Sunteți pe pagina 1din 328

Discourses in 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 203
Discourses in Interaction
Edited by Sanna-Kaisa Tanskanen, Marja-Liisa Helasvuo, Marjut Johansson
and Mia Raitaniemi
Discourses in Interaction

Edited by

Sanna-Kaisa Tanskanen
University of Helsinki

Marja-Liisa Helasvuo
Marjut Johansson
Mia Raitaniemi
University of Turku

John Benjamins Publishing Company

Amsterdam / Philadelphia
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

Discourses in interaction / edited by Sanna-Kaisa Tanskanen... [et al.].

p. cm. (Pragmatics & Beyond New Series, issn 0922-842X ; v. 203)
Includes bibliographical references and index.
1.  Discourse analysis.  I. Tanskanen, Sanna-Kaisa.
P302.D5675   2010
401’.41--dc22 2010038020
isbn 978 90 272 5607 2 (Hb ; alk. paper)
isbn 978 90 272 8749 6 (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

Acknowledgments vii

Discourse and the interactional turn 1

Marja-Liisa Helasvuo, Marjut Johansson and Sanna-Kaisa Tanskanen

part 1.  Dialogues between contexts

Contexts in context: Micro meets macro 13

Anita Fetzer

Communicative activity types as organisations in discourses

and discourses in organisations 33
Per Linell

Discourse and context in a historical perspective:

On courtroom interaction in Salem, 1692 61
Risto Hiltunen

part 2.  Constructing identity across genres

Pronominal choice in French conversational interaction:

Indices of national identity in identity acts 81
Linda R. Waugh

Constructing interpersonal relations in the discourse of Russian media 101

Marjatta Vanhala-Aniszewski

Who communicates in the media supported by the Russian Church? 115

Lea Siilin

“O England! England! She says – my Father – my Sisters – my friends! –

shall I ever see you more?”: Reporting in 18th-century correspondence 133
Minna Nevala and Minna Palander-Collin
 Discourses in Interaction

part 3.  Managing interpersonal relations

Power in Early Modern English courtroom discourse 153

Barbara Kryk-Kastovsky

“I desire to have some tyme to consider of it”: A pragmaphilological

approach to refusals and refutations in Modern-English trials* 173
Ana E. Martínez-Insua

Interactive aspects of computer-mediated communication:

‘Disagreement’ in an English and a German public news group 195
Sonja Kleinke

‘A little story, for food for thought.......’: Narratives in advice discourse 223
Loukia Lindholm

part 4.  Structures in interaction

Appropriateness in interpersonal communication 239

Maria Sivenkova

Filling the German vorfeld in written and spoken discourse 263

Augustin Speyer

Phatic expressions in French and German telephone conversations 291

Anja Smith

Index 313

We would like to thank the following people with whom we had the pleasure of inter-
acting while producing this collection.
First of all, we want to express our gratitude to all the contributors for their excel-
lent cooperation throughout the publication process.
We gratefully acknowledge the professional cooperation by everyone at John
Benjamins Publishing Company, from the acquisition editor to the series editor. The
anonymous reviewers deserve thanks for their thorough comments on the manuscript.
Finally, we thank Ellen Valle for language checking, Mari-Liisa Varila for desktop
editing and Ira Hansen for secretarial assistance.
The Editors
Discourse and the interactional turn

Marja-Liisa Helasvuo, Marjut Johansson

and Sanna-Kaisa Tanskanen

1. From the linguistic turn to the interactional turn

The focus of this volume is on discourses in interaction. All the contributions investigate
language use from interactional perspectives, although the name given to the object stud-
ied ranges from spoken language and conversation to interaction, dialogue, discourse
and communication. Coming from different approaches in pragmatics, interactional lin-
guistics, conversation analysis, discourse analysis and dialogue analysis, the articles nev-
ertheless share a similar starting point: they all represent a form of linguistic understand-
ing which has emerged within what we would like to call the interactional turn.
The humanities have witnessed several major turns in the past decades, including
the linguistic turn, characterising analytical philosophy in the twentieth century, and
the narrative turn, which grew out of French narrative theories from the 1960s on-
wards. Both of these turns have been adopted in various fields of the humanities to
become part of mainstream approaches in various disciplines, also allowing interdisci-
plinary considerations. Today, rather than a single turn there are several turns taking
place, including the hermeneutic, cultural and cognitive. This raises the following ques-
tions: What does a ‘turn’ actually indicate? What does the interactional turn mean?
And how is it related to the dialogical turn (see Linell 1997 and this volume)?
From a broader perspective, a turn can be regarded as an epistemological shift
taking place in response or contrast to a previous phase within a field of scientific
knowledge. A new turn implies a new understanding of the object of study, and how it
is conceptualised and explained. In the philosophy of science, a change, or turn, is
neither a normative nor an axiological concept; it is a descriptive one, referring to the
idea of progress in science (cf. Niiniluoto 2007). Put another way, according to Foucault
(1969), change as a discursive formation is a situation in which new knowledge about
a certain object emerges, giving rise to a group of statements which describe this object
in a novel and seemingly coherent manner. This, however, does not imply the existence
of a unified field of knowledge; inevitably, according to Foucault (1969), there are dis-
continuities and ruptures as well. The germs that lead to a shift in thinking or to a new
episteme are due to mutations and alterations in contextual conditions.
 Marja-Liisa Helasvuo, Marjut Johansson and Sanna-Kaisa Tanskanen

From a linguistic perspective, we can ask how a shift takes place in a certain do-
main of linguistics. Providing an exhaustive answer to this question is not possible
here; even defining the starting point for the interactional turn is not feasible. How-
ever, during the past half century a number of seminal works have been published,
paving the way for the interactional turn. In the late 1960s, the Palo Alto Group pro-
posed the premise that all behaviour is communication (Watzlawick, Beavin &
Jackson 1967). One of the main axioms of communication, the metacommunicational
axiom, postulates that one cannot not communicate (id.: 51). In the work of the Palo
Alto Group, communication is not simply approached from a theoretical perspective;
the analysis involves different types of interactional patterns. Other approaches sug-
gesting notions central to interactional analysis are frame analysis and the perfor-
mance of self by Goffman (1959, 1974) as well as the ethnography of communication
proposed by Gumperz and Hymes (1964). Yet another line of thinking is that provided
by discourse analysis, including studies of ideology (see van Dijk 1981) and critical
discourse analysis (see Fairclough 1992).
In his account of dialogism, Linell (1997) examines approaches to discourse, ac-
tion and cognition which take interactional and contextual features as their starting
point. He lists several historical roots and scientific fields which have contributed to
interactional approaches; in addition to those discussed above, he mentions social psy-
chology, socio-cultural semiotics, cultural psychology and dialogue analysis. His view
on the dialogical turn is closely related to the interactional turn.
Within discourse-functional linguistics, there has been an interest in the kinds of
linguistic resources used for various discourse tasks: for introducing new topics into
discourse and for establishing topic continuity (see e.g. Givón 1979, 1983), as well as
for managing referent introduction and tracking (see e.g. Chafe 1980, 1994). Research
on Preferred Argument Structure (Du Bois 1985, 1987) has investigated the discourse
motivation of morphosyntactic alignment patterns in expressing argument relations,
and has shown that they are motivated by discourse strategies concerning the manage-
ment of given and new information (see Du Bois et al. 2003). These studies can all be
seen as attempts not merely to describe the use of grammatical structures in their dis-
course contexts (e.g. the linguistic resources used to introduce new referents into dis-
course), but also to find motivations for discourse patternings (e.g. the tendency to
introduce new referents in certain syntactic roles rather than others).
In functional linguistics, grammar is seen as a crystallisation of recurrent patterns
of discourse (see esp. Hopper 1987, 1988, Givón 1979, Chafe 1994). As Couper-Kuhlen
and Selting (2001: 2) put it, linguistic forms are seen as “something ‘to do things with’
on situated occasions of use”; grammar provides resources for pursuing conversational
actions and activities, and is in turn shaped by these activities. Functional linguistics,
however, does not have a fully developed methodology for analyzing conversational
activities; thus there has been a growing interest in supplementing the grammatical
analysis of functional linguistics with the methodological tools of the sociological tra-
dition of conversation analysis (see e.g. Fox 1987, Ono & Thompson 1995). This has
Discourse and the interactional turn 

eventually led to the emergence of interactional linguistics as a new field of linguistics

around the turn of the millennium (see especially Selting & Couper-Kuhlen 2001).
Interactional linguistics can be characterised as a field of linguistics that combines
the traditions of conversation analysis and linguistic anthropology with discourse-
functional linguistics (see Couper-Kuhlen & Selting 2001: 1). Early studies in the con-
versation-analytic tradition (see e.g. Sacks, Schegloff & Jefferson 1974, Sacks 1992)
emphasised the role of everyday conversation as the primordial site of social order.
Conversation analysis has developed tools for the careful microanalysis of linguistic
elements in their conversational contexts, and methods for analysing not only how
speakers use linguistic structures but also how the co-participants orient themselves
toward the linguistic structures used.
An illustrative example is the idea of a referent as being “given” or “new”. In func-
tional linguistics this is mostly a subjective notion, based on the perspective of the
current speaker. For example Chafe (1994: 74) defines “new” information as some-
thing that the speaker assumes was previously inactive in the short-term memory of
the listener, and that s/he thus verbalises as new information. Conversation-analytic
work has shown how new information can also be examined from the perspective of
the co-participants, by focusing on the ways in which they show their orientation to
previous talk as “new information” by their use of response particles (see Sorjonen
2001). Thus what is “new” in discourse is negotiated between the co-participants. Dis-
course notions such as “given”, “new” or “identifiable” are intersubjective in nature.
Interactional linguistics, as well as other approaches to interaction such as socio-
pragmatics, dialogue analysis and the like, focus on a wide range of different types of
interaction. Since the early days of conversation analysis, when the focus was placed
on everyday conversations, interest has shifted towards the whole variety of interac-
tions, private or public, dyadic or polyadic, written or spoken, present-day or historical
ones. In this volume, discourse in interaction is understood as all forms of meaningful
semiotic human activity which are considered “in connection with social, cultural and
historical patterns and developments of use” (Blommaert 2005: 3). This will be further
commented on in the following sections.
Let us now move on to a presentation of the individual contributions that make
up this volume. The papers are divided into four sections: dialogues between contexts,
constructing identity across genres, managing interpersonal relations and structures in

2. Discourses in interaction

2.1 Dialogues between contexts

Our understanding of genre and the construction of identities through discourse has
changed. Above all, they are contextual in nature. Language use is indexical: in other
 Marja-Liisa Helasvuo, Marjut Johansson and Sanna-Kaisa Tanskanen

words, the meaning of language actions is situated in the context of use (Silverstein
2003). The contexts anchor the discursive actions in social and temporal settings; the
plurality of contexts can be seen as contextual resources (Linell 1997: 127–158) or as
different types of context – cognitive, linguistic, social and sociocultural (Fetzer 2004).
In her contribution to the present volume, Anita Fetzer argues that context can no
longer be seen as an analytic prime. Rather, it is analysed as a product of language use,
negotiated and interactionally constructed, anchored to intentionality.
Genre, an activity type or a communicative activity type – as Linell defines it in this
volume – is one of the main notions of an interactional approach which takes indexical
and dialogical premises as its starting point. According to Levinson (1992), an activity
type has a specific goal, it has certain constraints as to verbal contributions, and it de-
fines a type of frame for a situation. Within an activity type, the agent organizes her
action under different modalities. An activity type can be seen to have a goal that the
discourse community considers as its function (cf. Levinson 1992). In his article, Per
Linell outlines a ‘dialogical’ perspective on discourse, more precisely on spoken, inter-
actional discourse, which focuses on interaction and contexts. Central notions in a
dialogical account of talk-in-interaction are those of ‘communicative project’ and
‘communicative activity type’.
Risto Hiltunen deals with contextualisation from the perspective of historical dis-
course linguistics, using the Salem witchcraft documents as data. Hiltunen argues that
reconstructing the interplay between context and discourse is a prerequisite for under-
standing the strategic choices involved. Assessing the issue in the light of the Salem
trials involves examining the socio-historical, cultural, legal, scribal and communica-
tive contexts in terms of both the community and the individual.

2.2 Constructing identity across genres

In research on identity the widely accepted view of identity is not as a global, static
concept, possessing an ‘essence’, but rather as a dynamic entity, which is at the same
time both partial and multiple. Generally, identity is based on the relationship of the
agent with the other. There is, however, no one identity but a number of shifting,
changing identities, depending on the contexts and genres within which they are con-
structed; they involve a semiotic process of representation and are situated in relation
to different layers of ‘groupness’ and ‘categories’ (Blommaert 2005: 203–204).
Identity is approached in three articles, in different cultural contexts. Linda R.
Waugh’s contribution addresses the interactional co-construction of identity in infor-
mal, natural conversation in French, and shows how the construction of identity is
bound up with the co-construction of ideology. Particular attention is paid to the use
of indefinite, non-specific, generic pronouns in generic statements in one conversa-
tion, where pronouns act as indices of linguistics and national identity and at the same
time markers as of ideological stance.
Discourse and the interactional turn 

The section continues with two papers on Russian media discourse. Marjatta
Vanhala-Aniszewski investigates the relations between participants in an interaction
actualized in Russian media discourse. She discusses, first of all, the role and identity
of writer and reader, secondly, the force of the writer’s statements, which are meant to
influence the audience; in other words, his personal attitude toward the truth-value of
the propositions contained in the text. The investigation of Russian media texts shows
that news texts include quite a few markers indicating the explicit presence of interper-
sonal relations between the participants in the communicative act.
Lea Siilin’s material is a new church medium, established by the Russian Orthodox
Church in order to attract a new audience, including young people. The article ex-
plores the kinds of interaction typical of such media, and how interactional relations
are constructed between representatives of the Church and ordinary readers of Church
media. The results indicate that the texts analyzed partly demonstrate an institutional
discourse; the interaction therefore illustrates the difference between authoritative and
individual opinions.
To conclude this section, Minna Nevala and Minna Palander-Collin’s contribu-
tion focuses on reporting in a sample of eighteenth-century letters written by Fanny
Burney to various recipients, with the purpose of studying the occurrence of reporting
from a socio-pragmatic perspective and understanding the function of reporting in
the communicative situations in which it occurs. Their analysis pays attention to the
mutual relationship of the reporter and the recipient of the report, the subject matter
of the report, the identity of the person whose speech is reported, the form of the re-
porting frame and the reporter’s attitude towards the report or reported person.

2.3 Managing interpersonal relations

Since the early 1980s linguistics has seen a broadening of perspectives, in terms both
of methodological tools and of theoretical apparatus. There has been a shift in focus,
from a strict division between speaker and hearer to a flexible participant framework
where the co-participants take different roles, and from viewing language as an expres-
sion of the speaker’s subjective viewpoint to intersubjectivity. At the same time there
has also been a shift from looking at linguistic expressions as products, as static enti-
ties, to identifying the underlying activities and the linguistic resources utilized in
such activities (see e.g. Linell 1982, Chapter 6 on our conception of a sentence). The
analysis of text types, for example, previously tended to focus primarily on the func-
tions of a text as a product, a static entity, whereas the more recent discussion takes a
more processual approach to texts as instances of meaning-making activities. Texts are
also seen as manifesting the construction of interpersonal relationship between the
participants in the interaction and the type of cognitive and discursive activities that
the speakers actualize in the speech situation. Reflecting this shift in perspective, the
present volume includes four papers which either focus on discursive activities or
 Marja-Liisa Helasvuo, Marjut Johansson and Sanna-Kaisa Tanskanen

contribute to the discussion of theoretical and methodological issues in the analysis of

interpersonal relationships.
Barbara Kryk-Kastovsky’s paper on power in Early Modern English courtroom
discourse is a contribution to diachronic pragmatics. One of the issues addressed is the
applicability of the analytical tools employed by synchronic pragmatics to diachronic
pragmatic studies, which is by no means uncontroversial. Kryk-Kastovsky’s paper is an
attempt to apply the notion of power to the analysis of diachronic data, concentrating
on the ways in which power was exercised by the interrogators upon the interrogated
in the early modern English courtroom.
Ana E. Martínez-Insua discusses instances of refusing to answer and refuting al-
legations in Early Modern English trials. The purpose of the paper is to detect and
analyse the ways in which participants in trials use non-performative speech acts to
refute allegations and deny involvement in crimes. The data come from a corpus of
trial proceedings from the period 1560–1760.
The remaining two articles in this section deal with computer-mediated interac-
tion. Sonja Kleinke approaches the functions and use of ‘disagreement’ in her cross-
cultural study of interaction in public news groups in English and German. She ana-
lyzes the specific frame conditions prevailing in both groups, including the discourse
type ‘discussion’, the development of the social links emerging between the interac-
tants during the interaction, and (dis)agree­ment among the participants.
Loukia Lindholm’s paper focuses on the functions of narratives in advice-giving
messages in the context of computer-mediated discourse. Narratives in advice dis-
course have mostly been discussed as an advice-seeking strategy related to the state-
ment of a problem. However, narratives have not been fully explored in other aspects
of advice discourse, such as advice giving. The analysis shows how narratives function
as advice, alignment to advisees, and arguments for the advice given.

2.4 Structures in interaction

Vital to all studies of discourse are questions of how utterances are connected to each
other, what makes an utterance appropriate in a certain context, how it orients itself in
relation to the previous discourse, and how it paves the way for what follows. This has
been modelled in different ways, inter alia in terms of sequential organization (CA)
and communicative moves.
Maria Sivenkova’s paper investigates two types of communicative move in which
interlocutors negotiate the appropriateness of an utterance: prospective moves, in
which the speaker prepares the ground for his/her future speech act by checking on
the utterance’s appropriateness, and retrospective moves, in which the addressee sig-
nals to the speaker that his/her previous utterance is viewed as inappropriate. The re-
search is based on dialogues taken from contemporary drama and fiction in English,
French and Russian.
Discourse and the interactional turn 

Augustin Speyer discusses the filling of the sentence-initial position (vorfeld) in

German in terms of discourse-structural considerations. Several types of element
compete for this position, the distribution of which can be modelled by Stochastic
Optimality Theory. In dialogic discourse other considerations play a role as well, such
as the explicit marking of origo and rhetorical relation.
Anja Smith explores “phatic expressions” within the theoretical frameworks of
ethnomethodology, conversation analysis and speech act theory in her analysis of ev-
eryday telephone conversations in German and French. Her research compares the
ways in which German “ne” and French “hein” contribute to the organization of inter-
action on both a “procedural” and a socio-emotional level.

3. Concluding remarks

The foundations of this volume were laid at the Organization in Discourse 3: The Inter-
actional Perspective conference, held in Turku in August 2006. The fourteen papers in
the volume are revised versions of presentations given at the conference. What we be-
lieve was unique to the conference, and similarly characterises the present collection
of papers, is the great variety in the types of interaction analysed. Although the starting
points, goals and aims of studies of such diverse materials naturally differ, engaging the
studies in a dialogue can help reveal where old beliefs may be challenged and new
understandings may emerge.
Orienting herself towards a novel way of understanding a field of study, Kerbrat-
Orecchioni (2005: 14) notes that a discipline is defined by the way it conceptualises its
object of study, rather than by the approach it adopts in studying this object. She also
refers to particular forms of discourse as various types of discursive practices, taking
place in interactional contexts. The present volume adopts a similar position: discourse
is defined as an interactional activity, a meaning-making social activity, which takes
place within contexts and between agents who have goals within this activity.
To conclude: the title of this volume can be understood in two different but inter-
connected ways. First of all, according to the above definition of discourse, it refers to
the multiple ways in which the construction of meaning takes place in different types
of interaction. Secondly, it refers to a larger concept of discourse, as part of a discursive
formation, and to the ways in which the object of study is conceptualised. Within in-
teractional approaches to discourse, there are thus several interconnected, parallel, but
also varying and even divergent discourses on interaction.
 Marja-Liisa Helasvuo, Marjut Johansson and Sanna-Kaisa Tanskanen


Blommaert, Jan 2005. Discourse. Cambridge: Cambridge University Press.

Chafe, Wallace L. 1980. “The deployment of consciousness in the production of a narrative.” In
The Pear Stories: Cognitive, Cultural, and Linguistic Aspects of Narrative Production, Wallace
L. Chafe (ed.), 7–40. Norwood, N.J.: Ablex.
Chafe, Wallace L. 1994. Discourse, Consciousness, and Time: The Flow and Displacement of Con-
scious Experience in Speaking and Writing. Chicago: The University of Chicago Press.
Couper-Kuhlen, Elizabeth & Margret Selting 2001. “Introducing interactional linguistics.” In
Studies in Interactional Linguistics, Margret Selting & Elizabeth Couper-Kuhlen (eds), 1–22.
Amsterdam & Philadelphia: John Benjamins.
Dijk, Teun van 1981. Prejudice in Discourse. Amsterdam & Philadelphia: John Benjamins.
Dijk, Teun van 1977. Text and Context: Explorations in the Semantics and Pragmatics of Dis-
course. London: Longman.
Du Bois, John W. 1985. “Competing motivations.” In Iconicity in Syntax, John Haiman (ed.),
343–365. Amsterdam & Philadelphia: John Benjamins.
Du Bois, John W. 1987. “The discourse basis of ergativity.” Language 63: 805–855.
Du Bois, John W., Lorraine E. Kumpf & William J. Ashby (eds) 2003. Preferred Argument Struc-
ture: Grammar as Architecture for Function. Amsterdam: John Benjamins.
Fairclough, Norman 1992. Discourse and Social Change. Cambridge: Polity Press.
Fetzer, Anita 2004. Recontextualizing Context. Amsterdam & Philadelphia: John Benjamins.
Foucault, Michel 1969. L’Archéologie du savoir. Paris: Gallimard.
Fox, Barbara 1987. Anaphora and the Structure of Discourse. Cambridge: Cambridge University
Givón, Talmy 1979. “From discourse to syntax: Grammar as a processing strategy.” In Discourse
and Syntax: Syntax and Semantics, vol. 12, Talmy Givón (ed.). New York: Academic Press.
Givón, Talmy 1983. Topic Continuity in Discourse: A Quantitative Cross-language Study.
Amsterdam & Philadelphia: John Benjamins.
Goffman, Ervin 1974. Frame Analysis. An Essay on the Organization of Experience. Cambridge,
Mass.: Harvard University Press.
Goffman, Ervin 1959. The Presentation of Self in Everyday Life. New York: Anchor Books.
Gumperz, John & Dell Hymes (eds). 1964. “The Ethnography of Communication.” American
Anthropologist 66:6, Part 2.
Hopper, Paul J. 1987. “Emergent grammar.” Proceedings of the Thirteenth Annual Meeting, Berke-
ley Linguistic Society, 139–157. Berkeley, CA: Berkeley Linguistics Society.
Hopper, Paul J. 1988. “Emergent grammar and the a priori grammar postulate.” In Linguistics in
Context, Deborah Tannen (ed.), 117–134. Norwood, NJ: Ablex.
Kerbrat-Orecchioni, Catherine 2005. Le discours en interaction. Paris: Armand Colin.
Linell, Per 1997. Approaching Dialogue: Talk, Interaction and Contexts in Dialogical Perspectives.
Amsterdam & Philadelphia: John Benjamins.
Niiniluoto, Ilkka 2007. “Scientific Progress.” In Stanford Encyclopedia of Philosophy. http://plato. Accessed 10 Jan 2009.
Ono, Tsuyoshi & Sandra A. Thompson 1995. “What can conversation tell us about syntax?” In
Alternative Linguistics: Descriptive and Theoretical Modes [Current Issues in Linguistic The-
ory, 102], Philip W. Davis (ed.), 213–271. Amsterdam & Philadelphia: John Benjamins.
Discourse and the interactional turn 

Sacks, Harvey 1992. Lectures on Conversation vol. 1 & 2. Edited by Gail Jefferson. Cambridge:
Sacks, Harvey, Emanuel A. Schegloff & Gail Jefferson 1974. “A simplest systematics for turn-
taking in conversation.” Language 50: 696–735.
Selting, Margret & Elizabeth Couper-Kuhlen (eds) 2001. Studies in Interactional Linguistics.
Amsterdam & Philadelphia: John Benjamins.
Silverstein, Michael 2003. “Indexical order and the dialectics of sociolinguistic life.” Language &
Communication 23(3–4), 193–229.
Sorjonen, Marja-Leena 2001. Responding in Conversation. Amsterdam & Philadelphia: John
Watzlawick, Paul, Janet H. Beavin, & D. Jackson 1967. Pragmatics of Human Communication: A
Study of Interactional Patterns, Pathologies, and Paradoxes. New York: W.W. Norton & Co.
part 1

Dialogues between contexts

Contexts in context
Micro meets macro

Anita Fetzer

Adopting an integrated approach, the contribution demonstrates that context

can no longer be seen as an analytic prime. Rather than being looked upon as
an external constraint on linguistic performance, it is analyzed as a product of
language use, as interactionally constructed and as negotiated. The first part
categorizes context as social context, sociocultural context, linguistic context
and cognitive context, and the second part examines the dynamics of context
captured through a relational conception based on the premises of indexicality
and intentionality.

1. Introduction

Context has become a major field of research in different research communities and re-
search contexts, for instance in information technology, engineering and science, in the
social sciences, and in arts and humanities. Not only is the concept as such the object of
investigation but so is its application to diverse domains, such as natural- and non-natu-
ral language communication, computer-mediated interaction and information technol-
ogy, robotics, social-action based analyses or literary analyses. The heterogeneous nature
of context and the context-dependence of the concept itself has made it almost impos-
sible for the scientific community to agree upon one commonly shared definition or one
commonly shared accepted theoretical perspective, and frequently, only a minute aspect
of context is described, analyzed or formalized (cf. the interdisciplinary conferences on
context: Akman et al. 2001, Blackburn et al. 2003, Bouquet et al. 1999).
A similar tendency is manifest in the research paradigm of discourse analysis
where discourse is investigated as a theoretical construct (e.g., Brandom 1994, Foucault
1997, Linell 1998 and this volume) and as its manifestation in different types of dis-
course in the social sciences, and in arts and humanities, for instance.
Context- and discourse-based research extends its frame of reference by looking
beyond the actual object of investigation thus integrating relevant background infor-
mation regarding production and reception on the one hand, and order and power
structure on the other. Moreover, both research paradigms base their investigation on
 Anita Fetzer

the premise that their ‘object’ displays an inherent structure. And it is this structure, its
organization and content which are identified, described and analyzed in the respec-
tive research paradigms. Unlike the primarily synchrony-based examination of con-
text, discourse analysis tends to accommodate both diachronic (cf. Hiltunen, this vol-
ume) and synchronic perspectives stressing the fact that discourses are historically
grown. It needs to be pointed out, however, that discourse and context are not neces-
sarily conceived as two mutually exclusive, separate entities. Rather, discourse is fre-
quently considered to be contained in context, while context is seen as an unbounded
entity embedding discourse. Alternatively, context is seen as being presupposed or
indexically contained in discourse, and it is the bounded entity of discourse which
instantiates the reconstruction of context.
Context and discourse are frequently examined from a parts-whole perspective.
On the one hand, context and discourse are conceived as wholes and are analyzed ac-
cordingly. On the other hand, context and discourse are conceived as parts-whole con-
figurations, and it is the constitutive parts of that configuration which are identified,
described and analyzed. As a consequence of the dual perspective, an investigation of
context and discourse requires the accommodation of macro- and micro-oriented
viewpoints. While the former employs a top-down perspective accounting for the ‘ob-
ject as a whole’, that is institutional context or discourse genre, the latter employs a
bottom-up perspective accounting for the ‘object’s constitutive parts’, that is the im-
mediate context or the immediate associations and collocations of an object of dis-
course. Frequently, a meso layer is added to refine the analysis of the micro-macro
interface by making the bridging operations between the different layers more trans-
parent. In context-based studies the bounded concept of a speech event (Hymes 1974),
communicative project and communicative activity type (Linell 1998) can be assigned
the status of a meso category, and in discourse analysis the bounded concept of an
episode may account for the bridging function between a micro discursive object and
a macro discourse genre.
Because of their multifaceted nature and complexity, context and discourse can no
longer be looked upon as analytic primes but rather need to be seen as a multilayered
parts-whole configuration. That is, context contains sub-contexts and discourse con-
tains sub-discourses, and sub-contexts instantiate context and sub-discourses instanti-
ate discourse. To capture the interactive and dynamic nature of context and discourse,
an integrated approach informed by linguistics, psychology, sociology, linguistic an-
thropology, and cultural studies is required. Only then is it possible to cross and tran-
scend disciplinary boundaries and account for the inherently unbounded theoretical
constructs, which may become bounded when instantiated.
The focus of this contribution does not lie on the examination of discourse, but on
context. As the two constructs are intrinsically interwoven, however, discourse is men-
tioned whenever it is considered to be required for the current purpose. The paper is
organized as follows: the following part, context and contexts, examines context from a
top-down perspective and distinguishes between four types of context: linguistic
Contexts in context 

context, cognitive context, social context and sociocultural context. The third part,
context, contextualization and contextualization cues, concludes by looking at context
from a bottom-up perspective anchored to the pragmatic premises of indexicality and
intentionality of communicative action.

2. Context and contexts

The theoretical construct of context has been described not only as a multifaceted, but
also as a multilayered phenomenon. This outlook contains a number of different per-
spectives, which are examined in the following.
First, context is conceived as a frame whose job it is to frame content by delimiting
that content. The former operation assigns content the status of a figure and the latter
assigns the context surrounding the figure the status of a ground. At the same time the
delimited context of the figure is being framed and delimited by less immediately ad-
jacent contextual frames (or the ground). The nature of the connectedness between
those frames is a structured whole which is composed of interconnected frames
(Goffman 1986). The gestalt-psychological figure-ground scenario prevails in psycho-
logical and psycholinguistic perspectives on context. It has also been adopted to cogni-
tive pragmatics as is reflected in the relevance-theoretic conception of context as an
onion, metaphorically speaking. Sperber and Wilson not only point out the intercon-
nected nature of the layers but also stress the fact that their order of inclusion corre-
sponds to their order of accessibility (Sperber and Wilson 1986). This is of particular
importance to the cognitive operations of inferencing and to the calculation of impli-
catures, which are key operations in natural-language communication.
Second, context is seen as a dynamic construct which is interactionally organized
in and through the process of communication. This view prevails in ethnomethodology
and ethnomethodological conversation analysis (Garfinkel 1994, Goodwin and Duranti
1992, Heritage 1984, Schegloff 1992), interactional sociolinguistics (Gumperz
1996, 2003) and sociopragmatics (Fetzer 1999, 2004), where context is assigned the
dual status of process and product. The dynamic outlook is based on the premises of
indexicality of social action and on (joint) construction of a common context; that is,
meaning is not conceived of as autonomous but rather as relational, considering the
embeddedness and interdependence of linguistic expressions. To use Ochs’ words,
“(1) language systematically varies across social contexts, and (2) such variation is part
of the meaning indexed by linguistic structures. (...) The meanings so indexed are re-
ferred to as social meanings, in contrast to purely referential or logical meanings ex-
pressed by linguistic structures” (Ochs 1992: 337–338). Consequently, a thorough ex-
amination of context needs to go beyond the prevailing definition of context as a set of
propositions (Stalnaker 1999). To capture the indexicality of social action and the rela-
tional nature of social meaning, context needs to be conceived of as a complex dynamic
network, which undergoes a permanent process of structuring and re-structuring.
 Anita Fetzer

In those qualitatively-oriented paradigms context is intrinsically connected with

the concepts of adjacency pair, conditional relevance and turn-taking on the micro
level, with activity type (Levinson 1979) or speech event (Hymes 1974) on the meso
level, and with institutional talk on the macro level, whose order is captured through
context-independent and context-sensitive constraints and requirements. In institu-
tional interviews, for instance, there is a clear-cut division of labor anchored to the
turn-taking system and to the adjacency pair question/answer: the representative of
the institution has the right to ask questions while the client has the obligation to an-
swer those questions (Fetzer 2000).
Closely related to the conception of context as a dynamic construct is its rela-
tional conception which conceives it as a relational construct, relating communicative
actions and their surroundings, relating communicative actions, relating individual
participants and their individual surroundings, and relating the set of individual par-
ticipants and their communicative actions to their surroundings. Against this back-
ground, the meso domain of speech event can be seen as connecting individual action
with collective goals (Alexander and Giesen 1987) thus representing an important
bridging point in the micro-macro interface.
Third, context is seen as given, as is reflected in the presuppositional approach to
context which is also referred to as common ground or background information. Here,
context is seen as a set of propositions which participants take for granted in interac-
tion. This allows for two different conceptions of context: a static conception in which
context is external to the utterance, and an interactive one, in which context is im-
ported into the utterance while at the same time invoking and reconstructing context
(Fetzer and Fischer 2007). While the former has been refuted in pragmatics, which is
concerned fundamentally with context-dependent meaning and thus with communi-
cative action and its felicity in context, it still has a number of supporters in informa-
tion science, as is put succinctly by Levinson (2003: 33):
the idea that utterances might carry along with them their own contexts like a snail
carries its home along with it is indeed a peculiar idea if one subscribes to a defini-
tion of context that excludes message content, as for example in information the-
ory. Context is then construed as the antecedent set of assumptions against which
a message is construed. But it has long been noted in the study of pragmatics that
this dichotomy between the message and context cannot be the right picture.

The context-dependence of context is thus reflected in its statuses as (1) given and
external to the utterance, (2) re-constructed and negotiated in and through the process
of communication, (3) indexical, and (4) never saturated.
The connectedness between context and discourse has become apparent through-
out the analysis of context above. To shed more light on the connectedness, their
shared domains of reference, that is society, culture, cognition and language, are exam-
ined in context, and context as a whole is classified into social context, sociocultural
context, cognitive context and linguistic context.
Contexts in context 

2.1 Linguistic context

Linguistic context comprises the actual language used within discourse. Language is
composed of linguistic constructions (or parts) embedded in adjacent linguistic con-
structions composing a whole clause, sentence, utterance, turn or text. Thus, linguistic
context, co-text (de Beaugrande and Dressler 1981, Janney 2002) or verbal context
(Robinson 2006), denotes a relational construct composed of local and global adja-
cency relations. In the stance adopted here, the connectedness between a linguistic con-
struction (a part) and other linguistic constructions constituting a text (the whole) is
looked upon analogously to Searle’s conception of regulative rules and constitutive
rules (Searle 1969). That is to say, the rule-governed realization of linguistic construc-
tions in context constitutes an utterance act thus counting as a move within the game of
producing and interpreting utterance acts. At the same time, the utterance act counts as
a move within the game of producing and interpreting speech acts. While the rule-
governed realization of linguistic constructions is constrained by the rules of grammar,
the production and interpretation of speech acts are constrained by felicity conditions.
The production and interpretation of an utterance act is anchored to language’s
constitutive parts of syntax, morphology, phonology, semantics and pragmatics. While
syntax is composed of structural units, for instance constituents in traditional gram-
mar, phrases in functional grammar and generative grammar, groups in systemic func-
tional grammar or constructions in construction grammar, it is the linear ordering of
the individual parts within a sequence which constitutes their grammatical function.
The adverb really, for instance, realizes the grammatical function of a sentence adver-
bial with wide scope in the utterance really, Tim is strange while it is assigned the gram-
matical function of the adverbial of a subjunct with narrow scope in Tim is really
strange. Or, the proper noun Mark can realize the grammatical function of an object in
Sue met Mark and it can realize the grammatical function of a subject in Mark met Sue.
Thus, it is not the linguistic construction as such which is assigned a grammatical func-
tion. Rather, it is the positioning of a linguistic construction within a sequence which
assigns it a grammatical function. In the framework of constitutive rules, the position-
ing of a linguistic construction within a sequence counts as a move within the game of
producing utterance acts thereby assigning it a grammatical function.
The relational nature of linguistic context is also reflected in a sentence’s topologi-
cal units of pre-field, middle-field and post-field, and their respective sub-fields, which
are also conceived of in relational terms thus counting as further constitutive parts in
the construction of an utterance act. For instance, a change in the canonical word or-
der SVO in English in the utterance Sue called Peter to a non-canonical OSV Peter Sue
called with stress on initial O does not change the propositional meaning of the utter-
ance. From a discursive viewpoint, however, the fronting of the object signifies a con-
trastive set. That is, the speaker intends the message that Sue called Peter while at the
same time implicating that Sue did not call other not-named, but presupposed mem-
bers of the contrastive set, for instance Tom, Larry or Mark.
 Anita Fetzer

The investigation of syntax and syntactic structure from a context-anchored parts-

whole perspective has demonstrated that the whole, viz. the whole utterance, is more
than the sum of its parts, as the ordering of the individual parts constitutes additional
discursive meaning.
An investigation of morphology from a context-based perspective sheds further
insights into the morphological processes of inflection, derivation and compounding.
In that scenario, inflection, such as the inflectional morpheme s in English in the word
form [[run][s]], is looked upon as an indicator for the connectedness between indi-
vidual words in the context of a sentence by making explicit the status of the word as a
pluralized noun in the utterance he did five runs or as a verbal form indicating simple
present tense in Sue runs faster than Mark, for instance. Derivation and compounding
are analyzed as making explicit the connectedness between morphemes in the context
of a lexical expression, such as the prepositional verb [[take] [over]] and the com-
pound [[over][take]], or the derivational affix and free morpheme [ism] in [[contextual]
[ism]] and in there’s too much ism nowadays. As has been shown for syntax, an analysis
of morphology from a context-anchored parts-whole perspective leads to more re-
fined results in word formation with respect to inflection, derivation and compound-
ing. This is because not only the part but also its position in the sequence is required
for the constitution of its function.
An explicit accommodation of context from a parts-whole perspective in pho-
nology also leads to stimulating new insights, refining the results obtained within
the framework of more static notion of phonological environment. Here, assimila-
tion is looked upon as the adaptation of a part to its phonological context (or whole).
The following sequence, ten pencils, consists of two parts, [ten] and [penslz]. When
realized as a whole, the alveolar nasal [n] is adapted to its local phonological context
[p] and realized as another bilabial sound, viz. [m], in the phonological sequence
[tempenslz]. The context-anchored parts-whole perspective is further manifest in
the realization of a phonological form as a full or reduced form. For instance, the
preposition to in the sequence I walked to school can be realized as a reduced form
with a schwa-sound stressing the location to which the speaker went, namely
‘school’, or it can be realized by the monophthong [u] as a full form stressing the
direction of the movement.
A context-anchored perspective is also of importance in supra-segmental phonol-
ogy, and here in particular in the field of intonation with its unit of investigation, the
intonational phrase. In systemic-functional grammar, intonation is seen as a signalling
system (Halliday 1994) and in interactional sociolinguistics an intonational phrase is
assigned the status of a contextualization device (Gumperz 1996). For instance, the
intonational contour of a fall signifies the illocutionary force of a request in English
and the intonational contour of a rise signifies the illocutionary force of an offer. The
one-word utterance coffee realized with a fall contextualizes the utterance as a request,
while its realization with a rise contextualizes it as an offer. Again, a context-anchored
parts-whole perspective to the investigation of phonology leads to exciting insights by
Contexts in context 

showing that the phonological realization of an utterance contextualizes a speaker’s

communicative intention (Pierrehumbert and Hirschberg1992).
Semantics has been traditionally defined as the investigation of context-inde-
pendent meaning while pragmatics has been promoted as the investigation of con-
text-dependent meaning. From a parts-whole perspective, truth-conditional seman-
tics examines the meaning of a whole proposition by identifying its constitutive parts,
that is reference and predication. Whenever all of the constitutive parts are true, the
meaning of the whole proposition is true. In that frame of reference, the propositions
Peter does not do anything to change his life and Peter does not do nothing to change his
life do not share the same truth conditions and therefore are not identical. From a
pragmatics-based outlook, however, they may share the same communicative status
in communication.
Possible-worlds-anchored semantics restricts the investigation of a proposition’s
meaning from truth conditions which are valid in any context to that of one of its sub-
sets, a possible-world scenario. Here, the meaning of a proposition is true in a specified
scenario only. Discourse semantics focuses on anaphora resolution, cohesion and co-
herence, and lexical semantics examines the semantic meaning of lexical expressions,
such as large. Again, the explicit accommodation of a context-anchored parts-whole
perspective may lead to more refined results, as has been pointed out by Akman and
Alpaslan (1999: 10) in their examination of the meaning of the adjective large. In the
utterance Stephen built a large snowman the lexical meaning of the adjective large is
interdependent on the size of the discourse identity of Stephen. If Stephen is a toddler,
‘large’ denotes a size of about 1.20 metres, and if Stephen is an adult, ‘large’ denotes a
size of about 1.80 meters.
Speech acts are composed of propositional acts and illocutionary acts which are
composed of further constitutive acts, such as reference acts and utterance acts
(Searle 1969). Or, they are composed of locutionary acts, illocutionary acts and perlo-
cutionary acts which are composed of further constitutive acts, such as phatic acts or
rhetic acts (Austin 1980). All of a speech act’s constitutive acts – except the perlocu-
tionary act – are conventional in nature, and it is their performance in an appropriate
context which makes them count as a particular speech act. For instance, the utterance
I hereby request you to give me 100 Euros is composed of the illocutionary act of a re-
quest, which is signalled through the explicit performative I hereby request, and the
propositional act is composed of the reference act ‘you’ referring to the hearer and the
predication ‘give speaker 100 Euros’. Following Searle (1969), the performance of the
constitutive acts counts as a request if the generalized felicity conditions for a request
obtain, viz. normal in- and out-put conditions (speaker and hearer speak the same
language), propositional content conditions (reference to a future act), preparatory
conditions (speaker has the necessary social status to utter a request and make the
hearer comply with the act, and hearer has the requested sum of money), sincerity
condition (speaker means what he or she utters), and essential condition (the utterance
counts as a request). Adopting a context-anchored framework, Sbisà (2002) assigns
 Anita Fetzer

felicity conditions the status of context categories thus extending the domain of valid-
ity of in- and output conditions from ‘language’ to that of ‘language use in context’.
As has been shown for the domains of syntax, morphology and phonology, a con-
text-anchored parts-whole perspective to the examination of language and language
use achieves news insights into their rule-governed nature and into the nature of their
connectedness. This also holds for the fields of semantics and pragmatics, where the
whole, that is the whole proposition, discourse and speech act, is more than the sum
of its parts.
The explicit accommodation of linguistic context to the investigation of syntax,
morphology, phonology and semantics leads to more refined results regarding the
grammar-interaction interface. Not only does this allow for a systematic examination
of linguistic parts, such as syntactic, morphological and phonological constructions or
lexical expressions, and their connectedness with wholes, such as clauses, sentences,
utterances, propositions and texts, but also to a holistic outlook on grammar, which
may be supplemented by social, sociocultural and cognitive perspectives.
In the following, cognitive context which is a necessary condition for a cognition-
based theory of language and language use is examined.

2.2 Cognitive context

Cognitive context is not only of relevance to cognitive linguistics and cognitive prag-
matics, but also to the field of psychology, and here in particular to the psychology of
communication. Bateson (1972) conceives context along the lines of the gestalt-psy-
chological distinction between figure and ground and the related concepts of frame
and framing. Frame is seen as a delimiting device which “is (or delimits) a class or set
of messages (or meaningful actions)” (Bateson 1972: 187). Because of its delimiting
function, “psychological frames are exclusive, i.e. by including certain messages
(or meaningful actions) within a frame, certain other messages are excluded” and they
are “inclusive, i.e. by excluding certain messages certain others are included” (Bateson
1972: 187). The apparent contradiction is eradicated by the introduction of set theory’s
differentiation between set and non-set, which – like figure and ground – are not sym-
metrically related. To use Bateson’s own words: “[p]erception of the ground must be
positively inhibited and perception of the figure (...) must be positively enhanced”
(Bateson1972: 187). This leads him to the conclusion that the concept of frame is
metacommunicative, which also holds for context. Or in his own words: “the hypoth-
esis depends upon the idea that this structured context also occurs within a wider
context – a metacontext if you will – and that this sequence of contexts is an open, and
conceivably infinite, series” (Bateson 1972: 245).
Bateson explicitly connects set and non-set, frame and meta-frame, and context
and meta-context with a parts-whole perspective: “whenever this contrast appears in
the realm of communication, is simply a contrast in logical typing. The whole is always
in a metarelationship with its parts. As in logic the proposition can never determine
Contexts in context 

the meta proposition, so also in matters of control the smaller context can never deter-
mine the larger” (Bateson 1972: 267).
In frame analysis, Goffman (1986) uses frame as a metaphor for context, back-
ground and setting thus referring to the relational dimension of meaning: “I am not
addressing the structure of social life but the structure of experience individuals have
at any moment of their social lives” (Goffman 1986: 13). The relational conception of
frame is reflected in Goffman’s differentiation between primary framework, key and
fabrication. A primary framework provides a way of describing the event to which it is
applied. Key denotes “the set of conventions by which a given activity, one already
meaningful in terms of some primary framework, is transformed into something pat-
terned on this activity but seen by the participants to be something quite else” (Goffman
1986: 44).
Framework is connected with the activity of framing, key is connected with the
activities of keying, downkeying, that is opting for a more direct manner of expression,
upkeying, that is opting for a more indirect manner of expression, and rekeying, that
is opting for a different manner of expression. Fabrication is distinguished with respect
to self-induced and other-induced fabrications, and is defined as “the intentional effort
of one or more individuals to manage activity so that a party of one or more others will
be induced to have a false belief about what it is that is going on” (Goffman 1986: 83).
The concept of frame is fundamental to the construction of meaning: “In general,
then, the assumptions that cut an activity off from the external surround also mark the
ways in which this activity is inevitably bound to the surrounding world” (Goffman
1986: 249). While the connectedness between frame and framing, and amongst key-
ing, upkeying, downkeying and rekeying needs to be based on meta-representation,
framing also needs to be recursive: “Frame, however, organizes more than meaning; it
also organizes involvement” (Goffman 1986: 345).
Cognitive context is not only a key to the psychology of communication. It is also
of immense importance for language processing and the corresponding inference pro-
cesses involved. Relevance theory (Sperber and Wilson 1986) differentiates between
cognitive environment and cognitive context: the former refers to a set of facts, while
the latter refers to a set of premises, namely, true or possibly true mental representa-
tions. Constitutive elements of cognitive context are mental representations, proposi-
tions, contextual assumptions which may vary in strength, and factual assumptions.
Assumptions are read, written and deleted. In the meantime, contextual implications
are raised in strength, lowered in strength or erased from memory. Since cognitive
contexts are anchored to an individual but are also required for a cognitively based
outlook on communication, they must contain assumptions about mutual cognitive
environments. Thus, cognitive context is not only defined by representations but also
by meta-representations. To describe multilayered cognitive context, relevance theory
employs the onion metaphor and represents context as an onion with its constitutive
layers. What is of importance for language processing and inferencing is the premise
that the order of inclusion corresponds to the order of accessibility. This ensures that
 Anita Fetzer

both processes are ordered, and that their order is based on meta-representations,
meta-layers and meta-contexts.
In functional grammar, context also denotes a psychological construct, which
Givón (2005: 91) explicates in Context as Other Minds as follows:
First, we noted that context is not an objective entity but rather a mental con-
struct, the construed relevant ground vis-à-vis which tokens of experience achieve
relatively stable mental representation as salient figures. Whatever stability mental
representations possess is due, in large measure, to the classification of tokens of
experience into generic categories or types.

What is important to the investigation of cognitive context is the differentiation be-

tween types of experience and tokens of experience. While the former are of prime
relevance to language processing and inferencing, the latter are intrinsically connected
with practical reasoning and abduction, in and through which tokens are categorized
into types.
A similar type-token differentiation is manifest in Penco’s distinction between
subjective context and individual context (Penco 1999). The former refers to a cogni-
tive or epistemic representation of the world and thus to an individual set of beliefs
which may belong to an individual or a community, and the latter refers to an indi-
vidual’s set of beliefs and thus to their representation of the world. A necessary conse-
quence of Givón’s differentiation between type of experience and token of experience,
and Penco’s differentiation between subjective context and individual context is the
fact that cognitive context is both individual (or unique) and social (or typed). This
ambivalence is also reflected in van Dijk’s context model (van Dijk 2006), in which
context is allocated to the interface between social structure and discourse. Accord-
ingly, contexts are not objective or deterministic constraints of society or culture, but
subjective participants’ interpretations, constructions or definitions of aspects of the
social environment.
The token-type differentiation is not only of relevance to the micro domain. It has
been extended to the meso domain of genre thus filtering the production and interpre-
tation of utterance acts, as is succinctly put by Thibault (2003: 44):
Rather, genres are types. But they are types in a rather peculiar way. Genres do
not specify the lexicogrammatical resources of word, phrase, clause, and so on.
Instead, they specify the typical ways in which these are combined and deployed
so as to enact the typical semiotic action formations of a given community.

Cognitive context is a structured, multilayered construct which is indispensable for

language processing and inferencing. The nature of the connectedness between its
constitutive layers and subsystems is meta-communicative and meta-systemic.
In the following, social context is examined more closely.
Contexts in context 

2.3 Social context

Social context is often considered to comprise the context of a communicative ex-

change and is defined by deducting linguistic context and cognitive context from a
holistic conception of context. Constituents of social context are, for instance, partici-
pants, the immediate concrete, physical surroundings including time and location,
and the macro contextual institutional and non-institutional domains. Frequently,
language use in social contexts has been allocated to communicative performance
(or parole) which has been assigned the status of an individual and momentary prod-
uct. This has not only been denied by ethnographic, but also by artificial-intelligence
anchored studies (Gumperz and Levinson 1996, Recanati 1998).
The non-individualistic use of language is also manifest in the contextual phe-
nomenon of deixis and its realization as deictic expressions, that is temporal deixis,
local deixis, participant deixis, discourse deixis and social deixis. Unlike anaphora
resolution, which requires linguistic context, discourse deixis is informed by both lin-
guistic and social contexts (cf. Waugh, this volume).
The categories of speaker, hearer and audience are no longer conceived of as ana-
lytic primes. Rather, they denote interactional categories and have been refined by
Goffman (1981) with respect to their footing anchored to the participation frame-
work. In an actual speech situation, however, the interactional roles do more than
simply produce and interpret communicative contributions. In a social context, they
subcategorize into social roles and their gendered and ethnic identities, to name but
the most prominent ones. In institutional communication, the participants’ institu-
tional roles embody institutional power which manifests itself in their context-depen-
dent rights and obligations.
Social context has been further differentiated by van Dijk (1981) into general so-
cial context anchored to functional pragmatic coherence and particular social context
types anchored to contextual frames, assumed purposes and intentions. The context
types contain information about institutions, frame structures, settings, functions,
properties, relations, positions and frame conventions. To account for the micro-mac-
ro interface, van Dijk distinguishes between macro actions and micro actions.
The research paradigm of ethnomethodology investigates the interactional orga-
nization of society. It represents a micro sociological perspective par excellence, in
which the indexicality of social action is of key importance. Ethnomethodology fo-
cuses on intersubjectivity and examines the questions of how separate individuals are
able to know or act within a common world, and of how members (or participants)
negotiate or achieve a common context: “in an interaction’s moment-to-moment de-
velopment, the parties, singly and together, select and display in their conduct which
of the indefinitely many aspects of context they are making relevant, or are invoking,
for the immediate moment” (Schegloff 1987: 219). Here, common context is synony-
mous with social context, which, like linguistic context, classifies into micro social
context and macro social context. To refine the dichotomous micro-macro interface,
 Anita Fetzer

social contexts further classify into a number of intermediate layers, such as meso so-
cial context which denotes the delimiting frame of a particular speech event (Drew and
Heritage 1992, Sarangi and Slembrouck 1996).
The importance of social context to communication and its relevance is spelled
out by Hanks (1996: 235) as follows:
Hence it is not that people must share a grammar, but that they must share, to a
degree, ways of orienting themselves in social context. This kind of sharing – par-
tial, orientational and socially distributed – may be attributed to the habitus, or
relatively stable schemes of perception to which actors are inculcated.

But is it really social context which is at the heart of communication? In the following,
the connectedness between social context and culture is examined more closely.

2.4 Sociocultural context

Social context is frequently used synonymously with extra-linguistic context which

contains the participants of a communicative exchange, their physical and psychologi-
cal dispositions and the specific knowledge or assumptions about the persons involved,
the knowledge of the language and the conventions regarding appropriate use of lan-
guage, the knowledge of activity-types including communicative intentions and goals,
and general background knowledge. Of course, the immediate extra-linguistic context
is embedded in more remote extra-linguistic contexts, such as organizational contexts
and other socio-historically constituted contexts of institutions and (sub)cultures.
The synonymous use of extra-linguistic context and social context is, however, an
oversimplification as research in sociolinguistics, anthropology and cultural studies
has informed us. Rather, social context subcategorizes into different types of sociocul-
tural context which are defined by a particular perspective on social context. Against
this background, social context is conceived of as an unmarked type of context, and
sociocultural context is conceived of as a marked type of context in which extra-lin-
guistic variables, such as time, location or individual, are set and interpreted accord-
ingly. For instance, the distinction between monochronic time and polychronic time is
based on the differentiation between a linear, tangible and divisible conception of time,
where events are scheduled one at a time and where this schedule takes precedence
over interpersonal relationships. Polychronic time, by contrast, is characterized by
things occurring simultaneously. Here, interpersonal relationships take precedence
over a task-oriented outlook on communication (Hall and Hall 1989). This also holds
for a culture-dependent interpretation of space with respect to public space and pri-
vate space on the macro domain as well as on the micro domain, referring to interper-
sonal proximity (Goffman 1986, Hall and Hall 1989), and to a culture-dependent in-
teractional organization of gender (Ochs 1992). Consequently, the social-context
variables of time, space and identity obtain a culture-specific interpretation in differ-
ent social contexts, thus co-constructing different sociocultural contexts.
Contexts in context 

Culture provides the members of a speech community1 with a common configu-

ration and interpretation of extra-linguistic variables, such as time, space, institution,
individuals and their multiple roles, and of linguistic variables, such as linguistic con-
structions and conventionalized meanings, levels of pragmatic directness or preferred
sequential organizations. Metaphorically speaking, culture is a filter which allows us to
interpret social context in accordance with particularized sociocultural-context con-
straints and requirements.
A culture-dependent outlook on communication has been promoted by the eth-
nography of communication (Saville-Troike 1989), in particular by Hymes’ speaking
grid (Hymes 1974). Here, a communicative exchange is systematized with respect to
its constitutive components of situation (the physical setting and the psychological
scene), participants (speaker, hearer and audience, and their statuses in the participa-
tion framework), ends (the goal and the purpose of the speech event from a sociocul-
tural viewpoint), act sequence (how something is said with regard to message form
and what is said with regard to message content), key (mock or serious), instrumen-
talities (channels, i.e. spoken, written, e-mail, multi-modal), norms of interpretation
and forms of speech (vernacular, dialect, standard), and genre.
Hymes’s speaking grid and his notion of communicative competence has been
refined by Gumperz (1992) who explicitly connects the cognitive operation of infer-
ence with the sociocultural activity of conversation. His conception of conversational
inference represents a context-bound process of interpretation in which other’s inten-
tions are assessed, and in which self illustrates her/his understanding and comprehen-
sion through her/his response. Gumperz assigns language usage the status of actual
language practice, and he interprets evaluation as a social activity. Saville-Troike sum-
marizes his contribution to the ethnography of communication as follows: “Gumperz
builds on this in proposing the outline of a theory of how social knowledge is stored in
the mind, retrieved from memory, and integrated with grammatical knowledge in the
act of conversing” (Saville-Troike 1989: 131). Because of its cultural base, the meaning
that emerges in a conversation is different for participants if they are not members of
the same speech community.
Gumperz’s original contribution to the field of ethnography of communication
bridges the gap between (1) linguistic context on the one hand and sociocultural and
social contexts on the other, and (2) between linguistic, social and sociocultural con-
texts on the one hand, and cognitive context on the other. This is due to the fact that
language is seen as a socially situated cultural form and, therefore, as a specification of
the more general social and linguistic contexts. Moreover, Gumperz’s interactional so-
ciolinguistics accommodates a micro and macro outlook on communication which is

1. Culture is used synonymously with Hymes’ conception of speech community. It may com-
prise national (sub)culture, ethnicity- and gender-related (sub)cultures and work-related (sub)
culture, to name but the most prominent ones.
 Anita Fetzer

not only reflected in the participants’ communicative performance, but also in their
conversational inferencing:
It is useful to distinguish between two levels of inference in analyses of interpre-
tive processes: (a) global inferences of what the exchange is about and what mu-
tual rights and obligations apply, what topics can be brought up, what is wanted
by way of a reply, as well as what can be put into words and what is to be implied,
and (b) local inferences concerning what is intended with my one move and what
is required by way of a response. (Prevignano and di Luzio 2003: 14)

Interactional sociolinguistics thus represents a holistic context-based framework par

excellence. The closing section systematizes the results obtained. It demonstrates that
a relational conception of context based on the connectedness amongst context, con-
textualization and contextualization cue provides the tools to bridge the gap between
internal and external contexts, micro and macro contexts, marked and unmarked con-
texts, and context importation and context invocation.

3. Context, contextualization and contextualization cues

Interactional sociolinguistics has surfaced as a frame of reference which may not only
account for the different types of context examined above, but also for the micro-
macro interface and the respective bridging operations. This is due to the fact that
(1) language is seen as a socially situated form, (2) inference is given a context-depen-
dent interpretation, and (3) its unit of investigation, the speech activity, is a meso cat-
egory going beyond the micro domain of isolated speech acts, utterances or turns.
The accommodation of context is a fundamental premise for the analysis of natu-
ral-language communication, as is made explicit by Gumperz (2003: 119):
With respect to context, psychologists, cognitive scientists, and many linguists
who pay attention to context tend to define it almost entirely in extra-communi-
cative terms. I argue that, while these factors are, of course, significant, contextual
information is imported into the interpretative process primarily via indexical
contextualization cues, in the form of presuppositions of what the activity is and
what is communicatively intended.

To account for context, communication needs to be anchored to the basic pragmatic

premise of intentionality of communicative action (Brandom 1994, Searle 1995). Not
only are context and intentionality of relevance to the investigation of speech activities,
but so is the contextualization of a communicative contribution and its constitutive
parts. To use Gumperz’s (1996: 403) own words: “Since all interpretation is always
context-bound and rooted in collaborative exchanges that rest on shifting contextual
presuppositions, contextualization must be a universal of human communication”.
Contexts in context 

Table 1.  Micro and Macro Dimensions of Context

micro linguistic context macro linguistic context

–  constructions –  text
–  phrases
–  clauses
–  sentences/utterances

micro cognitive context macro cognitive context

–  local inferencing and reasoning – global inferencing and reasoning with update

micro social context macro social context

–  participants –  institution
–  location
–  time
–  .............

micro sociocultural context macro sociocultural context

–  situation –  genre (Hymes 1974)
–  participants
–  end
–  act sequence
–  key
–  instrumentalities
– norms of interpretation and forms
of speech (Hymes 1974)

Contextualization or the importation and invocation of contextual information is con-

nected with the fundamental pragmatic premise of indexicality and thus with the
meta-linguistic device of contextualization cue. Before the connectedness amongst
context, contextualization and contextualization cue is examined more closely, the re-
sults of the analysis of context and of its ‘objects’ obtained in the previous section are
schematized in Table 1.
The objects of context contained in context do not represent indexical devices as
such, in spite of the fact that they may be referred to with the help of indexical expres-
sions in discourse. So what is the difference between an ‘object of context’ and a con-
textualization cue?
Unlike the types of objects of context listed above, contextualization cues are me-
ta-linguistic indexicals which belong to a meta-system and
serve to highlight, foreground or make salient certain phonological or lexical
strings vis-á-vis other similar units, that is they function relationally and can-
not be assigned context-independent, stable, core lexical meanings. Foreground-
ing processes, moreover, do not rest on any one single cue. Rather, assessments
 Anita Fetzer

depend on cooccurrence judgments (...) that simultaneously evaluate a variety

of different cues. When interpreted with reference to lexical and grammatical
knowledge, structural position within a clause or sequential location within a
stretch of discourse, foregrounding becomes an input to implicatures, yielding
situated interpretations. (Gumperz 1992: 232)

Regarding their function in discourse, they “serve to retrieve the contextual presup-
positions conversationalists rely on making sense of what they see and hear in interac-
tive encounters. They (..) have no propositional content. That is, (..) they signal only
relationally and cannot be assigned context-free lexical meanings” (Prevignano and di
Luzio 2003: 9). So, contextualization cues import context into the speech activity and
they bring context out in the speech activity by channelling “inferential processes that
make available for interpretation knowledge of social and physical worlds” (Gumperz
1996: 383).
The classification of context into micro and macro linguistic, cognitive, social and
sociocultural context types and the accommodation of the meta-linguistic device of a
contextualization cue, which expresses relational meaning by signifying the nature of
the connectedness between the individual objects and context, is a first move towards
a theory of context as a dynamic construct. In order to account for the dynamics of
context in an appropriate manner, a further distinction needs to be introduced, name-
ly one between a default and a non-default context (Recanati 1989). Against this back-
ground, the different types of context introduced above are further differentiated into
default contexts, or unmarked contexts in functional-grammar terminology
(Givón 1993), and non-default (or marked) contexts. In functional-grammar terms,
the marked category is more complex regarding structure, less frequent regarding dis-
tribution, and harder to process, while the unmarked category is less complex regard-
ing structure, more frequent regarding distribution, and less hard to process
(Givón 1993: 178). The classification of context as default or non-default is performed
through conversational inferencing. Analogously to the Gricean implicature “infer-
ring also involves a two-step process in which the contextual ground, in terms of which
an assessment of what is perceived is made, must be first retrieved and related to stored
memories before an interpretation is arrived at”(Gumperz 1996: 383).
For the connectedness between contextualization and default/non-default context
this means that unmarked contexts are less complex regarding structure; that is to say
the process of contextualization does not need to accommodate “more” contextual in-
formation “than is required” (Grice 1975: 45) for the current purpose of the talk ex-
change. Furthermore, unmarked contexts are more frequent and thus defined by a high-
er degree of conventionalization, and they are less hard to process as both their
structural configuration and the higher degree of conventionalization are internally
documented in culture-dependent default frames, or stored memories in Gumperz’s ter-
minology. Marked contexts, by contrast, deviate from the default configuration by being
more complex regarding structure; that is to say, the process of contextualization needs
Contexts in context 

to accommodate more contextual information, thus going beyond the default frame.
Marked contexts are less frequent and therefore defined by a lower degree of conven-
tionalization, and they are harder to process as all of the surplus contextual information;
that is the relevant tokens, need to be administered and attributed to the appropriate
types so that all of the contextual information may be stored in the default frames.
In a nutshell, a dynamic theory of context needs to be anchored to the basic prag-
matic premises of intentionality, indexicality and contextualization. This requires the
accommodation of cooperation and collaboration on the one hand, and micro, meso
and macro referential domains on the other. In such a meta-systemic framework, con-
text can be delimited in spite of the fact that “this sequence of contexts is an open and
conceivably infinite series” (Bateson 1972: 245).


The author is grateful to the anonymous reviewers for helpful comments on the first
version of this paper. Any remaining errors are exclusively the author’s.


Akman, V. and Alpaslan, F.N. 1999. “Strawson on intended meaning and context”. In Modeling
and Using Context. 2nd International and Interdisciplinary Conference, Context’99, Proceed-
ings, P. Bouquet et al. (eds), 1–14. Heidelberg: Springer.
Akman, V. Bouquet, P. Thomason, R. and Young, R. A.(eds). 2001. Modeling and Using Context.
Third International and Interdisciplinary Conference, CONTEXT 2001, Proceedings. Spring-
er: Heidelberg.
Alexander, J.C. and Giesen, B. 1987. “Introduction.” In The Micro-Macro Link, J.C. Alexander, et
al. (eds), 1–42. Berkley: The University of California Press.
Austin, J.L.1980. How to Do Things with Words. Cambridge: Cambridge University Press.
Bateson, G. 1972. Steps to an Ecology of Mind. New York: Chandler Publishing Company.
Blackburn, P., Ghidini, C., Turner, R. M. and Giunchiglia, F. (eds). 2003. Modeling and Using
Context. 4th International and Interdisciplinary Conference, Context 2003, Proceedings.
Heidelberg: Springer.
Bouquet, P., Brezillon, P., Benerecetti, M., Castellani, F. and Serafini, L. (eds). 1999. Modeling and
Using Context. Second International and Interdisciplinary Conference on Modeling and Using
Context, Context’99, Proceedings. Heidelberg: Springer.
Brandom, R.B. 1994. Making it Explicit: Reasoning, Representing, and Discursive Commitment.
Cambridge: Harvard University Press.
De Beaugrande, R. and Dressler, W. 1981. Einführung in die Textlinguistik. Tübingen: Niemeyer.
Drew, P. and Heritage, J. (eds).1992. Talk at Work: Interaction in Institutional Settings. Cambridge:
Cambridge University Press.
 Anita Fetzer

Fetzer, A. 1999. “Non-acceptances: re- or un-creating context”. In Modeling and Using Context.
2nd International and Interdisciplinary Conference, Context’99, Proceedings, P. Bouquet et
al. (eds), 133–144. Heidelberg: Springer.
Fetzer, A. 2000. “Negotiating validity claims in political interviews”. Text 20(4): 1–46.
Fetzer, A. 2004. Recontextualizing Context: Grammaticality meets Appropriateness. Amsterdam:
Fetzer, A. and Fischer, K. 2007. “Introduction”. In Lexical Markers of Common Grounds, A. Fetzer
and K. Fischer (eds), 1–14. Oxford: Elsevier.
Foucault, M. 1997. Die Ordnung des Diskurses. Frankfurt: Fischer.
Garfinkel, H.1994. Studies in Ethnomethodology. Cambridge: Polity Press.
Givón, T. 1993. English Grammar: a Function-Based Introduction. Amsterdam: Benjamins.
Givón, T. 2005. Context as Other Minds. Amsterdam: Benjamins.
Goffman, E. 1981. Forms of Talk. Oxford: Basil Blackwell.
Goffman, E. 1986. Frame Analysis. Boston: Northeastern University Press.
Goodwin, C. and Duranti, A. (eds) 1992. “Rethinking context: an introduction.” In Rethinking
Context. Language as an Interactive Phenomenon, A. Duranti and C. Goodwin (eds), 1–42.
Cambridge: Cambridge University Press.
Grice, H.P. 1975. “Logic and conversation”. In Syntax and Semantics. Vol. III, M. Cole and J.L.
Morgan (eds), 41–58. New York: Academic Press.
Gumperz, J.J.1992. “Contextualization and understanding.” In Rethinking Context: Language as
an Interactive Phenomenon, A. Duranti and C. Goodwin (eds), 229–252. Cambridge:
Cambridge University Press.
Gumperz, J.J. 1996. “The linguistic and cultural relativity of inference.” In Rethinking Linguistic
Relativity, J.J. Gumperz and S.C. Levinson (eds), 374–406. Cambridge: Cambridge Univer-
sity Press.
Gumperz, J.J. 2003. “Response essay.” In Language and Interaction. Discussions with John J.
Gumperz, S. Eerdmans et al. (eds), 105–126. Amsterdam: Benjamins.
Gumperz, J.J. and Levinson, S.C. (eds). 1996. Rethinking Linguistic Relativity. Cambridge:
Cambridge University Press.
Hall, E.T. and Hall, M.R. 1989. Understanding Cultural Differences: Germans, French and
Americans. Yarmouth/Maine: Intercultural Press.
Halliday, M.A.K. 1994. Introduction to Functional Grammar. London: Arnold.
Hanks, W.F. 1996. “Language form and communicative practices.” In Rethinking Linguistic Relativ-
ity, J.J. Gumperz and S.C. Levinson (eds), 232–270. Cambridge: Cambridge University Press.
Heritage, J.1984. Garfinkel and Ethnomethodology. Cambridge: Polity Press.
Hymes, D.1974. Foundations in Sociolinguistics: An Ethnographic Approach. Philadelphia: Uni-
versity of Philadelphia Press.
Janney, R.W. 2002. “Cotext as context: vague answers in court.” Language and Communication
22(4): 457–475.
Levinson, S.C. 1979. “Activity types and language.” Linguistics 17:365–399.
Levinson, S.C. 2003. “Contextualizing ‘contextualization cues’.” In Language and Interaction. Dis-
cussions with John J. Gumperz, S. Eerdmans et al. (eds), 31–40. Amsterdam: Benjamins.
Linell, P. 1998. Approaching Dialogue. Amsterdam: Benjamins.
Ochs, E. 1992. “Indexing gender”. In Rethinking Context. Language as an Interactive Phenomenon,
A. Duranti and C. Goodwin (eds), 335–358. Cambridge: Cambridge University Press.
Contexts in context 

Penco, C.1999. “Objective and cognitive context”. In 2nd International and Interdisciplinary
Conference on Modeling and Using Context, Context’99, Proceedings, P. Bouquet et al. (eds),
270–283. Heidelberg: Springer.
Pierrehumbert, J. and Hirschberg, J. 1992. “The meaning of intonational contours in the inter-
pretation of discourse”. In Intentions in Communication, P.R. Cohen et al. (eds), 271–311.
Cambridge: MIT Press.
Prevignano, C. and di Luzio, A. 2003. “A discussion with John J. Gumperz”. In Language and
Interaction. Discussions with John J. Gumperz, S. Eerdmans et al. (eds), 7–30. Amsterdam:
Recanati, F. 1998. “Meaning and force: introduction.” In Pragmatics: Critical Concepts, A. Kash-
er (ed.), 126–154. London: Routledge.
Robinson, D. 2006. Introducing Performative Pragmatics. New York: Routledge.
Sarangi, S. and Slembrouck, S. 1996. Language, Bureaucracy & Social Control. London: Longman.
Saville-Troike, M. 1989. The Ethnography of Communication. Blackwell: Oxford.
Sbisà, M. 2002. “Speech acts in context”. Language and Communication 22(4): 421–436.
Schegloff, E. A. 1987. “Between macro and micro: contexts and other connections.” In The Micro-
Macro Link, J. Alexander et al. (eds), 207–234. Los Angeles: University of California Press.
Schegloff, E.A. 1992. “In another context”. In Rethinking Context. Language as an Interactive
Phenomenon, A. Duranti and C. Goodwin (eds), 191–228. Cambridge: Cambridge Univer-
sity Press.
Searle, J.R. 1969. Speech Acts. Cambridge: Cambridge University Press.
Searle, J.R. 1995. The Construction of Social Reality. New York: The Free Press.
Sperber, D. and Wilson, D. 1986. Relevance: Communication and Cognition. Oxford: Blackwell.
Stalnaker, R. 1999. Context and Content. Oxford: Oxford University Press.
Thibault, P. 2003. “Contextualization and social meaning-making practices.” In Language and
Interaction. Discussions with John J. Gumperz, S. Eerdmans et al. (eds), 41–62. Amsterdam:
Van Dijk, T. 1981. Studies in the Pragmatics of Discourse. The Hague: Mouton.
Van Dijk, T. 2006. “Discourse, context and cognition.” Discourse Studies 8(1): 159–177.
Communicative activity types as organisations
in discourses and discourses in organisations

Per Linell

This paper1 outlines a ‘dialogical’ perspective on discourse, more precisely on

spoken, interactional discourse . Its focus is on interaction and contexts. Central
notions in a dialogical account of talk-in-interaction are those of ‘communicative
project’ and ‘communicative activity type’. This implies a dynamic interpretation
of what ‘organisation(s) in discourse’ are about.

1. Introduction: A dialogical turn

In recent years, some scholars (e.g. Soler-Gallart 2004: 159) have argued that a so-
called ‘dialogical turn’ seems to be gaining ground in the human sciences, and perhaps
even more generally in society. This could be regarded as analogous to the ‘linguistic
turn’ that many proposed as a characteristic movement in the social sciences of the
1970s. That slogan referred to the idea that language was (at last) becoming acknowl-
edged as having a constitutive role for knowledge formation and social realities, and
for the explanation of mind, culture and society. If so, language does not just deal with
the expression of thoughts and the regulation of interaction.
However, in retrospect we find that the claims about the linguistic turn did not
quite capture the gist of what might be involved. Dialogue and dialogicality (notions to
be further explained below) are more fundamental than language, and language is
simply only one of the semiotic means by which humans are in dialogue with their
environments. My perspective on dialogue and dialogicality implies that we attend to
the pervasive impact of interactions and contexts on human sense-making, and to the
omnipresence (at various levels) of the other (i.e. other individuals or collectives, cul-
ture, etc.). Language and discourse (discourse defined as situated language use in talk
or text) are embedded in dialogue, rather than the other way around. Dialogue is not
just something secondary, something made possible merely by language (as many

1. This article is based on a keynote lecture given at OID3, Turku, Finland, August 9, 2006.
The research reported was made possible due to a grant from the Swedish Research Council
(no. 421-2004-1087).
 Per Linell

schools of theoretical linguistics would claim). Language and language-borne thought

cannot function by themselves, interdependent as they are with the world, with the
body, with other aspects of cognition and affect, and of course with human action and
interaction in the world ‘out there’ (A. Clark 1997). Moreover, sense-making in dia-
logue occurs before the infant has developed any ability to use language.

2. Dialogical theory

What else is dialogism or dialogical theory? In order to answer this question, we must
first define and understand ‘dialogism’ in contrast to an alternative, as a counter-theory,
to something else, namely ‘monologism’. What is monologism then? The brief answer
is this: the constituent theories of monologism are the information processing model
of cognition, the transfer model of communication, the code model of language, and
the theory of contexts as external to discourse and thinking (Linell 1998, 2009). Mo-
nologism is based on individualism (assumptions about the autonomous subject, who
– among many other things – initiates speech acts entirely on his or her own initia-
tive), but also on objectivism (the social world consisting of social structures, norms
and rules, with a causal impact on individuals).
By contrast, dialogical theory is based on quite different axiomatic assumptions.
Rather than thinking solely in terms of subjectivity and objectivity, it is based on as-
sumptions of intersubjectivity, and human interactions and (inter-)relations in and to
the world. Some of its points are the following:
a. The dynamics and action- or activity-basis of language use (languaging), thinking
and communication; these involve active situated meaning-making2, aided by so-
ciocultural resources and skills (including language, knowledge of the world, knowl-
edge of communicative activity types, etc.) that have been appropriated over time;
b. The interdependence of acts or utterances and their overarching communicative
projects or activities (i.e. in part-whole relationships);
c. Initiative-response-structure and reciprocity of contributions to discourse; the se-
quentiality of meaning-making;
d. Co-authoring in external dialogue (the interaction with the other), and the pres-
ence of other voices in one’s own discourse – in and through responsivity and ad-
dressivity – and in internal dialogue. This is sometimes called ‘multivoicedness’,
i.e. others’ discourses and ideas are reflected in a single speaker’s or thinker’s

2. Although the terms ‘sense-making’ and ‘meaning-making’ are often used interchangeably,
the former would arguably be more comprehensive, referring to all kinds of perceiving and
understanding the world. ‘Meaning-making’, by contrast, would then be reserved for more or
less conscious actions, related to acts of ‘meaning something’ (e.g. in language use), for which
the actor can be held accountable.
Communicative activity types as organisations in discourses and discourses in organisations 

These assumptions of dialogicality are clearly interdependent, and they all pertain to
interaction and contexts in different ways, and to the role of others. Individual persons
are portrayed as social beings, rather than as entirely autonomous subjects.
The terms ‘dialogicality’ and ‘dialogism’ may be in need of a few further remarks.
‘Dialogicality’ must be seen in terms of general and fairly abstract properties of human
sense-making, and must be distinguished from dialogical organisation, that is, the overt
interaction with two or more persons taking turns. Accordingly, dialogical theory
(alias dialogism) applies to written discourse as well. That is, while written texts are
(more) often monologically organised (and so are of course many spoken messages),
they too have the dialogical properties of responsivity, addressivity, belongingness to
genres, and sometimes also multivoicedness (e.g. Bakhtin, 1986). At the same time,
however, it must be pointed out that written texts easily lend themselves to monologi-
cal, product-oriented theorisation (Linell 2005).
‘Dialogism’, by contrast, refers here3 to a (meta)theoretical framework, and it has
many variants. With a suitably eclectic or ecumenical definition, many present-day
empirical approaches to talk-in-interaction could be called (at least partly) ‘dialogistic’
(Linell 1998: 40–54): Conversation Analysis, contextual discourse analysis, interac-
tional linguistics, interactional sociolinguistics, social semiotics and pragmatics, H.
Clark’s (e.g. 1996) social-psychological action theory of language use, neo-Vygotskyan
activity theory, discursive psychology, social representations theory (Marková 2003),
among others. So, what is relevant here is far more than the theories of Mikhail Bakhtin
and his “circle” (e.g. Brandist et al. 2004).4

3. Organisations in discourses?

I have used the plurals here – organisations in discourses, discourses in organisations

– in order to emphasise multi-dimensionality and multi-voicedness. In addition, we
have to deal with at least two sets of different questions:
a. How is discourse organised as a reflection of social interaction and sociocultural

3. One may note that Mikhail Bakhtin himself sometimes uses ‘dialogism’ in the sense here
proposed for ‘dialogicality’.
4. Some references to my own work on dialogical discourse analysis and dialogical linguistics
are Linell (1998) (Approaching Dialogue, on dialogical properties of discourse and interaction),
Linell (2005) (The Written Language Bias in Linguistics, on monologism in the language sci-
ences), Linell (2006) (‘Towards a dialogical linguistics’, on some dialogical properties of lan-
guage, such as grammatical construction methods and lexical meaning potentials), Linell (2009)
(Rethinking Language, Mind and World Dialogically, on dialogical thinking across disciplines),
and contributions to Marková & Foppa (1990, 1991) and Marková et al. (2007) (Chapter 4 on
communicative activity types).
 Per Linell

b. How is society constituted and organised through discourse?

In dealing with these questions, we are concerned with the micro-macro problem in so-
cial science (cf. Fetzer, this volume). In this paper, I shall start from the micro-end of
situated interaction, focusing on the communicative activity type as a bridging meso-
concept. (On organisations and corporate cultures as discourse, see the work of discourse
analysts such as Boden, 1994, and organisation theorists such as Czarniawska, 1997.)
But for a moment I will first be engaged with communicative projects in general.
This is an important part of a dialogical analysis of spoken interaction.

4. Communicative projects in discourse

Discourse is organised in terms of communicative projects. Let us, as an example, look

at an excerpt from a phone call between two persons, here called C and D. D, who
works with public transportation, has talked about some problems he has got in his
job. At this point, C comes in with a contribution in which he expresses sympathy
(lines 1–3):
(1): Swedish original: DE E INTE ALLS ROLIT (LiCTI: TTC 13)5
1. C: asså ja tycker synd om er ibland för de ä (.)
2. fan gammal skit ni får köra ni vet inte om ni
3. kommer hem en gång ibland
4. D: de här ja
5. C: aa
6. D: javisst
7. C: de ä ju
8. D: de här e ju värdelöst vettu
9. (1.0)
10. D: man sitter ju för fan å på nålar varenda da
11. man ska gå till jobbet
12. C: aa
13. D: de e inte alls rolit (.) man blir ju (.)
14. men dom gör ju inget åt dä heller va (.)
15. de e ju de
16. C: näe de ä ju de som e lessamt mä

5. The excerpts cited in this paper are drawn from Swedish data. LiCTI stands for Linköping
Corpus in Talk-in-Interaction. All names of persons, places, companies, etc. given in the ex-
cerpts are fictive.
After the respective excerpts, (close) English translations are given. All analyses were of course
based on the originals, not the translations. In this paper, however, I shall use the translations in
my comments. Note that line numbers do not always match completely between the Swedish
and English versions.
Communicative activity types as organisations in discourses and discourses in organisations 

17. (0.8)
18. ja har ju en kompis som jobbar på VT ((samma företag
19. som D)) i Bergåkra där
20. D: aa
21. C: vi bor i Sandskoga* (.) Burken ((* Stad i närheten av “Bergåkra”))
22. D: a just dä
23. C: å han eh känner väl också att de börjar bli lite eh
24. eländit ibland
25. D: amen dä men de e ju (.) om man säger de e ju så mycke va
26. C: ((suck)) de ä ju
27. D: de e ju inte eh (.) de e ju inte en eh grej
28. de e ju (.) de e ju jämt för fan
1. C: = y’know I feel sorry for you it’s only because it’s
2. (.) such old shit you’re driving around with, you
3. don’t know if you arrive home at all sometimes
4. D: this, yeah
5. C: yeah
6. D: yeah sure
7. C: it’s
8. D: it’s worthless y’know
9. (1.0)
10. D: it’s damn me it’s like sitting on pins an’
11. needles every day when you go to work
12. C: yeah
13. D: it’s not funny at all (.) you get y’know (.) but
14. they don’t do anything about it either y’know
15. (.) that’s how it is
16. C: no that’s what’s so sad too
17. (0.8)
18. y’know I have a mate who’s working for UR ((the
19. company)) in Charlestown up there
20. D: yeah
21. C: we live in Danby (.) the Can ((nickname of work-mate))
22. D: exactly ((Danby is a town close to Charlestown))
23. C: an he uh also feels y’know that it’s becoming a bit
24. miserable sometimes
25. D: yeah but it’s (.) so to speak it’s so much y’know
26. C: ((sighs)) it’s y’know
27. D: there’s not a uh (.) there’s not a thing it’s
28. y’know (.) hell it’s y’know all the time
 Per Linell

One can treat this episode as one comprehensive but local ‘communicative project’
(henceforth: CP), in which the two parties express their consensus on the evaluation
of the troubles D has told C about (which has occurred before the spate of talk cited
here). This is simultaneously an opportunity for C to express his sympathy, which is a
typical feature of ‘troubles-telling’ in conversation (Jefferson & Lee 1992). Barring
some details, the episode can be divided into three subsequences, which are constitu-
ent CPs within the whole CP (which, of course, is embedded in its turn within a over-
arching CP, which – again – is not cited here). The three constituent CPs are roughly
lines 1–8, lines 10–16, and lines 18–28, respectively.
The three constituent projects make up a three-part sequence of CPs in which the
parties take turns at introducing new subtopics. The first CP (lines 1–8), with C as the
main speaker, is about communicating sympathy with D’s misgivings about his cir-
cumstances at work, and D’s response to this is to confirm the evaluation (“it’s worth-
less”; line 8). The message that the second CP (lines 10–16), with D as the main speak-
er, is aimed at getting across is D’s daily anxiety at work (“sitting on pins an’ needles
every day”; lines 10–11) and his claim that the employer doesn’t seem to care about it
(“they don’t do anything about it either y’know”; line 14). C provides an appropriate
and sympathising response to this in line 16: ‘that’s what’s so sad”. This remark also
serves as a link to the third CP, with C as the instigator and perhaps main speaker;
C reports about a work-mate who has had similar experiences as D. This inspires D to
conclude the account with a couple of rather ‘extreme formulations’ (Pomerantz 1986)
(“there’s not a thing, it’s (i.e. it goes on) all the time”; lines 27–28). In and through this,
D formulates a conclusion, which, as it turns out, also becomes the closing of the whole
three-part episode, since C initiates a different topic directly afterwards (not shown in
the excerpt).
There are other parts of the sequence in (1) that could be analysed as separate, still
smaller CPs. For example, C’s reference to his work-mate (lines 18–19) receives a rath-
er uncommitting response from D (line 20), which seems to make C insert a some-
what more specific reference (mentioning the mate’s nick-name “the Can”; line 21).
Now, D indicates that he understands who is meant (“exactly”; line 22). Thus, lines
21–22 comes out as an extra CP, which revolves around repair.

5. Communicative projects: Some general principles

I will now proceed to a few more general remarks on communicative projects (CPs).
The concept of communicative project originates in the work of Alfred Schütz and
Thomas Luckmann (Schutz & Luckmann 1962; Luckmann 1995) (compare also ‘joint
project’ in H. Clark 1996: ch. 7). CPs can be characterised in terms of:
a. Dynamic progression: CPs emerge in the flow of participants’action, interaction
and migrating attention in communication, where they originate, develop, and are
Communicative activity types as organisations in discourses and discourses in organisations 

brought to completion (at least temporary completion). While in progress, CPs

are usually open-ended and multiply determinable.
Accordingly, CPs must be characterised in terms of dynamic movements and situ-
ated problem-solving. Moreover, while there may be conscious plans or intentions
involved in CPs, many are only marginally planned; thus, we can sometimes un-
expectedly find ourselves involved in doing or saying something that we had not
planned, or we can retrospectively discover that we have just done or said some-
thing which can be justifiably interpreted in ways we have not wanted to encour-
age. CPs vary in participants’ awareness of them.
b. Asymmetrical participation and collective accomplishment: A CP is dialogical: re-
sponsive, addressed, involving an implicit or overt co-action between two or more
parties. The notion is opposed to monological ideas (Searle 1969) that speech acts
are performed by autonomous individuals. CPs are (partially) shared between the
speaker and others; it takes two to communicate. As Bakhtin (1981: 293) said, “the
word in language is half someone else’s”. But partial sharedness is usually com-
bined with an asymmetric distribution of communicative labour: parties do differ-
ent things, and they contribute more or less, but usually they make mutually com-
plementary contributions (H. Clark 1996).
c. Nestedness: CPs are embedded within other, successively larger projects.6 The ‘utter-
ance act’, the ‘project’ of getting something said and understood linguistically, is al-
ways done in the service of an overarching communicative project: “Do you have a
match?” is said when the speaker wants a match to light a cigarette or a candle. A
communicative project is regularly, although sometimes only indirectly, embedded
within a larger non-communicative project. Hence, CPs are context-interdependent.
Large CPs may cover whole encounters, such as a doctor consultation or a job in-
terview (at the level of ‘communicative activity (type)’, see below) and series of
d. Variation in size: A corollary of the prior point is that CPs vary in extension, from
a speaker’s attempt at finding the right word in the appropriate moment or repair-
ing an occasional mishearing, to carrying out extensive and complex tasks that
may require a whole encounter or series of encounters with other people.
e. Multi-functionality: Typically, CPs are multiply purposeful and multi-functional.
This complexity applies to CPs at all levels, but perhaps most conspicuously to
those at more global levels.
Let us for a moment focus on local CPs, in which participants accomplish a communi-
cative task over a limited sequence. As an example of such a local project, we can take
the project of establishing the defendant’s stance on the issue of guilt (admission or
denial), which is a core project within a criminal court trial (Linell et al. 1993). It can

6. Nestedness is akin to the notion of ‘multiple framing’ of Bateson (1972) and Goffman
(1974). However, these authors often applied the notion to whole situations or communicative
activity types (as these will be discussed below).
 Per Linell

be expressed in different ways, as in the following examples (translations from Swedish

courtroom interactions):
(2) DE ERKÄNNER JA (LiCTI: Tema K: A5: J = judge, D = defendant)
1. J: jaha. erkänner eller förnekar John Sigurdsson
2. alla dessa gärningar?
3. D. ja de erkänner ja.
4. J: erkänner?
5. D: ja.
1. J: okay, does John Sigurdsson admit or deny
2. all these deeds? (JS = the defendant)
3. D: yes, I admit it.
4. J: admit?
5. D: yes.
(3) JA HAR JU ERKÄNT DE (LiCTI: Tema K: A36)
1. J: ja, e de riktit de här, John Eriksson?
2. D: ja, ja har ju erkänt de så–
3. J: du erkänner
1. J: okay, is it correct, this, John Eriksson?
2. D: yes, I have admitted it, haven’t I, so–
3. J: you admit it.
(4) DE STÄMMER (LiCTI: Tema K: A21)
1. J: ja, då ska ja fråga dej, John Gregersson, va din
2. inställning e till den här första gärningen?
3. D: de stämmer.
4. J: du erkänner?
5. D: ja.
1. J: okay, then I am going to ask you, John
2. Gregersson, what is your stance with regard to this
3. first act?
4. D: it is true.
5. J: you admit it?
6. D: yes.
A salient feature of the communicative project (CP) of admission/denial is that it re-
quires the participation of two parties, the defendant and the judge, who is chairing
the trial. In other words, the defendant can not accomplish the CP all by himself. Both
Communicative activity types as organisations in discourses and discourses in organisations 

parties enact their contributions in different ways (of which I have shown only a small
sample here). The judge may phrase his question in a way which is both relatively for-
mal and transparent with regard to the purpose of the CP, as in (2: lines 1–2). Alterna-
tively, his query may exhibit features of formal legal language but be less transparent,
as in (4: lines 1–3). A third option is to use more everyday language in a question that
is parasitic on the prosecutor’s prior reading of the charge (not shown here), as in (3:
line 1). Defendants’ replies also vary in various respects. It is characteristic that the
judge always uses a third turn to either confirm the admission (or denial) or ask for a
The CP of establishing the defendant’s stance in the court trial has several features
that appear to be generalisible to other kinds of local CPs. These features include the
following. First, a local CP is a joint accomplishment consisting of complementary ac-
tions, two parties making different contributions over a sequence of (mainly) initia-
tory and (mainly) responsive actions. Secondly, a local CP occurs in a dialogically es-
tablished (and in this case: normatively specified) position, here: directly after the
prosecutor’s reading the charge (which explains the initial okay’s7 and the anaphoric
references in the judges’ first turns in (2–4)). Thirdly, the participatory actions by par-
ties are typically asymmetrical; they are different in content, quantity and interactional
meaning (in our examples, J is the instigator, and he is more explicit than D). And fi-
nally, the local project is embedded, nested, within a much larger project, in this case:
the whole court trial, and ultimately within even larger projects, the judicial process in
the individual case and in general.
All the Examples (1–4) given so far have been focused on quite local communica-
tive projects. But there are much larger CPs, covering whole social encounters, or
series of encounters. Such comprehensive contacts may be conceptualised in terms of
their dominant purposes and projects, too. For example, Bredmar & Linell (1999), in
a basically conversation-analytic study, analysed series of encounters between mid-
wives and pregnant women within maternal health care. We argue in that article that
there are some recurrent themes running through all these encounters, for example,
the idea that pregnancy is a natural process and that various problems and symptoms
be conceptualised in terms of ‘normality’. A superordinate goal of this overall CP, at
least on the part of the midwife, is therefore reassurance: inducing self-confidence on
the part of the expectant mother in her capacity to carry through pregnancy and

7. Here, however, the English translation (okay) corresponds to several different Swedish re-
sponse particles (jaha, ja).
 Per Linell

6. Communicative activity types

A central notion in the analysis of talk-in-interaction is that of ‘communicative activity

type’. It is convenient to introduce it with some relevant quotations:
I assume that when individuals attend to any current situation, they face the ques-
tion: “What is it that’s going on here?” Whether asked explicitly, as in times of
confusion and doubt, or tacitly, during occasions of usual certitude, the question
is put and the answer to it is presumed by the way the individuals then proceed to
get on with the affairs at hand. (Goffman 1974: 8)

In particular, I take the notion of an activity type to refer to a fuzzy category

whose focal members are goal-defined, socially constituted, bounded, events with
constraints on participants, setting, and so on, but above all on the kinds of allow-
able contributions. Paradigm examples would be teaching, a job interview, a jural
interrogation, a football game, a task in a workshop, a dinner party, and so on.
(Levinson 1992: 69; italics in original)

Knowledge of the conversational activity entails expectations about possible goals

or outcomes for the interaction, about what information is salient and how it is
likely to be signalled, about relevant aspects of interpersonal relations, and about
what will count as normal behavior. (Gumperz 1982: 101)

Goffman, Levinson and Gumperz are all hinting at the same notion, what I will call
‘communicative activity type’ (CAT). Examples of CATs are criminal court trials, psy-
chotherapy sessions, calls for emergency assistance, classroom lessons, job interviews,
focus group discussions etc, but also various kinds of so-called ‘ordinary conversa-
tions’ (see also quotation from Levinson above). CATs are kinds of large overarching
communicative projects (CPs). At the same time, CATs – like other CPs – may them-
selves be embedded within projects and activities that are not primarily communica-
tive in nature. Also, CPs vary widely in size, e.g. that of lighting a candle (as in one
example above) to that of giving reassurance to a pregnant woman over a series of
maternal health care encounters (in another example above).
Some of the properties of the prototypical CAT are the following:8
i. it is related to a social situation and encounter, whose nature is recognised by par-
ticipants and often has a conventional name (i.e. there is a folk concept tied to it,
e.g. “job interview”);

8. There are many concepts and terms in the literature which have a considerable family re-
semblance with CAT: language game (Wittgenstein 1958), situated activity system (Goffman
1974), activity type (Levinson 1992 [1979]; Gumperz 1982; Sarangi 2000), speech event (Hymes
1972), activity, activity system (neo-Vygotskyan activity theory; Engestrom, 2000), activity lan-
guage (Allwood 2000, 2001), joint activity (H. Clark 1996), speech genre (Bakhtin 1986), com-
municative genre (Luckmann 1989, 2002; Marková 2001).
Communicative activity types as organisations in discourses and discourses in organisations 

ii. it is framed by specific expectations and purposes;

iii. although there are often different and sequentially ordered subactivities (= phases),
each instantiation is temporally contiguous within the situation, and involves (at
least partly) the same primary participants;
iv. some are linked to, and administered by, institutions, specific profession(al)s and
societal organisations.
CATs can be analysed in many conceptual and empirical dimensions. It is not feasible
to provide a comprehensive account here. Very roughly, however, one could group
dimensions under three headings, in three families of concepts (only some of which
are given here):
– framing dimensions demarcating the specific CAT: situation definitions in terms of
(prototypical) purposes and tasks, activity roles, scenes, times and medium, spe-
cific activity language (Allwood, 2000) and in general, the role of language (central
vs. subsidiary) within the overall activity,
– internal interactional organisations and accomplishments (within the specific
CAT): phase structure, core communicative projects, agenda, topics, turn organi-
sation and feedback patterns, topical progression methods (e.g. question designs),
dominance patterns, participant positionings, degree of (in)formality, and the role
of artefacts,
– sociocultural ecology (of the specific CAT to other CATs or activity systems): so-
ciocultural history, relations to societal organisations and societal sectors, to larg-
er activity systems and neighbouring activity types, positions in chains of com-
munication situations, as well as hybridities, and discrepancies in participants’
It is of course difficult, if not impossible, to sort these aspects into two or three families,
since they are so clearly overlapping and interdependent. But very roughly, framing
dimensions are basically pre-given, either as physical resources or as culturally deter-
mined premisses, which are “brought along” into new situations. However, they be-
come relevant only if participants invoke them in the actual interaction, that is, orient
to them (to varying extents) and make them relevant (again). Some of them can oc-
casionally be actively bracketed or even ignored in the single instance, and they may of
course change with time, within longer sociocultural traditions. (What is seen as a
proper court trial, for example, may vary across cultures and times.) Yet, on the whole,
they cannot be easily changed in the single instance.
The second family of aspects, the ‘interactional accomplishments’, are things that
are necessarily “brought about” in situ; they differ between concrete instances, al-
though there are of course emergent patterns that become characteristic of the com-
municative activities considered as types. The third category, the ‘sociocultural ecology’,
concerns larger surroundings, which (I argue) are necessary for the full understanding
of CATs, although these larger contexts have often been neglected, particularly in CA
studies (cf. Arminen 2000, 2005).
 Per Linell

By way of summary, CAT is a meso-concept, providing a link between situated mi-

cro-processes and societal macro-structures. It links the ‘interactional order’ (Goffman,
1983) with the ‘institutional order’ (Sarangi & Roberts, 1999), thus giving substance to
considerations of “organisations in discourses” and “discourses in organisations”.

7. An example: Phone calls in train traffic control

As an example of a CAT, I shall use phone calls between train drivers and train dis-
patchers. The official name for these phone calls, as applied to the Swedish national rail
net, is ‘safety calls’ (Sw. säkerhetssamtal), but I shall adopt the term ‘train traffic control
calls’ (TTC calls) (Andrén, Sanne & Linell 2010). According to written regulations,
such calls must be made in order to regulate certain tasks, such as confirming train
meetings on stations of non-automated single-track lines, entries into automated sec-
tions, changes of train identification numbers, time-table modifications, temporary
speed limitations, changes of routes and even destinations, etc. TTC calls are also used
for drivers’ reporting problems and emergencies or dispatchers’ passing on informa-
tion on movements of other trains on the same routes.
The above-mentioned are the official purposes and some of the factual functions
of the TTC calls. One significant feature of this kind of communicative activity is that
it requires formality in at least some of its constituent tasks. In general, a certain type of
communicative project can be considered to be “formal” if
a. it must obligatorily be carried out on every occasion when a specified type of situ-
ation is at hand, and
b. it must be carried out in a particular manner:
b.i. in a specific position within a pre-defined sequence, and
b.ii. in a certain linguistic form (“formal”) (although of course allowing for minor
variations in performance), irrespective of what the specific circumstances in
the individual cases are (i.e. even if some aspects would be completely pre-
dictable in the specific individual case, they have to be included).
As we will see, however, TTC calls are not free from informalisations in various as-
pects. I shall point to some of these in the following exemples (for a more systematic
account of the issue of informalisations, see Andrén et al., op. cit.).
The first example is a fairly straightforward case of a driver’s arrival notification
(here, and below, the speaker initial C stands for train dispatcher or line Controller,
and D for Driver).
1. ((telefonton/ringsignal))
2. C: fjärren i Järnberga
3. D: ja tjena föran på nittifem noll åtta
Communicative activity types as organisations in discourses and discourses in organisations 

4. C: hejsan hejsan
5. D: äntligen ankommit till Åkersby
6. C: äntligen inkommit till Åkersby ja
7. D: ja
8. C: eh tu- eh tu- eh tackar Svensson så mycke här då ja
9. D: Strand här ja
10. C: Strand ja
11. (0.5)
12. D: ja
13. C: tack hej
14. (0.5)
15. D: tack hej
16. C: hej
1. ((telephone rings))
2. C: the remote in Newbury
3. D: yeah hi (tjena), the driver on ninety-five zero eight
4. C: hi there hi (hejsan hejsan)
5. D: at last arrived Oakdale
6. C: at last arrived in Oakdale yes
7. D: yes
8. C: uh thou- uh thou- uh Svenson’s thanking so much
9. here yeah (ja)
10. D. Strand here yeah (ja)
11. C: Strand yeah (ja)
12. (0.5)
13. D. yeah
14. C: thanks bye
15. (0.5)
16. D: thanks bye
17. C: bye
This excerpt starts with the train dispatcher (or line controller = C) answering the
phone and identifying himself in line 2. (“The remote” is a literal translation of the col-
loquial Swedish expression fjärren, an abbreviation of fjärrblockeringscentralen, ‘the
remote blocking centre’, i.e. the dispatch centre.) The driver D acknowledges this by a
greeting and his own self-identification (giving his train service identification number;
line 3). The dispatcher then reciprocates the greeting in line 4, whereupon the driver
introduces his reason for calling in line 5. Lines 1–4 can be said to constitute the open-
ing of this particular call.
 Per Linell

In general, the opening sequence of a normal TTC call, considered as a CAT, consists
of a summons-answer pair, followed by two other mutual actions (identifications, greet-
ings), each often consisting of paired components. In terms of turn design and sequen-
tial organisation (the distribution of turns), we often find the following sequence:
1. recipient’s identification (cf. 5: line 2)
2. caller’s greeting + identification (5: line 3)
3. recipient’s greeting (5: line 4)
According to this pattern, we have a mixture of serial organisation (one action (by one
party) per turn) and moderately compressed, interlocking organisation (two actions
per turn). However, as we shall see, the sequence can often be considerably more com-
pressed than this: the parties can rush into the main activity, and there may be more
than two actions performed in the same turn.
The main activity of a TTC call is governed by one or several of a limited number of
reasons for calling (we will soon see a few examples). The chief categories of train dis-
patchers’ reasons for calling are either (a) mainly proactive: informing about decisions on
cancelled train meetings, changes of train identification numbers or destinations, etc.,
(which requires recipient actions from the driver, such as filling in a form with reading
back (repetition) according to a prescribed routine), or (b) mainly reactive to something
that has happened: reporting problems, incidents and emergency situations, providing
useful information on the development of problematic situations, e.g. informing the driv-
er about features of the overall traffic situation in the area, location of ongoing mainte-
nance work on the rails etc. The chief reasons for drivers to call, on the other hand, are
also either (c) mainly proactive: issuing a formal request of certain actions (measures)
and/or obtaining permission (clearance), for example, to pass a red light signal under
certain circumstances, or (d) mainly reactive: reporting the arrival at predefined locations
(stations) or the preplanned meeting with another train, or reporting problems, incidents
or emergency situations relating to one’s own train vehicle or the traffic situation affecting
the train service, i.e. aspects that can be presumed to be unknown to the dispatcher.
The sequential layout of the main activity usually covers more turns than the
opening sequence above:
4. stating the reason for calling, consisting of the caller’s message initiation, plus
(if the message is short, e.g. arrival notification as in Example (5)) its delivery;
5. (in cases of more complex tasks:) carrying out the main task, e.g. the dispatcher’s
dictation of a message to be entered by the driver on a pre-printed form and the
latter’s repetition of this (accompanied by writing) (as we will see, this often in-
volves several subphases); note that this necessarily involves a turn exchange se-
quence (both parties are active);
6. recipient’s acknowledgement of receipt of message, plus signing of it (by giving
one’s family name, usually accompanied by writing);
7. caller’s signing.
Communicative activity types as organisations in discourses and discourses in organisations 

These four turns, or turn exchanges, usually receive at least short acknowledgements
by the recipient underway.
Finally, the closing sequence is usually short and comprises two, or sometimes
three, exchanges of paired components:
8. mutual thanking
9. (optional) (unilateral or mutual, often routinised) assessments (e.g. A: that’s fine,
B: okay)
10. mutual leave-taking (bye, bye)
Even if Example (5) is a short call, it is relatively exemplary, with all the obligatory
phases present. However, in many of our TTC calls, there is quite often a compression
of actions, so that several prescribed actions are “batched” within the same turn, as in
Example (6), which is also an arrival notification:
1. ((telefonton/ringsignal))
2. C: fjärren Järnberga
3. D: ja hejsan de här va förarn på ått- åttisju arton då va de
4. ankomstanmälan i Åsta å mitt namn va Johansson
5. C: Johansson där och Englund här dö
6. (0.5)
7. D: Englund?
8. C: ja
9. (0.8)
10. D: ja tack
11. C: då tackar vi för dä
12. D: mm [hej]
13. C: [okej]
1. ((telephone rings))
2. C: the remote Newbury
3. D: yeah hi this was the driver on eigh- eighty-seven
4. eighteen then it was arrival report in Oldtown an’
5. my name was Johansson
6. C: Johansson there and Englund here du
7. (0.5)
8. D: Englund?
9. C: yeah
10. (0.8)
11. D: yeah thanks
12. C: then we say thanks for that
 Per Linell

13. D: mm [bye]
14. C: [okay]
Here, the driver manages to accomplish a greeting, his own self-identification, the
communication of the main message, and the signing in one single turn (lines 3–5).
The next example is a slightly more complicated call, in which the dispatcher gives
an order by dictating material which the driver is obliged to copy into a printed form,
called “S 16”:
1. ((telefonton/ringsignal))
2. C: Fjärren i Järnberga
3. D: mmja nittifem förtifem
4. C: ja hejsan hejsan (.) [hej]
5. D: [hej]
6. C: då [ska vi se]
7. D: [S sexton sa du]
8. C: S sexton ja precis=
9. D: .hja
10. C: baksidan där då
11. D: baksidan
12. C: ja (..) .hh till tåg nittifem förtifem då ja
13. D: ja
14. C: dagens datum noll fyra noll tre sjutton=
15. D: <noll fyra noll tre sjutton>
16. C: så hoppar vi ner till eh avdelning tjugoett där
17. D: tjugoett ja=
18. C: ja och Bertil då K-mötet med tåg nittifem åttitvå nitt-
19. nittifem sjuttiosex
20. D: <nittifem sjuttisex>
21. C: i Söderås bortfaller
22. D: <ja>
23. C: dä blir inställt idag där då så
24. D: a just dä
25. (0.5)
26. D: då ä de till tåg nittifem förtifem::: noll fyra noll
27. tre sjutton och sen tjugeett B (.) nittifem sjuttisex
28. i Söderås bortfaller
29. C: de ä korrekt ja klockan är tretton noll fyra då o ja
30. heter Quist
31. D: <Quist> (0.5) ja heter Eriksson
32. C: Eriksson ja
Communicative activity types as organisations in discourses and discourses in organisations 

33. D: mm
34. C: tackar så mycke där ja
35. D: tack ska du ha
36. C: a okej=
37. D: hej
39. C: hej
1. ((telephone rings))
2. C: the remote in Newbury
3. D: mm yeah ninety-five forty-five
4. C: yeah hi there (.) [hi
5. D: [hi
6. C: then [let’s see
7. D: [S sixteen you said
8. C: S sixteen yeah exactly=
9. D: = .yeah (.hja)
10. C: the back side there then ((referring to the sheet))
11. D: the back side
12. C: yeh (0.5) for train ninety-five forty-five then
13. yes
14. D: yes
15. C: date of the day zero four zero three seventeen=
16. D: <zero four zero three seventeen> ((writing))
17. C: then we hop down to uh section twenty-one there
18. D: twenty-one yes
19. C: yes and Bert ((i.e. “B”)) then K-meeting
20. with train ninety-five eighty-two ninet-
21. ninety-five seventy-six
22. D: <ninety-five seventy-six>
23. C: in Danby drops off
24. D: <ye:s>
25. C: that will be cancelled today there so
26. D: yes exactly
27. (0.5)
28. D: then it’s for train ninety-five forty-fi::ve zero
29. four zero three seventeen and then twenty-one B
30. (.) ninety-five seventy-six in Danby drops off
31. C: that’s correct yeah. the time is thirteen zero
32. four then an’ my name is Quist
33. D: <Quist> (0.5) my name is Erikson
 Per Linell

34. C: Erikson yeah

35. D: mm
36. C: thank you so much there okay
37. D: thank you too
38. C: yeah okay=
39. D: bye
40. C: bye
Example (7) begins with an abbreviated opening (the driver is calling back after an
interruption) and a mutual naming of the task (lines 7–8). The train dispatcher C then
guides the driver D to the appropriate slots of the form (lines 10 “back side”, 12 “train”,
15 “date”, 17 “Section 21”, 19 “Bert”) and dictates pieces of information to be entered,
and D repeats certain parts of this in a confirmatory manner (lines 11, 16, 18, 22).
After that, the driver makes a comprehensive repetition in one turn (lines 28–30),
which is acknowledged by C (lines 31–32), and then follow notification of the exact
time, signings, and the usual closing sequence.

8. TTC calls as a communicative activity type

If we set ourselves the task to account for TTC calls in terms of a CAT, there are of
course many more points to bring up than can be done here and now. I will select a few
of the points that were enumerated in Section 6.
Phase structure: A CAT analysis is concerned with the overall organisation of an
encounter and its embedded discourse. For example, doctor consultations in primary
care (Heritage & Maynard 2006) could be analysed into a number of phases that form
fairly comprehensive projects of their own. The authors argue that such consultations
exhibit the following phase structure (“overall structural organisation”): Opening, Pre-
senting complaint (reason for seeing the doctor), Examination (verbal, physical), Di-
agnosis, Treatment (discussion), Closing, with some phases being susceptible to fur-
ther subdivision.
In the case of our TTC calls, the actual performance consists of a limited set of sub-
activities that are sequentially realised, basically as follows (see also Section 7 above):
Message (reason for calling), e.g.
Order-giving + repetition (by installments)
Comprehensive repetition
Communicative activity types as organisations in discourses and discourses in organisations 

What I have called “message” here is obviously the main activity, that is, the core with-
out which there would not have been any TTC call. The openings and closings, and to
some extent the signings, are subsidiary to this main activity.
Sequence structure (turn sequences): Below the phase level, there are basically
paired actions, sometimes expanded into longer sequences. Some of these sequences
are often compressed into what we have called “batchings”. One can interpret these
sequences as realisations of local communicative projects.
Activity roles: The two parties have clearly different activity roles. Depending on
the reasons for calling, they complement each other in an asymmetrical division of
communicative labour: the train dispatcher decides, gives orders, the driver requests
permissions, repeats dictated orders, etc.
Agenda: There is a clear action agenda (rather than a fixed topic agenda), which
can be either driver-initiated or dispatcher-initiated, and in each of these two catego-
ries, the reason for calling can be either reactive (reporting something) or proactive
(instructing or requesting actions). Of course, the latter division must not conceal that
reporting actions have projective functions, too (providing necessary information for
the recipient’s future actions), and that instructing or requesting actions have respon-
sive aspects, too (some kind of problem must be solved).
Hybridity: The TTC calls have specific transactional functions, which are often
prescribed by rules and define the reasons for calling. But this is not all there is; these
talks do not only have transactional functions. TTC calls are also a special activity
embedded within a wider organisation or community with social relations between
members, and this leads to certain informalisations. The parties often know each other
personally, they have frequent contacts over the phone, and the phone calls are the
main channel for social contact between them.
Accordingly, there are various features of informalisation in the TTC calls. The use
of colloquial language and professional jargon belongs here, as well as the phenome-
non of batching. But in addition, there is often a mixture of transactional and social-
relational talk. When the CAT is actually implemented, and there is time available,
participants engage in various kinds of relational talk, which is only indirectly related
to professional tasks and work. Indeed, Example (1) above comes from one conversa-
tion in our corpus (TTC 13); C and D in (1) are in fact a train dispatcher and a train
driver, making a joint digression from the TTC main activity.
Sociocultural history: There are many things that could be brought up under
this heading (Sanne 2001; Andrén 2005). Let me just mention in passing that TTC
calls are historically modelled on military communication, and they still exhibit
some traces of this (as can perhaps be gleaned already from the few examples I have
given above).
 Per Linell

9. Communicative activities: Types and hybridities

Let us now return to the concept of communicative activity type as such, on a more
general level. As already indicated, the notion is akin to that of ‘communicative genre’
(Luckmann 1989, 2002; Marková 2001, 2003). However, ‘genre’ appears to be a concept
originating in studies of texts, literature and the arts, whereas communicative activities
are directly linked to actions, social situations and social encounters.
The framing of a CAT is a situation definition, a set of assumptions guiding parties’
expectations and interpretations of what may happen in the situated encounter. We
may also think of it as an interactional contract, usually implicit and sometimes rene-
gotiable, which governs participants’ rights and obligations in the communicative ac-
tivity of the situated encounter. While participants are guided and constrained by
CATs, they also actively recreate, negotiate, stretch and play with them. In other words,
they can bend the rules. So if CATs involve typification and structuring, agents are not
forced to act in accordance with prescribed rules, not even when formality (as defined
above) is the norm. It remains important to keep in mind that it is the actual interac-
tional patterns that ultimately make up the real CATs.
There is a ‘double dialogicality’ (Linell 1998) in communicative practices: we can
talk about dialogue in situations (à la CA), and dialogue between traditions (in a
Bakhtinian fashion). First, there are the situated interactions themselves: the interac-
tions there-and-then between participants, and between their orientations to framings,
the concrete situational environments and participants’ actual accomplishments. At
the same time, it is in and through these interactions that participants deploy and play
with CATs, i.e. with the situation-transcending (trans-situational) practices that con-
stitute sociocultural traditions. CATs bridge between events and recurrent practices.
As I noted earlier, ‘CAT’ should be regarded as a bridging meso-concept between
‘the interactional order’ and ‘the institutional order’ of talk in particular situation
types. On the one hand, you can not situate a concrete analysis of most phenomena in
talk-in-interaction, such as the use of certain sequence types, turn designs, phase or-
ganisations or action types (say different kinds of ‘formulations’; Drew 1998), without
determining the hosting CATs in which they appear. On the other hand, you can hard-
ly account for what goes on in a societal sector, such as society’s judicial apparatus, the
educational system (comprehensive schools, universities, etc), commercial companies
or even governments (cf. Boden 1994), without specifying what CATs they comprise.
A theory of CATs may seem to suggest that the world of talk consists of clearly
distinct social situation types. But this is just the point of departure. In real social life,
there are lots of mixed activities. Some activities have only been vaguely solidified into
types, and many are hybridities of types (Sarangi 2000). Although the topic of hy-
bridities cannot be covered in this paper, the point is so important that it must be
mentioned. Briefly, there are at least three kinds of hybridity:
Communicative activity types as organisations in discourses and discourses in organisations 

a. sequential type: In many kinds of social gathering, we have first CAT1, then CAT2,
then CAT3, etc. That is, the interactional encounter is different in different phases.
Phases are (by definition) their own (subordinate) CATs. Sometimes, phases are
independent activities. For example, a dinner party may consist of, say, mingling
with drinks, seated dinner, and dancing, but these “phases” are ordered, and have
a loose linking, e.g. they involve, for instance, the same participants. A more ex-
treme example is the circus varieté performance; yet, this too is temporally con-
tiguous, and has the same audience throughout.
b. frames within frames; Here we have two or several framings, one embedded with-
in the other. This applies, for example, to H. Clark’s (1996) role plays, i.e. the phe-
nomenon he calls ‘layering’. Empirically, such situations can be studied in cases
such as theatre performance rehearsals and education in drama. In such activities,
only one single framing tends to be oriented to in each single moment; for exam-
ple, when participants find themselves in preparations for the lesson, in episodes
of instruction or discussion of how to enact the play and the actual playing out of
selected scenes from the drama script, the respective activities belong to divergent
frames and different moments. Yet, these frames are all embedded within the the-
atre education lesson (Rönn 2009).
c. merged types: Here, parties are orienting to several frames simultaneously, some-
times trying to reach a kind of compromise. These activities encompass many
training and practising situations which simultaneously involve teaching and in-
struction; examples are practising a language and being taught that language at the
same time (Gustavsson 1988), simulated job interviews within unemployment
programs for young people (Linell & Persson Thunqvist 2003), and simulated pri-
mary care consultations with medical students in training (Thomassen 2005).
Well-established activity types too, such as police interrogations and the TTC calls we
have seen here, involve hybridities. In police interrogations, the police officer is not
only the interviewer working on the case, but sometimes also therapist, counsellor and
moral educator (Gunnerød 2005). Focus groups (Marková et al. 2007) oscillate be-
tween free discussions and decision-making activities, etc.

10. A note on the role of texts in CAT analysis

The main bulk of the contributions to this volume deal with written language or texts
(as well as electronic discourse). Mine is an exception, dealing only with talk-in-inter-
action. This is also the home base of CAT analysis. However, the use of texts is often an
integrated aspect of CATs, and this fact must be taken into account. In fact, as we have
seen, TTC calls involve some reading and writing, particularly in the instance of the
filling in and use of the “S16 form” in (7).
 Per Linell

The dialogical perspective would emphasise the situated and interactive use in the
dynamic production and interpretation of texts and images (cf. also Kress & van Leeu-
wen 1996): if you want: the study of text events and text practices, rather than texts per
se. Texts are written, read and used by participants in interaction, in different ways
within particular CATs and for various specific purposes. We often read and use texts
in ways that can only be understood as part of quite particular activities. Not all texts
are read in their entirety, nor can all reading activities be seen as autonomous, the lat-
ter being a situation approximated when, for example, we read a book without any
instrumental intentions to use it in particular ways.

11. CAT analysis: A two-step analysis?

Is there a specific analytic method that could be called “Communicative Activity Type
Analysis”? Would it be different from Conversation Analysis (CA)?
I do not think that such questions can receive a clear “yes” or “no” answer. CAT
analyses can study many specific details, although of course no single study can go into
all dimensions at the same time. CATs can be researched with the help of both qualita-
tive and quantitative methods (see, for example, as regards presidential news confer-
ences, Clayman et al. 2007). What one can definitely say is that a “CAT analysis” must
theorise the overall structures and functions of communicative situations in at least
some respects.
One must acknowledge that CA is the most worked-out methodology for analysing
talk-in-interaction in detail, and as such it lacks real competitors, and its descriptive
conceptual apparatus can hardly be ignored. Moreover, CA is dialogical too in several
respects. The analysis of CATs should be seen as a natural extension of CA. Most CA
studies do acknowledge the importance of CATs (Drew & Heritage 1992: 22), but they
hardly ascribe the same kind of central importance to the notion, as when we appro-
priate it as a (or even the) bridging meso-concept in discourse theory.
Many scholars have discussed the role of contexts in discourse analysis. It would
be wrong to propose that CA ignores contexts; rather, it insists on the requirement that
analysts invoke contextual resources in their analyses only if they show that partici-
pants make them relevant in and for their discourse. Yet, the understanding of many
CATs requires ethnographic or sociolinguistic knowledge on the part of analysts
(Cicourel 1981; Sarangi 2000; Arminen 2000, etc.). For example, we can hardly analyse
or understand TTC phone calls, in the absence of any knowledge of the CAT as part of
a wider context.
Orthodox Conversation Analysis, on the other hand, marginalises ethnographic
knowledge. At one level, this is in accordance with the strict demands for method-
ological rigour (Schegloff 1998: “discipline”). At the same time, this rather acontextual
approach may have to do with the fact that many CA practitioners have studied a
rather special family of CATs, namely, so-called ordinary conversations from the
Communicative activity types as organisations in discourses and discourses in organisations 

researchers’ own native culture. But TTC talks are not native for most discourse ana-
lysts. This also applies to court trials, primary-care doctor consultations etc.; in order
to analyse them, one must learn about their organisational surroundings.
Actual accomplishments remain the basic thing in CAT analysis, as it has been in
CA. But we must realise that the analyst’s task is different from the participants’, and the
analyst’s understanding will be enriched if the studies encompass organisational and so-
cio-historical contexts. Accordingly, CAT analysis may be thought of as a two-step analy-
sis, first a thorough close analysis of the data (which should be performed by scrutinising
the data many times), then the interpretation of the data in a wider context, dictated by
the theoretical interest and purpose of the study (e.g. Linell & Luckmann 1991: 18). I
would venture to contend that something like this is being may seem to claim done in
applications of CA too, even if the methodological meta-talk often the opposite.

12. Back to ‘dialogical theory’.

At last, a few words more about the dialogical bedrock of discourse analysis. One
might ask oneself: How basic is actually dialogue? The answer will depend on whether
we take ‘dialogue’ in a concrete or an abstract sense. For sure, dialogue in a concrete
sense, defined roughly as interaction through language and other semiotic means be-
tween two (or more) mutually co-present participants, as for example in a conversa-
tion, is important empirically, and it can serve as a model and metaphor more gener-
ally. But what is important and explanatory in an epistemology, or even ontology, for
the human sciences, is dialogicality on a more abstract level.
The essential (and mutually interdependent) properties of dialogicality at this ab-
stract level are roughly these:
Dynamics. Human life involves active problem-solving on different time scales, at
least on the meso- and micro-levels. The study of the microgenesis of interaction fo-
cuses on the development of communicative projects on a moment-to-moment basis.
This is to some extent the opposite of structural determinism; people are not struc-
tural dopes, simply enacting static and prescribed roles, but agents who orient to ex-
pectations that they have accumulated but also changed over time (Heritage 1984). In
doing so, they actively (though often routinely; Schegloff 1986) solve (major or minor)
communicative problems (tasks).
Interactionality. Sense-making in thinking and communication always occurs in
interaction with (actual or virtual) others and with the world. As I said in the begin-
ning, this implies a focus on intersubjectivity.
Contextuality. Dialogism includes a more radical theory of contexts and interac-
tion than most variants of linguistic pragmatics; it claims that contexts are always cru-
cially relevant, and co-develop with discourse.
Dialogical theory covers not only the externalised situated interaction (as in CA), but
also the internal dialogue in thinking, and the interaction within situation-transcending
 Per Linell

practices (through different kinds of orientations to them). Thinking is assumed to be

partly analogous to external interaction (internalisation of dialogue; cf. e.g. Billig
1987), and it is interdependent with (internal and) external artefact use; thinking be-
comes different when the thinker uses notation: notes/marks on paper, computer soft-
ware and interfaces, for numerical calculation or for the formulation of linguistic texts.
(Again, these are things that I have left out in this chapter.)
Dialogical theory is therefore an attempt at an integrated theoretical framework
for the understanding of mind, interaction and society, and of language, discourse,
communication and thinking, as well as interventions into the world. In lieu of a seg-
regated, autonomous linguistics, we must aim for interdisciplinarity.


Allwood, J. 2000. An activity-based approach to pragmatics. In Abduction, Belief and Context in

Dialogue: Studies in Computational Pragmatics, H. Bunt and B. Black (eds), 47–80. Amster-
dam: John Benjamins.
Allwood, J. 2001. Capturing differences between social activities in spoken language. In Perspec-
tives on Semantics, Pragmatics and Discourse, I. Kenesi and R. M. Harnish (eds), 301–319.
Amsterdam: John Benjamins.
Andrén, M. 2005. Formalitet och informalitet i säkerhetssamtal mellan lokförare och tågklare-
rare. Unpublished Master’s thesis from tema K, Linköping University.
Andrén, M., Sanne, J. M. and Linell, P. 2010. Striking the balance between formality and infor-
mality in safety-critical activities: Train Traffic Control Calls. Journal of Pragmatics, 42:
Arminen, I. 2000. On the context sensitivity of institutional interaction. Discourse & Society 11:
Arminen, I. 2005. Institutional Interaction. Studies of Talk at Work. Aldershot, England: Ashgate.
Bakhtin, M. M. 1986. Speech Genres and Other Late Essays. Transl. by V. McGee. Austin: Univer-
sity of Texas Press.
Billig, M. 1987. Arguing and Thinking: a rhetorical approach to social psychology. Oxford: Blackwell.
Boden, D. 1994. The Business of Talk: Organization in Action. Cambridge: Polity Press.
Brandist, C., Shepherd, D. and Tihanov, G. (eds). 2004 The Bakhtin Circle: In the Master’s Ab-
sence. Manchester: Manchester University Press.
Bredmar, M. and Linell, P. 1999. Reconfirming normality: The constitution of reassurance in
talks between midwives and expectant mothers. In Talk, Work and Institutional Order: Dis-
course in Medical, Mediation and Management Settings. S. Sarangi and C. Roberts (eds),
237–270. Berlin: Mouton de Gruyter.
Cicourel, A. 1981. Notes on the integration of micro- and macro-levels of analysis. In Advances
in Social Theory and Methodology: Toward an integration of micro- and macro-sociologies, K.
Knorr-Cetina and A. Cicourel (eds), 51–80. London: Routledge & Kegan Paul.
Clark, A. 1997. Being There: Putting Brain, Body, and World Together Again. Cambridge, MA:
MIT Press.
Clark, H. H., 1996. Using Language. Cambridge: Cambridge University Press.
Communicative activity types as organisations in discourses and discourses in organisations 

Clayman, S., Heritage, J., Elliott, M. and McDonald, L. 2007. When does the watchdog bark?
Conditions of aggressive questioning in presidential news conferences. American Socio-
logical Review, 72: 23–41.
Czarniawska, B. 1997. Narrating the Organization: Dramas of institutional identity. Chicago:
Chicago University Press.
Drew, P. 1998. An exercise in the comparative analysis of talk-in-interaction in different
(institutional) settings: the case of formulations. In Samtalsstudier, H. Lehti-Eklund (ed.)
(Meddelanden från Institutionen för nordiska språk och nordisk litteratur, B:19), 29–42.
Helsinki University.
Drew, P. and Heritage, J. (eds). 1992. Talk at Work. Cambridge: Cambridge University Press.
Engestrom, Y. 2000. From individual action to collective activity and back: developmental work
research as an interventionist methodology. In Workplace Studies: Recovering Work Practice
and Informing System Design, P. Luff, C. Heath and J. Hindmarsh (eds), 150–166. Cambridge:
Cambridge University Press.
Fetzer, A. 2010. (this volume) Contexts in context: micro meets macro.
Goffman, E. 1974. Frame Analysis. Cambridge, MA: Harvard University Press.
Goffman, E. 1983. The interaction order. American Sociological Review 48: 1–17.
Gumperz, J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.
Gunnerød, M. O. 2005. Politiavhøret som samtale. Hovedfagsoppgave i anvendt språkvitenskap,
Institutt for språk- og kommunikasjonsstudier, NTNU, Trondheim.
Gustavsson, L. 1988. Language Taught and Language Used. (Linköping Studies in Arts and Sci-
ence, 18). Linköping: Department of Theme Research.
Heritage, J. 1984. Garfinkel and Ethnomethodology. Oxford: Polity Press.
Heritage, J. and Maynard, D. (eds). 2006. Communication in Medical Care. Cambridge:
Cambridge University Press.
Hymes, D. 1972. Models of the interaction of language and social life. In Directions in Sociolin-
guistics: The Ethnography of Communication, J. Gumperz and D. Hymes (eds), 35–71. New
York: Holt, Rinehart and Winston.
Jefferson, G. and Lee, J. 1992. The rejection of advice: managing the problematic convergence of
a “troubles-telling” and a “service encounter”. In Drew and Heritage (1992): 521–548.
Kress, G. and van Leeuwen, Th. 1996. Reading Images: The Grammar of Visual Design. London:
Levinson, S, 1992. Activity types and language. In Talk at Work, P. Drew and J. Heritage (eds),
66–100. Cambridge: Cambridge University Press. First published in 1979 in Linguistics 17:
Linell, P. 1998. Approaching Dialogue: Talk, interaction and contexts in dialogical perspectives.
Amsterdam: John Benjamins.
Linell, P. 2005. The Written Language Bias in Linguistics: Its Origin, Nature and Transformations.
London: Routledge.
Linell, P. 2006. Towards a dialogical linguistics. In The XII International Bakhtin Conference:
Proceedings, M. Lähteenmäki, H. Dufva, S. Leppänen and P. Varis (eds), 152–167. Jyväskylä:
Department of Languages.
Linell, P. 2009. Rethinking Language, Mind and World Dialogically: Interactional and contextual
theories of human sense-making. Charlotte, NC: Information Age Publishing.
Linell, P., Alemyr, L. and Jönsson, L. 1993. Admission of guilt as a communicative project in
judicial settings. Journal of Pragmatics 19: 153–176.
 Per Linell

Linell, P. and Luckmann, T. 1991. Asymmetries in dialogue: Some conceptual preliminaries. In

Marková and Foppa (1991): 1–20.
Linell, P. and Persson Thunqvist, D. 2003. Moving in and out of framings: Activity contexts in
talks with young unemployed people within a training project. Journal of Pragmatics 35:
Luckmann, T. 1989. Prolegomena to a social theory of communicative genres. Slovene Studies
1–2: 159–167.
Luckmann, T. 1995. Interaction planning and intersubjective adjustment of perspectives by
communicative genres. In Social Intelligence and Interaction, E. Goody (ed.), 175–186.
Cambridge: Cambridge University Press.
Luckmann, T. 2002. On the methodology of (oral) genres. In Jagen och rösterna: Goffman, Vive-
ka och samtalet. (SIC, 42), P. Linell and K. Aronsson, K. (eds), 319–337. Linköping: Tema
Marková, I. 2001. Social representations and communicative genres. In Penser la vie: le social, la
nature. Mélanges en l’honneur de Serge Moscovici, F. Buschini and N. Kalampalikis (eds),
219–235. Paris: Éditions de la Maison des sciences de l’homme.
Marková, I. 2003. Dialogicality and Social Representations. Cambridge: Cambridge University
Marková, I. and Foppa, K. (eds). 1990. The Dynamics of Dialogue. New York: Harvester
Marková, I. and Foppa, K. (eds). 1991. Asymmetries in Dialogue. Hemel Hempstead: Harvester
Marková, I., Linell, P., Grossen, M. and Salazar Orvig, A. 2007. Dialogue in Focus Groups. Lon-
don: Equinox.
Pomerantz, A. 1986. Extreme case formulations: a way of legitimizing claims. Human Studies 9:
Rönn, M. 2009. “Det är inte förrän man gör det som man förstår!” Om kommunikativa hinder vid
en teaterhögskola. (‘“You can’t understand until you do it.” On Communicative Obstacles at
a Drama College’) (Stockholm Studies in Scandinavian Philology, New series, 48).
Sanne, J. 2001. Fart och spänning – banarbetare, lokförare och tågtrafikledare om risk och riskhan-
tering. Linköping: Tema T Rapport 27, 2001.
Sarangi, S., 2000. Activity types, discourse types and interactional hybridity: the case of genetic
counselling. In Discourse and Social Life, S. Sarangi and M. Coulthard (eds), 1–27. Harlow:
Sarangi, S. and Roberts, C. 1999. The dynamics of interactional and institutional orders in work-
related settings. In Talk, Work and Institutional Order: Discourse in Medical, Mediation and
Management Settings, S. Sarangi and C. Roberts (eds), 1–57. Berlin: Mouton de Gruyter.
Schegloff, E. A. 1986. The routine as achievement. Human Studies 9: 111–152.
Schegloff, E. A. 1998. Reply to Wetherell. Discourse & Society 9: 413–416.
Schutz, A. and Luckmann, T. 1962. The Structures of the Life-World. Vol. II. Evanston, Ill.: North-
western University Press.
Searle, J. 1969. Speech Acts. Cambridge: Cambridge University Press.
Communicative activity types as organisations in discourses and discourses in organisations 

Soler-Gallart, M. 2004. Reading to share: accounting for others in dialogic literary gatherings.
In Aspects of the Dialogue Self (International Cultural-historical Human Sciences, vol. 11),
M.-C. Bertau (ed.), 157–183. Berlin: Lehmanns Media.
Thomassen, G. 2005. Den flerbunnete treningssamtalen: En studie av samtaler mellom pasient
og student fra sykepleier- og medisinerutdanning. (Doctoral Theses at NTNU, 2005:61).
Trondheim: NTNU.
Wittgenstein, L. 1958. Philosophical Investigations. Translation by G.E.M. Anscombe of Philoso-
phische Untersuchungen (1953). 2nd edn. Oxford: Blackwell.
Discourse and context
in a historical perspective
On courtroom interaction in Salem, 1692

Risto Hiltunen

The article stresses the importance for historical discourse linguistics of

contextualising texts, using the Salem witchcraft documents of 1692 as an
example. A contextual analysis presupposes taking into account a number
of variables with a potential impact on the discourse. Here I discuss the
following issues: (a) the historical, social and cultural context, involving
not only the historical background and social structures of Salem, but also
considerations of the “invisible” world of witchcraft; (b) the legal context,
involving the genre conventions shaping the records; (c) the scribal context,
involving the representation of original spoken discourse in writing, and (d) the
communicative context, involving the discourse strategies of the actual trials.
Such considerations will be crucial both for the interpretation of the documents
and for understanding the verbal behaviour of the parties at the trials.

Pray do not accuse me, I am as clear as your worships; you must do right judgments.
(George Jacobs, Sr., SWP 475)

1. Introduction

The axiomatic statement by Mellinkoff (1963: vii), describing the law as “a profession
of words”, captures the crux of the inalienable relationship between language and law.
At the same time it draws attention to the importance of examining the relationship
between the two. Since the 1960s, the linguistic study of legal documents has emerged
as a field of study in its own right. During its relatively short history a shift of focus has
taken place in this research, whereby the earlier emphasis on the linguistic properties
of written legislative language has shifted to the spoken discourse of the courtroom.
Such a movement is another reflection of the trend in linguistics towards the study of
language use in real-life contexts.
 Risto Hiltunen

Like written law, the oral process of litigation in the courtroom is a manifestation of
a verbal activity. It is equally dependent on words, often on finding the most appropriate
word(s) for a given context. If the social importance of legislative writing is underscored
by its special linguistic form, setting it apart from texts in other domains, the institu-
tional nature of oral courtroom proceedings is similarly distinguished by its own verbal
conventions. In addition, the physical context of the courtroom itself provides an insti-
tutional frame for the interaction. Thus the verbal conventions of legal discourse and the
institutional circumstances together contribute to the construction and deconstruction
of meaning for the parties to a legal process (cf. Lakoff 1990: 92ff.; Maley 1994: 32).
In principle, court cases constitute testing-points for the relevance of legal state-
ments encoded in the law. Since the assessment takes place through language, legal
proceedings may be viewed as constituting chains of spoken linguistic events address-
ing the matter from the different perspectives of the participants (cf. Gibbons 1994: 3).
The proceedings are characterized by certain rituals, which have to be observed in
order for the procedural linguistic acts to be performed successfully. The process usu-
ally involves several stages, from the preliminary hearings of witnesses to the trial and
passing of the verdict. Consequently, the language of all parties, especially that of the
defendant, plaintiff, witnesses and judge, will be relevant to the study of courtroom
discourse as a communicative process.
Nowadays, especially with reference to the American legal scene, with such a pow-
erful adversarial system of justice, trials are often discussed in terms of two approach-
es, both of which involve looking at the process metaphorically. Thus, according to
Maley and Fahey (1991: 3), the more traditional view is to look upon the trial as a
battle between the parties, with the discourse organized in terms of strategic choices
relating to ‘defence’, ‘resistance’, ‘aggression’ and ‘attack’. The other, more recent, view
involves the metaphor of story-telling, whereby the trial is viewed as a discursive pro-
cess rather than a purely legal one and the decision will involve not so much the ap-
plication of law to facts as an evaluation of different narrative stories. As a result, the
party that is able to come up with the more convincing story will also be the likely win-
ner of the trial.
From a discourse-analytic viewpoint, an interesting question concerns the lin-
guistic correlates of a convincing case. A notable strand in the earlier research of trial
discourse addressed specifically the question of which linguistic features support cred-
ibility and which undermine it (O’Barr 1982). As a result, the distinction was intro-
duced between ‘powerful’ and ‘powerless’ speaking styles, characterized by certain
phonological, syntactic, semantic and lexical features. In a courtroom situation, a
powerful presentation is allegedly marked for example by narrative unity, as against
fragmentation in the more powerless one. Overall, powerless speech tends to be judged
as less intelligent, convincing and trustworthy than powerful speech (Levi 1990: 18).
In such an image-building process the role of the legal professionals will be significant,
for they may significantly contribute to reinforcing or undermining a witness’s credi-
bility, for example through their (cross-)examination strategies.
Discourse and context in a historical perspective 

On the other hand, although the distinction between powerful and powerless per-
formance is a helpful tool for analysing interaction in the courtroom for the linguist,
the court will need consider the ethical question of to what extent it allows presenta-
tional skills to interfere in its decisions, for looking at a trial in terms of powerful or
powerless performance alone may lead a biased interpretation.

2. Purpose

When we consider such aspects of the legal context from a historical perspective, a
number of questions arise for the investigator, ranging from basic issues concerning
data to ones of comprehension and interpretation. In studying earlier historical peri-
ods we are at the mercy of what contemporaries have considered important enough to
be recorded in writing at the time, and how well those records have survived to the
present day. In view of such contingencies, which are beyond our control, the docu-
ments of the Salem witchcraft trials form a particularly interesting source of data, com-
prising a rich and well-preserved collection of documents representing the different
phases of the legal process. In particular the surviving original manuscripts are a valu-
able resource, providing first-hand information that is not available from printed edi-
tions. Among other things, they yield important information about the recorders in-
volved in the process of creating the documents, in terms for example of the number of
recording hands, the division of labour among the recorders, and their individual writ-
ing habits. Recent research into such issues has shown how the evidence based on the
original documents may add significantly to our understanding of how, when and why
the documents came into being (Grund 2007a, 2007b; Hiltunen and Peikola 2007).
The purpose of this article is to address the issues of contextualizing historical
texts, using the Salem documents as a point of reference. ‘Context’ is understood in a
broad sense to refer to “all elements of a communicative situation, the verbal and non-
verbal context, the context of a given speech situation and the social context of the
relationship between the speaker and hearer, their knowledge, and their attitudes”
(Bussmann 1996: 100). As pointed out by Fetzer (1994: 2), context is particularly im-
portant for understanding what is meant by a particular turn or utterance in a given
situation, i.e. for capturing its pragmatic meaning. In order to be able to grasp prag-
matic meanings in discourses of the past, we need to contextualise the material as
fully as possible in terms of both text-external and text-internal information. This in-
volves looking at texts both from the ‘outside’, as products situated in a particular ex-
ternal context, and from the ‘inside’, as discourses reflecting their situational environ-
ment. Since contextual information about the past is likely to be defective and
fragmentary, the process of contextual reconstruction is correspondingly likely to be
problematic. In the case of Salem, we are talking about a series of events taking place
over three hundred years ago, which is a long enough period for many contextual fac-
tors to have either been changed or lost completely. At the same time, we are talking
 Risto Hiltunen

about an incident that has been much researched during the intervening time, espe-
cially as regards the language-external historical context. On the other hand, the mate-
rial remains under-researched as regards its use as data for the study of courtroom
discourse in its New England setting at the close of the seventeenth century. This as-
pect has only recently been brought up for discussion within the framework of “his-
torical courtroom discourse” (cf. Kryk-Kastovsky 2006 and 2007). Several of the re-
cent contributions have been inspired by the collaborative project for a new edition of
the entire Salem witchcraft material (see Rosenthal et al. 2009).

3. Synchronic and diachronic aspects

Text and context belong together and interact with each other. According to Widdow-
son (2004: 53), for example, “it is only when the linguistic features of the text are re-
lated to contextual factors that discourse is realized.” In everyday interaction, it is re-
markable how much contextual information we silently take for granted, without
necessarily being aware of its contextual implications. In reading about a news item, for
example, we focus in the first instance on what happened rather than why it happened.
As new items highlight actions, the text can afford to be much less specific about any
background information without risk of being misunderstood. As a rule, readers are
able to process the content adequately because they are capable of relating pieces of
information to previous discourses on the subject and to relevant functional and com-
municative frames. In other words, context tends to be an unmarked property of com-
munication by virtue of being based on a set of shared assumptions and a shared sche-
mata about the world in any community of interactants (cf. Widdowson 2004: 42).
The interplay between text and context applies equally to historical texts, but there
the picture is nevertheless essentially different. Not only may the time and place of
composition of the text be unknown, but information as to who wrote the text and for
what audience may also have to be inferred or hypothesised by the researcher
(see Carroll et al. 2003: 3). When such crucial background is missing, such basic infor-
mation as the function of a text may remain obscure for the modern reader. Recover-
ing contextual information thus often becomes a marked activity in the attempt to ac-
cess discourses of historical documents. The process usually involves a step-by-step
process of reconstruction whereby new pieces of contextual information supplement
an earlier view; this process, over a period of time, may result in new interpretations of
the material. However, there are likely to remain significant gaps in the process, which
may or may not be filled in the course of time. This is the more likely, the more ancient
the material. The Salem material, despite the large quantity of surviving documents,
also contains such gaps of missing contextual information. The records nevertheless
have a great deal to offer the present-day scholar. Let us next take a look at selected
issues of context and their impact on the discourse, proceeding from the macro-level
of the historical setting to the micro-level scene of the actual trials.
Discourse and context in a historical perspective 

4. Contexts of the Salem documents

4.1 Historical, social, and cultural aspects

The events leading to the trials that started in the Salem Village parsonage in January
1692 were unusual but not unique, for there had been similar incidents in New England
before (Demos 1982; Rosenthal 1995). The beginning of the crisis followed a recurrent
pattern, involving a group of young girls – in this case, Elizabeth Parris and Abigail
Williams – falling suddenly into violent ‘fits’, which were diagnosed as symptoms of
witchcraft and regarded as the Devil’s interference in the life of the village. Other girls
and young women soon showed similar symptoms, and before long they were blaming
their neighbours and friends for their afflictions, accusing them of bewitching them
(cf. Gibson 2003: 207). The next phase also followed an established pattern: those ac-
cused of causing the afflictions were women of low social status, some also known for
their erratic habits in the community. The first person to be named was Tituba, the
slave of the local minister Samuel Parris: a woman of uncertain ethnic background, but
called “an Indian woman” in the documents (Norton 2002: 20–21). She was joined by
two other women, Sarah Good and Sarah Osborne. They had aroused suspicion,
among other things, by being prone to using foul language and scolding their hus-
bands. Such deviant traits of behaviour made them potentially obvious targets to be
singled out as witches. The three women were examined in early March, 1692, by the
Salem magistrates John Hathorne and Jonathan Corwin.
In the meantime, the group of accusers acquired new members, who came up with
further names, and by the end of May 1692 there were already 68 people in prison. The
accusers no longer limited themselves to people of lower social status, but on the con-
trary implicated even highly respected members of the community (Hoffer 1997: 66–67).
At this point, the Governor of Massachusetts, William Phipps, set up a special court,
called the Court of Oyer and Terminer, to deal with the burgeoning number of witch-
craft cases. The court passed the first death sentence, on Bridget Bishop, on 2 June. The
crucial evidence against her, and others following her to the gallows, was ‘spectral’: in
other words, the court believed the accusers when they claimed that it was the spirit of
Bridget Bishop that had attacked and hurt them. In the courtroom the afflicted formed
an active group of accusers, seized by terrible fits as soon as the person they are accus-
ing was brought in, only to be restored to their normal state by the touch of the accused.
The communication of the accusers is mainly non-verbal and physical (i.e. they act
through fits), but they may also speak, for instance when they are asked whether they
recognize the accused person or when they tell the court that they are seeing the spectre
of the accused. The accusers could also be made unable to speak by the spectre of the
accused, especially when faced with a question they did not want to answer.
The anomalous behaviour of the accusers signals the crucial importance of the
religious context for the Salem material. The Puritan community was founded upon a
firm belief in the divine and a commitment to following God’s word. In accordance
 Risto Hiltunen

with their faith, people generally also believed in the existence of witchcraft. They
thought that the Devil made pacts with witches, whose spells, curses and evil eyes
could harm innocent believers. It was also common knowledge that witches were par-
ticularly inclined to choose children as their victims (Hoffer 1997: 36). For the au-
thorities, the fact that children were the first to be affected in Salem confirmed the
existence of witchcraft in the community. They were convinced that the physical tor-
ments of the afflicted were genuine, and thus accepted them as evidence of the guilt of
the accused. This is also one of the reasons why the incident at Salem developed into a
much bigger process than any of the earlier ones. The consequences are well known.
Altogether 185 individuals (141 women and 44 men) were accused. Nineteen (thirteen
women and six men) were hanged and one was tortured to death by pressing; 55 con-
fessed, thus avoiding death, and some 150 ended up in jail, where four adults and one
infant died (Hill 2000: xv).
The historical context of the Salem trials is that of seventeenth-century colonial
America, and the establishment of permanent settlements in New England as a result of
the Puritan mass exodus from England in 1630–1640. The arrival of the Pilgrim Fathers
in Plymouth in 1620, the settlement of Massachusetts Bay Colony in 1628 and of Salem
itself a year later were still within living memory when the witchcraft incident broke out
in 1692. Direct references to the historical background of the community, however, are
largely missing in the documents, partly due to the shared knowledge of the context,
partly to the fact that the documents are geared to the moment of recording. The only
obvious link to the larger context is the voice of the mother country, the English Crown,
present for instance in the opening and closing formulas of indictments. The opening
formula is typically in Latin, while the latter is in English (italics added):
(1) Anno RR’s & Reginae Gulielmi & Mariae Angliae & c Quarto Annoq’e Domini
and also for Sundry other Acts of Witchcraft by the said Mary Bradbury Com-
mitted and done before and since that time against the peace of our Sov’r Lord
& Lady King and Queen theire Crowne and dignity And the forme Of the Stat-
ute In that case made and Provided. (Mary Bradbury, SWP 115)
Regarding the social structure of the Salem community, its material foundation con-
sisted of farming, fishing, and trade. The documents also make clear the presence of
distinct social classes, including slaves. The witchcraft process cut across the whole
society, although it was the lower end of the social scale that was most affected. In
spite of extensive historical study of the Salem community in the light of early colonial
history and the witchcraft documents, there is still no consensus as to what ultimately
triggered the outbreak. Many potentially valid motives have been proposed, including
personal and social conflicts among social groups and individual families (Boyer and
Nissenbaum 1974), gender issues (Karlsen 1987), and the hostilities between the set-
tlers and the Wabanaki Indians in the frontier regions of New Hampshire and Maine
Discourse and context in a historical perspective 

(Norton 2002). In support of the latter argument, Norton adduces evidence from the
documents for the idea of “an alliance between Satan and the Wabanakis”, as conveyed
through the frequent reference to “the black man” in the records (Norton 2002: 59).
Another interesting example is an expression, common in the depositions and ex-
amination records, describing the way the spirit of the accused is attacking the depo-
(2) That Goody Johnson and Goody falkner appeared at the same tyme and
threatened to teare her in peeces if she did not doo what she then did. (Mary
Barker, SWP 60)
The phrase of “tearing to pieces” is thought by Norton (2002: 135–136) to have been
picked up by locals from eye-witness narratives by people who had seen scalped and
dismembered bodies of victims in the Indian raids. Such instances of a possible causal
connection with the Indian wars are very interesting in indicating how certain highly
specific details of the discourse may be significant in reconstructing the historical
course of events.
At the same time, the fact that such references are incidental rather than system-
atic reflects the fact that the documents were written for the legal purposes of the
court, whose main concern is to establish the guilt or innocence of the accused. The
documents are therefore silent as to the causes of the outbreak, unless we regard the
works of the Devil, which are described in considerable detail, as an explanation. For
those involved in the trials, however, it was precisely the Devil that lay behind the
events. The outbreak of witchcraft and other disasters hitting the community at this
time were seen as God-sent punishments for the sins of its people. The individual
voices in the documents were responding to the question of ‘what happens’ rather than
‘why it happens’. There was thus no need to look for other reasons, beyond noting the
obvious interference of the Devil.
Outside the legal context, there must also have existed less formal ways of talking
about witchcraft. In everyday situations such talk was probably widespread and consti-
tuted an important phase of discourse creation by the speakers. Such interaction, how-
ever, is difficult to recover from documents that were written predominantly according
to the conventions of formal genres. We get occasional glimpses of the existence of
such contexts for example in the depositions, where people relate accounts of their
personal witchcraft experiences. Such passages indicate indirectly what kinds of sub-
jects were discussed in the community. They must have included, along with the all-
important question of who in the neighbourhood was (or was not) a witch, the behav-
iour of suspect witches, the spectral visits of alleged witches to the neighbours’ houses,
the shapes of the Devil, and promises of serving the Devil and renouncing one’s bap-
tism. The question of how best to respond to the magistrates’ accusations, in case one
were to be charged and summoned before the court, would undoubtedly also have
been constantly debated. The following examples, involving the verb to tell, indicate
the occurrence of conversations about such issues preceding depositional statements.
 Risto Hiltunen

(3) Her brother urged her to confess, and told her that, in so doing, she could not
lie: to which she answered, `Good brother, do not say so; for I shall lie if I con-
fess, and then who shall answer unto God for my lie?’ (Martha Tyler, SWP 777)
(4) But being asked who told her there were threescore Witches, she answered she
Could not tell. Mary Warren affirmd that Her Father having promised to mow
the grass for her if he had time, which he not doeing she came to the house,
and told him he had better he had done it, presently after that Her Sister fell ill
and shortly after Her Mother was taken ill, and dyed (Alice Parker, SWP 623)
Such topics, as exemplified above, would also have provided a rich source for gossip,
perhaps the most effective way of spreading news in the community at the time (Norton
1992: 6). Anything that was newsworthy, especially if it was negative or implied an-
other person’s involvement in witchcraft, would quickly spread from house to house.
In the light of the documents, the proportion of misinformation that was circulated in
this way must have been considerable. Much of it, judging from the evidence provided
by the documents, was also fatal. For example, it is likely that the names of the people
who were accused of witchcraft by their fellow community members in the deposi-
tions and examination records were not brought up randomly, but were based on ru-
mours being passed around in the community.

4.2 Legal aspects

Witchcraft had been defined a criminal act in England ever since the Middle Ages, but
the Salem judges followed the 1604 Act of Parliament of King James I “against Conju-
ration Witchcrafte and dealinge with evill and wicked Spirits” (I Jas.I, c.12), which
extended the death penalty to acts of witchcraft (cf. Gibson 2003: 5). There are no ref-
erences in the Salem documents to this Act by name, but the use of certain phrases and
formulae for example in indictments, the most formal of the genres in the Salem mate-
rial, suggests an intertextual link between the two. For example, the effects of malefi-
cium are expressed in the Act with phrases, such as “use practise or exercise”, and
“killed destroyed wasted consumed pined or lamed”. The indictments make use of the
same expressions, sometimes omitting one or two of the items or changing their inter-
nal order.
It is also worth noting in this connection that in December, 1692, when the witch-
craft incident was essentially over in Salem, the text of the 1604 Act was introduced
officially with some modifications as a Bill with the same title, and “read orderly in
This house of Representatives and voted and passed in the Affirmative & Sent to his
Excellency the Governor & Councill for Consent”. The document is signed by William
Bond, the speaker of the House, and William Phipps, Governor of the Massachusetts
Bay Colony (cf. SWP 885–6). According to Rosenthal, the reason for this late publica-
tion of the text was that after the Court of Oyer and Terminer had ended, there was
still a need to reaffirm the reality of witchcraft (Rosenthal et al. 2009: 24).
Discourse and context in a historical perspective 

In terms of legal procedure, the trajectory for an accused person from a complaint
to the gallows took the following stages (with the number of surviving documents for
the major categories in brackets): (a) complaints (16 documents), (b) warrants/mitti-
muses (89), (c) examinations (67), (d) depositions (402), (e) indictments (81), (f) grand
jury inquests (which determined, on the basis of witness depositions and other docu-
ments, whether the charge warranted a trial (Billa vera) or not (Ignoramus), common-
ly indicated on the reverse of the relevant indictments), (g) trial records of the Court of
Oyer and Terminer, and (h) death warrants (cf. Grund, Kytö and Rissanen 2004: 149).
The records are more representative for some of these categories than others, with
depositions making up the greater part of the data. The documents for (g) are missing,
since the trial records of the Court of Oyer and Terminer have been lost (cf. Trask
1992: xx), and (h) is represented by only two documents (SWP 108, 378). The surviv-
ing data contain basically two kinds of documents. First, there are those written in
seventeenth-century legal discourse; these are interesting in terms of the definition of
witchcraft, the history of legal phrases from earlier sources, and the varied ways in
which individual recorders organize the subject matter in the documents. Secondly,
there are the documents representing oral interaction between the parties involved;
these consist mainly of examination records and depositions. Here I focus on this part
of the material, especially the examination records, which are the most interesting as
far as the voices of the Salem inhabitants are concerned. First, however, a few words on
the role of the recorders and on scribal hands will be in order.

4.3 Scribal aspects

The research for the new edition of the Salem records (Rosenthal 2009) has provided
new information about the recording hands and the written transmission of the docu-
ments. As a result, we now have a better understanding of the identity of the recorders,
their role and status in the community, and their working methods as recorders. A
surprisingly large number of people were involved in the process. So far more than two
hundred different hands have been identified. The output of individual recorders var-
ied: some produced tens of documents, others considerably fewer, some just one or
two. About a hundred hands, including some of the major ones, have been success-
fully identified by name, but more than half are still unidentified, and many are likely
to remain anonymous. The recorders’ technical skills vary considerably, from experi-
enced, professional hands to those with more limited writing skills. Research into this
aspect of contextualisation continues, with special reference to the identity of the re-
corders, by means of so-called scribal profiling; this involves philological and palaeo-
graphical analysis, focusing on recording techniques, practices of document produc-
tion, and the involvement of recorders in the legal process (for further discussion see
Grund, Kytö and Rissanen 2004; Grund 2007a, 2007b; Hiltunen and Peikola 2007).
 Risto Hiltunen

4.4 Communicative aspects

Some of the most exciting documents in the Salem material are the examination re-
cords. The purpose of an examination is to decide whether the person on trial is guilty
and is to be sentenced, or not guilty, in which case she/he is to be released. The form of
the Salem examinations suggests that the speech of the examiners and defendants is
reproduced as actually spoken. There are many striking examples that seem to have
captured genuine features of spoken language, such as those of “pidgin” English in the
speech of Candy (SWP 179): Candy no witch in her country. Candy’s mother no witch.
Candy no witch, Barbados. This country, mistress give Candy witch. (cf. Rissanen 1997).
However, such reporting is always relative to the accuracy of the recorders. As Trask
(1992: xx) points out, events are seen “through the eyes, writing style and prejudice of
the original recorder, whose perception of the reality of that time is not reality itself.”
While it was no doubt in the interest of the recorders to produce as accurate records of
the exchanges as possible, documents of this type always involve a degree of scribal
intervention (Grund, Kytö and Rissanen 2004: 150–151). In preparing the documents,
the recorders may have made use of a variety of techniques (see Grund 2007a: 3–5;
2007b: 124–126). Mostly, they were probably working on the basis of notes taken down
during the hearing. The clean copies would normally be prepared after the sessions,
using the notes and from memory. Although the intention would have been to record
all turns in full, in practice this was difficult to achieve. The rapid flow of speech would
in many cases have rendered the notes incomplete to start with. The occasional switch-
es to indirect discourse in the records, and the frequent use of summaries of interactive
turns, may also reflect such difficulties. Scribes sometimes also inserted their own
comments in the records. Interference of this kind may have affected the representa-
tion of the accused in the record, for better or worse. The recorders were not necessar-
ily impartial observers of the proceedings, but might also have a personal stake in the
case under consideration. Thus the documents represent the discourse in the court-
room as perceived through the eyes of the recorders. Nevertheless, they provide – as
Trask (1992: xxi) observes – the “best evidence” we have available in its original form.
Let us next take a look at the communicative context of the examination records, and
consider the discourse strategies employed by examiners and defendants.

4.4.1 Strategies of the examiners

A striking feature of the examination records is the magistrates’ frequent conviction
that the defendant is “guilty but unwilling to confess” (Archer 2002: 1). Determined to
extract confessions from the defendants, they maintain their conviction throughout
the sessions, often ignoring what the accused has to say in her/his defence. This nega-
tive presupposition structures both question and answer sequences to a considerable
extent: the questions posed by the magistrates, for example, are frequently framed as
accusations (Archer ibid.). The negative presupposition is most obvious in wh-ques-
tions, which are typically leading questions. In yes/no questions, which are more
Discourse and context in a historical perspective 

coercive, it is less obvious, but even there the context strongly suggests that the exam-
iner expects a positive answer. The following examples come from the examination of
Sarah Good (SWP 356):
(5) a. wh questions:
Sarah Good what evil spirit have you familiarity with
why doe you hurt these poor children
how come they are thus tormented
b. y/n questions:
have you made no contract with the devil
c. both
Sarah good doe you not see now what you have done why do you not tell us
the truth, why do you thus torment these poor children
These examples indicate the asymmetric power relations characteristic of the trials.
The examiner is in total control of the proceedings: he holds the floor, asks the ques-
tions, voices the accusations and issues directions to the defendant – often in a single
turn, as in (5c). The questioning strategy remains the same throughout the examina-
tion. In addition, very often the responses of the accusers, the audience, and even the
recorders strongly support the examiner’s standpoint and reasoning. In this situa-
tion of considerable psychological pressure, the options available to the defendant
are very limited.

4.4.2 Strategies of the defendants

In a modern context, there are various alternatives open to the defendant in respond-
ing to an allegation. According to Tiersma (1999: 150–151), one strategy is to contest
the relevance of the accusation, another to challenge the truth by denying the facts.
The defendant may also offer a counter-narrative that might require the court to re-
solve the dispute in the defendant’s favour. Finally, the defendant can admit that the
plaintiff ’s narrative is both legally adequate and true, but offer an excuse or justifica-
tion for their behaviour. These variations also occur in the Salem examination records.
For our purposes, however, it will be practical to consider them in terms of two alter-
native strategies: denial and confession. Defendants pleading ‘not guilty’

Trials where the defendant pleads ‘not guilty’ are usually more complex for the prose-
cution than confessions, because they entail a basic conflict between the truth para-
digms of the participants. When it comes to Salem, it needs to be remembered that
while in modern trials defendants will have lawyers speaking on their behalf, the ac-
cused in Salem had no access to such intermediaries. The proceedings are therefore
technically simpler for the prosecution, but correspondingly harder for the defendants.
It is both the subject matter of the trials and the special setup of the court in Salem that
 Risto Hiltunen

together account for the highly polarized division of the examinations into denials and
confessions. It is likely that most defendants would normally have made up their minds
as to whether to admit or deny the accusation of witchcraft in advance of their hearing.
As regards the examiners, the records leave no doubt that they were fully convinced of
the guilt of the defendants.
The act of denial has different variations in this scenario, which also involve differ-
ent communicative purposes. First, there are those who categorically reject the accusa-
tions, leaving little or no chance for the examiners to persuade them to confess. This
kind of determination entails a strategy of absolute face-saving. It is exemplified by the
records of such defendants as Martha Carrier (SWP 184), Martha Corey (SWP 248), and
George Jacobs, Sr. (SWP 474). Their defence offers examples of turns aimed at challeng-
ing the court by contesting the relevance of the questions and/or the truth of the accusa-
tions. The following example is from the examination of George Jacobs, Sr. (SWP 474),
whose desperate words were quoted at the beginning of this paper, indicating the frus-
tration of a defendant who realizes that his pleas will not be heard, his words will not be
considered, and that he is about to be pronounced guilty and sentenced to death:
(5) Here are them that accuse you of acts of witchcraft
Well, let us hear who are they, and what are they.
Abigail Williams – Jacobs laught.
Because I am falsely accused. – Your worships all of you do you think this is
Nay, what do you think?
I never did it.
who did it?
Don’t ask me.
Jacobs openly challenges his accusers by laughing at the court, asking a counter-ques-
tion, denying his involvement and indicating that he thinks the questions are irrele-
vant. Jacobs was hanged on August 19, together with three other men and one woman
(Rosenthal 1993: 108).
The second group of denials consists of records that indicate a more cooperative
attitude on the part of the defendants. This is manifested for instance in the way the
names of other persons are introduced into the discourse, with an implied allegation
of involvement. The examination of Sarah Good (SWP 356) exemplifies such a pattern.
The defendant first consistently rejects every accusation, whether explicit or implicit.
She even comes up with counter-arguments (but I am falsely accused). In (6), “H”
stands for “Hathorne”, the examiner, and “g” for “Good”, the defendant:
(6) (H) Sarah good doe you not see now what you have done why doe you not tell
us the truth, why doe you thus torment these poor children (g) I doe not tor-
ment them, H who do you imploy then (g) I imploy nobody I scorn it (H) how
came they are thus tormented, (g) what doe I know you bring others here and
Discourse and context in a historical perspective 

now you charge me with it (H) why who was it. (g) I doe not know but it was
some you brought into the meeting house with you (H) wee brought you into
the meeting house (g) but you brought in two more (H) Who was it then that
tormented the children (g) it was osburn (H) what is it that you say when you
goe muttering away from persons houses (g) if I must tell I will tell (H) doe tell
us then (g) if I must tell I will tell, it is the commandments I may say my com-
mandments I hope (H) what commandment is it (g) if I must tell you I will tell,
it is a psalm (H) what psalm (g) after a long time shee muttered over some
part of a psalm (H) who doe you serve (g) I serve god (H) what god doe you
serve (g) the god that made heaven and earth
In the midst of a series of denials, Sarah Good’s behaviour is provocative in the eyes of
the court. Her face-threatening act I scorn it, uttered baldly on record, further under-
mines her case in the eyes of the magistrates. She finally mentions the name of another
person (it was osburn), suggesting the person responsible for the afflictions. However,
this is also a weak argument, in the sense that while Good is denying her own involve-
ment in the afflictions, she simultaneously admits the existence of witchcraft. Here her
case differs from the consistent denials cited above. In Good’s case, the general nega-
tive attitude towards her is also registered by the recorder, according to whom her an-
swers were in a very wicked, spitfull manner reflecting and retorting aganst the authority
with base and abusive words and many lies, and a hearsay statement quoting the words
of the husband, testifying to her “bad carriage to him” (SWP 357).
In instances such as (5) and (6), it may be asked whether the “battle” metaphor is
at all relevant to examinations where the defendant’s voice is not heard, let alone lis-
tened to. In the Salem trials, the outcome of the battle appears to have been deter-
mined at the outset in favour of the examiner. In terms of whose story is the more
convincing, the situation is equally biased: the magistrates are willing to hear and ac-
cept only one version of the ‘story’, that which corresponds to their own predeter-
mined conviction. Defendants pleading ‘guilty’

When we turn to defendants pleading ‘guilty’, we find more strategic variation in the
discourse. One of the curiosities of the Salem trials is that it was only those who con-
sistently denied the charge of witchcraft that were executed: those that confessed were
saved and suffered ‘only’ imprisonment. The reason behind this unexpected course of
action may have been the idea of keeping the confessors alive as proof of the existence
of witchcraft in the community, and by inference as justification for the death sen-
tences that the court had passed (Rosenthal 1993: 152). As strategic alternatives, both
denial and confession involve face-saving. The choice came to depend on whether, in
addition to saving face, defendants also wanted to save their lives.
The records show a marked increase in confessions after the first few months of
the incident, when the “safety-in-confession concept” (Rosenthal, ibid.) became
 Risto Hiltunen

common knowledge. The execution of Bridget Bishop in early June is likely to have
served as an important signal in this regard; Bishop did not confess and was hanged,
whereas those who confessed at the same time were saved. Whether or not to confess
was no doubt a source of much discussion in the community and also a great deal of
anxiety and uncertainty among the defendants. One indication of the troubled cir-
cumstances is the case of Samuel Wardwell (SWP 783), who had initially confessed to
witchcraft but later retracted his confession (see Rosenthal, 1992: 155–156).
If denial was an uncooperative strategy from the examiners’ point of view, confes-
sion entailed cooperation. In the trials, confession could be reached through various
stages. In the documents, the process is recorded variously in either direct or reported
discourse. The mode of reporting depends to some extent on when the record was
made. The early confessions are usually in direct discourse, while the later ones tend to
be in reported discourse. In the most straightforward case, the accused immediately
pleads guilty. This is usually recorded in reported discourse (7), less commonly in di-
rect discourse (8):
(7) He confesses he has been in the snare of the devil three years (William Barker,
Sr., SWP 65)
(8) What say you? Are you guilty, or not? Speak the truth. I will speak the truth. I
have seen sights and been scared. I have been very wicked. I hope I shall be
better, if God will help me. (Abigail Hobbs, SWP 405)
Confession can also be introduced by a denial as the first step. In the later records such
instances usually also appear in reported discourse, indicating that the initial denial
was seen by the examiners only as an (un)necessary step preceding confession. For the
defendants, however, it was still an important face-saving device, as in the following
(9) After many questiones and negative answers returned and her Stricking Down
of severall of the afflicted persons with her looks, she was Desyred to tell the
truth in this matter She then said that ... (Mary Toothaker, SWP 767)
In the confessions reported in direct discourse, the initial stage is categorical denial;
gradually, however, as a result of persistent questioning, the defendant (who may al-
ready at the outset have decided to confess) will begin to make concessions. This phase
of the examination usually consists of a step-by-step procedure: the defendant admits
one detail at a time, until the magistrate is satisfied that the confession is sufficiently
comprehensive and convincing. Sometimes this takes a long sequence of turns to com-
plete, as in the case of Deliverance Hobbs (SWP 419–420). The accused first denies any
involvement in maleficium (10). As the examination proceeds, there is a shift from
negative to positive in the answers. This happens simultaneously with a change from
y/n questions to wh-questions, cf. (11). In the final phase (12) practically all questions
Discourse and context in a historical perspective 

of any type are answered in the positive, most importantly those concerning conspir-
ing with the Devil:
(10) Why do you hurt these persons? It is unknown to me.
How come you to commit acts of Witchcraft? I know nothing of it.
(11) It is said you were afflicted, how came that about? I have seen sundry sights.
What sights? Last Lords day in this meeting house & out of the door, I saw a
great many birds cats & dogs, & heard a voice say come away.
(12) Have you signed to any book? It is very lately then.
When was it? The night before the last.
Well the Lord open your heart to confesse the truth. Who brought the book to
you? It was Goody Wilds.
What did you make your mark with in the book? Pen and ink.
Who brought the Pen and Ink? They that brought the book, Goody Wilds.
Did they threaten you if you did not signe? Yes, to tear me in peices.
Was there any else in company? No, Sir.
This record contains a total of 41 questions. Of these, eight are overt denials, occurring
in the early part of the record, while some thirty can be interpreted as confessions. In
some instances the answer is not clear. At the end of the questioning, the magistrates
are satisfied that she has made a complete confession.
The records also show other paths to confession (cf. Doty and Hiltunen 2002),
including accusing other persons and blaming the Devil for what happened. Such
moves tend to be followed by an act of repentance and a firm promise not to commit
the sin of witchcraft ever again. Such defensive strategies typically minimise the in-
volvement of self and maximise the involvement of other. Some individuals manage
the questioning better than others, one of the most successful being the slave woman
Tituba. She was the first to confess, thereby setting an example of how to save one’s life
for those who wanted to do so, even at the risk of losing eternal salvation by making a
false confession. These individuals, however, are not generally remembered, with the
exception of Tituba, who was a key figure at the beginning of the outbreak. In the eyes
of posterity, it is those who maintained their innocence at all cost, even when faced
with the risk of losing their lives on the gallows, who are remembered as the martyrs
of Salem: in other words, the true “winners” of the trials.

5. Conclusion

The present discussion has focused on two principal aspects of discourse contextual-
ization in a historical perspective, with reference to the Salem witchcraft records from
1692: (a) the surviving trial data, and (b) the deployment of major discourse strategies
in the oral examination of the accused persons. Regarding the data as a whole, the
 Risto Hiltunen

Salem material exhibits a dense network of relationships between text-external and

text-internal worlds, in terms of three different contexts: (a) historical, social and cul-
tural, (b) legal, and (c) scribal. Naturally, the better our understanding of the external,
the more favourable the prerequisites for understanding the internal in the surviving
documents. One key to bridging the two is provided by the genres of writing emerging
in response to the communicative needs of the legal institution. These range from
highly standardized documents, where the institutional voice is predominant, to those
where the voices of the participants, especially the accused, are represented. The latter
materials are discussed in the light of selected examination records. These data open
up possibilities for observing how discourse is constructed in a context where the spe-
cific situational constraints of the courtroom determine strategic choices for both
prosecution and defence.


Archer, Dawn. 2002. “‘Can innocent people be guilty?’” A sociopragmatic analysis of examina-
tion transcripts from the Salem Witchcraft Trials.” Journal of Historical Pragmatics 3. 1–30.
Bussmann, Hadumod, 1996. Routledge Dictionary of Language and Linguistics, transl. and ed.
by Gregory P. Trauth and Kerstin Kazzazi. London and New York: Routledge.
Boyer, Paul and Stephen Nissenbaum. 1974. Salem Possessed: The Social Origins of Witchcraft.
Cambridge, Mass.: Harvard University Press.
Carroll, Ruth, Risto Hiltunen, Matti Peikola, Janne Skaffari, Sanna-Kaisa Tanskanen, Ellen Valle,
Brita Wårvik. 2003. “Introduction.” Discourse Perspectives on English: Medieval to Modern,
ed. by Risto Hiltunen and Janne Skaffari, 1–12, Amsterdam: John Benjamins.
Demos, John P. 1982. Entertaining Satan: Witchcraft and the Culture of Early New England.
Oxford: Oxford University Press.
Doty, Kathleen and Risto Hiltunen. 2002. “’I will tell. I will tell’: Confessional patterns in the
Salem Witchcraft Trials, 1692”. Journal of Historical Pragmatics 3. 299–335.
Fetzer, Anita. 2004. Recontextualizing Context: Grammaticality meets Appropriateness. Amster-
dam/Philadelphia: John Benjamins.
Gibson, Marion. ed. 2003. Witchcraft and Society in England and America, 1550–1750. Ithaca,
New York: Cornell University Press.
Gibbons, John. ed. 1994. Language and the Law. London & New York: Longman.
Grund, Peter. 2007a. “The Anatomy of Correction: Additions, Cancellations, and Changes in the
Documents of the Salem Witchcraft Trials.” Studia Neophilologica 79.3–24.
Grund, Peter, 2007b. “From Tongue to Text: The Transmission of the Salem Witchcraft Exami-
nation Records.” American Speech 82: 2.119–150.
Grund, Peter, Merja Kytö, and Matti Rissanen. 2004. “Editing the Salem Witchcraft Records: An
Exploration of a Linguistic Treasury.” American Speech 79. 146–66.
Hill, Francis, 2000. [1943]. The Salem Witch Trials Reader. New York: Da Capo Press.
Hiltunen, Risto. 1996 “’Tell me, be you a witch?’: Questions in the Salem Witchcraft Trials of
1692”. International Journal for the Semiotics of Law 2. 17–37.
Hiltunen, Risto and Matti Peikola. 2007. “Trial discourse and manuscript context: Scribal pro-
files in the Salem Witchcraft Records.” Journal of Historical Pragmatics 8: 1. 43–68.
Discourse and context in a historical perspective 

Hoffer, Peter C. 1997. The Salem Witchcraft Trials: A Legal History. Lawrence: University of Kan-
sas Press.
Karlsen, Carol F. 1987. The Devil in the Shape of a Woman. New York: W. Norton.
Kryk-Kastovsky, Barbara. ed. 2006. “Historical courtroom discourse: Introduction”. Special Is-
sue on Historical Courtroom Discourse. Journal of Historical Pragmatics 7:2. 163–179.
Kryk-Kastovsky, Barbara. ed. 2007. Special Issue on Historical Courtroom Discourse II. Journal
of Historical Pragmatics 8: 1. 1–68.
Kurzon, Denis. 1994. “Linguistics and legal discourse: An introduction.” International Journal
for the Semiotics of Law 7. 5–12.
Lakoff, Robin T. 1990. Talking Power: The Politics of Language [n.p.]: Harper Collins.
Levi, Judith N. 1990. “The study of language in the judicial process.” In Language and the Judicial
Process, edited by Judith N. Levi and Anne Graffam Walker, 3–35. New York: Plenum.
Maley, Yon and Rhondda Fahey. 1991. “Presenting the Evidence: Constructions of Reality in
Court.” International Journal of the Semiotics of Law 10. 3–17.
Maley, Yon. 1994. “The language of the law.” Gibbons 1994. 3–50.
Mellinkoff, David. 1963 The Language of the Law. Boston: Little, Brown, and Company.
Norton, Mary Beth. 2002. In the Devil’s Snare: The Salem Witchcraft Crisis of 1692. New York:
Alfred A. Knopf.
O’Barr, William, 1982. Linguistic Evidence: Language, Power, and Strategy in the Courtroom.
New York: Academic Press.
Rissanen, Matti. 1997. “‘Candy No Witch, Barbardos’: Salem Witchcraft Trials as Evidence of
Early American English.” Language in Time and Space. Studies in Honour of Wolfgang Vi-
ereck on the Occasion of his 60th Birthday, edited by Heinrich Ramisch and Kenneth Wynne,
183–93. Stuttgart: Franz Steiner Verlag.
Rosenthal, Bernard. 1993. Salem Story: Reading the Witch Trials of 1692. Cambridge: Cambridge
University Press.
Rosenthal, Bernard, Gretchen Adams, Margo Burns, Peter Grund, Risto Hiltunen, Merja Kytö,
Matti Peikola, Bejamic C. Ray, Matti Rissanen, Marilynne Roach, Richard Trask. eds 2009.
The Records of the Salem Witch Hunt, 1692. Cambridge: Cambridge University Press.
SWP = Boyer, Paul and Stephen Nissenbaum. eds 1977. The Salem Witchcraft Papers. Verbatim
Transcripts of the Legal Documents of the Salem Witchcraft Outbreak of 1692. I–II. New
York: Da Capo Press.
Tiersma, Peter M. 1999. Legal Language. Chicago: The University of Chicago Press.
Trask, Richard, B. 1997. The Devil Hath Been Raised: A Documentary History of the Salem Village
Witchcraft Outbreak of March 1692. Danvers, Mass.: Yeoman Press.
Widdowson, Henry G. 2004. Text, Context, Pretext: Critical Issues in Discourse Analysis. Oxford:
Blackwell Publishing.
part 2

Constructing identity across genres

Pronominal choice in French
conversational interaction
Indices of national identity in identity acts

Linda R. Waugh

This paper examines how national identities are co-constructed and can be the
site for struggle in interaction: in particular, how they are affected by general
socio-cultural patterns and ideologies, the topics under consideration, and the
positioning of the interactants. Through a fine-grained analysis of a conversation
in French between three monolingual French speakers and one bilingual speaker
of French and (American) English, shifts in the bilingual speaker’s national
identities are examined in relation to the topics discussed, the speaker’s own
ambivalence about his identities, and the cultural and symbolic capital of one of
the monolingual speakers, who has negative prejudices and stereotypes about
the U.S. and who seizes power in the interaction. The shifts are shown through
identity acts (a sub-type of speech acts) and the use of indefinite/non-specific
pronouns in making generic statements. In particular, the shifts go through three
stages and create an organization for the conversation.

1. Introduction: Identity

In this paper I will address the issue of the interactional co-construction of identity in
informal, natural conversation, taking as my point of departure recent literature on
identity which shows that identities are multiple, dynamic, and contextualized (Antaki
& Widdicombe 1988, Block 2006, Bucholtz & Hall 2004, 2005, Cerulo 1997, Coupland,
Nussbaum & Grossman 1993, Doran 2004, Goffman 1959, Goodwin & Duranti 1992,
Gumperz 1982, Joseph 2004, Milroy 2002, Mendoza-Denton 2002, Schiffrin 1996,
Tabouret-Keller 1997, van Dijk 1989).1 We also know from work in second language

1. I wish to thank members of the “Identity and Ideology Study Group” at the University of
Arizona for all the highly interesting conversations we’ve been having about the issues discussed
here, as well as the audiences at the Organization in Discourse Conference in Turku, Finland, at
my invited talk at Pennsylvania State University, at the Corpus and Applied Linguistics Confer-
ence in Flagstaff, Arizona, at the Conceptual Structure and Discourse in Language conference in
 Linda R. Waugh

acquisition and bilingualism that there are interesting issues around the construction
of a new identity in a new language, the maintenance of sometimes highly contradic-
tory identities, loss of identity in the L1, and so forth (Block 2007, Clyne 2004, Kramsch
2000, Norton 2000, Norton & Toohey 2002, Paulston 2004, Pavlenko & Lantolf 2000).
What we will be addressing here is national identity (see de Cillia et al 1999, Wodak et
al 1999, Joseph 2004), and we will examine how a participant’s national identities are
co-constructed/negotiated in interaction (Jacoby & Ochs 1995) and thus are dialogic
(Bakhtin 1981, Holquist 1990/2002, Wertsch 1991, Vice 1997). In particular, we will
focus on how they are affected by general social and cultural patterns, the topics under
consideration, and the contribution of other participants in the interaction. We will
also address how identities can shift as a result of all of these factors.
The ways in which participants make their identities known is through many lin-
guistic, semiotic, and other means. What we will study here is a sub-type of speech acts
that I call identity acts (see Waugh 2008). Some identity acts are explicit, in which the
participant overtly identifies him/herself with regard to a facet of identity, as in “I am
French”, an explicit national identity act. These are marked identity acts (see Waugh
1982) since they directly address the issue of identity. In addition, there are many oth-
er, more implicit, unmarked identity acts in which a participant’s identity is ascer-
tained by the listener through inferencing: e.g., “I lived in France for one year” (since
a listener could infer from this that I am probably not French, it is an implicit national
identity act).
In the identity literature there has been some attention given to personal pronouns
like I, you, we, they as important indicators of identity (Brown & Gilman 1964, da Fina
1995, Muhlhausler & Harre 1990, Sebba & Wooton 1998, Silverstein 1976, 2003, Wales
1996); note the pronoun “I” in the examples given above. In this paper, we will study
the use of indefinite (non-specific, generic) pronouns in making generic statements in
implicit identity acts. As we will see, these are ideal places to find evidence for identity,
because generic utterances with nonspecific referents allow speakers to make claims
without necessarily taking responsibility for them personally, but at the same time
their identity can be inferred.
The construction of identity is bound up with the co-construction of ideology
(see also Davison 2001, Gal 1988, Kroskrity 2000, Simpson 1993, van Dijk 1989) and
like identity, language (discourse) both reflects and constructs ideology. I take ideology
here in its globalizing, intellectualizing sense (Woolard 1998): positive and negative
taken-for-granted assumptions, beliefs and value systems that are shared by social
groups and are inscribed in power structures, political forces, socio-cultural constraints,
interactional norms, pragmatic conventions, personal histories and preferences, etc.
The data for this paper will be a conversation in French between three monolin-
gual French speakers and one bilingual speaker, to be called Karim, who is also highly

San Diego, California, and at the Second Language Acquisition and Teaching Roundtable meet-
ing at the University of Arizona, for their input on some of the ideas presented here.
Pronominal choice in French conversational interaction 

proficient in (American) English. As we will see, Karim shows co-constructed shifting

national identities (and their associated ideologies); his shifts go through three stages
and create an organization for one part of the conversation. Since there is no evidence
that this organization is done consciously or is the result of pre-planning, this is fur-
ther confirmation that conversational organization is also co-constructed/negotiated
by the participants as they interact with each other.

2. Method of analysis: Work with an authentic conversation

2.1 Some specifics about everyday conversational French

The conversation to be analyzed is taken from a corpus of authentic conversations, the

Corpus of Everyday Conversational European French (ECEF). Published findings on
ECEF relevant to this study (Waugh & Fonseca-Greber 2002, Fonseca-Greber &
Waugh 2003a, b, Waugh et al 2007) are that the dependent (unstressed) pronouns/clit-
ics are grammatical prefixes on the verb; they are transcribed with a hyphen [-] and
rendered by a period [.] between the pronoun and the verb in the English translation.
More importantly, the subject pronoun nous ‘we’ has all but disappeared in everyday
conversational French and has been replaced by on- ‘we’ (subject) in the speech of
educated, middle-class speakers in a variety of different settings, and is the spoken
standard form for ‘we’ (cf. Atlani 1984, Boutet 1986, Covenay 2000, Fonseca-Greber &
Waugh 2003a, b, Freyne 1990, Gadet 1992, 1997, Peeters 2006, Soll 1983, Stewart 1995,
Waugh & Fonseca-Greber 2002, Waugh et al 2007). The use of on- for ‘we’ (subject)
does not mean that nous is no longer used in French. Nous is used, but in other func-
tions, namely: as a direct/indirect object prefix on the verb; as an independent pro-
noun (used in isolation, for emphasis, contrast, or clarification; as a thematic or topic
marker; as the object of prepositions; etc.); and, sometimes along with on-, as what
some have called ‘subject doubling’ or a ‘reprise construction’ (Convenay 2004). In
Example (1), given below, there is a reprise construction, with on- as the subject, as
indicated by the form of the verb, and nous as an independent pronoun:
(1) Karim Nous on-est perdus! Nous on-a notre ville à nous <laugh> parce
‘We/Us, we’re isolated! We/Us, we.have our city for ourselves <laugh>
because we.have’ <on- here means ‘we students (who go to a univer-
sity in a small city far from big cities)’>
[In all the transcriptions in this paper, the following conventions are used: all the prop-
er names are pseudonyms; [..] indicates an overlap with either a preceding or a suc-
ceeding utterance by one or more of the interlocutors; <> + italics means a comment
by the author; <.>, <..>, <...> are pauses of different lengths; <–> means an ellipsis; (..)
are words supplied by the author to make the translation smoother even though there
 Linda R. Waugh

is no corresponding form in the French original; a [/] between forms in the translation
indicates alternative translations; and bold italics are used for forms of interest both in
the French original and in the translation.]

2.2 Meanings of the three indefinites

In the written and high formal spoken language, on is used for indefinite (generic,
non-specific, sometimes called impersonal) meaning and is virtually the only pronoun
so used. In the ECEF corpus there are three indefinite prefixes in the everyday lan-
guage, namely, tu- ‘you (familiar)’, ils- ‘they’, on- ‘one’ (Waugh and Fonseca-Greber
2002, Fonseca-Greber and Waugh 2003a, 2003b, Waugh et al 2007); it should be noted
that others (Ashby 1992, Covenay 2003b, Laberge & Sankoff 1980) have identified a
fourth indefinite, namely vous- ‘you (formal)’, but we have only a couple of instances in
our corpus and thus we will not study the uses of indefinite vous-.
Careful interpretive work with the ECEF corpus has established the meanings of
the three indefinites as follows. Indefinite tu- ‘you (familiar)’ is inclusive of the inter-
locutors as virtual participants in the verbal process and highly personalized because it
can express solidarity between the interlocutors; it strives for communicative involve-
ment of the addressee(s) in the topic of the utterance, just as it conveys involvement by
the speaker. It has an implied meaning of “suppose you were in the situation” (cf. Ashby
1992, Atlani 1984, Boutet 1986, Covenay 2003a, b, Freyne 1990, Laberge & Sankoff
1980, Le Bel 1991, Peeters 2006, Soll 1983, Waugh et al 2007). There is a parallelism
with indefinite ‘you’ in English, and thus in most cases, indefinite tu- in French is trans-
lated by indefinite ‘you’. Example (2) is one use of indefinite tu- from the corpus:
(2) Karim Même dans les villes comme Cincinnati ou je-sais pas des villes qui
sont un peu plus perdues que Washington tu-trouveras beaucoup
plus de gens obèses.
‘Even in cities like Cincinnati or I. dunno cities that are a little more
isolated than Washington you’ll.find many more obese people’
In this case, Karim has lived in Washington (D.C.) and knows it well, but it seems from
the rest of the conversation that he’s never been to Cincinnati, and it’s clear that his
interlocutors have never been to the U.S. But he’s including them, and himself, in a
virtual way as observers of the situation described and thus as involved in some way in
what he’s saying.
Whereas indefinite tu- is inclusive and expresses solidarity, ils- ‘they’ is exclusive
of the interlocutors as virtual participants in the verbal process; it tends to express
distance from the topic of the utterance and from the indefinite ‘others’ on the part of
the speaker. It tends to mean “some others, but not me and probably not you”. Often
the referent, while indefinite, can be narrowed through metonymy. Curiously, there is
very little reference to the use of ils- as an indefinite in the linguistic and sociolinguistic
literature on French (but see Atlani 1984), and yet its presence is clear in the data here
Pronominal choice in French conversational interaction 

and it plays a very strategic role. As with tu-, there is a parallelism with indefinite ‘they’
in English. Example (3) shows one use of indefinite ils- in the corpus.
(3) Michel: Dans les films américains, enfin je-sais pas [si c-est la réalité]
‘In American films, well I. dunno [if it.s real]’
Karim: [ouais, ouais]
‘[yeah, yeah]’
Michel: dans tous les feuilletons, ils-mangent [tout le temps devant la télé]
‘in all the TV shows/sitcoms, [all the time in front of the TV]’
Given the context of American films in the first utterance, and then the reference to TV
shows/sitcoms, which is interpreted contextually to mean American TV shows, the
referent of ils- is interpreted contextually, through metonymy, to be generic, indefinite,
non-specific Americans. They don’t include Michel, who obviously doesn’t know a lot
about Americans; they are “others” whom he would like to get information about.
Indefinite on- ‘one’ is still used in spoken French for indefinite/non-specific/ge-
neric meaning, but much less than before, much less than its use for ‘we’ and in some
studies less than the use of tu-/vous- for indefinite meaning (Ashby 1992, Covenay
2000, Fonseca-Greber & Waugh 2003a, b, Peeters 2006). In comparison with tu- and
ils-, on- is the unmarked (Waugh 1982) indefinite prefix with the widest variety of in-
terpretations. It can express solidarity, distance or neutrality; it can be inclusive, or
exclusive of the speaker and/or addressee, or neutral; it can express high, low or neu-
tral involvement. It has various translations in English, including ‘one’ (which is asso-
ciated with formal usage in American English but is not always formal in the French
context), ‘people’ in general, ‘someone’, ‘anyone’, as well as indefinite, generalized ‘we’,
‘you’ and ‘they’. Because of its flexibility, on- is often used as a face-saving device
(Covenay 2003a, b, Stewart 1995) since the speaker can always deny that a given inter-
pretation was meant, if challenged by the addressee.
In Example (4) below we have a nice contrast between indefinite tu- and indefinite
on-: tu- gives the solidarity and involvement discussed earlier, while on- refers to face-
less, anonymous others, who are contextually (through metonymy) identified as being
in France, but nothing else is known about them.
(4) Michel ouais mais je-veux dire, les les, excuse-moi, tu-traverses une foule
américaine plutôt qu’une française, tu-vois plus d’obèses
‘Yeah but I mean, the the, excuse me, you.go through an American
crowd rather than a French (one), you.see more fat people’
Karim Oui, oh c-est c-est clair! C-est sûr!
‘Yes, oh it.s it.s true! It.s for sure!’
Maher [Parce qu’en France, on-avait dit qu’y-avait moins de 3%]
‘[Because in France, one.(someone./they.) said that there. were less
than 3%]’
 Linda R. Waugh

In addition to these two uses of on-, there is a third use, vague on-, also called ambigu-
ous, indeterminate (Ashby 1992, Boutet 1986, Covenay 2003 a, b, Fonseca-Greber &
Waugh 2003a, b, Le Bel 1991, Peeters 2006, Stewart 1995). So, in Example (5) on- could
be interpreted contextually in its non-specific/neutral interpretation (‘one’, ‘anyone’),
in its non-specific/inclusive interpretation (‘we’ in its generic sense), or in its non-
specific/exclusive interpretation (‘they’). The meaning is not clear from the context,
but this doesn’t impede communication; and indeed this vague usage can be exploited
for its strategic value possibilities (Le Bel 1991, Peeters 2006, Stewart 1995). In Exam-
ple (5), Karim and Maher are talking about living on 180 francs per month in Tunisia,
where Karim has just spent a month.
(5) Karim  180 francs. Mais c-est, c-est beaucoup là-bas!
‘180 francs. But it’s, it’s a lot there!’
Maher  Ah ouais d’accord
‘Oh yeah o.k.’
Karim  On- on-fait pas mal avec. <...>
‘One./we./they. one./we./they.get along all right with (that). <...>’
In summary: tu- is used for inclusiveness/solidarity/involvement; ils- for exclusive-
ness/distancing; and on- for all of these plus a more neutral stance and for vagueness.
In terms of their communicative-pragmatic import and relationship to identity, we can
say that these indefinites can convey group membership (in-group vs. out-group),
solidarity/alignment vs. distance, and neutral stance vs. communicative involvement.

2.3 Conversation to be analyzed

The conversations to be analyzed took place in Southern France, in August 1999, in an

outdoor cafe. The four participants are all in their late teens. Karim is a French citizen
with a Tunisian ethnic background. At the time of the taping, he had been living for
some time in the U.S., had gone to high school in the Washington, D.C. area where his
parents live, had recently finished his Freshman year at an elite private university in the
U.S., and was on a summer vacation trip to France to see relatives and friends, after
which he was to return to the U.S. for his second year of college. His parents are quite
wealthy and upper middle class. His first language is standard French; his second lan-
guage (American) English. The day before the taping, he had come back from a month-
long stay in Tunisia to learn Tunisian Arabic.
The other participants are Maher, a long-time friend of Karim’s, who is a French
citizen of Moroccan ethnicity and speaks the stigmatized French dialect of the Arab
youth of immigrant families and lower socioeconomic background, and Sylvie, a
friend of Maher’s who had met Karim before only briefly; she is a French citizen of
Tunisian ethnicity and also speaks the stigmatized French dialect of the Arab youth.
The third participant is Michel, who knows Maher and Sylvie but is meeting Karim for
Pronominal choice in French conversational interaction 

the first time. He is a French citizen with no immigrant/ethnic background and speaks
standard French.
While the major focus of this paper will be Karim’s national identities, we will look
briefly at the issue of his linguistic identities first, because Michel brings up this topic
right after his arrival and the ensuing discussion provides background for an under-
standing of what happens later (for a fuller analysis, see Waugh 2008).

3. Karim’s national identities

3.1 Background: Karim’s linguistic identities

When Michel arrives and is told that Karim lives in the U.S., Michel says that he de-
tects an American accent in Karim’s French, although he has only heard Karim say five
words, and native speakers who have listened to the tape have a hard time detecting a
foreign accent in Karim’s French. Michel is using his cultural and symbolic capital as a
native speaker of standard French, which has high symbolic power in France (Bourdieu
1986, 1991, Duranti 2004, Hanks 2005, Thompson 1991). He seems to cling to the
ideology prevalent in France that no one who lives outside of France (and other
French-speaking areas) can speak French without a foreign accent and be a native
speaker. He is also, probably, reacting to the prevalent discourse in France about the
differences/rivalry between France and the U.S. and also the perceived threat to the
French language from the use of English as an international language, replacing French
in this respect (Ager 1999, Gordon 1978, Hagege 2006, Sanders 1993).
Karim’s reply to this face-threatening act (Goffman 1959, Brown & Levinson 1978)
by Michel is at first a show of surprise by what Michel says; Maher makes it clear that
he doesn’t agree that Karim has a foreign accent; Sylvie agrees with Michel but insists
that Karim speaks excellent French. Given Michel’s cultural capital as a native speaker
of standard French, Karim accommodates more to him than to the other two (Giles,
Coupland & Coupland 1991, Giles & Powesland 1997), and goes on to say that he has
a French accent in his English, thus attempting to save face and to underscore his
French identity. After Sylvie repeats that he speaks excellent French, Karim again in-
sists on his French identity by saying that he speaks French at home with his family
and when Maher says that he doesn’t give the impression (in French) that he speaks
American English, Karim says that in English, Americans know that he isn’t American,
thereby distancing himself from his American identity. But then, after Maher shows
his surprise at this statement, Karim says he speaks no language perfectly:
(6) Karim Ouais, ouais! C-est drôle je-parle, je-parle aucune langue parfaite-
‘Yeah, yeah! It.s funny I.speak, I.don’t speak any language perfectly.’
 Linda R. Waugh

This meta-linguistic identity act is very telling. Earlier, when talking with Maher and
Sylvie before Michel’s arrival, his bilingualism was a real achievement and he showed
pride in his high level of competence in English. Now, his bilingualism is problematic
and while we don’t know exactly what he means by not speaking any language ‘per-
fectly’, we do know that there is an overwhelming ideology among some language
learners (and teachers and researchers) that having an accent means that one’s lan-
guage, first or second, is imperfect (see Cook 2003, Davies 2003, 2004). That is, having
been put on the defensive by Michel, Karim loses, interactionally, his high communi-
cative competence in two languages in the face of a possible ‘accent’ in both, and that
undermines all of the abilities that he undoubtedly possesses in both. Now, this issue
of lack of confidence by bilinguals who may not feel that they are native speakers in
either language is not surprising. And it is very important for this conversation since
this admission by Karim creates a disequilibrium in Karim’s linguistic identities; and
we will see that this has a large effect on his national identities in the next part of the

3.2 Co-construction of Karim’s national identities

As the literature has shown, national identities are the product of discourse (Argenter
2000, Gumperz 1982, Hansen & Liu 1997, Jenkins 2004, Joseph 2004, Tajfel 1974); in
addition, work on national identity has shown that people participate in the idea of a
nation as represented in its national culture, thus creating a link between nation and
culture. This imagined national(-cultural) community is real to the extent that one is
convinced of it, believes in it, and identifies with it emotionally (cf. Anderson 1991,
Blommaert & Verschueren 1998, de Cilia et al 1999, Wodak et al 1999, Yamaguchi
2005). But, there are many questions about whether someone who is binational (has
two national identities) can also be bicultural (see e.g., Agar 1991, Hoffman 1989,
Kaplan 1993, Pavlenko & Lantolf 2000). As we will see, Karim shows a lot of uncer-
tainty about being binational and bicultural.
As said earlier, there are three stages in the conversation in the co-construction of
Karim’s national identities. Each stage is correlated with particular topics having to do
with America and France, and each is associated with explicit national identity acts as
well as uses of the indefinite/non-specific/generic prefixes in implicit identity acts.

3.3 Stage one: Karim is a Tunisian-American,

Michel is a good (real) Frenchman

After the statement given above about not speaking any language perfectly, Karim goes
back to a previous topic, namely, his stay in Tunisia learning Tunisian Arabic, and this
time he brings up the fact that there were four Americans in his class. In Example (7),
he says that two of them were ‘Tunisian-Americans like me’:
Pronominal choice in French conversational interaction 

(7) Karim <...> Et y-avait deux autres Américains euh Franç- euh Tuniso-
Américains comme moi.
‘<...> And there.were two other Americans uh Fren- uh Tunisian-
Americans like me.’
This explicit national identity act is in direct contrast with his earlier statement that
Americans could tell by his accent that he isn’t American. We can understand this
contextually and interactionally because earlier Karim had said that the participants in
the course in Tunisia broke up into clans ‘groups’ when class was over and although he
never tells his interlocutors the basis for the groupings, we can infer that it was prob-
ably nationality coupled with language. And we can also infer from what he says in
Example (7) that he was in a group with Americans who were Tunisian-Americans
(given the fact that this course in Tunisia was explicitly set up for the children of
Tunisian emigres). It’s telling that he says nothing about the Tunisian French in the
course, who undoubtedly were there, and his silence on this topic probably means that
he wasn’t in their group. In any case, the disequilibrium that we saw with his linguistic
identity discussed above is mirrored here by the way in which he formulates this na-
tional identity act since it contains two repairs, first on ‘Americans’, then on the half-
formed ‘Fren(ch)’, after which he finally settles on Tunisian-American. This last for-
mulation means that he’s not a ‘real’ American; he’s what is sometimes called a
‘hyphenated’ American (a hybrid American). This of course is consistent with his not
speaking English perfectly and not being a ‘pure’ American.
After a while, Karim abandons the topic of who was in Tunisia, since the others are
clearly not interested, and he turns to Michel to ask him where he’s from. Here is
Michel’s reply:
(8) Michel Moi je-suis de M[...], du nord de la France
‘I.m from M[..,.] in the north of France.’
Karim D’accord.
Michel un bon Français!
‘a good/real Frenchman!’
Michel’s answer immediately sets up a stark dichotomy between himself and Karim,
the self-identified Tunisian-American who speaks French with an American accent.
The ideology at work in this explicit national identity act is not hard to see: Michel is
not an immigrant, he’s from native French stock, he speaks standard (‘white’) French,
and he lives in the north of France, which for some is seen as the ‘center’ of French
culture. All of this means that he’s a better example of a Frenchman than French citi-
zens of North African heritage (like Maher and Sylvie) who live in the South, much
less someone who lives in the U.S. (Karim). It certainly means that Michel is again as-
serting his cultural and symbolic capital over Karim (Bourdieu 1986, 1991, Hanks
2005, Thompson 1991), this time as a bona fide Frenchman.
 Linda R. Waugh

Maher follows Michel’s remark by saying that Michel is from a town that’s very
isolated. In response, Karim then says that the city where the university he’s attending
is located is small, more isolated than Michel’s town, and far away from many large
well-known North American cities. In this implicit identity act, he uses on- ‘we’, from
which his interlocutors can infer that he identifies with this American town, even
though he’s only lived there for nine months; see Example (9), which is followed a sec-
ond later by Example (1) above, which includes the independent pronoun nous ‘we:
(9) Karim On-est à 5 heures de New York, à 6 heures de Washington, à 6 heures
de Montréal.
‘We.are 5 hours from New York, 6 hours from Washington, 6 hours
from Montreal.’
Michel has many prejudices and stereotypes, particularly negative ones, about the U.S.
In this part of the conversation, Karim constructs himself as an expert on the U.S.,
both the good and the bad, and Michel concurs with him, especially when what Karim
says conforms to his own negatives about the U.S. So, for example, when Michel says
that, aside from the big cities, there’s nothing in the U.S., whereas in France there are
(smaller) cities, Karim agrees, and says that in the U.S. one can drive through corn-
fields for a whole day. Then, when asked by Maher how often he saw his parents while
he was at the university, Karim answers that it’s about once every three months, which
is shocking to his interlocutors because this would not be the case in France where
adolescents his age would go home every weekend. So, once again, he is identifying
himself as being more American in his socio-cultural habits than French. And then,
when asked how he goes from his small town to Washington, he says he drives and it’s
six hours:
(10) Sylvie six heures!
‘six hours!’
Karim Ouais, non, mais ici ça fait long! Mais, mais nous six heures ça va.
‘Yeah, no, but here that’s long! But, but we/for us six hours that’s all
Sylvie Ouais ouais, c’est des trajets courts en fait pour eux, six heures!
‘Yeah yeah, those are short trips in fact for them, six hours!’
Karim: C-est pas court, mais c-est faisable. C-est qu’on- <–>
‘It‘s not short, but it’s doable. It’s that we.’ <–>
Karim goes on to explain that six hours of driving is doable because, whereas in France
there are lots of fous ‘crazies’ who drive around, American highways are calmer:
(11) Karim Aux Etats-Unis c-est beaucoup plus calme. Alors c-est facile, on-est
pas fatigué après avoir fait six heures six heures de route
‘In the US it.s much calmer. So it.s easy, we.aren’t tired after having
been six hours six hours on the road.’
Pronominal choice in French conversational interaction 

Notice that in these national identity acts, Karim explicitly uses nous ‘we’ (Example 10)
in contrast with ici ‘here (in France)’ and on- ‘we’ (Example 11) along with aux Etats-
Unis ‘in the U.S.’. Indeed, in Stage One of his national identity, Karim uses the indepen-
dent pronoun nous ‘we’, and the prefix on-, which in some contexts can be interpreted
as meaning ‘we’ and in others as ‘one, generalizing we’, for the U.S. The others consis-
tently use the independent pronoun eux ‘they/them’ and the prefix ils- ‘they’ for the
U.S. The dichotomy between Karim and the others is quite striking: he identifies as
being (Tunisian)-American and they don’t.

3.3 Stage two: Karim begins to withdraw from being American 

In Stage Two, Michel, in particular, brings up, or helps in the co-construction of, topics
about negative aspects of the U.S. He starts with the issue of the young age of driving
permits in the U.S.: 15 years old in some states, which seems really young in France.
Maher then talks about ‘really big cars’ (grosses bagnoles) in the U.S., and Karim agrees
that they’re ‘enormous things’, ‘boats’ (énormes trucs, bateaux). It is at this point that
Karim starts to take his distance from being American: he talks about a small Ford that
is found in France but doesn’t exist in the U.S. because it’s too small for Americans:
(12) Karim Ouais. C’est une Ford que pour-, pourtant non, ça existe pas, ils- c-
est trop petit pour les Américains!
‘Yeah. It’s a Ford that ye-, yet no, that doesn’t exist, they. it.s too small
for Americans!’
Here, Karim uses, for the first time, ils- ‘they’, as well as the distancing nominalization,
les Américains ‘Americans’, when speaking about the U.S. and Americans. This is an
implicit national identity act since it shows that he’s not including himself in the
‘Americans’, or at least in the Americans for whom the Ford is ‘too small’.
The others now react by saying that listening to Karim makes them want to travel,
to go to the U.S. Karim then goes on in the guise of the expert on America that he con-
tinues to construct for himself. He talks about violence in the U.S., characterizes
Washington as the former murder capital of the U.S., speaks about visiting a friend who
lives in Harlem but says he wasn’t attacked because he’d learned to avoid certain streets,
and describes segregation in U.S. cities. Then, he gives his view that in France there are
more ‘small robberies’ (petits vols) and other petty crimes whereas in the U.S.:
(13) Karim Si, si on-vole bon voilà on-te-tue, on-vole toute ta voiture
‘If, if they.steal well then, they.steal your whole car’
Karim tout, ta maison, mais on-, mais y-a moins de petits trucs tu-vois.
‘everything, your whole house, but they-, but there.are fewer small
things you.see’
 Linda R. Waugh

Here he uses on- ‘they’ to mean anonymous, faceless, alien others, Americans with
whom he obviously has no connection and no identification, whereas the indefinite
tu- (and its corresponding object prefix te- ‘you’ and possessive adjective ta (‘your’))
shows that he identifies more with those who are stolen from or killed and that he’s
asking his audience to place their solidarity with them. Again, he’s distancing himself
from certain U.S. socio-cultural patterns.
Karim then goes on to talk about the reasons for killing and violence in general,
especially the right to have a firearm, a gun, in the U.S. and the ease with which
Americans can buy guns. In recounting an incident where someone in Atlanta killed
12 or 15 people, he says:
(14) Karim <–> Bon, ça ça arrive et bon euh c-est c-est
une conséquence de de, c’est c-est complètement ridicule! Là-bas, y-a
y-a des enfants qui peuvent aller acheter leurs armes
‘<–> Well, that that happens and well uh it.s it.s
a consequence of of, it.s it.s completely ridiculous! Over there, there.
are there.are kids who can go buy their guns/weapons/fire arms’
Karim à quatorze ans dans le
‘at fourteen years old in the’
Karim [Ouais ouais à quatorze ans] dans le Montana.
C-est c-est ridicule on-a même pas besoin de permis, tu-vois c’est
juste on-s’en va, on-achète son arme et on-sort. C-est ridicule! <..> Et
après ils-sont étonnés d’avoir des trucs, des tueries comme ça!
‘[Yeah yeah at fourteen years old] in Montana. It.s it.s ridiculous one.
doesn’t even need to have a permit, y’see it’s just one.goes/they.go,
one.buys/ one’s/their gun and one./they leave(s). It.s ridicu-
lous! <..> And afterwards they.are surprised to have these things,
these killings like that one!’
Here Karim is even more explicit about taking his distance from being American. In his
first turn in this example, Karim shows his negative attitude with ‘it’s completely ridicu-
lous’. Then with ‘over there’ (là-bas, meaning ‘far away’) he takes a perspective that is
rooted in France, not in the U.S., showing that as against Stage One, he is now placing
himself in a position external to the U.S. (Simpson 1993). After saying, and repeating,
that 14-year-old kids can buy guns in Montana, he uses a series of on-, contextually
interpretable as meaning ‘one/they/anyone’ (‘others’) but not as ‘we’; this interpretation
is reinforced by his repetition of ‘it’s ridiculous’. And then in the next utterance he
makes his distance even clearer, with the use of ils- ‘they’ to mean Americans. Here, he’s
saying, ‘they’re astonished, but I’m not, because I’m not one of them, I know better’.
Karim’s prefixal usage is very different in Stage Two from Stage One: whereas in
Stage One he used on- to mean ‘we’ or the neutral ‘one/generalizing we’, in Stage Two
Pronominal choice in French conversational interaction 

he uses on- to mean ‘one/generalizing they’ in reference to faceless, anonymous others.

Americans are referred to with ‘they’; the nominalization ‘Americans,’ combined with
value judgments like ‘it’s ridiculous’, also show his distance from America and American
mores. In addition, by using tu- he attempts to establish non-American solidarity with
his interlocutors. At this point, at the very least he’s very distant from the Tunisian-
American identity of Stage One and from the praise he had for American socio-cultural
One could imagine some Americans also speaking this way to show that they are
critical of certain aspects of American society like big cars, violence, and the right to
buy guns, so at this stage Karim is not declaring himself to be non-American; but he is
preparing for what will happen in Stage three.

3.4 Stage three: Two kinds of Americans vs. Karim: “I’m from a French family”

There are several topics about cultural differences between the U.S. and France in Stage
Three: how much Americans watch TV, Americans eating in front of the TV, what
Americans eat, how obese Americans are, how little culture there is in the U.S. All of
these have to do with cultural stereotypes about Americans in France. As we will see,
Karim makes a difference between those Americans who conform to these stereotypes
and those who don’t.
On the topic of watching TV, Maher asks Karim if American young people watch
TV more than the French and he says yes:
(15) Karim C-est sûr, c-est sûr [ça dépend dans quel milieu]
‘It.s for sure, it.s for sure [that depends on what kind of people/what
This response is typical of Karim in Stage Three: he agrees with a particular stereotypi-
cal characterization of Americans, but then he just as quickly backs off from including
all Americans in the stereotype. Here, after agreeing that more young Americans watch
TV than the French, he mitigates his answer by saying that it depends on the milieu,
which could roughly be understood as meaning their socio-economic status, their
level of education, their type of employment, and so forth. After a short digression and
prompting from Maher to go back to the topic of watching TV, Karim elaborates fur-
ther about not only watching TV but also eating in front of the TV, but again he miti-
gates his answer with mais ça dépend qui ‘but that depends on who’. He then says:
(16) Karim <–> Euh vers, bon bien sûr, moi je-suis d’une famille française, alors
je-vais pas manger devant la télé! Et on-regarde pas tellement la télé
non plus. Surtout qu’y-a tellement de cochonneries à la télé!
‘<–> Uh, well of course, I.m from a French family, so I.m not going
to eat in front of the TV! And we.don’t watch TV very much either.
Especially since there.s so much junk on TV!’
 Linda R. Waugh

This explicit national identity act, ‘I’m from a French family’, followed by his identity
declaration that he wouldn’t eat in front of the TV, is preceded by ‘well of course’
(bon bien sûr), as if there could be no doubt that because he’s from a French family,
therefore (alors) he doesn’t eat in front of the TV. He goes on to align himself further
with his French family with the use of on-, which means ‘we, in my French family’,
when he says that ‘we don’t watch TV very much either’. And he uses the strong (slightly
vulgar) word cochonneries ‘junk’ (cochonneries contains the root cochon ‘pig’ and is
pejorative in this context) to characterize what’s on TV. What Karim is now insisting
on is that he might live in America, but his cultural habits (especially his familial hab-
its) are French. And in this way he is different from the ‘other’ Americans who don’t
adhere to the stereotypes.
This differentiation between Karim and the ‘other’ Americans is obvious again
when he goes on to talk about American eating habits:
(17) Karim Et humm mais heu <...> bon ils-, ils-peuvent bouffer n’importe quoi
c-est vrai, mais il-faut pas se-dire que tous les Américains [mangent]
‘And umm but uh <...> well they-, they-can eat/stuff themselves with
anything it’s true, but one shouldn’t believe that all Americans [eat]
(like that)’
Note the use of generic ils- ‘they’, the pejorative verb bouffer ‘stuff oneself with’ and the
denigrating n’importe quoi ‘anything’, all of which show how distant Karim is from be-
ing American. Once again, he nuances his claim by saying ‘but one shouldn’t believe
that all Americans eat (like that)’, with the nominalization ‘Americans’ showing that he
doesn’t include himself. At this point Maher picks up on what Karim is saying about
the fact that not all Americans conform to this stereotype and says it’s a cliché. This
puts into question the rather extreme formulations that Michel in particular has been
making, but it doesn’t stop him from continuing in this vein. He brings up the issue of
obesity in the U.S. and Karim agrees that there is much more obesity in the U.S. than
in France, but he then goes on to say that in big cities, like Washington, there are fewer
obese people because people are educated and therefore they eat better.
(18) Karim Heu alors quand quand on-est éduqué, on-fait plus attention à sa
ligne, y-a beaucoup de de magasins biologiques, les gens font atten-
tion! Ils-cuisinent bien, ils- mangent pas n’importe quoi, etc. Mmm
‘Uh so when when eduated, one.pays more attention to one’s
shape (of the body), there are many healthfood/organic food stores,
people pay attention! They.cook well, they.don’t eat just anything, etc.
Notice here that on- is interpretable as neutral and generic: it is coupled with the neu-
tral possessive adjective sa ‘one’s’ (not notre ‘ours’) and with the generic noun les gens
‘people’. Once again, Karim differentiates between two types of Americans: les gens
‘people (who live in Washington)’ vs. (earlier, not shown here) paysans ‘rural people,
Pronominal choice in French conversational interaction 

peasants, farmers’, which is a somewhat pejorative term. Moreover, his use of les gens
coupled with anaphoric ils- ‘they’ to refer back to les gens shows that again he excludes
himself and his family from their reference, even though presumably his family cooks
well and they don’t eat just anything. By using ils-, he doesn’t include himself, even
though he’s now talking about a positive aspect of American mores. Presumably he’s
implying that his family doesn’t cook like those other non-cliché Americans, since his
family’s food is more French(-like). Here we can point to the research by Lindenfeld
(2000), which shows that the French in the U.S. tend to keep their French cultural
habits at home even if they seem to be Americanized in other contexts.
After more conversation about obese people in the U.S., Karim says that in
Washington, you can tell the people from the city, who are not obese, from the
(American) tourists, who are, and he claims, as in Example (2) above, that even in cit-
ies like Cincinnati you’ll (with tu-) find more obese people. At the end of this, Michel
cites his aunt, who said that, apart from the big cities, there are uncouth people all over
the U.S. Maher interprets this as meaning that there’s a difference between the cities
and the countryside, and Karim picks up on that:
(19) Karim [L’Amérique] eh dans eh dans les grandes villes on-a beaucoup de
culture, on-a des musées surtout Washington, les musées, les pièces
de théâtre, euh les expositions, y-a, on-a des énormes, bon, on-a, bon
on-a la culture. Et dans dans la campagne, y-en-a pas! Ça ça fait une
grande différence au niveau des gens, vers, moi j-estime que le le
Français, je-sais pas, tu-vois la France, l’Académie Française tu-vois,
y-a, bon, moi je-dis pas que c-est bien, mais c-est c-est des standards
de culture au niveau de l’état, aux Etats-Unis on-a rien <emphasis>!
Alors euh c-est, le peuple est un peu <.> peu cultu-, peu cultivé!
‘[America] uh in uh in the big cities one.has lots of culture, one.has
museums especially Washington, museums, plays, uh expositions,
there.are, one.has enormous well, one.has, one.has culture. And in in
the countryside, there.isn’t any! That that makes a big difference at
the level of people, about, I think that the the French person, I.dun-
no, you.see, France, the French Academy you-see, there.are, well, I
don’t say that it.s good, but it.s it.s standards of culture at the na-
tional government level, in the U.S. one.has nothing <emphasis>! So
uh it.s the people are a little cult- <.> have little culture’
Here, for his French interlocutors, Karim praises the French, France, the French Acad-
emy, for standards of culture at the national government level; we can see that there is
a certain nostalgia, and a (stereotypical) idealism based on an imagined community of
the highly cultured French (Anderson 1991). In the U.S., on the other hand, on-
(neutral/vague use) has nothing; there’s little culture outside of the big cities like
Washington. Notice also here the use of the word peuple, which is pejorative in French
and recalls paysans used earlier.
 Linda R. Waugh

4. Conclusion

Karim has shifting national identities in interaction that are multiple, dynamic, con-
textualized, and quite unstable. Interactionally, there are three stages of his identities,
which are different in terms of their content, his identity, and his use of the personal
and indefinite pronouns, as well as his explicit and implicit identity acts. In summary
we can say that, in situating himself as either American or French, he goes from more
general social mores to more local, familial cultural habits in this discussion. In Stage
One, when he self-identifies with an explicit national identity act as “Tunisian-American”,
the topics are where he lives, where he went to high school and is going to college, the
way Americans drive, in addition to generalizations about Tunisians gleaned from his
recent sojourn in Tunisia to learn Tunisian Arabic. Here we see the use of on- ‘we’ to
mean Americans, including himself. In Stage Two, he distances himself from his
American identity with implicit national identity acts, when discussing certain aspects
of U.S. society, for example, big cars, killings, robberies, buying guns, violence. Here
we see the use of ils- ‘they’ for Americans and on- ‘one’ for generalizing statements. In
Stage Three, there is further differentiation and division into two types of Americans
vs. his French family when there is a discussion of some cultural aspects of U.S. more
tied to family and personal habits, for example, TV-watching, eating in front of the TV,
eating just anything, obesity. Once again, Karim self-identifies with an explicit
national(-cultural) identity act as being “from a French family”. Here we see the use of
on- ‘we’ to mean Karim and his family; ils- ‘they’ for Americans; and on- ‘one’ for gen-
eralizing statements.
What this close analysis has been able to show us is how Karim’s identity shifts as
influenced by the issue of his American accent in French and the obvious negative
prejudices and stereotypes of someone like Michel. Where he takes his distance from
his (Tunisian-)American identity is when there are cultural differences between the
U.S. and France. He’s quite comfortable with saying that in the U.S. driving long dis-
tances is easier but sometimes there are lots of cornfields, as against France where
there are lots of different landscapes in a short distance. But he takes his distance when
talking about big cars and violence in the U.S. And then he’s even more explicit when
saying that he doesn’t watch lots of TV, or eat in front of the TV, or eat just anything,
because he’s from a French family. So, as we see, these shifts in identity (and associated
ideologies) are dependent on topic. They are also correlated with the presence of an
interlocutor who immediately attacks him by claiming he has an accent, then asserts
symbolic power as a good Frenchman, and then shows many negative prejudices and
stereotypes about America. All of this makes it difficult for Karim to maintain his ear-
lier identification as a (Tunisian-) American. He responds to this by a more nuanced
identity as a member of a French family living in America who maintains certain
French cultural habits and speaks French at home and thus has not taken on a full
American identity.
Pronominal choice in French conversational interaction 

Karim shows ambivalence about being partly American and partly French, and
seems to be uncomfortable with the gap between the French culture of his family and
the American culture that he is part of. So, the shifts that we see may also be a reflec-
tion of his highly complex situation with hybrid identities, and of his own search for
which identities he wants to construct for himself and how to negotiate between the
cultural norms of his family and the surrounding national culture. Thus, we see in
microcosm that identity can be the site of a struggle within the individual and, as a
result, in interactions between native speakers and learners of a language.
I want to finish with a methodological point. As I hope I’ve shown, in doing this
kind of interpretive work, it’s necessary to do a very fine-grained interpretive analysis
of the data from an authentic conversation, seen as a dynamic process in which posi-
tions are taken but are malleable. This means understanding the conversation as it
unfolds for the participants, contextualizing the data and reinforcing the analysis with
everything that is known about the participants in the conversation, as well as the so-
cio-cultural, historico-political context in which the conversation takes place, and the
personal history of the participants (insofar as this can be known). Only then is it pos-
sible to understand the complexity of what Karim says about his identity and the ideo-
logical stances that he takes.
Thus, we have seen that identity is co-constructed and negotiated, as evidenced by
micro-processes in the interaction, in particular in identity acts, either explicit or im-
plicit. Identity is indeed emergent in interaction, the “outcome of intersubjectively ne-
gotiated practices” (Bucholtz 2004: 469).


Anderson, B. 1991. Imagined Communities: Reflections on the Origin and Spread of Nationalism,
2nd edition. London: Verso.
Antaki, C. and Widdicombe, S. 1998. “Identity as an achievement and as a tool”. In Identities in
Talk, C. Antaki and S. Widdicombe (eds), 1–14. London: Sage.
Argenter, J. 2000. “Cultural identity and heteroglossia.”. Estudios de Sociolinguistica 1, 27–39.
Ashby, W. 1992. “The variable use of on versus tu/vous for indefinite reference in spoken French.”
Journal of French Language Studies 2, 135–157.
Atlani, Francoise. 1984. ON l’illusioniste. In La langue au ras du texte, F. Atlani, L. Danon-Boileau,
A. Gresillon, J-L Lebrave, and J. Simonin (eds), 12–29. Lille: Presses Universitaires de Lille.
Bakhtin, M. 1981. The Dialogic Imagination: Four Essays. Austin: University of Texas Press.
Block, D. 2006. Multilingual identities in a global city: London stories. NY: Palgrave Macmillan.
Block, D. 2007. Second language identities. London: Continuum.
Blommaert, J. and Verschueren, J. 1998. “The role of language in European nationalist Ideolo-
gies.” In Language Ideologies: Practice and Theory, B. Schieffelin, K. Woolard, and P.
Kroskrity (eds), 189–210. NY: Oxford University Press.
Bourdieu, P. 1986. “The forms of capital.” Handbook of Theory and Research for the Sociology of
Education, 241–258.
 Linda R. Waugh

Bourdieu, P. 1991. Language and Symbolic Power. (Translation of Ce que parler veut dire.)
Cambridge, MA: Harvard University Press.
Boutet, J. 1986. “La référence à la personne en français parlé: le cas de on.” Langage et société 38:
Brown, R. and Gilman, A. 1964. “The pronouns of power and solidarity.” Reprinted in, Psycho-
linguistics, R. Brown (ed.). 1970.
Brown, P. and Levinson, S. 1978. Politeness: Some Universals in Language Usage. Cambridge:
Cambridge University Press.
Bucholtz, M. and Hall, K. 2004. “Language and identity.” In A Companion to Linguistic Anthro-
pology, A. Duranti (ed.), 369–394. Malden, MA: Blackwell.
Bucholtz, M. and Hall, K. 2005. “Identity and interaction: a sociocultural linguistic approach.”
Discourse Studies 7, 585–614.
Cerulo, K. 1997. “Identity construction: New issues, new directions.” Annual Review of Sociolo-
gy, 23:385–409.
Clyne, M. 2004. “Multilingualism.” In The Handbook of Sociolinguistics, F. Coulmas (ed.),
301–314. Oxford: Blackwell.
Cook, V. (ed.) 2003. Effects of the Second Language on the First. Clevedon, UK: Multilingual
Coupland, N., Nussbaum, and J. Grossman, A. 1993. “Introduction: Discourse, selfhood, and
the lifespan.” In Discourse and Lifespan Identity, N. Coupland and J. Nussbaum (eds).
Newbury Park, CA: Sage.
Covenay, A. 2000. “Vestiges of nous and the 1st person plural verb in informal spoken French.”
Language Sciences 22:447–81.
Covenay, A. 2003a. “ ‘All for on and on for all’: on and its competitors in spoken French.” Fran-
cophonie, 27:22–26.
Covenay, A. 2003b. “ ‘Anything you can do, tu can do better’: tu and vous as substitutes for in-
definite on in French.” Journal of Sociolinguistics 7, 164–191.
Covenay, A. 2004. “Subject doubling in spoken French: A sociolinguistic approach.” The French
Review 79, 96–111.
DaFina, A. 1995. “Pronominal choice, identity, and solidarity in political discourse.” Text 15,
Davies, A. 2003. The Native Speaker: Myth and Reality. Clevendon: Multilingual Matters.
Davies, A. 2004. “The Native speaker in applied linguistics.” In TheHandbook of Applied Linguis-
tics, A. Davies and C. Elder (eds), 431–450.
Davison, C. 2001. “Identity and ideology: The problem of defining and defending ESLness.” In
English as a Second Language in the Mainstream: Issue of Teaching, Learning and Identity, B.
Mohan, C. Leung, and C. Davison (eds). London: Pearson Longman.
DeCillia, R., Reisigl, M. and Wodak, R. 1999. “The Discursive construction of national identi-
ties.” Discourse and Society 10, 149–173.
Doran, M. 2004. “Negotiating between bourge and racaille: Verlan as youth identity practice in
suburban Paris.” In Negotiation of Identities in Multilingual Contexts, A. Pavlenko & A.
Blackledge (eds). Clevedon, Uk: Multilingual Matters.
Duranti, A. 2004. “Agency in language.” In A Companion to Linguistic Anthropology, A. Duranti
(ed.), 451–473. Malden, MA: Blackwell.
Fonseca-Greber, B. and Waugh, L. 2003a. “On the radical difference between the subject per-
sonal pronouns in written and spoken European French.” In Corpus Analysis: Language
Structure and Language Use, P. Leistyna and C. Meyer (eds), 225–240. Amsterdam: Rodopi.
Pronominal choice in French conversational interaction 

Fonseca-Greber, B. & Waugh, L. 2003b. “The subject clitics of European conversational French:
Morphologization, grammatical change, semantic change, and change in progress.” In A
Romance Perspective on Language Knowledge and Use, R. Nunez-Cedeno, L. Lopez and R.
Cameron (eds), 99–117. Amsterdam: Benjamins.
Freyne, M. 1990. “The strange case of on.” In Muhlhausler and Harre 1990, 178–192..
Gadet, F. 1997. Le français ordinaire. 2e edition. Paris: Armand Colin.
Gal, S. 1998. “Multiplicity and contention among language ideologies: A commentary.” In Schi-
effelin, Woolard, and Kroskrity, eds, 317–331.
Giles, H., Coupland, N., and Coupland, J. 1991. “Accommodation theory: Communication, con-
text, and consequence.” In Contexts of Accommodation: Developments in applied sociolin-
guistics, H. Giles, J. Coupland and N. Coupland (eds), 1–68. Cambridge: Cambridge Uni-
versity Press.
Giles, H., & Powesland, P. (1997). “Accommodation Theory.” In Sociolinguistics: A Reader, N.
Coupland and A. Jaworski (eds), 232–239. New York: St. Martin’s Press.
Goffman, E. 1959. The Presentation of Self in Everyday Life. NY: Doubleday.
Goodwin, C. and Duranti, A. 1992. Rethinking Context: Language as an Interactive Phenomenon.
Cambridge: Cambridge University Press.
Gumperz, J. 1982. Language and Social Identity. Cambridge; Cambridge University Press.
Hanks, W. 2005. “Pierre Bourdieu and the practices of language.” Annual Review of Anthropol-
ogy 34, 67–83.
Hansen, J. and Liu, J. 1997. “Social identity and language: Theoretical and methodological is-
sues.” TESOL Quarterly 31.
Holquist, M. 1990/2002. Dialogism: Bakhtin and his world. London: Routledge.
Jacoby, S. & Ochs, E. 1995. “Co-construction: An introduction.” Research on Language and So-
cial Interaction 28, 171–183.
Joseph, J. 2004. Language and Identity: National, Ethnic, Religious. New York: Palgrave Macmillan.
Kramsch, C. 2000. “Social discursive constructions of self in L2 learning.” In Sociocultural Theory
and Second Language Learning, J. Lantolf (ed.), 133–153. Oxford: Oxford University Press.
Kroskrity, P. 2000. “Regimenting languages: Language ideological perspectives.” In Regimes of
Language: Ideologies, Polities, and Identities, P. Kroskrity (ed.). Santa Fe, NM: School of
American Research Press.
Laberge, S. and G. Sankoff. 1980. “Anything you can do.” In The Social Life of Language, G.
Sankoff (ed.). Philadelphia: University of Pennsylvania Press.
Le Bel, Edith. 1991. “Le statut remarquable d’un pronom inaperçu.” La Linguistique 27, 91–109.
Lindenfeld, J. 2000. The French in the United States: An Ethnographic Study. Westport, CT: Bergin
& Garvey.
Mendoza-Denton, N. 2002. “Language and identity.” In The Handbook of Language Variation
and Change, J. Chambers, P. Trudgill, and N. Schilling-Estes (eds), 475–499. Malden, MA:
Oxford Univ. Press.
Milroy, L. 2002. “Conversation, spoken language, and social identity.” In Style and Sociolinguistic
Variation, P. Eckert, and J. Rickford (eds). Cambridge: University Press.
Muhlhausler, P. & Harre, R. 1990. Pronouns and People: The Linguistic Construction of Social and
Personal Identity. Oxford: Basil Blackwell.
Norton, B. 2000. Identity and Language Learning: Gender, Ethnicity and Educational Change.
Harlow, England: Longman
Norton, B. and Toohey, K. 2002. “Identity and language learning.” In The Oxford Handbook of
Applied Linguistics, R. Kaplan (ed.), 115–123. NY: Oxford University Press.
 Linda R. Waugh

Offord, M. 1990. Varieties of contemporary French. London:Macmillan.

Paulston, C. 2004. “Biculturalism: Some reflections and speculations.” In Intercultural Discourse
and Communication, Kieslien and C. Paulston (ed.), 255–276. Malden, MA: Blackwell.
Pavlenko, A. and Lantolf, J. 2000, “Second language learning as participation and the (re)con-
struction of selves.” In Sociocultural Theory and Second Language Learning, J. Lantolf (ed.),
155–178. Oxford: Oxford University Press.
Peeters, Bert. 2006. “‘Nous on vous tu(e)’: La guerre pacifique des pronoms personnels.”
Zeitschrift für romanische Philologie 122, 201–220.
Schiffrin, D. 1996. “Narrative as self-portrait: Sociolinguistic constructions of identity.” Lan-
guage in Society 25, 167–203.
Sebba, M. and Wooton, T. 1998. “We, they and identity: Sequential versus identity-related expla-
nation in code-switching.” In Code-Switching in Conversation: Language, Interaction and
Identity, P. Auer (ed.), 262–289. London: Routledge.
Silverstein, M. 1976. “Shifters, linguistic categories, and cultural descriptions.” In Meaning in
Anthropology, K. Basso and H. Selby (eds), 11–55. University of New Mexico Press.
Silverstein, M. 2003. “Indexical order and the dialectics of sociolinguistic life.” Language and
Communication 23, 193–230.
Simpson, P. 1993. Language, Ideology and Point of View. London: Routledge.
Soll, L. 1983. “Situer on ‘nous’ en français moderne.” In Studies in Descriptive Linguistics, F.
Hausmann (ed.), 7–15. Heidelberg: Julius Groos Verlag.
Stewart, M. 1995. “Personally speaking... or not? The strategic value of on in face-to-face nego-
tiation.” Journal of French Language Studies 5: 203–223.
Tabouret-Keller, A. 1997. “Language and identity.” In The Handbook of Sociolinguistics, F.
Coulmas (ed.), 301–314. Oxford: Blackwell.
Thompson, J. 1991. “Editor’s introduction.” In P. Bourdieu, Language and Symbolic Power, 1–31.
Cambridge, MA: Harvard University Press.
Van Dijk, T. 1989. Ideology: A Multidisciplinary Approach. London: Sage.
Vice, Sue. 1997. Introducing Bakhtin. Manchester: University Press.
Wales, K. 1996. Personal Pronouns in Present-Day English. Cambridge: Cambridge University
Waugh, L. 1982. “Marked vs. nmarked: A choice between unequals in semiotic Structure.” Semi-
otica 38, 299–318.
Waugh, L. 2008. “Shifting Linguistic Identities in Interaction: The Case of a French-English Bi-
lingual”. Readings in Language Studies, Vol. I: Language Across Disciplinary Boundaries, M.
Mantero, P. Miller, J. Watzke (eds), 211–228. St. Louis, MO: International Society for Lan-
guage Studies, Inc.
Waugh, L. & Fonseca-Greber, B. 2002. “Authentic materials for everyday spoken French: Corpus
linguistics vs. French textbooks.” Arizona Working Papers in SLAT 9,114–27.
Waugh, L., Fonseca-Greber, B., Vickers, C. and Eroz, B. 2007. “Multiple empirical paths to a
complex analysis of discourse.” Methods in Cognitive Linguistics, M. Gonzalez-Marquez, M.
Spivey, I. Mittelberg, and S. Coulson (eds), 120–148. Amsterdam: John Benjamins.
Wertsch, J. 1991. Voices of the Mind: A Sociocultural Approach to Mediated Action. Cambridge:
Harvard University Press.
Wodak, R., de Cillia, R., Reisigl, M. and Liebhart, K. 1999. The Discursive Construction of Na-
tional Identity. Edinburgh: Edinburgh University Press.
Woolard, K. 1998. “Introduction: Language ideology as a field of inquiry.” In Language Ideolo-
gies: Practice and Theory, B. Schieffelin, K. Woolard, and P. Kroskrity (eds), 3–47.
Constructing interpersonal relations
in the discourse of Russian media

Marjatta Vanhala-Aniszewski

In this paper the relations between participants in the interaction actualized

in the Russian media discourse are investigated. Attention will be paid, firstly,
to the role and identity of the writer and the reader and, secondly, to the force
of the writer’s statements meant to influence the audience, i.e. his personal
attitude to the truth of propositions of the text. The empirical data consist of
newspaper articles representing ordinary news reports. Concrete analysis of the
realization of the interpersonal relations is provided by applying the theory of
critical discourse analysis. This investigation of Russian media texts showed that
the news texts include quite a few markers indicating the explicit presence of
interpersonal relations between the participants of the communication act.

1. Introduction

As a result of increasing globalization and rapid spread of the new information tech-
nology in the post-modern world, the mass media have obtained new features. “Old”,
mainly written, forms of communication (newspapers, journals, radio etc.) now have
a serious competitor from modern mass media (TV, Internet and multifunctional
phones). Thus, sender-oriented means of communication have in many areas been
replaced by reader-oriented or, more precisely, interaction oriented means of commu-
nication. This development also characterises contemporary Russian media, which
have changed completely in the past two decades. I.A. Sternin (2003: 99–100, 105, see
also Kon’kov et al. 2004: 67–68) speaks about “the personification and an oralization
of the Russian media” and notes that those features typical of spoken language also
occur in media discourse; being an essential part of social life and, consequently, fo-
cusing on communication between people, media texts (discourse) tend more and
more to interact with the audience and establish social relations, i.e. links, with them.
In this paper the relations between participants in the interaction actualized in the
Russian media discourse are investigated. Attention will be paid mainly to the role and
identity of the writer and the reader, as well as to the interaction between them, on the
one hand, and to the force of the writer’s statements meant to influence the audience,
 Marjatta Vanhala-Aniszewski

i.e. his personal attitude to the truth of propositions of the text, on the other. The pur-
pose of the present study is to establish what sort of interpersonal rules govern the
Russian media discourse and which linguistic resources are preferred for realising
them. Consequently, the interpersonal relations occurring in media texts are analyzed
on both the macro and micro levels. As Fairclough (1995: 126) notes, analysis of the
relations and identities of the participants in media texts is important for understand-
ing the socio-cultural questions of a given society. Such a study is especially important
today when the power and influence of the mass media on different levels of societies
continue to increase.
The empirical data used in this study consist of Russian newspaper articles repre-
senting ordinary news reports (see discussion on news in van Dijk 1988; Pietilä 1995;
Scollon 1998; Heikkinen et al. 2005). The data (a total of 899 articles) were collected
during January–December 2005 from the electronically stored database Integrum
( which contains a large number of Russian newspapers, journals,
laws, etc. The data include articles from ten newspapers: Komsomol’skaia Pravda (KP),
Moskovskii komsomolets (MK), Izvestiia (I), Argumenty i Fakty (AiF), Kommersant (K),
Novaia gazeta (NG), Molodezhnaia gazeta from Ufa (MG/Ufa), Molodezhnaia gazeta
from Ul’ianovsk (MG/Ul), Molodezhnaia gazeta from Chita (MG/Ch) and Molodezh
Iakutii from Iakutsk (MIA).1 The common theme of these texts is Europe and, in a
broad sense, the European Union. The texts deal, among other things, with economic
and political questions. One of most important topics that year was the constitution of
the European Union, which has been widely discussed not only in member states of
the EU, but also in Russia, where EU enlargement is actively monitored. In analyzing
the texts and their topics, one must, however, not forget that “the ‘content’ of newspa-
pers is not facts about the world, but in a very general sense ‘ideas’” (Fowler (1991, 1).
In constructing these ideas the role of the writer, his beliefs and attitudes play an im-
portant role: the writer’s commitment to the content and his evaluation are able to
influence (and even to manipulate) the reader and his reception of information.

2. Investigating interaction

When interacting with other people, we use a multiplicity of communicative channels;

language (written and spoken), gestures (facial expressions), body language, etc. In
this qualitative study the interaction is studied only from the standpoint of how it is
realized in written media texts. The role of texts (language, discourses) in construction
of the social world is emphasized by several theories of social constructivism (see e.g.
Faigley 1985; Berkenkotter & Huckin 1993). The relation between the discourses

1. Newspapers Izvestiia, Argumenty i Fakty, Kommersant and Novaia gazeta were represented
only by the June 2005 numbers when the referendum of the constitution of EU was actively
discussed in the mass media.
Constructing interpersonal relations in the discourse of Russian media 

(texts) and the social practice is, however, two-way: discourses not only reflect the
social world but also construct it. This concept of the social nature of language has
been adopted by the systemic-functional theory for understanding language as a social
phenomenon, as a means of interaction between members of a certain society (Halliday
1978, 1985; Halliday & Matthiessen 2004). Based on this view, language and language
use must be analyzed as actions, behaviour between people where the meanings and
the functions of the language are most important, not the linguistic structures.
According to the systemic-functional model, texts are multi-functional, i.e. they
simultaneously fulfil three main social functions: ideational, interpersonal and textual.
This means, firstly, that texts represent the experience of the users (ideational func-
tion), secondly, they shape social relations between the participants of the communi-
cation act, as well as express their attitudes to the content of the text (interpersonal
function). Finally, the textual function organizes representations determined as a co-
hesive and coherent text (Halliday & Matthiessen 2004: 61, see also Fairclough 2003: 27).
In the present study, attention will be paid only to the interpersonal function of the
language to construct “a dialogue”, i.e. social relations between communicants. This
function has been investigated in various studies concerning metadiscourse and, in
particular, that of academic texts in different languages (see e.g. Crismore and Vande
Kopple 1997; Mauranen 1992, Hyland 2000, Vanhala-Aniszewski 2006).
Concrete analysis of the realization of the interpersonal relations of texts is pro-
vided by applying the theory of critical discourse analysis (Fairclough 1992, 1995, 2003;
Chouliaraki & Fairclough 1999; Wodak 1996; Wodak et al. 1999; see also van Dijk 1997).
As an approach it allows us to combine both the linguistic and social levels of language
use. Based on Halliday’s concept of systemic-functional grammar, this approach takes
into account how the writer’s native culture and the socio-cultural context (discursive
and social practices) influence his writing habits and consequently, his choices of lin-
guistic structures. Critical discourse analysis offers instruments for analysing texts in
detail: what kinds of linguistic means are used to construct the interaction and what
reasons may lie behind a certain realization of the participants in media texts. It is par-
ticularly interesting to determine the special features created by Russian traditions of
expressing relations between the communication partners in media text.
As stated above, the interpersonal function of the language is linked to the social
relations and social identities in the communication. The interaction between the par-
ticipants in media texts is actualized on two planes: personal and subjective position
markers (see relational and positional markers in Stillar 1998: 32–45). The first group
of markers includes three major categories: writers (reporters), readers and ‘other par-
ticipants’, mainly from the public domain (e.g. politicians) (Fairclough 1995: 125). The
relationship between these categories can be explicit or implicit, formal or informal,
close or distant.
On the linguistic level of the text, the participants in the interaction are realized,
first of all, by the personal markers (I, we, you can see, he, according to Smith). Hyland
(2000: 112–113) divides these markers into two subgroups: on the one hand, the person
 Marjatta Vanhala-Aniszewski

markers explicating the authorial presence in the text and on the other hand, the rela-
tional markers, drawing the reader into the communication process. In the classifica-
tion of Hyland the third person (more precisely the source of the information), briefly
discussed here, presents a textual function (see evidentials in Hyland 2000: 111; see
also Vande Kopple 1985, Mauranen 1992). Among the personal markers, as Wodak et
al. (1999: 45) note, the pronoun ‘we’ is the most complex because it can, except for the
word ‘us’, refer to other persons and groups of people, as will be seen below. In addition
to explicit highlighting, the participants of the communication can also be effaced, for
instance, by using passive and impersonal constructions, nominalizations, etc.
When transmitting information, the writer, knowingly or not, also mediates his/
her own authorial judgments, commitments, attitudes, etc. to the content, i.e. to the
truth status of the propositions. These subjective meanings can be actualized by valid-
ity and attitude markers. The validity markers include, on the one hand, hedges. These
express the writer’s commitment to the statements of the propositional content
(probably, might, perhaps) within an indication that the information given is presented
as an opinion. The writer announces that he is not certain about the truth value of his
claim (Lakoff 1972: 195). On the other hand, validity markers consist of emphatics
(see boosters in Hyland 2000: 111–112) conveying certainty about the proposition
(of course, clearly). Thus, validity markers express different degrees of the writer’s com-
mitment to the statement, either increasing or reducing its force. In contrast, attitude
markers (cf. attitudinals in Stillar 1998: 39), another group expressing the writer’s sub-
jective position, actualize his affective attitudes to the audience or to the information
in the text, for example, importance, surprise, agreement, etc. (unfortunately, surpris-
ingly, it is important that).
It must be borne in mind that the interaction of texts can be created in many other
ways. For instance, devices of spoken language such as colloquial vocabulary and syntax,
dialogues and special use of punctuation are frequently used to imply a familiar relation-
ship, and consequently a sort of interaction between the participants (see Smetanina
2002, Sternin 2003, Skovorodnikova 2004, Vanhala-Aniszewski 2007). In addition, some
discourse particles (Schiffrin 1987) are also able to express the writer’s certainty con-
cerning the common knowledge of the reader about the facts presented (after all, just).
In this study, however, these colloquial elements are not discussed in detail.
All the interpersonal elements mentioned are of great importance for the readers’
understanding of the context; because these elements make the text more informal, the
readers obtain from them commentaries, i.e. judgments reflecting the writer’s opinion
about and position on the content.2 These elements guide the reader, helping him to

2. Crismore and Vande Kopple (1997) investigated the effects of hedging (e.g., devices that
express the writer’s commitment to the truth value of the proposition being made) on reading
retention. Experimental groups read passages from both social studies and science textbooks
containing various types of hedging, while a control group read corresponding passages from
which all hedging had been removed. According to the results, students learned more from
reading science and social studies passages with hedging included.
Constructing interpersonal relations in the discourse of Russian media 

orient in the content. When using them, the writer draws the readers into an implicit
dialogue with himself (see Mauranen 1992: 156, 178–179). The writer must, however,
be careful not to overstate and invite readers’ rejection.

3. Organization of the interaction in Russian media texts

As mentioned above, the Russian media have undergone great changes since the 1990s.
Nowadays many newspapers and journals (e.g. boulevard journals and tabloids) that
did not exist in the Soviet time are being published. Some of them, e.g. Ogonek, have
changed completely since the perestroika (Kakorina 2003: 243; Zasurskii 2002:
221–231). In general, the publishing policy differs from that of the Soviet era. There are
no longer taboo subjects; and newspaper articles seem to centre on, e.g., the private life
of politicians and famous artists, etc. Thus, it is obvious that as a result of the enormous
changes in the society also the language, and in particular that of the media, has un-
dergone important changes: the “old”, quite formal and distant Soviet style of writing,
has been given up and instead, various features of the spoken language are becoming
overstated. One must note that as a result of the democratization of society, also the
language and the conventions for using it have become democratized; consequently,
conversational style and entertainment play an increasing role in media texts (Solganik
2003: 263, Kon’kov et al. 2004: 70–72, Kostomarov 1999: 78–109). Klushina (2003: 283),
who investigated changes in the discourse of the Russian media, states that the use of
language in contemporary media has indeed begun to include many individual fea-
tures; for example, having in Soviet time been a passive object of communication, the
reader is now understood as an equal participant in the act of communication.
Here the main emphasis will be placed on how the writer and the reader are re-
ferred to. The third person, the other participants – presented in a majority of the news
texts by politicians and different experts – are not studied in more detail. It is, never-
theless, necessary to note that the third person in the texts analyzed is drawn into the
interaction through presentation of explicit and direct dialogues between him and the
journalist. Skovorodnikova (2004: 68–69; see also Sternin 2003: 105) points out that
the use of mere dialogues and quotations from the speech of the interviewer has in-
creased since the Soviet era. Kon’kov (1995: 10) also notes that polyphony, i.e. the
presence of different voices, is typical of texts in the modern Russian media. In this
way, the writer, on the one hand, creates a direct connection between the reader and
the third person and, on the other hand, makes sure that the information is transmit-
ted accurately as the interviewee stated it. Such a practice of presenting facts and trans-
mitting information, however, can prevent the creation of a cohesive and coherent
text, although it may give the illusion of a direct interaction between the communi-
cants. Below, the interaction between the participants in media texts will be studied on
two planes: personal and subjective position markers.
 Marjatta Vanhala-Aniszewski

3.1 Personal markers

In general, in the texts analyzed here the interaction and the presence of the partici-
pants is not very important: the author is mostly missing or he is detached, implicit.
When transmitting news, the writer seems to remain in the background. The most
explicit way to highlight the writer is, of course, to use a personal pronoun. However,
in the newspapers studied the 1st person singular pronoun occurs in only a few cases,
firstly, when the writer is interviewing a third person and this interview is presented as
such. Secondly, the writer refers to himself with the pronoun ia ‘I’ only when referring
to action processes (see discussion of the classification of the processes in Stillar 1998:
22–25), for example: Ia poobshchalsia s paroi desiatkov evropeiskikh ekspertov i chi-
novnikov (MK 7.6.05) ‘I discussed with a couple of European experts and officials’; Etot
vopros ia zadal zamestiteliu predsedatelia Evropeiskoi komissii Margot Wallström –
(NG 6.6.05) ‘This question I made to Margot Wallström, the deputy chairman of the
European Committee’; Povtorius’, zakony,– sushchestvuiut uzhe okolo 10 let! (MG/Ul
14.1.05) ‘I repeat, the laws – have already existed for 10 years!’. On the contrary, when
expressing a subjective opinion, judgment, etc., ‘I’ is never used. Such defocusing of the
first person singular is, in general, typical of official written text in Russian (see e.g. on
the expression of the writer in Russian academic texts in Vanhala-Aniszewski 1997).
Instead of ia, to refer to the writer, the pronoun my ‘we’ is preferred. This pronoun,
as stated above, has different functions and references: it can be either an ‘addressee-
inclusive’ or an ‘addressee-exclusive we’, as well as something between a ‘speaker-inclu-
sive’ and a ‘speaker-exclusive we’. One can also use the terms ‘author’s plural’, ‘pluralis
modestiae’, ‘pluralis maiestatis’, etc. (Wodak et al. 1999: 45.) The last type, pluralis
maiestatis, is used in the media mostly when direct speech is presented, e.g. when a
politician speaks on behalf, for example, of the government, the party. Pluralis modes-
tiae is characteristic of academic texts (see e.g. Lapteva 1995). In contrast, the author’s
plural is frequent in Russian media texts when referring, firstly, to the writer alone:
Napomnim, chto v ES proizoshel raskol po voprosu o subsidiiakh – (MK 24.6.05.) ‘We
remind you that there was a break concerning the appropriation in the EU –’ and,
secondly, to some journalists of the staff including the writer. In addition, ‘we’ can also
include both the speaker and the addressee:
(1) No, dopustim, proizoshlo nemyslimoe, i Rossiiu vdrug priglasili v ES. No
nado li eto nam? (MK 7.6.05.)
‘But, let us assume that the unbelievable happens, and Russia suddenly is in-
vited to join the EU. But do we need this?’
The pronoun ‘we’ is able to refer very globally to ‘us, all Russians’ including the writer
and the reader:
Constructing interpersonal relations in the discourse of Russian media 

(2) Rossiiane tol’ko-tol’ko nachali osoznavat’ sebia evropeitsami. Nam budet slo-
zhnee lavirovat’ i v obshchemirovom politicheskom protsesse. (AiF 22.6.05.)
‘The Russian people just began to realize themselves as Europeans. It will be
complicated for us to maneuver in the world-wide political process.’
Here the main focus of ‘us’ is, of course, on the politicians, but also more widely, on ‘us
Russians’. The same concerns the example: Nam by vashi zaboty (I 9.6.05) ‘If we only
had your problems’ where the possessive pronoun ‘us’ refers to all Russians while the
referent of ‘you’ is not the reader(s), but the Europeans. The text deals with the system
of education in Europe under the headline My sporim ob obrazovanii, a evropeitsy o
ravenstve ‘We are arguing about education, while the Europeans argue about equality’.
In this sentence ‘we’ is understood as ‘our politicians, the decision-makers’ with the
exclusion of the writer. Solganik (2003: 266, see also Sternin 2003: 100) notes that in
such cases ‘we’ implies the irony of the writer observing the situation as an outsider.
The pronoun ‘we’ in the meaning of ‘Russians’ implies the reader (excluding the
speaker) as in the example Tak, stali li my ZHIT’ LUCHSHE? (KP 26.5.05) ‘Have we
really started to LIVE BETTER?’ The use of ‘us’ is near to the so-called paternalistic
‘we’ which refers solely to a ‘you’. The text continues with another ‘we’, implying only
the writer as the subject: Segodnia my podvodim itog nashemu serialu “Zhit’ stalo
luchshe?” ‘Today we shall draw our conclusions about our series “Life became better?”’.
In addition to the possibility of meaning all Russians, ‘we’ can also refer only to some
group, a territory with its inhabitants as in the next example, where the pronoun means
the people of Moscow (including the writer):
(3) Esli v etom dekabre my vnov’ “progolosuem nogami”, pravo bezrazdel’no
rasporiazhat’sia sud’boi stolichnoi vlasti budet na bliudechke prepodneseno
Kremliu. (MK 12.9.05.)
‘If we again “vote with our feet”, the right to make decisions concerning the
faith of the capital will be given to the Kremlin.’
Here, the writer, by assuming shared knowledge between himself and the reader, im-
plies togetherness and engagement in pursuing a common goal.
In the texts analyzed the writer is not very strongly aware of the reader, even
though by paying attention to him he has a better possibility to influence the reader
and his values. When referring to the audience directly, the 2nd person of singular ty
‘you’ is not preferred, but this pronoun is used in some cases in its generic meaning
(‘anybody’): Pokupai ili proigraesh’! (MK 8.6.05) ‘Buy and you will lose!’; Kogda u tebia
est’ vse ili pochti vse, ty ved’ ne postavish’ vse eto na kon neizvestno radi chego, net tak li?
(K 2.6.05) ‘Having everything or almost everything you will not risk your neck with-
out knowing for what, will you?’ The use of ty in referring to a concrete person is in
Russian typical only of intimate and familiar acts of communication, e.g. between the
members of a family. People can work at the same place for many decades, and still
 Marjatta Vanhala-Aniszewski

they turn to each other using the polite and distant pronoun of the second person
plural vy ‘you’.
In texts the reader is paid attention to by imperatives, but only when it is a ques-
tion of a process of concrete action. The imperative as well as the interrogative con-
struct a direct interaction between the writer and the reader. On the other hand, com-
mands in the imperative form may indicate the authority of the writer (Stillar 1998:
34), for example:
(4) Poprobuite populiarno ob’iasnit’ obychnomu grazhdaninu evropeiskoi
liberal’noi demokratii s normal’nymi zakonami, zachem pomimo etikh za-
konov nuzhna eshche kakaia-to obshcheevropeiskaia konstitutsiia, i vy, sko-
ree vsego, ne naidete dokhodchivykh argumentov. (K 2.6.06.)
‘Try to explain to an ordinary citizen of a European democratic country that
has liberal laws why another common European constitution is still needed,
and you, as is supposed, cannot find good arguments. You’ll find that there are
no arguments.’
When an opinion, an evaluation and the values of the reader are concerned, the in-
tended reader remains implicit. Simultaneously, the 2nd person plural vy ‘you’ is also
used only when speaking about concrete actions, for example, when giving instruc-
tions to do something: podrobnee o peregovorakh chitaite na str. – (I 15.6.05) ‘You can
read in detail about these negotiations on page –’; Tak chto, esli vy otkroete rublevyj
schet pod 12% godovykh i vyshe, schitaite, chto ot rosta tsen zastrakhovalis’ (KP 28.6.05)
‘If you open the account in rubles with the interest of 12 per cent your money is in safe’.
Such use, naturally, is not very typical of news that transmits events, facts, etc. In only
two cases was the audience referred to by using the noun chitatel’ ‘reader’: Spetsial’no
dlia chitatelei “KP” my uznali podrobnosti (KP 22.7.05) ‘Especially for the reader of
“KP” we found out details’. Here, the writer (‘we’) as well as the addressee (reader) are
referred to but over a distance.
In the examples above, the linguistic resources, more or less explicitly demonstrat-
ing the presence of the person, writer or reader, have been analyzed. In addition, the
“hidden power” of the writer is, for example, expressed by declarative sentences used
mainly to convince the addressee. The declarative meaning takes the truth value of the
proposition for granted, as a fact, and implies simultaneously some form of power. The
same is also true for tense: by using the present tense the writer includes a categorical
modality, i.e. takes the factuality for granted. He is also authoritative; the present tense
form implies that the writer is telling what the case is in no uncertain terms.
(See Fairclough 1992: 159.)
A special means of expressing personal relations and, in particular, of drawing the
attention of the reader is to use rhetorical questions and to offer the reader a possibil-
ity for inferring about the issue discussed (see Mauranen 1992: 180). Such questions
are frequent in the texts analyzed here. Their function is to express solidarity with the
reader concerning the content of the presented information:
Constructing interpersonal relations in the discourse of Russian media 

(5) Lozh’ dezorganizuet zhizn’ strany. Kakoi narod zakhochet rabotat’ vo imja bu-
dushchego Rossii, esli emu ne govoriat pravdy o nastoiashchem? (AiF 29.6.05.)
‘Falsity disorganizes the life of our country. How can people work for the fu-
ture of Russia if they are not told the truth about today’s life?’
Since the writer cannot expect an answer to the questions from an imagined reader, he
reacts himself:
(6) Kak vy dumaete, pochemu eti tsifry tak veliki imenno v Rossii? Otvet, po
mneniiu medikov, lezhit na poverkhnosti: slishkom mnogo p’em! (KP 2.3.05.)
‘What do you think, why are these numbers [statistics showing the amount of
alcohol drunk in Russia – MVA] so high? The answer, according to the doc-
tors, is that we drink too much!’
Here again ‘we’ is used as speaker-exclusive. On the other hand, the writer can choose
a less direct approach; that is, he uses a circumlocution, an effacement to refer to him-
self and/or the reader. In the media texts, one of the most popular grammatical means
is the use of modal auxiliary verbs (moch’ ‘can’, khotet’ ‘want’) and modal adverbs (nado
‘necessary’, nuzhno ‘necessary’, mozhno ‘possible’, neobkhodimo ‘necessary’, obiazatel’no
‘necessarily’) having epistemic as well as deontic meaning and able to refer not only to
the writer and reader, but also to people in general. In the example:
(7) Nado polagat’, chto seichas Zapadnaia Evropa s opaskoi otnesetsia k planam
Ukrainy po povodu skoreishego vstupleniia v ES. (AiF 1.6.05.)
‘It is can be supposed that Western Europe looks at the plans of Ukraine with
the writer is implied by using a structure nado polagat’ (‘it can be supposed’). In the
(8) Padaet uvazhenie k sobstvennoi strane. Kak mozhno otnosit’sia k strane i
vlasti, esli militsii ne doveriaiut 98% naseleniia, pravitel’stvu – 70%, a deputa-
tam – pochti nikto. (AiF 29.6.05.)
‘Respect for our own country is going down. What kind of relationship can
one have with the country if 98% of the citizens do not trust the policy, 70%
mistrust the government and almost nobody trusts the deputies?’
the modal adverb mozhno refers to ‘us, Russians’; here it is a question of an average
Russian. The same concerns the modal adverb mozhno in the next example:
(9) Rvenie novykh chlenov ES vpolne mozhno poniat’: delezh britanskogo piroga
s likhvoi okupil by liubye zhertvy. (K 20.6.05.)
‘The appetite of new members of the EU can be understood: the division of
the British “cake” would quite quickly compensate all their sacrifices.’
 Marjatta Vanhala-Aniszewski

Here again, it is ‘us’ in general who are implied, however, not just ‘any of us’ but rather
sophisticated ones who understand the European political situation.
The direct, concrete reader is effaced in Na otdykh mozhno otpravit’sia i s plastiko-
voi kartoi (KP 9.6.05) ‘One can go on holiday with a credit card’; No eto ne znachit, chto
nado idti i meniat’ vse den’gi na “zelen’” (MK 8.6.05) ‘That does not mean that one has
to change all one’s money to “bucks”’. The first text discusses the general possibilities of
the Russians to spend their holidays abroad. The writer gives some advice to the pas-
sengers by using an impersonal structure which includes the modal adverb mozhno
with reference to the addressee. In the second sentence, the adverb nado ‘necessary’
also implies the reader as a hidden subject. Although the question is of the action pro-
cesses, the writer does not use personal pronouns or other explicit elements in turning
to the addressee: his aim is to give advice in general, not only to one group, the readers.
On the other hand, a special subgroup in the audience can also be indicated by using
a modal adverb. In No po krainei mere k nemu nado byt’ gotovymi (I 9.6.05) ‘But at least
for this [joining of the Ukraine to the EU – MVA] one must prepare’ a subgroup ‘poli-
ticians’ is referred to.
The texts include various lexical elements able to refer either to ‘us together’ or
only to the reader. The use of the noun reader was already mentioned above. Other
means of lexicalization also occur, for example, Rossiiane ‘Russians’, nashi liudi ‘our
people’, rossiiskii chelovek ‘Russian person’. Such nouns may also include self-reference,
but their main function is, on the one hand, not to refer too directly, too personally,
too intimately and, on the other, still to transmit the feeling of togetherness of the
writer and the reader, i.e. solidarity between them (cf. ‘our people’).

3.2 Subjective position markers

The presence of the writer and his subjective position towards the content of the texts
is also realized by the validity markers expressing his commitment to the propositions.
Hedges express a view concerning the truth value of a proposition. They tone down
the certainty and indicate that the writer is presenting opinions rather than facts. The
modal verbs and adverbs mentioned above, as well as the epistemic adverbs, are used
in this function: vozmozhno ‘probably’, mozhet byt’ ‘perhaps, maybe’;
(10) Narod khochet nakonets poniat’ i drugoe. Mozhet byt’, samoe vazhnoe. Tol’ko
li v zlykh oligarkhakh prichina ego bednosti? (AiF 8.6.05.)
‘People want to understand something else. Maybe the most important: is it
possible that oligarchs are the reason for our poverty?’
(11) I dazhe odobrennyi priem Bolgarii i Rumynii, vozmozhno, budet otlozhen do
2008 goda. (I 17.6.05.)
‘Even the joining of Romania and Bulgaria to the EU probably will be put off
until 2008.’
Constructing interpersonal relations in the discourse of Russian media 

Here it can be seen, as Hyland (2000: 88) states, that in the communication act not
only ideas are transmitted but also the writer’s commitment, his epistemic status to
those facts as facts or interpretations (see also Fowler 1991). The commitment of the
writer, as well as his uncertainty, is also expressed by using the conditional mode and
such unipersonal verbs as kazhetsia ‘it seems’. In summary, hedges allow writers to
dispute their interpretation with the reader or at least to create the image of such an
Another group of validity markers, the emphatics, convey certainty; iavno ‘obvi-
ously’, navernoe ‘certainly’, iasno ‘clearly’, konechno ‘of course’, etc. These markers em-
phasize the force of the statement. Simultaneously they mark involvement and solidar-
ity with the readers (Hyland 2000: 87). Stressing shared information, they imply
togetherness and direct engagement with the readers:
(12) Navernoe, prav byl Saltykov-Shchedrin, kogda govoril, chto “rossiiskaia
vlast’–” (AiF 29.6.05.)
‘Certainly, Saltykov-Shchedrin was correct in noting that Russian authority is –’
Emphatics are used to accomplish the persuasive goals: Neiasno i to, kakoe reshenie
budet priniato otnositel’no dal’neishei ratifikatsii Evrokonstitutsii (I 2.6.05) ‘It is not ob-
vious what decision will be made concerning the ratification of the European constitu-
tion’; U nas narod, konechno, terpelivyi, no – (KP 3.3.05) ‘Our people, certainly, are
tolerant, but –’. Emphatics promote the value of the content of a statement: they seek
to convince the reader through the writer’s belief in this truth. They can, however, also
construct the writer as an authoritative person who is, in a mood of solidarity, counsel-
ling the reader. In this way, emphatics balance objective information with subjective
evaluation and interpersonal negotiation (see Hyland 2000: 101).
Compared to validity markers signalling the writer’s degree of commitment and cer-
tainty, the attitude markers indicate his emotional, affective relation to the proposition:
(13) V etom sluchae ochen’ vazhno gramotno rasporiadit’sia imeiushchimisia fin-
ansami – (MG/Ul 20.5.05.)
‘In this case it is important that one can competently put the budget in order –’
(14) Interesno, chto v Velikobritanii istoriia izuchaetsia cherez bytovoi podkhod,
“bez politiki”. (MIA 6.6.05.)
‘It is interesting that in Great Britain history is studied through everyday life,
“without politics”.’
(15) I chto udivitel’no: nesmotria na to, chto dokhod kazhdogo rossiiskogo zhil’tsa
v dva raza men’she, chem srednii mirovoi – (MG/Ch 21.4.05.)
‘It is most surprising that the income of every Russian inhabitant is two times
lower than that the world average –’
The texts analyzed are not characterized by affective markers. On the other hand, it is
understood that the readers do not expect extra subjective and emotional opinions
 Marjatta Vanhala-Aniszewski

where the news and the objective information, supposed to be included by them, are

4. Discussion

This investigation of Russian media texts indicated that, in particular, the news texts
include quite a few markers indicating the explicit presence of interpersonal relations
between the participants of the communication act. The writer mostly hides behind
‘us’, leaving open whether he alone or the editorial staff is responsible for the state-
ments in the text. Furthermore, in addition to the writer and the audience, ‘we’ can
include ‘all Russians’. In this case the pronoun is also able to exclude the writer and
refer only to the readers, or to some subgroups of readers, e.g. ‘politicians’, ‘officials’, etc.
This ‘we’ is the most frequent means of expressing the reader, who is mostly ignored.
Only rhetorical questions pay attention to the reader; but, in fact, also these elements
are used by the writer for creating a cohesive text: within the rhetorical questions the
writer moves from one subtopic to another in his text.
The use of the collective ‘we’ indicates the togetherness of ‘us’ typical of Russian
society. In political history as well as in the Orthodox religion the collective of the
people remains more important than the individual. In addition, the pronoun simul-
taneously indicates the informal and familiar relation of the writer (‘we together’, ‘you
and me’) to the addressee. On the other hand, the style and means of communication
are changing; and interpersonal expressions are more frequent and personal, e.g. in
TV and in the journals for the young.
In the effacement of the writer and the reader, modal verbs and adverbs are used.
Simultaneously, these elements are also able to transmit the commitment of the writer
to the content: they detach the personality and the individual’s responsibility for the
truth value of the proposition. Validity markers having the function of indicating the
epistemic modality and the certainty and commitment of the writer to the text content
are not very widely used in Russian news texts. This means that the writer tends to
express facts, not his own interpretation of the facts. This is confirmed by a small num-
ber of affective markers. As a result of the conversationalization of contemporary text
in the Russian media, one could expect more such signals; but again, we are dealing
with news texts.
In summary, one must conclude that the tendency of the media texts to the per-
sonification mentioned above does not concern news texts and the interpersonal rela-
tions presented in them. In news texts the writer as an individual actor as well as his
subjective attitude to the information and the audience still remain in the background.
The main attention is paid to transmission of facts, which, however, are described by
many colloquial and familiar structures, capable in a special way of constructing a
“bridge” between the writer and the reader.
Constructing interpersonal relations in the discourse of Russian media 


Berkenkotter, C. & Huckin, T.H. 1993. Rethinking genre from a sociocognitive perspective.
Written Communication 10 (4): 475–509.
Chouliaraki, L. & Fairclough, N. 1999. Discourse in Late Modernity: Rethinking Critical Dis-
course Analysis. Edinburgh: University Press.
Crismore, A. & & Vande Kopple, W. J. 1997. Hedges and readers: effects on attitudes and learn-
ing. In Hedging and Discourse: Approaches to the analysis of a pragmatic phenomenon in aca-
demic texts, S. Markkanen & H. Schröder (eds), 83–114. Berlin: Walter de Gruyter and Co.
Faigley, L 1985. Nonacademic writing. The social perpective. In Writing in non-academic set-
tings. L. Odell and D. Goswani (eds), 231–248. New York: Guilford Press.
Fairclough, N. 1992. Discourse as Social Chance. Cambridge: Polity Press.
Fairclough, N. 1995. Media Discourse. London: Edward Arnold.
Fairclough, N. 2003. Analysing Discourse. Textual Analysis for Social Research. London and New
York: Routledge.
Fowler, R. 1991. Language in the News. Discourse and Ideology in the Press. London and New
York: Routledge.
Halliday, M.A.K. 1978. Language as Social Semiotic: The Social Interpretation of Language and
Meaning. London: Edward Arnold.
Halliday, M.A.K. 1985. An Introduction to Functional Grammar. London: Edward Arnold.
Halliday, M.A.K. & Matthiessen, Ch. M.I.M. 2004. An Introduction to Functional Grammar.
Third edition. London: Edward Arnold.
Heikkinen, V. & Lehtinen, O & Lounela, M. 2005. Lappeenrantalaismies löi toista nenään baa-
rissa. Uutisia ja uutisia. In Tekstien arki. Tutkimusmatkoja jokapäiväisiin merkityksiimme, V.
Heikkinen (ed.), 231–258. Helsinki: Gaudeamus.
Hyland, K. 2000. Disciplinary Discourses: Social Interactions in Academic Writing. Harlow:
Kakorina, U.V. 2003. Sfera massovoi kommunikatsii: otrazhenie sotsial’noi differentsirovan-
nosti iazyka v tekstah SMI. In Sovremennyi russkii iazyk. Sotsial’naia i funktsional’naia dif-
ferentsiatsiia, L.P. Krysin (ed.), 241–276. Moskva: Iazyki russkoi kul’tury.
Klushina, N.I. 2004. “Iazyk publitsistiki: konstanty i peremennye.” Russkaia rech’, 3, 51–54.
Kon’kov, V.I. 1995. Rechevaia struktura gazetnogo teksta. Sankt-Peterburg: Sankt-Peterburgskii
gosudarstvennyi universitet.
Kon’kov, V.I, Potsar, A.N. and Smetanina, S.I. 2004. Iazyk SMI: sovremennoe sostoianie i ten-
dentsii razvitiia. In Sovremennaia russkaia rech’: sostojanie i funktsionirovanie, S.I. Bogdanov,
L.A. Verbitskaia, L.V. Moskovkin and E.E. Iurkov (eds), 67–81. Sankt-Peterburg: Sankt-
Peterburgskii gosudarstvennyi universitet.
Kostomarov, V.G. 1999. Iazykovoi vkus epohi. Sankt-Peterburg: Zlatoust.
Lakoff, G. 1972. Hedges: A study of meaning criteria and the logic of fuzzy concepts. In Papers
from the Eight Regional Meeting of Chicago Linguistic Society, Chicago: Chicago Linguistic
Society, 182–228.
Lapteva, O.A. 1995. “Kak pishut uchenye?” Russkaia rech’ 2: 55–62.
Mauranen, A. 1992. Cultural Differences in Academic Rhetoric. A Textlinguistic Study. Frankfurt
am Main: Peter Lang.
Pietilä, V. 1995. Kertomuksia uutisista, uutisia kertomuksista. Kirjoituksia kolmelta vuosikymme-
neltä. Tampere: Tampereen yliopisto.
 Marjatta Vanhala-Aniszewski

Schiffrin, D. 1987. Discourse markers. Cambridge: University Press.

Scollon, R. 1998. Mediated Discourse as a Social Interaction. A Study of News Discourses. Lon-
don: Longman.
Skovorodnikova, A. P. 2004. “Refleksy postmodernistskoi stilistiki v iazyke rossiiskikh gazet.”
Russkaia rech’ 6, 68–76.
Smetanina, S.I. 2002. Media-tekst v aspekte kul’tury rechi: interstilevoe tonirovanie izlozheniia.
Vestnik SPbGU. Ser. 2, (3) 18: 90–98.
Solganik, G.Ia. 2003. O iazyke i stile gazety. In Iazyk SMI kak ob’ekt mezhdistsiplinarnogo issledo-
vaniia, 261–268. Moskva: Moskovskii gosudarstvennyi universitet.
Sternin, I.A. 2003. Sotsial’nye factory i publitsisticheskii diskurs. In Massovaia kul’tura na rube-
zhe XX-XXI vekov: Chelovek i ego diskurs, 91–108. Moskva: Iazyki russkoi kul’tury.
Stillar, G.F. 1998. Analyzing Everyday Texts. Discourse, Rhetoric and Social Perspectives. London:
Vande Kopple, W. 1985. Some exploratory discourse on metadiscourse. In: College Composition
and Communication 36: 82–93.
van Dijk, T.A. 1988. News as Discourse. Hillsdale: Erlbaum.
van Dijk, T.A. 1997. Discourse as Social Interaction. Discourse Studies: A Multidisciplinary Intro-
duction 1. London: SAGE.
Vanhala-Aniszewski, M. 1997. Kirjoittaja-minä venäjänkielisessä tieteellisessä tekstissä.
Finlance. Vol. XVII, 70–79.
Vanhala-Aniszewski, M. 2006. The role of metadiscourse in Russian and Finnish academic texts.
In Wissenschaftliches Schreiben abseits des englischen Mainstreams/Academic Writing in
Languages Other than English, U. Doleschal and H. Gruber (eds), 127–141. Frankfurt am
Main: Peter Lang.
Vanhala-Aniszewski, M. 2007. Eurokaput. Venäläisen mediatekstin keskustelunomaistuminen.
In Puhe ja kieli. 27:1, 9–23.
Wodak, R. 1996. Disorders of Discourse. London: Longman.
Wodak, R., Cillia de, R., Reisigl, M. and Liebhart, K. 1999. The Discursive Construction of the
National Identity. Edinburgh: University Press.
Zasurskii, Ia.N. 2002. Zhurnalistika i obshchestvo: balansiruia mezhdu gosudarstvom, bizne-
som i obshchestvennoi sferoi. In Sredstva massovoi informatsii postsovetskoi Rossii, Ia.N.
Zasurskii (ed.), 195–231. Moskva: Aspekt Press.
Who communicates in the media
supported by the Russian Church?

Lea Siilin

The Russian Orthodox Church has established new Church media in order to
gain a new audience, e.g. young people. The aim of this study was to determine
what kind of interaction is typical of this media, and how interactional relations
between representatives of the Church and ordinary readers of Church media
are constructed. The material consisted of three texts published in the journal,
Foma (Issue 5, 2005). The analysis of the texts indicates that to some extent they
constitute an institutional discourse because the interaction expressed by the
priest, the bishop and the ordinary readers of the journal “Foma” demonstrates
the difference between authoritative and individual opinions. On the one hand,
the use of colloquial language shows a tendency toward a more casual interaction
between the clergy and the readers. On the other hand, interactional linguistic
features such as modality markers and attitude markers does not occur very
frequently in written interaction, which, however, may refer to a certain distance
between the debaters. Hence the official voice of the Church tends to prevail.

1. Introduction

This paper deals exclusively with media supported and sold by the Russian Orthodox
Church. During the past fifteen years, the position of the Russian Orthodox Church
(ROC) in Russian society has changed completely. Today the ROC is again regarded as
a social institution with which Russian civil society is willing to cooperate. In addition,
in the 1990’s there was a marked upsurge of interest in religion, which has often been
called “a religious Renaissance”. The dynamic role of Orthodoxy in society is also re-
flected in the increase in media supported by the Church.
Apparently, this phase is now coming to an end, and the Russian media have been
critical of the Church’s role in Russian society. People are interested in knowing what
attitude the Church takes toward various current social phenomena, for example, those
connected to the market economy. As a hierarchical institution, the ROC is facing a
new reality: its way of interacting with the people has until now been quite authoritar-
ian and old-fashioned. Traditionally, the ROC has not offered religious instruction
 Lea Siilin

outside churches and services. However, in order to influence contemporary society,

the Church has been forced to seek new strategies. This means establishing new Church
media and gaining a new audience, e.g. young people. It also means that the Church
can no longer preach the eternal truths unilaterally, but has to talk with believers,
members of the Church and unbelievers. Therefore, it is interesting to observe the in-
teraction between the addressers and addressees in media backed by the ROC.
It should be mentioned that there is little information available on the language
and discourse of the media of the Russian Church. This may be surprising but can be
explained as being due to the long history of atheism as the official ideology in the
Soviet Union.

1.1 Main objectives

The aim of this study was to determine (1) what kind of interaction is typical of the
media supported by the Russian Church, and (2) how interactional relations between
representatives of the Church and ordinary readers of Church media are constructed.
I will thus focus on interactional linguistic features such as modality markers, attitude
markers and commentary found in the data.
The main interest here is whose voice is dominant: do the representatives of the
Church express their own individual opinions and in what way, or are they only
mouthpieces for the official statements by the Church? The fact is that Patriarch Alexis
II, head of the ROC until 2008, uses the official voice of the Church together with the
Episcopal Conference. Taking into account this new situation in which the Church
needs to draw closer to its members, the hypothesis used here involves an assumption
that the interaction in the Church media is becoming more informal, more casual. If
this is true, the results of this study may help to explain how the Church is able to oper-
ate in new circumstances. Furthermore, they may lead to a better understanding of
interaction in hierarchical institutions, such as the Church.

1.2 Data

The material used for this study was texts published in the Orthodox Church journal,
Foma (literally ‘Doubting Thomas’). The corpus used for analysis is drawn from three
texts under the heading Orthodoxy light, which were published in Issue 5 (28), 2005
(pp. 14–31). Basically, these texts discuss the place of Orthodoxy in the market econo-
my and secularizing society.
The thematic entity to be analyzed includes different types of texts: (1) a letter
(= A), written by Nikolai, a 19 year-old student of journalism, who was recently bap-
tized, (2) an interview (= B) by the columnist Alla Mitrofanova with the priest Maxim
Kozlov as a comment (= C) to the letter of Nikolai. The title of the interview is Cheap
version of Orthodoxy or something about “the quality of religious services”, (3) a position
Who communicates in the media supported by the Russian Church 

paper (= D) of the bishop of Saratov and Vol’sk Longin under the heading Parishioners
and visitors. I have made a complete analysis of these three texts, which comprise a
total of 14 pages.
Foma is a non-traditional journal established in Moscow in 1997. It is important
to note that the publication committee of the ROC has given this journal its official
approval. The journal welcomes open discussion about doubts concerning Orthodoxy
and faith in God. It is addressed to anyone who wishes to know more about Ortho-
doxy. The journal is richly illustrated and contains about 100 pages per issue. The edi-
tor-in-chief is a professional journalist from Moscow University, Vladimir Legojda.
The letter sent to the editorial staff of the journal serves as a starting point for the
discussion about the so-called light version of Orthodoxy. Obviously, as an answer and
reaction to this letter, the journal decided to interview a priest and to ask for a state-
ment by a bishop. The beginning of the letter is a direct request addressed to the jour-
nal to fulfil its duty to dispel doubts, in this particular case doubts concerning the
questions “What role does money play in the life of the Orthodox Church?” and “Do
some people in the Church have special privileges?”
The texts to be analyzed consist of public, formal written discourse, which in two
cases takes the form of a monologue and in the third a dialogue (see Renkema 2004:
64–66). This means that we are examining the interactional aspect of a written dis-
course. The interview and position paper represent a descriptive, informative and argu-
mentative discourse. The letter, unlike the other texts, also contains a gentle biting irony,
because the student criticizes the Church in a slightly mocking tone. Therefore, the rep-
resentatives of the ROC take a defensive position and try to justify the Church’s policy.
From the standpoint of clerical hierarchy, the participants in the interaction have
unequal role identities. Power relations also occur in this context, because two writers
are representatives of the clergy, one is a lay member of the Church and one is a female
journalist working in the Church media. In texts representing different discourse
genres one can expect differences in the amount and degree of interaction. Here we
use the term speaker as a cover term for an acting, in this case also writing, person.

2. Theoretical background

The theoretical framework of the present study is based on critical discourse analysis
(CDA) which is concerned with social problems and the ideological meanings of lan-
guage. Overall, to my knowledge, this approach has not been used with regard to
Russian Church media. CDA is understood here as it has been developed by the British
researcher Norman Fairclough (e.g. 1989). This approach has adopted the principles of
Halliday’s systemic-functional grammar (Malmkjaer 2002: 103–104).
Before proceeding to study the particular realizations of interaction, it is impor-
tant to specify what is meant by interaction and interpersonal function. Language can be
seen as interaction; in other words, this means the speaker’s intrusion into the speech
 Lea Siilin

event. Consequently, interaction occurs as an exchange of information between two

persons, in which giving implies receiving and demanding implies giving in response
(Halliday 1985: 68). Naturally, the speaker can vary his own communication role, mak-
ing assertions, asking questions, giving orders and advice or expressing doubts.
CDA takes into account three major functional-semantic components of language
proposed by Halliday (e.g. 1976). One of these functions is interpersonal. The inter-
personal component expresses the speaker’s ‘angle’: his attitudes and judgments, his
encoding of the role relationships in the situation, and his motive in saying anything at
all. In the social context the interpersonal meaning corresponds to the tenor of the
discourse, which refers to who is taking part, the characteristics of the participants,
their situation and roles. The interpersonal function is the function “to establish, main-
tain, and specify relations between members of societies” (Halliday 1976: xix, Stillar
1998: 54).
As Fairclough (1993: 134) argues, “language use is always simultaneously consti-
tutive of social identities, social relations and system of knowledge and beliefs.” This
means that the choice of linguistic form in media discourse depends on these factors.
The content and texture are inseparable, and the connecting factor is meaning. There-
fore, analysis of linguistic text naturally focuses on meanings relevant for analyzing
interactional units (see Stillar 1998: 40–45).
Here I want to look in greater detail at what such units may be. I will approach this
question using Halliday’s classification (1976: 190, 1978: 132) of the interpersonal
function arranged by rank; this is useful because it allows us to relate linguistic units of
clause structure with realization of the interpersonal function. Within a clause, inter-
personal function is expressed firstly by mood, i.e. different types of speech function:
offer, command, statement and question, and modality. Secondly, within a verbal
group it is expressed by person. Thirdly, within a nominal group interpersonal func-
tion is expressed by the person’s attitude and role expressed using attitudinal modifiers
and intensifiers. Furthermore, within an adverbial and prepositional group it is ex-
pressed by comment, which means classes of comment adjuncts. Finally, it can be real-
ized by expressive words (see Palmer 1988).
Modality is the speaker’s assessment of probability and predictability (Halliday
1976: 211). On the one hand, it concerns the speaker’s attitude, in this case, towards his
own speech role as ‘declarer’; on the other hand, it is the attitude to the content being
expressed. As Halliday (1978: 46) points out, “the modal element expresses the par-
ticular role that the speaker has chosen to adopt in the situation and the role or role
option that he has chosen to assign to the hearer“.
According to Halliday’s classification (1976: 190, 212), interpersonal function is
expressed: (1) within a verbal group by person and polarity (positive/negative),
(2) within a nominal group by the person’s attitude and role expressed using attitudi-
nal modifiers and intensifiers, (3) within an adverbial and prepositional group by a
comment, which means classes of comment adjuncts; and finally, it can be expressed
(4) within words, e.g. expressive words (connotations to particular lexical items).
Who communicates in the media supported by the Russian Church 

In the modal system we deal with notions of possibility and necessity. As a matter
of fact, modality is a form of participation by the speaker in the speech event (Halliday
1976: 197). Modality has two different dimensions: on the one hand, ‘probable’, on the
other, ‘possible/certain’. Modality can be expressed by either or both of two elements,
one verbal (that is, verbal auxiliaries in finite form) and the other non-verbal forms
such as lexical items (possible), adverbs (perhaps), adjectives (it is possible that – imper-
sonal; I am sure, certain that – interpersonal), and nouns (there is a presumption that).
There are two types of modality: the first is an indicative, epistemic type, that in-
volves probability and usuality; it involves probability (may be) and usuality (sometimes).
Epistemic modality, being interpersonal, is concerned with knowledge, belief, truth,
and opinion in relation to proposition. The second type is an imperative, a deontic one,
that involves obligation (is expected to) and inclination (wants to) (see Halliday 1985:
86, 334–338, Palmer 1988). Deontic modality contains an element of will; it is con-
cerned with an action initiated by others expressing necessity, obligation or possibility
(obligation, permission, directives).

3. Interactional elements

I will start the presentation of results by describing the data of all three texts (see 1.2).
Indicative mood and assertive sentences are typical of all texts. First of all, the speaker
considers his claim to be true. A very strong degree of certainty is expressed in clauses
like: Vykhod zdes’ tol’ko odin: neobkhodimo prosveshchat’ liudei (D) ‘There is only one
solution: we must educate people’. The priest and the bishop use the most categorical
A high degree of certainty can also be expressed by introducing facts as self-evi-
dent or received knowledge: Izvestno, chto v tserkov’ nado khodit’. (B) ‘It is well known
that one should go to church’. There are also modal adverbs, which emphasize quite a
strong degree of certainty. In most cases this is expressed by such modal adverbs, called
emphatics, as konechno ‘certainly’, deistvitel’no ‘really, indeed’, poniatno ‘of course,
naturally’, kak pravilo ‘as a rule’. Both the priest and the bishop use them:
(1) a. Poniatno, chto eto byl ne vykhod (B)
‘Certainly, this was not a solution’
b. Kak pravilo, imenno takie liudi govoriat pro sebia (D)
‘As a rule, those people in particular talk to themselves’
c. Mozhno poniat’, kogda chelovek nakhoditsia v kakikh-to osobykh otnosh-
eniiakh s udalennym monastyrem (B)
‘One could understand it, if the person had an exclusive relationship to a
distant monastery’
 Lea Siilin

From the interactional point of view such “certain” statements, which are often com-
bined with deontic modality, are not of interest. Hence I will focus on the means of
toning down certainty, such as the use of modality markers and attitude markers. To
be honest, there are different means but not all of them occur frequently in the data.

3.1 Modality markers

The degree of certainty of a proposition can be modified by reservations related to

epistemic modality. Reservations belong to hedges, which are expressed by modal
auxiliaries. In Russian the modal auxiliary related to the epistemic modality is the verb
moch’ ‘to be able, can’. It is also possible to use an adverb vozmozhno ‘it is possible’ as
predicative, but it is not found in the data. However, the short form of the adjective
vozmozhnyi is used once in the question: Vozmozhen li takoi kompromiss? (C) ‘Could
such a compromise be possible?’ Overall, the modal auxiliary is not used very often:
(2) Ne dumaiu – chto o Rossii mozhno budet govorit’ kak o pravoslavnoi strane vo
glave s pravoslavnym chelovekom, dlia kotorogo ego vera opredeliala by obraz
ego bytiia (B)
‘I do not think that in the future it might be possible to talk about Russia as
an Orthodox country under the leadership of an Orthodox person for whom
faith would determine his way of being’
Modal auxiliary moch’ is also found in questions which come from the interview: Chto
eto mozhet byt’ za rasskaz? (C) Naprimer, eto mozhet byt’ rasskaz o SMI (B) ‘What
might such a story be? For example, it may be a story about the media’. Interrogative
sentences are related to epistemic modality because in questions the notion of possibil-
ity is present: Kak izbezhat’ soblazna? (C) ‘How is it possible to escape distortion?’. The
lesser extent of certainty is represented in rhetorical questions with an interrogative
particle li: Mozhem li my skazat’, chto nash narod uvidel eto i vrazumilsia? Net (B)
‘Could we possibly say that our nation saw this and realized it? No’.
Hedges are also expressed by mood, in this case by subjunctive mood (Palmer
1988: 21–23). There are some examples of the subjunctive mood, which is basically
used in two meanings: firstly, it refers to the fact that the proposition is less probable;
and secondly, it is used as a polite set expression:
(3) a. Khotelos’ by, konechno, chtoby nevotserkovlennyi chelovek sostavlial sebe
vpechatlenie o Pravoslavii ne po iarmarke (B)
‘Of course, I would not like to see a convert form his opinion of Ortho-
doxy based on the market place’
b. Kazalos’ by, student universiteta, ne p’iu, ne kuriu, ne kolius’ – i tut vdrug
potentsial’nyi zlodei (A)
‘One would think, I am a university student, I do not drink, smoke or use
narcotics, and suddenly I am a potential villain’
Who communicates in the media supported by the Russian Church 

c. ia by ne soglasilsia (D)
‘I would not agree’
Furthermore, hedges are expressed by impersonal verbs. In Russian, uncertainty can
be expressed by the impersonal verb kazat’sia ‘to seem, appear’ which is typical of both
the letter and the interview, but is not used in the position paper: Vot kazhetsia, i est’
khriastianskaia liubov’ (A) ‘Apparently, Christian love does exist’, Mne kazhetsia,
ochen’ vazhno (B) ‘It seems to me that it is very important’.
The most commonly used type of reservations are different kinds of modal ad-
juncts, i.e. adverbs like navernoe, naverno, veroiatno ‘probably’, mozhet byt’ ‘maybe,
perhaps’, budto by ‘apparently’. Modal auxiliary or the subjunctive mood can occur in
the same sentence with adverbs, as in the following Examples (b, c):
(4) a. Mozhno, navernoe, uvidet’ v Garri Pottere chto-to, chto ne vovse
otvratitel’no (B)
‘Perhaps it might be possible to see in Harry Potter something which is
not totally disgusting’
b. Ved’ eto seichas izvestnyi epizod, i mnogie, mozhet byt’, posledovali by ee
primeru (B)
‘Actually, now this is a known episode and many people would perhaps
like to follow her example’
c. Veroiatno, eto vazhno dlia togo, chtoby kazhdyi priobrel svoi lichnyi opyt
very (B)
‘It is probably important because everyone should have his own experi-
ence of faith’
d. Prichem stoiu budto by pered sledovatelem (A)
‘Moreover, I am apparently standing before an investigator’

3.2 Restrictions

There are many specifying adverbs, which are used as restrictions to indicate the range
over which the speaker’s claim is true. The restriction may concern a quantity such as
frequency and time. In this sense the Russian adverbs vsegda ‘always’, chasto ‘often’,
poroi, inogda ‘at times, sometimes’ are used in the following examples:
(5) a. Dlia missionera vsegda est’ soblazn vziat’ vyzhimki iz khristianstva (B)
‘For a missionary there is always a temptation to take advantage of Chris-
b. poroi dostatochno budet eto nastoiashchee prosto pokazat’ (B)
‘Sometimes it will be enough to show the genuine thing’
 Lea Siilin

These restrictive expressions are often combined with other adverbs of quality and
degree such as dovol’no ‘quite, fairly’, pochti, chut’ li ‘almost’, ne vpolne ‘not entirely’
khot’ ‘at least’. They give the impression of an approximate certainty:
(6) a. Takoi vybor vstaet dovol’no chasto (B)
‘Such choice comes up quite often’
b. Tot, kto khot’ nemnogo znakom s Evangeliem znaet (B)
‘And he who is at least somewhat familiar with the Gospel knows’
c. “Titanik” chut’ li ne posobie dlia nachinaiushchego khristianina (D)
‘“Titanic” has been interpreted almost as an instruction to a new Christian’
There are different examples of restrictions concerning quality. They occurred in such
expressions as v bolshoi stepeni ‘to a great extent’, zachastuiu ‘partly’, kak-to ‘some-
how’, nekii ‘some’:
(7) a. Eto v bol’shei stepeni pravda po otnosheniiu k religioznomu opytu (B)
‘This is to a great extent the truth with regard to religious experience’
b. Ia soglasen, chto sam narod pravoslavnyi zachastuiu vinoi etomu “ba-
zaru” (A)
‘I agree that the Orthodox themselves are partly guilty of this “bazaar”’
c. Nuzhno kak-to razdeliat’ v nashei zhizni bytovoe i sviashchennoe (B)
‘We somehow need the separate profane and sacred in our life’
Restrictions also include expressions of the speaker’s own opinion (a, b) or the manner
of expression (b, c), as in the following examples:
(8) a. na moi vzgliad – “eto vera mira sego” (D)
‘To my mind this is “a faith of this world”’
b. Po-moemu, tot fakt – eto v znachitel’noi mere sledstvie pobedy potrebitel’skoi
tsivilizatsii (B)
‘In my opinion the fact is – to a great extent – a result of the victory of
culture of consumption’
c. Chelovek, miagko govoria, sidit na dvukh stul’iakh (B)
‘This person, to put it mildly, is sitting on the fence’
Expressions like za iskliucheniem ‘with the exception of ’ and v printsipe ‘in principle’
can also be interpreted as restrictions: Takoi utilitarnyi podkhod v printsipe svoistven
sovremennomu cheloveku: ko vsei zhizni (D) ‘Such a utilitarian approach to the whole
of life is in principle typical of a modern human being’.
One type of restriction is the attributor, which represents a passing or casual refer-
ence. In this case we deal with an intentional mention of the Bible, either directly or by
implication. Allusion is used by both the priest and the bishop, especially by the priest.
For example: Propoved’ Gospod’ nachinaet svoiu propoved’ slovami: Pokaites’, ibo
Who communicates in the media supported by the Russian Church 

priblizilos’ Tsarstvie Nebesnoe (B) ‘The Lord begins his preaching with the words “Re-
pent for the kingdom of Heaven is upon you”’. This quotation from the Gospel of
Matthew (3:2) is printed in italics to let the readers know that it is a quote. In spite of
the Bible, the priest’s references are, for example, to the Russian philosophers Alexei
Losev, St. John Chrysostom and Metropolitan Anthony. In addition, the priest offers a
verse from a poem by the Russian philosopher Vladimir Soloviev (1858–1900) as a
text common to all readers. He starts this reference with the interactional expression:
Pomnite: Milyi drug (B) ‘Remember: Dear friend’. The bishop gives only one reference
to the Bible: V Novom Zavete skazano, chto – (D) ‘In the New Testament it is said that’
man should give water, feed the sick and poor, and visit them in prison. And on the
basis of these acts all will know that we are Christ’s disciples. This allusion refers both
to the Gospel according to Matthew (25: 36–37), and to the Gospel of John (13:35): “If
there is love among you, then everyone will know that you are my disciples.”
Finally, quotation marks may indicate a restriction in the meaning of a word.
Quotation marks are put around such words as deistvitel’no sviataia (B), ‘really sacred’,
pozhertvovanie (A) ‘donation’, vera mira vsego (D) ‘the faith of this world’, zakhozhane
(D) ‘visitors’, pravoslavie-lait (D) ‘Orthodoxy light’. This can also be associated with the
interactional relationship with the audience because they indicate the speaker’s atti-
tude toward these words.

3.3 Attitude markers

3.3.1 Reference to person

Attitude markers make a more direct reference to interaction. Here we examine
whether personal elements such as personal pronouns and finite verb forms with per-
sonal endings are highlighted or minimized in the data. It is interesting to note that,
with the exception of the journalist’s part of the interview, every text has examples us-
ing the 1st and 2nd person singular and plural personal pronouns ia ‘I’, my ‘we’, ty, vy
‘you’. However, there are differences between texts: the student quite often prefers to
highlight himself using the 1st person singular pronoun, but there are also many cases
where the personal pronoun is omitted. This is easy to explain because in his letter he
either describes what he saw at the Easter market or gives his own opinion of it: ia ni v
koem sluchae ne stremlius’ rugat’ Tserkov’ (A) ‘I do not try to blame the Church in any
circumstances’, Ia ne protiv (A) ‘I am not against’.
The bishop and the priest also sometimes use the 1st person singular pronoun and
thus emphasize their own opinions:
(9) a. Ia uveren, chto dlia vsekh chlenov Tserkvi dolzhno naitis’ kakoe-to delo (D)
‘I am confident that every member of the Church must find some activity’
b. ia podkhozhu k etoi probleme bez osobogo optimizma (B)
‘I am approaching this problem without particular optimism’
 Lea Siilin

c. No ia byl by kuda bolee rad, esli by etot tekst prozvuchal ne tam (B)
‘But I would be even happier if this text would not be heard in that place’
In the last examples we can see that the use of the pronoun is combined with the sub-
junctive mood and hedging. In the interview the 1st person pronoun is the most fre-
quent one, despite this, it does not always refer to the speaker. Both the priest and the
bishop use it to quote other people’s speech as a concrete example, which makes the
text more interactional. It gives readers the opportunity to identify themselves with the
person in the example. In such direct speech, 1st and also 2nd person singular per-
sonal pronouns with generic reference are used:
(10) a. takie liudi govoriat pro sebia: “Da, ia khozhu v tserkov’, reguliarno, pochti
kazhdyi den’. Vot kogda na rabotu edu ili s raboty – obiazatel’no zakhozhu
svechki postavit’ (D)
‘Those people used to talk to themselves: All right, I regularly, almost ev-
ery day, go to church. And going to work or returning home, without fail
I visit the church in order to light candles’
b. Kak znaki na doroge: tuda poedesh’ – golovu slomish’, tuda poedesh’ – tu-
pik, po etoi doroge mozhesh’ zabludit’sia (B)
‘They are like traffic signs on the road: if you go there, you will risk your
neck; if you go there, you will come to a dead end; on this road you can
get lost’
c. Mozhno tol’ko pomoch’ stat’ na nee i skazat’: von vidish’, skol’kie uzhe po
nei idut, – Idi po nei – i uznaesh’, chto takoe Pravoslavie. Smotria, konech-
no, kak poidesh’ (B)
‘You can only help to find the road and say: have a look, you see how
many already go along the road – Go and you will know what Orthodoxy
is all about. Of course, depending how you go’
Both the priest and the bishop use the 1st person plural pronoun in the collective and
interactional sense, referring both to themselves and to the readers:
(11) a. Inogda my sozdaem svoimi rukami vse to, chto est’ i v miru, no tol’ko s
pravoslavnoi okraskoi (D)
‘Sometimes we create with our own hands everything that is in the world
but with an Orthodox colouring’
b. My zabyvaem, chto Tserkov’ – eto vse my, veruiushchie liudi, kazhdyi che-
lovek (D)
‘We forget that the Church consists of all of us believers, of every person’
The personal pronoun we may have different references. Sometimes it is not easy to
differentiate between either the pronoun we, referring to Russians as a whole nation,
or to believers, members of the Church, as in the example: Otets Maksim, vot my
Who communicates in the media supported by the Russian Church 

govorim (C) ‘Father Maxim, we used to say’. Consequently, the corresponding posses-
sive pronoun our is also used in the same meaning: na vosstanovlenie nashikh sviatyn’,
chto otnialo u nas mnogo sil (D) ‘renewal of our holy places which took a lot of effort
from us’, tozhe nashi prikhozhane (D) ‘also our parishioners’. The pronoun we may also
refer to certain persons, as in this example: my ochen’ blizki s ottsom Andreem (D) ‘We
are very close to Father Andrei’.
In addition, the imperative mood and 2nd person plural pronoun are used in the
interactional function: Soglasites’, seichas uzhe trudno naiti takogo cheloveka (B) ‘Agree
with me that it is now difficult to find such a person’. In the next example the pronoun
you refers to only one person as a polite form of address:
(12) a. Postaraites’ dat’ im uvidet’ opyt podlinnoi tserkovnoi zhizni – (tol’ko tako-
go, chtoby Vy byli uvereny, chto eto monastyr’) (B)
‘Try to give them the opportunity to see real church life – but only that
which you certainly know to be a real monastery’
b. Pozhaluista, esli Vy zaniatoi delovoi chelovek – pishite pis’ma zakliuchen-
nym, otvechaite na ikh pis’ma. Ili posylaite im posylki, ili pomogaite de-
tiam v detskom dome, ili svoim sosediam. Ishchite podskazku sredi vsekh
form obshchestvennoi raboty (D)
‘If you are a busy businessman, please write letters to prisoners, answer
their letters or send them parcels or help children in children’s homes or
help your neighbours. Find a hint of what to do among all the forms of
social work’
The 2nd person singular in the imperative mood is occasionally used in the generic
meaning: Krestilsia – i do otpevaniia zhivi spokoino (B) ‘You have been baptized, so live
in peace until the funeral’, Zhivi spokoino, zhdi spaseniia dushi (C) ‘Live in peace, look
forward to salvation’.
The 2nd person plural pronoun is used regularly in the journalist’s questions as a
polite form of address. In this case the pronoun Vy ‘you’ is written with a capital letter.
In the Orthodox Church the position is to call a priest using the word Father and his
first name: Otets Maksim, v odnom iz svoikh vystuplenii Vy upotrebili (C) ‘Father
Maxim, in one of your public appearances You used’, Kak Vy otnosites’ k etomu iav-
leniiu? (C) ‘What attitude do You take to this phenomenon?’, kak Vy dumaete, chto i
kakim obrazom nuzhno rasskazyvat’ o Pravoslavii (C) ‘What do You think, what and
in which way should Orthodoxy be told about?’. The verbal form of the 2nd person
plural in the indicative mood is not used often: mozhete smelo nazyvat’ sebia pa­
lomnikom (A) ‘You can bravely call yourselves pilgrims’, Kak skazat’, chto, znaete, re-
biata, luchshe vam seichas ne zdes’ byt’ (B) ‘How can you say: you know, boys, it would
be better for you not to be here now’.
In Russian there is a special use of the verb form of the 1st person plural in the
imperative mood function. This form has a strong collective indication: let us do
something. Three times the priest repeats Let us remember after rhetorical questions to
 Lea Siilin

remind readers of what happened in 1988 and thereafter in the life of the Church:
Vspomnim eshche, chto v te zhe gody k nam khlynuli sektanty (B) ‘Let us remember
that in these years we had an invasion of members of different sects’.
Rhetorical questions are a clear interactional element used in most cases by the
priest, but incidentally also by the student and the bishop. Sometimes the answer is
also given in the text. But even if the answer appears, such questions allow readers to
think and formulate their own opinions. As a matter of fact, the priest uses these ques-
tions in a manipulative way to persuade readers and to justify the Church’s actions,
which were criticized in a public discussion. Here are some examples:
(13) a. Dumaete – grekh na dushe otkrylsia? Net (A)
‘Do You think that a sin [in my] soul was uncovered? No’
b. Mozhem li my skazat’, chto nash narod uvidel eto i vrazumilsia? Net (B)
‘Could we possibly say that our nation has seen this and realized it? No’
c. Mogli li my kak-to vospol’zovat’sia situatsiei 1988-go i posleduiushchikh let
luchshe, chem eto proizoshlo? Mozhet byt’, da (B)
‘Could it have been possible for us to take more advantage of the situa-
tion in 1988 and the following years than we did? Probably, it could’
d. Obrashchaetsia li k vere kto-to posle takikh meropriiatii? Bog vest’ (B)
‘Can someone find faith in God after such events? God knows’
Very affective exclamations with the particle zhe may have the meaning of a question
forcing readers to think about this issue: Ne otdavat’ zhe bylo im nashikh krestia­
shchikhsia! (B) ‘We could not leave to the sectarians our people who were willing to be
baptized, could we!’
I have the impression that also the particle ved’ ‘you see, you know, is it not?’ may
have an interactional meaning: by emphasizing the main statement of the sentence as
shared knowledge, it looks for verification from the reader. The priest quite often uses
this particle in wishing that the reader would agree with him on the matter under con-
sideration: Ved’ v liubom pravoslavnom khrame evkharistiia sovershaetsia odinakovo
(B) ‘You know, don’t you, that in every Orthodox Church the Eucharist is performed
in the same way’, No ved’ eto ne tsel’ (B) ‘You know, don’t you, that this is not the goal’,
Ved’ kogda Gospod’ poslal apostolov na propoved’ (B) ‘You know, don’t you, that when
the Lord sent apostles to preach’. I think there is something similar in the colloquial
expression vot chto ‘listen to me’: Vot chto vazhno dlia spaseniia (B) ‘Listen to me; this
is of great importance to salvation’. One would think that these interactional means are
related to the manipulative tone in the priest’s discourse.
The reference to the 3rd person singular and the plural is in most cases made using
the words chelovek, ‘man, person’, liudi ‘people’. In addition, other more specific words
such as khristianin ‘Christian’, veruiushchii ‘believer’ in singular and plural are used:
Esli kazhdyi khristianin budet ispolniat’ svoi dolg, Tserkov’ ispolnit svoiu missiiu v obsh-
chestve (D) ‘If every Christian did his duty, the Church would also accomplish its
Who communicates in the media supported by the Russian Church 

mission in society’. Reference to the 3rd person occurring in the imperative mood is
possible by using the particle pust’: Poetomu esli chelovek khochet uznat’ chto-to o Pra-
voslavii, pust’ reshitsia pereshagnut’ porog imenno khrama (B) ‘So if a person wants to
know something about Orthodoxy, let him be brave enough to cross the threshold of
the church’. The Orthodox Church is regarded both as an abstract organization and as
an active community of believers: V Tserkvi tak mnogo nastoiashchego (B) ‘In the
Church there are so many genuine things’.
In summary, the highlighting of oneself is typical of the student’s letter, which can
be seen in the use of the 1st person singular pronoun. The priest and the bishop also
use this pronoun, but they use it in examples to make them more interactional. The
same is also true concerning the 2nd person singular pronoun ty ‘you’ in the interview
of the priest by the columnist Alla Mitrofanova. Unlike the student, the priest and the
bishop also use the 1st person plural my ‘we’ as a solidarity marker. Interaction with
the audience can be found in the use of the 1st and 2nd person plural pronouns and in
some expressions in the imperative mood. They usually generalize the statement. In
my opinion an effective means of interaction is the use of rhetorical questions.
The question of highlighting or minimizing the speaker is related to the use of ac-
tive and passive voice and impersonal sentences. As a whole there are not very many
passive clauses in either the present or the past tense. For example: Vedutsia spory i o
tom– (D) ‘There are also debates about–‘, Tak slozhilos’ (D) ‘It turned out that’, Pravo-
slavie teper’ propoveduetsia otkryto (C) ‘Now Orthodoxy is openly preached’, U liudei
formiruetsia bolee pozitivnoe vospriiatie otdel’nykh predstavitelei tserkvi (B) ‘A more
positive image of the representatives of the Church is formed by people’, I trebuetsia
muzhestvo, chtoby skazat’ (B) ‘One needs to have the courage to say that’. It seems a bit
strange that there were only few examples of passive participles: I chto sdelano ne tak?
(C) ‘And what has not been done in an appropriate way?’
There are also some examples of impersonal clauses with the infinitive or with the
omission of the 3rd person plural. Most examples come from the interview:
(14) a. im prosto nechego delat’ (D)
‘There is simply nothing for them to do’
b. Tak legko sebia obmanut’ (B)
‘It is so easy to cheat yourself ’
c. Trudno emu chem-libo sovetovat’ (B)
‘It is difficult to give any advice to him’
d. tut mne pokazyvaiut, chto On prezhde vsego v kazhdom iz nas – somne-
vaetsia (A)
‘Here they show me that first of all He (God) suspects each of us’
e. Dukhovnye sovety na rynke ne daiut (B)
‘It is not appropriate to give spiritual advice in the market place’
 Lea Siilin

f. V khram prikhodiat dlia togo, chtoby perezhit’ vstrechu s Bogom (B)

‘They come to the church in order to live through a meeting with God’

3.3.2 Comment adjuncts

The speaker’s attitudes are expressed by comment adjuncts like the adverbs konechno
‘of course’, deistvitel’no ‘of course, indeed’, or the short form of adjectives like khorosho
‘good’, zamechatel’no ‘remarkable’, or set phrases like slava Bogu ‘thank God’, which are
used only by the priest and the bishop and only in single cases.
(15) a. Konechno, tsifra eta somnitel’naia (D)
‘Of course, this figure is questionable’
b. chto otnialo u nas mnogo sil, sredstv i energii, kotoruiu Tserkov,’ konechno
zhe, dolzhna raskhodovat’ na bolee vazhnye tseli (D)
‘It took a lot of effort, facilities and energy from us, which the Church, of
course, ought, after all, to spend on more important purposes’
c. Deistvitel’no, est’ khoroshie veshchi v okruzhaiushchem nas mire (D)
‘Indeed, there are good things in the world around us’
d. Slava Bogu, chto sredi tekh, kto eiu zanimaetsia (B)
‘Thank God that among them who work on it’
e. Khorosho, uzhe i tserkov’ vosstanovili – vse eto zamechatel’no (D)
‘Great, churches have already been renewed – all this is remarkable’
Furthermore, a long form or short form of an adjective such as interesnyi ‘interesting’
or nelegkii ‘difficult’, especially with the verb byvat’ ‘tend to be’, can be used as a com-
ment: Dal’she proiskhodit eshche bolee interesnoe (D) ‘Furthermore, something even
more interesting’ will happen, Pri takom razbrose mnenii naiti istinu byvaet nelegko
(D) ‘Having such different opinions, it tends not to be easy to find the truth’.
The most clearly expressed attitude is dissatisfaction. The priest and bishop are not
happy with contemporary tendencies in society and therefore use the expression k
sozhaleniiu ‘unfortunately’: Potomu-chto “Pravoslavie-lait”, k sozhaleniiu, stanovitsia
vse bolee rasprostranennym (D) ‘For some reason Orthodoxy light is, unfortunately,
becoming more and more common’. One special case of comments may be the use of
an exclamation mark, which can be replaced by the adverb unfortunately in the fol-
lowing sentence: Prishel (po svoei vole!) na Paskhal’nuiu iarmarku (A) ‘I came (of my
own volition!) to the Easter market’. Using affective verbs like porazhat’ ‘to strike’, one
can also make a comment, as in the following example: No porazhaet, s kakim entuzi-
azmom Tserkov’ (A) ‘However, it strikes [me] how enthusiastic the Church is’.
In fact, punctuation marks such as an exclamation mark and ellipsis can also be
seen as interactional elements. Ellipsis is sometimes used to draw the reader’s attention
to exceptional things: detei krestiat – chtoby noch’iu ne plakali, soboruiutsia – chtoby
spina ne bolela... (D) ‘they let their children be baptized so they would not cry, they
take part in prayer services for the healthy, not wishing to have backache...’. In some
Who communicates in the media supported by the Russian Church 

cases this punctuation mark is used to make the reader think of more similar exam-
ples: veruiushchie pravoslavnye khristiane (mat’, otets, brat, drug...) (B) ‘believing
Orthodox Christians (mother, father, brother, friend...)’.
The exclamation mark is used by the student and the priest in statements which
surprise the speaker and hopefully also the reader: On prezhde vsego v kazhdom iz nas
– somnevaetsia! (A) ‘First of all He (God) suspects each of us!’, No smotriu na matush-
ku – v rukakh u nee vmesto Evangeliia – kal’kuliator! (A) ‘I see that the nun has a cal-
culator rather than a gospel book in her hands!’ Such examples have a slightly ironic
tone. Irony is typical of the student’s letter, where he expresses his own doubts about
the Church: No pochemu mne nachinaet kazat’sia, chto organizatsiia, vziavshaiasia
privesti menia k Bogu, opredeliaet moi uroven’ very proportsional’no moei platezhe­
sposobnosti? (A) ‘But still, why do I start to think that the organization that undertook
to lead me to God determines the level of my faith proportionally to my solvency?’,
Chustvuiu sebia gastarbaiterom bez registratsii (A) ‘I feel I am a Gastarbeiter without
any registration’.
In the interview the priest answers those questions which the letter has raised, but
partly also in an ironic tone: Neponiatno: neuzheli tam pominovanie kachestvennee
budet, chem v moskovskikh monastyriakh i khramah? (B) ‘It is incomprehensible, is it
really true that remembrance in prayer is of higher quality in the market place than
in Moscow monasteries and churches?’ Neuzheli tam evkharistiia drugaia sovershaet-
sia, de liuks, vysshei kategorii? (B) ‘Is it really so that they perform another Eucharist,
de luxe, of the highest quality in the market place?’

4. Concluding discussion

Pointing out some characteristic features of each speaker may summarize the results of
the present study. In the letter, personal elements are highlighted by the explicit use of
the 1st person pronoun while certainty of statements is often toned down by using the
impersonal verb kazhetsia (‘to seem, appear’). The approach is clearly individual. The
student is a semi-professional writer, and this can be seen in his repeating adversative
sentences: I think, but they show me. Interactional elements involve a rhetorical ques-
tion and the emphasized use of punctuation marks. The ironic tone of the letter is as-
sociated with doubt and hesitation. The student is afraid of starting to worship com-
mercial stands instead of God.
The journalist has totally effaced herself. She uses a polite conventional address in
the questions, which in most cases are inquires about the addressee’s opinion: what do
you think? Of course, the text has been edited; but it still has the style of spontaneous
In the interview it is interesting to see how the priest, on the one hand, expresses
his own opinions and, on the other hand, speaks as a representative of the Church. It
is typical of this interview that there are both categorical statements about what is
 Lea Siilin

right, what is wrong, and toning down the certainty of statements with hedges. By us-
ing the 1st person plural pronoun “we” the priest indicates collective solidarity. From
an interactional point of view, he is quite active. The priest reacts to the questions made
by the journalist. He also makes direct references in a slightly critical tone to the issues
raised in the letter. In order to make the readers actively think about the issues dis-
cussed in the interview, he often uses concrete examples and rhetorical questions. In
these examples he uses colloquial expressions that indicate a less formal style. While
giving advice and making a clear distinction between appropriate and inappropriate
things, he acts as a priest, which can be seen in the last sentence of the interview: It is
not possible to be a too deeply believing person – it is possible to believe too superficially
– too deeply simply does not exist. Obviously, the priest intends to convince or persuade
readers by using allusions.
The position of the bishop is not directly connected to the letter. The bishop dis-
cusses the theme Light version of Orthodoxy on a more general level. He is clearly the
shepherd using the 1st person plural pronoun. The style used in the position is formal,
although he also uses some concrete examples. By means of this text he takes part in
the public discussion about the Church’s right to organize the teaching of religion at
public schools. This question is a burning issue in Russian society today; therefore the
bishop refers to the enemies of the Church and tells the readers that “we must be ready
to fight”. Both the priest and the bishop make comments and use modal auxiliaries
expressing necessity like must, ought to. Overall, the bishop prefers to give directives
reminding his readers that real Christians will be known by the fruits of their labour.
In general, it appears that these results indicate that the analyzed texts to some
extent constitute an institutional discourse (see Cameron 2001: 132–133). Assertive
statements and an imperative, deontic type of modality (obligation) demonstrate the
social status of the priest and the bishop. They are experts and the student is an inex-
perienced novice. Therefore, in the interaction we can see the difference between au-
thoritative and individual opinions. In spite of all this, the findings of this study lead to
the conclusion that in new Church media the voice of lay members of the Church will
also be heard more often.
I originally assumed that the general shift away from formal and hierarchical rela-
tions might also be reflected in the church media. The results show a tendency toward
a more casual interaction because the letter and the interview contain some examples
of colloquial language. The data include many kinds of interactional units, but they
were not as frequent as might be expected on the basis of previous research (see e.g.
Luukka 1992). Hence, this study does not allow us to conclude that there are some
specific interactional strategies typical of Church media.
Who communicates in the media supported by the Russian Church 


Cameron, Deborah 2001. Working with Spoken Discourse. London/Thousand Oaks/New Delhi:
Fairclough, Norman 1989. Language and Power. London: Longman.
Fairclough, Norman 1993. “Critical Discourse analysis and the marketization of public dis-
course: the universities’”. Discourse & Society 4 (2): 133–168.
Foma. Pravoslavnyi zhurnal. Moscow, 1997–. Available from Internet:
Halliday, M. A. K. 1976. Halliday: System and Function in Language: Selected Papers. Edited by
G. R. Kress. London: Oxford University Press.
Halliday, M. A. K 1978. Language as Social Semiotic: The Social Interpretation of Language and
Meaning. London: Arnold.
Halliday, M. A. K 1985. An Introduction to Functional Grammar. London: Arnold.
Luukka, Minna-Riitta 1992. Akateemista metadiskurssia. Tieteellisten tekstien tekstuaalisia, in-
terpersonaalisia ja kontekstuaalisia piirteitä. [Korkeakoulujen kielikeskuksen julkaisuja
N:o 46]. Jyväskylä: Jyväskylän yliopisto.
Malmkjaer, Kirsten (ed.) 2002. The Linguistics Encyclopedia. 2nd edition. London/New York:
Palmer, F. R. 1988. Mood and Modality. [Textbooks in Linguistics]. First published 1986.
Cambrigde: Cambridge University Press.
Renkema, Jan 2004. Introduction to Discourse Studies. Amsterdam: John Benjamins Publishing
Stillar, Glenn F. 1998. Analyzing Everyday Texts: Discourse, Rhetoric and Social Perspectives.
Thousand Oaks: Sage.
“O England! England! She says –
my Father – my Sisters – my friends!
– shall I ever see you more?”
Reporting in 18th-century correspondence

Minna Nevala and Minna Palander-Collin*

This study focuses on reporting in the letters of eighteenth-century writer

Frances (Fanny) Burney. Our purpose is to study the occurrence of reporting
from a socio-pragmatic perspective, with the aim of understanding the function
of reporting in all its communicative situations as well as in personal and
interpersonal functions. The results from the analysis show that reporting is
more typical in Burney’s letters to close recipients, and that the saying of those
closest are much more often reported than general sources like newspapers. Her
reporting is often accompanied with evaluative remarks that further highlight
the interactive nature of letters and emphasise the use of reporting to show
personal stance. Direct reporting also seems to appear in emotionally laden
contexts in particular.

1. Introduction

Much of what we say or write repeats what we have said in another context, heard from
other people or read somewhere. Such reporting is a universal phenomenon to the
extent that languages employ linguistic means to signal overtly that we are passing on
information that is borrowed or quoted from other texts (e.g. Coulmas ed. 1986;
Collins 2001: 1). There are, however, differences in “what is quoted when, how and
why” in different discourse contexts (Fairclough 1992: 118–119). For instance, both
scientific papers and everyday conversations report other people’s ideas, thoughts and
sayings, but they rely on different techniques to do so, and the whole purpose of the
reporting is also different.

* The authors gratefully acknowledge the financial support received during the writing of this
article from the Research Unit for Variation, Contacts and Change in English and the Academy
of Finland (Nevala) and the Department of English, University of Helsinki (Palander-Collin).
 Minna Nevala and Minna Palander-Collin

Our study focuses on reporting in a sample of eighteenth-century letters written

by Frances (Fanny) Burney to various recipients. Our purpose is to study the occur-
rence of reporting from a socio-pragmatic perspective with the aim of understanding
the function of reporting in the communicative situations in which it occurs. Our
analysis focuses on the role of the reporter, the reporting situation, the subject matter
of the report, the identity of the person whose speech is reported and the form of the
reporting frame. In this way, we hope to establish some of the purposes that reporting
serves in eighteenth-century communication.
Our previous research on reporting in Hester Piozzi’s eighteenth-century letters
showed that, in these letters, the occurrence of reporting was related to the nature of
the writer-addressee relationship: reporting was generally more frequent in letters to
close and intimate female recipients than in those to more distant or male recipients
(Palander-Collin & Nevala 2006: 135). Direct reporting in particular was almost re-
stricted to letters that were written to close recipients, and the topics reported using
direct quotes were often laden with emotion. The people whose words were most fre-
quently reported were also intimate family members, and Piozzi reported more on
what men had said. In Fanny Burney’s letters, some of these tendencies are repeated;
in particular, closeness between Fanny and her correspondents seems to encourage
reporting in her letters, and she reports most often on her close family members. In
Section 3.2 of this paper we shall, also, propose a model for dealing with the functions
of reporting in letters. The results of the application of the model given in Section 5
further show why close recipients should receive more reporting: the functions of re-
porting in correspondence are very (inter)personal.

2. Fanny Burney and her letters

Frances (Fanny) Burney, later D’Arblay, (1752–1840) has been called the mother of
English fiction. Her work includes the novels Evelina (1778), Cecilia (1782) and Camilla
(1796), which made her a respected and popular figure in the literary circles of the
time. In her letters, Fanny Burney talks about her literary pursuits, but mostly she
writes about everyday concerns, family and friends.
Most of the 65 letters we have analysed were written to Fanny’s father, sisters and
brother (Table 1).1 The Burney family was apparently a very close-knit unit, and Fanny
writes fondly about her siblings and father. Her father was the famous Dr Charles
Burney, a music teacher, musician and author of History of Music. Fanny carried out

1. The letter material used in the study is taken from the Corpus of Early English Correspon-
dence Extension (CEECE), currently under compilation in the Research Unit for Variation, Con-
tacts and Change in English at the University of Helsinki. The Extension covers personal cor-
respondence from 1653 to 1800, and in its current size (2.2 million words; November 2008)
contains 4923 letters from 308 writers.
Reporting in 18th-century correspondence 

extensive work for her father’s publication, and he was clearly an important influence
in Fanny’s life, making her abandon the publication of several unsuitable plays and
disapproving of her choice of a husband. In addition to these recipients, there are let-
ters addressed to two friends, Samuel Crisp and Marianne Waddington. Together with
her father, Crisp was an important influence in Fanny’s life. He was a cultivated bach-
elor, originally a friend of Charles Burney, who addressed Fanny as his “dear Fannikin”;
to Fanny he was her “Daddy Crisp”. Marianne Waddington became Fanny Burney’s
acquaintance and friend in the mid-1780s through Mrs Delany, who was a court fa-
vourite at Windsor and presented Fanny to the king and queen. Marianne, thus, be-
longed to the circle of people Fanny Burney got to know during her years at court as
second keeper of the robes to Queen Charlotte. Finally, there are a couple of letters to
Fanny’s sister-in-law Sarah Rose Burney, who was married to her brother Charles
(Harman 2001; Rogers 2004).
The majority of the letters we have analysed were written in the 1790s, when Fanny
was in her forties and had already established her literary fame. She married M.
D’Arblay in 1793 and had a son, Alexander, in 1794. Her husband and son, therefore,
appear only in letters written after 1793. Eleven letters date from the 1770s, when
Fanny was still a single woman in her twenties, writing her first novel, Evelina, which
was published anonymously, and living in her father and stepmother’s household.

Table 1.  Fanny Burney’s (FB) letters (1774–1800) to various recipients

Recipient Relationship to FB Word count Number of letters

Charles Burney Father   1590  2
Esther (Hetty) Burney Sister    498  2
Susanna Elizabeth Burney Sister    283  1
Samuel Crisp Friend   3361  6
Charles Burney Father 11247 14
Esther (Hetty) Burney Sister 12650 13
Susanna Elizabeth Phillips Sister   4491  4
Charlotte Ann Francis/Broome Sister   8121 10
Charles Burney Brother   1987  4
Sarah Rose Burney Sister-in-law   1526  3
Marianne Waddington Friend   4805  6
TOTAL 50559 65
 Minna Nevala and Minna Palander-Collin

3. Reporting speech and writing

3.1 Direct and indirect strategies

Traditional grammatical accounts see reporting as a syntactic phenomenon, which

means that reporting is understood as “a series of operations carried out on direct
speech to transform it into indirect speech” (Baynham 1991: 89). Direct and indirect
reporting are said to differ in their closeness to the original wording: ‘indirect report-
ing’ (Example 1) reports only content, not the original words, whereas ‘direct report-
ing’ is supposed to report the exact original wording (Example 2).
(1) I asked what he would send her in return
 (FB to Mrs Charlotte Broome, 1798, IV, 228)
(2) he answered ‘I will send her some breeches & spatterdashes – when I have
got some.-‘ (FB to Mrs Charlotte Broome, 1798, IV, 228)
Research on indirect and direct reporting in context, however, has shown that report-
ing, whether indirect or direct, seldom reproduces the original wording faithfully
(cf. Collins 2001: 2).2 It may be difficult to remember and reproduce utterances pre-
cisely, but more importantly the reporter uses the reported utterance in a different
context from the original.3 Therefore, the reporter is likely to alter the original wording
in one way or another and add her own point of view to the report. From the perspec-
tive of evidentiality, direct and indirect speech or writing may carry different truth
claims, as in direct reporting the reporter gives the impression of expressing both the
content and the wording of the original utterance, whereas in indirect reporting the
reporter appears only to paraphrase the meaning of the reported utterance as he un-
derstood it (Baynham 1996: 64).
Thus, it seems not to be the inaccurate vs. accurate wording as such, but other fac-
tors which condition the use of indirect and direct reporting. Direct reporting has, for
instance, been found to mark the climax of the story, and is said to be more vivid in
creating involvement in the text (Tannen 1989; Baynham 1991, 1996; Holt 2000), and
our analysis supports this view. Table 2 summarises the findings of some earlier studies.
In our discourse-pragmatic analysis the word ‘reporting’ signifies the conveying of
information that was supposedly uttered, thought or written by someone in a given
situation or source and is now repeated in a new situation; it also contains an overt

2. Clark & Gerrig (1990: 774) argue that quotations depict rather than describe selected as-
pects of the original utterance or conversation.
3. Johansson (2002: 255) uses the term ‘recontextualization’ of a reconstruction process which
concerns a situation in which a speaker quotes something another speaker has said in another
context. Here, a sequence of reported speech, or represented discourse as Johansson calls it,
takes full meaning in the new discursive context in which it is embedded.
Reporting in 18th-century correspondence 

Table 2.  Functional differences between indirect and direct reporting. (Based on
Palander-Collin and Nevala 2006: 128–129)

Indirect reporting Direct reporting

Huddleston 2002 – Reports only the contents of – Reports the actual wording of
the original the original
McGregor 1994 – Focus on the present speech – Focus on the original speech
situation and the reporter situation and the reported
Coulmas 1986 –  Reporter’s point of view – The reported speaker’s point
of view
Halliday & Matthiessen – Projected element has – Projected element has
2004 dependent status (parataxis) independent status (hypo-
→ less immediate and lifelike taxis) → more immediate and
– General situation reported –  Narrative registers, fictional,
– Specific situation quoted
Bell 1991 – The norm in news stories – Presents facts (strong
evidence in the potential libel
(in news media) – Puts the journalist in control – Distances and disowns,
of focusing the story absolves journalist and news
outlet of endorsement of what
the source said
– Stylistic reasons, adds to the
story the flavour of the
original speaker (e.g.
colourful quotes)
Holt 2000 & 1996 – Often background informa- – Recalls utterances that are the
(in casual conversation) tion, recall utterances made focus of a story
on another occasion from the – Used at the climax of the
one being described story
– Effective way of “recreating”
– Effective and economical way
of providing evidence -> air
of objectivity to the account
Tannen 1989 – Creates involvement

marker indicating reporting, i.e. a reporting frame (cf. Tannen 1989: 105). The frame
may contain a verb of saying/writing (Example 3), it may be in the form of a noun of
saying/writing (Example 4), or it may even indicate reporting as heard by someone
 Minna Nevala and Minna Palander-Collin

(Example 5). The reporting frame often indicates the reported person, but not neces-
sarily (Examples 5 and 6).
(3) My mother then asked young Mr. Beatson [= reporting frame] where he would
prefer to be, in the Boxes or the Pit? [= report]
 (FB to Samuel Crisp, 1776?, II,205)
(4) he has sent (& written) the Letter [= reporting frame] which exhorts the King
of Prussia to order the Duke of Brunswick to banish & drive from his Do-
minions All the Emigrants there in asylum! [= report]
 (FB to Susanna Elisabeth Phillips, 1798, IV,69)
(5) Sophy I hear [= reporting frame] is at Chelsea [= report]
 (FB to Esther Burney, 1797, III,321)
(6) The water is said [= reporting frame] to be excellent [= report]
 (FB to Dr Charles Burney, 1796, III,243)
Reporting in written discourse such as private letters may be affected by intertextual
elements. As a letter is both a response to previous letters as well as a cause for new
ones, series of letters can also be seen in terms of intertextual relations (Fitzmaurice
2002: 2). This means that sometimes the reported content and its source are already
known to the recipient on the basis of previous correspondence, and the source of the
reported utterance may be left out or the report may not be framed as such at all. As
Collins (2001: 2) notes, “though much of the information communicated in everyday
conversation comes at second or third hand, often it is neither meant nor perceived as
RS [reported speech].” In our analysis, however, instances of reporting without a re-
porting frame are not included.
The distribution of reporting clauses varies according to the type of discourse;
thus, reporting forms and functions can be said to be register specific. For instance,
Halliday & Matthiessen (2004: 444) state that present-day spoken English favours di-
rect reporting (‘paratactic projection of locutions’) over indirect reporting (‘hypotactic
projection of locutions’). In news media, on the other hand, indirect reporting is the
norm and direct reporting the exception, as e.g. Bell (1991: 209) has shown.4 Semino
and Short’s (2004) corpus-based study also shows that the presentation of speech,
writing and thought is used to a different degree and in different ways in prose fiction,
newspaper news reports and (auto)biography. Fictional texts contain more thought

4. Preference for direct or indirect reporting strategies can naturally be subject to change.
Interestingly, Collins (2006) has found that a diachronic change has taken place in the use of
direct and indirect reporting in Russian trial transcripts. In the fifteenth century, trials were
primarily oral confrontations, and so direct reporting was preferred in the transcripts. By the
seventeenth century, trial documents written by professional clerks were mostly transcripts of
direct testimonies given earlier and then verified at a court hearing. Collins (2006: 286) argues
that for this reason direct speech reporting proved to be no longer necessary and indirect re-
porting was preferred instead.
Reporting in 18th-century correspondence 

presentation than the other two genres, whereas press sections are characterised by
more speech presentation and (auto)biographical texts contain more writing presenta-
tion (Semino and Short 2004: 225). We have not counted the presentation of speech,
writing and thought separately, but in letters speech presentation seems to dominate,
although writing presentation is also common, while thought presentation is clearly
less common.

3.2 General and contextual functions

In order to analyse the discourse functions of Fanny Burney’s reporting segments, we

have roughly divided the instances of reporting in her letters into two main categories
according to their interactional purposes: general and contextual. The division is il-
lustrated in Figure 1.
The general function category includes instances which concern information about
e.g. news, events, historical and/or contemporary facts and literary quotations. These
reports are used in a general and non-personal manner, although if accompanied by
Burney’s evaluative remarks are contextualised for personal or interpersonal involve-
ment. Example (7) shows a report from a newspaper focusing on current news.
(7) To Day’s papers teem with the promise of great & decisive victories to the arms
of the Duke of Brunswick – (FB to Dr Charles Burney, 1792, I,230)
The contextual function category can further be divided into personal and interper-
sonal categories. As can be seen in Example (8), writer-oriented reporting, which
functions as a tool of self-involvement, includes cases in which Burney herself is the
initiating or receiving person. In other words, the reported clause concerns something
that has been said or written to her and/or about her, or, for example, her books or
physical appearance. Personal reporting supports the writer’s own emotions and opin-
ions, promotes his/her own gain, and may emphasise a certain (communicative) event


General Contextual

Personal Interpersonal

Figure 1.  Discourse functions of reporting

 Minna Nevala and Minna Palander-Collin

important to the writer. In Example (8), the topic is the money owed to Fanny by Mrs
(8) I have just received a message from Mrs. Boscawen, that she has now more cash
in her House belonging to me than she chuses to keep in there in these bad
times! – £60! (FB to Dr Charles Burney, 1795, III,142)
Contextual involvement can also be interpersonal. This category includes the instanc-
es of recipient- or reported person-oriented reporting in which Burney repeats some-
thing that has been said or written about either the recipient or a third person or per-
sons. Interpersonal reporting can involve quoting descriptions of events which have
involved other interactants in the correspondence, but also may express politeness,
courtesy, or impoliteness to the recipient or the person reported. By reporting some-
thing positive about the recipient/source of the report, the writer aims to strengthen
her relationship with other interactants in some way, as well as to receive confirmation
of her own opinions from the recipient. When repeating negative comments about the
recipient, the writer usually evaluates the situation for his/her own benefit (“I don’t
think that way, you can trust me but not the person reported”), expecting the recipient
to agree and to keep the primary source in the “out-group”.
In Burney’s letters, interpersonal reporting often involves expressing politeness or
repeating something good which has been said or written about the recipient of the
letter, as in Example (9), where Burney compliments her sister Esther by reporting Mrs
Locke’s positive comments about her. Unlike Hester Piozzi, Burney mostly confines
herself to positive reporting, and seldom quotes negative matters.
(9) Our sweet Mrs. Locke – with whom we are going to spend the Evening – tells me
you looked better than usual – & speaks with animated pleasure of what she
calls your kind reception of her, in the few minutes she spent with you. It
doubled her concern, she says, that she could see you no more.
 (FB to Esther Burney, 1797, III,321)
There are also cases which function both as general and as contextual. In Example (10),
Burney quotes her sister Susanna, who is currently in the middle of a rebellion in Ire-
land. She shares Susanna’s news with her other sister, Esther, but instead of confining
her reporting to bare facts about the tumultuous situation, she also emotionally in-
volves both herself and the recipient by quoting Susanna’s attitude to Esther: “she is sure
you must be a full sharer of the extreme anxiety the situation of Ireland has given me”.
(10) & she [Susanna Burney] charges me to make you, as well as our dear Father, to
whom I have this moment written, acquainted with its chief contents – for she
is sure you must be a full sharer of the extreme anxiety the situation of Ire-
land has given me, & she assures us all that she believes herself in perfect
safety, that all her neighbourhood are friendly & good & loyal, & all the
poor people about, & all the Major’s workmen are even kindly attached to
Reporting in 18th-century correspondence 

them. And – which is truly consolatory – that if the rebellion continues, the
Major has promised to take them to Dublin – whence, she says, the passage
is very short to Wales, &, should it still not be crushed, she & her 3 loves are
all to come over – (FB to Esther Burney, 1798, IV,142)

4. Reporting frequencies

In Fanny Burney’s letters, reporting can be said to be a relatively common phenome-

non, as 59 of the 65 letters we have analysed contain reporting. On average, a single
letter contains five instances of reporting. It is also typical for reporting to occur in a
sequence of several reported utterances; 57 per cent of reporting occurs in such se-
The letters in our sample were written in the 1770s and 1790s. Unfortunately, the
1770s sample is fairly small, but it appears that in her earlier letters, written as a young
woman, Fanny was less concerned with reporting what others had said, since the fre-
quency of reporting increases in her later letters, which she wrote in the 1790s as a
mature woman, wife and mother (Table 3). A frequency analysis of personal pronouns
in Fanny’s letters supports this idea: her 1770s letters contain significantly higher fre-
quencies of first-person singular (I) and second-person (YOU) pronouns than her let-
ters written in the 1790s, and in these later letters, she uses the third-person singular
(SHE/HE) and first-person plural (WE) pronouns significantly more frequently than
in her earlier letters. It seems that as a young woman her communication is more fo-
cused on herself and the recipient, and she builds rapport with the recipient by means
of self reference and addressee inclusion, whereas some twenty years later her atten-
tion is not predominantly focused on the self-addressee axis, and she includes third
parties in her letters, at least in her use of pronouns and reporting.
The frequency of reporting varies depending on the recipient, and it seems that
intimacy and closeness between the correspondents encourages reporting (Table 3).
Like Hester Piozzi, Fanny Burney reports most often to the correspondents closest to
her, such as her father Dr Charles Burney and sister Esther. Her other sisters also re-
ceive more reporting than her brother Charles, her friends Samuel Crisp and
Mrs Waddington, or her sister-in-law Sarah Burney, who receives very little reporting.
In fact, two of the five letters to Sarah Burney do not contain any reporting at all. All
in all, indirect reporting (83%) is far more common than direct reporting (17%), and
Mrs Waddington and Sarah Burney do not receive any direct quotes. This corroborates
our earlier findings as well: in Hester Piozzi’s letters the corresponding figures were
77 per cent for indirect and 23 per cent for direct reporting.
If reporting is typical in letters to close addressees, the most reported persons also
tend to be family members or otherwise close friends. All in all, Fanny reports the
words of c. 60 different persons, but many of these are reported only once or only to
one recipient. Only nine people are reported to three or more different recipients.
 Minna Nevala and Minna Palander-Collin

Table 3.  The frequency of reporting in Fanny Burney’s letters to various recipients

Recipient Indirect % Direct % TOTAL Fq/1,000 Fq/letter


1770s Samuel Crisp   10   56%  8 44% 18 5.4 3

Charles Burney    7   78%  2 22%  9 5.7 4.5
Sisters    1   50%  1 50%  2 2.6 0.7
Total 1770s   18   62% 11 38% 29 5.1 2.6
1790s Charles Burney   78   81% 18 19% 96 8.5 6.9
Esther (Hetty)   87   89% 11 11% 98 7.7 7.5
Burney (sister)
Charlotte Ann   46   87%  7 13% 53 6.5 5.3
Susanna Elizabeth   14   74%  5 26% 19 4.2 4.8
Phillips (Sister)
Charles Burney    8   73%  3 27% 11 5.5 2.8
Mrs Waddington   18 100%  0   0% 18 3.7 3.0
Sarah Rose Burney    5 100%  0   0%  5 3.3 1.7
Total 1790s 256 85% 44 15% 300 6.7 5.6
TOTAL 274 83% 55 17% 329 6.5 5.1

These are Fanny’s sister Susanna (43 mentions), her husband M. D’Arblay (33), her
friend Mrs Locke (15), her son Alexander (12), Miss Cambridge (12), her father
Dr Charles Burney (12), her brother James (7), her brother Charles (4) and her step-
mother Mrs Burney (4). Fanny also reports her own words in another context
(16 mentions) to all the recipients except Sarah Rose Burney and Mrs Waddington.
Unlike Hester Piozzi, however, Fanny Burney does not show any particular bias to re-
porting either men or women, as both are reported fairly equally.5 This may simply
result from the fact that Piozzi’s correspondents were mostly not family members,
whereas Fanny’s letters are predominantly addressed to her relatives.

5. In Palander-Collin and Nevala (2010) we have found that Fanny’s father Charles very rare-
ly reports on women’s sayings: the daughter Hetty is the most reported woman and she is re-
ferred to twice. Moreover, only 16 instances of reporting are attributed to women, which is
14 percent of the cases that can clearly be attributed to either men or women.
Reporting in 18th-century correspondence 

5. Reporting functions

5.1 General reporting

As regards the general and contextual functions of reporting in Burney’s letters, it

seems that she uses different reporting strategies for different purposes. In general,
Burney rarely uses reporting to highlight events or opinions of a general sort. Although
she reports the tumultuous situations in Ireland or in France at the time, she eventu-
ally converts the general into the personal or interpersonal by foregrounding or com-
menting on the matter and/or the persons reported and referred to who are close to
either her, the recipient or both.
(11) I am very impatient to know if the Invasion threat affects your part of Ireland.
Mr Oracle is of opinion the French soldiers will not go to Ireland, though
there flattered with much help, because they there can expect but little ad-
vantage, after all the accounts spread by the opposition of its starving condi-
tion: but that they will come to England, though sure of contest, at least,
because there they expect the very road to be paved with gold.
 (FB to Susanna Phillips, 1798?, IV,67)
In Example (11), the situation is similar to the one already presented in Example (10):
Burney is concerned with her sister, who lives in Ireland, and uses the indirect report-
ing strategy to convince both her sister and herself that there is no immediate threat
from the French soldiers in Ireland, since Mr Oracle thinks so.
Reports of general news are rarely from established sources such as newspapers;
quotes such as that in Example (7) are unusual in Burney’s letters. Most often the primary
person reported is at least an acquaintance, or, as in Example (12), a family servant.
(12) Three hundred Labourers, & poor Men, assembled yesterday in Dorking, de-
claring they would work no more, while bread was at such a price, unless
their wages were again raised. They made their application at Lord Leslie’s,
– who must have been much distressed how to act. This we heard from our
little maiden, who went thither on an errand.
 (FB to Dr Charles Burney, 1800, IV,401)
Here again Burney adds a personal comment after the indirect quote, expressing her
concern for Lord Leslie’s position in the matter with the words “who must have been
much distressed how to act”. By adding this comment, she also commits herself to
Lord Leslie’s in-group, as opposed to the three hundred labourers and poor men. Our
earlier study shows that these kinds of evaluative remarks were common in Hester
Piozzi’s letters, and it is not surprising to find them attached to Burney’s reporting
clauses either. We will discuss evaluative comments in more detail in Section 5.3.
 Minna Nevala and Minna Palander-Collin

5.2 Contextual reporting

Bearing in mind that the material for this study is personal correspondence, it is not
surprising that Burney uses reporting in contextual, i.e. personal or interpersonal,
functions more often than for the general purpose of conveying news. There are, how-
ever, some differences between the personal and interpersonal uses of reporting.
Burney’s reporting is most often personal when she writes to her father Charles, espe-
cially in her earlier letters. She may quote what has been said about herself or her
books, or who has read them and liked them. Example (13) shows an instance in which
Fanny first reports something she has heard from Mrs Boscawen and Miss Cambridge,
and then quotes Miss Cambridge’s answer, which in turn includes an embedded report
of Mrs Boscawen’s response to the matter.
(13) I had heard of such uncommon exertions from her, & of her principal Agent,
Miss Cambridge, that I had written to beseech not to have my Book a burthen,
& that the name, & honour of such a Bookkeeper, was all I desired. Miss
Cambridge writ me for answer –
‘Mrs. Boscawen laughs at the notion of her & I sitting with our hands before
us upon such a business. – she says she will not accept such a sinecure,
though you so graciously offer it.’ (FB to Dr Charles Burney, 1795, III,143)
Here, Burney uses an indirect reporting strategy in order to provide the background
for the entire discussion. She then turns to a direct quote from Miss Cambridge’s letter,
which highlights the impact Mrs Boscawen’s exact reported words have made on her.
Fanny reports news of her husband more often to her father than to other mem-
bers of the family. It is obvious from the tone in her letters that she values her father’s
opinion and trusts him enough to share her emotions towards the members of her im-
mediate family. Example (13) is an indirect quote from Fanny’s husband, to whom she
often refers with terms like “the gardener”, “my better half ”, “my Monsieur” or plain
“M. d’Arblay”.
(13) My philosophic Gardener [Monsieur d’Arblay], however, made his first chagrin
his last, & read this final sentence with perfect sang froid, telling me I should
be at least a little consoled that there were finer things said of me in the
Preface than in decency, he could have said himself.
 (FB to Dr Charles Burney, 1794?, III,62)
Fanny also writes about her son Alex to her father in more detail than to her sisters.
The next example is a quote of something little Alex has said in reply to his mother.
Burney again uses direct reporting to highlight the unique occasion of Alex’s first in-
dependent reply.
(14) Yesterday my Bab, for the first time, spoke in reply; & I deemed it an Epock
worthy Record to his dear Grandpapa. He was eating some bread, & let a piece
Reporting in 18th-century correspondence 

fall; I presented it to him, saying ‘Will Bab have it? –’ He looked irresolute, &,
to help him, I added ‘Say Yes, Mama. –’ He then turned round, with a look
more decisive, & instead of ecchoing me as heretofore, pronounced ‘Lo,
Mama!’ Lo stands for no. ... I gave him a bit of Cake, ... bidding him, at the
same time, repeat after me ‘Thank you, Mama. –’ He took it with great glee; ...
but called out ‘Dea Ock! -’ [‘Dear Locke!’]
 (FB to Dr Charles Burney, 1796, III,242–243)
Although Burney does discuss personal issues with her brother as well, the reporting
in the letters to her siblings appears to be more interpersonal and focused on their
children and other members of the Burney family. Example (15) is a report of a com-
pliment to Esther Burney, the recipient, made by Mrs Locke in a letter to Fanny
(cf. Example 9).
(15) Mrs. Locke, in a Letter from London during her first residence, says ‘I have had
the happiness to spend a comfortable half Hour with sweet Mrs. Burney &
her charming Girls –’ &c &c & in her only visit here between the journies, she
said ‘She is so amiable as to make light of her removal, because she must be
sure how I felt it for her, – but she chearfully declared it was to a very pretty
Apartment, eligibly & genteely situation, & would not let me be uneasy –’
 (FB to Esther Burney, 1798, IV,144)
In Section 3.2 we discussed instances which combine both general and contextual re-
porting functions. Indirect reporting can also be used as a strategy to combine infor-
mation about the original source of the report with a quote relating to the reporter,
Burney herself. This kind of mixing of personal and interpersonal functions is com-
mon in Fanny’s reporting. The letter introduced in Example (16) also includes a re-
ported passage from Mr and Mrs Barbauld, who in turn quoted something Mrs Chapone
had said.
(16) Mr. & Mrs. Barbauld, whose names I am sure you must know, (she was Miss
Aiken) called upon me last Week, & gave me a good account of dear Mrs.
Chapone, from whom they had heard of our trio visit, & to whom she had
kindly said That she had been agreeably disappointed of a party that she had
invited to meet us.  (FB to Esther Burney, 1798, IV,144)
Burney’s own positive attitude towards the original person reported, Mrs Chapone, is
shown in the phrases “dear Mrs Chapone” and “she had kindly said”. The reported pas-
sage from Mrs Chapone, “That she had been agreeably disappointed of a party that she
had invited to meet us”, reflects in turn Chapone’s own feelings.6
Burney’s interpersonal reporting also includes examples in which she quotes
something which relates to a third party, i.e. not to herself or the recipient, but who is

6. According to the OED, agreeably can also be used to mean “In a way that answers to cir-
cumstances or the nature of things; suitably, fittingly”, which seems to be the case here.
 Minna Nevala and Minna Palander-Collin

still somehow important to both. Example (17) shows a long stretch of indirect report-
ing of a discussion between Dr Burney, Mrs Ord, Mrs Smith and Mr Pepys.
(17) When we took leave, my Father told Mrs Ord that it gave him great pleasure
to say that he knew 2 or 3 Houses even in these Times, where Company
could be entertained & got together merely by conversation, unassisted by
Cards, &c. ‘Such Parties as Mrs Ord collects, said Mrs Smith, cannot fail in
regard to Entertainment.’ ‘And yet, answered Mr Pepys, I have known meet-
ings where equal pleasure has been proposed & expected, & where the in-
gredients have been equally good, & yet the Pudding has proved very bad.’
“‘True, returned my Father, for if the Ingredients are not well mixed, their
separate goodness does not signify; for if one is a little too sour, & another
a little too sweet, or too bitter, they counter-act each other: but Mrs Ord is
an excellent Cook, & takes care not to put clashing materials into one mess.’”
 (FB to Samuel Crisp, 1776?, II,203)
The quote, of course, shows not only Mrs Ord but also Fanny’s father in a favourable
light; he compliments Mrs Ord both to the hostess herself and to the other guests.
Again, Fanny uses direct reporting to emphasise, most probably, her father’s good
nature and witty words and her own agreement with the opinion presented in the

5.3 Evaluative remarks

As already mentioned, our earlier study shows that Hester Piozzi made evaluative re-
marks in connection with reporting. Burney also uses this discourse strategy, which is
often used to make the recipient more clearly understand what the purpose of the re-
porting is in a particular context. Her comments most commonly follow reporting
sequences which foreground her own positive and negative feelings about the person
or situation in question. Example (18) shows an instance of positive evaluation
(18) I heard from little Martin of your loyalty – & of his. – which he was properly
proud of. I think him very much improved.
 (FB to Charles Burney, 1798, IV,126)
Example (19), on the other hand, shows how an incident at the theatre has annoyed
Fanny. By using direct reporting she emphasises the “boldness” of young Miss Beatson
who, in Fanny’s opinion, should not have challenged her mother’s decision, and her
own disapproval of the whole event. To make sure the recipient understands her nega-
tive attitude, she adds a negative evaluative remark after the long quote.
(19) You may have heard me frequently mention what a very easy young Lady Miss
Beatson is, who has no Notion of doing any thing but what she likes, so, when she
Reporting in 18th-century correspondence 

heard this, she whispered me, ‘Well, if we part, you and I will go into the Boxes,
& Mrs. Burney & Bob shall go into the Pit – though Bob won’t like that.’
Accordingly, when we got to the Door, the man refused the Box Tickets. My
mother then asked young Mr. Beatson where he would prefer to be, in the
Boxes or the Pit?
‘In the Boxes, Ma’am,’ answered he, very composedly.
‘Well then, said Mama to Nelly, ‘you & Fanny will go into the Pit, & your
Brother & I into the Boxes, & we will meet by & bye.’
‘No, no,’ said Nelly colouring, ‘Not so, no, I shall go where ever Miss Burney
chooses, myself.’
Did you ever hear such boldness in your Life?
 (FB to Samuel Crisp, 1776?, II,205)
In making the comment, Burney also seeks confirmation of her own attitude, prompt-
ing the recipient to respond in a similar way. In her letters, evaluative commenting,
and direct reporting in particular, appears in emotionally laden contexts. This cor-
roborates the earlier findings in our research on Piozzi’s letters, that direct reporting is
most often used as a means of creating involvement in the situation. Yet another good
example of emotional involvement can be seen in the following quote, in which Fanny
comments to Esther on what their sister Susanna has written to Mrs Locke.
(20) O England! England! she says – my Father – my Sisters – my friends! – shall
I ever see you more? – I think I shall die of the joy – yet I feel as if I could not
outlive a disappointment!’ What touching words!
 (FB to Esther Burney, 1799, IV,350)
What we have presented here are examples of typical evaluative clauses, which are usu-
ally placed outside the reporting sequence itself. Dossena (2000: 311) calls this sort of
manipulation of the readers’ participation “emotional colouring”, also created by the
use of adverbs and pronouns. Burney uses commenting devices such as adverbs and
adjectives to increase the reader’s emotional involvement. In contrast to interjections,
these are usually embedded in the reporting frame and can be used to evaluate the
primary or secondary reported person and even the report itself.

6. Conclusions

In our discussion of reporting in Fanny Burney’s personal correspondence we have

shown that reporting often serves personal and interpersonal functions. Fanny reports
what has been said about herself, what she has said, what has been said about the re-
cipient or what mutual friends or relatives have said. Consequently, it is not surprising
that reporting is more typical in letters to close recipients, and that the sayings of those
closest to her are much more often reported than general sources like newspapers.
 Minna Nevala and Minna Palander-Collin

Often Fanny’s reporting is accompanied with evaluative remarks that further

highlight the interactive nature of letters and emphasise the use of reporting to show
personal stance. Direct reports in particular occur in emotionally laden contexts and
are accompanied by Fanny’s comments. Fanny’s evaluative remarks are mostly posi-
tive, which perhaps reflects her attitude towards letter writing and friendship, as ex-
pressed in the following quotation:
The basis of Letters, as of Friendship, must be kindness, which does not count lines
& words, but expressions & meaning, which is indulgent to brevity, puts a favour-
able construction upon silence, grants full liberty to inclination, & makes every
allowance for convenience. (FB to Mrs Marianne Waddington, 1797?, III,317)

The tendencies observed here corroborate our findings concerning reporting in the
letters of another eighteenth-century literary figure, Hester Piozzi. In Piozzi’s letters,
however, women correspondents received more reporting than men and men were
more often reported than women, whereas in Burney’s letters the gender of the recipi-
ent or the reported person was not important. This may result from differences in the
correspondences of these two women, since Piozzi’s letters were mostly addressed to
friends and acquaintances and Burney’s letters to relatives. It seems, on the basis of the
correspondences of Burney and Piozzi, that there are general tendencies in the nature
of writer-addressee relationships in which reporting tends to occur and in the func-
tions reporting serves in personal correspondence.
Moreover, our analysis of Fanny Burney’s letters shows that an individual’s life
span may be an important factor in the use of reporting. On measuring the frequency
of pronoun use, we found that Fanny’s early letters were more focused on herself (I) and
the recipient (YOU) than any third parties (HE/SHE). Reporting was also less frequent
in her earlier than in her later letters, which contained more reporting but also signifi-
cantly more references to third parties in the form of third-person pronouns.


Baynham, Mike. 1991. “Speech reporting as a discourse strategy: Some issues of acquisition and
use.” Australian Review of Applied Linguistics (ARAL) 14(2): 87–114.
Baynham, Mike. 1996. “Direct speech: What’s it doing in non-narrative discourse?” Journal of
Pragmatics 25: 61–81.
Bell, Allan. 1991. The Language of the News Media. Oxford, UK & Cambridge, Mass.: Blackwell.
Clark, Herbert and Richard J. Gerrig. 1990. “Quotations as demonstrations.” Language 66(4):
Collins, Daniel E. 2001. Reanimated Voices: Speech Reporting in a Historical-Pragmatic Perspec-
tive. Amsterdam & Philadelphia: John Benjamins.
Collins, Daniel E. 2006. “Speech reporting and the suppression of orality in seventeenth-century
Russian trial dossiers.” Journal of Historical Pragmatics 7(2): 265–292.
Reporting in 18th-century correspondence 

Coulmas, Florian, ed. 1986. Direct and Indirect Speech. Berlin, New York & Amsterdam: Mouton
de Gruyter.
Dossena, Marina. 2000. “Truth and murder will out: Reported speech and quoted speech in 19th
century accounts of the Highland Clearances.” In English Diachronic Pragmatics, Gabriella
di Martino and Maria Lima (eds), 299–318. Napoli: CUEN.
Fairclough, Norman. 1992. Discourse and Social Change. Cambridge: Polity Press.
Fitzmaurice, Susan. 2002. The Familiar Letter in Early Modern English. (Pragmatics and Beyond
New Series 95.) Amsterdam & Philadelphia: John Benjamins.
Halliday, M.A.K. and Christian M.I.M. Matthiessen. 2004. An Introduction to Functional Gram-
mar. Third edition. London: Arnold.
Harman, Claire. 2001. Fanny Burney. A Biography. New York: Alfred A Knopf.
Holt, Elizabeth. 1996. “Reporting on Talk: The Use of Direct Reported Speech in Conversation.”
Research on Language and Social Interaction 29(3): 219–245.
Holt, Elizabeth. 2000. “Reporting and Reacting: Concurrent Responses to Reported Speech.”
Research on Language and Social Interaction 33(4): 425–454.
Huddleston, Rodney. 2002. “11 content clauses and reported speech.” In The Cambridge Gram-
mar of the English Language, Rodney Huddleston and Geoffrey K. Pullum, 947–1030.
Cambridge: Cambridge University Press.
Johansson, Marjut. 2002. “Sequential positioning of represented discourse in institutional me-
dia interaction.” In Rethinking Sequentiality: Linguistics Meets Conversational Interaction,
Anita Fetzer and Christiane Meierkord (eds), 249–271. (Pragmatics and Beyond New
Series 103.) Amsterdam & Philadelphia: John Benjamins.
McGregor, William. 1997. Semiotic Grammar. Oxford: Oxford University Press.
Palander-Collin, Minna and Minna Nevala. 2006. “Reporting in eighteenth-century letters of
Hester Piozzi.” In Syntax, Style and Grammatical Norms: English from 1500–2000, Chris-
tiane Dalton-Puffer, Nikolaus Ritt, Herbert Schendl and Dieter Kastovsky (eds), 123–141.
(Linguistic Insights 39.) Frankfurt: Peter Lang.
Palander-Collin, Minna and Minna Nevala. 2010. “Reporting and social role construction in
eighteenth-century personal correspondence.” In Social Roles and Language Practices in
Late Modern England, Päivi Pahta, Minna Nevala, Arja Nurmi and Minna Palander-Collin
(eds), 111–133. Amsterdam: John Benjamins.
Rogers, Pat. 2004. “Burney, Frances (1752–1840).” In Oxford Dictionary of National Biography.
Oxford University Press. [, accessed 29 March
Semino, Elena and Mick Short. 2004. Corpus Stylistics: Speech, Writing and Thought Presentation
in a Corpus of English Writing. London: Routledge.
Tannen, Deborah. 1989. Talking Voices: Repetition, Dialogue, and Imagery in Conversational Dis-
course. Cambridge: Cambridge University Press.
part 3

Managing interpersonal relations

Power in Early Modern English
courtroom discourse

Barbara Kryk-Kastovsky

The paper discusses power in Early Modern English courtroom. A few levels
of linguistic analysis are examined for the occurrence of the exponents of
power. As regards lexicon, it is the specific vocabulary which creates the social
distance between the interrogators and the interrogated. My hypothesis is that
the vocabulary items include not only what I call “overt” exponents of power,
e.g. impolite expressions, but also “covert” ones, like an ironic use of politeness
markers or forms of address. Moreover, power can be reflected in the syntax of
the utterances, e.g. in the questions asked by the interrogators. Finally, many
(socio-)pragmatic devices contribute to the demonstration of power. Thus,
specific speech acts, e.g. covert orders can be disguised as overt requests, and the
inferencing strategies used by the interrogators can reveal the presuppositions
often contained in their utterances.


When it comes to real-life interaction, successful performance is not exclusively due to

the power inherent either in the user or in his or her words; to a high degree, this
power resides in the society, but mediated and negotiated in the institutional setting of
a particular societal context (Mey 1993: 148).

1. Introduction

This paper is a contribution to diachronic pragmatics, which studies language as it was

used in the past. It is therefore not surprising that the idea of exploring old written
texts for the traces of pragmatic phenomena has attracted many researchers, but left
many others skeptical. Both the adherents and the opponents of the new paradigm
have been aware of the problems they have to face. One of the issues to be addressed is
the applicability of the analytical tools employed by synchronic pragmatics to dia-
chronic pragmatic studies, which is by no means uncontroversial. To cope with the
 Barbara Kryk-Kastovsky

problem, Labov formulated “uniformitarian doctrine”, which assumes parallels be-

tween the linguistic changes which happened in the past with the changes that are still
in progress (Labov 1994: 21ff). Romaine has reversed Labov’s doctrine into “uniformi-
tarian principle” and claimed that the linguistic mechanisms which operate today are
comparable to those which operated in the past (Romaine 1982: 122). Although his-
torical (socio-)pragmaticists are aware of the problems with the analysis of old written
texts, they emphasize additional factors that must be taken into account. Raumolin-
Brunberg refers to the great challenge involved in any attempt “to transfer to diachron-
ic linguistics the socio-linguistic methods that have been used for the study of present-
day languages” (Raumolin-Brunberg 1996: 16). Also Arnovick in her study of past
speech acts points to the significance of cultural context and social reality in the inter-
pretation of illocutionary history (Arnovick 1999: 13). Along the same lines, Bertu-
celli Papi (2000) asks “Is a diachronic speech act theory possible?” in the title of her
contribution to the first issue of Journal of Historical Pragmatics 1(1). Her answer to
the question is rather vague, stating that research in the field of historical speech act
analysis should proceed very cautiously. A more concrete proposal is due to Jucker and
Taavitsainen (2000), who emphasize the necessity of analyzing past speech acts in a
multidimensional pragmatic space which they share with other speech acts. Numer-
ous other studies in the area of diachronic (socio)- pragmatics have been pursuing
Labov’s idea of historical linguistics thought of as “the art of making the best use of bad
data” (Labov 1994: 11), and the present analysis is another contribution to the ongoing
Bearing these problems in mind, I will try to show in this paper how the notion of
power, originally designed for solving problems in synchronic studies, can be applied
to the analysis of diachronic data. Power has been an important analytical tool in so-
cio-pragmatics ever since the seminal work of Brown and Gilman (1960), where it
is juxtaposed with the notion of solidarity, power signifying social asymmetry
(inequality), and solidarity social symmetry (equality). The motto of this paper em-
phasizing the societal nature of the notion of power will guide me through my explora-
tions in diachronic pragmatics where I will investigate how power was instantiated in
a doubly marked context: in its institutional dimension it is the context of courtroom
discourse, in its historical dimension it is 17th century England with its unique politi-
cal and social conditions. I will analyze two Early Modern English trial records: The
Trial of Titus Oates and The Trial of Lady Alice Lisle for the possible occurrence of ex-
ponents of power, specifically, how power was exercised by the interrogators on the
interrogated in the Early Modern English courtroom. I will look for the manifestations
of power at various language levels: the lexicon (forms of address, discourse markers),
syntax (especially the form of the questions used by the interrogators in the examina-
tion process), through socio-pragmatic devices, like specific combinations of speech
acts used by the interrogators. The question that arises at this point is the rationale
behind the choice of the language data and their suitability for the analysis of the no-
tion of power, which is the topic of the next section.
Power in Early Modern English courtroom discourse 

2. Why courtroom discourse?

There are a number of reasons why courtroom discourse proves to be a rich source of
various linguistic manifestations of power:
1. The turn-taking system used in court is much more rigid and less flexible than the
one operating in face-to-face conversation (cf. Levinson 1983: 301).
2. The institutionalized character of the court is reflected in the formulaic, if slightly
archaic and stilted language used by anyone who enters the scene (cf. Lakoff 1990:
92ff). Like Levinson, Lakoff also draws parallels between the courtroom, school, and
church, and refers to the jural proceedings as “courtroom ceremony” (Lakoff ibid.).
3. According to Danet, legal discourse is concerned with “the nature, functions and
consequences of language use in the negotiation of social order”. But, despite its
highly formulaic character, legal language can represent various styles ranging
from frozen through formal and consultative to casual, characterized by different
manifestations of power (Danet 1985: 276).
4. The verbal interaction in court exemplifies various questioning strategies, which
are of relevance here since “courtroom discourse is unilateral in that barristers
enjoy a one-sided topic control of discourse” (Luchjenbroers 1997: 477).
5. In the formal language used in court politeness may be automatically present,
i.e. presupposed (cf. Kurzon 2001, based on Fraser 1990). As I will show below, no
such restriction held in 17th century England when the use of epithets and foul
language was at the discretion of Lord Chief Justice, Attorney General, and other
interrogators exerting their power on those whose social roles did not allow for it
(witnesses, defendants). This tendency in the 17th century address system has
been pointed out by Nevalainen and Raumolin-Brunberg with relation to address
formulae in Early English correspondence which was characterized by “the over-
riding relevance of power as a factor determining the choice of address form”
(Nevalainen and Raumolin-Brunberg 1995: 563).
It follows from the above that the notion of power is an integral part of the courtroom
context, i.e. its presence is presupposed. Consequently, a question arises at this point
to what extent looking for the instantiations of power in court trial records might re-
sult in any novel observations. My claim is that despite the power-laden nature of the
courtroom context, old court trial records contain various intricate manifestations of
power which were due to the unique socio-historical conditions of a given period. In
what follows I will demonstrate that in the Early Modern English court these manifes-
tations were much more explicit than nowadays. In other words, in 17th century
English courts the power relations were explicitly marked by the use of direct speech
acts or even invectives, whereas in 20th century courts the interrogators’ techniques of
 Barbara Kryk-Kastovsky

exerting power on the interrogated are much more implicit and indirect1. For all these
reasons, it would be revealing to look at the power play in the Early Modern English
court between the interrogators and the interrogated.

3. The analysis

3.1 Turn taking and the use of discourse markers

Consider the following two instances of the interrogators demonstrating their power
to the interrogated by means of turn-taking strategies:
(1) L.C.J. Now, my Lord, we shall go to our Evidence to prove, that all this is ab-
solutely false. For Ireland went out of Town into Staffordshire. (TO)
Here Lord Chief Justice controls the discourse at two levels of linguistic analysis. At
the lexical level, the choice of the discourse marker now signals the judge’s power to
start a new turn in the proceedings, which is comparable to one of the uses of well,
called by Jucker a frame marker, i.e. introducing a new topic (cf. Jucker 1997: 92). No-
tice also the wording of the rest of the sentence where almost every lexical item is bi-
ased since it carries the presupposition that the witness’s deposition was false. Thus the
phrase our evidence (supplied by the interrogators) stands in opposition to your evi-
dence (given previously by the interrogated witness). Unfortunately for the witness, the
judge continues demonstrating his power in an even more direct manner by using
words carrying a connotation of finality and indisputability, like prove (a telic verb
which implies that the action tends towards and reaches a certain goal2) and the adjec-
tive false used to deny the credibility of the witness’s deposition, especially if reinforced
by the adverb absolutely. This single sentence could be considered a masterpiece of
what van Dijk (1993: 43–48) calls “elite discourse”. Even though his theory concen-
trates on racial and political issues of the modern world, it is obviously based on the
notion of power reflected in the elite discourse by means of negative other-presenta-
tion (cf. van Dijk 1993: 84ff). He further claims that verbal defamation was much more
explicit several decades ago, as opposed to the subtle ways others can be disparaged

1. See Kurzon’s remarks on the clear differences between the overpoliteness of the English
judges as opposed to the much more straightforward, sometimes even aggressive, ways of the
American judges. However, even the instances of impoliteness in American court are appropri-
ately mitigated and hardly ever performed bald on record, (cf. Kurzon 2001: 82–3). Therefore,
American judges do not employ direct criticism, let alone invective, as was the case in Early
Modern English court.
2. Compare one of the latest definitions of telicity due to Declerk “A situation type (...) is said
to be telic when the verb phrase describing it represents the situation as tending towards a natu-
ral (inherent) point of completion, i.e. a necessary terminal point, without which the situation is
not complete and at which it naturally comes to an end”, (cf. Declerk 2006: 60).
Power in Early Modern English courtroom discourse 

nowadays. This observation is consonant with my claims on the impoliteness of Early

Modern English courtroom discourse where the interrogators were not only impolite
but even downright offensive with regard to the interrogated, which is no longer pos-
sible in modern courts, where power is manifested much more indirectly, e.g. through
artificial politeness of the English judicial style which makes lawyers ridiculous
(cf. Kurzon 2001: 82).
Apart from the manifestation of power at the lexical level demonstrated by Lord
Chief Justice in (1), he also controls the discourse at the pragmatic level by performing
a shift of the illocutionary force of his utterance from a warning (that he will refute the
witness’s deposition on the basis of the evidence he has) to what Austin (1962) calls a
constative and Searle (1976) a representative, i.e. an assertion where the judge supplies
the relevant evidence in the next sentence.
Consider (2) where the judge’s power is reflected in his making a new turn by
means of a combination of two discourse markers, i.e. well and then. Here Lord Chief
Justice decides to change the topic of the investigation in view of the witness’s uninfor-
mative response which does not answer the judge’s question, hence it does not contrib-
ute to the progress of the proceedings:
(2) L.C.J. Who was that other person?
Mr. Carpenter. I did ask his Name, but he said, he did not know him.
L.C.J. Well then, when they came there on the Tuesday night, how did you
receive them? (AL)
Notice that in both excerpts the turn-taking is fully controlled by Lord Chief Justice,
i.e. the most prominent of the interrogators. Moreover, quantitative evidence confirms
my contention that, due to their social power, the interrogators in both trials have the
floor much more often than other participants, since they do not only address the in-
terrogated (whereas the opposite does not hold), but also communicate with each
other. Consequently, my data have demonstrated so far that the interrogators exercise
their power on the interrogated, which confirms the asymmetry of their mutual rela-
tion (cf. Lakoff 1990: 89, Hiltunen 1996). The statistical evidence corroborates the di-
viding line between the two groups of participants of the trials also in the case of Titus
Oates, whose number of turns is exceeded only by the chief interrogator, Lord Chief
Justice. This follows from Oates’ double social role as the defendant and the defence.
As an Anglican minister, Dr. Titus Oates was a man of a high social standing and was
allowed to represent himself; thus he had to secure a sufficient number of interven-
tions to achieve the ultimate goal, i.e. his own defence. In Table 1 the number of turns
is given in absolute figures followed by the respective percentages, the overall number
of turns being 424.
Analogous quantitative data can be established for The Trial of Lady Alice Lisle,
where the statistics reflects the leading position of the judge, and marks Dunne, the
chief witness, as the most active person among the interrogated. In this case the overall
number of turns is 282:
 Barbara Kryk-Kastovsky

Table 1.  The number of turns performed by the main participants of The Trial
of Titus Oates

The interrogators: The interrogated:

Lord Chief Justice 93 = 21.9% Mrs. Anne Ireland 26 = 6.1%

Titus Oates 67 = 15.8% Lord Aston 16 = 3.7%
Attorney General 60 = 14.1% Mrs. Duddle 15 = 3.5%
Solicitor General 35 =   8.2% Mrs. Graves 11 = 2.6%
Mr. Richardson 10 = 2.3%
Mrs Keeling 10 = 2.3%
Mr Fowler  9 = 2.1%
Mrs Fowler   9i = 2.1%

Table 2.  The number of turns performed by the main participants of The Trial
of Lady Alice Lisle

The interrogators The interrogated

Lord Chief Justice 137 = 48.5% Dunne 82 = 29%

Mr. Pollexfen    6 =   2.1% Mr. Carpenter 26 = 9.2%
Lady Lisle 17 = 6.02%
Mrs. Carpenter 14 = 4.9%

Thus, the qualitative and quantitative information reveals that the interrogators in the
two trials exerted their power on the interrogated overtly by means of turn-taking strat-
egies and through some of their formal exponents, like discourse markers. Some more
possible ways of manifesting power in the courtroom discourse are discussed below.

3.2 Forms of address

One of the most obvious (hence overt) manifestations of power is the way social supe-
riors address social inferiors and vice versa. In the courtroom context this social divi-
sion usually runs parallel to the dividing line between the interrogators and the inter-
rogated, respectively (even when one of the interrogated was a lord, in the peculiar
context of court trial he was the powerless discourse participant as opposed to the
powerful interrogators, cf. below). In the trials analyzed here the possible forms of ad-
dress range from deferential, through neutral, to supercilious and downright offensive.
The forms of address selected below have been divided according to the social roles of
the addressees (i.e. the interrogators vs. the interrogated vs. the third party). As could
be expected, the former were addressed with all reverence by means of official titles,
whereas in the case of the two remaining groups a variety of possible forms can be
Power in Early Modern English courtroom discourse 

noticed, from very polite ones (your Lordship), to epithets (you Blockhead). My data
reveal not only a high degree of formality and positive politeness in the forms of ad-
dress, but also an interesting historical development. The polite forms of address re-
ceived by the interrogators and the witnesses seem to be very conservative. They have
not changed since Early Modern English, whereas the other forms of address and epi-
thets (like Sirrah), have been lost or they sound obsolete, and/or socially unacceptable.
In what follows three different situational contexts will be discussed, i.e. addressing the
interrogators, addressing the witnesses/defendants with socially expected forms, and
addressing the witnesses with other forms of address.

3.2.1 Addressing the interrogators

The following examples illustrate a typical formal situation determined by a power
relation between the interrogated and the interrogators, i.e. “social inferiors upwards”.
Thus the former are addressed by the latter either by the form my Lord, which entered
English at the beginning of the sixteenth century as the form of address used for noble-
men (cf. Nevalainen and Raumolin-Brunberg 1995: 575), or by their official titles,
which express deference and social distance, cf. (3) and (4), respectively:
(3) Mrs. A. Ireland. No, my Lord, my Brother came home on foot, but we staid all
Night. (TO)
(4) L. C. J. Look on your Almanack, if you have any one of that Year, Mr. Attorney.
It is worth noting at this point that the use of appropriate forms of address in a par-
ticular social context could also be considered a case of what Watts calls “politic behav-
ior”, which he defines as “socio-culturally determined behavior directed towards the
goal of establishing and/or maintaining in a state of equilibrium the personal relation-
ships between the individuals of a social group” (cf. Watts 1992: 50). Thus, forms of
address and honorifics are according to Watts instances of socially regulated use and
can be considered polite only if they go beyond their normal usage as socially con-
strained forms of politic behavior. To put it in a nutshell, in the courtroom context the
interrogated had no other choice but to engage in the socially regulated politic behav-
ior, whereas the question of polite behavior did not arise.

3.2.2 Addressing the witnesses/defendants with socially expected forms

Consider the following examples:
(5) L. C. J. Have you any Questions to ask her, Mr. Oates? (TO)
(6) Mr. Sol. Gen. Pray will your Lordship give my Lord and the Jury an account,
when Mr. Ireland came to your House, and how far he travelled with you af-
terwards? (TO)
The socially expected and highly conventionalized forms of address in the Early Mod-
ern English period included professional titles or the form Mr./Mrs. followed by the
 Barbara Kryk-Kastovsky

person’s surname, which constituted another case of politic behavior. In 18th century
England the appropriate choice of lexemes was, along with the use of socially accept-
able topics of conversation and other socio-pragmatic factors, a sign of the speaker’s
“mental cultivation and polished manners, elegant refinement and neoclassical good
taste” (Watts 1992: 49). In most of the instances discussed here the social situation
would be opposite to the one exemplified in the section; i.e. it would be the case of
“social superiors downwards”. Thus, for example, a gentleman would be addressed by
a person in high office as Mr + surname, as in (5). However, intricate instances, like (6),
where a lord was questioned by the Solicitor General, required the use of the appropri-
ate address form, which led to an amusing situation. Even though both Lord Aston and
the Solicitor General received the form lord, it came in its appropriate variations indi-
cating the social roles of the participants of the jural theater on the two sides of the bar.
Thus, Lord Aston as a witness was one of “them”, i.e. the interrogated, hence he re-
ceived the more detached form your Lordship, whereas Lord Chief Justice, the chief
interrogator, is for the Solicitor General one of “us”, thus receives the conventionalized
form my Lord.

3.2.3 Addressing the witnesses with other forms of address

The situation changes dramatically when other forms of address received by the inter-
rogated in court are considered. These can be divided into several categories, all of
them being an overt manifestation of power constituting various degrees of face-
threatening acts directed at the addressee’s positive face. For instance, name calling
(invective) was regularly used by the infamous Judge Jeffreys during the court pro-
ceedings, which confirms his bad reputation of a verbal abuser. It is also consonant
with Nevalainen’s observation that invectives were more common in Early Modern
English courts than they are nowadays and their use was associated with controversial
judges like Jeffreys (cf. Nevalainen 1994: 321; Kryk-Kastovsky 2000; 2006). Consider
the following examples where all the forms of address are overtly offensive:
(7) L. C. J. Sirrah, I charge you in the presence of God, tell me true, What other
Persons did you see that Night? (AL)
(8) L. C. J. But you Blockhead, I ask you whether you did see any body else? (AL)
(9) L. C. J. Why, thou vile Wretch didst not thou tell me just now that thou pluck’d
up the Latch? (AL)
The unambiguously disparaging nature of the forms of address employed by Judge Jef-
freys in (7)–(9) is confirmed by the respective entries in SOED.3 Judge Jeffreys used the
above epithets as a manifestation of his power which allowed him to express his super-
cilious or even contemptuous attitude towards the defendant and the witnesses and to

3. Blockhead
1. A wooden head, a wooden block for hats and wigs.
2. Hence, an utterly stupid fellow.
Power in Early Modern English courtroom discourse 

constantly intimidate them. The judge’s communicative goals follow clearly both from
the lexical choices he makes and from the structure of his utterances. Notice how the
epithets are consonant with the form of his sentences and the communicative acts they
represent. Thus, in (7) the performative use of the verb charge followed by an impera-
tive construction form an indirect speech act which expresses an order, whereas the
indirect order in (8) employs ask in its performative function, and finally the negative
yes/no question in (9) is an obvious instant of exerting power on the witness
(such questions are most difficult to answer as they contain a presupposition, cf. the
discussion on types of questions in 3.3. below).
Another case of expressing power while addressing the interrogated was a use of
expressions which were overtly polite, but in the context of courtroom discourse could
gain negative connotations:
(10) L. C. J. Look you, Friend, you say you went with Col. Penruddock to search the
House, did you find any body there? (AL)
(11) L. C. J. What say you, good Woman; Did not your Lady sup there? (AL)
Notice that it is again Judge Jeffreys who addresses the witnesses with these forms, and
again they are inappropriate, since friend was used reciprocally among members of the
same guild or profession as of the Middle Ages, and later it was adopted by merchants
(Nevalainen and Raumolin-Brunberg 1995: 575). Thus, the use of the form of address
could only be ironic or condescending, since there was no solidarity relation between
the judge and the witness. As to Good Woman, the term could be used ironically, al-
though it could also be a variation on Goodwife, a form received by members of non-
gentry in those days.
Finally, the following is an example of indirect address where Lady Alice Lisle does
not receive her usual title Lady but instead is addressed with a form referring to her
present (rather precarious) social role, i.e. the prisoner. This strategy again illustrates
the power of the judge who in the court trial context can threaten Lady Alice Lisle’s
positive face by ignoring her high social status, thus stripping her of her dignity:
(12) L. C. J. Will the Prisoner ask this Person any Questions? (AL)

3.3 Examination strategies

The power of the interrogators is also manifest in the examination strategies used to
question the defendants and witnesses in the Early Modern English court. According

Sirrah. Now arch.

A term of address used to men or boys, expressing contempt, reprimand or assumption of au-
thority on the part of the speaker.
3. A vile, sorry, or despicable person; one of opprobrious or reprehensible character; a
mean or contemptible creature.
 Barbara Kryk-Kastovsky

to Luchjenbroers (1997), following Danet (1980: 520), questions can be defined in

terms of the degree of the factuality of the potential answers (ranging from open-
ended questions of ‘high’ fact value, through wh-questions, to restrictive yes/no ques-
tions of ‘low’ fact value). Consider the following examples quoted from Luchjenbroers
(1997: 482):
(13) a. Tell me about X? open-ended questions
HIGH fact value/least control
b. When did you do X? wh-questions
Where did you X?
... etc.
c. Did you X? yes/no questions
LOW fact value/most control
Predictably, the notion of power is also reflected in the questions asked in court. Bar-
risters have least control over the witness replies with the open-ended questions and
maximal control with yes/no questions, and it is the low-fact-value, maximal control
questions that are labelled ‘leading questions’, where the interrogators provide the facts
of a testimony and the witnesses either confirm or deny them. The crucial role of lead-
ing questions in court is manifest in the barristers’ ability to create their own versions
of events. This contributes to their control of the witnesses, since they do not only al-
ready know the answer but also gear the witnesses to what they want to hear. Leading
questions usually can take three forms: declaratives, accusatory yes/no forms, and al-
ternative questions, cf. (14a, b, c) respectively:
(14) a. You had some alcohol?
b. Did you have some alcohol?
c. Did you or didn’t you have some alcohol?, Luchjenbroers (ibid.).
Let’s have a look at the roles which different types of questions play in my corpus.

3.3.1 Wh-questions
Consider the following exchange at the beginning of The Trial of Titus Oates abound-
ing in wh-questions, which, together with yes/no questions are considered the most
frequent questioning strategies in court:
(15) Mr. Sol. Gen. Mrs. Ireland, pray where did you take your leave of your Brother
Mr. Ireland, who was executed in Summer 1678, and when?
Mrs. A. Ireland. I took my leave of him the beginning of August .
Mr. Sol. Gen. What day in August, do you remember?
Mrs. A. Ireland. The 3d of August.
Mr. Sol. Gen. Where was it?
Mrs. A. Ireland. In my own Lodging.
L. C. J. Where was your Lodging?
Power in Early Modern English courtroom discourse 

Mrs. A. Ireland. In Russel-street, Covent-Garden .

L. C. J. Now tell us again the time when it was?
Mrs. A. Ireland. It was on Saturday Morning, as I remember, the 3d of August,
the Saturday after St. Ignatius’s Day.
L. C. J. How come you remember so particularly, that it was then? (TO)
In view of the above remarks, the interrogator’s choice between the two major types of
questions is by no means random: it constitutes an instrument of power. Analogously
to Danet (1980) and Luchjenbroers (1997), Stubbs (1983: 106) claims that by selecting
one rather than the other construction, the speaker can predict and control the answer
s/he would like to receive. A yes/no question is chosen, if an unambiguous answer is
expected, which is often just monosyllabic. By contrast, a wh-question requires much
more information from the addressee, so that the length of the answer is unpredict-
able. However, it turns out that the class of wh-questions is by no means homogeneous
and can be subject to a further division. Notice that in (15), which contains mainly
wh-questions, all but the last one can be classified as what Culpeper and Kytö (2000)
call “relatively closed questions”, a sub-type of “information-seeking questions” which
seek very specific information, in this case concerning the place or time of the event.
In contrast, some wh-questions can be “relatively open” in that they require a more
extensive explanation of a particular issue. For instance, while asking the last question
in (15), i.e. How come that you remember so particularly, that it was then? Lord Chief
Justice certainly expects an elaborate answer from Mrs. Ireland.
If we leave the typology of questions aside and analyse (15) from the point of view
of speech act theory, the last wh-question asked by Lord Chief Justice likewise stands
out as compared to the other questions which are simple requests for information,
cf. Searle (1969: 69). Thus, they aim at finding out the answers, and this is what their
perlocutionary effects are. In contrast, the judge’s last utterance is not a simple question
any more but casts doubt on the witness’s testimony (by means of the expression how
come). It can be assumed that the judge used this elaborate construction not only to
signal his expectation of a longer answer, but also to act from his position of power to
express his negative feelings like disbelief and/or annoyance. This could signify his
emotional involvement in the questioning process, where the asymmetrical power rela-
tionship between the interrogators and the interrogated was analogous to the court
situation observable nowadays, (cf. Lakoff 1990: 87ff, Kurzon 1995: 59, Luchjenbroers
1997: 480). Notice also how (15) differs from (8) and (9) above. On the one hand, in the
two latter excerpts we can detect an obvious emotional involvement of judge Jeffreys,
who bullies and abuses the chief witness Dunne and the defendant Lady Lisle, respec-
tively, whereas his interlocutors stay detached and limit their answers to the necessary
minimum – another instance of imbalance of power. On the other hand, in (15) not
only are the answers given to the judge by Mrs. Ireland relatively detached both from
the present context (i.e. the court), and the past context (i.e. the events to which both
interlocutors are referring), but also the judge’s questions are hardly emotional. One can
 Barbara Kryk-Kastovsky

only speculate that the differences might heave had to do with idiosyncratic questioning
styles of the two judges, Judge Jeffreys obviously standing out in a negative way.

3.3.2 Yes/no questions

Apart from wh-questions another main examination technique employed in court are
yes/no questions. Danet (1980) calls them coercive questions which, along with de-
clarative questions, are more prevalent when serious offences are involved. Luchjen-
broers (1997: 482) emphasizes the maximal control that the interrogators have over
the contents of witness replies with what she calls “restrictive yes/no questions”. Addi-
tional evidence comes from Hiltunen (1996: 26ff) who labels questions which give the
accused two alternatives and ask him/her to supply the correct information “alterna-
tive yes/no questions”. Similarly Kurzon points out that polar interrogatives are nor-
mally used for unequivocal answers; however they may result in the asking of leading
questions, which presuppose the desired answer (cf. Kurzon 1995: 59). In sum, since
yes/no questions are linked to cases where serious offences are at stake, they can be
expected in cross-examinations and are certainly among the most obvious instru-
ments of power in court. (16) below contains a few such questions which together with
appropriate answers, create what Jucker & Taavitsainen call ‘pragmatic space’, an ana-
lytical tool in speech act analysis analogous to the semantic field. Following Jucker &
Taavitsainen (2000: 74), speech acts will be treated here as fuzzy concepts which show
both synchronic and diachronic variation in the pragmatic space, thus, they should
not be considered in isolation, but in relation to other speech acts. In (16) we can dis-
tinguish several combinations of acts, which for the sake of convenience, have been
divided into individual utterances given under (16a)–(16e) below:
(16) a. L. C. J. And you are sure he went out of Town the Saturday after?
Mrs. A. Ireland. Yes, I am sure he went out of Town then; for I asked him,
why he would go on a Saturday? And he told me, he would go but to
Standen that Night.
b. Mr. Sol. Gen. And that does hold, according to the Computation, to be the
3d of August .
c. Oates. My Lord, she is not positive in this, that he went out of Town the 3d
of August.
d. Mr. At. Gen. Yes, but she is; for she says, that she was the Wednesday be-
fore (which was St. Ignatius’s Day with him a little way out of Town.
e. L. C. J. And that it was Saturday after he went out of Town; and she gives
the Reason, that she entered into a Discourse with him, why he would go
on Saturday? And he made that Answer which you hear.
If we look at the consecutive turns of each participant of the discourse, (16a) is an ex-
ample of direct questioning by a yes/no question, which aptly illustrates the appropri-
ateness of the above-mentioned labels coined by linguists for this type of construction.
It is obvious how by virtue of its form the judge’s question forces Mrs. Ireland to an
Power in Early Modern English courtroom discourse 

answer where the choice is restricted to two alternatives only. Additionally, its form
starting with the Are you sure? clause casts doubt on the truth of the whole proposi-
tion. However, it does not confuse Mrs. Ireland, who again seems to be in control of
her emotions. Mrs. Ireland’s cooperation is additionally enhanced by her reply to the
entire question, including the Are you sure? clause. She might have done so since under
duress created by the power imbalance she decided to conform to the rules of court
investigation and spoke exactly to the point, (cf. Lakoff 1990: 90).
In (16b) Solicitor General’s argumentation is clearly performed from the position
of power since in his strongly affirmative statement he employs the emphatic do, which
has the intended perlocutionary effect of convincing the court that the witness is tell-
ing the truth. According to Searle (1969: 25), convincing someone is one of the perlo-
cutionary effects of the act of arguing.
The truth of (16b) is suspended by the defendant Oates in (16c). From his position
of power he casts doubt on the truth of Mrs. Ireland’s statement by questioning the
degree of her commitment to its contents. Oates’ subversive strategy is quite transpar-
ent here since, if he could prove that Mrs. Ireland had not told the truth, it would obvi-
ously be to his advantage in his double role of the defence and the defendant.
(16d) is a denial of the previous denial (16c) by means of an emphatic statement
geared at persuading the court that the witness is, after all, telling the truth. Once
again, it is the Attorney General who uses his position of power to direct the proceed-
ings towards the ultimate goal of the prosecution, i.e. proving that Mrs. Ireland is tell-
ing the truth, whereas Oates is lying and is thus guilty of perjury.
The series of illocutions and perlocutions is crowned by a final confirmation based
on additional evidence uttered by Lord Chief Justice, who appeals to Oates’ reason and
even his sense of hearing. Since the judge is the highest and most powerful instance in
court, the effect of his emphatic statement amounts to trying to convince the jury and
the audience that the witness is telling the truth and that, consequently, the defendant
Oates is lying.

3.4 A complex case: “Speech act network”

The next example of an examination, given below as (17), consists of the interrogators’
questions interwoven with the reactions which they cause in the addressee. I will call
this kind of complex discourse which combines interrelated illocutions and their cor-
responding perlocutions “a speech act network”:
(17) L. C. J. Are you sure he staid all Night?
Mrs. Duddle. I am sure he staid but one Night.
L. C. J. But what say you to that, Mr. Attorney? this Witness contradicts the
Mr. Just. Withins. Ay, plainly.
L. C. J. But mind my Question, Woman.
 Barbara Kryk-Kastovsky

Mrs. Duddle. Yes, my Lord.

L. C. J. Did he come home that Night he went on the Recreation?
Mrs. Duddle. I do not know.
L. C. J. But just now, you swore he staid out all Night?
Mrs. Duddle. No, my Lord.
L. C. J. Yes, but you did though; prithee mind what thou art about.
Mrs. Duddle. I do not say he, but I am sure his Sister and the Company staid
out that Night. I remember very well, he went the third Day after, which was
Saturday . And Mr. Jennison came to ask for him three Weeks after; and there
was a Person of Quality with him in the Coach; I think it was Sir Miles
Wharton. And he asking for him, they gave him an account, that they had not
heard from him since he went; which was then three Weeks after he was gone.
And I remember well, he did not come to Town again till a Fortnight before
L. C. J. How can you tell that?
Mrs. Duddle. My Lord, I can tell it very well: For I was almost every Night in
the Room where he used to lie; and there lay a Gentlewoman there, that I
L. C. J. What was her Name?
Mrs. Duddle. Mrs. Eagleston.
L. C. J. How come she to lie there?
Mrs. Duddle. Her Maid fell sick, and she chang’d her own Chamber, and lay
there all the time he was out of Town.
Oates. My Lord, is this good Evidence?
L. C. J. Ay, why not?
Oates. My Lord, I think she contradicts the other Witness: For she says he lay
out two Nights.
L. C. J. No, there you are mistaken too. (TO)
Notice that (17) starts with a question (a request/order for information) issued by a
disgruntled Lord Chief Justice. Due to its form (opening with the clause are you sure)
the question casts doubt on the truth of Mrs. Duddle’s previous deposition and signals
the judge’s position of power. The perlocutions resulting from this simple question
could include any of the following effects of the relation between the powerful and the
powerless in the courtroom context: eliciting an answer, scaring the witness, or even
forcing her to change the deposition. Here the judge’s order for information is followed
by Mrs. Duddle’s straightforward answer whose truth he immediately challenges. In
another demonstration of his power the judge accuses Mrs Duddle of contradicting
the testimony presented by another witness. This does not, however, have the intended
perlocutionary effect, since instead of changing her story, Mrs. Duddle goes on with it.
Interestingly, at this point Oates comes up with a meta-comment which repeats the
judge’s remark that Mrs. Duddle’s deposition contradicts that of another witness.
Power in Early Modern English courtroom discourse 

However, Oates’ meta-comment misfires since it is rejected by Lord Chief Justice.

Once again, it is the powerful position of the judge that allows him to control the flow
of the discourse. Despite Oates’ attempts to cast doubt on the truth of Mrs. Duddle’s
deposition, it is the judge’s statement that will influence the court in assessing the cred-
ibility of the witness’s deposition. Notice that a court decision, as opposed to an opin-
ion of a private person, had legal consequences, both for the witness (if we consider the
punishment in case she had been lying) and for the defendant Oates (who might have
been found guilty). In contrast, an analogous discourse outside the courtroom lacks
this institutional framework so that a lie would have just moral consequences and pos-
sibly be subject to repair.4
The analysis of (17) has corroborated my argument that the interpretation of
speech act networks which occur in old written documents cannot be achieved without
recourse to the knowledge of the socio-political conditions of a given period. The above
excerpt from The Trial of Titus Oates is an example of questioning a witness. Another
excerpt coming from The Trial of Lady Alice Lisle illustrates my point even more viv-
idly. It contains the final interrogation of the defendant, where the speech act network
reflects the judge’s verbal cruelty (an epitome of power abuse) in an effort to force the
defendant to change her testimony. If Lady Alice Lisle had confessed that she did har-
bour the rebels, she would have pleaded guilty of high treason. Since it is an excep-
tional case of a court examination, it has been included here to conclude my analysis:
(18) L. C. J. Have you any more to say for yourself?
Lisle. My Lord, I came but five days before this into the Country -
L. C. J. Nay, I cannot tell when you came into the Country, nor I do not care;
it seems you came time enough to harbour Rebels.
Lisle. I staid in London till all the Rebellion was past and over; and I never ut-
tered a good Word for the Rebels, nor ever harbour’d so much as a good Wish
for them in my Mind: I know the King is my Sovereign, and I know my Duty
to him, and if I would have ventured my Life for any thing, it should have been
to serve him, I know it is his due, and I owed all I had in the World to him: But
tho’ I could not fight for him my self, my Son did; he was actually in Arms on
the King’s side in this Business; I instructed him always in Loyalty, and sent
him thither; it was I that bred him up to fight for the King.
L. C. J. Well, have you done?
Lisle. Yes, my Lord.
L. C. J. Have you a mind to say any thing more?
Lisle. No, my Lord.

4. Compare the argument in a study on the semantic and pragmatic aspects of legal terminol-
ogy, especially focussing on the legal definition of the term ‘lie’ and a layperson’s understanding
of it, (cf. Kastovsky and Kryk-Kastovsky 2008).
 Barbara Kryk-Kastovsky

The speech act network in (18) starts with a question from Lord Chief Justice which is
not a request for information, but rather a meta-question (directed both at the defen-
dant and the audience) whether, under the serious circumstances of being accused of
high treason, Lady Alice Lisle still has something to say for herself (the presupposition
being that she does not). However, her reply is rejected by the infuriated judge as ir-
relevant to his question, whereupon he produces another meta-comment concerning
his negative attitude to the information offered by her. The meta-comment produces a
transparent relationship between him and the other participants of the trial. His claim
that he does not know and does not want to know when Lady Lisle came to the coun-
try is based on his presupposition that she had harboured the rebels, in other words,
that she is guilty of high treason. At this point Lady Lisle makes another attempt to
defend herself against the judge’s accusation. She tries the impossible and attempts to
undermine the judge’s presupposition by voicing more arguments which are supposed
to demonstrate that she could not have harboured the rebels. To achieve her aim she
employs the following arguments in order to convince the court that:
– since she was absent from the site of the rebellion at the time it happened, she is
not guilty;
– since she didn't support the rebels, she is not guilty
– since she is a loyal servant to the king, her sovereign, she is not guilty
– since her son fought for the king having being brought up by her in full loyalty to
the crown, she is not guilty.
The interrogation ends with Judge Jeffreys' caustic question whether Lady Lisle has
finished her testimony. It is worth noting how he employs the discourse marker well,
which both marks a new turn and expresses his scepticism towards what the defendant
has just said. This conforms with Culpeper & Kytö’s (2000) observation that well is one
of the verbal indicators of the judge’s control over the situation. Notice that the analyst
would not be in the position to assess the dramatic tension of this excerpt without ap-
propriate socio-historical knowledge both of the epoch and the persons participating
in the trial. Thus, the well-known verbal cruelty of Judge Jeffreys and his predilection
for irony, invective and other face-threatening acts aim at intimidating the defendant
and forcing her to change her testimony. Although, as has been shown by Jucker and
Taavitsainen (2000), this is what happened in the same court trial to one of the wit-
nesses, Dunne, Lady Alice Lisle firmly adhered to her original version. In the light of
these historical facts, the interpretation of this passage is even more dramatic, since the
reader realises, along with the actual participants of the trial, that the defendant’s final
words have the most binding perlocutionary effect, i.e. in the absence of evidence to
the contrary, she turns out to be guilty. Moreover, additional knowledge of the history
of the period allows the analyst to see this series of illocutions and perlocutions as re-
sulting in the gravest of legal consequences (Lady Alice Lisle was pronounced guilty
and was beheaded as soon as six days later).
Power in Early Modern English courtroom discourse 

4. Conclusion

My analysis of the notion of power in Early Modern English court trial records has a
number of implications. Firstly, it has confirmed that yet another analytical tool operat-
ing in synchronic pragmatics, i.e. power, is applicable to the analysis of diachronic lan-
guage data. Even more so, in the very unique context of courtroom discourse, power
seems to have been a more salient phenomenon in 17th century England than it is
nowadays, since in modern courts power is manifested in a much more covert fashion.
Secondly, it follows from my data that the notion of power can be reflected at dif-
ferent language levels, like the organization of discourse (e.g. turn-taking and the use
of discourse markers), socio-pragmatics (e.g. speech acts and their networks used to
achieve a variety of communicative goals relevant to the courtroom context), the lexi-
con (e.g. the choice of neutral vocabulary vs. terms biased with negative connotations),
and syntax (e.g. the form of questions). Since the language data analyzed here come
from written sources, there is unfortunately no possibility of obtaining any supra-seg-
mental information. Especially data concerning the intonation and pitch (and addi-
tional noises like laughter) used by the interrogators during the proceedings could
turn out to be useful in the search for additional, less covert exponents of power.
Finally, my exercise in diachronic pragmatics has demonstrated how essential it is
for the analyst to have access to sufficient information concerning the socio-historical
context of the times when the text was produced, otherwise many facts would remain
puzzling (e.g. the number of turns taken in the court trial proceedings by Titus Oates,
thanks to his powerful position, both in the English society and in the universe of
discourse of his trial). It is this kind of power which still plays a crucial role in various
contemporary discoursal practices, courtroom discourse being one of them.


Primary sources

Hargrave, Francis (ed.). 1730. A Complete Collection of State-trials and Proceedings for High
Treason and other Crimes and Misdemeanours from the Reign of King Richard II to the End
of the Reign of King George I. London: J. Walthoe Sen.

Secondary sources

Arnovick, Leslie. 1999. Diachronic Pragmatics. Amsterdam: Benjamins.

Austin, John. 1962. How to Do Things with Words. Oxford: Clarendon Press.
 Barbara Kryk-Kastovsky

Bertucelli-Papi, Marcella. 2000. Is a diachronic speech act theory possible? Journal of Historical
Pragmatics 1.1., 57–66.
Brown, Roger, and Albert Gilman. 1960. The pronouns of power and solidarity. In: Thomas
Sebeok (ed.). Style in Language. Cambridge, Mass.: MIT Press, 253–276.
Culpeper, Jonathan, and Merja Kytö. 2000. Data in historical pragmatics. Spoken interaction
(re)cast as writing. Journal of Historical Pragmatics 1.2., 175–199.
Danet, Brenda. 1980. Language in the legal process. Law and Society Review 14, 445–464.
Danet, Brenda. 1985. Legal discourse. In: Teun A. van Dijk (ed.). Handbook of Discourse Analy-
sis. Vol. 1. London: Academic Press, 273–291.
Declerk, Renaat. 2006. The Grammar of the English Verb Phrase. Berlin: Mouton de Gruyter.
Dijk, Teun A. van. 1993. Elite Discourse and Racism. Newbury Park: Sage.
Fraser, Bruce. 1990. Perspectives on politeness. Journal of Pragmatics 14, 219–236.
Hiltunen, Risto. 1996. “Tell me, be you a witch?” Questions in the Salem witchcraft trials of
1692. International Journal for the Semiotics of Law IX, 17–37.
Jucker, Andreas. 1997. The discourse marker well in the history of English. English Language and
Linguistics 1, 91–110.
Jucker, Andreas, and Irma Taavitsainen. 2000. Diachronic speech act analysis: Insults from flyt-
ing to flaming. Journal of Historical Pragmatics 1.1., 67–95.
Kastovsky, Dieter, and Barbara Kryk-Kastovsky. 2008. Crimes and misdemeanours: legal termi-
nology vs. lexical semantics and the role of pragmatics. In: Thormalen, Marianne (ed.).
English Now. Selected Papers from the 20th IAUPE Conference in Lund 2007. Lund: Centre
for Languages and Literature, Lund University. 277–291.
Kryk-Kastovsky, Barbara. 2000. Representations of orality in Early Modern English trial re-
cords. Journal of Historical Pragmatics 1.2., 201–230.
Kryk-Kastovsky, Barbara. 2006. Impoliteness in Early Modern English courtroom discourse.
Journal of Historical Pragmatics 7.2., 213–243.
Kurzon, Dennis. 1995. The right of silence: A socio-pragmatic model of interpretation. Journal
of Pragmatics 23, 55–69.
Kurzon, Dennis. 2001. The politeness of judges: American and British judicial behaviour. Jour-
nal of Pragmatics 33, 61–85.
Labov, William. 1994. Principles of Linguistic Change. Internal Factors. Oxford: Blackwell.
Lakoff, Robin. 1990. Talking Power. The Politics of Language. New York: Basic Books.
Levinson, Stephen. 1983. Pragmatics. Cambridge: Cambridge University Press.
Luchjenbroers, June. 1997. “In your own words...”. Questions and answers in a Supreme Court
Trial. Journal of Pragmatics 27, 477–503.
Mey, Jacob. 1993. Pragmatics. An Introduction. Oxford: Blackwell.
Nevalainen, Terttu. 1994. Ladies and gentlemen: The generalisation of titles in Early Modern
English. In: Francisco Fernandez, Miguel Fuster, and Juan Jose Calvo (eds). English His-
torical Linguistics 1992. Amsterdam: Benjamins, 317–327.
Nevalainen, Terttu, and Helena Raumolin-Brunberg. 1995. Constraints on politeness: The prag-
matics of address formulae in Early English correspondence. In: Andreas Jucker (ed.). His-
torical Pragmatics. Amsterdam: Benjamins, 541–601.
Raumolin-Brunberg, Helena. 1996. Historical sociolinguistics. In: Nevalainen, Terttu, and
Helena Raumolin-Brunberg (eds). Sociolinguistics and Language History. Amsterdam:
Rodopi, 11–37.
Power in Early Modern English courtroom discourse 

Romaine, Suzanne. 1982. Socio-historical Linguistics. Cambridge: Cambridge University Press.

Searle, John. 1969. Speech Acts. Cambridge: Cambridge University Press.
Searle, John. 1976. The classification of illocutionary acts. Language in Society 5, 1–24.
Stubbs, Michael. 1983. Discourse Analysis. Oxford: Blackwell.
Watts, Richard. 1992. Linguistic politeness and politic verbal behaviour: Reconsidering claims
for universality. In: Watts Richard, Sachiko Ide, and Konrad Ehlich (eds). Politeness in Lan-
guage. Studies in its History, Theory and Practice. Berlin: Mouton de Gruyter, 43–69.
“I desire to have some tyme to consider of it”
A pragmaphilological approach to refusals
and refutations in Modern-English trials*

Ana E. Martínez-Insua

Enclosed within a larger project on the variation experienced by the English

language in its recent history, the paper explores the use of indirect speech
acts of refutation and denial in speech-related Modern-English texts. The
corpus employed contains trial proceedings from the period 1560–1760, and
the conclusions drawn from its analysis provide an interesting insight into the
possibility of characterising this particular group of illocutionary forces, in the
particular context of trials, during the Modern-English period. The study also
points to the possibility of comparing the uses and values of indirect speech acts
in Modern-English and Present-Day English trials, tracing any possible course
of evolution and/or change.

1. Introduction

The piece of research presented in this paper forms part of a larger project on the
variation experienced by the English language in its recent history. The general aim of
the project is to apply theoretical frameworks of linguistic variation and textual cate-
gorisation to the linguistic description of Modern-English. At a more specific level, the
study presented here focuses on the analysis of the use and formal realisation of indi-
rect speech acts in speech-related Modern-English texts, more specifically, of indirect
refutations, denials and refusals in Modern-English court discourse.
The investigation is based on data drawn from a corpus consisting of Modern-
English dialogues, all of which can be claimed to reflect spoken conversation in some
way. The reason why the study is based on dialogic texts is that dialogue, by definition,
involves interaction, and indirect speech acts seem to play an important role in the
dynamics that are implemented in face-to-face interaction (see Culpeper & Kytö 1999:

* The research reported has been funded by the Spanish and the Galician Ministries of Edu-
cation and Science, grant numbers HUM2005-02351/FILO and 2006/XA-133, respectively.
These grants are hereby gratefully acknowledged.
 Ana E. Martínez-Insua

294). At the same time, the adequate interpretation of indirect speech acts is usually
favoured by factors that are closer to orality and to the language of immediacy than to
literacy and to the language of distance (see Jucker 2004: 13 on the orality-literacy
scale). In this line, factors such as the necessary contextualization, which is crucial for
the addressee’s capacity to infer and recover the intended meaning of the utterance,
turn out to be essential for the interpretation of indirect speech acts.
Historical pragmatics and the subsequent pragmatic approach to historical lin-
guistic data have been gaining force in the course of the last decade, since Jucker’s
groundbreaking Introduction to Historical Pragmatics (1995) established the frame-
work for classifying historical pragmatic research. Nonetheless, not much investiga-
tion has been carried out about indirect speech acts in the Modern-English period,
which makes the present study particularly attractive and worth undertaking.1
From this point of departure, the study is carried out keeping in mind that his-
torical explanations of any type cannot be seen as monolithic enterprises and that
the various aspects of dialogue forms may very well call for separate explanations
(see Fritz 1995). Heed is also paid to the need to carry out speech act analysis against
the background of the pragmaphilological framework,2 knowing that the pragmatic
basis for the texts of a given period (Modern-English, in this case) and the condi-
tions that govern communication between authors and their audience may be fun-
damentally different from those prevailing in modern times. In Jacobs and Jucker’s
(1995: 21) words,
The communicative status as an implicit speech act and thus the degree of conven-
tionalisation can be regarded as conditioned by history. Again, in order to under-
stand a speaker’ intention expressed in the speech act, it is important to know as
precisely as possible the historical circumstances of the utterance. For example, in
former times communicators used implicit speech acts to avoid the use of explicit
speech acts that were morally unacceptable or taboo (cf. Hartmann, 1977: 5).

2. Historical pragmatics and pragmaphilology

Being a relatively recent subfield of historical linguistics, historical pragmatics has

been defined as the study of historical data from a pragmatic perspective, the dia-
chronic study of pragmatic elements or the study of language change from a pragmatic

1. There are recent speech act analyses such as Kohnen’s (2007) and Taavitsainen & Jucker’s
(2007), which assess the possible connections between textual characterisation and speech act
analysis. Even if both refer to the need to bear in mind that speech acts may be indirect, they
study direct (mostly performative and descriptive) uses of speech act verbs exclusively.
2. Pragmaphilology, as defined by Carroll et al. (2003: 2), “is primarily a matter of synchronic
research, taking into consideration the contextual aspects of historical texts in which past stages
of the language are preserved”.
Refusals and refutations in Modern-English trials 

point of view (see Jucker 2000: 90). Being a pragmatic discipline, it aims at taking into
consideration the social situation in which linguistic phenomena take place, the par-
ticipants in the linguistic interaction, their social relationships, and the cognitive pro-
cesses they go through while producing and interpreting language (see Arnovick 1999:
9–14 for a brief account of the origins and scope of historical pragmatics). Knowing
that pragmatics tends to rely mostly on spoken data (considered to be more spontane-
ous and, therefore, a more direct reflection of the linguistic competence than written
language), it may appear to be impossible to approach historical data from a prag-
matic perspective, except for the very recent past (as all historical data are written
data). As a matter of fact, “[a]ny attempt to write a corpus-based illocutionary history
is faced with basic problems involving the methodology of historical pragmatics and
the design and use of historical corpora” (Kohnen 2004: 237).
Nevertheless, there is general agreement (see Jucker 2000: 90) on the fact that the
distinction between written and spoken language should not be seen as a clear-cut
dichotomy and one should be aware of the fact that samples of the language of imme-
diacy may be found in historical written sources.3 The fact that historical data have
only survived in the written code does not imply that such data cannot tell us anything
about the spoken language of the past. In the particular case of trial proceedings and
court records, for instance, what we face is a clear attempt to be an exact reproduction
of what people actually said in such situations. They are therefore, “likely to show more
features of orality even if the renderings are not one hundred percent faithful to the
spoken word in court” (Jucker 2000: 14).
Several approaches can be distinguished within the wide field of historical prag-
matics, one of them being the above mentioned Pragmaphilology, an approach “de-
voted to particular historical texts which are studied from a pragmatic perspective”
(Jucker 2000: 91). Pragmaphilological research is synchronic in that it focuses on one
particular stage of the language, and it is in this sense that the study presented here
might be located within such a framework. Although it aims at paving the ground for
a fully diachronic study of the use and realisation of indirect speech acts in the recent
history of English legal speech, it is, at this initial stage, fully synchronic and concen-
trates exclusively on the Modern-English period.
Space constraints do not allow us to go into details about most of the contextual
aspects that inevitably influence and condition the historical texts under analysis. Nev-
ertheless, bearing in mind that the types of documents analysed for this study are
“shaped by the social and political circumstances of the period in question” (Hiltunen
2002: 17), as much of this information as possible was taken into account when the

3. For further comments on the spoken-written distinction and the orality-literacy scale and
the blurring of orality and literacy for historical texts, see Carroll et al. (2003: 5–6) and Jucker
(2004: 13).
 Ana E. Martínez-Insua

texts at issue were studied.4 In line with this, the fact that we are dealing with a particu-
lar speech event (i.e. trials)5 could not be overlooked during the analysis of the texts.

3. Speech act theory and indirect speech acts

It does not seem necessary to develop, at this point, a full argumentation on the gen-
eral assumption that, whenever we use language, we do not simply report some state
of affairs but rather perform acts through words. Suffice it to say that there is general
agreement on the fact that speech acts are a central part of pragmatics, which is in turn
concerned with the meanings given to text elements by the participants in the dis-
course. As initially proposed by Austin (1962) in his seminal work How to do things
with words, our utterances are speech acts in that they are actions performed by saying
something. That is, they may actually change the interlocutors’ world or state of mind
in some small but significant way. Consequently, speech acts are communicative ac-
tivities that need to be “defined with regard to the intentions of the speaker [author:
their illocutionary force] and the effects achieved on a listener [author: their perlocu-
tionary effect]” (Arnovick 1999: 8). Similarly, it seems sufficient to point out that the
relationship between clause type and illocutionary force is not one-to-one but many-
to-many.6 As pointed out by Cutting (2002: 19ff) and many others, much of the time,
what speakers and writers actually mean is actually not in their words themselves but
in the meaning implied by those words. Such a lack of a predictable link between form
and function needs to be taken into account when working with speech acts, as it in-
evitably affects the methodological procedures to be adopted.7
Besides this, it is generally assumed that the functions of utterances have less to do
with the forms of such utterances than with the contexts in which they are used
(see Geis 1995), and “[s]peech acts are concerned with the speaker’s intention rather
than the content-meaning of the utterance” (Carter & McCarthy 2006: 680). The crucial

4. Notice that the present study refers to the trial section of the CED. Unfortunately, the Socio-
pragmatic Corpus (see Archer, 2005; 2007; Archer & Culpeper, 2003) was not available to the
author when the texts were analysed, which implies that all the annotation that has been added
to some of these texts afterwards could not be taken into account.
5. Speech event is understood here as “an activity in which participants interact via language
in some conventional way to arrive at some outcome” (Yule 1996: 57). “In many ways, it is the
nature of the speech event that determines the interpretation of an utterance as performing a
particular speech act” (Yule 1996: 47–48).
6. The label ‘illocutionary force’ is used in this study so as to cover the rather general concept of
‘intended meaning’ or ‘speaker’s communicative intention’. For a more complex view and charac-
terisation of illocutionary force as a notion made up of seven components, see Vanderveken (1985).
7. See Kohnen (2004) for comments on this lack of a one-to-one coincidence between form
and function as one of the basic methodological problems that corpus-based diachronic study
of speech acts is faced with. See also Cutting (2002) for comments on this fact.
Refusals and refutations in Modern-English trials 

role of the context could not, therefore, be disregarded in a study of this kind. Thus, the
following sections will attempt to evince that even if a refutation is similar to proving
in a passionate and polemical way that something is not true and a denial involves,
above all, saying that something is false, such speech acts may be performed either di-
rectly by using explicit performative verbs8 (e.g. refute, deny, refuse, etc.) or indirectly
(with the speaker’s avoiding to pronounce the corresponding speech act verb).
Classifying utterances into categories of direct and indirect speech acts is not an
easy task, partly because “much of what we say operates on both levels, and utterances
often have more than one of the macrofunctions (‘representative’, ‘commissive’, ‘direc-
tive’, ‘expressive’, and so on)”9 (Cutting 2002: 19). The labels ‘indirect speech act’ and
‘indirectness’ are used here from a basically (Systemic) Functional perspective, based
on the Searlian assumption that speakers using indirect speech acts want to communi-
cate a somehow different meaning from the apparent surface meaning. Dascal’s (1983),
Downing and Locke’s (2006) and Huang’s (2007) characterisations of indirect speech
acts are adopted as the guides for the recognition of the indirect speech acts of the
corpus, assuming that whenever there is an underlying pragmatic meaning, one speech
act is performed through another speech act. Downing and Locke distinguish two
slightly different types of indirectness. On the one hand, indirect speech acts may be
the result of a particular form or clause type being used to convey an illocutionary
force, or intended meaning, that is different from its basic one (with the help of the ap-
propriate intonation).10 On the other hand, indirect speech acts may also appear when
the words we use do not express the full illocutionary force of our intended speech act.
These are cases where addressees recover the intended meaning of the utterances bas-
ing their inference on the assumptions of cooperativeness, truth, relevance and cul-
tural knowledge, as pointed out above. Such a type of indirect speech act can be seen in
situational dialogues like the following one (from Downing and Locke 2006: 179):
(1) A: The door-bell’s ringing.
B: I’m in the bath.
A: Ok. I’ll go.

8. Performative verbs are those speech act verbs which are used by speakers to label speech acts
explicitly, as for instance in he denied that he was involved in any way (Carter and McCarthy 2006:
681). Using Yule’s (1996: 49) terminology, they are Illocutionary Force Indicating Devices.
9. Representative, commissive, directive, expressive are among the macro-classes initially offered
by Searle (1976) as a solution to classifying speech acts. See also Yule’s (1996) or Carter and McCa-
rthy’s (2006) classifications into constatives, directives, commissives, expressives and declarations.
10. As explained by Dascal (1983: 128), an indirect speech act “’means something else’ in the
sense that its final pragmatic interpretation differs from its utterance meaning either in its illo-
cutionary force, or in its propositional content, or else in both”. In Huang’s (2007: 110) words, “if
there is no direct relationship between a sentence type and an illocutionary force, we are faced
with an indirect speech act” [boldface in the original]. See, for instance, Cutting (2002: 19–20)
for some illustrative examples.
 Ana E. Martínez-Insua

Admittedly, the difference between both types of indirectness is mild indeed, as the
second one can be understood as a sub-type of the first one. Besides this, the analysis
of the corpus has been carried out on the basis that it is not always possible to make a
clear-cut distinction between one type of indirect speech act and another. Witness, for
instance, the example below as an illustrative case which may be either a request or an
invitation, or a combination of the two:
(2) Sit over here by my side.
This situation implied the need to pay close attention to the situational context (i.e. trials,
in this case) and the relationship between hearer and speaker (most frequently, prose-
cutors and magistrates vs. defendants), as in different situations or at different points of
a conversation any utterance may take on a different pragmatic force. Far from consid-
ering this a drawback, “[t]his indeterminacy of pragmatic meaning is not, in general, a
disadvantage, as it allows the interlocutors in a situation to negotiate the outcome of
any one utterance as they go along” (Downing and Locke 2006: 179).11

4. The corpus

The texts analysed for this study were drawn from the Corpus of English Dialogues
1560–1760 (hereafter, CED), 2006, a 1.2-million-word computerized corpus compiled
under the supervision of Merja Kytö (Uppsala University) and Jonathan Culpeper
(Lancaster University). The corpus contains texts which include constructed dialogue
(drama, comedy, didactic works and prose fiction) as well as texts which purportedly
record authentic dialogue (trial proceedings and witness depositions) (see Kytö and
Walker 2006).
Aiming to work with a sample of language as close as possible to the speech of the
time, only texts subsumed within the ‘authentic dialogue’ category of the CED were con-
sidered for our study. More specifically, only trial proceedings were analysed because, as
explained by the corpus compilers, “scribal intervention tends to be limited to speaker
identification or to explanatory comments on the proceedings” in them (see Kytö and
Walker 2006: 20).12 The 202,690 words contained in the texts selected for the study are
distributed through the periods covered by the corpus as Table 1 below specifies:

11. See Linell (1998) for further comments on the multifunctionality of discourse construc-
tions, the fact that most utterances fulfil several communicative functions and, at the same time,
some utterances are vague or ambiguous with respect to illocutionary function.
12. Trial records might not tell us exactly how people spoke in Early Modern-English, “but it is
interesting in itself to study how Middle English and Early Modern-English authors chose to
represent dialogues” (Jucker, Fritz & Lebsanft 1999: 16).
Refusals and refutations in Modern-English trials 

Table 1.  Distribution of words in the corpus

Period Word count

1   19,940
2   14,430
3   47,850
4   82,660
5   37,810
Total 202,690

According to Kytö and Walker’s (2006: 20) explanations, trial proceedings contain the
records of court proceedings written down by official scribes or any other observers
who were not otherwise involved in the proceedings. Such records collect the speech,
most frequently, in dialogue form, in question and answer format. Scribal intervention
tends to be limited in them, unlike in the case of witness depositions, for instance,
where the testimony is for the most part recorded as a third person narrative and there
is “considerable intervention on the part of the scribe” (Kytö and Walker 2006: 21).
Being made up of trial proceedings, the corpus exemplifies English legal discourse
(i.e. “language in legal contexts”, as explained by Hiltunen 2002: 3). Consequently, at
least some of the “numerous idiosyncratic linguistic characteristics of which the lan-
guage of the law is notorious” (Hiltunen 2002: 5) may be observed in the language
used in the proceedings under analysis. Space does not permit all of such characteris-
tics to be mentioned here, suffice it to refer to their presence at the levels of lexis, syn-
tax and discourse.13
As will be seen in the following section, trials “are very much about stories or nar-
ratives” (Tiersma 1999: 147) and this obviously affects their structures and illocution-
ary forces. One possible approach to the trial situation involves the metaphor of story-
telling, in that “the party that is able to come up with the more convincing story will

13. As explained by Hiltunen (2002: 5–11), in legal discourse it is frequent to find common
words with uncommon meanings, Latin and formal words and phrases. On certain occasions,
legal discourse becomes unclear and suffers from wordiness as a result of the speaker’s desire to
achieve precision of expression. Similarly, the desire for making the legal text precise and ex-
haustive, as well as each sentence a self-contained and context-free unit usually explains the
syntactic complexity so often observed in legal texts. At the discursive level, the marked legal
character of the texts under analysis is evinced by the abundance of illocutionary forces such as
those of ordering, permission, prohibition, requesting, etc. See also Danet’s (1985) discussion on
the features of legal discourse.
 Ana E. Martínez-Insua

also be the likely winner of the trial” (Hiltunen 2002: 13).14 From this perspective, it
seems interesting to analyse, not only the linguistic features that are likely to support
and undermine credibility (Hiltunen 2002: 13), but also the speech acts that serve
speakers to perform such communicative intentions.

5. Analysis of the corpus

5.1 Methodology

Before referring to the findings, a note on methodological questions seems in order at

this point. Once the texts were selected, the next step to take was to identify and clas-
sify the indirect speech acts they contained. Unlike in other types of corpus studies, it
was not possible to resort here to any kind of automatic search for the retrieval of the
examples to be analysed. In pragmatic studies like this, it is absolutely necessary to
read and consider full texts, not only because “[w]ith speech acts there is no predict-
able link between form and function and consequently no systematic and reliable way
of retrieving relevant forms” (Aijmer & Altenberg 2004: 5), but also because of the
need to pay close attention to contextual factors. Consequently, no other method apart
from microanalysis or close reading was employed for the retrieval of the indirect
speech acts of the texts.
As explained above, the label indirect speech act is applied in this study both to
those cases in which a particular form or clause-type is used to convey an illocutionary
force (or intended meaning) that is different from its basic one, and also to those utter-
ances which do not express the full illocutionary force of the intended speech act. It is
in this second type of indirectness that addressees (and contemporary readers of the
records, by extension) recover the intended meanings of the utterances by relying not
only on the assumptions of cooperativeness, truth, relevance and cultural knowledge,
but also on the common ground provided by the context of the situation.15
The taxonomy employed for the classification of the speech acts of the corpus is
based on Wierzbicka’s (1987) semantic dictionary of English Speech Act Verbs. The
dictionary provides a reasonably complete list of speech act verbs with a relatively high
frequency of use in ordinary English, and a systematic investigation of their meanings.
Being semantic, it strongly emphasises meaning and meaning relations, with the

14. Another main approach to the trial situation implies the consideration of the trial as a
‘battle’ between the parties, which conveys the organization of the discourse in terms of strate-
gies such as ‘defence’, ‘resistance’, ‘aggression’ and ‘attack’. See also Archer (2005) for further
comments on the features of Modern-English courtroom.
15. “[O]n Searle’s view, a speaker’s performing and an addressee’s understanding an indirect
speech act always involves some kind of inference” (Huang 2007: 112).
Refusals and refutations in Modern-English trials 

crucial ambition to show explicitly how a given verb is related to other/s, rather than
to provide unconnected definitions of the different verbs.
In Wierzbicka’s (1987: 11) own words,
The method of identifying meanings in their decomposition into parts is of course
well known in linguistics, and it is not totally alien to the lexicographic tradition.
(...) it is generally assumed, however, that (exhaustive) decomposition is appli-
cable only to a very small fraction of the vocabulary, and that it is certainly not
applicable to complex and elusive concepts such as those encapsulated in speech
act verbs. (...)
Certainly, the meaning of a speech act verb cannot be portrayed by means of just
two or three words; it can, nonetheless, be fully justified in terms of a few simple
sentences. The meanings of two or more different speech act verbs can then be
The dictionary, as an attempt to portray the semantic links between related verbs,
does not list them in alphabetical order, but rather presents them as a network of
various groups of verbs that maintain some kind of semantic relationship. This
structure results from the assumption of the multidimensional nature of the links
between different speech act verbs and the impossibility to reflect all such links
through linear arrangements. Admittedly, the groupings offered in the diction-
ary could be considered as partly arbitrary, just like the order of both individual
verbs and groups of verbs. Far from being undesired, this possible arbitrariness
aims at reflecting that “[t]he whole vocabulary of speech act verbs constitutes a
network of interrelated networks, and there is no way it can be neatly divided
into non-arbitrary classes” (Wierzbicka 1987: 28). Wierzbicka’s categorisation of
speech act verbs and their definitions helped us to detect non-performative and
implicit speech acts of refusal and denial.17

5.2 Discussion

It has already been pointed out that the texts analysed mostly develop as dialogues
between judges/prosecutors/lawyers/plaintiffs and defendants/accused/witnesses.18

16. See also Leech (1983: 207–ff) as a previous source of comments on the problems that may
be found when classifying speech act verbs and their possible solutions.
17. Archer (2005: Appendix 2) provides a refined taxonomy of force sub-categories, partly
based on Stenström’s (1984) and Wierzbicka’s (1987). Following Archer’s terminology, this study
focuses on evades, refusals to answer, disclaims and utterances that do not confirm/oppose
18. For the sake of brevity, no distinction will be made between solicitors and barristers in this
paper and the general term lawyer will be used instead. Similarly, defendant or interrogated will
generally refer to those being accused and/or interrogated, and the labels ‘law language’ and
‘legal language’ will be used interchangeably to refer to the language of the courtroom. See
 Ana E. Martínez-Insua

These dialogues, ultimately aimed at the search for the truth, appear as discovery pro-
cesses that include taking live testimony of potential witnesses (usually called deposi-
tions, in legal language), posing questions (interrogatories), and even requests for
documents or others kinds of evidence. There are certain cases of negotiation and
bargaining between participants. Most commonly, the prosecution offers the accused
less severe penalties if they forgo trial and plead guilty.
In general, it may be said that, in trials, prosecutors and lawyers have the burden
of proving the facts of the story alleged. A primary way to do so is to call witnesses to
testify and, most typically, such witnesses have their own stories to tell, each of them
forming part of the larger narrative that is in dispute. This is something that could be
clearly observed in the corpus under analysis. On many occasions, defendants are al-
lowed to narrate facts and they insert, within their narratives, indirect refutations and
more or less indirect accusations against other people, or even external factors, as re-
sponsible for the crime at issue. It is not uncommon to find that in these narratives
defendants try to evince and maximize other people’s involvement in the facts, as the
responsible agents of the crimes, thus trying to refute allegations and deny their own
responsibility for the criminal facts. In this sense, on many occasions, the accusations
and the refutations inserted in these narratives constitute indirect speech acts, in that
they are not explicitly presented as denials and/or accusations, nor are the correspond-
ing speech act verbs (accuse, refute, deny) used by narrators. Example (3) below illus-
trates this with Grafton’s indirectly refusing to answer by stating lack of courage to take
an oath, and indirectly opposing the accusation by stating that “there was nothinge
spoaken against the King, nor against the State”.
(3) [$KING’S ADVOCATE.$] “I require you, and the Court requireth you, to
take your oath to answere to matters of your owne fact as farre as you know,
and are bound by law.” [$GRAFTON.$] “An oath is a matter of an high nature,
and must not be taken rashlie, I dare not therefore take this oath. We have
done nothinge against the law: it was noe Conventicle: there was nothinge
spoaken against the King, nor against the State, I dare not take the oath, and I
am no ringleader of any to evill.” (B D2THIGHC)19

Archer (2005, Chapter 3) for a detailed account of the roles of judges, prosecutors, witnesses for
the Crown, witnesses for the defence, defendants, attorney general, prosecuting counsel and
defence counsel in Early Modern-English courtroom.
19. The reference code that appears in between brackets at the end of each example contains
information about the text from which the example has been drawn. The encoding characters
contained in some of the examples are those used for coding purposes in the CED. The following
combinations of characters are used for text-level coding:
(^......^): font other than the basic font
(\..........\): foreign language
[{......{]: editorial emendation
Refusals and refutations in Modern-English trials 

As pointed out, in general, the witnesses/defendants in the corpus have a story -or part
of a story– to tell. Yet they are not always permitted to testify in narrative form. Some-
times, the examination of witnesses occurs in a somehow more rigid question-and-
answer format, which at certain points implies a higher degree of pressure on the de-
fendant and a lesser margin of freedom, as Example (4) illustrates:
(4) [$Mr. Serj. (^Gapper.^) $] Well, as you were coming back from fishing, what
happened then?
[$ (^John Egglestone.^) $] By that Time we had got half way in the Meadow,
we saw (^Joseph Redding^) and Mr. (^Annesley^) running, and (^Joseph
Redding^) out run Mr. (^Annesley^), and came up to my Father first.
[$Mr. Serj. (^Gapper.^) $] When they came up what was the first Thing they
[$ (^John Egglestone.^) $] (^Redding^) took my Father by the Collar, and
demanded the Net, and he refused to deliver the Net.
[$ (^Court.^) $] Did you see him take him by the Collar?
[$ (^John Egglestone.^) $] Yes, my Lord.
[$Serj. (^Gapper.^) $] What became of the Net afterwards?
[$ (^John Eggleston.^) $] My Father threw it into the River.
[$Mr. Serjeant (^Gapper.^) $] How far were you from the River then?
[$ (^John Egglestone.^) $] I was about two Yards from the River. After the Net
was thrown into the River, (^Annesley^) came up with his Gun, and swore,
God damn your Blood, deliver your Net, or you are a dead Man; and he fir’d
off before he received any Answer from my Father. (B D5TREDDI)
The different formats of the trials under analysis seem to be connected to certain dif-
ferences at the communicative level. In those trials where the interrogated are allowed

[\..........\]: editorial comment

[^.......^]: corpus compilers’ comment
[$.........$]: running text other than direct speech
[^..^]: text on the line omitted
[^–^]: text in the same sentence omitted
 s explained by the compilers, the main line division of the source text is preserved in the cor-
pus, except in the case of words hyphenated across the line boundary and in cases where the line
width was too long. In those cases where the line division of the source text was not maintained
(e.g. for reasons of clarity), this was noted in a corpus compilers’ comment. The examples pre-
sented here maintain the paragraph division of the CED, and turns of different speakers are
presented as they appear in the corpus. As regards the line width of the CED, it must be taken
into account that the maximum width of a line in the corpus is 70 spaces. If a line in the source
text is longer, the line is cut at 70 or before in the corpus, and a hash (‘#’) inserted to indicate
that the remainder of the line follows below. This convention appears in some of the examples
presented here.
 Ana E. Martínez-Insua

to narrate events, defendants themselves decide how to tell their story and how to pres-
ent the events (which are generally past events the defendants allege to be true while
their truth remains to be established at trial). Quite on the contrary, in trials and inter-
rogatories which adopt the question-and-answer format, control is allocated to ques-
tioners who may resort to leading and goal-driven questions so as to influence the way
in which witnesses remember something and/or force the interrogated to confess. In
these cases, the power of questioning evinces the inequality of participants in terms of
authority and sometimes also in terms of their social status (as a general rule, inter-
rogators have a higher social level than the interrogated do).20
As hinted above, it was observed that, very frequently, apparently simple narra-
tives and descriptions of events (in answer to the judge/lawyer’s questions) are actually
used by the interrogated as accusations as well as self-justifications and/or self-excuses.
The extract below exemplifies the speaker’s attempts on self-justification.
(5) [$ (^Mr. Oates.^) $] My Lord, when (^Mr. Coleman^) was upon his Exami-
nation before the Council-board, he saith, I said there that I # never saw him
before in my Life: I then said I would not swear that I had seen him before in
my Life, because my sight was bad by Candle-light, and Candle-light alters
the sight much, but when I heard him speak I could have sworn it was he,
but it was not then my Business. I cannot see a great way by Candle-light.
In view of the analysis of the corpus, it might be said that these are some of the ways in
which defendants try to refute allegations and deny involvement in crimes when they
are allowed to narrate events. Expectedly enough, when facing accusation, defendants
search for self-justification and quite frequently resort to blaming other people as a
means of excusing themselves. On certain occasions they simply and overtly charge
other people (Examples (6) and (7) below), while on some others they explain the
cause of the offence/crime as provoked by external factors over which they had no
control (as in Example (8), for instance) (see Blum-Kulka, House & Kasper, 1989: 21).
(6) [$ (^Franklin exam.^) Saith, That he provided a white Powder, which was
Poison, for my Lady called it (^Arsenick^); which, as my Lady did afterwards
tell him, was sent to Sir (^Thomas Overbury^) in a Letter.$] (B D2TCARR)
(7) After # this Mr. (^Rookwood^), and Mr. (^Lowick^), and I, had a Meeting at
# (^Red-Lyon-Fields^), where we did discourse about the Matter. I did often
declare against it, That it was so barbarous a thing, that no Man of Honour
almost wou’d be guilty of it, but Major (^Lowick^) answered, That we were to
obey Orders, for sure Sir (^George Barclay^) wou’d not undertake a # thing of
that nature without Orders (B D4TROOKW).

20. Archer (2005) provides an in-depth study of questioning procedures in courtrooms of the
period 1640–1760.
Refusals and refutations in Modern-English trials 

(8) [$ (^L. Chief Just.^) $] Did you not accuse Sir (^George # Wakeman^) by
name, and that he accepted his Reward?
[$ (^Mr. Oates.^) $] Yes, then (^I^) did accuse him by name.
[$ (^L. Chief Just.^) $] Why did you not accuse Mr. (^Coleman^) # by
[$Mr. Oates.$] (^For want of Memory; being disturbed and wearied in sitting
up two nights, I could not give that good account of Mr.^) # Coleman (^which
I did afterwards, when I consulted my Papers; and when I saw Mr.^) Coleman
(^was secured, I had no need to give a farther # Account^) (B D3TCOLEM)
As Example (4) evinced, things may be slightly different when the question-and-an-
swer format is adopted, as it is more rigid and implies a higher degree of pressure on
the defendant. The differences between the question-and-answer format and the nar-
rative one affect not only the testimonies themselves, but also the way in which defen-
dants refute allegations and/or deny involvement.
Yet, not all defendants’ reactions to accusations and charges are complete denials
of responsibility. It has been observed that reactions may range from such blunt deni-
als of liability to self-humbling on the speaker’s part. Even if they are a minority in the
corpus under analysis, confession speech acts such as the one exemplified in (9) below
have been attested in the corpus. Within this reduced group of confession speech acts,
it has been observed that, although on certain occasions defendants choose indirect-
ness so as to admit their responsibility, most commonly, these speech acts are realised
as direct or performative, like in the case of (9):
(9) [$ (^Attorney.^) $] Are you Guilty of adhering to, and comforting the Queen’s
Enemies, and the conveying of the (^French^) Packet and Money that was
sent to relieve them?
[$ (^Hickford.^) $] I confess my self Guilty. (B D1THICKF).
Generally speaking, while the acceptance of responsibility is viewed by the hearer as an
apology, denial of responsibility somehow evinces the speaker’s rejection of the need
to apologize (see Blum-Kulka, House & Kasper 1989: 21). It has been observed in the
corpus that it is precisely in these cases of rejection of the need to apologize that defen-
dants resort more frequently to indirect speech acts of refutation, denial, self-excuse
and self-justification, trying to blame other people and/or provide themselves with
alibis. As explained by Archer (2005: 343), speech acts of denial, refutation, rejection
or disagreement “invite the inference that those holding the opposing opinion to
S [author: the speaker] are/may be lying. For this reason, they also tend to suggest
personal involvement on the part of S”.
Examples (10) and (11) illustrate the speaker’s use of disagreement as a (self-)
 Ana E. Martínez-Insua

(10) [$ (^L. C. J.^) $] How came you to have so many seized in your # house?
[$ (^Colledge.^) $] My Lord, here is (^Elizabeth Hunt^), the # Maid by whom
they were taken in, and who can give you an account of it. I cannot deny but
that they # were in my house; but that I was the Author, or did take them in,
is as great a # mistake as ever was made. Call (^Elizabeth Hunt^) . I do not
know whether (^Curtis^) # be in Town; but this I am confident, he was
Examined before the King and Council, and he # and his Wife denyed it.
(11) After # this Mr. (^Rookwood^), and Mr. (^Lowick^), and I, had a Meeting at
# (^Red-Lyon-Fields^), where we did discourse about the Matter. I did often
declare against it, That it was so barbarous a thing, that no Man of Honour
almost wou’d be guilty of it, but Major (^Lowick^) answered, That we were to
obey Orders, for sure Sir (^George Barclay^) wou’d not undertake a # thing of
that nature without Orders. (B D4TROOKW)
It has also been found that refusal to answer was a frequent reaction on the defendant’s
part (see Table 3 below). Thus, when requested to give a clear answer or take an oath,
rather than openly or performatively refusing to do so, many defendants alleged lack
of knowledge about the facts, lack of courage or capacity to provide an answer, need
for further consideration upon the facts, or even mental derangement. Following Sten-
ström (1984: 77), Archer (2005: 341) labels these as “conscious avoidance manoeu-
vres”, where the speaker’s non-compliance is evinced in his/her avoidance to provide
(part or all of) the knowledge that the other participant/s is/are seeking. Extracts (12)–
(15) below provide illustrative examples of this. In (12) the speaker evades the ques-
tion asking for “some tyme to consider of it”. In this way, the speaker does “not pro-
vide/express ‘yes’, ‘no’, or value of missing variable, and do[es] not answer in such a way
that one can be inferred” (Archer 2005: 341), at the same time that he/she signals his/
her non compliance. Example (13) illustrates how the speaker invokes his ignorance of
the articles so as to evade the question. Something similar is exemplified in (14), where
the speaker alleges mental derangement due to imprisonment, and his subsequent in-
capacity to answer. Finally, (15) is slightly different in that the speaker does not ac-
knowledge the authority of the jury and thus refused to answer.
(12) William Granger of S = t = Margarett’s in Westminster. He being called, THE
BISHOP OF LONDON spake unto him saying,$] “Granger! You look like a
man of fashion: will you take your oath to answere to the articles according to
your knowledg, and as farre as you are bound by law?”
[$GRANGER.$] “I desire to have some tyme to consider of it.”
(13) [$KING’S ADVOCATE speaketh to another of them,$] “You are required to
take your oath to answer the Articles put in against you.” [$PRISONER.$] “I
cannott sweare, because I know them not in certeinty.” (B D2THIGHC)
Refusals and refutations in Modern-English trials 

(14) [$Cler.$] Robert Tichborn, (^hold up thy hand, art thou guilty of this horrid
Act of Treason, or not guilty^) . [$ (^Tich.^) $] My Lord, I have for some
space been kept close # prisoner, and I am altogether unable in the Law to
speak for my self. [$Jud.$] (^You must plead guilty, or not guilty.^)
(15) [$ (^Lord President.^) $] The (^Court^) expects you should give them a final
Answer, their purpose is to adjourn till Monday next, if you do not satisfie
your self, though we do tell you our (^Authority^); we are satisfied with our
(^Authority^), and it is upon (^Gods Authority^) and the (^Kingdoms^),
and that Peace you speak of will be kept in the doing of Justice, and that’s our
present Work. [$ (^The King.^) $] Let me tell you, if you will shew me what
lawful (^Authority^) you have, I shall be satisfied; But that you have said sat-
isfies no reasonable man. (B D3TCHARL)
Tables 2 and 3 below respectively display the general frequency and distribution of
indirect speech acts in the corpus as well as the frequency and distribution of indirect
speech acts of refutation and denial through the periods distinguished in the corpus.
Given that the samples corresponding to the five periods are unequal in size, the fre-
quencies have been normalized and expressed in terms of frequencies per 1,000 words.
The Chi-square test has been applied to the frequencies so as to assess whether the dif-
ferences in their distribution through the periods are significant or not. Notice that

Table 2.  Distribution of indirect speech acts in the corpus (raw frequencies and the inci-
dence counted per 1,000 words)

Period ISA ISA per 1000 words

(1560–1599)   59 2,95
(1600–1639)   81 5,61
(1640–1679) 149 3,11
(1680–1719)   95 1,15
(1720–1760)   30 0,79
Total 414 2,04
Degrees of freedom = 4; chi-square = 185.667; p = ≤ 0.001. The gen-
eral distribution is significant. It remains significant when considering
periods 1 and 2; periods 1, 2 and 3; and periods 3, 4, and 5. The distri-
bution is not significant when periods 4 and 5 are considered.
 Ana E. Martínez-Insua

Table 3.  Distribution of the six most frequent indirect speech acts of refutation and de-
nial in the corpus (raw frequencies and the incidence counted per 1,000 words):

Period/Speech act Refuse Refute/deny Reject Contradict Disagree Request Totals

1   21 21
(1560–1599) 1,05
2   15   16    1 32
(1600–1639) 1,03 1,10 0,06
3   24    9    7    1    1 42
(1640–1679) 0,50 0,20 0,14 0,02 0,02
4   18    2    2 22
(1680–1719) 0,21 0,02 0,02
5    5    5
(1720–1760) 0,13
  62   48    7    3     1     1 122
Total 0,30 0,23 0,03 0,01 0,004 0,004
Degrees of freedom = 4; chi-square = 104.559; p = ≤ 0.001. The general distribution is significant. It remains
significant when considering periods 1 and 2; periods 1, 2 and 3; and periods 3, 4, and 5. The distribution is
not significant when periods 4 and 5 are considered.

caution is needed when analysing and drawing conclusions from the distribution of
the speech acts through the different periods, not only because of the reduced size of
the corpus, but also because the frequency counts and the classification of the SAs it-
self rely ultimately on interpretative decisions. Admittedly, this is, to a certain extent,
inevitable in any study about the pragmatics of historical data.21

6. Concluding remarks and further research

Despite the need to be cautious when drawing conclusions from this exploratory anal-
ysis, some general comments may be made about the tendencies and facts attested.

21. It must be said that most of the indirect speech acts attested in the texts were used by the
defendants and the interrogated, rather than the members of the jury. This is not completely
unexpected given that indirect speech acts constitute one of many forms of politeness. Besides
politeness, factors such as lack of familiarity, the formality of the context or the existence of so-
cial distance have been claimed to favour the use of indirectness (Cutting 2002: 20). In this con-
nection, trials illustrate a context where social distance is usually manifest in terms of social
status, education, class and occupation. Given that social distance may give speakers power and
authority, it is not surprising that those with the less dominant role (i.e. interrogated/defendant)
are those who tend to use indirectness more commonly.
Refusals and refutations in Modern-English trials 

It has been observed that defendants tend to resort to indirect speech acts when
they want to refute accusations and charges and deny liability for crimes. Rather than
explicitly using performative speech act verbs such as refute, deny, refuse, etc., they try
to insert their refutations and denials within apparently neutral discourse (especially
when they are allowed to narrate events). Together with such refutations and denials,
excuses, self-justifications and even accusations were attested in the analysis of indi-
rect speech acts. The use of indirectness in this kind of communicative event seems to
be connected with the interlocutors’ different positions within the social hierarchy as
well as with their different degrees of authority and power in the institutional and/or
social sphere. Thus, indirectness seems to be a good resource for maintaining the po-
liteness required by the situation and the inferiority position occupied by defendants.
The findings also reveal that certain defendants used indirect speech acts so as to
refuse to answer and/or vow at certain points of their trials. Rather than refuse bluntly
to provide the required information, they sometimes alleged lack of knowledge, in-
ability to answer, or even need for further consideration. If indirect refutations and
denials were more easily inserted in narrations, indirect refusals to answer and/or vow
are, in turn, more commonly observed in interrogatives that adopt the question-and-
answer format.
It is important to mention that multifunctionality is detected under certain cir-
cumstances. On certain occasions, the categorisation of the indirect speech acts was
somehow problematic as the same utterances were meant to exonerate speakers and
deny their responsibility at the same time as they blamed other people. This may seem
inconsistent at first sight, but it is not, however, different from what actually happens
in human communication, where the functionality of language becomes evident at all
levels. The procedure in these cases was to include the example at issue within two or
more groups, as required.
Finally, it is necessary to highlight that the frequencies displayed in Table 2 seem
to suggest a marked decline in the usage of indirect speech acts of denial and refutation
as we approach the contemporary period. The chi-square test reveals significant differ-
ences in the general distributions of ISA and ISA of refutation and denial.22 In view of
the results, we may say that the decline observed in the use of the speech acts under
analysis is in fact significant in general terms.23
Further comparative research will be necessary so as to allow us to make such a
decline an extensive conclusion, but we may point out, in any case, that this finding

22. Notice that, as the chi-square test is inappropriate if any expected frequency is below 1 or if
the expected frequency is less than 5 in more than 20% of the cells under analysis, the results
displayed in Tables 2 and 3 refer to the total number of ISA per period and the total number of
ISA of refutation and denial per period, respectively.
23. The Chi-square test result obtained when contrasting the frequencies obtained for the 4th
and the 5th periods evinced that the distribution is not significant, which seems to suggest a
certain homogeneity of these two periods versus the rest of the periods.
 Ana E. Martínez-Insua

actually reveals that the use of direct and indirect speech acts underwent changes in
the course of the 200-year lapse covered by the corpus. It seems plausible to connect
this drop in the frequency of use of indirect speech acts to possible changes in the
format of trials. Apparently, while earlier trials usually presented the narrative format
discussed above, where the interrogated were allowed to narrate their own version of
the events, they seemed to adopt gradually the more rigid question-and-answer for-
mat, which progressively granted the interrogated less freedom to narrate events. Such
evolution at the discursive level of trials might very well be related to the evolution that
has been observed at their syntactic level by Hiltunen (2002). In his analysis of some
characteristic syntactic properties of English law language, and drawing on Gustafs-
son’s (1975) previous investigation on the topic, Hiltunen explores a number of possi-
ble syntactic changes in progress in English law language during a period of some
twenty-five years and concludes that not only do sentences seem to grow shorter, but
also the proportion of dependent clauses inserted in medial position decreases nota-
bly. It seems reasonable to assume that changes may be detected similarly in legal dis-
course as regards the (non-)use of speech acts, and that the motivation for such chang-
es is to be sought in the contextual changes affecting the format of trials.24
The present paper is an exploratory and preliminary account of the analysis of the
use of certain indirect speech acts in Modern-English trials. First of all, as regards the
taxonomy employed for the classification of the indirect speech acts of the corpus, it
may be helpful to take as a point of reference the categorisation of ‘forces’ offered by
Archer (2005) in her book on questions and answers in the English courtroom
(1640–1760), rather than Wierzbicka’s (1987) more general one. Side by side with this,
further research will be necessary in the connection that has been pointed out between
changes in the frequency of indirect speech acts and changes in the structure of trials.
In order to establish some kind of possible parallelism, further examination of trials,
their format, their conventions, etc. will be required.
It might also be interesting to expand the corpus in various ways, enlarging not
only the number of words analysed but also the text types and, ideally, the periods
covered. Witness depositions, for instance, being records of the spoken testimonies of
witnesses, also have prototypically oral traits and properties characteristic of spoken
interaction. In this sense, it will be very interesting to expand the corpus so as to in-
clude Modern-English witness depositions and analyse to what extent the presence of
indirect speech acts in one and the other register differ from each other. Within such a
comparative/contrastive framework, it will be compelling to take into consideration
that in witness depositions the testimonies are for the most part recorded as a third-
person narrative, with a considerable intervention on the part of the scribe (in both the

24. Finally, as regards the reactions (perlocutionary effects) provoked by the indirect speech
acts analysed in this study, we may say that, expectedly enough, as they are produced largely by
interlocutors lacking authority within the context of trials, most of them are unsuccessful in that
they do not attain their intended effect (exculpating defendants, avoiding punishment, etc.).
Refusals and refutations in Modern-English trials 

reproduction of the speech and the inclusion of more or less legal formulae). Thus,
scribal intervention might be regarded as a possible variable generating difference be-
tween both text types.
Undoubtedly, the field of interest might also be expanded and the study might
become focused on the use of (in-)direct speech acts throughout other genres outside
the legal environment (e.g. drama), in an attempt to detect possible differences and/or
similarities among registers. Widening our perspective by including other speech-
based texts in the corpus under analysis would leave us in a much better position to
make generalisations about the use of indirect speech acts in English language. We
should not forget, however, that the genres under analysis should be dialogic and this
kind of texts is not available for non-contemporary periods.
Expanding the corpus towards posterior periods would allow us to analyse the
evolution and possible changes in the use of indirect speech acts in the recent history
of English. This would shed light on the possibility of tracing any possible course of
evolution and/or change, as well as help us to characterise indirect refutations, denials
and refusals through the recent history of English.


Aijmer, Karin & Bengt Altenberg. 2004. “Introduction” In Advances in Corpus Linguistics. Papers
from the 23rd International Conference on English Language Research on Computerized Cor-
pora (ICAME 23), Karin Aijmer and Bengt Altenberg (eds), 1–11. Amsterdam: Rodopi.
Archer, Dawn. 2005. Questions and answers in the English courtroom (1640–1760). A socioprag-
matic analysis. Amsterdam/Philadelphia: John Benjamins Publishing Company.
Archer, Dawn. 2007. “Developing a more detailed picture of the English courtroom (1640–1760):
Data and methodological issues facing historical pragmatics”. In Methods in Historical
Pragmatics, Susan M. Fitzmaurice and Irma Taavitsainen (eds), 185–217. Berlin/New York:
Mouton de Gruyter.
Archer, Dawn and Jonathan Culpeper. 2003. “Sociopragmatic annotation: New directions and
possibilities in historical corpus linguistics”. In Corpus linguistics by the Lune: Studies in
honour of Geoffrey Leech, Andrew Wilson, Paul Rayson and Tony McEnery (eds), 37–58.
Frankfurt: Peter Lang.
Arnovick, Leslie K. 1999. Diachronic Pragmatics. Amsterdam/Philadelphia: John Benjamins
Publishing Company.
Austin, J.L. 1962. How to do things with words. Oxford: Oxford University Press.
Blum-Kulka, Shoshana, Juliane House and Gabriele Kasper. 1989. “Investigating cross-cultural
pragmatics: an introductory overview” In Cross-Cultural Pragmatics: requests and apolo-
gies. Shoshana Blum-Kulka Juliane House & Gabriele Kasper (eds), 1–34. Norwood: Ablex
Publishing Corporation.
Carroll, Ruth, Risto Hiltunen, Matti Peikola, Janne Skaffari, Sanna-Kaisa Tanskanen, Ellen Valle and
Brita Wårvik. 2003. “Introduction” In Discourse Perspectives on English. Risto Hiltunen and
Janne Skaffari (eds), 1–13. Amsterdam/Philadelphia: John Benjamins Publishing Company.
 Ana E. Martínez-Insua

Carter, Ronald and Michael McCarthy. 2006. Cambridge Grammar of English. Cambridge: Cam-
bridge University Press.
Culpeper, Jonathan and Merja Kytö. 1999. “Modifying pragmatic force. Hedges in Early Mod-
ern-English dialogues” In Historical Dialogue Analysis. Andreas H. Jucker, Gerd Fritz and
Franz Lebsanft (eds), 293–312. Amsterdam/Philadelphia: John Benjamins Publishing
Cutting, Joan. 2002. Pragmatics and Discourse. a resource book for students. London and New
York: Routledge.
Danet, Brenda. 1985. “Legal discourse” In Handbook of Discourse Analysis, Volume 1 Disciplines of
discourse. Teun A. Van Dijk (ed.), 273–291. London: Harcourt Brace Jovanovich, Publishers)
Dascal, Marcelo. 1983. Pragmatics and the Philosophy of Mind I. Thought in language. Amsterdam/
Philadelphia: John Benjamins Publishing Company.
Downing, Angela and Philip Locke. 2006. English Grammar. A university course. New York:
Routledge (2nd edition).
Fritz, Gerd. 1995. “Topics in the history of dialogue forms” In Historical Pragmatics. Andreas H.
Jucker (ed.), 469–498. Amsterdam/Philadelphia: John Benjamins Publishing Company.
Geis, Michel L. 1995. Speech acts and conversational interaction. Cambridge: Cambridge Univer-
sity Press.
Gustafsson, Marita. 1975. Some Syntactic Properties of English Law Language. (Publications of
the Department of English, 4). Turku: Department of English.
Hartmann, Claudia. 1977. “Implizite Äusserungen im Rahmen einer historisch-pragmatischen
Analyse” Linguistische Berichte, 48: 47–56.
Hiltunen, Risto. 2002. “Be it enacted...: English legal discourse in focus” Studies in Modern-
English, 18: 1–21.
Huang, Yan. 2007. Pragmatics. Oxford: Oxford University Press.
Jacobs, Andreas and Andreas H. Jucker. 1995. “The historical perspective in pragmatics” In His-
torical Pragmatics. Andreas H. Jucker (ed.), 3–33. Amsterdam/Philadelphia: John Benja-
mins Publishing Company.
Jucker, Andreas H. (ed.). 1995. Historical Pragmatics. Amsterdam/Philadelphia: John Benjamins
Publishing Company.
Jucker, Andreas H. 2000. History of English and English Historical Linguistics. Düsseldorf: Ernst
Klett Verlag.
Jucker, Andreas H. 2004. History of English and English Historical Linguistics. Stuttgart: Klett.
Jucker, Andreas H., Gerd Fritz and Franz Lebsanft. 1999. “Historical Dialogue Analysis. Roots
and traditions in the study of the Romance languages, German and English” In Historical
Dialogue Analysis. Andreas H. Jucker, Gerd Fritz and Franz Lebsanft (eds), 1–33. Amster-
dam/Philadelphia: John Benjamins Publishing Company.
Kytö, Meria and Terry Walker. 2006. Guide to A Corpus of English Dialogues 1560–1760. Stock-
holm: Uppsala Universitet.
Kohnen, Thomas. 2004. “Methodological problems in corpus-based historical pragmatics. The
case of English directives” In Advances in Corpus Linguistics. Papers from the 23rd Interna-
tional Conference on English Language Research on Computerized Corpora (ICAME 23),
Karin Aijmer and Bengt Altenberg (eds), 237–248. Amsterdam: Rodopi.
Kohnen, Thomas. 2007. “Text types and the methodology of diachronic speech act analysis”. In
Methods in Historical Pragmatics, Susan M. Fitzmaurice and Irma Taavitsainen (eds),
139–166. Berlin/New York: Mouton de Gruyter.
Leech, Geoffrey N. 1983. Principles of Pragmatics. London: Longman.
Refusals and refutations in Modern-English trials 

Linell, Per. 1998. Approaching dialogue. Talk, interaction and contexts in dialogical perspectives.
Amsterdam/Philadelphia: John Benjamins Publishing Company.
Searle, John. 1976. “The Classification of Illocutionary Acts”, Language and Society 5: 1–24.
Stenström, Anna-Britta. 1984. Questions and responses in English conversation. Malmö: Liber
Taavitsainen, Irma and Andreas H. Jucker. 2007. “Speech act verbs and speech acts in the his-
tory of English”. In Methods in Historical Pragmatics, Susan M. Fitzmaurice and Irma
Taavitsainen (eds), 107–138. Berlin/New York: Mouton de Gruyter.
Tiersma, Peter Meijes. 1999. Legal Language. Chicago: The University of Chicago Press.
Vanderveken, Daniel. 1985. “What is an illocutionary force?” In Dialogue. An interdisciplinary
approach. Marcelo Dascal and Hubert Cuyckens (eds), 181–204. Amsterdam and
Philadelphia: John Benjamins Publishing Company.
Wierzbicka, Anna. 1987. English speech act verbs. A semantic dictionary. Sydney: Academic
Press Australia.
Yule, George. 1996. Pragmatics. Oxford: Oxford University Press.
Interactive aspects of computer-mediated
‘Disagreement’ in an English
and a German public news group

Sonja Kleinke

This study is placed in the context of current pragmatic research on public news
group Internet communication as well as Watts’ notion of ‘emergent networks’
in the context of linguistic politeness. It investigates the functions and use of
‘disagreement’ in public news group Internet interaction in English and German.
The paper starts with an analysis of the specific frame conditions holding in an
English and a German discussion group including the development of ‘emergent
networks’. Part two deals with the communicative functions of disagreement in
the data and investigates their distribution, strategies of mitigation, preference
organization, and communicative effects. Finally two closed discussion threads
are analyzed in detail, with regard to their developing ‘emergent networks’ and
their interactional effects.

1. Introduction

Public news groups on the Internet have become a widely used medium for personal
electronic interaction on a broad range of topics. Numbers of participants reaching up
to 205,763 in a discussion on the ‘Struggle of Cultures’ in Forum SPIEGEL ONLINE in
February 20061 are highly indicative for their growing relevance as a public medium
for personal interaction. This study aims at gaining further insights into the functions
and use of ‘disagreement’ and cross-cultural aspects of the channel- or medium rules
of public news group Internet interaction in English and German. It is to be placed in
the context of current pragmatic research on the characteristics of Internet communi-
cation, the framing and preference organization of utterances, as well as Watts’ notion
of ‘emergent networks’ in the context of politic and polite behavior. The study is based
on two sets of 60 randomly chosen successive postings in two public news groups of

 Sonja Kleinke

each, BBC-TALK (BBCT) and Forum SPIEGEL ONLINE (SPON), on two topics si-
multaneously selected for discussion between spring 2004 and 2006. The first part of
the paper is devoted to an analysis of the specific frame conditions holding in both
groups including the discourse type ‘discussion’, and the development of the social
links emerging between the interactants during the interaction (‘emergent networks’
cf. Watts 2003) in such discussion boards.
The second part of the paper will deal with the communicative functions of dis-
agreement in the data focusing on ‘(dis)agreement with the propositional content’,
‘meta-pragmatic evaluations of the talk exchange’, and ‘(dis)agreement directed at the
person of the addressee’ (cf. Kohnen 1987 and further developed for Internet discus-
sion boards Kleinke 2007). A closer look will be taken at their distribution, strategies
of mitigation, their preference organization, and their communicative effects. Differ-
ences in each of these areas between the English and the German groups will be dis-
cussed in light of earlier cross-cultural observations on communicative styles in
English and German in general (cf. House 1998: 71), and more specifically in English
and German public Internet news groups (cf. Kleinke 2007 and 2008).
Finally two of the four groups, BBCT II and SPON II, will be analyzed with regard
to ‘emergent networks’ discussing their interactional effects in these groups as well as
the merits and limits of Watts’ approach for this type of interaction.

2. BBCT and SPON: The discussion frames of two public message boards

The notion of frame, referring in an interactional understanding2 “to a sense of what

activity is being engaged in, how speakers mean what they say” (Tannen and Wallat
1993: 60), has long played a crucial role in the linguistic description of ‘agreement’ and
‘disagreement’ – also in the context of preference organization.3 For the analysis of the
structure and functions of ‘disagreement’ in BBCT and SPON, four aspects of the situ-
ational framing of the communicative activities have proved to be predominantly rel-
evant and will be discussed in turn briefly: ‘discussion’ as a special type of interaction
with its own preference rules, some general characteristics of computer-mediated com-
munication relevant for the understanding and interpretation of single instances of
disagreement in these discussion boards, the specific conditions of interaction in each
of the two groups, as well as the emergence of networks between the participants.

2. For a more general treatment of the concept of ‘Frame’ and framing in discourse, see
Tannen (1993).
3. Cf. among others Atkinson and Heritage 1984: 53; Hayaski 1986: 230; Kangasharju 2002:
1452; Jacobs 1987: 234; Pomerantz 1984; Kotthoff 1993: 199ff.; Tannen 2002: 1652; Garcia 1989:
303; Thurmair 1995: 200.
Interactive aspects of computer-mediated communication 

2.1 ‘Discussion’ as a special type of interaction

‘Discussion’ can be characterized as a specific type of confrontational interaction in

which speakers take turns presenting each other, and possibly other participants and
bystanders, with contrasting views on the issue to be dealt with in the talk exchange.4
Linguistic descriptions of ‘disagreement’ have often focused on their specific rules for
‘preference organization’. As opposed to the findings for ordinary conversation in un-
specified contexts, in which ‘agreement’ is characterized as an expected, unmarked
and therefore ‘preferred’ second turn (cf. Sacks 1987: 54–59; Levinson 1983: 307–308),
in confrontational discourse expectations change. Dissent and disagreement turn into
preferred second turns whose absence may be “noticeably absent” in Atkinson and
Heritage’s (1984: 6) sense. According to Hayashi (1996: 230), “agreement may gener-
ally constitute a preferred response to assessment, [however,] it does not in a case like
an argument, where the absence of disagreement is interpreted as a failure to defend
one’s position”. Patterns of disagreement as a ‘preferred second turn’ do not, however,
function quite as straightforwardly as that. Kotthoff (1993: 195) observes that “the
technical turn shape concept interacts with assumptions on normality – and every
form of normality is culturally defined”. The concept of culture is to be interpreted here
in its widest sense, including the concept of ‘frame’, and as such goes hand in hand with
Gruber’s (1996: 52) claim that any kind of propositional dissent between the interlocu-
tors in confrontational discourse in natural conversation also has a negative effect on
their social relationship. This, of course, holds true for the discussion of highly sensi-
tive political topics involving ideological convictions as well, as we find them in BBCT
and SPON. According to Gruber, the ultimate goal in confrontational discourse is not
to find a commonly accepted solution for a problem, but rather to pursue and defend
one’s own position. Thus, by focusing on structural violence such as turn-taking viola-
tions, Gruber (1996: 62) assumes that ‘polite’ behaviour is often suspended in conflict
communication, where ‘impolite’ behaviour may well be felt to be appropriate by the
parties involved. This, however, seems to hold true only to some extent for the discus-
sion groups investigated here. Meta-pragmatic utterances such as in (1) indicate that
the actual framing of the discussion is not agreed upon per se by the participants, but
has to be negotiated during the talk exchange:5

4. Based on Klein 1981, Kotthoff (1989: 188) points out that one characteristic of ‘argumenta-
tive discourse’ is a quaestio that is meant to be solved in a way that participants don’t oppose
each other anymore and the ‘point at issue’ is no longer controversial. (But cf. Gruber 1996:
53–54 (based on Vuchinich 1990), who stresses that negotiating a point under discussion always
also immediately changes the social relations between the participants and enhances the social
prestige of the participant prevailing in a discussion).
5. Cf. also Schütte (2002).
 Sonja Kleinke

(1) ... Please say something substantial? This isn’t the playground (though it does
look like it at times)... [BBCT II – 55]6
What a stupid and offensive post [BBCT I – 9]
Muss man danach noch weiterlesen? [SPON II – 69]
Changes in the framing from a subject-oriented discussion towards a conflict-based
communication of dissent is often experienced by the participants as a process in
which single users want to manipulate the social hierarchy of the group in their own
interests. Due to the specific conditions of computer-mediated communication, the
conditions of interaction in each of the two groups, and the emergence of networks
between the participants, it is in such types of situations that the expression of dis-
agreement gets a reinforcing impact, which, in the end, may lead even to the virtual
mobbing of a single user.7

2.2 Computer-mediated communication and the specific

conditions of interaction in BBCT and SPON

Due to its hybrid nature, computer-mediated communication, with its merging of

characteristics of oral and written communication, formal and informal styles and
personal and public communication (cf. Crystal 2001: 24 ff.; Thaler 2003: 29f.;
Dürscheid 2004: 154; Kleinke 2008), is particularly liable for a broad range of framing
options users can choose from. The linguistic material analyzed here can be placed on
a continuum ranging from conceptual orality to conceptual literacy, which is exploited
by its users in various ways.8 Unlike natural face-to-face communication, computer-
based interaction in public news-groups depends on the medium conditions of writ-
ten interaction and thus lacks non-verbal, para-verbal as well as prosodic signals and,
even and more importantly, opportunities for immediate backchanneling.9 However,
unlike Bazermann (2004: 320f.) states for writing, opportunities for repair are not
quite as limited. Users suspecting that they may have been misunderstood may easily
produce a repair. Technically, public discussion groups on the Internet – sometimes
also called discussion boards or forums – are a special type of electronic discourse.
Participants produce “two-directional texts in which one person using a keyboard
writes language that appears on the sender’s monitor and is transmitted to the monitor
of a recipient, who responds by keyboard” (Davis and Brewer 1997: 1). Thus the text

6. All examples from the four discussion groups have been anonymized for ethical reasons.
Quotations from posts are kept to a minimum, focusing on the relevant passage of text. All pas-
sages chosen here for illustration are direct quotations with no alterations of the spelling, typing
errors, etc. Omissions are marked as usually by three dots in succession.
7. Cf. Kleinke (2007).
8. Cf. Dürscheid 2004: 154; Koch and Oestereicher 1994: 587f.; Kleinke 2008.
9. Cf. Crystal 2001: 24ff.
Interactive aspects of computer-mediated communication 

produced is relatively permanent, preceding discussions even reaching back a few

years can often be read in the archive of the forum. Also repeated readings of single
postings and repeated contemplation of the issues dealt with are possible. Unlike di-
rect face-to-face interaction, the length of a single turn is practically not limited
(even though turns longer than 100 lines are seen as unduly long and violating the
‘netiquette’ in most boards).
SPON and BBCT are public discussion boards accompanying other well-known
broadcast media. In SPON, the topics for discussion are suggested and arranged in
groups by the provider, who also hosts the discussion. The host usually refrains, how-
ever, from selecting single contributions, which are generally not subject to censorship
unless they violate the terms of use. In BBCT users self-select a topic of discussion
within a pre-arranged macro-topic provided by the moderator. In April 2006 BBCT
hosted more than 40 topically different message boards such as Culture, Lifestyle, Lo-
cal, Sport, Teens, TV and radio, etc., in which more than 300 macro-topics were dis-
cussed simultaneously.
SPON currently hosts eleven different central message boards (Diskussion, Politik,
Wirtschaft, Medien, Kultur, Gesellschaft, Wissenschaft, Uni + Bildung, Sport, Treffpunkt
und Schule), each split up into several single discussion groups focusing on a fixed
‘macrotopic’. In the macrotopic ‘politics’ participation reached 153,596 posts sent by
47,754 users in 120 parallel discussion boards on 759 different topics at the time of
BBCT and SPON are both asynchronous boards in which participants do not
communicate with each other simultaneously. Postings are sent to the central web-
based address of the provider, where they are saved and can be viewed and read upon
request by the participants (and the public) in the sequence in which the provider re-
ceived them. A posting can be directed at a specific preceding post or the general
topic of the discussion. Additionally, users can contact each other via E-mail with their
E-mails being publicly visible on the screen within a posting, as an extra separate post-
ing, or hidden from public view and exclusively directed at a specific addressee. Each
contribution shows the date of posting, its topical affiliation, the name of the poster
and, in SPON II and BBCT I, their status as ‘new’ or ‘experienced’ (including the num-
ber of previous postings in BBCT I). The discussion of a single topic can last for sev-
eral months. As in other message boards, postings appear on the screens of all users in
linear chronological order.11 Hardly anything can be said about the specific framing
conditions concerning the users in terms of their social, institutional, cultural or eth-
nic group and gender affiliation. Participants are completely anonymous and often use
nicknames, and even seemingly ‘real names’ can not be traced back to authentic ‘real-
life’ identities in BBCT and SPON.

10. Cutoff date: 12 September 2006, http://iiforum.

11. Cf. Thimm and Ehmer 2000: 221.
 Sonja Kleinke

For this study two times sixty successive contributions have been randomly cho-
sen from BBCT and SPON from spring 2004 and spring 2006. The discussions dealt
with the fight against international terrorism after the terrorist attacks in Spain in
spring 2004 and the public debate on the Danish Mohammed cartoons in spring 2006.
In 2004 BBCT was still operating with shorter topically connected threads. Four such
complete threads on ‘Terrorist Attacks in Spain’ containing 5, 10, 25 and 20 postings
were chosen for analysis in BBCT I. In SPON I the first 60 postings of a debate on
‘Terror in Europa – wie effizient sind die Abwehrstrategien?’, which at that time already
comprised 378 postings, were chosen for analysis. In the second data set taken in
spring 2006, BBCT had turned to longer discussion threads as well. BBCT II contains
the first 60 postings of a discussion on ‘The Muslim Topic’ which had a total of
126 postings by then. In SPON II (‘Kampf der Kulturen II’ – dealing with the same
topic as BBCT), the 60 postings ranging from number 50 to 110 were picked for anal-
ysis.12 During the period of time analyzed, about 30 users were participating each in
BBCT I and in SPON I and II. Only in BBCT II had the number of participants de-
clined to 18. This, however, had no impact on the total number of words contained in
the 60 postings, which marginally decreased from 3420 in BBCT I to 3408 in BBCT II.
All in all, users in SPON I (8760) and II (5504) produced more than twice as many
words in their 120 postings. The average length of the single posts increased only mod-
erately in BBCT II, and it also rose slightly in SPON II, due to two very long contribu-
tions. The greater length of the postings in SPON hints at the fact that these users
orientate their structural layout more towards the written pole of an oral – written
continuum than do the users of BBCT. This tendency is confirmed by the use of con-
ventionalized formal discourse units such as greetings, salutations and signatures. As
stated for English message boards by Crystal (2001: 144), explicit greetings are often
left out in BBCT. In SPON there is a development in the same direction, with more
than half of the postings being introduced by a greeting or other form of address in
SPON I and just eight of them in SPON II. In SPON II – as in BBCT – users are more
often not explicitly addressed. Instead, reference to a preceding turn is achieved by
quotations. With one single exception explicit leave-taking is not used at all in BBCT.
Additionally, in SPON the number of salutations (from 36 to 8) and leave-takings
(19 to 8) decreased, thus contributing to a less personal climate in the group.13

12. Cf.,
13. For other indicators of an orientation towards the written register and a less personal climate
in the same discussion boards, and especially so in SPIEGEL ONLINE, cf. also Kleinke (2008).
Interactive aspects of computer-mediated communication 

2.3 Disagreement in the context of ‘politic behavior’ and the emergence

of networks as framing conditions in BBCT and SPON

Due to its direct impact on the social relationships between interlocutors (cf. Gruber
1996: 52), disagreement plays a very special role in the joint construction of an equilib-
rium of power and distance between the participants in a talk exchange. More recent
approaches to linguistic politeness, such as Watts (2003), have focused their attention
on how ‘politeness’ is understood by ordinary language users. Interlocutors participate
in a talk exchange with a knowledge of what is appropriate in this specific type of in-
teraction. Based on that knowledge they produce what Watts (2003: 144) calls ‘politic
behavior’: “that behaviour, linguistic and non-linguistic, which the participants con-
struct as being appropriate to the ongoing social action”. A sense of what is ‘appropri-
ate’ can be modified and reconstructed in a social interaction, with single participants
imposing their own sense of ‘appropriateness’ on it. ‘Politic behaviour’ is to be distin-
guished from linguistic behavior that is liable to be interpreted as ‘(im)polite’, the latter
referring to “payment ... in excess of what is ordinarily required by the politic behaviour
in the social interaction” (Watts 2003: 152). How does that relate to the ‘framing of
discourse’ in the two discussion groups investigated here? The question is whether and
how users can negotiate ‘appropriateness’ by adjusting their respective individual ‘ha-
bitus’ and, at the same time, constantly create a (possibly changing) sense of appropri-
ateness in their postings? To answer this question another must be raised first: Can
news group communication be seen as a ‘social practice’ in Bourdieu’s and Watts’
sense? Watts (2003: 149) stresses that “[s]ocial practice is carried out within social
fields, and individuals and groups are defined by their relative position in them”. The
relative position of an individual or group in the social field is determined by their
social capital. In ‘real life’ each individual possesses three types of capital (‘material
capital’ such as “money, property, goods, stocks”, ‘cultural capital’ such as “educational
qualifications, skills, knowledge” and ‘social capital’ realized in a network of relation-
ships and by their quality (cf. Watts 2003: 149–150; Bourdieu 1986). Public Internet
news groups are certainly ‘social fields’, but due to the full anonymity of their partici-
pants, they are, partly deliberately, deprived of major aspects of the social context of
‘natural’ social fields. The interlocutors systematically interact with each other without
respective knowledge of their material and social capital outside the discussion board
they engage in. Only their cultural capital (including their linguistic resources) is dis-
played – though in differing degrees depending on the activity of a user in a group
(or other groups users may simultaneously be engaged in). Being completely reduced
to the wording of the postings, negotiation of ‘appropriateness’ in the sense of Watts’
‘politic behavior’, as well as ‘politeness 1’ (the individual’s understanding of what is
(im)polite in a specific situation), becomes difficult. Public Internet discussion groups
are still developing as a field of communication and are often subject to fast and radical
changes. Due to fluctuation and the large numbers of participants in a discussion
thread, individual concepts of ‘appropriateness’ and users’ individual understanding of
 Sonja Kleinke

‘what frame they are in’ constantly shift.14 Negotiations of ‘appropriateness’ in the two
groups investigated proved difficult, also from the perspective of ‘how they are practi-
cally pursued by the participants’. Here the concept of ‘emergent networks’ – Watts’
extension of Milroy’s social network theory – becomes crucial. In the same way as the
exchange of objects between ‘givers’ and ‘receivers’ creates social links between both
parties, “socio-communicative verbal interaction entails the establishment, reestab-
lishment and reproduction of social links between the interactants, which emerge dur-
ing the interaction” (Watts 2003: 154). Watts calls such networks ‘emergent networks’
and distinguishes them from ‘latent networks’ (the social networks already construct-
ed through previous interaction in the sense of ‘objectified structures’). ‘Latent net-
works’ are described as not ‘real’ but ‘rather imagined’ and “may (or may not) influence
the construction of emergent networks” (Watts 2003: 154). In the discussion boards
studied here users usually only have knowledge of their own ‘latent networks’. This may
shape their communicative style and understanding of the frame they are engaged in,
but it is exclusive to the producer of a post and hidden from all other participants in
the same way as the material and social aspects of their ‘capital’ discussed above. Al-
though the participants of a discussion board are all linked via participating in the
same type of activity, their networks are usually uniplex (or ‘open’); i.e. users to whom
one participant forges network links do not normally know each other in ‘real life’.
Regarding the number of participants forged by the links in emergent networks,
Watts (2003: 154–155) distinguishes three types of emergent networks for natural con-
versation: ‘unidirectional networks’ with links established only to one other partici-
pant, ‘ambidirectional networks’ with links connecting two participants reciprocally15,
and ‘multidirectional networks’ in which links are directed to more than one partici-
pant. Because of the public nature of the discussion, unidirectional and ambidirec-
tional links are always at the same time also multidirectional; i.e. any posting sent to
the discussion board, regardless if it is explicitly directed at one single user or referring
to one specific previous post, is at the same time subject to public reception in the
group and maybe also outside the group by guests just ‘lurking’.16 In natural conversa-
tion emergent networks are established by exchanging (linguistic) values. This, how-
ever, is not what normally happens in BBCT and SPON. Most of the links established
between participants are unidirectional. Often one posting directed at the whole group
is referred to explicitly by another user without the initial poster directly replying to it
– at least in due reading time17 (cf. Figure 1 and 2, in which ambidirectional emergent

14. See also Schütte (2002) and Storrer and Waldenberger (1998).
15. Cf. Schütte’s (2004: 63) dense, dialogical interactive networks („dichte, dialogische bzw.
interaktive Netzwerke“).
16. Cf. Clayman (2002).
17. In SPON II the gap between an initial posting of an emerging network reaching out of the
chosen interval and the next posting directly referring to it is 35 pages of text in one of the
four cases.
Interactive aspects of computer-mediated communication 

networks (AENs) are marked in bold print18). Figures 1 and 2 represent the discussion
threads of BBCT II and the relevant passages of SPON II schematically. Links to boxes
with dotted lines indicate that the posting refers to a previous one, outside the interval
of investigation, due to the discussion having already been going on for some time.19
The links pointing to previous postings in Figure 1 construct a further emergent net-
work in addition to the ones displayed in bold.
The letters in the boxes of Figures 1 and 2 represent the users participating in the
discussion at that time. The numbers refer to the number of the posting in the running
The question of ‘when an emergent network has been established in natural con-
versation’ has not yet been resolved (cf. Watts 2003: 155). Not much has been found
out about what types of emergent networks promote a true negotiation of equilibrium.
For this analysis I assume that as soon as at least three postings are related in an ambi-
directional way, an ‘emergent network’, in which ‘appropriateness’ of the mode of inter-
action can be negotiated, has been established. This is the case in all examples dis-
cussed by Watts and it makes sense, since first speakers, confronted with a challenging
turn, possibly directed at the reorganization of the structure and content of the net-
work, must at least take their chance to respond to that challenge. Theoretically speak-
ing, posters, of course, have the chance to reply to any (challenging) response. This,
however, is not the rule in the two message boards studied. At least for SPON, but to
some extent also for BBCT, one can doubt – due to their narrow range – whether the
formation of emergent networks is “in the focus of attention” of the participants as
Watts (2003: 155) observes for natural interaction.20 There are enormous input rates
with about 25 pages of text produced within a few hours in SPON II and less than a day
in BBCT II. Thus, within the two runs of sixty posts studied for each of the four threads,
more complex ambidirectional links and also unidirectional links (produced with a
considerable time lag) can often be hardly detected if a user attempts to follow the
whole discussion thoroughly, which is crucial for the framing of these news boards. In
their latent networks, participants have the tendency to construct their politic behavior
as being in a “state of equilibrium” (Watts 2003: 155).21 This equilibrium is endangered

18. The passage analysed was not taken from the beginning of the discussion, but from 50th
posting onwards. Therefore, it is not always clear which posting a participant refers to. The ques-
tion marks in the boxes in Figure 2 indicate such missing information.
19. In order to check if choosing an intermediate section of the thread has an influence on the
number of the emergent networks constructed, a control schema of the first 50 postings of the
same thread has been produced. With five instances it showed about as many AENs as the sec-
tion presented here.
20. Schütte (2004: 63) also observes the uniplex character of threads in German professional
message boards.
21. For argumentative discourse activities, but outside a ‘politeness’-context, also Kotthoff
(1989: 188) points out that disagreement is not simply accepted by the participants, but starts
attempts to seek agreement.
 Sonja Kleinke



D4 C14 B20

B9 C12 B21




E6 B11 B24 C26


F7 A37 G43 P46 G47

C8 A38 C48

E10 E39 E49 B50 B51

A13 P46 G47 A52 G57

H17 F25 A28 F31 Q58 G61

O44 A60
A33 B35
F16 Q53 G5 Q56

I27 F29 B30 J32


N42 N45


Figure 1.  The network structure in BBCT II

Interactive aspects of computer-mediated communication 

A50 C52
D53 L72
E57 D
F68 W92
J47 J100
H59 F60 J62

/4 J64 C67 P67 Q77 X94 L86 A105

A75 F85 H83
H65 V90
C84 G104
M70 F89
K82 F89 I102
L69 O74
M11 I96

L73 S79

Figure 2.  The relevant passages of the network structure in SPON II (Sent postings, their
senders and order of their arrival)
 Sonja Kleinke

C55 H59 J64

P76 H83 P87
Figure 3.  Emergent Unidirectional and Ambidirectional Networks of User H in SPON II

whenever participants attempt to reconstruct the rules for ‘politic’ behavior in an on-
going talk exchange. Based on Bourdieu, Watts (2003: 155) points out that “[c]hanging
the value and/or structure of network links in an emergent network is thus equivalent
to the exercise of power by a member of the network and is what Bourdieu meant by
the term ‘symbolic violence’”. According to Watts (2003: 156), the construction and
reproduction of emergent networks is essential for (re)establishing ‘equilibrium’ in a
talk exchange. The majority of postings in the groups studied are, however, not linked
in AENs. Thus, in practice, hierarchical positions can less easily be (re)negotiated than
in natural (face-to-face) conversation because they often would have to be simultane-
ously negotiated with several posters. Instead, users mostly concentrate on one or two
replies for further discussion. Figure 3 illustrates this from the perspective of user H in
SPON II, who replies to postings of C and P, simultaneously receiving responsive post-
ings from P, J, C, X and Q.

3. Communicative functions, the sequential order

and preference organization in BBCT and SPON

3.1 Disagreement in discourse

Before looking into the details of how disagreement is used in the two message boards,
I’ll briefly explain how I define the term and what discursive roles can be assigned to it.
I will use ‘disagreement’ in a wider sense, denoting any non-supportive turn in which
speakers express dissent with what has been said in a previous turn within the immedi-
ate talk exchange. Hardly any other conversational move seems to be as ambivalent
when it comes to evaluating its discourse role, both by natural language users and re-
searchers. On the one hand, there is widespread consensus that instances of ‘disagree-
ment’ are face-threatening acts (cf. e.g. already Leech’s 1983: 138 ‘Maxim of Agreement’;
indirectly also Brown and Levinson 1987: 112; Infante and Wigley 1986: 61; Kotthoff
1989: 189; Holtgraves 1997: 225; Gruber 1996: 52). On the other hand, Kotthoff
(1989: 197–198) points at cases in which tough ‘pro’ and ‘con’ is perceived as entertain-
ing and acted out as in a game; implied is a high degree of involvement which clearly
signals respect to the addressee.22 In the context of the media, Luginbühl (1999: 66)

22. Cf. Schiffrin 1984; Blum-Kulka 2002: 1574.

Interactive aspects of computer-mediated communication 

observes personal traits such as verbal aggressiveness as entertaining and highly valued
in politicians. Also, the fact that disagreement in confrontational discourse is seen as
the ‘preferred second turn’ in an adjacency pair (cf. above), shows how deeply practices
of displaying it have been entrenched into our conversational practice or linguistic ‘ha-
bitus’. In the following we shall take a closer look at how users in the public news boards
BBCT and SPON handle the issue of ‘disagreement’ in their discussions.

3.2 Communicative functions and use of disagreement in BBCT and SPON

Starting from a conceptual rather than structural definition of disagreement, the study
looks at any type of linguistic structure in which participants express dissent with what
has been said in a previous turn within the immediate talk exchange. It soon became
apparent that the subject of dissent is by far not always the propositional content of
previous postings. Often open dissent and more or less explicit disagreement is directed
towards the personality of other users (or rather their virtual identities) or the way in
which other members of the group express their opinions. Even though there is no
waterproof distinction between these three types of communicative functions, they will
be dealt with one by one in the following, pointing at areas of overlap where they arise.

3.3 Propositional disagreement

In propositional disagreement, speakers assess the propositional content of a previous

posting in an explicitly negative way.23 This may happen without using any mitigating
devices, such as ‘downgraders’, ‘politeness markers’ (please, excuse me, bitte, Entschul-
digung), ‘hedges’ (kind of, somehow, irgendetwas, irgendwie), ‘understaters’ (a little bit,
bissken, etwas), ‘downtoners’, (just, only, nur, lediglich) or ‘(-) committers’ (think, seem,
obviously, glaube, scheint, offenbar).24
In the data, sometimes very subtle strategies of mitigation can be found. These
include the ones already discussed in Brown and Levinson (1987: 112f.) as sub-strate-
gies of positive politeness (claiming common ground, the repetition of common assump-
tions, claiming common group identity, token agreement, pseudo agreement, white lies,
the hedging of opinions, joking, providing accounts, etc.). In addition, speakers may sim-
ply resort to indirect speech or violations of the Gricean maxims of quality, relevance
or manner exploiting a range of cognitive mechanisms.25

23. Cf. also Kleinke (2007).

24. See e.g. House (1982) and House and Kasper (1981) or, from a different perspective, dis-
cussing them as ‘expressions of procedural meaning’, Watts 2003: 168ff.
25. For a systematical approach to the description of cognitive mechanisms such as ‘profil-
ing’, the ‘windowing of attention’, etc. involved in the exploitation of the Gricean Maxims
cf. Kleinke 2010.
 Sonja Kleinke

Focusing on its interpersonal impact, Kohnen (1987: 201–202) states that propo-
sitional disagreement is also implicitly directed at the personality of the addressee,
since the fact that Speaker 2 (a user responding to a previous post) considers some-
thing as ‘true’ is being negatively evaluated. This, however, is very likely to have an
impact on the link-quality of possibly emerging networks and on the ‘social capital’ a
user or Speaker 2 may possibly gain in the discussion group. Kohnen observes this
mechanism for natural conversation, stating that if Speaker 2 immediately at the be-
ginning of an interchange and in a generalizing way (cf. (2) and (3)) postulates that
Speaker 1’s lines of thought are wrong, other participants will be induced not to check
the plausibility of their successive arguments and to accept Speaker 2’s competence
hastily(cf. Kohnen 1987: 205):
(2) ...About your second point, people can protest about what they like and what
they want. They do have that freedom. [BBCT II – 54]
(3) @ <Name> ..., Irgendetwas stimmt doch hier nicht [SPON I – 2]
Speakers in the two discussion boards basically use three structural patterns of propo-
sitional disagreement. They either state a counter-assumption to a previous post with-
out explicitly negating its propositional content (cf. (2)), use negation particles, nega-
tive prefixes or other lexical means such as no, not, never, nein, nicht, kein, un-, wrong,
falsch, etc. in order to directly negate a statement from a previous posting or its truth
as in (3) – (5) (sometimes we also find a combination of the two as in (6)), or they use
rhetorical questions as a third option:
(4) ... Nicht jeder Deutsche identifiziert sich mit der Bildzeitung ...[SPON II – 56]
(5) ... No. This is not the case, thus showing that Muslims do not even follow the
facts.” [BBCT II – 5]
(6) but it was muslims temselves who brought the whole thing to the fore not the
media who just reported the events [BBCT II – 54]
In rhetorical questions the conceptual mechanism of disagreement is more complex
and more subtle. What is put to question and not agreed with is an implicature or
sometimes even a chain of implicatures (cf. (7) – (8)):
(7) A: Why did Muslims protest in London? The cartoons weren‘t published in
England. [BBCT II – (1)]
Implicature 1:
[Muslims are to protest in London only if the cartoons were published in
Implicature 2:
[The cartoons were not published in England, hence
A is against Muslims protesting in London]
F: Hiya, Is there a law that is against protesting something that happened in
another country? Cos if there was, i didnt know. [BBCT II – (7)]
Interactive aspects of computer-mediated communication 

F indirectly questions Implicature 1 by putting one of the cognitive domains of ‘being

allowed to do something’ (‘the existence of a law against it’) in question and thereby
asking implicitly ‘Why do you think Muslims should only protest in London if the
cartoons were published in England?’ (again using a less central domain of ‘something
being allowed’, namely ‘to believe something is allowed’) in order to express disagree-
ment with Implicature 2.
Direct negation using negation particles is the most direct face-threatening strat-
egy of these three.26 Speakers 2 refer to the same domain as Speakers 1 and contradict
their proposition in the most direct way. However, in BBCT and SPON only about half
of the postings using this strategy are mitigated. Furthermore, in SPON three of the
overall six cases in which users place their disagreement with a preceding turn in ca-
nonical position (one of the markers of a ‘preferred second turn’) occur in this group.
This fits in with the special framing of the discourse and Kotthoff ’s and Luginbühl’s
observations on the entertaining impact of disagreement. As a tendency (not in terms
of statistical significance) it also fits in with cross-cultural aspects discussed by Kotthoff
(1989: 197–198), ascribing a high level of enjoyment in confrontational discourse for
(German) men.
Counter-assumptions and rhetorical questions are less easy to put on a scale of
directness and possible face-threatening impact. The proposition of counter-assump-
tions, when stated in the way Kohnen (1987: 205) describes, directly claims the oppo-
site of what a Speaker 1 has said. From an interactional, speech act- and functional
perspective, however, they leave the speaker free choice as to whether to react or not.
Cast in the syntactic shape of a statement, they do not, in principle, expect a reply or
response. Like direct negations, counter-assumptions in both groups are mitigated only
half of the time on average, though with some variation between the four discussions.
For rhetorical questions, their directness and possible face-threatening impact is
difficult to judge, too. As far as their propositional content is concerned, they are less
direct (cf. (7) above). Nevertheless they are cast in the syntactic form of an interroga-
tive. In natural conversation they often do not genuinely request information or expect
an answer from a potential addressee and are frequently, in terms of their response-
eliciting communicative function, not even truly directed at an addressee.27 On the
other hand, even as responses to a posting of Speaker 1, they may (sometimes also in
addition to the function of ‘disagreement’) initiate or just state a further topic for
discussion,28 and thus at least invite other speakers to provide answers, in addition to
the one often already provided by Speaker 2 (cf. again F’s turn in (7)). In this sense they
are ‘intruding’ on other users’ personal space. Brown and Levinson (1987: 211 and
223f.) handle rhetorical questions as an ‘off record strategy’, leaving it up to the

26. Cf. also Pomerantz 1984: 74.

27. Cf. Meibauer 1986: 169; Ilie 1994: 38.
28. But cf. Ilie (1994: 35) distinguishing rhetorical questions from ‘conducive questions’.
 Sonja Kleinke

addressee to ‘decide how to interpret it’ and hint at their capacity to express criticisms,
but ‘leaving the answer hanging in the air’.
The preferences for either of these strategies are distributed differently in the two
groups (cf. (8) and (9) for examples taken from both groups):
(8) a. ‘direct negation’:
... Das hat <NAME> auch nie behauptet. ... (SPON I – 60)
b. ‘counter assumption’:
Allein, dass man fordert, Menschen muslimischer Religion sollten öffentli-
che Bekenntnisse irgendwelcher Art ablegen, ist eine Frechheit ... (= Quote
from a previous posting)
Es wäre die effektivste Maßnahme schlechthin .... (SPON I – 53)
c. ‘rhetorical question’:
... Ist es zu viel verlangt, auch von denen einen Beitrag zu fordern?
(SPON I – 50)
(9) a. ‘direct negation’:
Sorry to be pedantic the danish newspaper did not take the decision two
weeks ago it took the decision 5 months ago ... (BBCT I – 49)
b. ‘counter assumption’:
..., people can protest about what they like and what they want.
(BBCT I – 16)
c. ‘rhetorical question’:
How many do you want? (BBCT I – 24)
Users in SPON significantly prefer direct negations to counter-assumptions and rhe-
torical questions, with 74 out of 126 instances (corresponding to 59%). In BBCT par-
ticipants use this strategy in only about 43 per cent of all cases. A completely reverse
picture emerges for rhetorical questions. In BBCT they comprise about 35 per cent of
all cases of propositional disagreement. Users in SPON resort to this strategy signifi-
cantly less often, with 11 per cent of all instances. Counter-assumptions occur slightly
more often (roughly 30% of all instances of propositional disagreement) in SPON than
in BBCT, with about 22 per cent. Summarizing, one can say that users of SPON have a
clear preference for direct negations and counter-assumptions. Both strategies are
propositionally oriented and express disagreement fairly directly. However, interac-
tionally speaking, they are ‘self-content’ in that they do not explicitly induce potential
addressees to participate further. By contrast, users of BBCT distribute the three types
more evenly on a scale (cf. Figure 4).
Rhetorical questions are, interactively seen, more open. Though ‘off record’, they
are not fully self-content, but often topic-initiating; thus possibly inducing users to
join the discussion. In this respect, users of both groups seem to polarize their interac-
tive behavior: in SPON clearly towards less interactive strategies (which goes hand in
Interactive aspects of computer-mediated communication 

Direct negation
10 Counter-assumption
0 Rhetorical question

Figure 4.  Propositional disagreement – direct negations, counter-assumptions and rhe-

torical questions in BBCT and SPON

hand with findings on their more systematic exploitation of features of written

discourse)29, and in BBCT more evenly distributed between the highly confrontation-
al strategy of direct negation and the off record and more interactionally oriented
strategy of rhetorical questions30.

3.4 Personal disagreement

Personal disagreement is directed at a negative evaluation of personal traits of the in-

terlocutors. Kohnen (1987: 201ff.) observes for natural conversation that the attitudes
of the addressees are being presented as questionable, wrong and in need of change.
This often happens without any mitigating strategies. Speakers 1 thereby adopt a su-
pervising position and assign members of the group a marginal position in the net-
work by reducing their social capital in claiming deficiencies in their cultural capital
– thereby negatively influencing the quality of their network links. Other than in prop-
ositional and meta-pragmatic disagreement, where the opportunity to address a mem-
ber of the group directly is given but not always exploited, personal disagreement is
almost always directly and explicitly addressed to a specific user indicated by user-
names or direct quotations from a previous posting. Here the fact that the networks
are always multidirectional and postings are potentially read by the whole group and

29. Cf. Kleinke (2008).

30. Totally in tune with the findings on the structure of rhetorical questions in a much larger
study (analysing about 900 postings from BBCT on a broad range of topics), the overwhelming
majority of rhetorical questions were wh- or yes-no questions. In the data studied here, there
was just one rhetorical tag-question in propositional disagreements (cf. Kleinke in preparation).
This suggests that the different frequencies of rhetorical questions do not reflect language related
differences between English and German concerning the use of tag-questions.
 Sonja Kleinke

possible ‘guest readers’ becomes especially relevant to the impact of the face-threat31
(cf. (10) and (11)):
(10) ... Es sind Menschen wie Sie, die die Menschenrechte zu einer frommen Illu-
sion verkommen lassen. [SPON II – 108]
(11) ..., Grow up KID!!! [BBCT II- 33]
As a tendency, personal disagreement is used far more often in the English-speaking
group BBCT with 44 per cent of all instances of disagreement vs. 27 per cent in SPON.32
However, users in BBCT produce instances of personal disagreement significantly
more often (p<0.001) in a mitigated form (73%) than without mitigation (27%), and
they also have the tendency to mitigate more often than users of SPON (54% of all
cases). However, one has to bear in mind that although strategies of mitigation are often
judged to be face-protecting, this effect is not always automatically reached (cf. (12)):
(12) ... Sind Sie in Matrix sitzen geblieben? [SPON II – 89]
Here Speaker 2 uses a rhetorical question to express his or her disagreement. The neg-
ative assessment of (the virtual identity of) Speaker 1 has to be inferred via an implica-
ture. Nevertheless, the incident is a severe threat to Speaker 1’s positive face. This al-
teration in the use of rhetorical questions occurs in both groups. Other than in
propositional disagreement, rhetorical questions expressing personal disagreement
are mostly used not to mitigate, but rather to aggravate the face-threatening impact of
a posting (cf. (12) – (14)):
(13) ..., I bet you don’t have a clue about that either huh? [BBCT II – 20]
(14) ... That enough fact for you? [BBCT II – 15]
The aggravating effect results from these questions being propositionally aimed at par-
ticularly sensitive areas of the positive face of other users. Of course, if those had not
been put in the form of a rhetorical question the positive face of the addressees would
probably have suffered even more. Nevertheless, there seems to be some justification
in not counting these devices as explicitly face-protecting in the normal sense of a
mitigating device.33

3.5 Meta-pragmatic disagreement

Meta-pragmatic disagreement also often implies the implicit negative evaluation of

the poster of a message. According to Kohnen (1987: 209), in instances of meta-prag-
matic disagreement Speakers 1 are implicitly claimed not to be able to express them-

31. See also Kleinke (2007) in the context of linguistic violence.

32. The result of a χ2 –test showed almost significant results with p<3.4.
33. Cf. also Meibauer 1986: 122.
Interactive aspects of computer-mediated communication 

selves correctly or appropriately. Heisler, Vincent and Bergeron (2003: 1623) judge the
potentially face-threatening impact of meta-pragmatic disagreement (involving a neg-
ative judgement of the speaker) as more severe than the one implied in propositional
disagreement. In the two groups studied, it is used to express a negative evaluation of
the general ‘discourse mode’ (cf. (15)), the negative evaluation of the relevance of a
posting criticizing the capability of another user to write to the point (cf. (16) and
(17)), and a negative evaluation of the clarity of a posting, in which other users’ ability
to express themselves appropriately is challenged (cf. (18)):34
(15) ... Eine putzige Abfolge von Erklärungen entschlossener, “aufgeklärter”, kluger
und herzensguter Foristen. ... [SPON II – 71]
... Btw: Why so many abusive posts these days? ... [BBCT I – 1]
(16) <Name>, I don’t care when they were printed – its not the time of the action
that upsets us, it’s the action itself ... [BBCT II – 57]
(17) ... aber das spielt keine Rolle. [SPON II – 93)
(18) ... Who is we? [BBCT II – 20]
Overall, users of both groups resort to this strategy less often than to propositional and
personal disagreement and about to the same extent in both groups (though with some
variation in each group between the first and the second set of data). There is, however,
a tendency to use them in reverse mitigation patterns, with BBCT using them more
often in a mitigated form (in accord with what has been observed for personal dis-
agreement) and participants in SPON using them more often in an unmitigated form.
Users of both groups seem to be aware of their potential face-threatening impact but
handle this potential in opposite ways. More data would have to be studied in order to
see if this tendency hints at an area of cross-cultural differences in the same way, as for
instance the preference for propositional and personal disagreement and their respec-
tive practices of mitigation. Figure 5 summarizes the overall use of the three commu-
nicative types of disagreement in this study.

Prop.D. Pers. D. MPD

Figure 5.  Propositional, personal and meta-pragmatic disagreement in BBCT and SPON

34. Cf. Kleinke (2007).

 Sonja Kleinke

All in all, participants in both groups disagree about equally often, but users in SPON
resort significantly more often to strategies of propositional disagreement, contradicting
the content of previous postings; whereas personal disagreement, complaining about
personal traits of other users or their virtual identities, is more often used in BBCT.
Both groups also deviate in their mitigation patterns. Throughout, the users in
BBCT express their disagreement more often in the form of a rhetorical question,
reaching a significantly higher overall score (p<0.001) in this pattern than users in
SPON. The amount of rhetorical questions differs most in instances of personal dis-
agreement. Though there is the possibility of aggravating a personal disagreement by
a rhetorical question (cf. again (12) – (14)), this technique is not systematically ex-
ploited in the two groups.
The two discussion groups show a reciprocal distribution of structural markers for
dispreferred second turns. In BBCT users mitigate more (67% of all instances), but
place their disagreement more (frequently) to the front of a posting (and significantly
more often – with p<0.001 in a χ2-test – in initial canonical position) than do users in
SPON. The latter mitigate less (52% of all instances), but almost exclusively postpone
instances of disagreement to a non-canonical position in a posting.35

4. The construction of equilibrium in ambidirectional

emergent networks (AENs) in BBCT II and SPON II

Focusing on BBCT II and SPON II, let us now finally look at some striking aspects in
the way users in both groups handle distortions of the discussion mood by renegotiat-
ing equilibrium in AENs. As outlined earlier, in these two discussion boards, and es-
pecially in SPON, true AENs are not systematically built up. Figures 1–2 show six in-
stances of true talk-exchanges with participants reciprocally taking turns several times
in BBCT (26 out of 60 postings, amounting to about 43% of all postings), and just four
of them in SPON (12 out of 60 postings amounting to 20% of all contributions). This
is quite in tune with the different ways users in the two groups exploit techniques of
conceptual orality36, and also with the communicative function of propositional, per-
sonal and meta-pragmatic disagreement (including rhetorical questions) and their
distribution described above. The two groups handle disagreement in these AENs in
slightly different ways. In BBCT the ratio of disagreements per posting (1.56) is the
same in the 26 posts occurring in AENs as in the rest of the discussion. In SPON there
are 1.5 instances per turn, a slight decrease in instances of disagreement in the 12 post-
ings belonging to AENs, compared with 1.76 per turn outside them. This seems to hint
at a greater reluctance of the SPON users to negotiate disagreement directly in virtual

35. For more details on the macro-structure of turns in BBCT and SPON, see Kleinke (2008).
36. Cf. the use of formal discourse markers, deictic expressions, devices of text-cohesion and se-
lected morphological and syntactic structures in BBCT and SPON described in Kleinke (2008).
Interactive aspects of computer-mediated communication 

1:1 interpersonal contact. In both groups, instances of mitigation do not deviate from
the rest of the discussion, though they are a bit more frequent with 0.75 mitigations
per single disagreement in BBCT II than with 0.5 mitigations per instance in SPON II.
Users in SPON use propositional disagreement even more often in the twelve postings
that are part of AENs, with an average of 1.1 per turn as compared to an average of 0.53
instances per turn throughout the discussion. By contrast, personal disagreement in
SPON occurs only 0.1 times per posting in the AENs, but with 0.38 instances per post-
ing almost four times as often in the remaining sections of the data. In BBCT proposi-
tional disagreement is a little underrepresented in the 23 postings belonging to AENs,
with only 0.52 instances per posting on average as compared to 0.7 instances for the
rest of the discussion. Personal disagreement, however, seems to be reinforced in these
networks in BBCT, although also only slightly, with an average of 1 instance per post-
ing compared to 0.8 instances in the remaining posts. Quite in tune with these tenden-
cies, the interactional functions of ambidirectional emerging networks in the two
groups do not seem to be precisely the same.
In the AEN in Figure 6 L 72 starts with a quotation from an earlier posting by X
(X42), in which X asks a topic initializing rhetorical question expressing disagreement
with a topic from the domain of public discourse concerning an issue dealt with in the
immediate talk exchange. L72 takes up X42’s topic, thereby adhering to the positive
face wants of X, establishing common ground in granting X that a question worth-
while thinking about and being discussed has been raised, as well as admitting explic-
itly that he or she cannot provide an answer to it either (‘warum sie das nicht tun, kann
ich nicht sagen’), except offering a ‘qualified guess’ (‘vielleicht aus Bequemlichkeit’).
L72 thereby makes an attempt at least to fulfil X’s desire for information. In addition,

L72 (06.03.2006, 17:21):

Zitat von X42: Wieso gehen unschuldige Männer nicht auf die Strasse um zu zeigen, dass
sie nicht mit den Mishandlungen, welche manche Frauen – die Frauenhäuser sind voll
– erleiden müssen, einverstanden sind?
Warum sie das nicht tun, kann ich nicht sagen (vielleicht Bequemlichkeit), aber dass
sie in Deutschland unter Generalverdacht stehen ist ein Fakt.
X95 (06.03.2006, 22:15):
Zitat von L72: Warum sie das nicht tun, kann ich nicht sagen (vielleicht Bequemlich-
keit), aber dass sie in Deutschland unter Generalverdacht stehen ist ein Fakt.
Wird es deswegen richtig?
L101 (07.03.2006, 08:51):
Zitat von X95: Wird es deswegen richtig?
In der Praxis (und damit realen Leben) ist Recht haben und Recht bekommen auch

Figure 6.  Ambidirectional emerging network between users L and X in SPON II

 Sonja Kleinke

L72 employs the same strategy a second time, by replying to a rhetorical question asked
by X42 in an earlier posting (X42: „...Hier wird eine ganze Religionsgemeinschaft vor-
verurteilt – alles Extremisten – und wer das nicht sein möchte, soll auf die Straße ge-
hen?“). The reply is an agreement („...aber dass Sie in Deutschland unter Generalverdacht
stehen, ist ein Fakt“). However, also adhering to the conditions of the discussion frame of
this specific discussion, the agreement is placed in non-canonical position, disjunct from
the line of thought followed at the beginning of the turn and thus hardly recognizable as
a true agreement anymore. In any case it is one that structurally qualifies as a dispreferred
second turn. X95 replies with another rhetorical question and thus uses not only an off-
record strategy, whose face-threatening impact is relatively weak, but also one that is in-
teractionally oriented in inviting L to keep involved in this emerging ambidirectional
network. The disagreement involved in X95’s rhetorical question (‘wird es deswegen
richtig?’) is based on a chain of implicatures and thereby highly mitigated (cf. (19)):
(19) .... Wird es deswegen richtig? [SPIEGEL ONLINE II – 95]
I1: L believes that ‘unter Generalverdact stehen’ is an excuse for not protesting
against women being victims of violence.
I2: This is not an excuse. L72 is wrong in believing this.
L101 pays X the courtesy of remaining involved in this AEN, despite the implied prop-
ositional disagreement in X95, which, of course, always also indirectly involves a neg-
ative assessment of the speaker (maybe just because it was mitigated and explicitly
invited L to justify his or her beliefs). But this is about as far as L is prepared to go, since
in L101 a highly mitigated personal disagreement follows (“In der Praxis (und damit
realen Leben) ist Recht haben und Recht bekommen auch zweierlei”). L resorts to a
strategy of establishing ‘common ground’ by uttering a ‘common place’ which, how-
ever, again is based on a chain of implicatures (cf. (20)):
(20) ... In der Praxis (und damit im realen Leben) ist Recht haben und Recht bekom-
men auch zweierlei. [SPIEGEL ONLINE II L101]
I1: X does not consider that ‘Recht haben’und ‘Recht bekommen’ is not the
same in real life.
I2: X is naïve in assuming that this could be so in his/her case.
This is the last turn in this AEN. By having admitted that L does not know any more
than X in L72 and having more or less explicitly agreed with X, L has taken on a com-
paratively weak position in the network. In L101 he or she corrects this impression by
offering a turn that can hardly be subject to discussion, thus determining by him-/
herself when this network and equilibrium has been established. All this, however,
happens in a very addressee-oriented, generally friendly atmosphere.37

37. Another strategy users in SPON II have chosen to establish positive relations of closeness
among the participants in these AENs is to develop a separate off-topic thread which both par-
ties follow in addition to pursuing the ongoing topical discussion.
Interactive aspects of computer-mediated communication 

Let us turn now for illustration to an AEN in BBCT in which mutual disagreement
is reinforced by establishing equilibrium of power between B and C, who respectively
use gradually aggravating face-threatening devices (cf. Figure 7).
B2 replies to the first posting in the discussion (A1) using a fused form of an inter-
rogative which merges a counter-assumption with a true request for information. Stat-
ing that B assumed something else than A (“I thought that some British papers re-
printed them in the name of ‘free speech’”) and simultaneously casting this statement
as an interrogative, B reduces the face-threatening impact of his or her counter-as-
sumption by claiming ‘common ground’ with possible addressees who may have
thought otherwise, in not indicating that B’s information is necessarily right. At the

B2 Feb 13, 2006:

Quoted message from A1: Why did Muslims protest in London?
The cartoons weren’t published in England.
God bless
I thought some British papers re-printed them in the name of ‘free speech’?
C5 (Feb 13, 2006):
Quoted message from B2:
I thought some British papers re-printed them in the name of ‘free speech’?
No. This is not the case, thus showing that Muslims do not even follow the facts.
B9 (Feb 13, 2006):
Quoted message from C5:
This is not the case, thus showing that Muslims do not even follow the facts.
I am all Muslims now? I think you ought to get you Bible off the shelf and read about
the good samaritan. They too were demonised by people like you, and one samaritan
turned out to be good and the rest were not.
C12 (Feb 13, 2006):
Quoted message from B9:
... I am all Muslims now? I think you ought to get you Bible off the shelf and read about
the good samaritan. They too were demonised by people like you, and one samaritan
turned out to be good and the rest were not.
Is the work of one danish cartoonist mean he is “all danish people”? Muslims seem to
think so...
B21 (Feb 13, 2006):
Quoted message from B9:
Is the work of one danish cartoonist mean
You seem to think so too.

Figure 7.  Ambidirectional emerging network between users B and C in BBC-TALK II

 Sonja Kleinke

same time, casting this counter-assumption as an interrogative opens it up from a self-

content move towards a device seeking true interaction.
C5 directly responds to this invitation by providing a straightforward answer
which from an interactional perspective fulfils at least two functions. On the one hand
C adheres to B’s positive face wants in providing a response explicitly directed towards
B and offering information B has implicitly asked for (“No. This is not the case” ...). On
the other hand, this response is an unmitigated negation of B’s proposition with the
highest potential face-threatening impact of the structures used systematically in these
message boards (cp. above). However, C disturbs the equilibrium at this point. C not
only poses an unmitigated propositional disagreement, but also produces an extension
which directly and blatantly threatens the face of possible Muslim participants in the
discussion and (presumably intentionally) also the face of B, who appears to be known
by C as a Muslim in the further course of the interaction (“... thus showing that Muslims
do not even follow the facts”). B9 responds to that face-threatening act (FTA) by pos-
ing a chain of FTAs on C’s positive face. Initially B9 poses a personal disagreement,
which, however, is mitigated by being cast in a rhetorical question and thus requiring
a chain of implicatures in order to be identified as such (cf. (21)):
(21) B: I am all Muslims now?
I1: C believes that if one Muslim thinks that something is true all Muslims do so.
I2: This is an unjustified generalization and not appropriate.
I3: C is generalizing in an inappropriate fashion (on this occasion and maybe
also more generally).
In addition, B alludes (likewise) to C’s religious beliefs, questioning their justification
and indicating C’s missing religious education (“...I think you ought to get you Bible off
the shelf and read about the good samaritan. They too were demonised by people like
you, ...”). C12 responds by a semantic and syntactic parallelism, using the offensive
strategy of miming B9 and thereby giving the level of personal confrontation an ad-
ditional spin (“Is the work of one danish cartoonist mean he is “all danish people”?”...).
In addition, another threat at B’s positive face is produced by C’s elaborating on the
face-threatening impact of the rhetorical question him- or herself and explicitly ver-
balizing the implicit meaning of the mimicry. This move also has a negative interac-
tional effect. Users’ answering their rhetorical questions themselves deprives these
questions of their interactional capacities and turns them into self-content moves.
Thereby C here clearly signals to B that he or she feels that equilibrium of power (to his
or her own advantage) has been re-established and that the AEN has been completed,
thus indicating that no further response by B is desired. This, however, is not accepted
by B21, who at least reverses the face-threat insinuating that what has (wrongly) been
stated about B by C is in fact equally true for C himself (“You seem to think so too.”).
This personal disagreement is constructed as a statement and thus indicating from an
interactional perspective that B for his or her part now judges the AEN as completed
and therefore has no further desire to get a reply from C.
Interactive aspects of computer-mediated communication 

Thus, in principle, we can say that in these discussions AENs appear to have a re-
inforcing impact which can work in contrasting directions. On the one hand, they may
be used in order to reassure users of an underlying friendly minded social relationship
which is not meant to be seriously undermined by sometimes rather hostile sequences
in the discussion. On the other hand, they may be used to construct precisely this
harsh and hostile atmosphere.

5. Conclusions

The functions and use of disagreement in these English and German Internet-based
public discussion groups differ on a range of levels, partly cross-patterning among the
groups as well as with observations on the preferred or dispreferred use of disagree-
ment as a second turn in confrontational and non-confrontational discourse. The Eng-
lish group, BBCT, shows a preference for personal disagreement, mitigation and ca-
nonical or preferred placement of disagreement in the postings. The German group,
SPON, prefers propositional disagreement, mitigates less, and delays instances of dis-
agreement. This reconfirms earlier cross-cultural observations on natural and com-
puter-mediated conversation (cf. House 1998, and Kleinke 2007, 2008) attributing an
orientation towards the propositional content, explicitness and directness for German
speakers, and an orientation towards the addressee, implicitness and indirectness also
on the deeper level of the interactional management of disagreement for English
speakers. More data will have to be analyzed in order to find out in the future if, and if
so how, users of different message boards exploit the conversational strategies of con-
structing ambidirectional emergent networks systematically.


I would like to thank Renée Flibotte-Lüskow for her support as a native speaker. All
remaining faults and inconsistencies are, of course, my own responsibility.


Bazerman, C. 2004. “Speech Acts, Genres, and Activity Systems: How Texts organize Activity
and People.” In What Writing Does and How It Does It. An Introduction to Analyzing Texts
and Textual Practices, C. Bazerman and P. Prior (eds), 309–339. Mahwah, NJ/London:
Lawrence Erlbaum.
Blum-Kulka, S., Blondheim, M. and Hacohen, G. 2002. “Traditions of Dispute: From Negotia-
tions of Talmudic Texts to the Arena of Political Discourse in the Media.” Journal of Prag-
matics 34: 1569–1594.
 Sonja Kleinke

Bourdieu, P. 1986. “The Forms of Capital.” In: Handbook of Theory and Research for the Sociology
of Education, J. G. Richardson (ed.), 240–258. Connecticut: Greenwood Press, Inc.
(Originally published as Bourdieu, P. 1983. “Ökonomisches Kapital, kulturelle Kapital, so-
ziales Kapital.” In: Soziale Ungleichheiten (Soziale Welt, Sonderheft 2), R. Kreckel (ed.),
183–198. Göttingen: Otto Schartz & Co. Translated by Richard Nice).
Brown, P. and Levinson, S.D. 1987. Politeness. Cambridge: Cambridge University Press.
Brown, P. and Levinson, S.L. 1987. Politeness. Cambridge: Cambridge University Press.
Clayman, S.E. 2002. “Disagreements and Third Parties: Dilemmas of Neutralism in Panel News
Interviews.” Journal of Pragmatics 34: 1385–1401.
Crystal, D. 2001. Language and the Internet. Cambridge: Cambridge University Press.
Davis, B.H. and Brewer, J.P. 1997. Electronic Discourse. Linguistic Individuals in Virtual Space.
New York: State University of New York Press.
Dürscheid, C. 2004. “Netzsprache – ein neuer Mythos.” In Internetbasierte Kommunikation, M.
Beißwenger, L. Hoffmann and A. Storrer (eds), 141–157. Osnabrück: OBST, Osnabrücker
Beiträge zur Sprachtheorie 68.
Garcia, C. 1989. “Disagreeing and Requesting by Americans and Venezuelans.” Linguistics and
Education 1: 299–322.
Gruber, H. 1996. Streitgespräche. Zur Pragmatik einer Diskursform. Opladen/Wiesbaden:
Westdeutscher Verlag.
Hayashi, T. 1996. “Politeness in Conflict Management: A Conversation Analysis of Dispreferred
Message from a Cognitive Perspective.” Journal of Pragmatics 25: 227–255.
Heisler, T., Vincent, D. and Bergeron, A. 2003. “Evaluative Metadiscoursive Comments and
Face-work in Conversational Discourse.” Journal of Pragmatics 35: 1613–1631.
Heritage, J. and Atkinson, J. M. 1984. “Introduction” In Structures of Social Action. Sudies in
Conversation Analysis, J.M. Atkinson and J. Heritage (eds), 1–15. Cambridge, London, New
York, New Rochelle, Melbourne, Sydney, Paris, Cambridge University Press and Editions
de la Maison des Sciences de l’Homme.
Holtgraves, T. 1997. “YES, BUT ... Positive Politeness in Conversation Arguments.” Journal of
Language and Social Psychology 2: 222–239.
House, J. 1982.“Gambits in deutschen und englischen Alltagsdialogen. Versuch einer pragma-
tisch-kontrastiven Analyse.” Grazer Linguistische Studien 17/18: 110–132.
House, J. 1998. “Kontrastive Pragmatik und interkulturelle Kompetenz im Fremdsprachenun-
terricht.” In Kontrast und Äquivalenz. Beiträge zu Sprachvergleich und Übersetzung, W.
Börner and K. Vogel (eds), 62–83. Tübingen: Narr.
House, J. and Kasper, G. 1981. “Politeness markers in English and German.” In Conversational
Routine. Explorations in Standardized Communication Situations and Pre-patterned Speech,
N. Coulmas (ed.), 157–185. The Hague: Mouton Publishers.
Ilie, C. 1994. What Else Can I Tell You? A Pragmatic Study of English Rhetorical Questions as
Discoursive and Argumentative Acts. Acta Universitatis Stockholmiensis. Stockholm Stud-
ies in English LXXXII: Almquist&Wiksell International.
Infante, D. and Wigley, C.J. 1986. “Verbal Aggressiveness: An Interpersonal Model and Mea-
sure.” Communications Monographs 53: 61–69.
Jacobs, S. 1987. “The Management of Disagreement in Conversation.” In Argumentation: Across
the Lines of Discipline, Eemeren, van F.H. (ed.), 229–239. Dordrecht: Foris Publishers.
Kangasharju, H. 2002. “Alignment in Disagreement: Forming Oppositional Alliances in Com-
mittee Meetings.” Journal of Pragmatics 34: 1447–1471.
Interactive aspects of computer-mediated communication 

Kleinke, S. 2003. “Geschlechtsrelevante Aspekte sprachlicher Unhöflichkeit im Internet.” In

Frauen und Gewalt: Interdisziplinäre Untersuchungen zu geschlechtsgebundener Gewalt in
Theorie und Praxis, A. Hilbig, C. Kajatin and I. Miethe (eds), 197–213. Würzburg.
Kleinke, S. 2007. “Sprachliche Strategien verbaler Ablehnung in öffentlichen Diskussionsforen
des Internets.” In Verletzende Worte. Die Grammatik sprachlicher Missachtung,. S. Kitty
Herrmann, S. Krämer and H. Kuch (eds), Bielefeld: transcript Verlag. 311–336.
Kleinke, S. 2008. “Zur sprachlichen Gestaltung deutscher und englischer Diskussionsbeiträge in
den öffentlichen Internetforen BBC-TALK und Spiegel ONLINE.” In Stil ist überall, aber
wie bekomme ich ihn zu fassen?, Reihe Stil – Kreativität – Variation – Komparation. V. Fuchs
and K. Störl (eds), Frankfurt am Main u. a.: Peter Lang. 101–118.
Kleinke, S. 2010. Speaker activity and Grice’s maxims of conversation at the interface of prag-
matics and cognitive linguistics. Journal of Pragmatics 42: 3345–3366.
Kleinke, S. “Responses to Rhetorical Questions in English and German Public News Groups in
the Internet.” (under review).
Koch, P. and Oesterreicher, W. 1994. “Schriftlichkeit und Sprache.” In Schrift und Schriftlichkeit.
Writing and its Use. Ein Interdisziplinäres Handbuch internationaler Forschung. An Interdis-
ciplinary Handbook of International Research, H. Günther and O. Ludwig (eds), 587–604.
Berlin/New York: de Gruyter.
Kohnen, T. 1987. Zurückweisungen in Diskussionen. Die Konzeption einer Sprechhandlungstheorie
als Basis einer empirisch orientierten Konversationsanalyse. Frankfurt am Main u. a.: Lang.
Kotthoff, H. 1989. “Stilunterschiede in argumentativen Gesprächen oder Zum Geselligkeitswert
von Dissens.” In Stil und Stilisierung, V. Hinnenkamp and M. Selting (eds), 187–202.
Tübingen: Niemeyer.
Kotthoff, H. 1993. “Disagreement and Concession in Disputes: On the Context Sensitivity of
Preference Structures.” Language in Society 22: 193–216.
Leech, G. 1983. Principles of Pragmatics. London and New York: Longman.
Levinson, S. 1983. Pragmatics. Cambridge.
Luginbühl, M. 1999. Gewalt im Gespräch. Verbale Gewalt in politischen Fernsehdiskussionen am
Beispiel der „Arena“. Bern u. a.: Peter Lang.
Meibauer, J. 1986. Rhetorische Fragen. Tübingen: Max Niemeyer Verlag.
Pomerantz, A. 1984. “Agreeing and Disagreeing with Assessments: Some Features of
Preferred/Dispreferred Turn Shapes.” In Structures of Social Action, J.M. Atkinson and J.
Heritage (eds), 57–102. Cambridge and New York: Cambridge, London, New York, New
Rochelle, Melbourne, Sydney, Paris, Cambridge University Press and Editions de la Maison
des Sciences de l’Homme.
Sacks, H. 1987. “On the Preference for Agreement and Contiguity in Sequences of Conversa-
tion.” In Talk and Social Organization, G. Button and J.R.E. Lee (eds), 54–69. Multilingual
Matters LTD. Clevedon.
Schiffrin, D. 1984. “Jewish Argument as Sociability.” Language in Society 13: 311–335.
Schütte, W. 2002. “Normen und Leitvorstellungen im Internet: Wie Teilnehmer/-innen in News-
groups und Mailinglisten den angemessenen Stil aushandeln.” In Soziale Welten und kom-
munikative Stile. Festschrift für Werner Kallmeyer zum 60. Geburtstag, I. Keim and W.
Schütte (eds), 339–362. Tübingen: Gunter Narr.
Schütte, W. 2004. “Diskursstrukturen in fachlichen Mailinglisten: Zwischen Einwegkommunika-
tion und Interaktion.” In Internetbasierte Kommunikation, M. Beißwenger, L. Hoffmann and
A. Storrer (eds), 55–75. Osnabrück: OBST, Osnabrücker Beiträge zur Sprachtheorie 68.
 Sonja Kleinke

Stegbauer, C. 2000. “Begrenzungen und Strukturen internetbasierter Kommunikationsgrup-

pen.” In Soziales im Netz. Sprache, Beziehungen und Kommunikationskultur im Internet, C.
Thimm (ed.), 18–38. Opladen/Wiesbaden: Westdeutscher Verlag.
Storrer, A. and Waldenberger, S. 1998. “Zwischen Grice und Knigge: Die Netiketten im Inter-
net.” In Medium Sprache: Forum Angewandte Linguistik 34, H. Strohner, L. Sichelschmidt,
M. Hielscher (eds), 63–77. Frankfurt am Main u. a.
Tannen, D. (ed.) 1993. Framing in Discourse. New York/Oxford: Oxford University Press.
Tannen, D. 2002. “Agonism in Academic Discourse.” Journal of Pragmatics 34: 1651–1669.
Thaler, V. 2003. Chat-Kommunikation im Spannungsfeld zwischen Oralität und Literalität. Berlin:
Verlag für Wissenschaft und Forschung.
Thimm, C. and Ehmer, H. 2000.“’Wie im richtigen Leben ...’: Soziale Identität und sprachliche
Kommunikation in einer Newsgroup.” In Soziales im Netz. Sprache, Beziehungen und
Kommunikationskultur im Internet, C. Thimm (ed.), 220–239. Opladen/Wiesbaden:
Westdeutscher Verlag.
Vuchinich, S. 1990. “The Sequentil Organisation of Closing in Verbal Family Conflict. In: Con-
flict Talk. Sociolinguistic Investigations of Arguments in Conversations, A. D. Grimshaw
(ed.), 118–139. Cambridge. Cambridge University Press.
Watts, R.J. 2003. Politeness. Cambridge: Cambridge University Press.
‘A little story, for food for thought.......’
Narratives in advice discourse

Loukia Lindholm

The use of narrative in advice discourse has been primarily discussed as an

advice-seeking strategy related to the statement of a problem. This paper
focuses on the functions of narrative in advice-giving messages in the context
of computer-mediated discourse. Drawing on data from asynchronous peer-
to-peer discussions, the analysis shows that advice-givers employ narratives
for argumentative purposes, such as supporting assessments and advice, or for
relational work. The paper also reports on instances of tellings found in the
data that fall under the scope of non-canonical narrative activities which recent
research has brought into focus (Georgakopoulou 2004, 2006).

1. Introduction

Narratives occur in everyday talk and across many different contexts. Telling a story is
used to establish different interactional modes (Quasthoff & Becker 2005) and fulfill a
range of different communicative purposes (Norrick 2000). The importance of contex-
tual particularities that shape the functions and structure of narratives has been em-
phasized in recent research (De Fina 2005; Georgakopoulou 2004; Norrick 2000; Ochs
& Capps 2001).
Advice discourse is an interactive communicative process where one person as-
sumes or is given the role of the advisor and another assumes the role of the advisee
(Hudson 1990). In the case of advice discourse, narratives have been principally dis-
cussed as an advice-seeking strategy related to the statement of a problem (DeCapua
& Dunham 1993; Locher 2006). Advice-seekers often tell stories to provide back-
ground information on a problematic situation they are facing. Although it has been
pointed out that telling a story is employed by both advice-seekers and advice-givers
(see DeCapua & Dunham 1993), the functions of narratives in advice-giving are none-
theless less explored. The fact that studies of advice have almost exclusively focused on
institutional settings may have contributed to this imbalance. In such settings, advi-
sors are experts in a specific domain and the practice of advice-giving sets up a frame
for bringing forward their expertise. Advice-giving delivered by experts is mainly
 Loukia Lindholm

realized through particular communicative strategies such as overt recommendations

and imperatives (Heritage & Sefi 1992; Hudson 1990; Kiuru et al. 2004) or “proposed
actions as alternatives” (Leppänen 1998). Narratives do not appear to be a preferred
advice-giving strategy in institutional settings.
This article focuses on the functions of narratives in the practice of advice-giving in
everyday discourse in an online advice-giving context. The goal is to determine what
narratives do in such a context by examining how advice-givers organize their narratives
within advice-giving messages and how these narratives are structured. The point of
departure is the view of narratives as “contextualized activities” (Quasthoff & Becker 2005)
that are part of the surrounding discourse (Ochs & Capps 2001). At the same time, this
article puts in focus a less researched context of occurrence of the narrative, namely that
of online interaction (see Georgakopoulou 2004; Suzuki & Calzo 2004).

2. The corpus of online discussions

For the exploratory purposes of this study, the material for analysis consists of a se-
lected number of asynchronous electronic discussions which are part of a larger corpus
collected during 2005–2006 from an online advice forum with the aim to investigate
advice discourse. The data comprise discussion threads collected from a publicly avail-
able advice forum on an online site dedicated to parenting, []1. Specifically,
27 discussions comprising 200 messages were extracted from the larger corpus of
219 discussion threads that amount to 1296 messages. These electronic discussions ap-
pear in the form of threads, that is, sequences of responses to an initial message2 posted
by a participant. This website features numerous parenting-related forums and is pro-
filed as a community that offers information, support and advice3 to parents and par-
ents-to-be. The participants, who are in their majority women, seek and give advice on
any general parenting-related topics such as parenting styles, discipline issues and edu-
cation as well as more specific matters pertaining to interracial, interreligious, large or
one-parent families. The regular socialization of the participants as well as their com-
mon interests and goals have given rise to a distinct linguistic repertoire and online

1. For purposes of protecting the online participants’ anonymity, a pseudonym is used in

place of the website’s real name. Message headers and signatures are removed, discussion titles
are paraphrased and all names or nicknames that appear in the examples are altered. All altered
names are contained within brackets [ ].
2. In the advice forum, each message that initiates a discussion appears with a discussion title,
the poster’s screen name, the number of replies to the thread-inititating message and the date
the last reply was originally posted. An additional attribute is the option to place an emoticon
next to the discussion title in order to signal the tone of the message.
3. A disclaimer in the forum states clearly that the advice given in the message boards should
not be considered a substitute for medical attention or diagnosis.
Narratives in advice discourse 

rituals within the online community. This online aggregation of parents and parents-to-
be constitutes what in Wenger’s (1998) terms is defined as a “community of practice”.
The examples of the posted messages have not been edited for spelling or gram-
mar. Wherever necessary, discursive move labels are placed in the examples for ex-
planatory purposes.

3. Method

The analysis, which is essentially qualitative, involved a close reading of the advice-
giving replies to the advice-seeking messages. The methodological toolkit included an
analysis of the discursive moves of the reply messages, the localization of the narratives
within these messages and the analysis of the narratives in relation to their internal
structure and position within the message.
More specifically, the corpus of the discussion threads was analyzed in terms of
the sequence of the discursive moves in the posted messages following Locher’s (2006)
model of analysis for investigating advice discourse. A move is defined here as a rhe-
torical unit that contributes to the ongoing interaction and serves a particular local
function in the context it occurs. Locher’s study of online advice examines the se-
quence of discursive moves within online question-answer exchanges in letter format
in order to identify the strategies used in the social practice of advice-giving. Such
analysis is useful for the purposes of the present study since it allows mapping out the
structure of the advice-giving messages, and can therefore reveal how narratives are
organized within them. In Locher’s study, the discursive move which is identified as
“narrative” is found only in the corpus of the question letters (advice-seeking messag-
es) while a move identified as “own experience” occurs in the corpus of the answer
letters (advice-giving messages). The “own experience”4 move is defined as a personal
anecdote, which can reflect both present and past experiences (Locher 2006: 62).
In the advice-giving data, the discursive move of “narrative” was identified based
on the criterion of the minimal narrative, namely the temporal sequence of two events
connected by temporal juncture (Labov & Waletzky 1967). The narratives found were
then analyzed according to Labov’s (1972) narrative structure. Having this as a starting
point, the analytical framework was further enriched by taking into account Ochs &
Capps’s (2001) narrative dimensions of tellability, which refers to what makes a narra-
tive worth telling in a given context, and embeddedness, which refers to whether a
narrative is embedded in the surrounding discourse or detached from it. A micro-
analysis of teller strategies was also used to address the contextual particularities of the
data (Norrick 2000: 5–6).

4. Locher (2006: 67) points out that in her data the advisor cites a “personal” experience, even
though it is a fictional persona constructed by a team of experts under the alias of Lucy.
 Loukia Lindholm

All thread-initiating messages in the sample are “advice-seeking” messages that

typically contain the following discursive moves: statement of a problem, elaboration
of the problem in question and request for advice. The reply messages that were coded
as “advice-giving” contained at least one advice move; that is, a suggestion or recom-
mendation (or a series of such) for a future course of action.

4. Narratives in advice-giving

The number of narratives identified in the sample of the advice-giving messages was
16. These narratives, which vary in length and range from brief tellings to lengthy,
Labovian narratives, were found to be thematically relevant to the topic of discussion
in the thread. Specifically, they revolve around personal and/or third person past ex-
periences that reflect situations similar to the ones the advice-seekers are facing at the
time of the discussion. Consider Figure 1, which shows an example of a threaded dis-
cussion that follows.
Figure 1 illustrates the typical structure of an asynchronous threaded discussion.
The first message is always the discussion initiating message to which multiple par-
ticipants may give a reply. In turn, these reply messages can receive their own replies
by participants, and so on. This is demonstrated by indenting the reply messages to the
right and placing them under the message they correspond to.
In the corpus, all initiating messages are advice-seeking; that is, they feature a
problem statement and a request for advice, among other moves. Advice-seekers may
employ narratives to give background information and elaborate on the problem at
hand. These narratives can then trigger the recount of similar incidents and experi-
ences by the advice-givers. In the example of a discussion thread in Figure 1, the ad-
vice-seeking message (message #1) includes a request for advice for the advice-seeker’s
abusive behavior towards her child and a narrative of a particular incident that serves
as an example of her behavior. In one of the replies (message #2), an advice-giver

Discussion #19: A mother’s abusive behavior

Message #1: Advice-seeking (features a narrative on the advice-seeker’s abusive behavior)
  -> Message #2: Advice-giving (features a third person narrative on parental abusive behavior)
  -> Message #3: Advice-giving
  -> Message #4: Advice-giving
   -> Message #5: Update by advice seeker
     -> Message #6: Advice giving
-> Message #7: Advice giving
-> Message #8: Advice giving
Figure 1.  Distribution of narratives in an example discussion thread
Narratives in advice discourse 

Assessment/Advice Narrative Assessment/Advice

Figure 2.  Discursive moves that precede and follow a narrative in the advice-giving

includes a lengthy narrative that involves a similar incident to which she was a by-
stander and proceeds with advice-giving moves on how the advice-seeker can remedy
her behavior. Such narratives, which share thematic relevance with the advice-seekers’
accounts of events, are constructed as “response” narratives showing how the advice-
giver understands the preceding narrative and comments on it (Norrick 2000).
Next, the examination of the sequential organization of the advice-giving mes-
sages that contain narratives revealed that assessment and advice moves are found to
precede and/or follow these narratives (Figure 2).
One instance was found where the advice-giving message contains a narrative that
is preceded only by a greeting move (see Example 7 in Section 4.2).

4.1 Supporting advice

Narratives that are found to precede or follow advice moves are used to support advice.
Example (1) is a reply to an advice-seeker’s request for advice on the impulsive behav-
ior of her daughter at kindergarten:
(1) A little story, for food for thought....... by the way this is a true story.
While workin at the daycare center, I was in the before and after school
room....we had your typical children, the quiet shy kids, the roudy trouble
makers and the in betweens.
After a class in college, the lecture was about steriotyping children and what
happens and how it happens.
I was fairly new to this classroom, and had already figured out who was who
and what was little boy was always being blamed for this mess or
that mess....and I became quite aware that sometimes he was being sent to
clean a mess he had nothing to do because the children had fingered him as
the mess maker..... <orientation>
So one day this little boy was not there, and while the children were plaing, I
went and made a mess of the leggos....I left it.... come clean up time, the lego
mess was still there and when the lead teacher asked who made that mess, the
children all named this one little boy.....the teacher immediately started repre-
manding and lecturing this boy about the mess and sending him on his way
to clean it took a full 3 minutes of her lecturing and ordering him
before she realized he wasnt even in the room......then she went looking for
him, assuming he was hiding in the bathroom.........
 Loukia Lindholm

Once she realized he was not even there that day (I had to point out to her the
roll sheet that indicated he was not there that day) I explained to her and the
other kids that they are too quick to blame someone else, and instead of finger
pointing, maybe they should just quietly clean up the mess and stop making a
big deal over who made the mess.....
I doubt the little boy realized why he suddenly was not the one blamed for
everything any more....... <narrative>
If the teachers are not “teaching” appropriate social skills either to your dd or her
peers, I would say this is not the appropriate center for your dd to be involved
with..... <advice> she is still young enough you can find better care, and the
switch should not be too traumatic for anyone but you.... <support for advice>
JMHO Discussion #16: A troublesome five-year old
The advice-giver marks off the telling of a story with a metatextual comment that an-
nounces and authenticates at the same time the story that follows “A little story, for
food for thought....... by the way this is a true story.” After the orientation section,
where a lecture about stereotyping children is mentioned, the advice-giver continues
with an incident about a supposed trouble-maker at kindergarten. The experiment the
advice-giver did with messing the leggos proved that the teachers not only reinforced
the children’s habit of blaming the little boy for everything but also participated in do-
ing so. Using external evaluation, “(I had to point out to her the roll sheet that indi-
cated he was not there that day) I explained to her and the other kids that....” the ad-
vice-giver makes the meaning of the narrative more obvious. They were “ too quick to
blame someone else”. The advice move that follows “If the teachers are not “teaching”
appropriate social skills either to your dd or her peers, I would say this is not the ap-
propriate center for your dd to be involved with.....” states the main point of the telling:
the advice-seeker should change daycare centre. In the original advice-seeking mes-
sage, the advice-seeker/parent presents the impulsive behavior of her daughter at kin-
dergarten as the problem in focus. As we can see in the reply message above, the ad-
vice-giver shifts the focus from the child’s behavior and emphasizes from the very start
the purpose of the story that follows “A little story, for food for thought...” offering an
alternative perspective of looking at the problem, namely that the problem lies with
the daycare staff stereotyping the children and reinforcing such an attitude.
While advice-givers opt for narratives of personal or third-person experience in
their replies, another type of story employed to support advice is the parable, as shown
in Example (2).
(2) Good luck and may God help you through this phaseESPECIALLY through
these holidays..... <offering support>.....stay strong, and focus on your re-
maining two kids....they want you and need you too..... <advice>
Just like the prodicle son in the Bible<metatext>.....he tells dad “hey I cant
wait until you are dead, so give me my inheritance now”
Narratives in advice discourse 

Dad did and let the son go....he prayed for his son constantly, and son went
and did some pretty stupid the end, he realized he needed his dad
and came home..... <narrative>
Your challenge at that point? forgiveness....not holding harbor the hurts she
has bestowed up on you.... <advice>
Hang in here and vent as often as you need. <advice> We will listen, and even
understand. <offering support>  Discussion #20: Problems with teens
Any familiarity with the motif of the prodigal son parable is sufficient to understand
that the narrative advocates forgiveness. The advice-giver goes on with the telling of
the parable and after “exiting” it she continues with an advice move “Your challenge at
that point? forgiveness....not holding harbor the hurts she has bestowed up on you...”
that states explicitly the point of the narrative.
Some of the narratives that support advice feature events that lead to an undesir-
able outcome. Such narratives leave the implication that the advice-seeker may experi-
ence a similar outcome if she does not follow the advice. Thus, these narratives can be
used as implicit warnings. Example (3) shows a reply to an advice-seeker’s concern
over her son’s learning difficulties that are met with indifference by his teachers who
refuse to put him in a help class:
(3) The process of starting up special education can take forever and can be very
frustrating. But the teacher could do things to help him while you were wait-
ing if she wanted to work with him and you. <assessment> I agree with
[Laura’s] suggestions and would be in the school first thing Monday morning
to see about getting him moved. <agreement with other’s advice> <advice> I
left my oldest dd in a calss were my concerns were being blown off and it cost
her a year of school because she fell father behind. <narrative>
Good luck and please keep us updated on how things are going.
<farewell><request for update>Discussion #5: Child with learning disability
As we see in Example (3), the advice-giver offers a telling of a personal past experience
that emphasizes the potential repercussions of not moving the boy to the help class.
The evaluative action includes a passive progressive “...were my concerns were being
blown off...” which reveals that the mother was repeatedly voicing her concerns but the
teachers were dismissing them. This also serves as a comparison to the advice-seeker’s
constant appeal to the teachers for help. What follows is the most reportable event of
the narrative: the advice-giver’s daughter lost a year of school. The narrative ends with
the explicative that indicates why this happened “...because she fell father behind”.
What makes this narrative tellable is therefore the danger of losing a school year.
Example (4) illustrates another instance of a narrative that supports advice through
delivering a warning.
(4) You do need to get legal doccuments going<advice>....for your safety as well
as the best interest of the child...bringing the courts into it is not a negitive
 Loukia Lindholm

thing unless it is being used to shut the other parent out... <support for ad-
I had a friend who was being abused by her husband....he beat her so badly
that she ended up in the hospital.....while she was in the hospital, he went to
the courts and filed a restraining order on her, telling the courts she had beat
him up, and gained 100% full custody of the baby. Since she was in the hospi-
tal, she was unable to defend herself, and was totally unaware of this going on
in the first place.
When he went to the hospital, picked her up, brought her home, then called
the police and had her thrown in jail for violation of the restraining order.....
the one she had no clue about.
So when she got out, she went to see about getting her dd out of the hands of
the abuser, and found she had no parental rights, the courts dissolved her
rights while she was in the hospital. <narrative>
I am telling you this, because you never know what someone can do behind
your back. And if they use the court system you have not a leg to stand on.
So the two of you need to sit down, maybe have a mediator present to keep
things unemotional. <advice>
Discuss what you want for your child....not what you want for you or what he
wants for him, but what the two of you want for your child.....then negotiate
how that is going to happen....then get it in a legal doccument/have it carved
in stone. <advice> Discussion #23: Spouse abuse
The narrative in Example (4) involves a third-person past experience with a very dra-
matic turn. The negative outcome of the story “So when she got out, she went to see
about getting her dd out of the hands of the abuser, and found she had no parental
rights, the courts dissolved her rights while she was in the hospital” stresses the impor-
tance of following the advice that was given before “You do need to get legal doccu-
ments going.... for your safety as well as the best interest of the child...”. The coda sum-
marizes the point of the narrative and delivers a warning “I am telling you this, because
you never know what someone can do behind your back. And if they use the court
system you have not a leg to stand on”. The advice move that follows repeats more or
less the advice given before the advice-giver “entered” the narrative.

4.2 Supporting an assessment

This function concerns narratives which follow an assessment made by the advice-
giver in the reply message. The discursive move of assessment involves the evaluation
of the questioner’s particular situation by referring to the questioner’s text (Locher
2006: 62–63). In the advice-giving data, narratives that occur after assessments serve as
arguments that support the advice-giver’s assessment of the advice-seeker’s situation:
Narratives in advice discourse 

(5) Personally I don’t see a problem with your son being scared of dogs. <assess-
ment> Heck, the last person who had just told me “oh, she won’t bite. she’s
never bitten ANYONE” was pulling her off of me 30 seconds later as she had
jumped for my throat and hooked onto my arm that I’d swung around to pro-
tect my neck, and wouldn’t let go. <narrative>
Last year I ran from a weiner dog, and have done some other embarassing
things. <story abstract>I trust no strange dog whatsoever and encourage my
DS-15 the same. When he was 2 he almost lost an eye from a dog that I’d
known for 5 years. <story abstract> We now have 7 large dogs and 2 puppies.
They’re the best ever and sooo non-scary. But would I ever guarantee anyone
that none of them would EVER bite in ANY situation? NO. That’s because
they’re only animals. If anyone ever came here and asked me to put the dogs
up I’d do it immediately and never think twice about it. And I wouldn’t bother
with a psychologist. I believe he’ll outgrow fear of “all” dogs and will learn to
love one of his own one day.
Good luck Discussion #7: Child scared of dogs
In Example (5), the reply message opens up with an assessment of the advice-seeker’s
concern: her son’s fear of dogs. The advice-giver evaluates it as unproblematic and goes
on with a telling of a personal incident where the advice-giver was the victim of a dog
attack. The narrative is set off by an evaluative marker “Heck” followed by an orienta-
tion section with embedded evaluation in the form of quoting the dog owner’s words
“oh, she won’t bite. She’s never biten ANYONE”. The dramatic telling of the dog attack
that follows constitutes in itself an evaluative action that underlines the danger of com-
ing in contact with dogs. At the same time, the dog owner’s words of reassurance come
in stark contrast with what followed. This is manifested in the use of evaluative devices
such as the use of intensifiers and negation in the dog owner’s words. Also, the absence
of resolution adds more dramatic effect. The advice-giver presents herself in the posi-
tion of the victim while the dog owner is assigned the blame making this telling into a
“polarized” narrative (Labov, 1997). The story abstract “Last year I ran from a weiner
dog, and have done some other embarassing things” that follows the main narrated
incident serves as an example of the advice-giver’s mistrust of dogs. She continues with
an assertion “I trust no strange dog whatsoever and encourage my DS-15 the same”
followed by another story abstract “ When he was 2 he almost lost an eye from a dog
that I’d known for 5 years”. These story abstracts function as additional evidence to
support the advice-giver’s assessment.
In Example (6), another advice-giver in the same discussion thread makes an as-
sessment of the behavioral pattern of dogs in comparison with cats:
(6) Forget the shrink. Have him choose a stuffed animal. <advice>My dd grew up
with cats. They are differnt. Cats don’t try to lick your face. Or jump up on
you. In fact when you go to pet a cat – if they don’t know you they back up.
<general information> Dogs are more social and aggressive. <assessment>
 Loukia Lindholm

When she was about 2–3 she had a medium/large dog jump on her (being
friendly) but it scared the hell out of her. she remained afraid of most dogs
except for stuffed animals... until she physically got taller!!! now she will deal
with small dogs. she is still leary about larger ones – and I am ok with that.
<narrative>Until she gets to be 5 feet tall and her face is out of reach of dogs
it’s ok. The small dogs she likes best are also about cat size.
It is totally ok for your son to be unsure after an incident. <assessment> I did
like the suggestion of taking him to a pet shop where he can look at them –
and they can’t do anything to him– just like we go to a zoo for.. We love look-
ing at large animals in cages – don’t weJust let him know that animals are re-
ally not mean. They just don’t know better. They don’t want to hurt. They just
can’t talk. But if he really listens... they will talk to him.....
I love all animals.... but there is a time and a place.
 Discussion #7: Child scared of dogs
The external evaluation “(being friendly)” that refers to the dog ties back to the assess-
ment of dogs as being more social and aggressive. A juxtaposition is created between
the little girl “she was about 2–3” and the “medium/large dog” that jumped on her
while the most reportable event is presented with an evaluative comment “it scared the
hell out of her”. What is interesting here is that the resolution part of the narrative “she
remained afraid of most dogs except for stuffed animals...” ties back to the advice move
in the opening of the message “Forget the shrink. Have him choose a stuffed animal”.
Other narratives that support assessments in the data are those that implicitly of-
fer a diagnosis. Narratives that fulfill this particular function are found in responses to
requests for help and advice to identify the cause of a particular problem, usually of
medical nature. Such narratives establish similarities of symptoms and conditions to
the advice-seeker’s problem along with a diagnosis supported by a third-party expert:
(7) Hello! When my son was about that same age (he’s 10 years old now), he used
to do the same thing. He would wake up crying in the middle of the night for
several nights and my husband and I did everything to try to calm him down.
Finally, I took him to the doctor and told her everything that he does. She said
that he is having night terrors. She said a lot of babies go through that and they
will grow out of it. My son did it for about a month or so and then it all
stopped. Discussion #8: Infant crying at night
Example (7) consists of a greeting and a narrative move alone. At this point, it has to
be stressed that a previous poster in this discussion thread had already offered a poten-
tial diagnosis (night terrors) but with no reference to any personal experience. The
narrative in the message above supports the previously suggested diagnosis of night
terrors through establishing similarities “When my son was about that same age
(he’s 10 years old now), he used to do the same thing” with the advice-seeker’s problem
and continuing with a story that results in the diagnosis delivered by an expert.
Narratives in advice discourse 

4.3 Doing relational work

Offering support and expressing empathy is a very common feature in the advice-
giving messages. In their responses, advice-givers often underline the fact that they
have faced or are facing situations similar to those the advice-seekers are in. This is
reflected in the narratives they offer when they engage in the social practice of advice-
giving. As a result, sharing stories of similar experiences is important within the com-
munity. In the data, there is an instance of a narrative of which the purpose is rela-
tional work rather than being used in any advisory manner:
(8) [Mary],
I have been having a LOT of the same problems with my 12 yr old daughter. I
can’t believe the way she talks to me – if I ever talked to my mother that way (...)
Like tonight, she and her sister were leaving for a church retreat. Couldn’t find
her everday shoes and I wasn’t letting her wear her week old K Swiss $45.00
shoes to an outdoor retreat! I talked and asked questions for quiete some time
and she just blew up then insisted that I started off yelling at her not the other
way around, etc. I need to find one thing she and I can connect on. I know she
loves me I just think she’s overwhelmed by all that has happened and is co-
cooning herself in anger – and I need to learn not to react no matter how de-
pressed I am or how much pain etc. <narrative> (...)
I know this wasn’t my OP5 but I am going to continue watching the responses
looking for help. <apology>And [Karen], know that you are not alone! I am
not a single parent but in a way I am as DH is out working from 8 a.m. to
midnight trying to cover the loss of my income.
I love having this spot to come to to get and sometimes give advice. I can’t re-
ally talk on the phone so the computer is my lifeline and [] mes-
sage boards have become a daily spot for me!
Hang in there [Mary] – you are not the only one going through this and if we
can support and vent to each other rather than at our kids we can teach them
patience and understanding <support> Discussion #12: Child out of control
In Example (8), the participant is aware that she is violating the etiquette of the forum
by not responding to the advice-seeking message directly “I know this wasn’t my OP
but I am going to continue watching the responses looking for help”. Her contribution
is geared towards the presentation of her own problematic situation that resonates the
problem of the advice-seeker. However, there is no concrete advice offered but an ex-
pression of support to the original advice-seeker and a justification of her participation
in the forum “I love having this spot to come to to get and sometimes give advice. I
can’t really talk on the phone so the computer is my lifeline and [] message
boards have become a daily spot for me!”.

5. The abbreviation OP stands for “original post”.

 Loukia Lindholm

4.4 Stories in the making?

In the data, there are instances of accounts of events that do not fit the strict require-
ments of a minimal narrative. These are references to accounts of events in the past,
present or future that are elliptical in nature and far from the Labovian narrative struc-
ture. Stories which are announced but left to be told at a different time and story ab-
stracts are such examples.
Advice-givers often refer to stories of similar experiences that they are willing to
share but do not proceed with doing so because of the space and/or time constraints of
the interaction at hand. These stories are not realized but rather put forward to be told
later and through a different communicative mode, such as email:
(9) Your story touched me because I have had many “bully” experiences with my
3 older children. Each situation and bully had to be handled differently! It
would take 4 post to tell you all the stories I have. If you want to email me I
will be more than happy to share my stories with you. (...)
 Discussion #3: Bullied child
As we can see in Example (9), the advice-giver opens up the telling of bully stories but
leaves it for a different interaction at a not-specified time that depends on the advice-
seeker’s will. In most cases, these story-previews or “(bids for) stories to be told” are
left for a future interaction (see Georgakopoulou 2004).
On many occasions, advice-givers opt for abstracts of stories that provide a sum-
mary of a significant incident (Example 5, Last year I ran from a weiner dog, and have
done some other embarassing things). Some messages in the data include both a story
and abstracts of stories that feature similar events. For instance in Example (5), the
story of the dog attack is followed by two story abstracts on the same theme. In that
case, the story abstracts together with the story are used as arguments to further sup-
port a previous assessment.
Also, requests for updates of the advice-seeker’s situation and the events that fol-
lowed the implementation of the advice given are found in the data. These requests,
which usually occur at the end of the advice-giving message, invite tellings of events
(Example 3, Good luck and please keep us updated on how things are going).

5. Discussion

The narratives found in the data feature past experiences with the advice-giver as the
protagonist or bystander (Examples 1, 5, 6), third-person stories of which the advice-
giver claims knowledge (Example 4) as well as presumed familiar stories (the parable
in Example 2). Most of the narratives identified in the data fall into the category of
“parental narratives” that are told from the perspective of the parent and involve an
event that happened to their own child (Peterson 2004). Other advice-givers offer
Narratives in advice discourse 

a.  supporting advice

b.  supporting assessments
c.  doing relational work
Figure 3.  Functions of narratives in the response messages

tellings of events that took place in their workplace e.g. the daycare worker in Example
(1). Apart from that, participants share similar experiences which help to establish
common ground and hone community ties. In the advice-giving context under inves-
tigation, delivering advice is of major importance for sustaining the goals and princi-
ples of the community. The functions of narratives in the advice-giving messages can
be summarized as follows:
The use of narratives to support assessments and advice suggests that their use is
not that of story-telling as such (primary use) but a secondary use, that of argumenta-
tion (see Virtanen 1992). Also, an important feature of such narratives that requires
further investigation is their point of exit. Specifically, advice-givers may use the coda
for the delivery of their advice (Examples 2 and 4).
Even when narratives of personal experience are mainly used for argumentative
purposes, they can also do relational work through the sharing of similar experiences
and establishing common ground. In other words, relational work is not a trait that
appears only in “off-advice” narratives such as in Example (8). It is rather an overarch-
ing function that is in line with the principles and goals of the forum in question.

6. Conclusion

Advice-givers turn to a pool of personal experiences as parents or professionals, para-

bles and third-person experiences to construct narratives that address specific topics
in the social practice of advice-giving. This article has briefly illustrated the functions
of narratives in an online advice-giving context. As we have seen, these are embedded
narratives; that is, their thematic content is relevant to the topic of the interaction at
hand and their position within messages is immediately tied to the preceding and fol-
lowing discursive moves in terms of sequential organization. It was shown that advice-
givers employ these narratives of personal and/or third-person past experience mainly
for argumentative purposes: to support assessments and advice directed to advice-
seekers. As this was a preliminary glimpse into the functions of narratives in advice-
giving, further analysis is required in a larger corpus of advice-giving data. Finally, it is
important to extend the scope of investigation to the role of identity construction in
the advice-givers’ narratives and the relationship between narratives in advice-giving
messages and those in update messages by advice-seekers.
 Loukia Lindholm


DeCapua, Andrea and Dunham, Joan F. 1993. “Strategies in the discourse of advice.” Journal of
Pragmatics 20 (6): 519–531.
De Fina, Anna. 2005. “Who tells which story and why? Context and narrative in an Italian
American club.” Paper presented at the 9th International Pragmatics Conference, Riva del
Garda, Italy, July 2005.
Georgakopoulou, Alexandra. 2004. “To tell or not to tell? Email stories between on- and off-line
interactions.” Language@Internet 1. Available:
cles/2004/36/[2006, 15 June].
Georgakopoulou, Alexandra. 2006. “The other side of the story: Towards a narrative analysis of
narratives-in-interaction.” Discourse Studies 8 (2): 235–257.
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:
Interaction in Institutional Settings, Paul Drew and John Heritage (eds), 359–417. Cambridge:
Cambridge University Press.
Hudson, Thom.1990. “The discourse of advice giving in English– ‘I wouldn’t feed until spring no
matter what you do’.” Language and Communication 10 (4): 285–297.
Kiuru, Päivi, Poskiparta, Marita, Kettunen, Tarja, Saltevo, Juha and Liimatainen, Leena. 2004.
“Advice-giving styles by Finnish nurses in dietary counseling concerning type 2 diabetes
care.” Journal of Health Communication 9 (4): 337–354.
Labov, William and Waletzky, Joshua. 1967. “Narrative analysis: Oral versions of personal expe-
rience.” In Essays in the Verbal and Visual Arts, June Helm (ed.), 12–44. Seattle: University
of Washington Press.
Labov, William. 1972. Language in the Inner City: Studies in the Black English Vernacular.
Philadelphia: University of Pennsylvania Press.
Labov, William. 1997. “Some further steps in narrative analysis.” Journal of Narrative and Life
History 7 (1–4): 395–415.
Locher, Miriam A. 2006. Advice Online. Amsterdam and Philadelphia: John Benjamins.
Leppänen, Vesa. 1998. “The straightforwardness of advice: Advice-giving in interactions be-
tween Swedish district nurses and patients.” Research on Language and Social Interaction 31
(2): 209–239.
Norrick, Neal R. 2000. Conversational Narrative. Storytelling in Everyday Talk. Amsterdam and
Philadelphia: John Benjamins.
Ochs, Elinor and Capps, Lisa. 2001. Living Narrative. Cambridge, Mass.: Harvard University Press.
Peterson, Carole. 2004. “Mothers, fathers, and gender: Parental narratives about children.” Nar-
rative Inquiry 14 (2): 323–246.
Quasthoff, Uta M. and Becker, Tabea. 2005. “Different dimensions in the field of narrative inter-
action.” In Narrative Interaction, Uta M. Quasthoff and Tabea Becker (eds), 1–11.
Amsterdam and Philadelphia: John Benjamins.
Suzuki, Lalita K. and Calzo, Jerel P. 2004. “The search for peer advice in cyberspace: An exami-
nation of online teen bulletin boards about health and sexuality.” Applied Developmental
Psychology 25 (6): 685–698.
Virtanen, Tuija. 1992. “Issues of text typology: Narrative- a ‘basic’ type of text?” Text 12 (2):
Wenger, Etienne. 1998. Communities of Practice: Learning, Meaning, and Identity. Cambridge:
Cambridge University Press.
part 4

Structures in interaction
Appropriateness in interpersonal

Maria Sivenkova

The paper investigates a variety of communicative moves in which interlocutors

negotiate appropriateness of an utterance. Two types of such moves are analysed:
prospective (the speaker prepares the ground for her future speech act by
checking on the utterance’s appropriateness) and retrospective (the hearer
signals that the speaker’s previous utterance is viewed as inappropriate). Four
types of appropriateness explication are discussed: role, interpersonal, temporal
and locative. The study shows that there exist correlations between different
types of prospective and retrospective communicative moves, and that they
demonstrate both universal and culture-specific features. The article builds
on the author’s PhD dissertation and is the first in a series of papers aimed at
investigating the phenomenon of appropriateness in various discursive spheres.

1. Explicit references to appropriateness in dialogue

On the way to a successful communicative interaction the speaker and the hearer meet
with many obstacles: their utterances can be mispronounced, misheard, misinterpret-
ed, etc. But there is yet another type of communicative failure related to the inappro-
priate use of utterances in context. Following (Fetzer 2004: 20), appropriateness is
viewed in the paper as a relational concept which reflects “the connectedness between
a communicative action, its linguistic realization and its embeddedness in linguistic
and social contexts.” According to transactional theories of discourse processing
(Schegloff et. al 1996), acts of communication are normally coordinated between the
interactants, and appropriateness as one of the guiding principles of the social world
and an important characteristic of successful interactions (Fetzer 2004) often mani-
fests itself in discourse becoming the subject of negotiation between the speaker and
the hearer. This article focuses on a range of metacommunicative turns containing
references to appropriateness designed to monitor various “organisational issues” of
interpersonal communication.
The study is based on dialogues taken from the 20–21 century drama, fiction and
screenplays in English, French and Russian. Each of the three samples consists of
 Maria Sivenkova

250 excerpts containing conversational turns related to the (in)appropriateness of an

utterance. My view on the possibility of using literary dialogue for linguistic analysis,
following many researchers (Burton 1980; Simpson 1996; Bucks 1997 among others),
is as follows: although constructed dialogue and naturally occurring conversation are
not identical types of communication, the analysis of dialogue in drama and fiction
can cast light on the organization of spontaneous speech.

2. Direction and structural types of references to appropriateness

An important distinction to account for the variety of types of references to

(in)appropriateness is the direction of communicative moves. As this research shows,
there are two types of communicative turns under analysis: initiating/prospective moves
in which the speaker prepares the ground for a future speech act by checking on the
utterance’s appropriateness (see Example (1)) or apologising in advance for its possible
inappropriateness (see Example (2)) and responding/retrospective moves in which the
hearer signals to the speaker that his/her previous utterance is viewed as inappropriate
(see Example (3)) or the speaker him/herself acknowledges the inappropriateness of
his/her previous contribution (see Example (4)).

(1) “Do you mind if I ask you a per-

sonal question, Mr. Archer?”
“Go ahead.”
“Are you interested in Miss Dew?
You know what I mean. Seriously.”
(R. MacDonald. Far Side of the
(2) “Mon capitaine... Vous m’excuserez (2) “Commander... You will excuse
de me mêler de ce qui ne me re- me for meddling in matters that
garde pas, mais... Vous tombez ici don't concern me, but... You find
sans savoir... Je voudrais vous yourself here, unaware ... I wanted
prévenir...” (L. Aragon. Aurélien); to warn you...” (L. Aragon. Aure-
(3) “Going out with each other?”
“What’s it to you?” Joe asked po-
(M. Keyes. Last Chance Saloon);

1. All Russian and French examples are translated by the author.

Appropriateness in interpersonal communication 

(4) − Au fait, pourquoi n’est-tu plus (4) “As a matter of fact, why are you
pratiquante? not a church-goer anymore?”
− C’est venu comme ça. Je n’ai pas “It happened like that. I haven’t re-
renoncé à ma foi. Je ne peux pas nounced my faith. I can’t talk
en parler. about that.”
− Pardonne-moi, c’est indiscret. “I’m sorry, it is indiscreet.”
(Z. Oldenbourg. Les Amours (Z. Oldenbourg. Strayed Love).

I will argue that there exist correlations between various types of prospective and ret-
rospective references to (in)appropriateness; i.e. for each type of preparatory moves
there is a corresponding type of retrospective turns demonstrating an utterance’s fail-
ure to be appropriate. Interestingly, the number of retrospective turns outnumbers the
prospective ones comprising approximately two thirds of each of the samples under
study (see Table 2.1). This indicates, in my view, that the corrective function of appro-
priateness explication in discourse prevails over its preventive function. In other words,
interactants generally prefer to signal inappropriate performance of an utterance post
factum rather than take preventive measures to avoid potential inappropriateness.
As Examples (3) and (4) show, apart from the direction of communicative moves,
there is yet another important dimension related to the retrospective turns to take into
account: self-repair (see Example (4)) vs. other-repair (see Example (3)). Although the
self-repairing type of references to inappropriateness is not numerous (Engl. 7%;
Rus. 7%; Fr. 4% of the total number of examples), it seems to prove the point that the
negotiation of appropriateness in interpersonal communication is viewed as a joint
project of all the participants of an interaction. Further analysis of the self-repair/oth-
er-repair distinction indicates that both self-initiated (see Example (5)) and other-
initiated repairs (see Example (4)) are present among references to inappropriateness.
(5) “The Indians never divorce.”
“They haven’t seen my first wife either.”
“She was unpleasant? I’m sorry, I’m not prying. It’s not important.” (J. Graham.
The Testament).

Table 2.1  The ratio of prospective versus retrospective moves related to appropriateness

sample English Russian French

direction of moves
prospective moves 30% 32% 27%
retrospective moves 70% 68% 73%
 Maria Sivenkova

In Example (5), a series of redressive verbal actions (an apology I’m sorry, an acknowl-
edgement of overstepping the bounds and a promise not to do it again I’m not prying,
as well as a mitigator It’s not important) performed by the speaker immediately follows
her own intrusive question about the hearer’s wife. In Example (4), the speaker ac-
knowledges her inappropriate verbal behaviour only after the hearer’s comment indi-
cating displeasure with the speaker’s intrusiveness (Je ne peux pas en parler – ‘I can’t
talk about that’).
As other-initiated other-repairs constitute the majority of our retrospective ex-
amples of references to inappropriateness, further in the article I mainly provide ex-
amples of this type.
One more important aspect of prospective references to (in)appropriateness is
related to their dependence on the communicative turn for which they function as
preparatory moves. Along with references to (in)appropriateness that constitute a
single communicative turn (e.g. Can I ask you a personal question? − see Example 1),
there are also shorter references incorporated in larger units that serve mostly as dis-
claimers (e.g. It’s none of my business, but...; Если не секрет − ‘If it is not a secret’; Sans
indiscrétion − ‘Without indiscretion’). They are often highly clichéd and cannot be
used independently of the propositional content they are designed to mitigate. Be-
sides, as will be shown later in the article, negotiating (in)appropriateness can some-
times cover several communicative moves in a long stretch of dialogue, making it a
truly interactive activity.
To sum it up, references to (in)appropriateness in interpersonal communication
can take a variety of forms ranging from short conversational routines to longer
stretches of dialogical discourse, both prospectively and retrospectively directed, per-
formed by either speakers or addressees of an utterance.

3. Functional types of appropriateness explication in dialogue

Among a long list of elements of context that may be viewed as inappropriate and
consequently made explicit by the participants of an interaction, four items seem to
be of great relevance in dialogical discourse: appropriate interpersonal relationships
between the speaker and the hearer, appropriate roles of interactants, appropriate
time and space of the conversational contribution. As this analysis shows, the appro-
priateness of these elements of context is negotiated in dialogical discourse on a
regular basis, which may account for the clichéd nature of many of the conversa-
tional turns under study. These four elements constitute four types of appropriate-
ness explication, each type being represented by both initiating and responding
communicative moves.
Appropriateness in interpersonal communication 

3.1 References to role (in)appropriateness

The first two types of conversational turns referring to the appropriateness of an ut-
terance are related to the participants of an interaction. As Stubbs succinctly puts it,
“any situation of talk is a microcosm of basic social and personal relationships”
(Stubbs, 1983: 61). To secure the felicitous outcome of most interactions, both the il-
locutionary force and the propositional content of the speech act should be in compli-
ance with the interactants’ roles and the type of interpersonal relationship between
the speaker and the addressee. The relevance of the conversational turns referring to
role and interpersonal (in)appropriateness may be illustrated by the following quan-
titative data: such turns represent the two most numerous types in the samples, ac-
counting for approximately two thirds of all the examples under study. In fact, refer-
ences to role (in)appropriateness constitute 32 per cent of the total number of English,
37 per cent of Russian and 33 per cent of French examples; whereas references to in-
terpersonal (in)appropriateness comprise 39 per cent of English, 33 per cent of
Russian and 30 per cent of French examples. They are also the most diverse types in
terms of their surface structure.
Conversational turns referring to role (in)appropriateness show that to be felici-
tous a speech act needs to be compliant with the speaker’s and addressee’s various
roles: communicative (Clark and Carlson 1982), status-based, position-based, situa-
tional (Gerhardt 1980; Karasik 2002), gender (Coates 2003; Baxter 2006), etc. A par-
ticular role may empower the speaker to perform a speech act, deprive him/her of the
right, or be subject to negotiation between the interactants. Compare (6):
(6) “<...> you’ve got to know what our people are saying outside – they’re saying if
you were the President of the United States, <...> you’ll be like the ones before –
the President of the whites –”
“That’s enough from you, Julian, that’s quite enough. You remember who you
are and who I am, and that I’m the one who’s still in charge of seeing you think
right and behave right...” (I. Wallace. The Man).
Example (6) illustrates the situation when Julian’s family role (Dilman’s son, a minor)
hinders the successful realisation of his speech acts conveying criticism of his father’s
political decisions. Dilman refuses to embark on an otherwise important political dis-
cussion by making an explicit reference to Julian’s and his own roles.
Furthermore, the roles that are subject to negotiation in discourse can be indicat-
ed explicitly (I am the lawyer, you are the witness, I ask the questions, you give the an-
swers; Как ты разговариваешь с отцом?! − lit. ‘How are you talking with your fa-
ther?!’; Vous n’êtes pas mon confesseur que je sache − ‘You are not my confessor, as far as
I know’), be associated with a certain illocutionary force (e.g. the speaker who has no
authority issues a command – see Example (7)) or they may be derived from the con-
text of an interaction (see Example (8)).
 Maria Sivenkova

(7) −  Vadeau, fous-lui la paix! (7) “Vadeau, leave him alone!”

−  Non. Il obéira! “No. He will obey!”
− Ce n’est pas toi qui commande, “It is not you who commands
ici! (B. Clavel. La saison des here!” (B. Clavel. The season of
loups); wolves);
(8) [хирург (говорящий) задает (8) [the surgeon (the speaker) asks
пациенту (адресату) вопрос, не the patient (the addressee) a
относящийся к question that does not relate to
профессиональной the medical encounter]:
компетенции врача]:
− Погодите-ка... Можно “Wait a minute... May I ask a pri-
частный вопрос? Правда ли vate question? It is rumoured that
поговаривают, что Пугачева Pugacheva is Boris Eltsin’s illegiti-
− незаконнорожденная дочь mate daughter. Is it true?”
Бориса Ельцина? “Fibs.” (A. Nikonov. Stalking Un-
− Легенды (А. Никонов. noticed).
Подкравшийся незаметно).

Example (8) illustrates a simple technique the speaker uses to back up the speech act
that is beyond his authority: his question is preceded by a pre-question that qualifies
the speech act to follow as a private one, thus asking for the hearer’s permission for the
forthcoming breach of roles (see Schegloff 1980 and Goldschmidt 1998 among others
on pre-sequences and their discursive functions).
As my data show, the communicative moves in which role appropriateness is ne-
gotiated by participants of an interaction take different forms in terms of their direct-
ness/indirectness, varying from the hearer’s explicit protests against the speaker’s inap-
propriate verbal behaviour (e.g. How dare you!; Я запрещаю так со мной
разговаривать! − ‘I forbid you to talk to me like that!’, Что ты себе позволяешь?!
− lit. ‘What do you permit yourself?!’; C’est le comble! – ‘That’s too much!’) to indirect
means of signalling role inappropriateness (e.g. I’ll make notes after the lecture – used
to indicate the hearer’s annoyance at the speaker’s patronising tone; Tu ferais mieux de
penser à ton projet – ‘It would be better for you to think about your project’ to mean
I’m the boss here).
The diversity of communicative moves referring to role inappropriateness can be
further illustrated by taking into account such an important parameter as the type of
speech act used by interactants to pave the way for a forthcoming conversational con-
tribution or indicate role inappropriateness. I have found that such moves can be rea-
lised in the form of the following speech acts:2

2. This analysis of types of speech acts is based on a modified version of Searle’s seminal clas-
sification of illocutionary acts (Searle 1969), with the speech act of question viewed as a separate
Appropriateness in interpersonal communication 

1. statements:
I’m not your errand boy, You are not my superior; Не властен – ‘I have no authority’,
Ты не говоришь, а слушаешь и отвечаешь на вопросы – ‘You will not talk, but listen
to and answer the questions’; У тебя слова нет – ‘You have no permission to speak’;
Je ne te demande rien – ‘I’m asking you nothing’, Je n’ai pas de comptes à te rendre – ‘I
do not report to you’, Tu n’as pas le droit de dire cela – ‘You have no right to say this’.
2. questions:3
Who might you be to give orders in the house?, You want to tell me how to run my busi-
ness?; Где ты такую власть взала? – ‘Where did you get the power?’, А у вас есть
на это право? – ‘Do you have the right?’; C’est toi qui m’interroges? – ‘Is it you who
interrogates me?’, Tu me donnes des ordres, toi? – ‘Is it you who gives me orders?’, De
quel droit? – ‘By what right?’.
3. directives:
Please don’t swear in front of the children, Stop with the third degree, Let’s go in a private
office; Вы не командуйте! – ‘Don’t give orders!’, Не смей так с матерью
разговаривать! – ‘Don’t you dare talk like that to your mother!’, Я хотела бы
поговорить с вами наедине – ‘I’d like to talk to you alone’; Je te prie de te taire – ‘I ask
you to keep silent’, Tais-toi – ‘Shut up’, Parle plus bas – ‘Speak quieter’.
4. reproaches:
That’s supposed to be my line, A person of your eminence in our profession must know
that I cannot answer that question; Ну как ты можешь?! – ‘How can you?!’, Что ты
себе позволяешь?! – ‘What do you permit yourself?!’, Я ведь вас просил не говорить
об этом при жене – ‘I asked you not to talk about that before my wife’; Vous auriez pu
ne pas me dire cela en public – ‘You shouldn’t have said that in public’, On ne parle pas
comme ça de ses parents – ‘One shouldn’t talk like that about one’s parents’.
5. apologies:
Sorry; Извините – ‘I am sorry’; Pardonne-moi – ‘Forgive me’.
Furthermore, the communicative moves referring to role (in)appropriateness differ in
the degree of the interactants’ confidence in the distribution of roles. Their choices
regarding the acceptance of a role in discourse are as follows: asserting a role, checking
on it or stating the breach of roles.

type (Kobozeva 2000) and reproaches regarded as a distinctive class to account for the difference
in the degree of explicitness of criticism between references to role and interpersonal
(in)appropriateness realised as statements and those conveyed as reproaches.
3. It should be noted that most interrogative sentences referring to role and interpersonal in-
appropriateness in this study fall in the category of rhetorical questions and imply a negation of
the speaker’s right to perform the utterance she/he has generated (You want to tell me how to run
my business? = You have no right to tell me how to run my business, Ça vous regarde? = Ça ne vous
regarde pas, etc.).
 Maria Sivenkova

As this research shows, statements of the breach of roles are by far the most popular
form of references to role (in)appropriateness (approx. 90% in the three samples).
Compare Example (9) in which the hearer complains about the breach of communica-
tive roles:
(9) [a conversation between two lawyers representing the opposite sides]:
“We’re suing for five million dollars”
“No, you’re not. Because your client is not going into court. I just paid her a
visit. <...>”
“You had no right to talk to Connie Garrett without my being present”
(S. Sheldon. Rage of Angels).
In Example (9), the hearer mentions that the roles are being breached in a matter-of-
fact manner. Neither mitigation, nor aggravation of negative impact on the hearer
occurs in the interaction. However, both types of illocutionary force modification are
possible in references to role inappropriateness. In fact, when interactants acknowl-
edge this type of inappropriateness, the communicative turns often include a mitigat-
ing device to reduce the risk of face-threat. Compare Example (10) in which Schwarz’s
ironic remark about Mitya’s low social status is downplayed with the help of an apol-
ogy (Excuse me), and two disclaimers − the first one highlighting the hearer’s profes-
sional qualities (I have great respect for you as a good storekeeper), and the second one
acknowledging the possibility of a mistake in the attribution of the addressee’s status
(если я не ошибаюсь – ‘if I am not mistaken’):

(10) [Митя просит Шварца (10) [Mitya asks Schwarz to let

разрешить внуку Шварца Schwarz’s grandson interrupt his
прервать длительные занятия long music class and go for a
музыкой и отправиться на walk]:
Шварц (со смешком). Вот как? Schwarz (with a snicker). Really?
Вы меня извините, Митя, я вас Excuse me, Mitya, I have great re-
очень уважаю как хорошего spect for you as a good storekeeper
кладовщика, и всякое такое.... and all that... But, if I am not mis-
Но если я не ошибаюсь – я имею в taken – I mean music – Professor
виду музыку, – так профессор Stolyarsky is not your last name?!
Столярский это не ваша (A. Galitch. Sailor’s Silence).
фамилия?! (А. Галич. Матросская

Interestingly, Example (11) can serve as an illustration of the opposite tendency to ag-
gravate the negative impact of references to role inappropriateness. In this example, a
negative evaluation of the addressee (Tu es une sotte – ‘You are a silly woman’) is added
Appropriateness in interpersonal communication 

to the statement of the breach of roles (je n’ai pas de comptes à te rendre – ‘I have no
accounts to give you’).

(11) – Je voudrais savoir... <...> si rien, (11) “I’d like to know ... <...> if nothing
dans sette lettre, n’est de nature à lui in this letter is able to cause him
causer quelque peine, à lui?... pain?...”
Anabel la regarda avec ironie. Anabel looked at her ironically.
– Excusez-moi, maîtresse, murmu- “I’m sorry, mistress,” murmured
ra Bessie. Bessie.
– Tu es une sotte, et je n’ai pas de “You are a silly woman, and I have
comptes à te rendre (P. Benoit. Le no accounts to give you.”
Lac Salé). (P. Benoit. The Salted Lake).

As my data show, evaluative remarks in statements of this kind are mostly negative if
they are hearer-oriented (as in Example (11)) and usually positive if they are speaker-
oriented. Compare Example (12) in which the concierge, being accused of careless-
ness, defends himself and boosts his self-presentation by explicitly rejecting the wom-
an’s right to ‘lecture’ him (Vous n’allez pas me donner des leçons en plus! – ‘You are not
going to lecture me!’) and emphasising his professional experience (Il y a vingt ans que
je fais ce métier! – ‘I’ve been in this job for twenty years!’):

(12) La mere: <...> C’est tout de même (12) Mother: <...> But, anyway, it is you
vous qui en êtes responsable, non? who are responsible, aren’t you?
Le concierge: Oh, ça va! Vous n’allez Concierge: Oh, it’s too much! You
pas me donner des leçons en plus! are not going to lecture me! I’ve
Il y a vingt ans que je fais ce métier! been in this job for twenty years!
(J. Anouilh. L’Arrestation). (J. Anouilh. The Arrest).

When the interactants’ interpretations of their roles do not coincide, special efforts
need to be undertaken to come to terms regarding the distribution of roles. As can be
seen from Example (13), the role-asserting process can cover several communicative
moves in a rather long stretch of dialogue, making it a truly interactive activity.
(13) “How many wives have you had?”
“How many have you had?” Junior shot back, then looked at his lawyer for ap-
proval. <...>.
“Let me explain something to you, Mr. Phelan,” Nate said without the slightest
irritation. “I will go over this very slowly, so listen carefully. I am the lawyer, you
are the witness. Do you follow me so far?”
Troy Junior slowly nodded.
“I ask the questions, you give the answers. Do you understand that?”
 Maria Sivenkova

The witness nodded again.

“You don’t ask questions, and I don’t give answers.” (J. Grisham. The Testa-
As Example (13) illustrates, the witness’s (Troy Junior) interpretation of the lawyer’s
(Nate) question as a private one and consequent refusal to answer it forces the lawyer
to initiate a series of verbal actions that assert the lawyer’s communicative rights.
To further illustrate the diversity of references to role (in)appropriateness, Exam-
ple (14) shows that the distribution of communicative roles can be accomplished by
making a reference to the genre of interaction (informing, which assigns the addressee
the role of a passive listener, rather than discussing, which would give her a possibility
to participate in the dialogue on equal terms). By assigning the communicative roles
in this way, the husband (the addressee) tries to avoid an argument with his wife
(the speaker), as can be seen in (14).

(14) − Tu m’as l’air faraud. Tu trembles (14) “It seems there is nothing to brag
du maigre des fesses et tu voud- about. You are trembling. And
rais jouer au p’tit soldat. <...> you’d like to play soldier. <...>”
− Je n’ai pas à discuter avec toi, “I’m not going to enter a discus-
Maïa. Je t’informe seulement sion with you, Maya. I’m only in-
d’une chose, et j’aurais mieux forming you, and it would be
fait de ne rien dire (L. Guilloux. better if I told you nothing.”
Le Sang Noir). (L. Guilloux. The Black Blood).

Checking on roles is the type of references to role (in)appropriateness in which interac-

tants demonstrate the lowest degree of confidence in the distribution of roles in con-
trast to the role-asserting moves and the statements of the breach of roles that have
been analysed above. When interactants are in doubt about role appropriateness, such
communicative moves help them clarify how roles are assigned. Compare (15):
(15) [two businessmen discuss a project]:
“The fee we are discussing is .5 of one percent per annum based on the assets
under management <...>.”
“Are you in a position to make the commitment?” (P. Erdman. The Billion
Dollar Sure Thing).
To sum it up, references to role (in)appropriateness demonstrate a variety of illocu-
tionary and epistemic forms of realisation and can be accompanied by conversational
turns serving to mitigate or aggravate negative impact on the hearer.
Appropriateness in interpersonal communication 

3.2 References to interpersonal (in)appropriateness

Moving on to the second type of appropriateness explication in dialogue, conversa-

tional turns referring to interpersonal (in)appropriateness are designed to establish a
correspondence between the relationship shared by the participants of an interaction
and the speech act performed. If the speaker regards the forthcoming speech act as
potentially threatening to his/her rapport with the hearer, he/she prefaces the speech
act with a pre-question (see Example (1)) or disclaimer to lessen the threat (see Ex-
ample (2)). The failure to do so may result in the hearer’s complaint about the speaker’s
Similar to communicative moves referring to role (in)appropriateness, explicit
references to interpersonal (in)appropriateness can be performed in a variety of ways,
taking the form of the following speech acts:
1. statements:
That’s none of your/mine concern/business, It’s entirely your affair, It’s hardly any concern
of mine; Это вас не касается – ‘It’s none of your concern’, Это не ваше/не мое дело
– ‘It’s none of your/mine business’, Не твоего ума дело – lit. ‘It’s not a business for your
intellect, Это мое (личное) дело – ‘It’s my (personal) business’; Ça ne vous/me regarde
pas – ‘It’s none of your/mine concern’, C’est mon/ton affaire – ‘It’s my/your affair’,
Chacun son affaire – ‘To each his own’.
2. questions:
Do you mind if I ask you a personal question?, What’s it to you?, What does it have to do
with you?, What makes it your business?; Можно задать тебе личный вопрос? – ‘May
I ask you a personal questions?’, Какое вам дело? – ‘What is it to you?’, Ничего, что я
так говорю? – ‘Is it all right that I say so?’; Ça vous regarde? – ‘Does it concern you?’,
Qu’est-ce que ça peut vous faire? – lit. ‘What can this do to you?’, Qui t’a demandé de
t’occuper de ça? – lit. ‘Who asked you to occupy yourself with this?’, Serait-ce indiscret?
– ‘Would that be indiscreet?’.
3. directives:
Don’t talk to me like that, Don’t be flip, Mind your own business, I’d rather you minded
your own business, Take care you don’t go too far; Не будь грубой – ‘Don’t be rude’,
Уважай вкусы других людей – ‘Respect other people’s tastes’, Не суй нос не в свое
дело! – lit. ‘Don’t poke your nose into a business which is not yours’; Mêlez-vous de ce
qui vous regarde/votre affaires! – lit. ‘Meddle with things that concern you/your affairs’,
Prenez garde et ne me poussez pas à bout! – ‘Beware and don’t exasperate me!’.
4. reproaches:
How dare you (talk to me like that)?, What kind of question is that?; Разве можно говорить
такое/задавать такие вопросы? – ‘How can you say so/ask such questions?’, Ты что
себе позволяешь? – ‘What do you permit yourself?’’, Как ты смеешь такое говорить?
– ‘How dare you say such things?’; Comment peux-tu dire une chose pareille? – ‘How can
 Maria Sivenkova

you say a thing like that?’, Vous avez une façon odieuse de vous mêler de ce qui ne vous
regarde pas – ‘You have a disagreeable manner to meddle with what does not concern
you’, Ta question n’est guère delicate – ‘Your question is not very delicate’.
5. apologies:
I shouldn’t have said that, I suppose it was an intrusive question; Извини меня, я не
должна была этого говорить – ‘I’m sorry, I shouldn’t have said that’; Je m’excuse
d’avoir posé la question – ‘I apologise for having asked the question’, Pardonne-moi, c’est
indiscret – ‘Forgive me, it’s indiscreet’.
In contrast to role inappropriateness, however, the retrospective turns in which the
hearer signals the interpersonal inappropriateness of the speaker’s utterance, despite
their face-threatening character, are not usually accompanied by mitigating devices to
minimise face-threat, Examples (16) and (17) being rare exceptions. In Example (16),
the reference to interpersonal inappropriateness is mitigated by means of the reverse
polarity tag (is it?) added to the negative stem clause (That’s really none of your con-
cern). By expressing the idea in the form of a disjunctive question, the speaker is at
least formally soliciting the hearer’s agreement with the proposition negated by the
stem clause (see McGregor 1995 among others on functions of English tag questions).
Similarly, in Example (17), the negative impact of the reference to inappropriateness is
downplayed with the help of the tag non.

(16) “And how have you been lately, Duchess?”

“That’s really none of your concern, is it?” (S. Sheldon. If Tomorrow Comes);
(17) “Notre couple vous occupe un peu (22) “Our couple occupies you a bit
trop, non?” (F. Sagan, Les merveil- too much, isn’t it?” (F. Sagan.
leux nuages). Wonderful Clouds).

One way to account for the presence of numerous mitigating devices in responsive
turns referring to role inappropriateness and their absence in references to interper-
sonal inappropriateness may be that interactants assess role conflicts and violations of
interpersonal boundaries as possessing different degrees of face-threat and thus re-
quiring different strategies of face-threat redress. To assume that role inappropriate-
ness is perceived by participants of an interaction as a more serious offence in contrast
to interpersonal inappropriateness is one in a range of possible interpretations. It
seems to be supported by the existence of a recurrent three-part sequential pattern that
contains an apology as an immediate redressive action typical of many references to
role inappropriateness:

role-breaching utterance + acknowledgement of role breach + apology.

Appropriateness in interpersonal communication 

Compare Example (18) containing a series of questions that are perceived as inquisi-
tive by one of the interactants, a consequent statement of the role breach Stop with the
third degree, an apology Sorry, a justification Just curious, and a promise to redress pos-
sible discomfort caused by the insistent questioning We must bring you out for a wel-
coming drink:
(18) [A group of friends is “interrogating” a new tenant]:
“And you’re here to work?”
“I’ve just moved from Cork,” Tom said. “I’ve got a new job here.”
“Doing what?” <...>
“Guys, stop with the third degree,” Holly reprimanded them.
“Sorry,” said Joan. “Just curious. We must bring you out for a welcoming drink.”
(C. Kelly. Just Between Us).
Curiously, this sequential pattern is almost nonexistent among my examples of re-
sponsive turns referring to interpersonal inappropriateness.
The contrary explanation for lack of mitigating devices in retrospective references
to interpersonal inappropriateness may be that the hearer’s resentment at certain in-
trusive utterances is so great that he/she goes for immediate redress and wastes no
time to mitigate the forthcoming turn (see Example (19)).
(19) With a thin, superior smile the network chief said, “I’m not in the middle of a
messy divorce action. You are. <...> You also have debts <...>, so you desper-
ately need a continuing income; otherwise you’ll be a personal bankrupt and the
next thing to a pauper.”
Raising his voice, he objected, “That’s insulting! It’s an intrusion on my per-
sonal privacy.” (A. Hailey. The Evening News).
Interestingly, retrospective references to role and interpersonal inappropriateness can
be formulated as the hearer’s complaints about the speaker’s improper tone. Сf. Ex-
ample (20) in which an accusation (You ruined the game) triggers Cable Guy’s com-
plaint about Steven’s tone, which is supported by an explicit reference to the interac-
tants’ social roles (friends):
(20) Steven: You ruined the game.
Cable Guy: I don’t appreciate your tone, Steven. That’s not the way friends
speak to each other. (J. Apatow. The Cable Guy).
As this study shows, such conversational turns often follow the speaker’s critical or
ironic remark about the hearer, a reproach, the use of low-register words, etc. which
the hearer perceives as unacceptable in terms of his/her role and the type of relation-
ship between the interlocutors. There exist a well-developed field of clichés that spe-
cialise in rendering the idea of role inappropriateness by means of references to im-
proper tone: I don’t like/appreciate your tone of voice, I object to your tone, Don’t talk to
me in that tone of voice!, Don’t you say ... in that ungrateful/patronising/disembodied
 Maria Sivenkova

tone, That’s not the tone of voice to use with your mother/father, I would suggest not tak-
ing that tone; Ваш тон неуместен – ‘Your tone is inappropriate’, Мне не нравится
твой тон – ‘I don’t like your tone’, Не семей разговаривать со мной таким тоном!
– ‘Don’t you dare talk to me in that tone!’, Почему вы разговариваете со мной таким
тоном? – ‘Why are you talking to me in that tone?’, Ваш ироничный/
запанибратский/осуждающий/назидательный/менторский/прокурорский тон
неуместен – ‘Your ironic/scornful/condescending/categorical/commanding/famil-
iar/unceremonious/accusing/didactic/mentoring/prosecutor’s tone is inappropriate’;
Arrête de me parler sur ce ton! – ‘Stop talking to me in that tone!’, Je n’aime pas le ton que
tu emploies avec moi – ‘I don’t like the tone you use with me’, Je ne vous permets pas de
me parler sur ce ton – ‘I don’t permit you to talk to me in that tone’, Et tu oses me parler
sur ce ton? – ‘Do you dare talk to me in that tone?’, Pourquoi ce ton ironique/condescen-
dant/moqueur/méprisant? – ‘Why use this ironic/condescening/mocking/contemptu-
ous tone’.
As this list of clichés reveals, there are a number of interesting differences in
their surface structure in the three samples. First, the range of Russian attributes that
can be combined with the noun тон – ‘tone’ is significantly more varied, which may
indicate greater salience of this segment of interactional grammar in Russian cul-
ture. Secondly, English examples show a tendency to mitigate possible face-threat by
means of various lexical, grammatical, and syntactical means: I don’t like the tone of
your voice one bit, I would suggest not taking that tone, I don’t like your tone, but I see
your point; Although I am very grateful..., I don’t appreciate your tone. Although
similar attempts to mitigate the negative impact of such clichés occur in the Russian
and French samples as well, they are much less frequent (e.g. Я бы на твоем месте
не стала разговаривать таким тоном – ‘I wouldn’t talk in that tone if I were
you’; Arrete s’il te plait ce ton condescendant – ‘Please stop talking in that conde-
scending tone’).

3.3 References to locative (in)appropriateness

Turning to the third type of references to (in)appropriateness − locative − one ob-

serves how the need for a speech act to fit the surroundings in which it is performed
is made manifest in dialogical discourse. This group represents the least quantita-
tively relevant type of appropriateness explication, according to these data (Engl. 4%,
Rus. 5%, Fr. 8% of the total number of examples). The retrospective moves are often
clichés: It’s not the place to...; Не здесь – ‘Not here’; Здесь не место для ... − ‘It is not
the place to...’; Pas ici – ‘Not here’, Nous ne sommes pas bien ici pour parler – ‘It’s not
convenient to talk here’, Ce n’est pas le lieu de parler – ‘It’s not the place to talk’, etc.
Compare (21) and (22):
Appropriateness in interpersonal communication 

(21) [the interlocutors – Mortimer and his wife − are at their friends’ and have just
participated in a quarrel]:
Mortimer: Come away, darling. I can’t say here what I want to say
(W.S. Maugham. The Constant Wife);
(22) Orphée: Dites-lui qu’elle n’est pas (22) Orpheus: Tell her that she is not
comme les autres croient, qu’elle est what others think she is, she is
comme moi je sais qu’elle est! what I know her to be!
Dulac: C’est trop compliqué à Dulac: It’s too complicated to
expliquer dans une gare <...>. explain at a railway station <...>.
(J. Anouilh. Eurydice). (J. Anouilh. Eurydice).

In Example (21), Mortimer’s complaint to his wife about the improper place of the
communicative interaction serves as a prospective reference to locative inappropriate-
ness and is designed to postpone further communication till the family get home.
However, it also demonstrates that boundaries between various types of references to
appropriateness can sometimes be blurred, since Mortimer’s turn can also be inter-
preted as a reference to the inappropriateness of communicative roles, as an attempt to
avoid potential witnesses/overhearers of their conversation. The retrospective refer-
ence to locative inappropriateness in Example (22) seems to present a more clear-cut
case: Dulac does not wish to start a serious conversation with Orpheus, since a railway
station is an inappropriate place for such type of interaction. However, a second inter-
pretation (role inappropriateness caused by the presence of potential overhearers) can-
not be completely ruled out either.

3.4 References to temporal (in)appropriateness

The last type of references to inappropriateness highlights the importance of time fac-
tor for the successful realisation of many speech acts. Temporal (in)appropriateness is
often negotiated in interpersonal communication with the help of numerous highly
clichéd initiating moves (Do you have a minute?, Got time for a word?; Ты не
торопишься? – ‘Are you in a hurry?’, Я некстати? – ‘Am I disturbing you?’; Vous
avez une/deux/cinq minutes? – ‘Do you have one/two/five minutes?’, Tu es très occupé?
– ‘Are you very busy?’) and responsive conversational turns (I am busy, This is not the
time for...; Не сейчас – ‘Not now’, Нашли время для... – iron. ‘You have found the time
for...’; Ce n’est pas le moment – ‘It’s not the right moment’, Le moment est mal choisi –
‘The moment is badly chosen’). They account for approximately one fourth of all refer-
ences to (in)appropriateness (Engl. 25%, Rus. 25%, Fr. 29% of the total number of ex-
amples). Such communicative moves termed “checks on availability” in (Edmondson
and House 1981: 116) are viewed by the authors as the strategy “designed to remove
the possibility that the hearer will later cut the conversation, or complain that he can-
not talk at this time.” As Goffman puts it, by performing such conversational turns the
 Maria Sivenkova

speaker opens up “a channel of communication which stays open beyond the hoped-
for reply that ratifies the opening” (Goffman 1981: 49). Compare Example (23) that
features a prospective reference to temporal appropriateness and Example (24) that
shows a probable communicative consequence of absence of a check on the interlocu-
tor’s availability:

(23) Teddie: Do you mind if we talk it over now?

Elizabeth: No.
Teddie: It makes me feel rather shy and awkward. <...> (W.S. Maugham. The
(24) − Non, Julien... Dis-le. Dis-le. (24) “No, Julien... Say it. Say it.”
− Je t’en supplie, Geneviève! Ce n’est “I beg you, Genevieve! It’s not the
pas le moment (N. Calef. Ascen- right moment” (N. Calef. Lift to
seur pour l’échafaud). the Scaffold).

3.5 Some combinations of references to (in)appropriateness

This description of (in)appropriateness explication in dialogue would be incomplete

without mentioning that various combinations can be formed on the basis of the four
types of references, with temporal and locative, on the one hand (see Example (25)),
and interpersonal and role appropriateness, on the other hand (see Example (26)),
showing preference for combined usage.

(25) − Vous voulez me parler (25) “Would you like to talk to me

d’Aurélien? about Aurelien?”
− Oui, mais... <...> pas ici... pas “Yes, but... <...> not here... not
maintenant... now...”
(L. Aragon. Aurélien); (L. Aragon. Aurelien);
(26) “Are forgotten god-daughters allowed to ask very personal questions?”
“Of course.”
“Was there any truth in that play you wrote?” (J. Fowles. Daniel Martin).

One possible interpretation for this co-occurrence may be related to the similarity of
functions such moves perform: conversational turns referring to temporal and locative
appropriateness mainly serve to open an exchange, whereas communicative moves
explicating interpersonal and role appropriateness are designed to help the speaker
obtain the addressee’s “sanction” for a specific speech act, mostly question or state-
ment. Furthermore, time and space belong to the most important characteristics of
any activity, including communication, which makes it natural for the respective refer-
ences to co-occur in one communicative move.
Appropriateness in interpersonal communication 

4. The symmetry of prospective and retrospective

references to (in)appropriateness

It is notable that most retrospective moves containing references to temporal inappro-

priateness and many of those referring to interpersonal inappropriateness are negative
transformations of the corresponding prospective turns assisted by several classic
transformational operations, mostly deletions, insertions and pronoun substitutions.
The correlations highlight, in my view, the preventive-corrective relationship between
the symmetrical turns (see Table 4.1).
As Examples (27)−(28) show, both references to temporal (in)appropriateness are
centred around the lexeme time making part of the proposition it’s time (to do smth.),
with the prospective turn taking the form of an interrogative sentence, and the retro-
spective one presenting the negated variant of similar propositional content. In Ex-
amples (29)−(30), the expression it’s none of one’s business performs the function of a
disclaimer that prepares the hearer for possible discomfort in the prospective turn,
and a signal indicating that interpersonal boundaries have been overstepped in the
retrospective correlative. Compare (27)–(30):

(27) “Is now a good time?” She heard, (28) “Let us confer with Mr. Huggins
and looked up to find Joe Roth and Mr. Young.”
standing over her. “Look, Marlowe, this is not the
“For what?” time.” (R. Chandler. The Long
“Expenses.” (M. Keyes. Last Chance Goodbye);
(29) “Fintan, I have to ask you some- (30) “And who gets the five pesos, eh?”
thing. It’s none of my business, but “That’s none of your business.”
I’m going to ask anyway. Have you (G. Greene. The Power and the
had an HIV test recently?” Glory).
“Tara, you’re overreacting.”
(K. Keyes. Last Chance Saloon).

Although correlations between other types of prospective and retrospective commu-

nicative moves referring to (in)appropriateness are not as explicit as in Examples
(27)−(30) above, the links between most preparatory and responding moves can be
traced through the same or cognate lexemes, as can be illustrated by Examples
(31)−(34) (see Table 4.2):
In Example (31), the speaker prepares the interlocutor for subsequent role inap-
propriateness by classifying the addressee as a stranger. In Example (32), the same role
is mentioned as an avoidance strategy to justify the hearer’s refusal to answer the ques-
tion. Similarly, the cognate lexemes indiscrétion and indiscret in Examples (33)−(34)
underline the preventive-corrective relationship between the symmetrical turns.
 Maria Sivenkova

Table 4.1  The symmetry of some prospective and retrospective moves related to appro-

Prospective temporal (in)appropriateness: Retrospective temporal (in)appropriateness:

Я некстати? − ‘Am I disturbing you?’ Ты некстати – ‘You are disturbing me’

Avez-vous le temps? – ‘Do you have time?’ Je n’ai pas le temps – ‘I do not have time’
Tu es très occupé? – ‘Are you very busy?’ Je suis très occupé – ‘I am very busy’

Prospective interpersonal (in) Retrospective interpersonal (in)

appropriateness: appropriateness:

It’s none of my business, but That’s none of your business

I am not prying, but You are prying

Table 4.2  The symmetry of prospective and retrospective moves related to appropriateness

Prospective role (in)appropriateness Retrospective role (in)appropriateness

(31) “I really should not be discussing this (32) “What can you tell about Philip Cas-
with a stranger, but it is too exciting to tle?”
keep to myself. <...>” (S. Sheldon. If “Things I’d better not tell strangers.”
Tomorrow Comes). “Sorry.” (K. Vonnegut. Cat’s Cradle).
Prospective interpersonal Retrospective interpersonal
(in)appropriateness: (in)appropriateness:
(33) Alcacer: Sans indiscrétion, qu’y a-t-il (34) − Combien tu pèses?
dans cette sacoche, dont vous vous as- − Tu es indiscret (D. Nerincx. Anach-
surez tout le temps? (H. le Monther- ronismes);
lant. La Mort qui fait le trottoir); (34) “How much do you weigh?”
(33) Alcacer: Without indiscretion, what is “You are indiscreet.” (D. Nerincx.
there in that bag, which you are check- Anachronisms).
ing all the time? (H. le Montherlant.
Death as a Streetwalker).

To conclude, the symmetry of prospective and retrospective references to

(in)appropriateness of an utterance highlights two main functions of appropriateness
explication in dialogical discourse: preventive, designed to avert the deterioration of
interpersonal relationships (interpersonal appropriateness), avoid role conflicts (role
appropriateness), prevent untimely interactions in a wrong place (temporal and loca-
tive appropriateness); and the corrective function available for dialogue participants to
remedy inconsistencies that may arise in any of the above parameters of their interac-
tions. The finding that corrective function prevails over preventive one suggests, in my
view, potential for enhancing the effectiveness of interpersonal communication.
Appropriateness in interpersonal communication 

5. Cross-cultural differences in English, Russian, French

As this research shows, the communicative moves checking on appropriateness/estab-

lishing inappropriateness display more similarities than differences in the three lan-
guages under study. The similarities lie in the presence of the same set of main sub-
classes of examples and similar quantitative data. As far as language-specific features
of the conversational turns are concerned, they are related to the surface structure of
several types of communicative moves and certain quantitative and qualitative differ-
ences that become obvious if a more detailed analysis of the four main types of refer-
ences to (in)appropriateness is undertaken. Several such differences are briefly dis-
cussed below.
Among my Russian examples of prospective interpersonal (in)appropriateness an
important role belongs to the conversational routine Если не секрет – ‘If it is not a
secret’ signalling possible violation of interpersonal boundaries, which is often incor-
porated in rather blunt questions presenting the danger of face loss on the part of the
addressee and/or speaker. Compare (35):

(35) Матушкин: Если не секрет, (35) Matushkin: If it is not a secret,

какую должность вы считали what position would you consider
бы достойным для себя worthy for you to get? Come on.
получить? Ну? Только честно. Be honest.
Зарубин (после паузы): Вашу Zarubin (after a pause): Yours.
(В. Арро. Сад). (V. Arro. The Garden).

In contrast to the Russian Eсли не секрет present in about 40 per cent of prospective
checks on interpersonal appropriateness, the frequency of the corresponding conver-
sational turns If it is not a secret, Si ce n’est pas un secret in the other two samples is
much less significant (approx. 5%).
Another important way to acknowledge interpersonal inappropriateness in the
same subsample is to use imperatives of the type Простите за нескромность/
бестактность/нескромный вопрос (‘Forgive [my] immodesty/tactlessness/immod-
est question’) serving as disclaimers to the speaker conscious of the dangers related to
interpersonal inappropriateness (45% of Russian checks on interpersonal appropriate-
ness). The same function − mitigation of possible negative effect of an imposing
question/statement on the addressee − seems to be performed by means of such dis-
claimers as Sauf votre respect (‘With all due respect’), Sans indiscrétion (‘Without indis-
cretion’) in the French sample (approx. 60%) and several types of but constructions in
the English sample: It’s none of my business, but..., No offense, but..., It’s hardly any con-
cern of mine, but... (approx. 80%). As my data suggest, the Russian clichés appear to be
more direct in accepting responsibility for the problems arising in discourse (tactless-
 Maria Sivenkova

ness, immodesty), whereas the corresponding English and French conversational rou-
tines convey the idea in a more indirect way.
Among dialogue excerpts with communicative moves featuring temporal appro-
priateness, there is a difference in the French sample related to the syntactic structure
of such conversational turns: along with the checks on availability in the form of a
question As-tu une minute? − ‘Do you have a minute?’ there are also statements regard-
ing the addressee’s availability that make the communicative move sound more asser-
tive by leaving the interlocutor less chance to avoid the forthcoming interaction
(e.g. Vous avez quelques instants à m’accorder – ‘You have several moments to spare me’,
Vous avez bien une minute – ‘You do have one minute’). They comprise about 15% of
prospective checks on temporal appropriateness and are absent from the Russian and
English samples.
As far as retrospective communicative moves are concerned, an interesting differ-
ence has been found in the domain of role inappropriateness in the Russian sample.
There are numerous examples of the inappropriateness of communicative roles (about
35% of retrospective role inappropriateness explication in Russian) in which the more
powerful interactant deprives the addressee of the right to speak (a parent vs. a child,
a higher-status participant vs. a lower-status participant, etc.). Compare (36):

(36) Леночка: Есть и другие (36) Lenochka: There are other plea-
удовольствия, мама. sures, Mother.
Олег: Барахло покупать. Oleg: To buy junk.
Клавдия Васильевна: Когда Klavdija Vasil’evna: When elders
старшие разговаривают, тебе are talking, you’d better keep
лучше помолчать, Олег. quiet, Oleg. (V. Rozov. In Search
(В. Розов. В поисках радости). of Joy).

The correlatives in the other two samples Tais-toi/Je te prie de te taire – ‘Shut up/I ask
you to shut up’, Shut up (and be polite!) differ both qualitatively and quantitatively: they
comprise an insignificant part of the total number of English and French examples of
inappropriateness (5%) and seem to be related to interpersonal rather than role inap-
propriateness. This finding may indicate that there are different mechanisms of dia-
logue termination in the three cultures under study, with role inappropriateness viewed
as a sufficient reason to prevent an interactant from further participation in dialogue
in Russian, and other rules regulating dialogue completion in English and French.
Based on the data, I conclude that there is a vast potential for cross-cultural stud-
ies related to the phenomenon of appropriateness in discourse. I subscribe to the view
that by comparing various conversational routines across languages, one may discover
important cultural scripts that would facilitate cross-cultural communication and for-
eign language teaching (Wierzbicka 2003).
Appropriateness in interpersonal communication 

6. Conclusion

This paper has attempted to show that such elements of context as interpersonal and
role relationships between interactants, time and place of an interaction not only serve
as important theoretical premises casting light on the nature of human communica-
tion, but are also routinely negotiated by interlocutors in one of two possible ways:
prospectively (the speaker attempts to pave way for an appropriate speech act or re-
duce the negative impact of her/his inappropriate contribution) and retrospectively
(the hearer rejects the speaker’s previous utterance as inappropriate). The correlations
between prospective and retrospective communicative moves appear to provide sup-
portive evidence for the conclusion that the existence of prospective checks on tempo-
ral, locative, interpersonal and role appropriateness is an influential factor in enhanc-
ing the effectiveness of interpersonal communication. The variety of conversational
turns related to various types of (in)appropriateness underlines the versatility of the
phenomenon, with two most numerous and varied types of references − interpersonal
and role appropriateness − indicating the priority of social and personal relationships
in interpersonal communication.
The outlook of this study is to investigate the phenomenon of appropriateness in
other discursive spheres (e.g. computer-mediated communication, political discourse)
and communicative cultures, which would show how universal such references to ap-
propriateness in dialogical discourse are, and would allow discovering related cross-
cultural and interdiscursive differences. Besides, further research can be undertaken
to appreciate the role of such meta-communicative maneuvers among other contextu-
alization cues on which interlocutors draw to negotiate their way though interactions
(Gumperz 1982: 131). In addition, by taking into consideration other contextual pa-
rameters (e.g. deference, group identity, activity type) new findings in the area of inter-
actional grammar can be obtained, allowing interlocutors to improve communication
skills and avoid failures in communicative encounters.


Baxter, J. (ed.). 2006. Speaking Out: The Female Voice in Public Context. Basingstoke: Palgrave
Buck, R.A. 1997. “Towards an extended theory of face action: Analyzing dialogue in E.M. For-
ster’s A Passage to India” Journal of Pragmatics 27 (1): 83–106.
Burton, D. 1980. Dialogue and Discourse: A Sociolinguistic Approach to Modern Drama Dialogue
and Naturally Occurring Conversation. London: Routledge.
Clark, H.H., and Carlson, T.B. 1982. “Hearers and speech acts” Language 58 (2): 332–371.
Coates, J. 2003. Men Talk: Stories in the Making of Masculinities. Oxford: Blackwell.
Edmondson, W., House, J. 1981. Let’s Talk and Talk about it: A Pedagogic Interactional Grammar
of English. Munich: Urban and Schwarzenberg.
 Maria Sivenkova

Fetzer, A. 2004. Recontextualizing Context: Grammaticality Meets Appropriateness. Amsterdam/

Philadelphia: John Benjamins.
Gerhardt, U. 1981. “Toward a Critical Analysis of Role” Social Problems 27 (5): 556–569.
Goffman, E. 1981. Forms of Talk. Oxford: Blackwell.
Goldschmidt, M. 1998. “Do me a favor: A descriptive analysis of favor asking sequences in
American English” Journal of Pragmatics 29 (2): 129–153.
Gumperz, J.J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.
Karasik, V.I. 2002. Jazyk social’nogo statusa [Language of Social Status]. Moscow: Gnozis.
Kobozeva, I.М. 2000. Lingvističeskaja semantika [Linguistic Semantics]. Moscow: URSS.
McGregor, W. 1995. “The English ‘Tag Question’: A New Analysis, is(n’t) it?” In On subject and
theme: a discourse functional perspective, Hasan R., Fries P.H. (eds), 91–121. Amsterdam,
Philadelphia: Benjamins.
Schegloff, E. 1980. “Preliminaries to Preliminaries: ‘Can I Ask You a Question?’” Sociological
Inquiry 50 (3–4): 104–152.
Schegloff, E.A., Ochs, E. and Thompson, S.A. 1996. “Introduction”. In Interaction and Grammar,
E. Ochs, E.A. Schegloff and S.A. Thompson (eds), 1–51. Cambridge: Cambridge University
Simpson, P. 1996. Language Through Literature: An Introduction. Lodon: Routledge.
Searle, J. 1969. Speech acts: an essay in the philosophy of language. Cambridge: Cambridge Uni-
versity Press.
Stubbs, M. 1983. Discourse Analysis: The Sociolinguistic Analysis of Natural Language. Chicago:
University of Chicago Press.
Wierzbicka, A. 2003. Cross-Cultural Pragmatics: The Semantics of Human Interaction. 2nd ed.
Berlin/New York: Mouton de Gruyter.

Literary sources

Anouilh, J. 1988. L’Arrestation. Paris: La Table Ronde.

Anouilh, J. 1992. Eurydice; Roméo et Jeannette. Paris: La Table Ronde.
Apatow, J. The Cable Guy [Electronic resource]. Mode of access:
Aragon, L. 1992. Aurélien. Paris: Gallimard.
Arro, V. 1987. Koleja. Leningrad: Sovetskij pisatel.
Benoit, P. 1973. Le lac salé. Paris: Albin Michel.
Calef, N. 1990. Ascenseur pour l’échafaud. Moscou: Vysshaja Shkola.
Chandler, R. 1988. The Long Goodbye. New York: Vintage Books.
Clavel, B. 1997. La saison des loups. Paris: Robert Laffont.
Erdman, P. 1978. The Billion Dollar Sure Thing. New York: Pocket Books.
Fowles, J. 1989. Daniel Martin. London: Picador.
Galitch, A. 1999. Matrosskaya tishina. Ekaterinbourg: U-Faktoria.
Greene, G. 2003. The Power and the Glory. London: Penguin Classics.
Grisham, J. 1999. The Testament. New York: Island Books.
Guilloux, L. 1992. Le Sang Noir. Paris: Gallimard.
Hailey, A. 1991. The Evening News. London: CORGI Books.
Kelly, C. 2003. Just Between Us. London: HarperCollinsPublishers.
Appropriateness in interpersonal communication 

Keyes, M. 2000. Last Chance Saloon. London: Penguin Books.

Krentz, J. A. 1993. Wildest Hearts. London/Toronto/Sydney: Pocket Books.
MacDonald, R. 1996. The Far Side of the Dollar. New York: Vintage Books.
Maugham, W. S. 1974. Three Comedies of W. Somerset Maugham: The Circle, Our Betters, The
Constant Wife. New York: Pocket Books.
Montherlant, H. 1987. La Mort qui fait le trottoir. Paris: Gallimard.
Nerincx, D. Anachronismes [Electronic resource]. Mode of access:
Nikonov, A. Podkravshijsja nezametno [Electronic resource]. Mode of access:
Oldenbourg, Z. 1987. Les Amours égarées. Paris: Gallimard.
Parks, A. 2003. The Other Woman’s Shoes. London: Penguin Books.
Rozov, V.S. 1983. Izbrannoe. Moskva: Iskusstvo.
Sagan, F. 1980. Les merveilleux nuages. Paris: Julliard.
Sheldon, S. 1988. If Tomorrow Comes. New York: Warner Books.
Sheldon, S. 1988. Rage of Angels. New York: Warner Books.
Vonnegut, K. 1988. Cat’s Cradle. New York: Laurel.
Wallace, I. 1964. The Man. New York: Simon and Schuster.
Filling the German vorfeld in written
and spoken discourse*

Augustin Speyer

The sentence-initial position (vorfeld) in German is filled in accordance with

discourse structural consideration. Several types of elements compete for this
position. Their distribution can be modelled by Stochastic Optimality Theory.
It is filled in general by phrases that have at least one of the following functions:
scene-setting elements, contrast, topic. In conflict cases the functions are ranked
scene-setting >> contrast >> topic. In dialogic discourse other considerations
play a role, too, such as the explicit marking of origo (deictic center) and
rhetorical relation. Two additional constraints, dann-vorfeld and subject-
pronoun-vorfeld, can be fit in; the ranking including those would be dann,
scene-setting >> subject-pronoun, contrast >> topic. After insertions, the re-
introduction of the topic is sensitive to which of the participants started the
insertion: topics re-introduced by another participant than the one starting the
insertion pattern rather with contrastive elements.

1. Introduction

In the history of German linguistics, the field model of German clause structure has
proven to offer a remarkably accurate description. The field model in the form as it
can be found in introductory works such as Grewendorf et al. (1987) is schematized
in Figure 1.

* This paper is the extended version of a paper presented at the conference Linguistic Evidence
2 (February 2006, Tübingen, Germany) and the conference Organization in Discourse 3: The In-
teractional Perspective (August 2006, Turku, Finland). I want to thank the audience members and
three anonymous reviewers for their extremely helpful comments, but especially Keelan Evanini,
Irene Rapp and Joel Wallenberg for their invaluable help. All remaining errors are mine.
 Augustin Speyer

Linke Rechte
Satzklammer Satzklammer
(Vorvorfeld) Vorfeld Mittelfeld Nachfeld

– finite verb – rest of verbal complex**

– complementizer – the entire verbal complex
– coordinators 1 phrase n phrases n phrases (?)
– Left-dislocated (right-dislocated
material material)

Figure 1.  Field model

We are interested mostly in the vorfeld. There are no syntactic constraints on what can
stand in the vorfeld (apart from the requirement that it is normally only one phrase),
such that we can conclude that the decision about what ultimately stands in the vorfeld
is due to discourse structural requirements. I have argued elsewhere (Speyer 2008)
that vorfeld-movement can be modelled accurately as constraint interaction in the
sense of Optimality Theory (OT; Prince & Smolensky 1993). Vorfeld-movement in
declarative clauses (nota bene: I do not refer to operator-driven wh-movement in ques-
tions) is not necessarily a process that takes place in narrow syntax, as the semantic
effects on Logical Form (LF) are negligible,1 but seems to be a surface-oriented phe-
nomenon, possibly taking place in the early stages of Phonetic Form (PF). If this is so,
it is legitimate to use a surface-oriented approach, such as Optimality Theory. Section
2 briefly summarizes the vorfeld-facts in written discourse as exposed in Speyer (2008).
Section 3 applies these findings to dialogical, spoken discourse, thus investigating how
speaker interaction influences the question of vorfeld-movement. A short summary
ends the paper in Section 4.

** The whole verbal complex is presumably generated in clause-final position. If the left sen-
tence bracket is already occupied by a complementizer (which presumably is also generated there),
no part of the right sentence bracket can move. If the left sentence bracket is empty, the finite part
of the verb form is moved there; if the verb form is only one word, the verb form as a whole moves
there. The left sentence bracket cannot be left empty. The left sentence bracket corresponds to C,
the vorfeld corresponds to Spec, CP in generative terms (den Besten 1977; Vikner 1995).
1. Preferred readings of scopally ambiguous sentences arise independently of whether one
quantified phrase is in the vorfeld or not. It is true that Alle Studenten haben ein Buch gelesen
(all students read one book) is ambiguous between an ∃ > ∀ reading and an ∀ > ∃ reading,
whereas Ein Buch haben alle Studenten gelesen (One book all students read) strongly encourages
the ∃ > ∀ reading, but the same goes for the vorfeld-less version dass ein Buch alle Studenten
gelesen haben. So the effects are independent of vorfeld-movement.
Filling the German vorfeld in written and spoken discourse 

2. Written discourse

I report here on the ‘second corpus’ used for Speyer (2008). The corpus consisted of
501 V2-declarative sentences of subliterary prose (newspaper articles, essays in con-
cert programs, essays for reading on the radio). These genres were chosen because they
all represent a middle stylistic level (what you might call ‘utility prose’, texts that are
produced for a special, ephemeral occasion and are therefore unlikely to delve into
literary refinement, but are at the same time aimed to be easily readable), and because
the texts are easily obtainable. The questions of concern for us are: what can stand in
the vorfeld in the first place, and what are the preferences?

2.1 What do we find in the vorfeld?

Taking only sentences in which the vorfeld is filled with a referential phrase (405 tokens
of 501),2 82 per cent (364 tokens) have one of the following three types of element in
the vorfeld:
– topics (discourse-old entity; ‘what the sentence is about (see e.g. Strawson 1964;
Reinhart 1982), basically coextensive with Center (see Walker, Joshi & Prince
1998; ex.(1)),
– contrastive elements, i.e. elements that stand in a poset relationship (that is, a par-
tially ordered set relationship, such as subset, but trivially understood as mere set
membership, see e.g. Prince 1999) to a set in the discourse universe that is evoked
either by this first mentioning of one of its members, or that have been evoked
earlier (see ex. (2)). The set can also be referred to exhaustively (e.g. first sentence
in ex. (2)),
– scene-setting elements, i.e. elements that specify the temporal and local situation
under which the proposition is evaluated (ex. (3), see Jacobs 2001).
(1) [topic Verteidigungsminister Peter Struck (SPD)] hat gestern
defence-minister Peter Struck (SPD) has yesterday
sein Sparprogramm bekannt gegeben. [topic Er] sieht darin
his cut-expense-plan known given he sees therein
auch einen Schritt zur Reform der Bundeswehr
also a step to-the reform of-the federal army
‘Minister of Defence Peter Struck (SPD) proposed his program for cutting
expenses yesterday. He sees it also as a step towards a reform of the Federal
Army.’  (StZ 1,1–2)

2. This equals 73 per cent of the total number of V2-sentences (364 of 501).
 Augustin Speyer

(2) [contr. Bisherige sozialdemokratische Vorzeigeminister] wollen

Former social-democrat present-ministers want
nicht mehr über sich verfügen lassen. [contr. Clement]
not more over themselves order allow Clement
verabschiedet sich, [contr. Struck] lehnt den Posten
takes-leave himself Struck declines the post
des Außenministers ab(...)[contr. Schröder] selbst hat
of-the foreign minister ptc. Schröder himself has
eine andere “Lebensplanung”. [contr. Manche] werden
another life-plan. Some become
gar nicht mehr genannt
ptc. not more mentioned
Set M:M= Bisherige soz.dem. Vorzeigemin.; M = {..., Clement, Struck,