familial in-
heritance patterns thus come to the fore, as part of G2’s expertise. As we see in turns
21ff, G2 stresses that it is the mother’s side which has direct consequences for AF’s
current and future genetic status. The two family lines are portrayed in terms of
proximal vs distant relations as far as genetic inheritance is concerned. The discus-
sion here anticipates which of the family members will need to be contacted for the
diagnostic/predictive work to be conducted smoothly, and how AF’s decision to
test might have implications for these family members vis-à-vis others.
Extract 3 (continues from extract above)
01 G2: I mean looking at that it does look like there might be something
being passed down the fam- through the family (.5) the other
thing that we do is we can do a calculation where we compare the
ages of the women who have had breast cancer to to those who
haven’t had breast cancer (1.0) and so we can actually from that
we can work out what we think the chances are (.) that there
might be a gene running in the family
02 AF: erm

4. According to Labov and Fanshel (1977), ‘A’ events refer to information which is known to
the speaker, ‘B’ events refer to information which is known to the addressee and ‘A/B’ events
refer to information known to both parties. Following this, if the speaker makes a statement
about a ‘B’ event, this will be heard by the addressee as a request for confirmation.
Healthcare interaction as an expert communicative system 

03 G2: and what we think your own risk of having breast cancer in your
04 AF: yeah
05 G2 see if it’s increased somewhat
06 AF: yeah
07 G2: now I did this for your (.) for your family and we do it using a com-
puter (.5) but it’s essentially just a sort of complicated calculation
08 AF: yes
09 G2: and your own risk is higher than that of the general population =
10 AF: =I thought it would be ((tense laugh)) I’m not surprised
11 G2: thought yeah yeah and it’s it’s your risk is about thirty percent basi-
cally I [think] so you’re ^^^^^^a thirty [percent]
12 AF: [so that’s] high high is it? Or::::
13 G2: it’s (.5) it is
14 AF: yeah
15 G2 significantly high I mean anything that we’re- I mean w-
16 AF: erm
17 G2: has to be taken with a pinch of salt because it is
18 AF: yeah::::ah
19 G2: just based on a on a sort of mathematical calculations so it’s not a
(.5) a
20 AF: yeah:::ah
21 G2: figure [that is set] in stone or anything [like that] that-
22 AF: [no:: no::] [erm]
23 G2: so it would mean that sort of three times out of ten you would have
a chance of (.) breast cancer (.) but then again seven times out of ten
you won’t [develop] breast cancer
24 AF: [yes:::] (.5)
25 G2: so you- (.5) so your chance is about three times as high as the gen-
eral population
26 AF: oh that’s nice ((laughing)) [hhhh hhhhh hhhhh] hah hah
27 G2: [^^^ ^ na:::::::::] (.5)
28 G2: but it (.) it sounds like that’s not a lot (.5) not a big (.5) of a [of a yeah
29 AF: [oh it’s
no shock no no]
30 G2: (you thought that it’d go up higher) (.5) you would be (1.0) we would
think about seriously think about looking for a gene in your family
[(^^ ^^)]
31 AF: mhm
 Srikant Sarangi

32 G2: to see if we could find an alteration because you your your risk is
sufficiently high enough for us to think about that
33 AF: right
34 G2: but (.5) doing that test has a lot of implications in it in itself (1.0)
The above episode begins with a return to the results frame, which is strategically
initiated in order to make progress with diagnosis. In turn 07, the computerised
risk figure is ascertained, similar to how medical test results are routinely given,
although a gene test result is yet to be obtained for establishing AF’s at-risk status.
G2 uses mitigation (‘just a sort of ’), which is subjected to further de-escalation in
turn 21 (see below). The risk figure is worked up in a staged manner: beginning
with a general statement such as ‘your own risk is higher’ (turn 09). This is met
with a lack of surprise from AF in turn 10 (notice the absence of a news marker),
almost as if she has been anticipating such an outcome. An interactional misalign-
ment becomes evident. G2 then reformulates the risk in percentage terms. AF
immediately asks if she falls into the category of high risk – the frame of reference
being one’s risk is either high or low in relation to one’s own individual circum-
stances (see Adelswärd and Sachs [1998], Edwards et al [2002], Sarangi 2002 on
how people find it difficult to make sense of individual risk in terms of absolute
numbers). When AF asks whether her risk is high, she seems to allude to the no-
tion of ‘high’ as it applies to her as an individual, and not ‘high’ in a comparative
sense such that her risk is higher vis-à-vis the general population. G2 responds by
labelling AF’s at-risk status as ‘significantly high’ (turn 15), but this is immediately
downgraded when G2 draws attention to the abstract basis of the ‘mathematical
calculation’ and underscores the fact that the 30 percent figure is not ‘set in stone’
(turns 17, 19 and 21). G2 then recasts the 30 percent figure in a different language:
‘three times out of ten you would have a chance of breast cancer’ (turn 23). And
within the same turn he de-escalates this risk assessment without actually altering
the risk figure (‘but then again, seven times out of ten, you won’t develop breast
cancer’). The frame again shifts to AF’s risk vis-à-vis the general population. Meta-
phorically speaking, the dynamics of escalation/de-escalation is achieved by treat-
ing AF as a dice, thrown into the air and then allowing it to land on a chance ter-
ritory. An interactional pattern emerges as far as escalation and de-escalation of
risk is concerned: rather than further escalate a given risk figure in the next avail-
able turn, G2 consistently chooses to de-escalate this risk assessment before addi-
tional escalation is done in a staged manner.5
AF’s response to the high risk figure (turn 26) is another instance of interac-
tional misalignment, which comes across as indifferent, or even sarcastic, tinged

5. For a similar analysis of escalation and de-escalation devices in the legal context, see
Goodwin (1994).
Healthcare interaction as an expert communicative system 

with nervous laughter, perhaps as a device to release anxiety. Seen from G2’s view-
point, this marks a dispreferred response, and he has to interactionally adjust to
this situation. We note, in turn 28, G2 formulating AF’s perspective (this seems to
be a variation of what Maynard [1991] regards as perspective display series) before
shifting the frame to diagnosis which will involve ‘looking for a gene in your fam-
ily’ (turn 30). AF has to meet the eligibility criteria for genetic tests, and here G2
has to draw on his expert awareness of local service provision at an organisational
level (Wood et al 2003). The risk figure of 30 percent is regarded by G2 as optimal,
even though AF does not seem to think of this as a high enough risk.
About 12 turns are omitted, where AF latches on to G2’s remark in turn 34 to
talk about a television documentary she had watched recently. This concerned the
story of two sisters, with susceptibility for breast cancer. The test results showed
one of them as positive, who then opted for mastectomy to remove all the tissues
from her breast. AF concludes with the remark: ‘the risk was virtually taken away
for her’. This interjection is significant because, as we will note in the extract below
(see turn 36 in particular), AF may already be considering drastic intervention
measures if her test results were to be positive. For her, mastectomy can be a means
of weeding out potential risk once for all.
Extract 4
01 G2: [yeah] [yeah] yeah (.5) yeah I mean the thing about doing the the
gene test is that if if it shows that you do have an alteration of a gene
02 AF: mhm
03 G2: then your your risk will actually be a lot higher than thirty percent
– it’ll be more like eighty percent
04 AF: ho God ((nervous laughter))
05 G2: so it’s actually-
06 AF: ((sighs))
07 G2: yeah so (.5) so I mean that’s [it’s-]
08 AF: [so it’s] virtually definite then
09 G2: well [still (.) there’s still twenty percent chance that it isn’t] so- but it’s
10 AF: [mmhm]
11 G2 a lot higher [^^^ ^^^^ ^^^^^]
12 AF: [(I mean) ^^^^ ^^^ ^] ^^^ ^^ ^ ^
13 G2: whereas if- if we showed that you didn’t have an alteration in the gene
then your chance of (.) you your chance (would go down again)
14 AF: mhm
15 G2: to the general population so-
16 AF: mhm so if I did have an alteration in the gene what (would happen)
 Srikant Sarangi

17 G2: well - essentially at the moment I would advise you to be involved to

see a breast surgeon and to so you’re and to first thing examine
18 AF: yeah
19 G2: yourself regularly [every] month - make sure you you’re seen-
20 AF: [yes:::] I’ve had a mammogram already yeah
21 G2: right make sure you’re seen by a breast surgeon
22 AF: mhm
23 G2: regularly and also be involved in a screening programme
24 AF: but you can only go as regularly as they allow [you can’t you]
25 G2: [yes exactly]
26 AF: I mean you can’t do any more [can you no]
27 G2: [that’s right] and there’s- at the moment the evidence that mammo-
grams help in the under fifties [isn’t] isn’t definite anyway
28 AF: [mhm] [mhm]
29 G2: [we don’t] know whether it actually works you know
30 AF: mhm (.5)
31 G2: so (there may) the best the best (.5) (plan) is still very much self-
examination and reporting symptoms at an early stage because
32 AF: mhm
33 G2: the earlier the earlier any cancers are picked up the the better the
chance of a cure (.5) and if we did the gene test [(and)]
34 AF: [yes::::::]
35 G2: and it showed that you’re at high risk we wouldn’t actually do any-
thing different in terms of- because we can’t do anything el- we can’t
do anything more in terms of screening that would already- you al-
ready have maximum screening for the moment
36 AF: yea:::::h so there [wouldn’t be] you wouldn’t have like an option to
have (.) surgery or anything
37 G2: well (.5) there would be an option for - but that’s quite a serious (.5)
undertaking basically (.5) to have a mastectomy is (.5) but I mean
that’s something (.5) that is you know some (.5) something to:::
think about for a long time beforehand
38 AF: mhm
39 G2: one of the things about- (1.0) the the testing procedure itself is be-
cause of the you know the implications are such as these we we - it’s
not the sort of thing we just say you know - give us your arm we’ll
take some blood and send the results in the post that sort
40 AF: mhm
41 G2: of thing I mean (.5) we’d see you back here for two sessions of an hour
42 AF: mhm mhm
Healthcare interaction as an expert communicative system 

43 G2: at least you know or as long as it takes basically

44 AF: AF: yeah
45 G2: and then give you the results on the third session (.) so it’s not the
sort of thing that we do:::: (.5) (off the cards)
46 AF: righ::t
47 G2: (.5) and the purpose of today is really just to explain this to you and
give you the option of you know of that it that it’s just to let you
know that we can do a test
48 AF: erm (.5)
First we notice escalation of risk by G2 as a way of justifying future action. The hypo-
thetical scenario makes it possible to talk about worst possibilities, while also encour-
aging AF to reflect on her feelings and coping mechanisms. AF’s remark in turn 08 –
‘so it’s virtually definite then’ – calls for a de-escalation move on G2’s part in turn 09
(‘there’s still twenty percent chance that it isn’t so’). These are like exit strategies when
the risk talk hits the upper ceiling. In the hypothetical mode, G2 offers both ends of
the spectrum, as he draws attention to the possibility of AF not having an alteration
(turn 13). It is interesting that AF chooses to engage with the ‘positive test result’ sce-
nario (turn 16) and test out the possible intervention options available. In turns 17ff,
G2 slides into typification, although apparently he continues to address AF in his use
of ‘you’ form. In turns 24 and 26, AF uses the generic second person pronoun ‘you’ to
signal her disappointment at accessing the screening programme optimally. With re-
gard to mammograms, G2 alludes to scarce resources and issues of eligibility as much
as outlining the lack of evidence of clinical benefits for mammograms under fifties,
using his expert knowledge status. This leads AF to explore in a prospective way alter-
native options such as surgery (turn 36). In turn 37, G2 signals his hesitation, high-
lighted by the dispreference marker ‘well’ and the accompanying statement that ‘that’s
quite a serious undertaking basically to have a mastectomy’. Rather than pursuing this
intervention frame any further, G2 instead chooses to return to the genetic test pro-
tocol and to make obvious the ways in which this differs from other medical test situ-
ations (turn 39), followed by a further definition of the genetic counselling protocol.
Some twenty turns are omitted, when AF reports her discussions with the
specialist nurse prior to this clinic appointment. While G2 goes to considerable
length in emphasising that she does not have to rush into a decision of any sort, AF
signals explicitly that she ‘would want the test’. Our final extract is as follows:
Extract 5
01 G2: (.5) the way that we actually (need) do the test would be to take
some blood of someone in your family who actually has this condi-
tion or has had [the condition]
02 AF: [so my mother] (that)
 Srikant Sarangi

03 G2: it would be your mother or- ideally your mother in fact (yeah)
04 AF: mhm
05 G2: because she is the closely most closely related
06 AF: erm
07 G2: to you (.5) and the way- it works is that (1.0) each of us has - you
you’d have how your mother’s got two copies of this gene (.) and if
we assume (.) that she (.5) that one of her one of the copies has
08 AF: mhm
09 G2: altered for her to have had the condition or the disease
10 AF: yeah
11 G2: and she could have either passed on the normal copy of the gene to
you or the altered copy of the gene to you so there’s a fifty-fifty chance
(.5) and if she’s passed on the normal copy for you (.) of the gene to
you then you haven’t got it- your risk is the same as the general
12 AF: erm
13 G2: population - if she’s passed on the (.) the altered copy then your risk
is [much higher]
14 AF: [higher]
15 G2: yeah so that’s - that would be the situation
16 AF: right
17: G2: and the same would also apply to your sister (1.0)
18 AF: my sister doesn’t seem to (^^^^^ ^^^^^^^ ^^) ((laughter)) I [don’t
19 G2: [right
well] I mean different people have different reactions to the whole
thing and some people are not interested at all and that’s fine be-
cause at the end of the day what we’re testing for is something that
we we can’t do that much about (you see)
20 AF: mhm mhm (.5)
21 G2: so it’s it’s not you know that’s why that’s why we spend so (.5) long
22 AF: mhm
23 G2: through the implications of it (1.0) how wo- how do you think your
mother would feel about giving a a blood sample
24 AF: oh she’d hate it but she’d do it [she] hates needles [she’d] absolutely
hate it
25 G2: [right] [right] okay (.5)
26 G2: ‘cause there are- I mean the other implication is that there’s also a
higher risk of ovarian cancer as well in these women
27 AF: mhm
Healthcare interaction as an expert communicative system 

28 G2: so that’s something else

29 AF: =Go:::::::::d [oh god he he heh hhe hheh heheh] ((sigh)) o::::::::::h
dear oh dear
30 G2: [alright it g- it g- and it gets worse but-] (.5) but I think
that’s about all (the bad news that) there is
31 AF: o:::::::::::h ((falling tone)) [really] I’m so glad I came
The consultation more or less returns to where it had begun. G2 first spells out
which of the family members could be involved in the testing process. We
again have an escalation/de-escalation routine (turns 07, 09, 11 and 13). In
turn 11, we have a repetition of the explanation concerning autosomal domi-
nant genetic conditions, but articulated in everyday terms. G2 now speaks the
language of high/low risk as AF had initiated. In turn 19, G2 admits the failure
to cure (‘what we are testing for is something that we we can’t do that much
about’). In what follows, he justifies why genetic counselling as an activity is
information-rich by dealing with implications of not only testing and test re-
sults for individuals concerned, but also in considering consequences for fam-
ily members. In turn 23, the intervention frame is introduced with the enquiry
about AF’s mother’s feelings for giving a blood sample. The question is framed
ambivalently, possibly requiring a more than matter-of-fact response that is
offered by AF in turn 24. In turn 26, G2 mentions other risk scenarios, which
makes AF display her anxiety and nervousness, but she is overall pleased to
have accessed the counselling service to know about her current and future
risk status.
The interaction continues for another 30 turns where AF begins to talk about
her cousin who recently died of bowel cancer. G2 offers to look at this cousin’s
history to see if this might have any consequences for AF’s at-risk status. The clo-
sure is accomplished by setting out a procedure for taking blood sample from
AF’s mother.
Throughout the encounter, two things become apparent: first, the frames
shift constantly between family history, diagnosis, possible test results and in-
tervention, which may be typical of genetic counselling dealing with predictive
testing. There is a constant shifting between retrospective and prospective sce-
narios: often retrospective accounts have a prospective orientation. The idea is
to selectively guide the client to provide information that would lead towards a
diagnosis or prognosis rather than just being a mission in fact finding. Sec-
ondly, the escalation and de-escalation of risk is accomplished in a rhythmic,
dance-like fashion whereby every escalation move is followed by a de-escala-
tion move.
 Srikant Sarangi

6. Conclusion

Focusing on the professional/institutional context, I have suggested that interaction

as an expert communicative system is a necessary condition of healthcare encoun-
ters for understanding risk, coping with uncertainty, evaluating evidence, making
decisions etc. It is not always possible to discern from interaction itself what deci-
sions are eventually made, but the interactional context, which is contingent and
constantly unfolding, plays a significant part in communicative processes and out-
comes. In developing a model of activity analysis, I have tried to show that there is
more to interactional trajectories than sequential organisation of turn-taking.
In the words of Lynch and Sharrock (2003: xxxix): ‘Although the sequential
procedures that make up what conversation analysts call “talk in interaction” are
evident in, and important for, the organisation of practices in a variety of social
institutions, it is not enough to say that, for example, a jury deliberation or a med-
ical diagnosis is an “organisation of talk”’. The case study of genetic counselling
demonstrates the dynamic frame shifts between accessing services, explaining
conditions, taking family history, discussing testing protocol, diagnosis, prognosis
and treatment. There are rhetorical aspects to interaction – here patterns of escala-
tion and de-escalation of risk – which calls for interactional mapping to be matched
with thematic mapping of the entire encounter.
Activity analysis is one possible way to engage with the micro-level of interac-
tion against the backdrop of professional/institutional realities. The activity analy-
sis framework proposed here is intended to be eclectic, not ‘in the frequent use of
that term when it is equated with some magpie collection but in the sense of choos-
ing ideas or principles from diverse systems of thought toward the formation of a
coherent, integrated, whole system’ (Perlman 1957: xii). There remains the issue of
consensus about definitions and scope of concepts such as frames, footing and
alignment if our analytic work has to have external validity.
Interaction in the professional setting is not a skill, but a knowledge system,
laminated with expertise and authority that we as interaction analysts need to un-
derstand in order to navigate our analytic apparatus. The need for collaborative
interpretation is paramount, given that interpretation of themes in professional
contexts is not as straightforward as it might seem (Sarangi 2007). In my view,
interaction analysis, especially with regard to professional and institutional set-
tings, has to be recipient-designed in practically relevant ways in order to facilitate
potential uptake (Sarangi 2005a, Sarangi 2007). What professionals may want
from interaction analysts is not directives about how to conduct social interaction
in their communities of practice, but theoretical and analytic insights for reflec-
tion as they are likely to apply such insights about interaction selectively, in the
same way as they deal with theories and models of scientific and technical
Healthcare interaction as an expert communicative system 

knowledge. Analytic sensibility of the kind emerging from conversation analysis,

interactional sociolinguistics and other traditions of pragmatics and discourse
analysis has no doubt demonstrated the dynamic workings of language in interac-
tion, including context-specificity as well as stable patterning of interaction types
across contexts. We may ask, with specific reference to the positioning of discourse
and interaction analysts, whether our analytic enterprise is more of an intellectual
rather than professional expert-like endeavour. I am inclined to see our analytic
endeavour as one that maps on to other-oriented professional expertise rather
than becoming an exercise in pure intellectualism.


Transcription Conventions
(.): micropause;
(…): pause exceeding one second;
((gap)): indicates an interval of longer length between speaker turns and an ap-
proximation of length in seconds;
underlining: indicates increase emphasis as in stress;
-: indicates cut-off of prior word or sound;
[text in square brackets]: overlapping speech;
((text in double round brackets)): description or anonymised information;
(text in round brackets): transcriber’s guess;
=: a continuous utterance and is used when a speaker’s lengthy utterance is broken
up arbitrarily for purposes of presentation.


Adelswärd, Viveka and Sachs, Lisbeth. 1998. “Risk discourse: recontextualisation of numerical
values in clinical practice”. Text 18 (2): 191–211.
Agar, Michael. 1985. “Institutional discourse”. Text 5 (3): 147–168.
Atkinson, Paul and Housley, William. 2003. Interactionism. London: Sage.
Bakhtin, Mikhail. 1986. Speech Genres and Other Late Essays. Trans. Vern W. McGee, edited by
Caryl Emerson and Michael Holquist. Austin: University of Texas Press.
Bales, Robert Freed. 1950. Interaction Process Analysis. Mass: Addison-Wesley.
Blumer, Herbert. 1969. Symbolic Interaction. Englewood Cliffs: Prentice-Hall.
Boden, Deirdre. 1994. The Business of Talk: Organisations in Action. Cambridge: Polity Press.
Byrne, Patrick S. and Long, Barrie E. L. 1976. Doctors Talking to Patients. London: The Royal
College of General Practitioners.
Candlin, Christopher N. and Candlin, Sally. 2003. “Healthcare communication: A problematic
site for applied linguistics research”. Annual Review of Applied Linguistics 23: 134–154.
 Srikant Sarangi

Cicourel, Aaron V. 1974. Cognitive Sociology: Language and Meaning in Social Interaction. New
York: The Free Press.
Davis, Alan G. 1982. Children in Clinic: A Sociological Analysis of Medical Work with Children.
London: Tavistock.
Department of Health 2001. The Expert Patient: A New Approach to Chronic Disease Manage-
ment for the 21st Century. London: Department of Health.
Drew, Paul and Heritage, John (eds) 1992. Talk at Work: Interaction in Institutional Settings.
Cambridge: Cambridge University Press.
Edwards, Adrian, Elwyn, Glyn and Mulley, A. J. 2002. “Explaining risks: turning numerical data
into meaningful pictures”. British Medical Journal 324: 827–830.
Ferrara, Kathleen W. 1994.Therapeutic Ways with Words. New York: Oxford University Press.
Fisher, Sue and Todd, Alexandra D. (eds) 1983. The Social Organisation of Doctor-Patient Com-
munication. Washington DC: Centre for Applied Linguistics.
Foucault, Michael. 1970. L’Ordre du discours. Paris: Gallimard.
Freidson, Eliot. 1970. Profession of Medicine: A Study of the Sociology of Applied Knowledge. New
York: Dodd, Mead and Company.
Freidson, Eliot. 1994. Professionalism Reborn: Theory, Prophecy and Policy. Cambridge: Polity Press.
Garfinkel, Harold. 1967. Studies in Ethnomethodology. Englewood Cliffs, NJ: Prentice Hall.
Gee, James P. 1997. “Thematised echoes”. Journal of Narrative and Life History 7 (1–4): 189–196.
Glassner, Barry. 1980. Essential Interactionism: On the Intelligibility of Prejudice. London:
Routledge & Kegan Paul.
Goffman, Erving. 1971. Relations in Public: Microstudies of the Public Order. Harmondsworth:
Penguin Books.
Goffman, Erving. 1974. Frame Analysis. New York: Harper & Row.
Goffman, Erving. 1981. Forms of Talk. Oxford: Blackwell.
Goffman, Erving. 1983. “The interaction order”. American Sociological Review 48: 1–17.
Goodwin, Charles. 1994. “Professional vision”. American Anthropologist 96 (3): 606–633.
Green, Judith L. and Wallat, Cynthia. 1981. “Mapping interactional conversations – a sociolin-
guistic ethnography”. In Ethnography and Language in Educational Settings, Judith L. Green
and Cynthia Wallat (eds), 161–195. Norwood, NJ: Ablex, 161–195.
Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.
Gumperz, John and Cook-Gumperz, Jenny. 1982. “Introduction: language and the communica-
tion of social identity”. In Language and Social Identity, John J. Gumperz (ed), 1–21.
Cambridge: Cambridge University Press.
Heath, Christian. 1986. Body Movement and Speech in Medical Interaction. Cambridge:
Cambridge University Press.
Heath, Shirley Brice. 1979. “The context of professional languages: an historical overview”. In
Language in Public Life, James E. Alatis and G. Richard Tucker (eds), 101–118. Washington,
DC: Georgetown University Press.
Heritage, John and Sefi, Sue. 1992. “Dilemmas of advice: aspects of the delivery and reception of
advice in interactions between health visitors and first-time mothers. In Talk at Work: In-
teraction in Institutional Settings, Paul Drew and John Heritage (eds), 359–417. Cambridge:
Cambridge University Press.
Heritage, John and Maynard, Douglas W. (eds) 2006. Communication in Medical Care: Interaction
between Primary Care Physicians and Patients. Cambridge: Cambridge University Press.
Hughes, Everett C. 1958. Men and their Work. New York: Free Press.
Healthcare interaction as an expert communicative system 

Illich, Ivan. 1977. Disabling Professions, with contributions from Irving K. Zola, John McNight,
Jonathan Caplan and Harley Shaiken. London: Marion Boyars.
Johnson, Terence J. 1972. Professions and Power. London: Macmillan.
Johnson, Terence J. 1993. “Expertise and the state”. In Foucault’s New Domains, Mike Gane and
Terry Johnson (eds), pp 139–152. London: Routledge.
Labov, William and Fanshel, David. 1977. Therapeutic Discourse: Psychotherapy as Conversation.
New York: Academic Press.
Labov, William and Waletsky, Joshua. 1967. “Narrative analysis: oral versions of personal experi-
ence”. In Essays on the Verbal and Visual Arts, June Helm (ed) 12–44. Seattle: University of
Washington Press.
Larson, Magali Sarfatti. 1977. The Rise of Professionalism: A Sociological Analysis. Berkeley:
University of California Press.
Levinson, Stephen C. 1979. “Activity types and language”. Linguistics 17 (5/6): 365–399.
Levison, Arnold. 1974. Knowledge and Society: An Introduction to the Philosophy of the Social
Sciences. Indianapolis: Bobbs-Merrill.
Lynch, Michael and Sharrock, Wes. 2003. “Editors’ Introduction”. In Harold Garfinkel, Michael
Lynch and Wes Sharrock (eds), vii–xlvi. London: Sage.
Maynard, Douglas W. 1991. “The perspective-display series and the delivery and receipt of diag-
nostic news”. In Talk and Social Structure, Deirdre Boden and Don H. Zimmerman (eds),
164–192. Cambridge: Polity Press.
McHugh, Peter. 1968. Defining the Situation: The Organisation of Meaning in Social Interaction.
Indianapolis: Bobbs-Merrill.
McIntosh, Jim. 1978. “The routine management of uncertainty in communication with cancer
patients”. In Relationships between Doctors and Patients, Alan G. Davis (ed), 106–131.
Farnborough: Saxton House.
Mead, George H. 1934. Mind, Self, and Society: From the standpoint of a Social Behaviourist.
Edited with Introduction by Charles W. Morris. Chicago: University of Chicago Press.
Mehan, Hugh. 1979. Learning Lessons. Cambridge, Mass.: Harvard University Press.
Merton, Robert K. 1957. Social Theory and Social Structure, revised edition. New York: Free Press.
Mishler, Elliot G. 1984. The Discourse of Medicine: Dialectics of Medical Interviews. Norwood,
NJ: Ablex.
Mishler, Elliot G., Clark, Jack A., Ingelfinger, Joseph and Simon, Michael P. 1989. “The language
of attentive patient care: A comparison of two medical interviews”. Journal of General Inter-
nal Medicine 4 (July/August): 325–335.
Moore, Alison, Candlin, Christopher N. and Plum, Gunther. 2001. “Making sense of viral load:
one expert or two?” Journal of Culture, Health and Sexuality 3 (4): 429–450.
Peräkylä, Anssi. 1995. AIDS Counselling: Institutional Interaction and Clinical Practice.
Cambridge: Cambridge University Press.
Peräkylä, Anssi and Vehviläinen, Sanna. 2003. “Conversation analysis and the professional
stocks of interactional knowledge”. Discourse & Society 14 (6): 727–750.
Peräkylä, Anssi, Ruusuvuori, Johanna and Vehviläinen, Sanna. 2005. “Professional theories and
institutional interaction”. Communication & Medicine 2 (2): 105–109.
Perlman, Helen H. 1957. Social Casework: A Problem Solving Process. Chicago: The University of
Chicago Press.
Polanyi, Michael. 1958. Personal Knowledge: Toward a Post-Critical Philosophy. London:
Routledge & Kegan Paul.
 Srikant Sarangi

Prior, Lindsay. 2003. “Belief, knowledge and expertise: The emergence of the lay expert in med-
ical sociology”. Sociology of Health & Illness 25: 41–57.
Roberts, Celia and Sarangi, Srikant. 2002. “Mapping and assessing medical students’ interac-
tional involvement styles with patients”. In Unity and Diversity in Language Use, Kristyan
Spelman Miller and Paul Thompson (eds), 99–117. London: Continuum.
Roberts, Celia and Sarangi, Srikant. 2005. “Theme-oriented discourse analysis of medical en-
counters”. Medical Education 39: 632–640.
Rueschemeyer, Dietrich. 1986. Power and the Division of Labour. Stanford: Stanford Univer-
sity Press.
Ryle, Gilbert. 1949. The Concept of Mind. London: Hutcheson.
Sarangi, Srikant. 2000. “Activity types, discourse types and interactional hybridity: The case of
genetic counselling”. In Discourse and Social Life, Srikant Sarangi and Malcolm Coulthard
(eds), 1–27. London: Pearson.
Sarangi, Srikant. 2001. “On demarcating the space between ‘lay expertise’ and ‘expert laity’”. Text
21 (1/2): 3–11.
Sarangi, Srikant. 2002. “The language of likelihood in genetic counselling discourse”. Journal for
Language and Social Psychology 21 (1): 7–31.
Sarangi, Srikant. 2004. “Towards a communicative mentality in medical and healthcare prac-
tice”. Communication & Medicine 1 (1): 1–11.
Sarangi, Srikant. 2005a. “The conditions and consequences of professional discourse studies”.
Journal of Applied Linguistics 2, 3, 371–394.
Sarangi, Srikant. 2005b. “Interactional expertise in healthcare encounters”. Communication &
Medicine 2 (2): 103–104.
Sarangi, Srikant. 2005c. “Activity analysis in professional discourse settings: The framing of risk
and responsibility in genetic counselling”. Hermès 41: 111–120.
Sarangi, Srikant. 2007. The anatomy of interpretation: Coming to terms with the analyst’s para-
dox in professional discourse studies. Text & Talk 27, 5, 567–584.
Sarangi, Srikant. [in press]. “Practising discourse analysis in healthcare settings”. In Qualitative
Methods in Health Research, Ivy Bourgeault, Raymond DeVries and Robert Dingwall (eds).
London: Sage.
Sarangi, Srikant, Bennert, Kristina, Howell, Lucy, Clarke, Angus., Harper, Peter, and Gray, Jona-
thon. 2004. “Initiation of reflective frames in counselling for Huntington’s Disease predic-
tive testing”. Journal of Genetic Counselling 13 (2): 135–155.
Sarangi, Srikant, Bennert, Kristina, Howell, Lucy, Clarke, Angus., Harper, Peter, and Gray, Jona-
thon. 2005. “(Mis)alignments in counselling for Huntington’s Disease predictive testing:
Clients’ responses to reflective frames”. Journal of Genetic Counselling 14 (1): 29–42.
Sarangi, Srikant and Roberts, Celia. (eds) 1999. Talk, Work and the Institutional Order: Discourse
in Medical, Mediation and Management Settings. Berlin: Mouton de Gruyter.
Sarangi, Srikant and Slembrouck, Stefaan. 1996. Language, Bureaucracy and Social Control.
London: Longman.
Scheff, Thomas J. 1968. “Negotiating reality: Notes on power in the assessment of responsibility”.
Social Problems 16 (1): 3–17.
Schutz, Alfred. 1964. Collected Papers vol II. The Hague: Nijhoff.
Shils, Edward. 1968. “Intellectuals”. In International Encyclopaedia of the Social Sciences vol. 7,
Edward Shils (ed), 399–415. New York: Macmillan.
Silverman, David. 1987. Communication and Medical Practice: Social Relations in the Clinic.
London: Sage.
Healthcare interaction as an expert communicative system 

Silverman, David. 1997. Discourses of Counselling: HIV Counselling as Social Interaction.

London: Sage.
Simmel, Georg. 1950. The Sociology of Georg Simmel. Translated and edited by Kurt H. Wolff.
New York: The Free Press.
Stehr, Nico. 1994. Knowledge Societies. London: Sage.
Stokes, Randall and Hewitt, John P. 1976. “Aligning actions”. American Sociological Review 41:
Strong, P. M. and Davis, Alan G. 1978. “Who’s who in paediatric encounters: morality, expertise
and the generation of identity and action in medical settings”. In Relationships Between
Doctors and Patients, Alan G. Davis (ed), 48–75. Farnborough: Saxton House.
Tannen, Deborah. (ed) 1993. Framing in Discourse. New York: Oxford University Press.
Tannen, Deborah and Wallat, Cynthia. 1993. “Interactive frames and knowledge schemas in
interaction: Examples from a medical examination/interview”. In Framing in Discourse,
Deborah Tannen (ed), 57–76. New York: Oxford University Press.
Ten Have, Paul. 1989. “The consultation as a genre”. In Text and Talk as Social Practice: Dis-
course, Difference and Division in Speech and Writing, Brian Torode (ed), 115–135.
Dordrecht: Foris Publications.
Tuckett, David, Boulton, Mary, Olson, Coral, and Williams, Anthony. 1985. Meetings between
Experts: An Approach to Sharing Ideas in Medical Consultations. London: Tavistock.
Turner, Roy. 1972. “Some formal properties of therapy talk”. In Studies in Social Interaction,
David Sudnow (ed), 367–396. New York: Free Press.
Voloshinov, V. N. 1987. “Discourse in life and discourse in art”. In Freudianism, I. Titunik and N.
Bruss (eds), 93–116. Bloomington: Indiana University Press.
Voysey, Margaret. 1975. A Constant Burden: The Reconstitution of Family Life. London: Routledge
& Kegan Paul.
Wilson, Thomas P. 1971. “Normative and interpretive paradigms in sociology”. In Understand-
ing Everyday Life, Jack D. Douglas (ed), 57–79. London: Routledge & Kegan Paul.
Wood, Fiona, Prior, Lindsay and Gray, Jonathon. 2003. “Translations of risk: decision making in
a cancer genetics service”. Health, Risk & Society 5 (2): 185–198.
Interacting with difficulty
The case of aphasia

Elizabeth Armstrong and Alison Ferguson

Edith Cowan University, Australia and University of Newcastle, Australia

Following brain damage after stroke, a language disorder known as aphasia

can significantly interfere with conversational flow due to both the individual’s
difficulties with word-finding and grammar, as well as their conversational
partner’s ability to deal with such difficulties. This chapter explores the kinds
of analyses used to date to illuminate the nature of interactions involving
people with aphasia and their conversational partners, and the difficulties they
encounter. In particular, Conversation Analysis (1977) and Systemic Functional
Linguistics approaches (Halliday 1994) are discussed, and the challenges
involved in applying such frameworks to clinical populations are examined.

In analysing talk as it occurs in everyday situations, observations can be made, and

hypotheses can be devised which add to our knowledge of individuals’ access to
meaning making and the way(s) in which speakers communicate with each other.
Current analyses of everyday conversations extend beyond the formalist analyses
that in the past often depended on monologic or literary texts, and challenge our
notions of language competence and what constitutes it. For many years, research-
ers and clinicians concerned with individuals with communication disorders
viewed their communication abilities and associated problems from a formalist
perspective of language. There was a focus on individuals’ lexical and syntactic
skills, and their performance on decontextualised language tasks was deemed to
reflect their underlying language competence. More recently, through the influ-
ence of pragmatic perspectives, discourse analysis, and conversation analysis, the
individual’s performance in everyday talk has become of increasing interest and
focus. Research into this area presents numerous challenges and benefits as it pro-
vides information on conversations that are inherently problematic, for different
reasons, across different disorders. In this chapter, we compare the psycholinguistic
and sociolinguistic contributions to the understanding of one particular
 Elizabeth Armstrong and Alison Ferguson

communication disorder, namely aphasia (a loss or impairment of language due to

specific brain damage e.g., stroke). We also contrast the contribution made by the
two very different approaches of conversation analysis and systemic functional
linguistics to an understanding of the effect of aphasia on everyday talk. Our ob-
servations drawn from this frame of reference draw attention to the insight that
not all the ‘problems’ associated with aphasia are directly related to the individual’s
brain damage, but instead arise from the contextual demands of the interaction.
We suggest that this recognition raises the important corollary that opportunities
for maximizing effective communication arise from working within this social
conversational context.
Aphasia has been chosen as the focus of this chapter not only because it is the
main focus of our own clinical research, but as it provides a specific instance of a
disorder in which breakdown in lexicogrammar is the essential characteristic, in
the presence of relatively intact ‘pragmatic’ skills (i.e., the aphasic speaker appears
to be trying to participate in the conversation/topic at hand, experiencing frustra-
tion when ‘meanings’ cannot be realized successfully). Aphasia can be contrasted
with other disorders that involve less disruption of lexicogrammar, and greater
disruption of so-called ‘pragmatic’ aspects of communication e.g., the ‘inappropri-
ate’ changing of topics, or the discussion of contextually inappropriate topics of
those with traumatic brain injury, or the confabulatory language characteristic of
dementia or schizophrenia (where the speakers often lack insight into their verbal
behaviours). While analysis of these latter disorders yields equally informative
data and interpretations, it is important to examine each disorder individually as
much as possible, with the aim being to focus on specific characteristics that can
be contrasted eventually with characteristics of other disorders, in order to estab-
lish their relative significance and to explore potential associations between proc-
esses and structures.

1. ‘Problems’ in conversation

Conversations can become problematic for a number of reasons and everyday dif-
ficulties are well documented in otherwise ‘normal’ speakers’ interactions. Misun-
derstandings can occur between interlocutors, leading to the need for clarification
and correction, or these can sometimes lead to conflict, where one person may be
offended by a misinterpretation of another’s comments, and even terminate the
interaction. Speakers can also experience ‘slips of the tongue’ (Cutler 1982) where
individual words or phrases are produced incorrectly, leading to the need for self-
correction and repair. Both the leading proponents of the two key approaches to
interaction analysis that we will be considering in this chapter – Conversation
Interacting with difficulty 

Analysis and Systemic Functional Linguistics - suggest that it is surprising that

there are not more problems in conversation, given the complexity of language.
However, while Schegloff, Jefferson and Sacks (1977) would suggest that interac-
tional machinery accounts for this relative ease of interaction, Halliday (1994)
would suggest the reason we understand each other so well is because of redun-
dancy and context.
For speakers with communication disorders and their conversational part-
ners, such difficulties as noted above can become amplified several times over, as
the interactants attempt to negotiate meanings under less than optimal conditions.
The effects of brain damage on a person’s interactive skills can be numerous. In
aphasia, the lexicogrammar is impaired, so that the speaker with aphasia has dif-
ficulty making the right lexical choices and even finding the correct lexical items,
and may also have difficulty in putting words together to form clauses, and estab-
lishing logical connections between the clauses. In severe cases, the disintegration
of language is so great that a listener can perceive little overall structure to the
discourse of the person with aphasia, in which case it is judged to be totally inco-
herent. The kinds of impairments following more diffuse brain damage, such as
that suffered after a traumatic brain injury or associated with dementia, are less
specific to the lexicogrammar, and involve more difficulties in linking the lexico-
grammar to the interpersonal context. In this regard, the speaker may not appear
to be having difficulty forming clauses or texts from a lexicogrammatical perspec-
tive, but the language is not always appropriate to the situation. These latter diffi-
culties are often described as ‘cognitive-communication’ disability, and thus dif-
ferentiated from the focus of this chapter on aphasia.
In the case of aphasia, the brain-damaged speaker struggles to use his/her re-
maining linguistic resources to convey meanings, whilst the conversational part-
ner uses all possible resources – both linguistic and extralinguistic knowledge, in-
corporating contextual variables – to try and make sense of what the other speaker
is saying, and to attempt to get the conversation back ‘on track.’ Often in the inter-
action between an aphasic speaker and communication partner, the person listen-
ing to the aphasic partner must become even more active than usual in his/her role
of co-construing the discourse, as inordinate amounts of ‘guessing’ and ‘filling in
the blanks’ have to occur to assist comprehension of what is being said. Converse-
ly, the aphasic speaker may also have difficulties processing what is being said to
him/her because of receptive language difficulties (also part of aphasia), but have
limited resources to use to interpret/repair such problematic instances.
An interactional approach to language provides those interested in the dis-
course of people with communication disorders a way of examining not only how
an aphasic person ‘copes’ and functions during conversation i.e., utilizes his/her
skills to convey meaning given diminished linguistic resources, but also two other
 Elizabeth Armstrong and Alison Ferguson

perspectives. One involves examining the effects of different conversational part-

ners (and hence different styles of interacting) on the aphasic person’s use of his/
her resources, and the second involves examining the ways in which the conversa-
tional partners ‘handle’ the aphasic person’s contribution to the conversation. Ob-
viously these two perspectives are inextricably interwoven and should be seen to-
gether as the ways in which both parties negotiate what can potentially be
problematic meaning-making by both partners (given that aphasic individuals
may also have comprehension difficulties, making it problematic at times for them
to follow what the non-brain damaged person is saying).

2. Analyses/frameworks used to date

Historically, the study of communication disorders has been approached from a

number of perspectives, reflective of the multi-disciplinary interest in language.
Formal linguistic and psycholinguistic frameworks have been perhaps the most
commonly utilised, aphasia research focusing for many years primarily on the lexi-
cogrammatical difficulties evidenced by aphasic speakers at the word and sentence
levels, and more recently in the context of discourse, but usually monologic dis-
course. The focus of analysis in this case is very much an intra-individual one, with
the language ‘disorder’ identified solely in the language of the aphasic speaker. Lan-
guage is viewed from a synoptic perspective in this paradigm, where it is seen as a
set of rules followed (or not) within the mind of the individual speaker. However,
in more recent years, the focus has expanded somewhat to include the ways in
which lexicogrammatical difficulties manifest themselves in an interactional con-
text, and the role of the interlocutor in co-constructing the discourse with the apha-
sic partner. Such an approach uses an inter-individual perspective, where language
is viewed as a dynamic phenomenon, designed and used for social purposes.
The earliest clinical analyses of interactions used a variety of ‘pragmatic’ anal-
yses that are still used, largely based on the work of Searle (1969) and Grice (1975).
Such analyses stem from the perspective that pragmatics involves ‘language in use.’
Pragmatic rules are often seen as separate from linguistic rules that govern pho-
nology, the lexicon and syntax – elements that are impaired in aphasia, for exam-
ple, as discussed in the work of Prutting and Kirchner (1983), Murray and Chapey
(2001), and Myers (1999). However, Carlomagno and colleagues (2000) challenge
this view in a paper that examines both the current and potential use of so-called
pragmatic principles in the treatment of aphasic individuals. While they acknowl-
edge the historical separation of pragmatics and lexicogrammar in many aphasia
assessment and treatment paradigms, they highlight the need to examine both
pragmatic ‘functions’ and aspects of the lexicogrammar as they interact. Ulatowska
Interacting with difficulty 

and Olness (2000) support this view. Traditionally, language therapy often relied
on word-finding activities involving picture naming tasks, or grammatical activi-
ties such as sentence construction tasks. Carlomagno and colleagues suggest that
therapy focus on both formal aspects of the lexicogrammar, as well as its use in
interactions. For instance, they highlight the necessity of taking context into ac-
count when using co-reference. They discuss the effects that working within a con-
versational framework can have on specific aspects of the lexicogrammar, for ex-
ample, Wertz and colleagues (1981), Springer and colleagues (1991).
Another approach has been based on Conversation Analysis (CA) and has
been used by a number of researchers to address the patterns of interaction occur-
ring in aphasic conversations (Beeke et al. 2007a). Much of the work in applying CA
to aphasia has been influenced by the pioneering work of the linguist, Charles
Goodwin (1981), whose analysis of interactions involving a speaker with severe
expressive aphasia (his father) and others illustrated the co-construction of mean-
ing (Goodwin 1995), and the emergence of communicative competence through
the joint work of the partners, rather than seeing (in)competence as an attribute of
an individual (Goodwin 2003a). Two main areas preoccupy the research using con-
versation analysis to study aphasia, firstly investigating how communication effec-
tiveness might be facilitated through self and other repair (Ferguson 1994, 1998;
Lesser 2003; Lindsay and Wilkinson 1999), and secondly the design and evaluation
of conversational therapy programs (Boles 1998; Booth and Swabey 1999; Whit-
worth et al. 1997; Wilkinson et al. 1998). Parallel with these developments, a number
of aphasia researchers have been exploring the applications of more eclectic frame-
works such as found in the interactional sociolinguistics of Gumperz (1982), and
within these broader frameworks have considered issues of power in interactions
involving a person with aphasia (Simmons-Mackie and Damico 1999).
Systemic functional linguistic principles (SFL) (Halliday and Matthiessen
2004) have also been used to describe aphasic interactions in terms of speech func-
tions (Ferguson 1992b), and generic structure potential (Ferguson and Elliot
2001), with Armstrong (1991; 1993) discussing the implications of breakdowns in
cohesion for interactions. Such applications will be explained in more detail below,
but it has been the link between lexicogrammar and context, central to the work of
Michael Halliday in systemic functional theory, that has attracted the authors to
description of disordered conversations using this framework. As Eggins & Slade
(1997/2004) note:
“Structural-functional approaches ask just what is conversational structure, and
attempt to relate the description of conversational structure to that of other units,
levels, and structures of language.” (p.43).
 Elizabeth Armstrong and Alison Ferguson

It is this relating between levels of language, looking for how one is realized by
another, through which SFL offers a different perspective from other forms of in-
teractional analysis. The following section will explore this further.

3. Connecting lexicogrammar with context

One of the recent developments in the application of CA principles to aphasic

discourse has been an increasing focus on examining the links between the lexico-
grammar and context during conversation. In an attempt to become more specific
regarding the links between lexicogrammar and the aphasic speakers’ and their
conversational partners’ ability to negotiate meaning, utilizing a CA framework,
Beeke and colleagues (2003a; 2003b; 2007b) have explored ways in which the pat-
terns of agrammatic constructions of an aphasic speaker during interactions may
well be attempts to ‘manage’ the conversation successfully in light of difficulties at
the lexicogrammatical level. Beeke and colleagues discuss phenomena such as
‘fronting’ or ‘left dislocation’ (e.g., a noun phrase being used at the beginning of a
clause, then elaborated upon by more simple constructions in aphasic talk). Such
constructions deny canonical order of clauses, but function in normal discourse to
highlight or thematise particular aspects. In aphasic discourse, however, Beeke
and colleagues describe such usage as a means by which aphasic speakers are able
to convey meaning in the presence of lexicogrammatical difficulties. For example,
when discussing her participation in past wedding preparations, the aphasic
speaker begins with the temporal noun phrase, ‘July no June….three tier wedding
cake I make it.’ Because of the presence of the noun phrase, the listener does not
misinterpret the lack of tense marker otherwise required for indicating the past.
Another phenomenon noted was termed a ‘sequential proposition’ pattern. This
involved the construction of a proposition without the normal conjoining ele-
ments, but interpretation depending on a contextual sequence. For example, the
aphasic speaker just mentioned was discussing making a wedding cake for her
friend, and having to buy particular hexagonal cake-tins for this task:
Jane: oh what sort are they havin’
Connie: m tuh hexagon shape m tuh [I
Jane: [ you got tins
for that
Connie: no no uh – I have the tins
Jane: you gotta get ‘em
Connie: yeah yeah
Jane: oh right
Interacting with difficulty 

Connie: m Edgerton green cake shop

Jane: oh is there one there?
Connie: yeah
Beeke, Wilkinson and Maxim (2003b) also reported on differences found between
grammatical usage in sentence tests, and structured elicited narratives to that found
in conversational interaction. In the more structured situations, their speaker dem-
onstrated ability to construct SVO type structures. However, few utterances in the
conversational sample conformed to the SVO type format. In contrast, the phe-
nomena of fronting and sequential construction of a proposition were contained in
the conversational data, but rarely occurred in the structured elicited data. Such
data suggest that studying language as it occurs in normal interaction may well tell
us very different things about the language of aphasic speakers than data gained in
either decontextualised or monologic tasks. Beyond this implication, as Lesser
(2003) points out, there is increasing recognition that what is observable in aphasic
language use is as much the speaker’s adaptation to communication difficulty as
reflecting what psycholinguistics would describe as the ‘underlying linguistic im-
pairment’ (Beeke et al. 2007a; Heeschen and Schegloff 1999).
This kind of analysis appears to go beyond previous CA approaches to aphasia
in as much as it looks at the semantic function of the patterns found a little more.
For example, in the texts from which the above brief examples were drawn, a phe-
nomenon known as fronting was identified, with its potential purpose also pro-
posed i.e., highlighting or thematising. In the next section of this chapter, we ex-
plore further research within a systemic functional framework, which addresses
some of these issues.

4. SFL’s perspective on contextual issues

While it is clear that CA is certainly concerned with the context of utterances, it

appears at times that this ‘contextualisation’ is either local to the immediately sur-
rounding utterances only, or global in terms of its description of predetermined
patterns of interaction, that do not relate to specific contexts, but can be generalized
across a variety of genres e.g., the notion of adjacency pairs. SFL, on the other hand,
characterizes context in terms of register and genre. Numerous linguists and socio-
linguists have written extensively on context, but Halliday’s notion of context of
situation (Halliday and Hasan 1985), expanding on the work of Firth (1957) pro-
vides a solid framework by which to characterize contexts as belonging to particu-
lar registers and genres. The framework enables the analyst to explore language in
context in a systematic way, examining the three aspects ‘field’ (referring to what is
 Elizabeth Armstrong and Alison Ferguson

occurring and what is being discussed), ‘tenor’ (referring to the relative status of the
interlocutors) and ‘mode’ (referring to the channel of communication e.g., written,
spoken) that relate to the three social functions of language: i) to understand and
represent our environment and experience of that environment ii) “to represent
our experience to each other’; and iii) to organise our enactments and representa-
tions as meaningful text” (Martin and Rose 2003:€6). These latter three purposes are
known as the experiential, interpersonal, and textual metafunctions of language. At
an even higher level, one can characterize genres, which are said to “represent a
system of staged goal-oriented social processes through which social subjects in a
given culture live their lives” (Martin 2002:€56–57) in terms of particular patterns
and structures of language required for particular social purposes.
The importance of this higher level of abstraction of context is that discourses
can be seen as having relatively predictable structures, and generalizations about
behaviours can be made beyond the single instantiation of any one event. Sys-
tematicity is central to such a framework, so that a finite set of options is available
to speakers in particular situations, and a level of probability is possible to gauge
in terms of how a particular situation will unfold at various linguistic levels.
When working with data obtained from individuals with communication dis-
orders, such abstraction can be very useful. Firstly, it provides some notion of
‘norm’ for particular genres, about which judgements can be made re pathology/
no pathology. Conversely, it can allow the analyst to isolate particular contextual
variables that may be affecting the behaviours observed. For example, in the anal-
ysis of field, tenor and mode, variables of field can be isolated (topic, nature of in-
teraction), tenor (social distance between conversational participants, familiarity)
and mode (e.g., face to face discourse, telephone interaction, SMS messaging).
This framework has proved useful in a series of studies on individuals who have
neurogenic communication disorders (Armstrong et al. 2005/2007; Armstrong
and Mortensen 2006; Ferguson 2007; Ferguson and Armstrong 2004; Togher 2000;
Togher and Hand 1998; Togher et al. 1997). Findings such as the fact that familiar-
ity has a significant impact on interactions for these individuals (Ferguson 1994;
Togher et al. 1997), and that written vs. spoken modes invoke different use of
grammatical resources (Mortensen 2003) are examples of the ways SFL can inform
clinical practice. While the effects of familiarity may be fairly obvious i.e., speakers
interact differently when levels of shared knowledge and comfort within the inter-
action vary, it is the ways in which language varies that is of interest. Such phe-
nomena as differences in modality manifest themselves in specific ways in interac-
tions involving individuals with brain damage. For example, in investigating
individuals with traumatic brain injury, Togher and Hand (1998) found that these
individuals were less able to use modality for politeness purposes in service en-
counters with unfamiliar persons than their non-brain-damaged counterparts.
Interacting with difficulty 

Such phenomena may account at least partially for the perceived bluntness of this
group and their perceived lack of social skills.
The importance of context and an understanding of how meaning is ‘con-
structed’ in an interaction between someone with a communication disorder and
their conversational partner can be viewed as central. SFL’s multistratal and multi-
functional dimensions enable the analyst to examine speakers’ utterances from
multiple perspectives and in so doing, add a depth to analysis not always available
in unidimensional analyses. For example, when exploring aspects such as gram-
matical structure, the role of certain structures in creating meaning is examined.
The use of modality has been examined for its role in the aphasic speaker’s ability
to negotiate conversations and in politeness (Ferguson 1992b). Many grammatical
analyses used in aphasia research tend to quantify features such as the amount of
clausal embedding, or use of tense endings, without looking at the effects of these
on overall discourse meaning. Similarly, ‘repair’ has been addressed without
directly exploring the role of lexicogrammar in the behaviour (Lindsay and
Wilkinson 1999; Perkins 1995). Ferguson’s study of repair (1992a; 1993), however,
incorporated the role of cohesion between turns in the analysis of guessing tech-
niques used by conversational partners (see below).

5. Lexical retrieval in interaction

While aphasia is a heterogeneous disorder, the symptom common to all types and
degrees of severity of aphasia is anomia or word-finding difficulty. Armstrong’s
research considers the availability of lexicogrammatical resources for cohesion for
people with aphasia and demonstrates the difficulties aphasic speakers have pro-
viding clear reference (often using pronouns and demonstratives without explicit
antecedent referents) and producing the variety of lexical sense relations that as-
sists in creating a coherent text. These difficulties in providing explicit cohesive
ties linking ideas within the text place an increased load on the communication
partner as the partner attempts to build a coherent view of the meanings being
conveyed, but to-date most research in aphasia has considered anomia as a prob-
lem of the individual with aphasia rather than recognizing its impact on the inter-
action as a whole.
When communication partners converse with anyone who is having word-
finding difficulty, they frequently will supply the word or prompt the speaker to be
able to find the word. This ‘supplying words’ is a particular feature of conversations
involving a person with aphasia. Naturally, world-knowledge and shared knowl-
edge are major contributors to assist with the guesses made by the partner, but
another important resource is the co-text that immediately precedes the word
 Elizabeth Armstrong and Alison Ferguson

being sought. Ferguson (1992a) combined conversation analysis methodology

with systemic functional linguistics to explore the role of co-text in supplying
words for people with aphasia (Armstrong 1987, 1991). Ferguson looked at
21 conversations - 7 conversations between an aphasic speaker and spouse, 7 con-
versations between an aphasic speaker and their friend or neighbour, and 7 con-
versation between the spouse and the friend or neighbour, and identified moments
in the conversations where significant trouble and repair arose, in that these mo-
ments extended over at least 5 turns, and involved at least two turns of hypothesis-
forming as defined by Bremer and colleagues (Bremer et al. 1996). In the following
Example 1, a typical moment of trouble in the interaction has occurred, in which
‘A’ (a 60 year old man who had suffered a left cerebrovascular accident 4.5 years
prior to this conversation being taped) is talking about fishing with ‘LF’ (his 33 year
old son-in-law).
Example 1: Word-finding difficulty in interaction
LF You ever been deep sea fishing?
A Oh yeah I been outside, um, Port Stephens. Went out with Maxie, and
Teddy. He said um, … he said, I’ll show you how to, hood [PARAPHASIC
ERROR] the r-rod up.€ And I was shrowing [PARAPHASIC ERROR] to
throw it in. And it was a very very rough sea, and I just X the XXX [UNIN-
TELLIGIBLE SYLLABLES]. I said I can’t feel anything. Listen how do, don’t
you feel that tug? Yes. He said, hook ‘em! And I hooked, and next minute
they brought him in. Two pound, …um.. christ, can’t think of it now.
LF Was it bream?
A No.
LF Snapper?
A …Flathead, flathead, yeah
LF Flathead! Deap sea, and you got a flathead!
A That’s right. Well, we’re on the bottom, and we got about sixty flathead.
That’s where we were.
This type of guessing sequence obviously also occurs in interactions not involving
an aphasic interactant, though in this example the son-in-law’s request for further
confirmation of the type of fish (‘flathead’) may well reflect uncertainty that the
sought-after word had in fact been found, as his world knowledge of fishing sug-
gested that this would be an unusual catch in deep sea fishing. His experience with
aphasic interaction would alert him to the higher probability of misrecognition of
target words and the common occurrence of semantic errors involving close se-
mantic associates.
In Ferguson’s study, the text immediately preceding moments of significant
trouble was analysed by identifying the semantic relationships of the words
Interacting with difficulty 

supplied to those sought based primarily on the systemic linguistic work on cohe-
sion (Halliday and Hasan 1976) – namely, lexical association of superordinate,
cohyponym, and repetition, as well as other collocational relationships. There were
57/63 guesses involved at these moments of trouble for which it was possible to
identify immediately preceding co-textual resources. Of the 57 guesses, 31 (54%)
were accepted by the partner (i.e., either explicitly accepted, or no further repair)
and 26 (46%) were rejected by the partner. Lexical associates proved to be more
strongly represented in the immediately preceding co-text for rejected guesses
(54% of resources were lexical associates) than for accepted guesses (23% of re-
sources were lexical associates). The converse was found for general collocation,
which was more strongly associated with accepted guesses (26% of resources) than
rejected guesses (4% of resources). This research suggests that the frequency with
which words co-occur in the language provides an important potential resource
for people in interactions involving repair of word-finding, and illustrates the use-
fulness of analysing the interaction as a unit rather than separating the contribu-
tions of the aphasic speaker from those of the communication partner.

6. Interpersonal resources

It is often remarked that people with aphasia have marked preservation of social
and interpersonal interaction skills, despite even severe to profound levels of apha-
sic impairment. While most of the research into aphasic communication has tend-
ed to focus on the difficulties in communication, both conversation analysis and
systemic functional linguistic perspectives have turned the focus of research to the
remaining strengths in communication, and so have illuminated some of the means
by which people with aphasia do manage to communicate despite their difficulties.
For example, how might people with aphasia manage the linguistic resources for
modality that allow speakers to modulate and shade meanings and manage the
politeness work required for interpersonal interaction? Ferguson considered the
linguistic resources available to speakers with moderate-severe aphasia for polite-
ness (Ferguson 1992b), and found that the aphasic speakers (in a series of short
role plays) made use of all available resources (including such devices as shifts in
verb tense, negation, comment adjuncts, use of interrogative forms for giving com-
mands). The extent of modulation remained weak e.g.€using forms such as ‘can’
rather than the more modulated ‘could’, but modulation was clearly an available
resource for interaction, even in the presence of significant aphasic difficulty.
Partners of people with aphasia accommodate for the communicative diffi-
culty in diverse ways, and Ferguson’s conversation analysis research suggested that
communication partners may simply make more use of personal interactive styles
 Elizabeth Armstrong and Alison Ferguson

for conversational repair, since they were observed to adopt similar patterns of
repair with both partners with and without aphasia (Ferguson 1994). However, at
least for some people, the presence of communication difficulty appears to lead to
‘over-accommodation’ for the difficulty (Giles et al. 1992; Giles and Smith 1979).
The following Example 2, provides an illustration of both the preserved story-
telling skills of a person with aphasia (in a continuation of the conversation from
the same dyad as described in Example 1) as he tells his story of the fish that got
away, as well as providing an example of overt ‘other-repair’ which would be con-
sidered marked in an interaction between father and son-in-law if aphasic diffi-
culty was absent.
Example 2. Other-repair of trouble in aphasic interaction
A You could see him. It was just on dusk.
LF And you had him on your line?
A Yeah, he was there, break the, you could see him, um, out about, ah seven
metres. X..
LF How far?
A ..About..
LF ..Seven what, metres, feet?1
A I don’t now, I was going to say metres, but X
LF Don’t worry about how far. But he was close enough to see.
A I’ll say.
LF Was he in the waves?
A Yes. And there was three, lifesavers there, XX come on get him! Bring him
in! They seen him too! I said, uh-oh, he’s gone. What!.. Couldn’t believe it.
In Example 2, the son-in-law attempts to clarify the distance being talked about,
firstly through an attempt to clarify the specific term used (‘metres’, ‘feet’), then
appears to let that correction go in order to re-establish the flow of the story, but
then seeks to make the clarification through establishing other information (‘close
enough to see’, ‘in the waves’). Arguably, the closeness of the fish to the beach
makes little difference to the point of the story (that the fish got away), and such
other-repair may have covert ‘therapeutic’ purposes that are rarely seen in non-
aphasic interaction, i.e., an attempt by the partner to provide practice and achieve-
ment of word-finding for the person with aphasia.

1. The metric system of measurement was introduced in Australia during the 1970s, and so
the 60 year old man would be more familiar with imperial measures, while the son-in-law would
be more familiar with metric. However, many people in Australia continue to use ‘inch’ and ‘feet’
measures as general descriptors in casual conversation while using the metric system in accurate
measurement situations.
Interacting with difficulty 

Sometimes over-accommodation can take the form of a marked intonation

pattern adopted when talking with the person with aphasia in a similar fashion to
what is sometimes described as ‘elder-speak’ (Kemper et al. 1998). An in-depth
case report of such accommodation using intonation was provided by Ferguson
and Peterson (2002), in which a 51 year old female neighbour used a marked into-
nation pattern when addressing comments to a 75 year old man but not when talk-
ing with his 72 year old wife. To consider this intonation pattern with reference to
the social interaction, Brazil’s framework (1981) was used for the categorisation of
pitch contours (falling, rising, fall-rise, rise-fall) with reference to information roles
(contrastive, additive, equivalence) and social roles (congruence/divergence). Fer-
guson and Peterson found that the speaker adopted intonation contours associated
with marked patterns of divergence and contrast, and the researchers hypothesised
that this pattern of communication was designed to facilitate the auditory compre-
hension of the person with aphasia (as this was significantly impaired). This social-
semantic analysis of the interaction allows us to see further than the observation of
the behavioural anomaly, and this deeper level provides important opportunities
for us to go beyond description towards insights into potential explanations.

7. Interpreting ‘problems’ as pathological

When examining aphasic texts, there is an inherent temptation to examine how

they differ from those of non-brain damaged speakers - and in the case of interac-
tions, how conversations between an aphasic speaker and their partner differ from
conversations between partners with no communication impairment. There are ob-
vious benefits in approaching aphasic discourse from this perspective, and various
aphasiologists have recently highlighted aspects of ‘normal’ discourse so as to high-
light the need for clinicians to be aware of stylistic variation across normal speakers
before labelling the verbal behaviours of individuals with brain damage as ‘patho-
logical’ (Armstrong 2002; Mackenzie 2000; Van Leer and Turkstra 1999; Youse et al.
2001). Much of this data, however, relates to monologic discourse. Schegloff (2003)
encourages aphasiologists to pursue such an approach with respect to conversation,
and highlights the fact that phenomena observed in ‘aphasic’ conversations often
occur in ‘normal’ conversation as well. With this in mind, Schegloff also notes the
importance of making systematic contrasts between interactions with and without
speakers who have language pathology, since without these contrasts it becomes all
too easy to account for problems that arise in the talk in terms of pathology.
The gathering of data from non-brain-damaged speakers is important both clin-
ically and theoretically, in order to establish more comprehensive notions of which
patterns of interaction are qualitatively different from normal, and which patterns
 Elizabeth Armstrong and Alison Ferguson

may perhaps be simply quantitatively different. However, there is some debate as to

how useful such comparisons with normal data are if repair is not just quantitatively
different but also qualitatively different in interactions involving aphasic speakers.
As previously discussed earlier in this chapter, Wilkinson, Beeke and Maxim (2003)
describe what they see as adaptations made by an aphasic speaker that they suggest
are qualitatively different, for example, ‘fronting’ (bringing key words to the first
position in an utterance). The main thrust of their argumentation lies with the rec-
ognition that such use of language is adaptation by the speaker to the communica-
tive needs of the listener, rather than evidence of underlying agrammatism, and
there is considerable support for this stance (discussed elsewhere in this chapter).
However, their research fails to recognise how the aphasic speaker is making use of
both the available discourse-semantic resources for thematising key information
(commonly used by normal speakers), either as a clause theme, or as a hyper-theme
for a grouping of clauses (Martin 1992:€437–449), i.e. the aphasic speaker is making
use of a ‘normal’ resource, rather than displaying ‘pathological’ adaptation.
In addition, interesting data has been collected by some researchers
(e.g.€Croteau et al. 2004) that include attitudinal information on what interactions
were like before the neurological incident – again important information of both
clinical and theoretical significance. It is common for behaviours to be labelled
pathological merely because some neurological damage has been identified. And
yet, there is much variation in conversational style, so speakers’ and their partners’
perceptions of differences are just as important as currently observed ‘difficulties.’
The other perspective from which to view the interactions of individuals with
aphasia is not to compare them with ‘normal’ interactions, but to analyse them in
relation to themselves alone i.e., the negotiation of meaning within the couple and
its ‘success’ or otherwise. This perspective is perhaps more amenable to and com-
mon in clinical practice, where the aim of any intervention is to improve interac-
tions. A clinical intervention program designed by Locke and colleagues (Locke
et al. 2001a; Locke et al. 2001b) emphasises the focus on the aphasic person and
their conversational partner, where a collaborative approach is used to establish
which behaviours are problematic, and which facilitate in cases of conversational
‘trouble’ i.e., where some breakdown in communication occurs. Such a collaborative
approach is being increasingly used in conversational treatment programs (Boles
and Lewis 2003; Fox et al. 2009; Hopper et al. 2002; Turner and Whitworth 2006).

8. Brain & language interaction

In researching aphasia there is an almost irresistible urge to draw implications and

relationships between the language data and potential brain functional organization.
Interacting with difficulty 

Indeed, for much of the early history of aphasia research, brain and language rela-
tionships were the major focus of those seeking an understanding of localization
of function. With the advent of technology allowing accurate imaging of lesion
sites and dynamic brain function in the second half of the twentieth century
(Hillis 2002), work using aphasic data explicitly drew away from localization as a
goal, and focused upon theoretical modelling of components and processes within
the linguistic system. Seidenberg articulates the concern of many working in the
field when he points out that using brain-damaged performance to make infer-
ences about normal function is analogous to trying to find the best way to the
airport by looking at the back roads after there has been a blockage on the freeway
route (Seidenberg 1988). Even within information processing perspectives involv-
ing the design of neural networks that incorporate the real-time processing con-
straints known to operate in human neurophysiology, the search has been to un-
derstand how language may be learned and how it functions within undamaged
systems (Harley 1996; Plaut 1996), rather than relying heavily on inferences from
biologically damaged systems.
As discussed by Schegloff (Cmerjrkova and Prevignano 2003), this focus on
abstract theoretical modelling within cognitive science sees language as an inter-
action of separable components, with any contextual variables either ignored, or
treated as another factor that again can be separated. For those working with apha-
sia, this modular perspective has been both useful and limiting. The cognitive neu-
ropsychological approach to aphasia has been useful in allowing for close defini-
tion of aphasic deficits and clear articulation of therapy targets (Byng 1993).
However, as Lesser (2003) notes, such modelling may provide insights into serial
(conscious, controlled) linguistic processing but not parallel (automatic, contextu-
alized) linguistic processing. Hence, the problems encountered in the generaliza-
tion of therapy gains to real-world contexts are to be expected, when the contex-
tual demands of language use are not incorporated within both assessment and
therapy for aphasia.
Researchers in the field of aphasia have at times suggested, as Lesser (2003)
does, that pathological language use may offer a ‘window’ into psycholinguistic
processing. However, Heeschen and Schegloff (1999) argue against looking at
agrammatic errors as ‘windows into the brain’ with regard to linguistic processing,
arguing instead that such manifestations represent contextual adaptation. How-
ever, we raise the question as to whether other models of language use may have
the potential to offer insights into real-time, real-world linguistic processing. At
this stage we are intrigued by the way that the systemic functional linguistic mod-
el suggests a view of semantic networks based on a probabilistic model of associa-
tion (Hasan et al. 2005/2007) and query the extent to which such a probabilistic
model is compatible with emerging neural network theories investigating semantic
 Elizabeth Armstrong and Alison Ferguson

access and storage impairments (Gott and Plaut 2002). We ask whether such mod-
els may provide a way through to future understandings of the semantic retrieval
problems that rest at the heart of aphasia.
There is also the question as to whether research into the social interaction of
people with aphasia can illuminate our understanding of brain and language rela-
tionships. Lesser argues that the methodology of Conversation Analysis does not
allow such insights, given its particular focus on what she describes as “surface
data” (Lesser 2003:€152), and its atheoretical approach. A separate but related ques-
tion is whether or not the study of aphasic interactions might inform an under-
standing of normal social interaction (i.e., of the social machinery, rather than the
individual cognitive processing). Lesser suggests that the inclusion of speakers
with pathology will necessarily result in abnormal interactional patterns. In re-
sponse, we argue, in line with Goodwin (Goodwin 2003b: 17) and Perkins (2003)
that the presence of communicative difficulty exposes the workings of conversa-
tional collaboration. Conversational collaboration is always present in every inter-
action, but typically is as invisible as the technique of a team of professional jug-
glers – dropping the ball alerts us to the interactive processes involved.

9. The value of interactional analysis for clinical purposes

In undertaking interaction analysis of conversations of individuals with commu-

nication problems and their partners, the analyst may have any one of a number of
goals. As noted previously, analyses of language addressing language ‘skills’ as they
appear in monologic discourse tend to focus on the speaker’s ability to use linguis-
tic resources to convey meaning from a largely ‘synoptic’ perspective. Interaction
analyses, looking at dialogic interactions directly, illuminate the success or other-
wise of the aphasic speaker’s language attempts/utterances through analysis of the
response of the interlocutor and tend to be dealt with from a more dynamic per-
spective, focusing on how the text unfolds during the interaction. Similarly, inter-
action analyses can reveal the way(s) in which an aphasic speaker might use the
other speaker’s utterances to ‘scaffold’ his/her own (Perkins 2003), so that notions
such as ‘coherence’ do not reflect one speaker’s contribution, but are constructed
through the dialogue between the two speakers.
Interactional analysis necessarily drives clinicians away from sentence-level
formal assessment tasks, towards more natural assessment methods. There is a
constant tension between the need for maximising the validity of sampling through
natural observation and the clinical demands of cost/time-effective methods and
sufficiently controlled conditions to allow for comparison over time (to ascertain
change associated with recovery and/or therapy) and for comparisons across
Interacting with difficulty 

speakers (to allow for research development of therapy methods). Ramsberger and
Menn (2003) describe a naturalistic story-retell task involving the aphasic speaker
telling another person about an episode of the television show I Love Lucy, and
this represents a more naturalistic sampling method than similar story-retell tasks
involving picture description, for example the ‘Cookie Theft’ picture from the
Boston Diagnostic Aphasia Examination (Goodglass et al. 2000), or the drawn
cartoon stories, such as the ‘Cat Story’ (Ulatowska et al. 1981), and more recently
using a well-known Finnish cartoon (Korpijaakko-Huuhka 2004). Ferguson has
used an interactive task in which the person with aphasia watches a mock car-ac-
cident enacted using model cars, and then provides an eye-witness account to an-
other person (Ferguson 1993, 2002), while Togher has developed a more interac-
tive task in which the target speaker and partner are left alone to work out what
might be the use of an unusual object and their conversation audio-recorded for
later analysis (Kilov et al. 2009; Power et al. 2004). Researchers working solely
within the conversation analysis paradigm have consistently used real-world con-
versation sampling, although still often in a clinical setting, for example, in the
work of Klippi and Laakso (Klippi 1996; Laakso 2003; Laakso and Klippi 1999). As
Schegloff (2003) points out, any targeted assessment creates a ‘frame’ which draws
our focus towards events within that frame and tends to reduce our focus on what
is happening outside that specific context. Certainly, elicitations such as just de-
scribed allow the interaction analysis a wider view than that obtainable through
formal testing frames, but any elicited observational opportunities need also to be
supplemented with natural observations of the speaker in their everyday interac-
tions to enhance the validity of the assessment process.
The use of interaction analyses allow us to move from a notion of ‘error’
(as resulting from some problem resting in the individual, which we as outside
observers identify) to the notion of ‘trouble’ (as arising in the interaction rather
than in either individual, and as identified by the interactants themselves). This is
a fundamental paradigm shift for the focus of therapy for communication diffi-
culty, as it challenges notions of effective and successful communication, and how
we might characterise linguistic competence. For example, Anward (2003) sug-
gests that the act of repairing provides an opportunity for the person with aphasia
to display social competence. This shift has profound implications for the decision
as to whether or not to provide therapy (i.e., who is deciding whether there is a
problem that requires therapy?), who should be involved in the therapy (i.e., one
or both or more communication partners?), where should therapy be undertaken
(i.e., in clinics or in natural settings?), what should be the goals of therapy (i.e., are
goals defined with reference to the individuals or for the interaction?), and how
are outcomes of therapy to be measured (i.e., how will clinicians know when ther-
apy should finish?).
 Elizabeth Armstrong and Alison Ferguson

10. Toward the future

The work on interaction analysis that has been applied to-date in the field of com-
munication disability in general, and in the study of aphasia particularly, has had a
strong clinical focus. Contemporary studies relate to how useful interaction analy-
sis will be to the mission of speech pathologists, i.e., to maximize communication
for people with communication disability. Thus, a range of approaches which might
be considered as falling within the umbrella of ‘interaction analysis’ have shaped
the research into the nature of communication disability and its consequences for
people’s everyday lives, as well as underpinning the development of clinical tools
for assessment and therapy, and in widening the scope of therapeutic practice.
Looking to the future, at a more theoretical level, it is questionable whether
contemporary studies integrate understandings at a broader social-cultural level
with interaction analyses, and there are limits to which contemporary studies at-
tempt to integrate understandings of brain-language relationships. It may be that
such theoretical boundaries are beyond the scope of approaches within interaction
analysis, but as speech-language pathologists, we find that the nature of our work
demands that we are ‘boundary-riders’ along these ‘borders’ of interaction analy-
sis. We are working with people faced with social stigma and socio-cultural no-
tions of disability in a world where language is highly valued, and in which the loss
of language results in disempowerment - yet our approach to the analysis of these
aspects of communication disability is, at best, guided by general understandings
only, and our interventions are often disconnected with our interaction analyses.
We are also working with people whose language difficulties arose from brain im-
pairment, and so we know a lot about which parts of the brain have been affected
for them, and we necessarily draw on the neurophysiological and psychological
understandings of brain function in order to understand and conceptualize the
‘how’ of their language processing, e.g.€why they might be slower to formulate re-
sponses, why perseveration may occur. These everyday demands prompt us to
view the future directions of interaction analysis as resting with the ability of dif-
ferent perspectives to model and explain linguistic processing in the context of
both the instance of interaction and the wider culture in which communication is
playing a part.


Anward, Jan. 2003. “Own words: On achieving normality through paraphasias.” In Conversation
and Brain Damage, Charles Goodwin.(ed). Oxford: Oxford University Press.
Interacting with difficulty 

Armstrong, Elizabeth. 1987. “Cohesive harmony in aphasic discourse and its significance in
listener perception of coherence.” In Clinical Aphasiology: Conference Proceedings, Robert
H. Brookshire.(ed). Minneapolis, MN: BRK Publishers.
Armstrong, Elizabeth. 1991. “The potential of cohesion analysis in the analysis and treatment of
aphasic discourse.” Clinical Linguistics & Phonetics 5(1):39–51.
Armstrong, Elizabeth. 1993. “Aphasia rehabilitation: A sociolinguistic perspective.” In Aphasia
Treatment: World Perspectives, Audrey J. Holland and Margaret M. Forbes.(eds). San Diego,
CA: Singular.
Armstrong, Elizabeth. 2002. “Variation in the discourse of non-brain-damaged speakers on a
clinical task.” Aphasiology 16(4/5/6):647–658.
Armstrong, Elizabeth, Ferguson, Alison, Mortensen, Lynne and Togher, Leanne. 2005/2007.
“Acquired language disorders perspective.” In Continuing Discourse on Language: A Func-
tional Perspective, Jonathan Webster, Ruqaiya Hasan and Christian Matthiessen.(eds). Lon-
don: Equinox.
Armstrong, Elizabeth and Mortensen, Lynne. 2006. “Everyday talk: Its role in assessment and
treatment for individuals with aphasia.” Brain Impairment 7(3):175–189.
Beeke, Suzanne, Maxim, Jane and Wilkinson, Ray. 2007a. “Using conversation analysis to assess
and treat people with aphasia.” Seminars in Speech & Language 28(2):136–147.
Beeke, Suzanne, Wilkinson, Ray and Maxim, Jane. 2003a. “Exploring aphasic grammar 1: A
single case analysis of conversation.” Clinical Linguistics & Phonetics 17(2):81–107.
Beeke, Suzanne, Wilkinson, Ray and Maxim, Jane. 2003b. “Exploring aphasic grammar 2: Do
language testing and conversation tell a similar story?” Clinical Linguistics & Phonetics
Beeke, Suzanne, Wilkinson, Ray and Maxim, Jane. 2007b. “Grammar without sentence structure:
A conversation analytic investigation of agrammatism.” Aphasiology 21(3/4):256–282.
Boles, Larry. 1998. “Conversational discourse analysis as a method for evaluating progress in
aphasia: A case report.” Journal of Communication Disorders 31:261–274.
Boles, Larry and Lewis, Mimi. 2003. “Working with couples: Solution focused aphasia therapy.”
Asia Pacific Journal of Speech, Language and Hearing 8:153–159.
Booth, Susan and Swabey, Donna. 1999. “Group training in communication skills for carers of
adults with aphasia.” International Journal of Language & Communication Disorders
Brazil, David C. 1981. “Intonation.” In Studies in Discourse Analysis, R.Malcom Coulthard and
Martin Montgomery.(eds): Routledge and Kegan Paul.
Bremer, Katharina, Roberts, Celia, Vasseur, Marie-Therese, Simonot, Margaret and Broeder, Peter.
1996. Achieving Understanding: Discourse in Intercultural Encounters. London: Longman.
Byng, Sally. 1993. “Hypothesis testing and aphasia therapy.” In Aphasia Treatment: World Per-
spectives, Audrey Holland and Margaret M. Forbes.(eds). San Diego, CA: Singular.
Carlomagno, Sergio, Blasi, Valeria, Labruna, Ludovica and Santoro, Anna. 2000. “The role of
communication models in assessment and therapy of language disorders in aphasic adults.”
Neuropsychological Rehabilitation 10(3):337–363.
Cmerjrkova, Svetla and Prevignano, Carlo L. 2003. “On conversation analysis: An interview
with Emanuel A. Schegloff.” In Discussing Conversation Analysis: The Work of Emanuel A.
Schegloff, Carlo L. Prevignano and Paul J. Thibault.(eds). Amsterdam: John Benjamins.
Croteau, Claire, Vychytil, Anne-Marie, Larfeuil, Celine and Le Dorze, Guylaine. 2004. “’Speak-
ing for’ behaviours in spouses of people with aphasia: A descriptive study of six couples in
an interview situation.” Aphasiology 18:291–312.
 Elizabeth Armstrong and Alison Ferguson

Cutler, Anne. 1982. Slips of the Tongue and Language Production. New York: Mouton.
Eggins, Suzanne and Slade, Diana. 1997/2004. Analysing Casual Conversation. London: Cas-
sell(1997)/Equinox Publishing(2004).
Ferguson, Alison. 1992a. “Conversational repair of word-finding difficulty.” In Clinical Aphasiol-
ogy, Margaret L. Lemme.(ed). Austin, TX: Pro-Ed.
Ferguson, Alison. 1992b. “Interpersonal aspects of aphasic communication.” Journal of Neuro-
linguistics 7(4):277–294.
Ferguson, Alison. 1993. “Conversational repair in aphasic and normal interaction.” In Depart-
ment of Linguistics. Sydney: Unpublished PhD thesis, Macquarie University.
Ferguson, Alison. 1994. “The influence of aphasia, familiarity and activity on conversational
repair.” Aphasiology 8(2):143–157.
Ferguson, Alison. 1998. “Conversational turn-taking and repair in fluent aphasia.” Aphasiology
Ferguson, Alison. 2002. “(Abstract) Information exchange: Describing dimensions which might
respond to conversation therapy. Paper presented at 10th International Aphasia Rehabilita-
tion Conference, Brisbane 24–26€July, 2002.” Brain Impairment 3(2):174.
Ferguson, Alison. 2007. “Multiparty interactions in aphasia.” In Clinical Aphasiology: Future Di-
rections, Martin J. Ball and Jack S. Damico.(eds). Abingdon, Oxford: Psychology Press.
Ferguson, Alison and Armstrong, Elizabeth. 2004. “Reflections on speech-language therapists’
talk: Implications for clinical practice and education.” International Journal of Language &
Communication Disorders 39(4):469–477.
Ferguson, Alison and Elliot, Ngiare. 2001. “Analysing aphasia treatment sessions.” Clinical Lin-
guistics & Phonetics 15(3):229–243.
Ferguson, Alison and Peterson, Peter. 2002. “Intonation in partner accommodation for aphasia:
A descriptive single case study.” Journal of Communication Disorders 35:11–30.
Firth, John R. 1957. “Personality and language in society.” In Papers in Linguistics 1934–1951:
Oxford University Press.
Fox, Sarah, Armstrong, Elizabeth and Boles, Larry. 2009. “Conversational treatment in mild
aphasia: A case study.” Aphasiology 23(7–8):951–964.
Giles, Howard, Coupland, Nikolas, Coupland, Justine, Williams, Angie and Nussbaum, Jon.
1992. “Intergenerational talk and communication with older people.” International Journal
of Aging & Human Development 34(4):271–297.
Giles, Howard and Smith, Philip.€ 1979. “Accommodation theory: Optimal levels of conver-
gence.” In Language and Social Psychology (pp.45–65), Howard Giles and Robert N. St Clair.
(eds). Oxford: Basil Blackwell.
Goodglass, Harold, Kaplan, Edith and Barresi, Barbara. 2000. Boston Diagnostic Aphasia Ex-
amination (BDAE-3). 3rd Edition: ProEd.
Goodwin, Charles. 1981. Conversational Organization: Interaction between Speakers and Hear-
ers. New York: Academic Press.
Goodwin, Charles. 1995. “Co-constructing meaning in conversations with an aphasic man.”
Research on Language and Social Interactions 28(3):233–260.
Goodwin, Charles. 2003a. “Conversational frameworks for the accomplishment of meaning in
aphasia.” In Conversation and Brain Damage, Charles Goodwin.(ed). Oxford: Oxford
University Press.
Goodwin, Charles. 2003b. “Introduction.” In Conversation and Brain Damage, Charles Goodwin.
(ed). Oxford: Oxford University Press.
Interacting with difficulty 

Gott, Stephen J. and Plaut, David C. 2002. “The impact of synaptic depression following brain
damage: A connectionist account of “access/refractory” and “degraded-store” semantic im-
pairments.” Cognitive, Affective & Behavioral Neuroscience 2(3):187–213.
Grice, Herbert P. 1975. “Logic and conversation.” In Syntax and Semantics: Vol.3, Speech Acts,
Peter Cole and Jerry L. Morgan.(eds). New York: Academic Press.
Gumperz, John J. 1982. Discourse Strategies. New York: Cambridge University Press.
Halliday, Michael A.K. 1994. An Introduction to Functional Grammar. 2nd Edition. London: Arnold.
Halliday, Michael A.K. and Hasan, Ruqaiya. 1976. Cohesion in English. London: Longman.
Halliday, Michael A.K. and Hasan, Ruqaiya eds. 1985. Language, Context, and Text: Aspects of
Language in a Social-Semiotic Perspective. Geelong, VIC: Deakin University.
Halliday, Michael A.K. and Matthiessen, Christian. 2004. An Introduction to Functional Gram-
mar. 3rd Edition. New York: Arnold.
Harley, Trevor A. 1996. “Connectionist modeling of the recovery of language functions follow-
ing brain damage.” Brain and Language 52(1):7–24.
Hasan, Ruqaiya, Cloran, Carmel, Williams, Geoff and Lukin, Annabelle. 2005/2007. “Semantic
networks: The development of clausal and textual perspectives on theme.” In Continuing
Discourse on Language: A Functional Perspective, Ruqaiya Hasan, Christian Matthiessen
and Jonathan Webster.(eds). London: Equinox.
Heeschen, Claus and Schegloff, Emanuel A. 1999. “Agrammatism, adaptation theory, conversa-
tion analysis: On the role of so-called telegraphic style in talk-in-interaction.” Aphasiology
Hillis, Argye E. 2002. “Preface to special issue: New techniques for identifying the neural sub-
strates of language and language impairments.” Aphasiology 16(9):855–858.
Hopper, Tammy, Holland, Audrey and Rewega, Molly. 2002. “Conversational coaching: Treat-
ment outcomes and future directions.” Aphasiology 16(7):745–761.
Kemper, Susan, Ferrell, Patrice, Harden, Tamara, Finter-Urczyk, Andrea and Billington, Cather-
ine. 1998. “Use of elderspeak by young and older adults to impaired and unimpaired listen-
ers.” Aging, Neuropsychology, and Cognition 5(1):43–55.
Kilov, Andrea, Togher, Leanne and Grant, Susan. 2009. “Problem solving with friends: Dis-
course participation and performance of individuals with and without traumatic brain in-
jury.” Aphasiology 23(5):584–605.
Klippi, Anu. 1996. Conversation as an Achievement in Aphasics (Studia Fennica, 6). Helsinki:
Suomalaisen Kirjallisuuden Seura.
Korpijaakko-Huuhka, Anna M. 2004. “Aphasic speakers’ linguistic choices in a cartoon-story task.”
University of Helsinki, Finland (Dissertation Abstracts International, ISSN 1042–7279).
Laakso, Minna. 2003. “Collaborative construction of repair in aphasic conversation: An interac-
tive view on the extended speaking turns of persons with Wernicke’s Aphasia.” In Conversa-
tion and Brain Damage, Charles Goodwin.(ed). Oxford: Oxford University Press.
Laakso, Minna and Klippi, Anu. 1999. “A closer look at the ‘hint and guess’ sequences in aphasic
conversation.” Aphasiology 13:345–363.
Lesser, Ruth. 2003. “When conversation is not normal: The role of conversation analysis in lan-
guage pathology.” In Discussing Conversation Analysis, Carlo L. Prevignano and Paul J.
Thibault.(eds). Amsterdam: John Benjamins.
Lindsay, Jane and Wilkinson, Ray. 1999. “Repair sequences in aphasic talk: A comparison of
aphasic-speech and language therapist and aphasic-spouse conversations.” Aphasiology
 Elizabeth Armstrong and Alison Ferguson

Locke, Sandra, Wilkinson, Ray, Bryan, Karen, Maxim, Jane, Edmundson, Anne, Bruce, Caroline
and Moir, Diana. 2001a. “Supporting Partners of People with Aphasia in Relationships and
Conversation (SPPARC).” International Journal of Language & Communication Disorders
Locke, Sandra, Wilkinson, Ray and Bryant, Karen. 2001b. Supporting Partners of People with
Aphasia in Relationships and Conversation. Oxon, UK: Speechmark.
Mackenzie, Catherine. 2000. “Adult spoken discourse: The influences of age and education.” In-
ternational Journal of Language & Communication Disorders 35:269–285.
Martin, James R. 1992. English text: System and Structure. Amsterdam: John Benjamins.
Martin, James R. 2002. “Meaning beyond the clause: SFL perspectives.” Annual Review of Ap-
plied Linguistics 22:52–74.
Martin, James R. and Rose, David. 2003. Working with Discourse: Meaning beyond the Clause.
London: Continuum.
Mortensen, Lynne. 2003. “Reconstructing the writer: Acquired brain impairment and letters of
community membership.” Sydney: Macquarie University.
Murray, Laura and Chapey, Roberta. 2001. “Assessment of language disorders in adults.” In Lan-
guage Intervention Strategies in Aphasia and Related Neurogenic Communication Disorders,
Roberta Chapey.(ed). Baltimore, MD: Lippincott, Williams & Wilkins.
Myers, Penelope S. 1999. Right Hemisphere Damage: Disorders of Communication and Cognition.
San Diego, CA: Singular.
Perkins, Lisa. 1995. “Applying conversation analysis to aphasia: Clinical implications and ana-
lytic issues.” European Journal of Disorders of Communication 30:372–383.
Perkins, Lisa. 2003. “Negotiating repair in aphasic conversation: Interactional issues.” In Conver-
sation and Brain Damage, Charles Goodwin.(ed). Oxford: Oxford University Press.
Plaut, David C. 1996. “Relearning after damage in connectionist networks: Toward a theory of
rehabilitation.” Brain and Language 52(1):25–82.
Power, Emma, Anderson, Alison, Togher, Leanne and Werry, Kate. 2004. “Chatting with their
peers: Analysis of discourse in people with Huntington Disease in a residential care facility.”
In 26th World Congress of the International Association of Logopedics & Phoniatrics.
Brisbane, 29€August - 2€September.
Prutting, Carol A. and Kirchner, Diane M. 1983. “Applied pragmatics.” In Pragmatic Assessment
and Intervention Issues in Language (ch.2), Tanya M. Gallagher and Carol A. Prutting.(eds).
San Diego, CA: College-Hill.
Ramsberger, Gail and Menn, Lisa. 2003. “Co-constructing Lucy: Adding a social perspective to
the assessment of communicative success in aphasia.” In Conversation and Brain Damage,
Charles Goodwin.(ed). Oxford: Oxford University Press.
Schegloff, Emanuel A. 2003. “Conversation analysis and communication disorders.” In Conver-
sation and Brain Damage, Charles Goodwin.(ed). Oxford: Oxford University Press.
Schegloff, Emanuel A., Jefferson, Gail and Sacks, Harvey. 1977. “The preference for self-correc-
tion in the organization of repair in conversation.” Language 53(361–382).
Searle, John R. 1969. Speech Acts. London: Cambridge University Press.
Seidenberg, Mark S. 1988. “Cognitive neuropsychology and language: The state of the art.” Cog-
nitive Neuropsychology 5(4):403–426.
Simmons-Mackie, Nina and Damico, Jack S. 1999. “Social role negotiation in aphasia therapy:
Competence, incompetence, and conflict.” In Constructing (In)competence: Disabling Eval-
uations in Clinical and Social Interaction (ch.14, pp.313–342), Dana Kovarsky, Judith F.
Duchan and Madeline Maxwell.(eds). Mahwah, NJ: Lawrence Erlbaum.
Interacting with difficulty 

Springer, Luise, Glindemann, Ralf, Huber, Walter and Willmes, Klaus. 1991. “How efficacious is
PACE therapy when ‘language systematic training’ is incorporated?” Aphasiology 5:391–399.
Togher, Leanne. 2000. “Giving information: The importance of context on communicative op-
portunity for people with traumatic brain injury.” Aphasiology 14:365–390.
Togher, Leanne and Hand, Linda. 1998. “Use of politeness markers with different communica-
tion partners: An investigation of five subjects with traumatic brain injury.” Aphasiology
Togher, Leanne, Hand, Linda and Code, Chris. 1997. “Measuring service encounters in the trau-
matic brain injury population.” Aphasiology 11:491–504.
Turner, Sonja and Whitworth, Anne. 2006. “Clinicians’ perceptions of candidacy for conversa-
tion partner training in aphasia: How do we select candidates for therapy and do we get it
right?” Aphasiology 20(7):616–643.
Ulatowska, Hanna K., North, Alvin J. and Macaluso-Haynes, Sara. 1981. “Production of narra-
tive and procedural discourse in aphasia.” Brain and Language 13:345–371.
Ulatowska, Hanna K. and Olness, Gloria. 2000. “Discourse revisited: Contributions of lexico-
syntactic devices.” Brain and Language 71(1):249–251.
Van Leer, Eva and Turkstra, Lyn. 1999. “The effect of elicitation task on discourse coherence and
cohesion in adolescents with brain injury.” Journal of Communication Disorders 32(5):327–348.
Wertz, Robert T., Collins, Michael J., Weiss, David, Kurtzke, John F., Friden, Thomas, Brook-
shire, Robert H., Pierce, James, Holtzappled, Pat, Hubbard, D.J., Porch, Bruce E., West,
Joyce A., Davis, Larry, Matovitch, Violet, Morley, Gerald K. and Resurreccion, Ernesto.
1981. “Veterans Administration cooperative study on aphasia: A comparison of individual
and group treatment.” Journal of Speech and Hearing Research 24:580–594.
Whitworth, Anne, Perkins, Lisa and Lesser, Ruth. 1997. Conversation Analysis Profile for People
with Aphasia (CAPPA). London: Whurr.
Wilkinson, Ray, Beeke, Suzanne and Maxim, Jane. 2003. “Adapting to conversation.” In Conver-
sation and Brain Damage, Charles Goodwin.(ed). Oxford: Oxford University Press.
Wilkinson, Ray, Bryant, Karen, Lock, Sandra, Bayley, Kate, Maxim, Jane, Bruce, Caroline,
Edmundson, Anne and Moir, Diana. 1998. “Therapy using conversation analysis: Helping
couples adapt to aphasia in conversation.” International Journal of Language & Communica-
tion Disorders 33(Suppl):144–149.
Youse, Kathleen, Stout, Connie and Bosworth, Karyn E. 2001. “Cohesion in discourse of college-
educated and non-college-educated adults: Implications for young adults with brain injury.”
Advances in Speech Language Pathology 3(2):97–108.
Ecologies of gesture

Jürgen Streeck
The University of Texas at Austin

This chapter presents a heuristic of ways in which hand-gestures participate

in communicative interaction. Rather than conceiving gesture as part of
language, an ecological perspective is proposed, and gestures are examined both
in relation to sequences of social actions and to the ecological context. This
broader account of gesture requires a different conception of the human body
than is commonly presupposed by researchers of “nonverbal communication”: a
model of the body not as an instrument of expression, but as a skilled (knowing)
inhabitant of worlds.

1. Context

In 1992, at a conference on “the contextualization of language” organized by Auer

and di Luzio (Auer & di Luzio 1992), hand gestures were analyzed by several pre-
senters as contextualization cues, that is, as “surface features of message form … by
which speakers signal and listeners interpret what the activity is [and] how seman-
tic content is to be understood” (Gumperz 1982:€57). The theory of contextualiza-
tion cues suggests that these cues signal contextual frames in terms of which social
meanings of utterances are understood, for example the perceived social relation-
ship between speaker and listener. Gestures serve as contextualization cues when
they indicate aspects of illocutionary force, discourse structure, or turn-taking or-
ganization; we can refer to this type of gesticulation as “pragmatic”.
Gumperz’ notion of contextualization is grounded in the tradition of context
analysis, a term first proposed by Gregory Bateson (1971). Bateson had noticed
that there are communicative behaviors which do not convey content, but rather
instruct recipients how a bit of behavior is to be taken, for example as “play”. Such
meta-messages, which are typically conveyed by small behavioral forms – a raised
eye-brow, a certain incline of the head – convey contextual frames of moments of
interaction. A familiar example of body behavior that functions in this context-
defining capacity are shrugs, i.e., motions of the eyebrows, hands, and/or shoulders,
 Jürgen Streeck

which have in common the formal feature of raising or retraction and which, as a
family, convey a stance of detachment or indifference towards a proposition or a
reported event. Hand gestures can frame the current or next verbal act by the
speaker or specify the kind of uptake that he or she seeks for the utterance cur-
rently underway. Heath, in his contribution to The Contextualization of Language
showed that hand-gestures, in their moment-by-moment coordination with
speech, form part of “the famework to which subsequent action will be addressed
in the turn by turn organization of talk” (Heath 1992). Streeck and Hartge showed
that certain conventional hand gestures, when they are made in the forefield or at
the beginning of turns at talk, “preview” the action that the speaker is planning to
take in that turn (Streeck & Hartge 1992).
While this is a mode of gesturing which closely corresponds to Bateson’s notion of
contextual frames – bits of behavior are marked (framed) as being of a certain mode
or type–, Bateson also proposed an ecological and systemic understanding of contexts
of interaction, for example when he suggested that no unit of behavior ever stands
apart from the context by which it is specified, but rather is “part of the ecological
subsystem called context”; it is not “the product or effect of what remains of the con-
text once the piece which we want to explain has been cut out from it” (Bateson 1972:€338),
but rather participates in establishing or sustaining the multimodal framework of be-
havioral interrelationships from which it accrues its locally specific meaning. Differ-
ent modes of gesticulation (e.g.€depicting, pointing, displaying an illocution) differen-
tially direct and focus the attention of the participants and in this fashion also
contribute to the structuring of the interaction. This will be clarified below.
While Bateson only made reference to G.H.Mead’s analysis of interaction in
an opaque remark in his contribution to The Natural History of an Interview – “our
concept of communication becomes interactional”, he notes, “and our intellectual
debt is to G.H. Mead” (Bateson 1971:€20) – Mead’s conception of human symbols
and mind as originating in a generic conversation of gestures (Mead 1909, 1934) is
an important foundation for the empirical study of gesture. In Mead’s account of
the interactional phylogenesis of human cognitive and symbolic functions ges-
tures play the most significant part. Mead regarded gestures as early parts of acts,
an idea which he took from Darwin (Darwin 1955 1872). But in contrast to Dar-
win, who thought about gestures as affect expression, Mead suggested that they
facilitate the mutual anticipation of social acts: a gesture is a foreshadowing, not an
expression. Gestures enable individuals to navigate the rapids of potentially harm-
ful moment-by-moment interaction by allowing them to adapt to what is coming
before the fact (cf.€Streeck 2009b).
We must be aware that the foreshadowing of imminent actions is not the only
purpose for which gestures are made and that Mead’s account would not cover and
explain all conversational gestures. Nevertheless, Mead’s model of the conversation
Ecologies of gesture 

of gestures gives us a starting point for analyzing gestures – namely in terms of how
they contextualize the next moment (see McDermott & Roth 1978) – that is not
very different from a conversation-analytic approach (e.g.€Goodwin 1986; Good-
win & Goodwin 1986; Heath 1986; Schegloff 1984). The analogy between Mead’s
account of interaction and conversation analysis and his implicit influence on CA
are rarely acknowledged. Evidently, Mead was a thinker of the first half of the 20th
century, not a micro-analyst with a microphone and video-camera. But what Mead
and Schegloff have in common is an understanding of interaction as sequentially,
turn-by-turn-, incrementally designed social action. Conversation analysts thus
have analyzed gestures by reference to where in the turn at talk and action-se-
quence gestures are being and what “jobs” they do and what projections they make
in these positions. Schegloff (1984), observing the frequent pre-positioning of hand
gestures relative to their “lexical affiliates”, analyzed them for how they bring an
element of talk into play and thus mark the opening up of this element’s “projec-
tion space”. While the foreshadowing of a “lexical affiliate” is a function typically
enacted by descriptive (referential) gestures, other modes of “forward-gesturing”
(Streeck 2009b) can be productively analyzed as components of action formation.
Schegloff (2007) has described conversational organization as a set of specific or-
ganizations each of which addresses a fundamental problem that conversational-
ists need to cope with at each and every point in their interaction. Among these is
the “action-formation problem”:
how are the resources of the language, the body, the environment of the inter-
action, and position in the interaction fashioned into … particular actions, …
recognizable by recipients …, – actions like requesting, granting, complaining?
(Schegloff 2007: xiv)

One can distinguish several relevant positions at which gestures are made during the
unit-by-unit unfolding of turns and sequences of action. For example, a first relevant
position is prior to turn beginning, near the possible completion of someone else’s
turn.1 Intending next speakers can rely on body motion, including gestures, to attract
attention to themselves and signal their intent to take the next turn (Kendon 1970).
Once the attention of others is attracted, the gesture can also display what kind of
talk the turn will be used for: a story (Streeck & Hartge 1992), a question, a confes-
sion (Streeck 2009b), etc. Other common displays in the forefield or at the beginning
of turns show aspects of the beginning turn’s design (e.g.€that it will be a multi-unit
turn). Or the gesture displays the speaker’s stance in relation to what is about to be
said (as in the example of shrugs). A common type of turn-final gestures are acts of

1. Concrete examples and analysis of the phenomena that are glossed over here are found in
Streeck 2007, 2009 a,b, and Streeck & Hartge, 1992.
 Jürgen Streeck

“handing over” which model talk as a transaction involving physical objects (see
Müller 2003; Cienki & Müller 2009). “Giving” or “offering” hands can gradually turn
into receptacles ready for something to be placed in them. In other words, they can
serve to solicit response (Streeck 2007). Generally, gestures of the hand are well-
adapted to dealing with turn-taking matters because they can be performed and
understood concurrently with speech, free from the pitfalls of overlapping talk.

2. Ecologies of gesture: Body and environment

One ecology to which gestures of the hand contribute and owe their intelligibility,
then, is the production (or “formation”) of sequences and turns of (conversational)
social action. Hand gestures can be part of the design of these turns and actions
and, by virtue of their positioning within them, provide recipients with a “forward-
understanding”, i.e. an anticipation, of what will come next. But this is only one
relevant dimension in terms of which gestures of the hands are made and under-
stood. Another was pointed out by Goffman, in a rarely cited statement:
While the substratum of a gesture derives from the maker’s body, the form of the ges-
ture can be intimately determined by the microecological orbit in which the speaker
finds himself. To describe the gesture, let alone uncover its meaning, we might then
have to introduce the human and material setting in which the gesture is made. … The
individual gestures with the immediate environment, not only with his body, and so
we must introduce this environment in some systematic way (Goffman 1964:€164).

Symbolic acts of the hands, in other words, are also made within real, physical settings,
and they can couple with specific aspects of the scene (Goodwin 2007) and illuminate
it. I have found a heuristic of six ecologically different modes of gesturing useful:
1. gestures physically linked to the (tangible) environment at hand;
2. gestures elaborating the (visible) world in sight;
3. gestures that depict actual, imaginary, and abstract worlds;
4. gestures that construe ideational content (gestural concepts);
5. gestures that embody and construe communicative acts by the gesturer
(e.g.€concurrent acts of speech); and
6. gestures by which transactions are managed, including those that regulate the
behavior of co-interactants.
Other gesture ecologies could presumably be identified, but in the meantime this
heuristic enables us to take note of the fact that hand gestures not only embody
meaning and mediate communication in heterogeneous ways, but also bring the
communicating body in contact with the world in a variety of distinct modes.
Ecologies of gesture 

What is at issue in part is how visual and tactile (or haptic) features of gestures
intersect. Consider gesture practices of the first variety. These involve direct tactile
contact – or at least near-contact – between the gesturing hand and the immediate
physical setting, the world at hand. Gestures of this kind, which involve immediate
physical contact, are abundant in many work settings, including those where peo-
ple cooperate by means of inscriptions on paper. One can think of two car-me-
chanics exploring with their fingers a dent in a fender to determine how to remove
it: exploratory motions become gestures, which can display information, such as
the texture of the surface. People can virtually share tactile experience by gestural-
izing the motions through which this experience is gathered. Car-mechanics (and
practitioners of many other professions) also disassemble and mock-disassemble
complex objects (e.g.€car-parts) and make the connectedness and separability of
subparts visible by motions of their hands. Or think of a group architects leaning
over a blueprint, one of them emphasizing by a motion of the hand the curvature
of a (planned) wall. All of these practices engage the interactants’ hands in their
actual, material involvement with the material world. In each case, the world at
hand is elaborated and made intelligible with the help of gestures.
A concrete example of this first variety comes from the interaction between a
car-mechanic and his apprentice. The apprentice explains the function of a bolt in
the trunk of a car to his boss. He points to tracks on the two sides of the trunk and
moves his hands back and forth, thus showing how the bolt moves while holding
the trunk-cover in place as it is being opened or closed. Notice how truncated the
apprentice’s verbal explanation is; without the gesture, it would be incomprehensi-
ble. (Black dots in the transcripts indicate the moments captured in the drawings.)
(1) 1 Mechanic What this bolt here for, need to go inside?

Figure 1
2 Apprentice Oh no, there’s- an ( - - - - • - - - - - - - )
3 Mechanic Something to hang from this?
4 Apprentice For the cover right here.
5 Mechanic Okay. ( - - ) You just inspect everything.
 Jürgen Streeck

In this example, an object at hand is elaborated by the gesture: what cannot be

gleaned from the immobile object – how its parts behave and connect when in
motion – is being added by the motion of a properly configured hand. The conti-
guity between object and hand secures the reference of the gesture.
Such reference can be more tenuous in the second variety, gestures that elabo-
rate the world in sight. The senses are implicated in different ways when people rely
on motions of the hands to make sense of the currently visible world, the part of
the immediate environment that is available to their eyes, but not their hands. In
this case the hands are used to visually augment visual experience, to provide ad-
ditional structure, layers of meaning that annotate what can presently be seen.
In the following extract, the mechanic, interacting with one of his senior em-
ployees, is instructing him to switch two tires of a car which is located in another,
distant section of the shop.€He uses pointing gestures in the process. After sum-
moning the employee with an index-up gesture (Fig. 2.1), he poins with his index-
finger (Fig. 2.2), which the recipient acknowledges with a pointing gesture of his
own (Fig. 2.3). The mechanic, taking this response as insufficient evidence that the
employee has correctly identified the tire that is to be replaced, then makes a sec-
ond pointing gesture (Fig. 2.4), carefully lowering his head so that his line of vision
can be inferred, and adds “this side” (line 5).

Figure 2.1
1 Mechanic (summons Cedric with an ‘index up’ gesture)

Figure 2.2
Ecologies of gesture 

2 Mechanic The • front right side tire

Figure 2.3
3 the fro:nt •

Figure 2.4
4 Cedric ( )
5 Mechanic • this side
6 Cedric Okay
7 Mechanic you need to switch with that tire
8 Cedric Okay.
Gestures of the hands that focus attention on an object in the tangible or visible
environment often also provide an aspect under which an object is to be per-
ceived.2 Differences in hand-shape, for example, can indicate whether an object is
to be seen as an individual, as an examplar, as a set or collection, and so on. Add-
ons to pointing gestures can indicate that there is something beyond the horizon,
and tracing motions can select and emphasize boundaries in the landscape.
Through gestures of both the first and the second ecological variety then, the local
setting of the interaction – the place that is currently inhabited by the interactants
– is structured and its meaning for the ongoing activity clarified. These gestures, in
turn, derive their visible significance from their coupling with the context that is
available to the senses.

2. See Kendon, 2004:€167; Streeck, 2009a: Ch.4.

 Jürgen Streeck

Other modes of gestural communication are predicated upon a turn away

from the environment of the interaction, a shift of attention from the world to the
gesturing hands themselves, which are not oriented outward, but rather do their
work in the center of the interactional huddle. These are the gestures that produce
mock-up worlds, representations of whatever actual or virtual (imaginary or ab-
stract) scenes that the talk is about. Depiction is a distinct gestural practice, tight-
ly organized and firmly supported by linguistic units, for example demonstratives
and deictic adverbs such as like this (Streeck 2002), and visually attended by both
speaker and recipient (Streeck 1993; Gullberg & Kita 2009). Rather than the ges-
ture directing attention to what can be seen, other deictics, including gaze direc-
tion, indicate that what needs to be seen is the gesture.
In the following extract gestures of the hand enable the recipient to imagine an
absent world. They depict what the spoken utterance describes. Here, an architect
(MJ) talks to a student researcher (C) about a potential building site that he has
just visited. While his talk gives a vivid description of the scene, it is the gestural
structuring of the space in front of him that ultimately enables the interlocutor to
build an adequate representation of the site that he describes. (Only a selection of
the gesture movements that the architect makes are included in this transcript.)
(3) 1 C So y’all went to a new space today.
2 MJ (- - - - - ) We went to a- well.
3 It’s gonna be (. ) our (. ) latest (. ) job hopefully.
4 C This is the Newman house is that the one?
5 MJ ( - - ) Ah-
� uh- Nieman.
6 Nieman. Rick Nieman.
7 C Nieman?
8 MJ Yeah. So (. ) we had to go check o:n

Figure 3.1
9 how the • power-cabels come across because
Ecologies of gesture 

10 thi- the initial idea I had for the- ( - - )

Figure 3.2
11 the house is to actually do it in three terrace • levels.=
12 C = Mh hm.
13 MJ Because the side almost has this three (. ) ter-

Figure 3.3
14 It has two very disti:nct ( - • - )
15 terrace levels and then
16 we’ll probably try’n pull one out of the very to:p.
17 Absolutely fanta:stic.
(11 lines deleted))
29 And on this bottom one ( - - )

Figure 3.4
 Jürgen Streeck

30 there’s this natural • grotto ( - - - - - - - ) right down here.

Figure 3.5
31 And it’s this very tiny, • intimate-
32 C Hmm.
33 MJ piece, with the with the water runoff coming down and
34 dripping.
The drawings show that, in contrast to the previous interactions, the gesturer here
focuses his gaze on his own hands. The architect’s gaze upon his hands is in fact
quite sustained, owing to the centrality of his hand motions to the complex spatial
representation that he is building. An intermittent glance by the gesturer to the
gesturing hands, however, is generally characteristic of hand-gestures made in the
depictive mode (Streeck 1993). The focus of attention of the entire interacting en-
semble is upon the gestures, which represent some world beyond the setting of the
interaction, rather than highlighting and elaborating features of the current scene.
This is what, quite simply, is meant by ecology here: a gesture’s relations to the
participants’ attentional field, that is, the field which they perceive, construe, or
imagine. Depictive gestures structure the participants’ imagination, pointing and
other visually indexical gestures the setting that they can see, and tactile and haptic
gestures the setting that they can touch, handle, and, more often than not, see as
well. In other words, the three types of gestures evoke figures on different kinds of
ground and they focus cognitive attention in different ways: they either augment
and organize what is available to the interactants’ senses or they give structure to
their imagination.
Depiction is not the only purpose for which pictorial gestures are made. Only
depictions, by definition, show what something looks like, and it may be for this
reason that speakers who provide depictions with their hands, look at them at
some point. The gaze to the hand is a display of the gesture’s depictiveness.
Equally pictorial (if not always in a narrowly visual sense) is the conceptual
mode of gesticulation, that is, performances of embodied schemata that structure
content and represent something�as something (Goodman 1968). To emphasize its
Ecologies of gesture 

nature as a distinctly manual mode of cognitive grasping, I label this mode of ges-
turing ceiving, from Latin cap, “take”, “take hold of ” (as in con-ceive and con-cept).
Gestures in both the depictive and conceptual modes are “iconic” or “imagistic”
(Beattie 2004). But whereas depictive gestures are made deliberately and, at least
briefly, looked at by the gesturer, conceptual gestures are more like spontaneous
bodily insights, creative manual figurations of content that is in need of a form.
This is most visible when the speaker makes a concrete gesture to construe idea-
tional content for which s/he is simultaneously searching a lexical representation.
Gestures produced in this mode are not attended (i.e. looked at) by their makers.
Depiction and conceptualization by gesture, in other words, are two related modes
which are distinguished by the different frameworks of visual attention within
which they operate: depictive gestures are made in or near the center of visual at-
tention, conceptual gestures emerge from the periphery of the attentional field:
they are spontaneous “body thoughts” (Strathern 1996).
In example 4, the mechanic enacts the concept cranking. He is presently taking
issue with a customer’s account of engine trouble that she experienced; he points out
that her car did, in fact, “crank”. (In other words, both battery and starter were work-
ing.) As he utters the word, he performs an eye-brow flash, a facial display of recog-
nition, while rotating the index-finger of his right hand, which is elevated to his ear.
The enactment is a multi-modal performance of “hearing cranking” (i.e. achieving
a knowing recognition of a sound as being by the turning of the starter).
(4) 1 Mechanic When she came to start the car, the car flooded her.

Figure 4
2 She crank•in’. ( - - ) She misdescribed to me.
This is not a pantomimic depiction of someone’s behavior in the talked-about situ-
ation; it is not a depiction at all, but rather a synthetic embodiment of a concept: the
mechanic enacts it frequently along with the word crank, independently of whether
this word is used in descriptive, abstract, or proscriptive contexts. In other words,
the gesture is a concrete cept (manual concept) that corresponds to a verbal concept,
not to a physical entity or event, as depictive gestures do. Many gestures made in the
mode of ceiving depict what we traditionally call the vehicle of a metaphor (Cienki
 Jürgen Streeck

& Müller 2009). The speaker’s body supplies a sensorimotor schema that structures
some phenomenon or abstract domain and thereby renders it intelligible.
The remaining modes of gesturing do not relate to the world around the inter-
action or the world that is being talked about, nor do they construe the content of
the conversation in other ways. Rather, they relate to the interaction itself. They
have been called “pragmatic” (Kendon 2004) or “interactive” (Bavelas et al. 1992)
gestures. I distinguish between gestures that relate to the communicative actions
of the speaker/gesturer and those that relate to the current, anticipated, or desired
actions of the addressee. In one mode the gesturer displays (aspects of) what he or
she is doing in the (current or imminent) utterance, in the other s/he directs others
what to do. It is quite possible that this distinction is not tenable or needs to be
refined, so that, for example, gestures that formulate communicative relationships
or engagements can be accomodated. But the ecological difference between these
modes of gesturing and the ones discussed above should be clear enough.
An example of the first of the two types, gestures that display communicative
acts, is the following, from the same interaction as Example 4.
Here, the mechanic concludes his explanation of the customer’s engine trou-
ble, hinted at in Example 4, with a “handing over” or “palm-up/open hand” ges-
ture (cf.€Müller, 2003).
(5) 1 Hussein all the car smoke,
2 wasting gas,

Figure 5
3 no po•wer.
4 Jürgen Oh.
The gesture marks the consequence of the antecedent conditions that the speaker
lists in this sentence, and at the same time marks the end of the utterance and thus
the transfer of the turn to the interlocutor, who produces an understanding token,
oh. Pragmatic gestures appear to be both the most frequent variety of conversa-
tional gestures and the type that is most often conventionalized, and among prag-
matic gestures the open-handed “handing over” is a particularly common unit.
The pragmatic mode of gesturing is not perceptually different from ceiving,
i.e. gestural conceptualization. What differs is what is being conceptualized, and
Ecologies of gesture 

thus, how the gesture is processed3: pragmatic gestures do not frame what is being
talked about, but aspects of the process of talking.
Gestures of the other subvariety refer to actions of the addressee, for example
by soliciting a response or by proscribing a responsive action that the other could
take. An example is the following gesture with which its maker seeks to stop the
interlocutor from proceeding further with her talk: that forthcoming talk is figu-
ratively held at bay. Bev is talking about a movie which she recommends Rani go
and see; Rani seeks to keep her from revealing too much about the plot.
(6) 1 Bev It’s really good.

Figure 6
2 Rani Don’t • tell me.
This gesture is understood as relating to the interlocutor’s actions; this is its pri-
mary coupling.
Hand gestures thus mediate processes of sense-making in a number of quite
different fashions, and a multitude of heterogeneous practices are available, appar-
ently in any culture, to deal with gesture’s diverse tasks.

3. The body in communication

The diversity of ways in which motions and configurations of the hands participate
in communicative processes and their different couplings with components of the
communicative situation gives us pause to reconsider the way in which the body is
conceptualized in research on language and social interaction. Until recently, most
research and model-building concerning “nonverbal communication” was predi-
cated upon an expressive understanding of the human body: various body parts or
regions (the forehead, the face, the hands) or the body in its structured configura-
tion (posture) were treated as expressive of the communicator’s “internal” disposi-
tion, of mental imagery or intent. The body was thus attributed an ancillary role
(as maker of sign-bodies) in activities governed by semiotic relations, or subsumed

3. For more on this see Streeck, 2009a: Ch. 8.

 Jürgen Streeck

under abstract categories such as agent or subject. The body was treated as if it had
nothing of its own to contribute to sense-making. Even though gestures are con-
ceived as “material carriers of thinking-for-speaking” (McNeill & Duncan 2000),
the mindfulness of the human body – the fact that it knows the world on its own
terms – was rarely taken into account. But especially when we study gestures in the
context of activities that involve manual action, as in a car-repair shop, the fact
that they themselves are bodily acts becomes important: gestures can no longer be
reduced to expressive forms that are only incidentally made by the hands, but their
manual nature and their frequent origin in, and indexical ties to, practical actions
of the hands (as well as the tools and objects involved in them) are central to how
these gestures mean.
The empirical study of multimodal interaction meets up here with developing
lines of research and theorizing in fields of study such as cognitive science
(Wilson 2002), cognitive linguistics (Johnson 1987), and “communities of prac-
tice” or praxeology (Hanks 1996). In these fields, attempts have been made to re-
conceptualize the body as a cognitive entity and make sense of it in terms of how
it inhabits and acts in worlds (Streeck, Goodwin and LeBaron, to appear). Much
of the recent work on bodily components of human communication continues to
be based, if not explicitly then in terms of research methods, on a dualist concep-
tion of body and mind. Gestures are seen as the external dimension or “sign vehi-
cles” of mental content or process. In my own work, I take a rather different tack,
working from an understanding of the human body as the primordial organ and
site of human cognition, as proposed by phenomenological philosophers
(Heidegger 1962 (1926); Merleau-Ponty 1962; Polanyi 1958). These have argued
that we must understand human understanding by finding it, in the first place, in
concrete, practical, physical activity in the world. This congrues with recent work
in anthropology (Hastrup 1995; Ingold 2000; Keller & Keller 1996; Strathern
1996), philosophy and linguistics (Varela, Thompson, & Rosch 1991), educational
psychology (Lave 1988), and sociology (Connerton 1989; Harper 1987; Mauss
1973 (1935)), which works from the fact that the human mind – and the symbols
that it relies upon – are embodied. In cognitive science, there is a growing accept-
ance of the idea that certain higher-level cognitive abilities are rooted in lower-
level sensorymotor skills (Wilson 2002). Recent findings indicate that the con-
cepts underlying language are stored in the brain in terms of action-perception
loops (Barsalou 1999; Glenberg 2002). Brain structures that control movement
are also involved in perception, especially perception of other humans (Wilson &
Knoblich 2005). More broadly, the raison d’être of brains is motor control, which
makes thinking an internalized form of movement (Llinàs 2001). Lakoff & John-
son have fittingly entitled their second book Philosophy in the Flesh (1999). In
their work, however, the body appears almost exclusively as a natural, universal
Ecologies of gesture 

entity: there is no recognition of the fact that bodily experience is itself culturally
Gesture is symbolic body action evolved from the body’s practical engagement
with the world. It exemplifies that traditional separations between internal and
external, or ideational and material resources, and between cognition and com-
munication, embodied communication and practical work, between language use
and practical action, are obsolete. In the most characteristic situations of human
communication – symbolic, multimodal communication lodged in practical co-
operation – they are one. Gestures and languages, then, are not only means for
interacting and communicating – let alone for sharing information–, but for in-
habiting worlds together and managing interaction in the process. They have
evolved from and facilitate distributed cognition (Hutchins 2006).
However, there are limitations even to advanced, non-dualist conceptions of
the body which squarely situate the mindful body in the world. For this world, as
it is commonly construed, is a thing-world, not an interpersonal or intercorporeal
one. This might be the legacy of Heidegger with his emphasis on equipment and
manipulation, or related to the current interest in professional communication
and contexts of material work (Lave 1988). But human bodies do not only acquire
worldly skills in their engagement with objects or matter, but also through their
unmediated contact with other human bodies, and this contact – especially during
childhood – is also subject to cultural regulation and thus difference. For example,
de León, studying the socialization of young children in Maya-speaking Tzotzil
communities in Southern Mexico (de León 2000), has demonstrated that these
children experience social interaction far more frequently and intensely than chil-
dren in the industrialized countries of the Northern hemisphere through the me-
dium of direct physical (tactile) contact. Distinct patterns of intercorporeality dur-
ing childhood have implications for the kinds of participatory competence that the
child acquires. We need a much richer understanding of the body in its culturally
specific modes of participation, resonance, and agency, not so much to solve meth-
odological problems – these, I believe, have large been resolved – but to close the
gaps between the various concurrent, non-dualist conceptions of humanity that
are currently emerging at such a rapid pace.

4. Historical dimensions

An ecological approach to interaction which seeks to come to terms with the mul-
timodal nature of human action and interaction – including the fact that human
action more often than not involves tools and is situated in artefactual,
i.e. human-made, environments – comes up against the issue of the historical
 Jürgen Streeck

constitution of this world. Of course, the historical constitution of the human

mind – via the socio-culturally specific, human-made psychological tools that it
uses – is at the core of Vygotsky’s conception of the human mind (Vygotsky 1978).
This, I think, is the other big theoretical challenge that we face: to rethink the phe-
nomena that we study – the behaviors, signs, and bodily practices – in terms of
their historical constitution.
Humans are adaptive systems. Internal processes are structured so that they
integrate well with structures in the environment, including the structures of hu-
man-made symbolic environments. Major changes in cognitive and communica-
tive media result in transformations of cognitive architecture (Donald 1991). In
any moment of communication – i.e., symbolically mediated interaction – his-
torically evolved or sedimented structures, practices, and generalized experiences
– concepts – are in play and may subtly be altered in the process. Our field awaits
a kind of theoretical reframing that is dynamic and accomodates uninterrupted
change, instead of implicitly or explicitly assuming the static nature of symbolic
structures, grammar, or conversational organization. For example, given the im-
portance of sequence organization in conversational interaction – and, as we have
seen, the understanding of gestures – the question arises where these action se-
quences that conversational organization manages come from: how old they are,
how wide-spread, how variable. Some of these “speech act institutions” (D’Andrade
1984) may well have developed from interaction formats that, without being in-
nate, predate the arrival of speech:
Social institutions … are often robust, with deep histories, and their fortunes are
subject to patterns of cultural evolution on a time-scale different from the ephem-
eral interactions that nevertheless instantiate them (Enfield & Levinson 2006:€27).

Levinson has reflected on the origin and history of the parts of the human interac-
tion engine (Levinson, 2006) and argued that
interaction appears to have detailed universal properties. … The … cultural sys-
tems that have been studied reflect very similar, in some cases eerily similar, sub-
systems (Levinson 2006:€46).

Still, these properties are cultural, not biological, properties, notwithstanding the
fact that they are predicated on a human biology that makes the production of
culture – adaptive, local culture – possible.
This challenge, too, is a theoretical one: we usually do not need to reckon with
the historicity of the phenomena when we seek to describe the organization of
social interaction and the deployment of linguistic and gestural components in
interaction. It appears to me that the real challenges – or the more interesting chal-
lenges – at this point are not ones of research methodology – of identification,
Ecologies of gesture 

description, analysis, and synthesis of interaction and language units. While a

great deal of work still needs to be done, the convergences of various hitherto in-
dependent strands of language and interaction research demonstrate that the study
of interaction and language, or of talk and interaction, has developed into a robust,
normal science, vast differences in approach notwithstanding. But something im-
portant remains unexplained when we abstract away from issues that we cannot
explain with our own methodology or that fall outside the purview of our discipli-
nary knowledge: the intelligence and resonance of the bodies that organize their
dealings with one another without any conscious “mind” even being aware of the
fact (which is what usually takes place in the realm of gestural communication) or
the historical evolution and ongoing change of all the component practices and
resources that the interaction order is made of.

5. Dynamical systems

One abstract framework in terms of which these multiple ecological parameters

can be brought together in a coherent fashion is dynamical systems theory. Dy-
namic systems theory conceives communicative encounters “in terms such as en-
gagement and disengagement, synchrony and discord, and breakdown and repair”
(Shanker & King 2002:€607):
in a dynamic system, all of the elements are continuously interacting with and
changing in respect to one another, and an aggregate pattern emerges
from this mutual co-action (loc.cit.)
Social interaction is a “dynamic dance” (King 2004), an “activity of mutual at-
tunement” (Shanker & King 2002:€ 606), within which “each partner must
cooperate, moment by moment, in creating the coordination” (King 2004:€1). As
Streeck & Jordan (2009) have argued, each participant in this “dynamic dance”
constitutes an embodiment of the multiple contexts in which it has to sustain itself.
The “dynamic dance” is one of fluid, multi-scale, joint sustainment in multiple con-
texts simultaneously. A gesture is part of a speaking turn and embodies in its form
the immediate context in which this turn is produced: for example, by being a
gesture of rejection it embodies the offer or invitation that preceded it. But it may
simultaneously be about and thus embody the specific, historically developed re-
lationship between two people and be a move in the institutional context of that
relationship.€And finally it is a component in– and through its enactment sustains
– a historically evolved symbolic code: it may be a conventionalized gesture.
 Jürgen Streeck

Interaction … is never ‘about’ just one level of context.... Rather, it is simultane-

ously ‘about’ all of the scales of embodied context the participants bring to bear
during the interaction. Embodied action (including speech) always contributes to
the sustaining of multiple nested contexts at once (Streeck & Jordan 2009:€454).


Auer, Peter, & di Luzio, Aldo (eds.) 1992. The Contextualization of Language. Amsterdam:
Barsalou, Lawrence W. 1999. “Perceptual symbol systems.” Behavior and Brain Sciences, 22 (4):
577–609; discussion 610–560.
Bates, Elizabeth, Camaioni, Luigia, & Volterra, Virginia.1975. “The acquisition of performatives
prior to speech.” Merril-Palmer Quarterly, 21: 205–226.
Bateson, Gregory. 1971. “Communication.” In The Natural History of an Interview. Norman Mc-
Quown (ed.), (Vol. 95 – Series XV. Chicago: Micro-fiche. University of Chicago Library.
Bateson, Gregory. 1972. Steps to an ecology of mind. New York: Ballantine.
Bavelas, Janet, Chovil, Nicole, Lawrie, Douglas A., & Wade, Allan. 1992. “Interactive gestures.”
Discourse Processes, 15: 469–489.
Beattie, Goeffrey. 2004. Visible Thought: The New Psychology of Body Language. London:
Cienki, Alan, & Müller, Cornelia (eds.). 2009. Metaphor and Gesture. Amsterdam: Benjamins.
Connerton, Paul. 1989. How Societies Remember. Cambridge: Cambridge University Press.
D’Andrade, Roy G. 1984. “Cultural meaning systems.” In Culture Theory. Essays on Mind, Self,
and Emotion. Richard A. Shweder & Robert A. LeVine (eds.). 88–121. Cambridge:
Cambridge University Press.
Darwin, Charles. 1955 (1872). The Expression of Emotions in Animals and Man. New York: Phil-
osophical Society.
de León, Lourdes. 2000. “The emergent participant: Interactive patterns in the socialization of
Totzil (Maya) infants.” Journal of Linguistic Anthropology, 82: 131–161.
Donald, Merlin. 1991. Origins of the Modern Mind. Cambridge, MA: Harvard University Press.
Enfield, N. J., & Levinson, Stephen C. 2006. “Introduction: Human sociality as a new interdisci-
plinary field.” In Roots of Human Sociality. Culture, Cognition and Interaction. N.J. Enfield
& Stephen C. Levinson (eds.). 1–37. London: Berg.
Glenberg, Arthur M., & Kaschak, Michael P. 2002. “Grounding language in action.” Psycho-
nomic Bulletin & Review, 9(3): 558–565.
Goffman, Erving. 1964. “The neglected situation.” American Anthropologist, 66(6, Part 2): 133–136.
Goffman, Erving. 1971. Relations in Public. Microstudies of the Public Order. New York:
Basic Books.
Goffman, Erving. 1983. “The interaction order.” American Sociological Review, 48: 1–17.
Goodman, Nelson. 1976 (1968). Languages of Art. 2nd ed.. Indianapolis: Hackett.
Goodwin, Charles. 1986. “Gesture as a resource for the organization of mutual orientation.”
Semiotica, 62 (1–2): 29–49.
Goodwin, Charles. 2007. “Environmentally coupled gestures.” In Gesture and the Dynamic Di-
mension of Language: Essays in Honor of David McNeill. Susan D. Duncan, Justine Cassell
& Elena T. Levy (eds.). 195–212. Philadelphia: Benjamins B.V.
Ecologies of gesture 

Goodwin, Charles., & Goodwin, Margorie. H. 1986. “Gesture and coparticipation in the activity
of searching for a word.” Semiotica, 62 (1–2): 51–75.
Gullberg, Marianne, & Kita, Sotaro. 2009. “Attention to Speech-Accompanying Gestures: Eye
Movements and Information Uptake.” Journal of Nonverbal Behavior, 33: 251–277.
Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.
Hanks, William F. 1996. Language and Communicative Practices. Boulder, CO: Westview Press.
Harper, Douglas. 1987. Working Knowledge. Skill and Community in a Small Shop.€ Chicago:
University of Chicago Press.
Heath, Christian. 1986. Body Movement and Speech in Medical Interaction. Cambridge:
Cambridge University Press.
Heath, Christian. 1992. “Gesture’s discreet tasks: Multiple relevancies in visual conduct and in
the contextualisation of language.” In The Contextualization of Language. Peter Auer &
Aldo diLuzio (eds.). 101–128. Amsterdam: Benjamins.
Heidegger, Martin. 1962 (1926). Being and Time. New York: Harper and Row.
Hutchins, Edwin. 2006. “The distributed cognition perspective on human interaction.” In Roots
of Human Sociality. N.J. Enfield & Stephen C. Levinson (eds.). 375–398. London: Berg.
Johnson, Mark. 1987. The Body in the Mind. Chicago: University of Chicago Press.
Keller, Charles M., & Keller, Janet D. 1996. Cognition and Tool Use. The Blacksmith at Work.
Cambridge: Cambridge University Press.
Kendon, Adam. 1970. “Movement coordination in social interaction: Some examples described.”
Acta Psychologica, 32: 100–125.
Kendon, Adam. 2004. Gesture: Visible Action as Utterance. Cambridge: Cambridge University Press.
King, Barbara. J. 2004. The Dynamic Dance: Nonvocal Communication in African Great Apes.
Cambridge, MA: Harvard University Press.
Lakoff, George, & Johnson, Michael. 1999. Philosophy in the Flesh. Chicago: University of
Chicago Press.
Lave, Jean. 1988. Cognition in Practice. Cambridge: Cambridge University Press.
Levinson, Stephen C. 2006. “On the human interaction engine.” In Roots of Human Sociality.
N.J. Enfield & Stephen C. Levinson (eds.). 39–69. London: Berg Publishers.
Lllinàs, Rodolfo R. 2001. I of the Vortex. From Neurons to Self. Cambridge, MA: MIT Press.
McDermott, Ray P. & Roth, David. 1978. “Social organization of behavior: Interactional ap-
proaches.” Annual Review of Anthropology, 7: 321–345.
McQuown, Norman. 1956. The natural history of an interview. Chicago: Chicago Manuscripts in
Cultural Anthropology.
Mauss, Marcel. 1973 (1935). “The techniques of the body.” Economy and Society, 21: 70–88.
Mead, George H. 1909. “Social psychology as a counterpart to physiological psychology.” Psy-
chological Bulletin, 6: 401–408.
Mead, George H. 1934. Mind, Self and Society. Chicago: University of Chicago Press.
Merleau-Ponty, Maurice 1962. Phenomenology of Perception. London: Routledge.
Müller, Cornelia. 2003. “Forms and uses of the Palm Up Open Hand.” In The Semantics and
Pragmatics of Everyday Gestures. The Berlin Conference. Cornelia Müller & Roland Posner
(eds.). 234–256. Berlin: Weidler.
Polanyi, Michael. 1958. Personal Knowledge. Towards a Post-Critical Philosophy. New York:
Harper Torchbooks.
Schegloff, Emanuel A. 1979. “On the relevance of repair to syntax-for-conversation.” In Dis-
course and Syntax.Vol. 12. Talmy Givón (ed.). 261–288. New York: Academic Press.
 Jürgen Streeck

Schegloff, Emanuel A. 1984. “On some gestures’ relation to talk.” In Structures of Social Action. J.
Maxwell Atkinson & John Heritage (eds.). 266–295. Cambridge: Cambridge University Press.
Schegloff, Emanuel A. 2007. Sequence Organization in Interaction. A Primer in Conversation
Analysis. Cambridge: Cambridge University Press.
Schutz, Alfred. 1982. Collected Papers. Vol. 1–3. The Hague: Martinus Nijhof.
Shanker, Stuart G., & King, Barbara J. 2002. “The emergence of a new paradigm in ape language
research.” Behavioral and Brain Sciences, 25(3): 605–656.
Strathern, Andrew J. 1996. Body Thoughts. Ann Arbor: University of Michigan Press.
Streeck, Jürgen. 1993. “Gesture as communication I: Its coordination with gaze and speech.”
Communication Monographs, 60: 275–299.
Streeck, Jürgen. 1995. “On projection.” In Social Intelligence and Interaction. Esther Goody (ed.).
87–110. Cambridge: Cambridge University Press.
Streeck, Jürgen. 2002. “A body and its gestures.” Gesture, 2, 1: 19–44.
Streeck, Jürgen. 2008. “Laborious intersubjectivity: Attentional struggle and embodied com-
munication in an auto-shop.” In Embodied Communication in Humans and Machines. Ipke
Wachsmuth, Manuela Lenzen & Gunter Knoblich (eds.). 202–228. Oxford: Oxford Univer-
sity Press.
Streeck, Jürgen. 2009a. Gesturecraft. The Manufacture of Meaning. Amsterdam: Benjamins.
Streeck, J Jürgen 2009b. “Forward-gesturing.” Discourse Processes, 45 (3/4): 161–179.
Streeck, Jürgen., & Hartge, Ulrike. 1992. “Previews: Gestures at the transition place.” In Peter
Auer & Aldo di Luzio (eds.). The Contextualization of Language. 138–158. Amsterdam:
Streeck, Jürgen. & Jordan, J.Scott 2009. “Communication as a dynamical system: The impor-
tance of time-scales and nested contexts.” Communication Theory, 19: 445–464.
Streeck, Jürgen, Goodwin, Charles & LeBaron, Curtis (eds.) to appear. Embodied Interaction.
Language and Body in the Material World. Cambridge: Cambridge University Press.
Vygotsky, Lew. 1978. Mind in Society. Cambridge: Harvard University Press.
Wilson, Margaret. 2002. “Six views of embodied cognition.” Psychonomic Bulletin and Review, 9:
The neglected listener
Issues of theory and practice in transcription
from video in interaction analysis

Frederick Erickson €
University of California, Los Angeles

How phenomena are arranged in a transcript always incorporates a theory of

the phenomena represented by it. The playscript transcripts of conversation
analysis embody a logocentric view of interaction: they arrange interaction
around the sequential progress of talk and obscure relations of mutual influence
between the speaking behavior of speakers and the listening behavior of
listeners. A transcription system that adapts musical scores is better able to
capture the embeddedness of talk in interaction and the coordinated timing
of instrumental acts that provides an independent resource of interactional
organization. In this way it contributes to a renewed effort toward the study of
space, time, and visual phenomena in social interaction. The chapter includes
examples of €quasi-musical transcription whose notation €shows both the
sequential and the simultaneous occurrence of verbal and nonverbal actions of
speakers and listeners.

1. Introduction

It seems to me that the two most pressing issues for future work in the study of
language in interaction (or of interaction in relation to language use) concern
(1) empirical work and theory development that connect studies of larger scale
and longer term social processes with local scenes of face to face interaction – the
“structuration” or “macro-micro” issue in relation to oral discourse, and (2) the
tendency over the last generation to emphasize the study of talk’s local conduct
over other aspects of what happens locally during the course of face to face inter-
action. To discuss the first issue requires much more room than is available here,
and I have tried to do that in a book recently published (Erickson, 2004). Accord-
ingly I won’t even try to address that issue here, but will simply mention it by title.
The second issue is the focus of the comments that follow. I should make it clear to
 Frederick Erickson

the reader at the outset that I think my own work has contributed somewhat to the
very problem I discuss below. That is, in part, because the study of talk is so intel-
lectually important and empirically intriguing that there is a strong temptation to
give it central focus – and also because many scenes of interaction are primarily
constituted by the talk that is taking place in them (many, admittedly, but not all
– and there’s the rub.). What I say here I do not want to be interpreted as implying
that I think the past half century’s efforts in the study of naturally occurring talk
have not made significant progress from where we were in the late 1950’s. Rather,
I see that work as providing shoulders on which scholars can now stand to look
toward a slightly more distant edge of the horizon.

2. The selectivity of transcription

Over the past fifty years audiovisual recordings (formerly sound cinema film and
more recently videotape) have been increasingly used as a primary information
source in studies of language use in social interaction. As Ochs has noted in a clas-
sic paper (1979), all transcription approaches presuppose theoretical commit-
ments. This is because any transcription is inherently incomplete – the multidi-
mensional complexity of speech and nonverbal behavior involved in social
interaction is such that it cannot be fully represented in any single transcript.
Rather, transcription is necessarily selective, emphasizing some aspects of speech
or nonverbal behavior over others. This selectivity is especially apparent in “play-
script” transcription of speech. Using a succession of lines on a full page as a way
of organizing a transcript can show many details of the real-time conduct of talk,
lexically and (with special diacritical markings added) such transcription can even
show a good deal about speech prosody. But it tends to privilege in its representa-
tion the speaking activity of speakers over the listening activity of listeners that is
occurring simultaneously with the speaking. Thus the on-line influence of listen-
ers upon the on-line production of talk by speakers tends to be obscured.
In other words, “incipit verbum,” whatever its status may be as a theological
proposition appearing at the outset of the Gospel According to John (Jn 1.1: “In the
beginning was the Word...”) is not an appropriate programmatic foundation for
the study of social interaction. Such interaction is a pre-human phenomenon and
thus its organization and conduct is, at least in part, phylogenetically prior to the
evolution of the human capacity for speech. Experientially for its participants, so-
cial interaction begins with space and time, and with vision in addition to hearing,
usually before any words are uttered. Social interaction in circumstances of co-
presence among participants happens in particular places, in real time, and as the
participants are able to monitor one another’s actions visually as well as auditorially.
The neglected listener 

(Indeed one of my teachers, Edward T. Hall (1966) showed not only how within
the course of social interaction we judge interpersonal distance visually in a vari-
ety of ways, but how the whole sensorium is involved as we sense distance audito-
rially, kinaesthetically, tactilely, olfactorily, and even thermally, as we monitor in-
tuitively the body heat of those engaged with us in interaction face to face.)

3. Social interaction as an ecosystem

As we speak during the course of social interaction we are not doing so on the
telephone, nor are we talking to one another through a keyhole; we have available
visual information on the listening activity of listeners on-line, while we speak. I
emphasize while here (see Condon and Ogston 1967) because the information
speakers have about listening reactions of audience members comes to us instant-
ly through vision – it is available in present moments as we speak, and everything
that interactional participants are doing together, verbally and nonverbally, con-
tains potentially significant information for the interpretation of what is going on
in the interaction, whether those are the interpretations of analysts or of the par-
ticipants themselves. These are key insights taken from the then cutting edge field
of cybernetics that informed the pioneering approach to the analysis of social in-
teraction developed a half century ago by Bateson, Fromm-Reichman, McQuown,
Birdwhistell, and others in the interdisciplinary group that produced the unpub-
lished monograph “The Natural History of an Interview” (see also Ruesch and
Bateson 1951). This approach has been called “Context Analysis” by Kendon, who
observes “One comes to recognize that what a person may be saying, for example
while he is saying it, may be shaped by information he is taking in from his recipi-
ent; and by the same token, how the recipient is behaving, where he is placing his
headnods, his smiles and frowns, and how he is patterning his visual attention,
may also be shaped by, even as it is shaping the activity of the speaker.” (Kendon,
1990:€29). As I have said elsewhere along similar lines, to be engaged in social in-
teraction is like climbing a tree that climbs you back in the same time. (Erickson,
In his classic essay “The neglected situation” Goffman called face to face inter-
action an “ecological huddle” (Goffman 1964:€135). Such an ecology is only pos-
sible to maintain when the participants in interaction not only occupy the same
spatial and temporal setting but construct it through their conjoint action. As
McDermott has said (1976:€36), “people in interaction constitute environments for
each other.” The conduct of interaction must be organized socially; that is, the ac-
tivities of the various participants must relate in ways that take account of one
another. This is a local ecology of sustained mutual attention and influence.
 Frederick Erickson

A necessary condition for such mutuality is that participants in interaction

must be in the same place, and that situatedness involves both objective and sub-
jective aspects; phenomena that are present in the scene and the phenomenology
of the participants – their patterns of attention. Participants must orient to one
another within the spatial setting that is given in the built environment (size of the
setting as defined by walls, stationary placement of furniture or tools) as well as
within the spatial setting that is established in their patterns of interpersonal dis-
tance. They also must orient to one another within a common temporal setting
that is established by the timing of their speech and body motion (by “timing” here
I mean actual duration in “real time”, not the compressed time of analytic synop-
sis). This spatial and temporal situating of interaction permits common expecta-
tions among participants of where action will happen and when it will happen, in
immediate and next moments. Actions contain cues for their own interpretation.
Bateson (1956) called these “metamessages” and Gumperz (1982) called them
“contextualization cues.” Moreover, as conversation analysts have elegantly shown
in their studies of speech, particular communicative actions, as part of an overall
contextualization cueing process project in the course of their doing toward turn-
ing points and (eventually) toward concluding points, all of which will occur in
future time. Communicative actions, both verbal and nonverbal, are also done si-
multaneously in concert with the actions of others. If participants take action out-
side the frames of timing and of spacing that their conjoint action is creating and
sustaining– action that happens too soon or too late, action toward there rather
than toward here – what happens are interactional “stumbles” – awkward mo-
ments of behavioral asymmetry in which abrupt posture shifts and gestural and
verbal re-starts take place. Such action can be characterized as inadequately social;
i.e. the social organization of face to face interaction can be conceived as resting
upon a foundation of time in conjoint use and space in conjoint use.

3. Playscript transcription privileging speech over other semiotic media

It is fair to say that over the last forty years the analysis of social interaction has
focused primarily on speech. Tremendous strides have been taken in the study of
oral discourse within immediate social interaction – I think in particular here of
the study of indexicality in talk that has developed within linguistic anthropology
(c.f. Silverstein, 1976, 1992 and Hanks 1996:€176–183, 230–236) and of the work of
conversation analysis (see the review by Goodwin and Heritage 1990 and the re-
cent volume edited by Prevignano and Thibault 2003). But everything is a tradeoff.
Progress in the study of naturally occurring speech carries with it a tendency to-
ward “logocentrism,” an overemphasis on speech by which the larger “whole” of
The neglected listener 

social interaction – its collective temporal/spatial/bodily organization – gets short

shrift analytically.
Routine transcription practices manifest this central focus on speech at the
expense of additional aspects of “situatedness” in interaction and of additional
semiotic media by which meaning is communicated in interaction. One of these,
the transcription system developed by Jefferson in collaboration with Sacks and
Schegloff (as exemplified in Sacks, Schegloff and Jefferson 1974 and as discussed in
Heritage and Atkinson 1984) has currently become almost a standard approach in
transcribing speech. It was developed on the IBM Selectric typewriter and is an
ingenious use of that machine’s capacities to portray, in relatively simple and direct
ways, features of speech that had previously been ignored in the analysis of the
social conduct of speaking or were previously only able to be represented by labo-
rious and arcane procedures of phonetic transcription. Among these features are
hesitations and re-starts, in-breaths, “sound-stretching” of vowels, volume empha-
sis in syllables, the timing of pauses, speech that overlaps between speakers, and
speech that alternates between speakers with no gap and no overlap.€ Jefferson’s
system has an overall playscript format which shows turns taken by successive
speakers by dropping down vertically at the left margin of the typed (printed)
page. (Playscript formats of one sort or another, whether they follow Jefferson’s
transcription conventions or not, are now so ubiquitous as to appear to be the only
way to transcribe speech for purposes of analysis and publication.)
The problem with any approach to transcription is that it illuminates and fore-
grounds some aspects of social interaction and leaves other aspects either un-rep-
resented entirely or in the background of analytic attention. A transcription ap-
proach is an intellectual tool, and like all tools it is designed for particular uses. A
screwdriver does one kind of work very well; a chisel does another. It is possible to
use a screwdriver as a chisel, and vice versa, but the results are not nearly so satis-
factory (and the work is accomplished in not nearly so efficiently a way) as when a
particular tool is used in the manner for which it was designed.
Because of this transcription of verbal and nonverbal behavior is not merely a
theory-neutral empirical procedure in the study of social interaction Rather, it is a
manifestation of theoretical commitments. (For elaboration on this point see the
discussion cited at the outset of this paper, in Ochs 1979). It follows that to criticize
a particular transcription approach for certain limitations is not to deny its utility
as a tool but to point out that its best uses are particular to the theoretical presup-
positions out of which it was developed. All transcription approaches have their
limits and attendant strengths, as Ochs argues persuasively. It is up to the analyst
to recognize those limits, to use them to advantage when appropriate and, when it
becomes necessary, to design other tools for other uses (thus adopting certain
other limits in order to maximize certain other utilities).
 Frederick Erickson

It needs to be recognized that playscript transcription approaches do privilege

speech at the expense of the other kinds of communicative behaviors that are in-
volved in interaction. Moreover, it is not just that playscript transcription presents
various aspects of speech in relatively great detail and presents nonverbal activity
in less detail, it is that such transcription makes it appear as if the primary actions
of agents that are occurring in the interaction are those of speaking; that it is talk
that moves social interaction along and that everything else that is happening is in
some way a consequence of the actions of talk, as if what in the overall scene is not
talk were a tail being wagged somehow by the talking.

4. Alternative approaches to “multimodal” transcription

As a corrective to the inherent logocentrism in playscript transcription, alternative

transcribing procedures can show the simultaneous actions of listeners and speakers
and can identify more clearly than can playscript approaches the relations of the ac-
tions of the interactional participants in real time and actual space. Because of limi-
tations of length in this discussion I will mention only a few examples of these pro-
cedures here. One is the postural positioning diagrams and line drawings presented
by Scheflen (1973) in his monograph length analysis of a single family therapy ses-
sion. Another is the diagrams presented in the collection of essays on instances of
dyadic and small group interaction by Kendon (1990). Another is the published ver-
sion of Goodwin’s doctoral thesis on relations between gaze and speech (Goodwin
1981) and his recent analyses of professionals’ learned “ways of seeing” (Goodwin
2000a, see also van Leeewen and Jewitt 2000 more generally as a resource, the edited
volume in which Goodwin’s 2000a essay appears). Another is the attempts that I and
others have made to represent the timing patterns of verbal and nonverbal behavior
in interaction using quasi-musical notation (Erickson 2003, 2004).
What many of these attempts have in common is that they display the various
actions of interactional participants in relation to a common time line which
shows “real time. (Even an ordinary transcript of speech, when it does not display
speech over a time line but rather in some form of playscript transcription,
presents the speech in a way in which it can be read faster than life – outside the
actual durations of its performance.) In the early work using audiovisual records,
e.g.€that of Scheflen and Kendon, this time line was established through slow mo-
tion analysis of cinema film using a number code for frames of cinema film, each
frame of film being exposed at 1/24 of a second. More recently, in analysis based
on videotape, digital clock-face time codes in minutes, seconds, and microsec-
onds provide the evidence for the timing of nonverbal and verbal phenomena,
The neglected listener 

and there are also now available various computer programs for the analysis of
motion and speech.
In such transcription the time line is presented horizontally on the printed
page and the behavior of separate individuals is overlaid vertically across the time-
line, as in the schematic representation below, which shows various verbal and
nonverbal actions of individuals A, B, and C – these could be turns at speech, sin-
gle word utterances as an exclamation or as a listening reaction, head nods, hand
and arm gestures (on this, see especially Kendon 1997), shifts in postural position,
glancing toward and then away from a computer screen or a tire pressure gauge,
reaching for a book on a desk or a plate of food on a dinner table, putting a forkful
of food in one’s mouth:
0 5 10 15 20 25 30 35 40 45
A: xxxx yy xxx zzzzz zzzz yyy xxxx bbb
B: zzzzz xxxx aa xxxxx zz xxxxxxx
C: xx y xxxx zzzzz c
Figure 1.╇ “Horizontal” transcription of verbal and nonverbal behavior on a time line.

In the case of quasi-musical notation (see Erickson 2003, 2004, 2008) the time-
line is not indicated by an ordinal scale displayed horizontally on the page, as in
the example above, but as on a musical page. In such pages the horizontal ar-
rangement of “bar lines” presents each “measure” as approximately equal in du-
ration. A closer representation of actual timing values is provided by the musical
notes within each two beat “measure” – half notes, quarter notes, eighth notes,
sixteenth notes, and thirty-second notes. (When the simultaneous actions of a
set of individuals are displayed on the musical page what results is a diagram
that appears like an orchestra score, and overall such a score resembles the ordi-
nally scaled time line version in the schematic diagram shown above.) An illus-
tration follows.
The quasi-musical notation in the example above shows eight successive two-
beat measures (10–17) from a larger transcript published and discussed in greater
elaboration in Erickson (1992 and 2004). Those discussions emphasized the con-
tent of talk, which is the beginning of a collective complaint sequence in which the
family discusses how much things cost nowadays. The transcript represents the
talk and eating behavior that occurred during a family dinner table conversation.
Seated at the table are the father (Fa), the mother (Mo), a guest (G), the oldest
brother who is in junior high school (B-1), the next oldest brother (B-2), the next
oldest brother (B-3), the sister who is in 3rd grade, and the youngest brother (B-4).
The transcript has a horizontal strip on which verbal and nonverbal activity of
each participant can be shown.
 Frederick Erickson

Figure 2.╇ Quasi-musical notation

The neglected listener 

Food had been served by the mother on each person’s plate (fried chicken,
potatoes, and a cooked vegetable). A stainless steel salad bowl sat near the middle
of the table and family members could, if and when they wished, reach for the
bowl, bring it to a position in front of their plate, and help themselves to a serving
of salad – lettuce and tomato from the family garden. A large metal fork and spoon
were placed in the bowl to enable food transfer to one’s plate. In order to get food
from one’s plate into one’s mouth it was necessary to spear a bit of food on one’s
plate with a small fork and then raise it to one’s lips. Because the plates were ce-
ramic and the salad bowl was metal, when the metal utensils hit the surface of the
plate or bowl they made small clicking sounds, which were audible on the video
recording. Thus the onset of food transport motions could be seen and heard to
begin with a click, followed by movement of the utensils toward either one’s plate
(in the case of salad service transfer) or to one’s lips (in the case of transferring one
bit of food to one’s mouth). These sounds and motions of food transfer are indi-
cated in the transcript. “FP” with an eighth note shows the point at which one of
the participants put their fork to their plate, making a click. “FM” indicates the
point at which the fork reached the mouth of the eater. A horizontal line below the
speech line indicates a forkful of food being held motionless in mid-air by an eater
(or moving only very slowly forward), and a jagged line indicates accelerated for-
ward motion of the fork toward the eater’s mouth. (Thus in measure 10 the oldest
brother’s fork is shown as motionless in mid-air, with motion toward his mouth
beginning at the onset of measure 11 and the fork reaching his mouth on the be-
ginning of the second beat of measure 11. Notice also at the beginning of measure
10 the father’s fork hit his plate at the beginning of the first beat. He did not pick
up food at that point, and his fork hit his plate again at the beginning of the second
beat of measure 11. The reader is encouraged to look for other FP-FM occurrences
throughout the eight measures shown above.)
The transcript also shows how the eating behavior of various family members
is temporally coordinated with the speaking behavior that was occurring. Usually
syllables that appear at the beginning of each of the two beats in a measure are
pronounced with volume stress. Thus in measures 10–11 as the oldest brother says
“seventy five dollars goes in a day” he did it with the following stress pattern: “sev-
enty dollars goes in a day.” This clause is initiated with the syllable “se” and is
completed with the word “day” and those syllables both appear at the beginning of
the first beat in each successive measure. In other words, the three volume-stressed
syllables in the clause appear at a regular time interval in relation to one another
– marking an underlying cadence. (The reader should now look across other meas-
ures to see the cadence patterns that are apparent in the verbal transcription that
uses quasi-musical notation.)
 Frederick Erickson

Having identified cadential patterning in the speech of the various partici-

pants, now look for the ways in which the nonverbal food transport behavior ar-
ticulates in cadence with the speech behavior that is occurring simultaneously and
successively. Notice in measure 10 that as the oldest brother utters most of a clause
he holds his fork motionless in mid-air, in front of his face. Then on the first beat
of measure 11 he begins to move his fork toward his mouth. The fork reaches his
mouth exactly at the beginning of the next beat, the second beat of measure 11.
Also in measure 11 the transcript shows that exactly on the beginning of the sec-
ond beat in simultaneity with the oldest son’s putting his fork to his mouth, the
father put his fork to his plate. As that was happening the guest uttered the second
of five successive laugh sounds – the second one receiving volume stress – and si-
multaneously B-3 uttered the first stressed syllable in the word “dollars.” This ca-
dential pattern – with stressed syllables or onsets of speech tending to occur, more
often than not, “on the beat” – is further marked at the beginning of measure 10 as,
simultaneously with B-1’s uttering the “se” syllable of “seventy,” both the father
and the guest put their forks to their plates in audible simultaneous clicks.
As a final way of illustrating the interdigitation of timing patterns in both
speaking and eating we can turn to the transcript’s representation of what happens
with the salad bowl. At the very beginning of measure 11 the mother had reached
across to the center of the table and picked up the salad bowl. She then moved it
relatively slowly to a position in front of her plate, setting the bowl down (with an
audible click) exactly on the beginning of the second beat of measure 12. In meas-
ure 13 she had reached for the serving fork and spoon in the bowl (however, the
transcript does not show this). Then at the very beginning of measure fourteen the
metal fork and spoon hit the bottom of the steel bowl with an audible click, on the
beat. After that the mother gathered pieces of lettuce and tomato between the fork
and the spoon in the bottom of the bowl. The fork and spoon then remained mo-
tionless in the bottom of the bowl during measure 15 while the oldest son was
saying his next major utterance, which was a question addressed across the table
to his mother. Immediately after he began to say “foo” in “food shoppin’” at the
beginning of measure 16 (notice that with that fricative and vowel he had disam-
biguated the utterance “how about last time you went/” which without a gram-
matically completing next word was still ambiguous, in a state of incompletion.
That clause could have been completed with the words “bowling” or “clothes shop-
ping”, but in this case the “foo” indicated that the word in production was “food”).
mother begin to lift the fork/spoonful of salad out of the bowl, which she then
transported toward her plate, with the fork and spoon touching the surface of the
plate with an audible click exactly at the beginning of the first beat of measure 17,
in time with the overall cadence pattern. In other words, the transcript shows the
The neglected listener 

mother having “put on hold” in mid-course her transport of salad from the salad
bowl to her plate while her oldest son was addressing her with a question. And
then, as her son completed his utterance and she had quickly brought the salad to
her plate on the next cadential beat, she replied to her son’s question with an iron-
ic rejoinder, “well, we won’t talk about that.”
The transcript’s depiction of the mother’s temporal accommodation of her
salad transport to the timing of her son’s uttering his question to her is analogous
to what that son (B-1) did in measures 10 and 11. Notice that as he said the utter-
ance “seventy-five dollars goes in a day” he held his forkful of food in mid-air until
he had completed the utterance with the final syllable “day,” which also completed
the grammatical unit. It was on the next “beat” of measure 11, and exactly on that
cadential beat, that the son then put the forkful of food to his mouth.
This kind of temporal articulation and accommodation of two different kinds,
or strata, of orderings of interaction at the dinner table – the enacted social order of
eating together and the enacted social order of conversing together – cannot be
shown in playscript transcription nearly so clearly as it can by some sort of “horizon-
tal” multimodal transcription that displays selected aspects of both speaking behav-
ior and eating behavior. The quasi-musical transcript shown and discussed above is
one way to transcribe to show the inter-digitation of talking and eating. (N.B. A
video clip of this example of a dinner table conversation can be viewed on the inter-
net by accessing the homepage of my faculty website at the University of California,
Los Angeles. That address is <>)
The previously discussed multimodal transcript did not emphasize spatial or
postural relationships between the various interactional participants and between
individuals and particular objects in their own personal surrounds as individuals
make use of those objects in their interaction. These can be shown concretely by
line drawings (see especially Scheflen 1973 for examples ) or more abstractly by
flow charts of various kinds. Sometimes the spatial displays are tied to a time line
and sometimes it becomes necessary to represent spatial relationships and tempo-
ral relationships separately in diagrams. Hybrid approaches are also possible. It is
now quite simple to import photographs from digital video into playscript tran-
scription of speech (as is especially characteristic in the recent work of Goodwin)
and the effect of this is to “interrupt” to some extent the logocentric implications
of the playscript transcription.
Formerly it was very expensive to print analytic charts in which information
about interactional behavior was displayed “horizontally” on a time line, or in
musical typescript. With the advent of digital typesetting, however, those costs
have decreased substantially. In addition, digital video recording and analysis soft-
ware also makes it logistically easier and less expensive than it used to be to collect
and analyze audiovisual records of social interaction.
 Frederick Erickson

6. Conclusion

The “multi-modality” of the semiotic means by which humans communicate with

one another in social interaction has become more significant theoretically. As
social interaction is considered as an ecosystem of mutual influence among par-
ticipants (Goodwin in press calls this a “semiotic ecology”), the time may be ripe
for renewed effort toward the study of space, time, and visual phenomena in social
interaction. Scholars seem to be attending more and more to “multi-modal” as-
pects of the organization of social interaction, as if they were rediscovering the
analytic and theoretical insights of the early “context analysts,” (see for example
recent publications on “multi-modal discourse analysis” such as van Leeuewen
and Jewitt 2000, Kress and van Leeuwen 2001, Levine and Scollon 2004, and the
essay on action and embodiment by Goodwin 2000b. See also the special issue of
Musicae Scientiae 1999/2000 on the musicality of social interaction, as well as van
Leeuwen 1999 and Auer et al 1999.)
Thus the prospect seems promising for nonverbal and temporal aspects of in-
teraction to receive more attention in relation to speech than in the recent past. I
hope that listening activity in relation to speaking activity, and other aspects of tem-
poral and spatial organization in the conduct of social interaction by which infor-
mation is made available to participants visually as well as auditorially, will be
treated in the future as phenomena of serious research interest, providing better
balance with our continuing analytic interests in the organization and uses of talk.


Auer, Peter, Couper-Kuhlen, Elizabeth and Müller, Frank. 1999. Language in time: The rhythm
and tempo of spoken interaction. New York and Oxford: Oxford U. Press.
Bateson, Gregory. 1956. “The message ‘this is play.’” In Group Processes. B. Schaffner (ed.). New
York: Josiah Macy, Jr. Foundation.
Condon, William and Ogston, William D. 1967. “A segmentation of behavior.” Journal of Psychi-
atric Research 5: 221–235.
Erickson, Frederick. 1986. “Listening and speaking.” Language and linguistics: The interdepend-
ence of theory, data, and interpretation. Georgetown University Roundtable on Languages
and Linguistics 1985. In Deborah Tannen and James Alatis (eds.). Washington, DC:
Georgetown University Press.
Erickson, Frederick. 1992. “They know all the lines: Rhythmic organization and contextualiza-
tion in a conversational listing routine.” In The contextualization of language. Peter Auer
and Aldo di Luzio (eds.). 365–397. Amsterdam/Philadelphia: John Benjamins Publishing
The neglected listener 

Erickson, Frederick. 2003. Some notes on the musicality of speech. In Georgetown University
Roundtable on Languages and Linguistics 2001. Deborah Tannen (ed.). Washington DC:
Georgetown University Press.
Erickson, Frederick. 2004. Talk and social theory: Ecologies of speaking and listening in everyday
life. Cambridge: Polity Press.
Erickson, Frederick. 2008. Musicality in talk and listening: A key element in classroom dis-
course as an environment for learning. In Communicative musicality: Exploring the basis of
human companionship.€Stephen Malloch, S. and Colwyn Trevarthen (eds.). 449–463. Ox-
ford: Oxford University Press.
Goffman, Erving. 1964. The neglected situation. In The ethnography of communication. John J.
Gumperz and Dell Hymes (eds.). Special issue of American Anthropologist 66: 6, Part 2:
Goodwin, Charles. 1981. Conversational organization: Interactions between speakers and hearers.
New York: Academic Press
Goodwin, Charles. 2000a. “Practices of seeing”. In Handbook of visual analysis. T. van Leeuwen
and C. Jewitt (eds.). London: Sage Publications.
Goodwin, Charles. 2000b “Action and embodiment within situated human interaction.” Journal
of Pragmatics 32: 1489–1522.
Goodwin, Charles. in press. “Constructing Meaning Through Prosody in Aphasia.” In Prosody in
Interaction. Barth-Weingarten, D. Reber, E. and Selting, M. eds. Amsterdam: John Benjamins
Goodwin, Charles, and Heritage, John. “Conversation analysis.” Annual Review of Anthropology
19: 283–307
Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge U. Press.
Hall, Edward T. 1966. The Hidden Dimension. Garden City, NJ: Doubleday.
Heritage, John. and Atkinson, D. Maxwell. (1984) “Introduction.” In Structures of social action:
Studies in conversation analysis. J.Maxwell Atkinson and John Heritage (eds.). Cambridge:
Cambridge U. Press.
Hanks, William 1996. Language and communicative practices. Boulder, CO: Westview
Kendon, Adam. 1990. Conducting interaction: Patterns of behavior in focused encounters. Cam-
bridge: Cambridge U. Press.
Kendon, Adam. 1997. “Gesture.” Annual review of anthropology. 26: 109–128.
Kress, Gunther and van Leeuwen, Theo. 2001 Multimodal discourse: The modes and media of
contemporary communiction. London: Arnold.
Levine, Philip.€and Scollon, Ron. (eds.) 2004. Discourse and technology: Multimodal discourse
analysis. Georgetown U. Press.
Musicae Scientiae 1999/2000. Special issue of Musicae Scientiae: Journal of the European Society
for the Cognitive Sciences of Music 1999/2000.
Ochs, Elinor. 1979. “Transcription as theory.” In Elinor Ochs and Bambi B. Schieffelin (eds.)
Developmental pragmatics. New York: Academic Press.
Prevignano, Carlo and Thibault, Paul. 2003 Discussing conversation analysis: The work of Emanuel
A. Schegloff. Amsterdam and New York: John Benjamins.
Ruesch, Juergen and Bateson, Gregory. 1951. Communication: The Social Matrix of Psychiatry.
New York: Norton.
Sacks, Harvey, Emanuel A. Schegloff, and Gail Jefferson. 1974 A simplest systematics for the
organization of turn-taking in conversation. Language 50: 696–735.
Scheflen, Albert. 1973. Communicational structure: Analysis of a psychotherapy transaction.
Bloomington: Bloomington U. Press.
 Frederick Erickson

Silverstein, Michael. 1976. “Shifters, verbal categories, and cultural description.” In Meaning in
anthropology. K. Basso and H. Selby (eds.). Albuquerque: School of American Research.
Silverstein, Michael. 1992. “The indeterminacy of contextualization: When is enough enough?”
In The Contextualization of Language. P. Auer and A. di Luzio, (eds.) Amsterdam and
Philadelphia: John Benjamins.
van Leeuwen, Theo. 1999. Speech, music, sound. New York: St. Martin’s Press.
van Leeuwen, Theo. and Jewitt, Carey. 2000. Handbook of visual analysis. London: Sage Publications
Dialogical dynamics
Inside the moment of speaking

John Shotter
University of New Hampshire

As I see it, all communication begins in, and continues with, our living,
spontaneous, expressive-responsive (gestural), bodily activities that occur in the
meetings between ourselves and the others and othenesses around us. Indeed,
as living, embodied beings, we cannot not be responsive in some fashion to
the expressions of others (spoken, written, or otherwise), and to other kinds
of events, occurring in our immediate surroundings. In this article I outline
methods for exploring the unfolding dynamics of our utterances in their speaking
and how they can give rise to a ‘shaped’ and ‘vectored’ sense of our moment-by-
moment changing placement within the situation of our talk – engendering in us
both unique anticipations as to what-next might happen along with, so to speak,
‘action-guiding advisories’ as to what-next we might do.

“On the one hand it is clear that every sentence in our language ‘is in order as
it is’. That is to say, we are not striving after an ideal, as if our ordinary vague
sentences had not yet got a quite unexceptionable sense, and a perfect language
awaited construction by us. – On the other hand it seems clear that where there
is sense there must be perfect order. So there must be perfect order even in the
vaguest sentence” (Wittgenstein 1953, no.98).

Much attention in linguistics has been paid to patterns of already spoken words, to
the spatial shapes or forms of already completed acts of speaking.1 Instead, in this
article, I want to focus attention on people’s words in their speaking, on what hap-
pens in the course of a person’s utterance, and on the dynamic ways in which

1. “If I had to say what is the main mistake made by philosophers of the present generation...
I would say that it is that when language is looked at, what is looked at is a form of words and
not the use made of the form of words” (Wittgenstein, 1966, p.2).
 John Shotter

people make use of words in the course of their other actions, as well as on the
subtle details of how, as their use of words unfolds in responsive relation to those
to whom they are addressed, people adjust their expressions accordingly.
This, however, is an approach I came to only slowly. When I first began to
study language (Shotter 1968), I was strongly influenced by Chomsky (1957, 1965).
Central to his whole approach then was idealization: “Linguistic theory,” he said,
“is concerned primarily with an ideal speaker-listener...” (p.3). For, as he then saw
it, due to “memory limitations, distractions, shifts of attention and interest, and
errors (random or characteristic) in applying his knowledge of the language in
actual performance” (p.3), our utterances, actual acts of speaking were too disor-
derly to study. Indeed: “A record of natural speech will show numerous false starts,
deviations from rules, changes of plan in mid-course, and so on. The problem for
the linguist, as well as for the child learning the language, is to determine from the
data of performance the underlying system of rules that has been mastered by the
speaker-hearer and that he puts to use in actual performance” (p.4). In other
words, his central concern in his theory is with discovering the “mental reality”
underlying actual behavior. Now, with reference to Wittgenstein’s (1953) remark
above, my interest has switched completely to a focus on the deviations that
Chomsky, rightly, sees as standing in the way of conducting a “natural scientific”
investigation into the workings of language.2
In the dialogical approach to interaction and interaction analysis I now take, I
have been centrally influenced, not by their theories – as they all especially eschew
idealizations – but by certain specific utterances or expressions in the writings of
Wittgenstein, Vygotsky, Bakhtin, and Voloshinov, as well as drawing influences
from expressions in Merleau-Ponty’s and Garfinkel’s writings. As I see it, all com-
munication begins in, and continues with, our living, spontaneous, expressive-re-
sponsive (gestural), bodily activities that occur in the meetings between ourselves
and the others and othenesses around us. Indeed, as living, embodied beings, we
cannot not be responsive in some fashion to the expressions of others, as well as to
other kinds of events, occurring in our immediate surroundings – hence my

2. “The more narrowly we examine actual language, the sharper becomes the conflict between
it and our requirement. (For the crystalline purity of logic was, of course, not a result of investiga-
tion: it was a requirement.) The conflict becomes intolerable; the requirement is now in danger
of becoming empty. – We have got on to slippery ice where there is no friction and so in a certain
sense the conditions are ideal, but also, just because of that, we are unable to walk. We want to
walk: so we need friction. Back to the rough ground!” (Wittgenstein, 1953, no.107). In other
words, what Chomsky (1965) thinks of as “degraded” examples of language use – which makes
the learning of the ideal principles of syntax difficult for a child – exemplify for Wittgenstein the
rich details (many of) which contribute to the unique sense we make of a person’s expressions.
For him, idealizations strip out essentials, not what is inessential.
Dialogical dynamics 

emphasis above on my own embodied responsiveness to the writings (the

expressive-responsive utterances) of others. It is this that leads me to the view that,
as I see it, abstract and general theories as such are of little help to us in the unique
living of our unique lives together, whether as ordinary people or as professional
practitioners. While the specific words of another person, uttered as a ‘reminder’
at a timely moment, can exert a crucial influence in the development and refine-
ment of our ongoing activities and practices. Primarily, as I see it, it is the living,
bodily activity of speaking to other people responsively that matters, that ‘moves’
them, not the simply displaying before them of patterned forms. And such spon-
taneously responsive occurrences can, as Garfinkel (1967) puts it continually oc-
cur “for ‘another first time’” (p.9).

1. Joint participation in a flow of spontaneously responsive

activity: Shared ‘feelings of tendency’

In taking this approach, I was influenced early on by a remark of Vygotsky’s (1962)

about the “basic laws governing human development.” As he saw it, “one of them
is that consciousness and control appear only at a late stage in the development of
a function, after it has been used and practiced unconsciously and spontaneously.
In order to subject a function to intellectual and volitional control we must first
possess it” (p.90). In other words, long before we are individually consciously
aware of deliberately acting to achieve a goal, we are nonetheless coming to act
unconsciously and spontaneously in ways intelligible to those around us. And, as
was clear from all the rest of Vygotsky’s work, while we might possess as an aspect
of our biological inheritance a great range of ‘lower’ psychological functions, the
gradual growth of our voluntary ability to organize them into ‘higher’, more com-
plex forms, comes about through other, already competent members of our verbal
community, ‘in-structing’ us verbally in how do so: “All the higher psychic func-
tions are mediated processes, and signs are the basic means used to master and
direct them. The mediating sign is incorporated in their structure as an indispen-
sable, indeed the central, part of the total process. In concept formation that sign
is the word, which at first plays the role of means in forming the concept but later
becomes its symbol” (1962:€56, my emphasis) – and one person’s words, their bod-
ily voicing of an utterance, their expressions, can exert this immediate and sponta-
neous (gestural) effect on (and in) another person. And later, the speaking of their
words to ourselves is “the means by which we [can come to] direct our mental op-
erations, control their course, and channel them toward the solution of the prob-
lem confronting us” (Vygotsky 1962:€58).
 John Shotter

Thus, as I see it, from a dialogical point of view, our intellectual lives are not
primarily based in picture-like mental representations, i.e., inner structures of
only a formal (static patterned) kind, but in ‘inner’ dialogically-structured move-
ments, in a dialogical dynamics giving rise to unfolding movements which shift
this way and that in a distinctive fashion, movements whose ‘shape’ can be ‘felt’ or
‘sensed’ but not pictured, or known at all in a propositional form. These sensed or
shaped ‘inner movements’ are, I take it, the “linguistic intuition[s] of the native
speaker” against which such linguists as Chomsky (1965:€19)3 test their theories of
syntactic structure, or other such general and abstract features of our use of lan-
guage – what he now calls their “I-language” (Chomsky: 2000).4
But here, in talking of such sensed or felt ‘inner movements,’ i.e., of thought as
not in any way separate from feeling, I am taking an approach toward these issues
very different in kind to Chomsky’s: mine is a practical-descriptive kind of ap-
proach rather than a theoretical-explanatory one, as I will explain in a moment.
For, in line with Vygotsky’s comments quoted above, from a dialogical point of
view, our inner intellectual lives can be seen as consisting in an ‘orchestrated’ in-
tertwining of many different kinds of influence: conscious and unconscious ones,
cognitive and affective, deliberate and spontaneous, biologically given and cultur-
ally developed ones, and in fact, as we shall see, many others of a much more oc-
casional or momentary kind that are at work in the immediate practical surround-
ings of a particular utterance.
As William James (1980) noted in his famous “The Stream of Thought” chap-
ter, we have failed in the past, in discussing the nature of such dynamic forms, to
register “the transitive parts” of the stream and succumbed to an “undue empha-
sizing of [its] substantive parts [i.e., its resting-places]” (p.237). In so doing, we
have tended to confuse “the thoughts themselves... and the things of which they
are aware... [But, while] the things are discrete and discontinuous... their comings
and goings and contrasts no more break the flow of thought that thinks them than
they break the time and space in which they lie” (p.233). To break the strangle-
hold of this compulsion upon us, James entreats us thus: “Now what I contend for,

3. “... there is no way to avoid the traditional assumption that the speaker-hearer’s linguistic
intuition is the ultimate standard that determines the accuracy of any proposed grammar, lin-
guistic theory, or operational test, it must be emphasized, once again, that this tacit knowledge
may very well not be immediately available to the user of the language” (Chomsky, 1965, p.21).
4. “... where I is to suggest ‘internalized’ (in the mind/brain) and ‘intensional’ (in that the
procedure is a function of enumerating structural descriptions, considered in intension with a
particular description)” (Chomsky, 2000, p.70). But he adds later, that he also uses “...’I” to sug-
gest ‘internal’, ‘individual’, since this is a strictly internalist, individualist approach to language”
(Chomsky, 2000, p.118) – while I would class my inquiries as ‘internalist’ also, it is in his ‘indi-
vidualist’ orientation that he and I most obviously diverge.
Dialogical dynamics 

and accumulate examples to show, is that ‘tendencies’ are not only descriptions
from without, but that they are among the objects of the stream, which is thus
aware of them from within, and must be described as in very large measure large
measure constituted of feelings of tendency, often so vague that we are unable to
name them at all5 “(p. 246). And, in being aware of them from within, i.e., of the
transitory parts of the inner stream of thought occurring within us, we find that as
they unfold they provided us with both a ‘shaped’ and a ‘vectored’ sense of our
moment-by-moment changing placement in our current surroundings. In short,
we find such responsive feelings as engendering in us both unique anticipations as
to what-next might happen along with, so to speak, ‘action-guiding advisories’ as
to what-next we might do – in Wittgenstein’s (1953) terms, they can provide us
with an immediate sense of how to “go on” in our current, practical circumstances.
Elsewhere (Shotter 2005), I have explored such transitory understandings and ac-
tion guiding anticipations extensively.
But what I must do here, is to note that from a dialogical point of view, it is
(mostly) out in the larger flow of inter-activity taking place between people when-
ever they meet, the flow within which they as individuals are ‘participant parts’,
that the momentary dynamic stabilities of interest to us occur – not within the
‘stream of thought’ hidden inside people’s heads. I say ‘mostly’, as the back and
forth flow of movement in such inter-activity, its ‘rhythm’, is of such a kind that, if
we are being spontaneously responsive to the embodied expressions of others,
then we are, so to speak, ‘resonating in tune with them’. And to the extent that we
are jointly participating in this common rhythm with our whole being, both our
‘inner’ feelings and our ‘outer’ expressions share in, or partake of, it too.
In such a short article as this, I cannot explore at any length the extensive array
of complex issues arising out of our participation in dialogically-structured reali-
ties (but see Shotter 1996, 2003). But I do want to emphasize here one of its most
important consequences: it is only in such meetings – and not in the heads of in-
dividuals – that we can find the starting points for our analyses. For it is in such
meetings that we can find the beginnings of our language games. As Wittgenstein
(1980) puts it: “The origin and the primitive form of the language game is,” he says
(p.31), a reaction; only from this can more complicated forms develop.€Language
– I want to say – is a refinement, ‘in the beginning was the deed’ (quoting Goethe)’.”
“The primitive reaction may have been a glance or a gesture, but it may also have
been a word” (1953:€218). “But what is the word ‘primitive’ meant to say here?” he
asks, “Presumably that this sort of behavior is pre-linguistic: that a language-game

5. “The truth is that large tracts of human speech are nothing but signs of direction in thought,
of which direction we nevertheless have an acute discriminatory sense, though no definite sen-
sorial image plays any part in it whatsoever” (p.244).
 John Shotter

is based on it, that it is the prototype of a way of thinking and not the result of
thought” (1981: no.541). In other words – and this is a point of especial importance
to practitioners – it is in the once-off, fleeting reactions occurring at the beginning
of our meetings with others, that we can find the beginnings of the uniquely new
ways of thinking required if we are to come to a grasp of the particular, concrete,
never-before-encountered person or circumstances now confronting us. These
moments, then, are moments of common reference, shared foundational moments
that can function as shared starting points for the further exploration of the unique
person or circumstance before us.

2. Developing our practices in practice

Indeed, to go further, we clearly do not need to be able to explain our everyday

utterances and actions scientifically, i.e., analyze them into a certain set of ele-
ments that combine in repetitive patterns to produce observed outcomes, to be
able, through everyday reflection and inquiry, to improve them, to gain a more
deliberate command of them. And to make this claim is not to reject the value of
science in our lives. It is simply to note such facts, for instance, that in the course
of their everyday involvements with them, in being spontaneous responsive to
their children’s actions in a living, bodily, expressive manner, parents can (infor-
mally) teach their children, not only their mother tongue, but also countless other
aspects of acceptable and intelligible behavior, without having any idea of the laws
by which their children’s minds and bodies are governed. In other words, at work
here in the spontaneous, living bodily interactions occurring unceasingly between
all of us, not just parents with their children, is another kind of process of under-
standing and of acting expressively, quite different from that at work when we act
deliberately and individually as scientists6 – a process that comes into play in, and
can only come into play, in our living meetings with the others and othernesses
around us, a dialogical form of understanding. Scientific understandings do work
in terms of static, picture-like, inner mental representations, but our everyday,
spontaneous, living understandings do not seem to work in this way. If Bakhtin
(1986) is right, they work in terms of in inner, dialogically-structured movements,
a dialogical dynamics, that gives rise to distinctive ‘movements’ whose shape can
be felt or sensed but not pictured, i.e., not known in a propositional form.

6. “Can only logical analysis explain what we mean by the propositions of ordinary language?
Moore is inclined to think so. Are people therefore ignorant of what they mean when they say
‘Today the sky is clearer than yesterday’? Do we have to wait for logical analysis here? What a
hellish idea!” (Wittgenstein in Waismann, 1979, pp.€129–130).
Dialogical dynamics 

Crucial to this process, is the realization that there are, as Wittgenstein (1953)
puts it, “countless different kinds of use of what we call ‘symbols’, ‘words’ ‘sentences’”
(no.23), and that, besides people’s talk about states of affairs, in which something
is pictured or portrayed, we need also to understand (among its many other uses)
the expressive use of our embodied talk. Indeed, unless we can understand how
others as 1st-persons, as ‘I’s’, manifest or exhibit crucial aspects of their ‘inner’ lives
to us, e.g., their surety and confidence, their uncertainty or humility, their pom-
posity and arrogance, their respect or contempt for us, their anxiety and sadness,
and so on, in the present moment of their acting, we cannot understand how, so to
speak, to ‘relate’ to them.
Tom Andersen (1996), a world-renowned Norwegian family therapist, in
characterizing his therapeutic attitude to his client’s utterances, their words, com-
ments as follows: “The listener who sees as much as he or she hears will notice that
various spoken words ‘touch’ the speaker differently. The speaker is touched by the
words as they reach his or her own ears. Some words touch the speaker in such a way
that the listener can see him or her being moved... one example may clarify this.
A woman had felt sad for a long while related that she could never ask for help,
even when she was sick. Help had to be given by others, not asked for by her. “Be-
cause,” she explained, “independence was a big word in my family. We were supposed
to be independent.” [JS: The voice of her father and mother at work in her – see the
final sentence in this quote below.] A shift in her face and a drop in the voice when
she uttered the word ‘independent’ indicated the meaningfulness of the word. When
she was asked: “If you looked into that word ‘independence’, what might you see?” she
first said that she did not like the word very much. Asked what she saw that she did
not like, she put her hands to her face and said, weeping: “it is so hard for me to talk
about loneliness... yes, it means staying alone.” As she told how hard it had been to
stay alone in order to fulfill all expectations of her being independent, she cried and
her body sank in resignation. She talked for a long while without interruption and
started to wonder if she would be able to fulfill those expectations. Being more and
more eagerly involved in her own discussion, her voice raised, and her neck and
shoulders raised, and she talked more and more angrily as the idea of being-in-the-
world as independent was forcefully challenged.
Asked what her mother would see in the word, she replied that she would see
strength; her father would also see strength, but of another kind. Her sister and
grandmother would also see what she did” (p.212).
In another case, Andersen (1996) brings to light further influences of (other’s)
words on us: “One woman who had been hospitalized at a mental hospital for a year
finally came to family therapy. Besides herself and her family and the family therapist,
the doctor-in-chief at the hospital and her nurse contact at the ward were present.
When she was asked if she had been given any diagnosis, she said: “a manic-depressive
 John Shotter

psychosis.” When she was asked if that diagnosis made any difference, she said it
changed her life. She could no longer laugh and be happy nor be sad and cry, because
she could see on the faces of those around her that they thought she might go manic or
she might become depressed. She therefore had a new inner voice speaking to her all the
time: “Don’t be happy and don’t be sad! Don’t laugh and don’t cry!” (pp.123–124).
As Vygotsky (1962) noted above, words are “the means by which we direct our
mental operations, control their course, and channel them toward the solution of
the problem confronting us” (p.58). But clearly, as both Andersen’s two cases cited
above suggest, and as Wittgenstein suggests throughout his later work, not all the
words we learn from others orient us appropriately, many can also disorient or
mislead us. Thus, as Wittgenstein (1953) puts it, with respect to the kind of ‘prob-
lems’ – or better, difficulties of orientation – we often face (but are often uncon-
scious of) in our own human affairs: “A simile that has been absorbed into the
forms of our language produces a false appearance, and this disquiets us. ‘But this
isn’t how it is!’ we say. ‘Yet this is how it has to be!’” (no.112), i.e., this is how it has
‘to be’ if we are to be intelligible to, and accepted by, those around us as competent
members of our social group.€“‘But this is how it is – ’ I say to myself over and over
again. I feel as though, if only I could fix my gaze absolutely sharply on this fact,
get it in focus, I must grasp the essence of the matter” (no.113). To cure ourselves
of such bewilderments, we require “a battle against the bewitchment of our intel-
ligence by means of language” (no.109). But in this struggle, “there is not a philo-
sophical method [JS - a methodology], though there are indeed methods, like dif-
ferent therapies” (no.133).

3. Methods for exploring how to improve our practices

from within our conduct of the practices

If we cannot work in terms of general theories, in terms of representations, to re-

fine our practices in the unique circumstances of their performance, what can we
do? The stance toward spoken words suggested above – that we think of the ges-
tural meaning of our words-in-their-speaking as ordinarily working to draw our
attention to the existing connections between our utterances and their circum-
stances – suggests that they can also be used extraordinarily, to draw our attention
to how we do in fact make such connections. To repeat yet again Vygotsky’s (1962)
comment that words are “the means by which we direct our mental operations,
control their course, and channel them toward the solution of the problem
confronting us” (p.58), it is clear that we can use certain ways of talking as
Dialogical dynamics 

psychological ‘tools’7, so to speak, as instruments or implements through which to

responsively influence, not only the behavior of others, but our own as well. For
these instructive forms of talk can, in practice – in terms of the feelings of tendency
they can engender in us if we can dialogically engage with them – ‘move’ us to do
something we would not otherwise do. Thus, in ‘gesturing’ or ‘pointing’ toward
something in our circumstances, they can cause us to relate ourselves to our cir-
cumstances in a different way.
This way of talking, however, involves a special form of talk, one that like po-
etic forms of talk uses quite ordinary words (no special technical words denoting
technical concepts) juxtaposed in unusual, and thus ‘striking’ or ‘arresting’ combi-
nations. Hence his remark about his style of philosophical writing: “philosophy
ought really to be written only as a poetic composition” (Wittgenstein, 1980, p.24).
It is the ‘poetic’, ‘gestural’ function of his ‘instructive’ forms of talk that is their key
feature. This is what gives them their ‘life’, their function ‘within’ our lives.
He calls the remarks he uses to draw our attention to what is, in fact, already
know to us, ‘reminders’. For, as he says: “Something that we know when no one
asks us, but no longer know when we are supposed to give an account of it
[cf.€Augustine], is something we need to remind ourselves of ” (Wittgenstein, 1953,
no.89). We are bringing to light what we are already doing spontaneously, in order
later to do it deliberately.8
The ‘poetic methods’ he uses in his own writings work, first: (1) To arrest or
interrupt (or ‘deconstruct’) the spontaneous, unconscious flow of our ongoing ac-
tivity, and to give “prominence to distinctions which our ordinary forms of lan-
guage easily make us overlook” (1953: no.132). Just as a mother might say to her
young son: “Stop, look, see the dirty footmarks you’ve just made... wipe you feet
next time you come it,” so his remarks draw our attention to unnoticed, but pos-
sibly to further socially significant features in our actions.
(2) Then, we can note that is his talk is full of such expressions as “Think of...,”
“Imagine...,” “It is like...,” “So one might say...,” “Suppose...,” and so on, all designed
“to draw someone’s attention to the fact that he [or she] is capable of imagining
[something]” (1953, no.144). They show us other possibilities present in a circum-
stance, where, in imagining something new, a person is “now... inclined to regard
a given case differently: that is, to compare it with this rather than that set of pic-
tures. I have changed his way of looking at things” (Wittgenstein 1953: no.144), he
says. In other words, prospectively, he draws our attention to the fact that, from

7. But Mead (1934) also says, regarding gestures: “They became the tools through which the
other forms responded” (p.44, my emphasis).
8. See Vygotsky’s (1962) account of one of the “basic laws governing human development”
quoted at the beginning of this article.
 John Shotter

our position of involvement in things, there are always other possibilities available
to us, other possible ways of ‘going on’.9
This suggests to us a third method that is sometimes important: (3) By the
careful use of selected images, similes, analogies, metaphors, or ‘pictures’, he also
suggests new ways of talking, new idioms, that not only orient us toward sensing
otherwise unnoticed distinctions and relations for the first time, but which also
suggest new connections and relations with the rest of our proceedings. This is
closely connected with a fourth: (4) By the use of various kinds of objects of com-
parison, e.g., other possible ways of talking, other “language games” both actual
and invented, etc., he tries “to throw light on the facts of our language by way of
not only similarities, but also dissimilarities” (1953: no.130). For, by noticing how
what occurs differs in a distinctive way from what we otherwise would expect,
such comparisons can work, he notes, to establish “an order in our knowledge of
the use of language: an order with a particular end in view; one of many possible
orders; not the order” (1953: no.132) – where again, the goal is to achieve a kind of
understanding that is useful in the ‘going on’ of a practice, the overcoming of a
‘disorientation’, of ‘not knowing one’s way about’.10
Such images, similes, and metaphors, etc., cannot represent any already fixed or-
ders in our use of language, for, by their very nature, in being open to determination
only in the context of their occurrence, they do not belong to any such orders. But what
such invented concepts can do for us – in artificially creating a fixed order where none
before existed – is to make aspects of our situated use of language publicly discussable
and accountable. They provide a practical resource: a way of talking that works to draw
our attention, in different ways in different contexts, to what otherwise we would not
know how to attend. Other ways of talking, other relational stances or style (i.e., orien-
tations), will function to bring out other connections. One can imagine many different
aims. Though what is at stake in them all, is not so much the grasp by isolated indi-
viduals, of an inner ‘mental picture’ of a state of affairs, but a grasp of the actual, practi-
cal connections between aspects of our own communicative activities - influences that
are present and at work in ‘shaping’ what we say in a particular circumstance.

9. Again, taking a practical stance toward the problems we face in human affairs, Wittgenstein
(1953) suggests that “a philosophical problem has the form: ‘I don’t know my way about’”
(no.123). In practice, to understand something simply means being able to continue appropri-
ately out in the practice, in the judgment of others, irrespective of what might be happening in
one’s head as an individual.
10. But note, the order produced is not the order, but merely an order. It is the imposing of a
single order of connectedness onto ongoing, complex, multi-dimensional, still developing, hu-
man phenomena, in the service of achieving a final explanation, that renders their living, dy-
namic nature rationally-invisible.
Dialogical dynamics 

These methods all contribute to the achievement of what he calls a “perspicu-

ous representation” (German: übersichtliche Darstellung).” (5) It is as if the task is
like the task we face in coming to feel ‘at home’ in a new town or city, so that we
‘know our way around’ within it without getting lost. Thus, as he puts it: “A main
source of our failure to understand is that we do not command a clear view of our
use of words. – Our grammar is lacking in this sort of perspicuity. A perspicuous
representation produces just that understanding which consists in ‘seeing connec-
tions’,” he says (Wittgenstein 1953: no.122). Where the kind of understanding he is
after here is a practical understanding, an understanding that allows one to ‘go on’
in practice with one’s activities. In other words, a “perspicuous presentation” is a
presentation in which “problems are solved, not by giving new information, but by
arranging what we have always known” (Wittgenstein 1953: no.109) – so as we can
move easily from one part of a landscape of possibilities to another. Thus, if we are
‘to find our way about’ inside our own linguistic forms of life, we need to grasp, to
sense, their inner ‘landscape’, their ‘grammatical geographies’, so to speak.
We might call all the methods above, ‘positive’ methods, in that they can per-
haps lead us into new ways of acting that are a refinement or elaboration of our old
ways. But some of his methods are, so to speak, ‘negative’ in that they are aimed at
preventing us from ever again taking certain paths. (6) Thus, with respect to our
temptation to look for hidden inner mental processes, he remarks: “Try not to
think of understanding as a ‘mental process’ at all. – For that is the expression
which confuses you,” he says (Wittgenstein 1953). “But ask yourself: in what sort
of case, in which circumstances, do we say, ‘Now I know how to go on’...” (no.154).
For, if it is our task ‘move about’ anywhere within a particular landscape of possi-
bilities, anywhere where there might be new connections, new relationships, etc.,
to be made. For the task is not to map the old already established paths, but to trail
blaze new ones.
Lacking the appropriate sensibility to notice these relational phenomena, we
have felt in the past that th