Dialogue Analysis 2000: Selected Papers from the 10th IADA Anniversary Conference, Bologna 2000 [Reprint 2011 ed.] 9783110933253, 9783484750258

Citation preview

Beiträge zur Dialogforschung

Band 25

Herausgegeben von Franz Hundsnurscher und Edda Weigand

Dialogue Analysis 2000 Selected Papers from the 10th I ADA Anniversary Conference, Bologna 2000

Edited by Marina Bondi and Sorin Stati

Max Niemeyer Verlag Tübingen 2003

Bibliografische Information der Deutschen Bibliothek Die Deutsche Bibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.ddb.de abrufbar. ISBN 3-484-75025-1

ISSN 0940-5992

© M a x Niemeyer Verlag GmbH, Tübingen 2003 http://wmv.niemeyer.de Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Printed in Germany. Gedruckt auf alterungsbeständigem Papier. Druck und Einband: Digital PS Druck AG, Birkach

Table of Contents

Preface

VII

Part I - Bologna 2000 Round Table Franz Hundsnurscher, Introductory Remarks Frantisek Danes, A Retrospect and a Prospect of Dialogue Studies Edda Weigand, Dialogue Analysis 2000: Towards a Human Linguistics Robert Maier, A Process and Structure Conception of Dialogue Valerij Dem jankov, Strategies of Understanding in Dialogue Bernd Naumann, Chaotic Dialogues. Can Ideas Formulated in Complexity Theories be Applied to Dialogue Analysis? Karin Ajimer, Discourse Particles in Contrast John Sinclair, Can We Have a Conversation with a Computer? Malcolm Coulthard, What Did They Actually Say? A Forensic Linguist's Evaluation of Police Evidence Svetla Cmejrkovä, Media Dialogue as a Genre of Public Oral Discourse

3 9 15 29 39 51 63 79 93 107

Part II - Papers Carla Bazzanella and Alberto Baracco, Misunderstanding in IRC (Internet Relay Chat) Massimo A. Bonfantini, Augusto Ponzio, Susan Petrilli, Trialogo sui dialoghi . . . .

119 133

Regine Borderie, Dialogues de groupe: Balzac, Zola

145

Adeline Cattani, II disputator cortese e il disputator polemico

155

Laurentia Dascälu Jinga, Metadiscursive Triads

163

Marianne Doury, Argumentation et "mise en voix". Les discours quotidiens sur Γ immigration Paula Fatur-Santos, Interacting with a Frontal Lobe Patient: What about Pragmatics? Gerd Fritz, Dialogical Structures in 17th Century Controversies Renata Galatolo, Les strategies de changement de footing dans le temoignage commun au tribunal: une resource pour la construction de credibilite

173 185 199 209

VI

Table of Contents

Julia Gillen and Angela Goddard, Medium Management for Beginners. The Discursive Practices of Undergraduate Novice Users of Internet Relay Chat, Compared with Those of Young Children Using the Telephone

219

Luis a G. Granato and Maria Leticia Möccero, Elicitation Utterances in Written Dialogues

231

Luca Greco, Activites de description et pratiques de cadrage dans l'interaction medicale par telephone

243

Cornelia Hie, Parenthetically Speaking: Parliamentary Parentheticals as Rhetorical Strategies

253

Liliana Ionescu-Ruxändoiu,

Politeness Strategies in Oral Dialogues

265

Ludmila Kastler, Accords et desaccords: une dialectique communicative

271

Jiri Kraus, Argumentation in Czech Political Debates

277

Gerda Lauerbach,

Opting Out of the Media-Politics Contract. Discourse Practices

in Confrontational Television Interviews Margareta Manu-Magda,

283

Ein Sonderfall des sozialen Dialogs: der interethnische

Dialog

295

Mirka Maraldi et Anna Orlandini, Exclamatives, interrogatives-exclamatives et le jeu de la negation Michel Marcoccia,

305

La politesse dans les forums de discussion sur l'lnternet. Regies

externes, manifestations discursives et commentaires metacommunicatifs

315

Liana Pop, Actes vs operations. Vers de nouveaux outils dans l'analyse du dialogue Anne Salazar Orvig, Elements pour une analyse de la connivence dans le dialogue

327 339

Donna Shalev, Yes (and No) in Ancient Literary Greek

351

Milena Srpovä, Dialogue entre cultures differentes: le role de l'ethos collectif . . . .

361

Grazieila Tonfoni, Information Transport through Dialogue. Α System for Reducing Fuzziness due to Culture and Context Shifts

371

Elda Weizman, News Interviews on Israeli Television: Normative Expectations and Discourse Norms

383

Larissa Wunderlich, Treaties: A Comparative Analysis of a Complex Dialogic Action Game

395

Preface

This volume has a special position in the IADA Dialoganalyse series as it collects the proceedings of a conference that celebrated the tenth anniversary of the Association in the year 2000. As organizers of the conference we felt the need to make a choice between a focus on past achievements and future research in the discipline. On the one hand we could review and assess the work of previous conferences and workshops organized by the Association, to include a history of the 250 membership and of the countries they represent. On the other hand we could outline possible trends for future research in the first decade of the new century and millennium. The final decision opted for a middle path: "The Tenth Anniversary might be an apt occasion to look both backwards and forwards - what has occupied us in the past ten years and what will be expected of us and has to be done in the future" (Hundsnurscher). The Association has often been characterized as - and sometimes criticised for - lacking in both a) a set of theoretical principles that guide and inform research practices, and b) a uniform methodological tradition of "regulae ad directionem ingenii". Even though we do not yet have a definitive set of "Bologna theses" these matters were discussed at length in the Round Table sessions on general, theoretical and methodological issues. The discussions are synthesized by Edda Weigand in her contribution to this volume. The choice of what we have called a middle path - "a retrospect and a prospect" (Daneä) - is reflected in most of the contributions, though in many of them there is a greater focus on past achievements. Many of the themes dealt with in this volume are well known to us and yet often presented in a new light ranging from Maier's "structural view of language", to Dem'jankov's "strategies of understanding in dialogue", to Aijmer's comparative analysis of pragmatic particles in Swedish and English. Basic concepts, such as communication itself, are also explored from different perspectives by some contributors: John Sinclair, for example, while focusing on "conversation with a computer" raises a number of issues concerning over-simple models of communication itself and highlights some of the key features of human conversation: participants have their own personal agendas, they maintain a coherent stance, they monitor the discourse as it takes shape and express evaluation in manners that are simultaneously collaborative and competitive. Similar concerns, with a literary slant, are expressed by Bernd Naumann (organizer of the Erlangen Workshop). Methodologically the dominant approaches are those of discourse analysis and conversation analysis. A new interest emerges in the application of dialogue analysis to areas of dialogic interaction in specific contexts. This is the case in forensic linguistics, as exempli-

VIII

Preface

fled by Coulthard's analysis of different types of dialogue in police interviews, and similarly in Cmejrkovä's study of the role of interviews in media discourse: "Interviewing a person is a highly individualised practice. There is no simple hunting method. It depends whether you are hunting a lion or catching a butterfly". This volume includes contributions that address both traditional areas of dialogue analysis such as politeness, and more recent areas of interests such as argumentation. A great many of the papers explore specific genres or communicative situations: attention is drawn to different types of interaction in the field of media communication (with particular emphasis on computer-mediated interaction and news interviews) and institutional and professional interaction. Other contributions deal with fictional dialogue, scientific controversies, and pathologies of speech. The (interpersonal and (inter)cultural dimensions of interaction also emerge as areas of research of increasing interest. Contributions to the conference were numerous and the majority are published in this volume. Those that are not included have been or will be published elsewhere. Although published with some delay, we hope that these proceedings will be of interest to the many disciplines and perspectives involved in the study of dialogue. Within these pages, the reader will not find a singular approach to the study of dialogue, an inevitable limit of a publication of this kind, but they will find a clear picture of the existing "state of the art", and for this the editors are grateful to all the contributors.

Bologna, July 2002

Marina Bondi Sorin Stati

Part One Bologna 2000 Round Table

Franz Hundsnurscher

Introductory Remarks

Ten years ago the International Association for Dialogue Analysis was founded by Sorin Stati here in Bologna. This was a grand idea and a courageous enterprise, and we all have reason to be grateful to Professor Stati, who has taken it into his hands and put it to the test. We all know the society has been a success. Over the years the society has brought together a great number of people from all over the world and has provided stimulus and encouragement to many students and scholars with an interest in dialogical matters. The tenth anniversary might be an apt occasion to look backward and forward - to discuss what has occupied us in the past ten years and what will be expected of us and has to be done in the future. When the Society was founded in 1990 here in Bologna, the pragmatic turn in linguistics was in full bloom and it looked as if the shift from monologistic speech act theory to dialogue theory would lead on to a paradigm change in linguistics altogether. In this last decade, at least to my mind - and I can only, of course, give my subjective view - the issues have indeed become clearer and one may well notice progress in the fields that have been covered by numerous papers given at our big conferences in Bologna 1990, Basel 1992, Paris 1994, Prague 1996 and Birmingham 1999 with six smaller conferences in between in Bologna, Toulouse, Lugano, Erlangen and Tel Aviv. At one of our pre-conferences in Bochum in 1988 Sorin Stati pointed out to me that in doing dialogue analysis we take our stand at the 'Carrefour de communication', right at the centre of human affairs, where the action is. Indeed it is from such a stance that we can take a broad outlook on the things going on in all directions and see how people get along with each other in the pursuit of their goals. Dialogue analysis certainly is a 'carrefour de linguistique' because from there we can watch the continuous flow of ideas and the emergence of new modes of linguistic thought in progress. As members of a society for Dialogue Analysis, people will not only expect us to be Masters of Dialogue, who set standards of how to converse with each other successfully and in an amiable way, but also to be able to tell others what dialogue is all about and how one should go about analysing, criticising and improving it. So one central point, to my mind, is and always will be 'Discours de la Methode' - the proper way of doing Dialogue Analysis. It is the age-old problem of how to reconcile Empiricism and Rationalism which is of crucial importance in dialogue analysis. On the one hand we have all sorts of authentic discourse which can be observed and documented by modern technical means - tape

4

Franz Hundsnurscher

recording and video devices - a n d then transcribed, scrutinised and interpreted in detail. On the other hand, we have a variety of traditional methodologies - the philological tradition, the structuralist tradition, the tradition of generative grammar and the tradition of ordinary language philosophy and on top of it the tradition of Ethnomethodology - but all these methodologies have been developed with a background of different theories of language and with specific problems in mind - most of them not exactly problems of dialogue. We may, of course make use of all these methodologies, as has been done in a variety of ways, but if Dialogue is to be our focus of interest we should not rely too much on old methods but be careful in developing a guiding methodology and keep on refining for the purpose of dialogue analysis. What we need is an interactive model of verbal communication where speakers talk to hearers that in their turn turn into speakers. Trivial as this many sound, if we really make this the focus of our research, it will have far-reaching consequences. Most of the work in dialogue research is being done by observation; this is of course, necessary because we still have to discover a lot of facts about conversations. The problem is what to do with the material once we have recorded it and transcribed it. We have to treat it as evidence for certain underlying patterns of verbal interaction that make up our communicative competence. A lot of work has been done in distinguishing and identifying such patterns and studying them in detail and hopefully we will arrive at a system to describe and analyse them on different levels of generality. These patterns will have their exemplifications in varying situations and will be combined and interwoven in authentic conversation. This is the center of work in DA, as I see it. One remarkable trend in the last decade is that work in dialogue analysis has had an effect on linguistics in general. If it is true that dialogue is the core of language, then a dialogical view of verbal communication is bound to yield new and essential insights into the fabric of language and this will have consequences in other fields of linguistic thought and research; I shall give a few examples where I find a dialogue-analytic perspective has yielded some interesting results. The focus of analysis in syntactic studies has been on the well-formedness of isolated sentences, and much insight has indeed been gained by the rigorous and explicit modelbuilding done by Chomsky and his followers. Although they are more interested in universal and neuro-biological aspects of language in general, many things have become clearer in detail about specific natural languages. Yet with respect to sentence complexity and sentence-combination the advantages of a dialogical view have to be realised. Most complex sentences, especially conjunctional clauses can be more plausibly explained as anticipatory queries of a dialogue partner:

Introductory Remarks

5

Spl "Monkeys do not go to heaven" Sp2 "Why is that so?" Spl "Well, they are not human". Sp2 "Now I see: Monkeys do not go to heaven because they are not human". Complex sentences are condensations of dialogue sequences. In classical Speech Act Theory working with the simple scenario of a speaker talking to a hearer, the main focus is on single initial speech acts: the hearer is only considered as one who is exposed to the speech act and passively lets it take effect on him. This is the reason why the special nature of sequence-dependent speech acts have not been given much attention; and hence speech act theory can only be considered as a first step in the analysis of verbal communication. The hearer as second speaker has his own repertoire of speech acts, so to react in a systematic way to what has been said, and this is what is expected from him by S p l . Sequencesensitive speech acts as second, third or fourth moves like agreeing, justifying, objecting, rejecting, insisting, admitting are indicative of the many devices that have been developed in natural languages for the dialogue game. Texts may be considered as complexes of sentences that are brought into a linear order due to the conditions of monologue inherent in the secondary system of writing. Texts can in many respects be reconstructed as dialogues with an interested listener. One has only to think of narratives where many things can be queried: how it all started, how it went on, why it took this turn or what happened next, what special things were involved, what the point of it was and so on. Written texts can be analysed and reconstructed on the basis of underlying dialogical patterns. Let us finally look at Rhetoric. From the start Rhetoric has been devised for the public speaker, and its monological characteristics have become even more pronounced when Ancient Rhetoric (Aristotle) turned into Literary Rhetoric (Cicero, Quintilian). What has been missing for a long time is a Rhetoric of Dialogue, working out how to engage in conversation, keep a good conversation going and unfold the various sides of a matter in a discussion, how to be cooperative and polite in conversation: all this needs more than just elocution. Dialogue analysis as we see from these examples is not only faced with the task of working in its own field, but is called upon to revise and re-conceive the linguistic concepts of other disciplines, too. If we look at the overall trends that have surfaced in the last decade, not only in the manifold contributions at our meetings but also at other congresses and in the international discussion on pragmatics, we could point out eight different domains with respect to subject:

6 -

Franz Hundsnurscher The first domain - the one with the most numerous contributions - is private conversation in all its facets: I shall not go into this, because this will probably be everybody's casual talk between meetings, in party conversation, social exchange at the cocktail bar, at table, at work, on a journey und so on as the conversational main stream.

There are seven more that I shall shortly comment on: -

Specific Dialogues in Institutions

-

Controversial discussions in society

-

Mass-Media-related dialogue

-

Computer and dialogue

-

Literary dialogues

-

Historical dialogue forms

-

Contrastive studies in culture-specific phenomena of conversation

Each of these domains exhibits a highly complex and fascinating world of forms and problems for analysis. Institutions like Courts of Justice, Administrative Bodies, Firms, Clinics, Churches, Universities and Schools etc. are prone to develop special hermetic forms of communication, that work under special conditions with institutionally defined purposes. The implications of these conventions are not easy to explain to the public, and by investigating these highly specialised domains of dialogue analysis we can do a good job in bringing light there and drawing public attention to some of the consequences. Quite remarkable is the interest that scholars of linguistic pragmatics take in political and social controversies e.g. about matters of ecology, of politeness or in precarious types of social intercourse connected with rumour, gossip and scandals in public and private life. The mass-media are gradually developing into an all-dominating form of communication. At the Tel Aviv Congress, members of IADA were welcomed by the Dean of the Faculty of Philosophy at the Hebrew University, Jerusalem. He made a remark that had a profound impression on me. He said: We all know that in our Western Society freedom of Speech for the individual is no longer the problem; we need no longer fight for it; we now have to fight for access to the stage and for freedom of the stage. Dialogue in the media and through the media is a real problem, much more than a decade ago. Just think of the multitude of talk shows, round tables, family comedies and other series where all kinds of dialogue are organised and put on stage under the conditions and for the purposes of the media. The consumer of these programs is immersed in dialogue, but is he taking part in the dialogue? The concept of dialogue underlying these performances is bound to have some aftermath with respect to the modes of talking in everyday life.

Introductory Remarks

7

We have at the moment a wave of enthusiasm that keeps us surfing in the Internet. Modem technical devices make it possible in theory for every person living on this planet to communicate with any other person and enter into conversation with them. New forms of dialogue are being created under these conditions and change the way people use language. A highly interesting branch of research in dialogue analysis is literary dialogue. Because these are prefabricated texts organised on certain conceptual and aesthetic principles it would be a mistake to treat them just as examples of normal dialogues and to analyse them according to the same categories and criteria as we do with spontaneous every-day conversation. Nevertheless, as linguists, we can learn a lot from the classical dialogue forms of Ballads, Drama, radio feature, conversations in novels, philosophical and didactic dialogues because the basic forms are worked out there in a stylised and precise way and dialogue strategies, forms of escalation and solution are brought to the point in an effective way. It is somehow amazing how little reflection one finds in scholarly work on literature concerning problems of dialogue typology and dialogue structure. In connection with the analysis of literary dialogues another interesting field of research has emerged: investigating the origin and development of dialogue types in a historical perspective. A new volume on this topic has just appeared in our series 'Contributions to Dialogue Research' with Niemeyer, Tübingen. Dialogues are in a special way mirrors of ways of life that undergo specific change under the social conditions of the age. Scarce as the sources may be, there are many old documents that show us how, e.g. conversion talks and didactic talks were performed, what course juridical procedures took, how theological dispute and scientific controversies were conducted, how saloon talk was organised. This is a very promising field that not only may shed light on our own dialogue conventions but can also make us better understand some conversational remarks we find in old literary texts. One will of course have to keep in mind the conditions of the codifying tradition every historical text is subject to, because everything written has been written down under certain conditions and for certain purposes and only behind this veil of literacy can we get a glimpse at reality. Investigations into culture specific principles and forms of conversation have been of great interest in the last ten years. It has been shown, for instance, that business talks between German and Norwegian partners have their problems. In one culture one listens and sometimes asks a simple question, whereas in the other culture one makes objections at every point and tries to clear it up on the spot, and if the other does not object, everything is supposed to be clear. John Gumperz has pointed out another interesting perspective. Cultural differences may be the origin of problems, e.g. between immigrants and local institutions; discrimination and frustration can be the result of different communicative conventions.

8

Franz Hundsnurscher

These, I am sure, are not all the trends that can be found in dialogue research. I hope that this short outlook on 10 years work in Dialogue Analysis is another source of encouragment for our Society and for our President to enter into another decade of successful work.

Frantisek Danes

A Retrospect and a Prospect of Dialogue Studies

1. Introduction: Dialogue analysis means science based on mutual understanding

A browse through the long series of the red volumes of Dialogue Analysis convincingly reveals an increasing "boom" in dialogue studies (and not alone in the framework of the IADA activities, to be sure). We find there a broad diapazon of topics (from the classical to the most up-to-date ones), of rather different kinds of dialogue, of different text genres (from research articles to essays and theoretical or methodological reflections), as well as a pleiade of authors from various countries, employing diverse methodological approaches. This does not suggest, however, that our work represents a variegated mosaic of contributions: on the contrary. Even a mere look at the titles of the particular volumes reveals a firm hand and prospective views of the organizers: Methodology of Discourse Analysis, Concepts of Dialogue (considered from the perspective of different disciplines). Future Perspectives of Dialogue Analysis. The two comprehensive volumes from the last Congress in Prague are divided into six thematic sections, presenting a number of topical and partly newly emerging thematic domains, such as "Dialogue and Institutions", "Dialogue in Politics", "Mass Media and Electronic Communication". The fact that two of the topics were taken up and discussed as general themes of the next two IADA sessions ("Rhetoric and Argumentation" and "Dialogue and Mass Media") witnesses the coherence of the IADA agenda. The thematic field "Dialogue" is certainly immense, with a rich inner differentiation. It comprises a considerable range of objectives, theories and methods as well as diversity of academic disciplines drawing upon the field and contributing to it. This state of affairs inevitably finds reflection in the scientific production of IADA. For all that, the scholars of our Association endeavour to take up mutual communication and arrive at mutual understanding - an inevitable condition for scientific progress. I would only subscribe to and emphasize the happy formulation of Edda Weigand (in the Preface to the Stati's Festschrift) that "beyond different scientific schools, DIALOGUE means science based on humanity and cross-cultural understanding".

10

Frantisek Danes

2. A critical examination of two authoritative methodological paradigms in present-day dialogue studies

Recently Jifi Kraus aptly remarked that the development of a field of research is largely contingent upon texts of authorities, who raise topical themes as well as methodological approaches. It would be easy to adduce names of authoritative scholars from particular currents in the broad field of discourse studies. But in the context of my talk I will touch only some of them, whose influential ideas deserve, in my view, to be critically discussed just at the present. It is hard to find any contemporary approach to discourse studies which does not more or less explicitly refer to the works by Goffman and Grice as well as to the later Brown and Levinson's theory of politeness, developing some ideas of the two pioneers.

2.1. Grice's Cooperative Principle and Conversational Maxims With the name of Grice two ideas are connected, namely the Co-operative Principle and Conversational Maxims. Grice regards the assumption of cooperation of the partners as the basis of any communication and presupposes that they never fail to attempt a coherent and effective dialogue. Nevertheless, this very strong presumption of the essential, if not unexceptional rationality of human behaviour is hardly acceptable for those who have enough experience from empirical research. Thus Marcelo Dascal (1998: 18) very aptly remarked that even though in many cases communication seems to be at odds with the old comfortable notions of cooperation and rationality, it is still possible, and he concludes that it is no longer feasible to base one's theory of communication upon unexamined principle of (instrumental) rationality. A similar argument formulated E. Weigand (1998: 39). But in spite of this counterevidence, some scholars obstinately try to rescue the idea of their master at all costs, even by means of tricky formulations. Thus Brown and Levinson (1987) defend the presumed universal validity of the Cooperative Principle maintaining that "the assumption of cooperative behaviour is actually hard to undermine, tokens of apparent uncooperative behaviour tend to get interpreted as in fact cooperative at a deeper level (5)". In a similar way the same authors defend Grice's Maxims: They admit that "the majority of natural conversations do not proceed in such a brusque fashion at all", "[...] one powerful and pervasive motive for not talking Maxim-vise is the desire to give some attention to the face". I find this to be a rather curious assumption: if polite conduct infringes Conversational Maxims, then politeness and the normal efficient course of

A Retrospect and a Prospect of Dialogue Studies

11

conversation are incompatible. Nevertheless, we are told that "even in such departures from the Maxims, they remain in operation at a deeper level" (95). Our intuition says that there must be something wrong in the premises of the argument and I will try to show their fallacies. In the set of Grice's well-known Maxims of Quality, Quantity, Relevance, and Manners I recognize the following problematic points. First, they have evidently a character of moral precepts or commandments such as "Do not lie", "Be not loquacious", "Behave decently" and the like, in essence a kind of didactic counsel rather than a scientific statement based on empirical investigation. Second, even when taken seriously, their application would meet with interpretational problems connected with the vagueness of expressions like

"Do not say less/more than required",

"Be relevant".

How could a

speaker plausibly decide what is in fact required or relevant to his partner(s)? Relevance is a gradual and changeable quality. In other words, since the whole field of "conversation" contains a very high number of varieties of the rather different types of "conversations", any attempt at a purely rational construction of generally valid primitive "Maxims" is doomed to failure. Consequently, also a presumed cross-cultural or universal validity of them appears as rather questionable.

2.2. B r o w n a n d L e v i n s o n ' s concept of Politeness Let us now take up the problem of politeness,

a highly important phenomenon of conduct

(not only of the verbal one, to be sure). It does not represent a new theme and for a long time it has been studied mainly by students of East-Asian and other "exotic" cultures. Nevertheless, politeness was drawn into the focus of attention of the wider scholarly audience due to Robin L a k o f f s essay from 1972 on the pragmatics of politeness, followed by contributions of G. Leech (1977: 1983) and of Brown and Levinson (1978: 1987). Among them, the last named approach gained wide publicity and politeness has become an enormously influential paradigm in discourse analysis. The politeness theory of Brown and Levinson is built primarily around Goffman's notion of face. He found inspiration in Dürkheim's dichotomy positive versus negative religious rites or rituals and retaining the ritualistic frame he introduced the notion o f f a c e , defining it as "an image of self delineated in terms of approved social attributes" (306). Following Goffman, the two authors define negative face as "the want of every adult member that his actions be unimpeded by others" and positive face as "the want of every adult member that his actions be desirable to at least some others". Certain kinds of interlocutor's acts intrinsically threaten face ("face threatening acts" is the central notion of the theory) and politeness

(again a negative and a positive one) is seen in the speaker's en-

12

Frantisek Danes

deavour to mitigate or minimize in communication the possible impact of such an act, by the use of different strategies (the employment of the term strategy in contradistinction to rule or maxim reveals an interactional approach). While that section of the book which offers a taxonomy of these strategies represents, in my view, the most valuable part of it, some points of the underlying theoretical assumptions appear disputable. First of all, our intuitive understanding of politeness comprises a far more extensive domain of the universe of conversation than that restricted sector of the "mitigation of facethreatening acts" or "strategic conflict avoidance". Moreover, we all feel "politeness" as something that is in essence of a positive nature. A severe criticism of "face" in this direction expressed R. Schmidt (1980): "This theory represents an overly pessimistic, rather paranoid view of human social interaction in language, viewing politeness as a response to threats of face rather than as an essentially positive phenomenon." The unsharp outlines of the field called politeness is also revealed by the terminological diversity: along with politeness we find in English courtesy, deference, tact, respect, regard, considerateness and some others, focusing on this or that aspect of the many-sided phenomenon. As J. Hoffmannova (1967) aptly suggested, we have to assume a rich set of qualities that dominate the whole behaviour of participants in interaction. In this context, politeness appears as a bundle or a fuzzy complex of qualities rather than one simple quality item. In general, it appears to me advisable to work, in our analytical practice, with complexes and subcomplexes of particular qualities (also in view of the certain vagueness of their identification or specification) on the principle "which goes with which" rather than with isolated units. Brown and Levinson's model has found several followers and is widely applied. E.g. G. Myers (1989, etc.) tried to extend its principles to scientific texts, so that this kind of discourse would appear as a fundamentally dangerous and antagonistic endeavour (cf. Hyland 1998: 67). On the other hand, the approach suggested by G. Leech (1983) seems to me to reflect the field of politeness in a far less one-sided and distorted way. He proposed six Maxims of Politeness: tact, generosity, approbation, modesty, agreement, and sympathy, and accounts for their flexible hierarchy, changeable according to different discourse types and different cultures. Another disputable feature of the criticized model is its presumed universality, nevertheless, in fact, with a strong anglocentric bias. This point has found a number of critics, viewing it as inappropriate for speakers of several different cultures. The authors of the model, in their rejoinder (attached to the second edition of their work, 1987: 13) admit that their universalistic account might be thought as an "inexcusable cultural denudation, or worse, ethnocentric projection". But, they maintain, "despite the rich cultural elaboration, the core ideas have a striking familiarity". They are certainly right that very probably in all cultures the interactants more or less often mitigate their polemical speech acts (cf. the notion of "hedging"). But what I and others criticize and dismiss is the idea that a mere

A Retrospect and a Prospect of Dialogue Studies

13

fragment of the rich domain of politeness should be taken as the politeness, in the full and broad sense of the term. Also the choice of the expression face in connection with politeness deserves to be commented. Brown and Levinson mention that their selection of this metaphor was partly derived from the English folk usage. It is true that in English (as well as in other languages) this word occurs in a number of phraseological idioms (many of them were interestingly listed by Cmejrkovä in her essay "The stony face of a S f m x " (1993)). But that reading of face, which is relevant in our discussion, namely "self-respect, a good name, dignity, prestige", is in fact a translation of one Chinese idiom (even Goffman mentioned that he was following Chinese usage). This metaphorical expression seems to bear a mythical connotation, which I find not very appropriate for the European cultural sphere (in spite of the true statement by Karl Popper that from historical perspective a myth may contain important anticipations of scientific theory). At last I will touch on a problematic point of a philosophical nature. Brown and Levinson's model is built as a fully rational and closed system, and the course of discourse is expected to be a totally purposive and strategical activity. The authors adhere to Aristotle's "practical reasoning", which - like in standard logic - "guarantees inferences from ends or goals to means that will satisfy them" (54). But in the same way as in other spheres of human life, also in their communicative interaction people behave partly in non-rational ways and exhibit all their abilities (including emotion) in different ways and with variable results, without a guarantee of success. (Cf. the discussion of a similar approach of Grice above.)

3. Conclusion: Conflict or Cooperation? A plea for a humanistic stance

I have arrived at the concluding part of my paper. It is very difficult and risky to make prognosis of the future development of a scientific discipline. Thus I have to content myself with expressing some of my expectations, hopes and wishes. From the methodological viewpoint, it seems to be primarily desirable to avoid onesided, narrow, and unduly simplifying or reductionist approach, exaggerating one aspect only. I would advocate a complex, integrative view, in all cases and all the time keeping in mind the very rich and complicated contents and structure of the vast phenomenon of DIALOGUE. Studying language usage as a component of the complex human communication requires a holistic and process-oriented approach. One of such approaches may be seen in

14

Frantisek Danes

Edda Weigand's conception o f the "Dialogic Action Game". It is broadly, undogmatically and humanly based, with "human beings at the centre". I dare to characterize it as functional,

since it sounds as a reformulation o f Vilem Mathesius postulate o f linguistic

functionalism, namely to take systematically into account the speaker and the hearer, always to see them behind the words. This humanistic stance (as he called it) and the accent on understanding as the very aim o f communication is also in line with Weigand's "democratic, humane and civilized background", rejecting "the demagogic direction and the direction h o m o homini lupus " (1999: 65). She is certainly not alone with her opinion. Thus also de Beaugrande is convinced that "cooperative and constructive uses o f communication should prevail over confrontational and destructive ones". This dictum may also be seen as a reply to Sorin Stati's question (1998: 3) "Conflict or/and cooperation?" Let us remember that people have not only their faces but also their hearts.

References

Beaugrande, R. de (1996): New Foundations for a Science of Text. - Norwood, N.J. tmejrkova, S. (1993): The stony face of a Sfinx. - In: Dialoganalyse IV/1, 41-46. Tübingen: Niemeyer. Dascal, Μ. (1998): Types of polemics and types of polemical moves. - In: Dialoganalyse VI/1, 1534. Tübingen: Niemeyer. Goffman, Ε. (ed.) (1967): Interactional Ritual. - New York: Anchor Books. Grice, H. (1975): Logic and conversation. - In: P. Cole, J. Morgan (eds.) Syntax and Semantics, Vol. 3,41-58. New York: Academic Press. Hoffmannovä, J. (1997): Linguistics, style, and dialogue analysis. - In: Dialoganalyse V, 257-262. Tübingen: Niemeyer. Hyland, Κ. (1998): Hedging in Scientific Research Articles. - Amsterdam/Philadelphia: Benjamins. Lakoff, R. (1972): The pragmatics of modality. - In: P. Peranteau et al. (eds.) Papers from the Eighth Regional Meeting, 229-246. Chicago: University of Chicago. Leech, G. (1983): Principles of Pragmatics. - London: Longman. Myers, G. (1989): The pragmatics of politeness in scientific articles. - Applied Linguistics 10, 1-35. Schmidt, R. (1980): Review of "Questions of Politeness: Strategies in social interaction". RELC Journal! I, 100-114. Stati, S. (1998): Le texte argumentativ - In: Dialoganalyse VI/1,3-14. Tübingen: Niemeyer. Weigand, E. (1998): Emotions in dialogue. - In: Dialoganalyse VI/1, 35-48. Tübingen: Niemeyer. Weigand, E. (1999): Rhetoric and argumentation in a dialogic perspective. - In: E. Rigotti (ed.) Rhetoric and Argumentation, 53-70.Tübingen: Niemeyer.

Edda Weigand Dialogue Analysis 2000: Towards a Human Linguistics

1. Dialogue Analysis 2000 and the scientific claim of the new millennium

A decade of Dialogue Analysis within the framework of our International Association reminds us to reflect on the results we have achieved. Moreover, after more than two millennia of classical Western thinking, we are called on to relate our analyses to general changes in scientific theorizing in other disciplines. The year 2000 invites us not only to look back but also to give our science future guidelines. I am going to take up these claims: sketching the state of the art in Dialogue Analysis with respect to the background of Western science and trying to redefine linguistics as a genuine human science. Looking around at other disciplines beyond the limits of linguistics, such as physics, biology, neurosciences, there seems to prevail a general methodological rethinking, a departure from classical theorizing to a new way of addressing the object-of-study. The object in science has always been the complex. The ways of addressing the complex however have changed significantly in time. At the beginning of the new millennium, it is indeed this question of how to address the complex which apparently calls for a new response. Classical theorizing as it has prevailed since Aristotle and as it still pervades Western thinking has never really addressed the complex. Classical thinking starts from the premiss that there is only one way of explaining the complex, namely by reduction to rules. If an explanation claims to count as a theory it has to be based on a closed rulegoverned system. This type of classical theorizing is characterized by total abstraction from complexity, i.e., it starts with methodology. Finally it is time to free ourselves from the underlying methodological fallacy and to focus again on our complex object and the way it functions. It is the object-in-function which will tell us how to construct methodology. It might seem strange that it is precisely the so-called exact natural sciences which can give us some ideas for how to proceed. Our complex object-of-study is not at all defined by order and definite rules. It contains a Principle of Uncertainty, analogous to the Principle of Uncertainty in quantum physics, which calls to mind that our object represents a mix of order and disorder, general regularities and individuality, even chance (cf. Weigand 2002a). The question is how the complex and the simple are interrelated, or, to express it with Gell-Mann (1994) with reference to physics, how the quark, the elementary

16

Edda Weigand

particle, and the jaguar, the perfect wildcat, are interrelated. Indeterminacy of meaning is present at the very outset and may increase exponentially to chaos as Prigogine (1998) taught us with respect to modern chemistry. Neuroscience unmasks 'Descartes' error' in demonstrating that different human abilities are intrinsically integrated (cf., e.g., Damasio 1994). Recent research in neurophysiology on so-called mirror neurons seems to confirm that perception and doing, namely perceiving and making a gesture, are signalled in the same way by the brain (Rizzolatti/Arbib 1998, Weigand 2002b). Integration seems to be a basic feature of every discipline, not only of linguistics where 'Integrational Linguistics' has already exposed the orthodox linguistic view as the language and communication myth (Harris 1981).

2. The state of the art in Dialogue Analysis

Against this background of a general rethinking in science, I will try to sketch the state of the art in Dialogue Analysis from my point of view with special reference to the research done in our Association and presented in the papers of this Round Table. The first point to be made is that classical theorizing still remains. Classical theorizing has not stopped with the pragmatic turn. The essential point in defining the orthodox view is based on the fact that it starts with methodology and reduces the natural object to an artificial one by establishing an own methodological level of competence as a rulegoverned closed system. In this way, Dialogue Grammar has focused on what I would like to call the 'deep structure' of dialogue. I remember Hundsnurscher's programmatic article of 1980 in which he postulated a method contrasting with Conversational Analysis by substituting so-called well-formed dialogues for the natural object of authentic texts. The point of rules is also focused on by Sorin Stati (1982) in his book Ί1 dialogo'. We will not forget that rules remain an important methodological technique. However, in dialogue considered as dialogic interaction, rules are tools used by human beings and thus dependent on their individual decisions. What we thought to be the great merit of Dialogue Grammar, namely, that it complied with the methodological conditions of generative grammar, turns out, in my opinion, to be the main obstacle we have to overcome. Thus we are not departing from the hard line, we are departing from the simple which avoids the complex. Martinet (1975), a long time ago, showed us the right way by urging us 'not to damage the integrity of the object by methodological exigencies'. Excluding constitutive features of our natural object, human dialogic interaction, such as cognition and the fact that we are always different human beings interacting in the action game, cannot be the

Dialogue Analysis 2000: Towards a Human Linguistics

17

right method for describing and explaining our object. We have to accept that our object consists of a mix of various integrated variables ranging from order to disorder. The attempt to describe it by total abstraction from disorder can only result in a theory myth. Problematizing orthodox theorizing and accepting the complex mix of order and disorder that our object-of-study represents is in my opinion the challenge we have to tackle when facing the new millennium. It requires us to recognize the object-in-function and to derive from it an adequate new way of theorizing. Most of the approaches used in Dialogue Analysis recently have tried to address the object-in-function in the belief that it is the authentic text that has to be analysed and dealt with. Karin Aijmer (1996) demonstrated convincingly that discourse particles can be comprehended in their multiple variety and multifunctionality only with the support of large text corpora. Large text corpora may be used for the analysis of verbal phenomena like discourse particles; they must however not be identified with our object-in-function. Text corpora are usually analysed from the observer perspective and contain only empirically registrable means. These are heavy restrictions imposed on human dialogic interaction which do not allow our object-in-function to be identified with text corpora. What is it about human dialogic interaction that cannot be gained from text corpora? That is the question. Or to put the question from John Sinclair's perspective (in this volume): What is it about a conversation that seems alien to computers? The conclusions to be drawn seem evident. We must not think that it suffices to analyse authentic texts as an observer. There are various cognitive phenomena influencing dialogic interaction which are not registered in text corpora and which can be understood only from inside the Action Game. Various contributions of this Round Table refer to these cognitive phenomena which are constitutive for dialogic interaction. Power, for example, is dealt with by Michael Metzeltin, identity is the topic of Robert Maier's contribution, or emotion is addressed by Jackie Schön. Svötla Cmejrkova and Adriana Bolivar also deal with aspects of dialogue which cannot totally be figured out from authentic texts. It is these aspects which are among the prospective research objectives indicated by Frantisek Danes. The cognitive level is addressed by Valerij Dem'jankov. In dialogic interaction we use different communicative means, empirical verbal and perceptual means and cognitive means which must not be separated. 'Integration is the name of the game' as Marcelo Dascal calls it. According to recent research in the cognitive sciences we have to account for the integration of different dimensions from the very outset. Linguistics therefore can no longer be considered a science of language in the narrow sense but has to be comprehended as a science of a complex human ability which integrates the verbal, cognitive and perceptual dimensions. Further conclusions result from the fact that it is different human beings, different individuals interacting. This property is focused on in the contributions by Malcolm

18

Edda Weigand

Coulthard (e.g., 1985: 145) pointing to individuality and by Bernd Naumann (in this volume) pointing to chance and chaos. Dialogue at the beginning of the new millennium indeed has to be seen as an open system ranging from order to disorder, from rules to principles of probability, from conventions to presumptions (cf. Weigand 2000a).

3. Some examples

Before revisiting the fundamentals of dialogue, let us first analyse some authentic examples which demonstrate basic features of our object-in-fiinction.

3.1. Meaning is not defined The orthodox view of language and communication is, in its strict version, based on socalled pattern transference (Harris, e.g., 1981). The pattern model or the model of fixed codes starts from the hypothesis that meanings are defined and understanding can be presupposed. Dialogue can thus be outlined in advance and 'transferred' to the interlocutor as a fixed pattern of defined possibilities, which is meant and understood in the same way by the speaker and the interlocutor. Dialogue thus is achieved by simply doubling the speaker side. In contrast to the pattern view, dialogic interaction in the action game is considered to be interaction between different human beings. The means they use are not restricted to explicit verbal means. Let us look at authentic examples like (1)

If you are homeless, you will find a home in Hong Kong because there all are homeless, (heard on German television, translated into English)

(2)

Change is the only constant in the life of a company. (The Economist', March 25th-31th 2000, p. 115).

We immediately notice that we are not decoding verbal signs with fixed meanings but primarily using cognitive means in order to understand what these examples mean. We do not reject these utterances because they seem to contain a contradiction between two defined signs: to be homeless and to find a home, or change and constant. We accept these utterances as quite natural and negotiate meaning and understanding in dialogue.

Dialogue Analysis 2000: Towards a Human Linguistics

19

An interesting authentic example in this respect is the following which again confirms the basic feature of indeterminacy of meaning. The situation is that of an exchange between the organizer of a conference S and two of his chairpersons Ε and F: (3)

S (to F) Ε S

Sie waren nicht streng als Diskussionsleiter. Also muß ich morgen strenger sein. Nein, nicht strenger, streng!

(in English S (to F) Ε S

translation) You were not strict as chairperson. So tomorrow I have to be stricter. No, not stricter, strict!

Neither the adjective nor the grammatical category of the comparative appear to have a definite meaning. Both are used relatively to the extent that the category of the positive might be stronger than the comparative. If we consider dialogue as a process of negotiation we are no longer forced to keep literary and everyday action games separate but can describe them in a unified model. They only take different positions on the same scale between order and disorder. It is simply not the case that everyday action games can be restricted to rule-governedness and well-formedness whereas creativity and innovation would be reserved for literary texts (Harris 1981: 153). Creativity and innovation are constitutive features of everyday conversation insofar as meanings are constantly made and remade by the interlocutors in the process of dialogue (Toolan 2000).

3.2. Different communicative means are integrated It is not a new insight that we do not communicate with verbal means only. Nevertheless, the view that dialogue is based on the same communicative competence for speaker and interlocutor seems to prevail. Such a view can only be understood by reference to the fixed pattern model and to the belief that native speakers of a language all have the same competence. If we presuppose the fact that we are different individuals and that we communicate not only with verbal but also with perceptual and cognitive means - i.e. by presumptions, associations, moment-to-moment decisions, creatively making and remaking meaning and understanding it becomes impossible to start from the view of an ideal or well-formed pattern competence. We have to start from human beings. There are no utterances independent of the speakers, and the minimal dialogically autonomous unit is to be considered the action game which combines different communicative worlds of different interlocutors. Integration is to be taken as a key concept for the action game insofar as it reflects a basic human condition, namely that different abilities are integrally

20

Edda Weigand

used. Even if we wanted, we could not separate our abilities of speaking, thinking, and perceiving. W e make use of these abilities as communicative means in dialogic interaction. As we are different individuals the means, especially the cognitive ones, used by different interlocutors are also different. We accept utterances like our examples (1), (2) and (3) and play our part, our cognitive part, to make them understandable, coherent. We do not know exactly what the interlocutor means and understands. Problems of understanding can be clarified in so-called metacommunicative dialogues. Metacommunication also confirms that meaning and understanding are negotiated in dialogue. The orthodox hypothesis of fixed codes has to be replaced by meaning indeterminacy (Harris 1981: 55 ff.). The model of negotiation is substituted for the model of pattern transference (Weigand/Dascal 2001). There is a special type of action game which is based precisely on the differences between the cognitive background of the interlocutors, namely action games between experts and laymen, and it seems strange that this type of action game did not cause us to problematize the view of pattern transference earlier. Maybe the reason is that dialogue in the pattern view is - even if unconsciously - restricted to verbal means in the belief that cognitive differences could be verbally clarified. What is going on in expert-laymen dialogues is however not only a question of more or less information. To take it in this way represents a technique of self-defence of the orthodox view. The point I want to emphasise in this respect is however another. Languages for specific purposes, for instance, the language of law, have the tendency to define their meanings precisely because in ordinary language use meanings are not defined, not unequivocal. The tendency to make meanings definite can however be fulfilled only within certain limits. Thus, for instance, in linguistics we can define concepts like the 'phoneme', the 'morpheme', because they are concepts introduced within the artificial system of language as sign system. If we consider language as a natural phenomenon, we are confronted with different lines of argumentation and terms used differently depending on the author, such as the terms 'dialogue', 'discourse', 'action', 'action game', etc. Moreover, we know from the language of law, that even seemingly defined juridical terms have to be applied to particular situations and individual conditions. Thus in the end it is again individual human beings, in this case the judges, who decide on the basis of their particular cognitive backgrounds.

3.3. Concepts of probability are constitutive components in the action game On the level of action, it is always the individual speakers who decide how to mean and how to understand the utterance. They cannot refer to rules only. In the complex range between order and disorder they orientate themselves by Principles of Probability. These

Dialogue Analysis 2000: Towards a Human Linguistics

21

Principles make use of other methodological techniques such as rules and conventions but also suppositions, presumptions, moment-to-moment decisions, etc. (cf. Dascal 1994). In dialogic action games we negotiate our positions regarding specific states of affairs which intrinsically contain concepts of probability such as preferences or habits. It is not only the fact that we are different human beings interacting with each other but also the fact that our way of life is based on probability, fluctuation and approximation that dialogic interaction can only be considered as an attempt to come to an understanding. It is the open-endedness of life which requires human dialogic interaction to be based on an open system of Principles of Probability. In this respect let us analyse another authentic example. The situation is the following: The mother enters the room where the daughter is playing the piano: (4)

Mother Daughter Mother

You are playing the piano again. Shall I stop it? No, it doesn't matter. I'm going to work outside.

This example demonstrates several points: In most cases, intonation is not so clear that we can decide what the utterance means. In the end, the speaker alone knows what he/she meant, as in our case the mother with her first utterance. Linguistic rules do not tell the interlocutor how the utterance is meant. The daughter must use cognitive means in order to come to an understanding. This understanding however is based on probability. The daughter cannot refer only to generalised cultural evaluations of playing the piano; she has to include the particular situation and the individual attitude of her mother, i.e., she has to make assumptions which refer to usual preferences. Preferences however represent concepts of probability which carry the risk of misunderstanding. The mother usually prefers the daughter to play the piano when she does not have to work. It is this preference the daughter is thinking of, understanding the mother's utterance as a reproach. However, in this particular situation, the mother intends to work in the garden and corrects the daughter's misunderstanding with her second utterance. I hope it has become clear from this example that open points carrying the risk of misunderstanding are not a disturbing factor which can be ignored but are constitutive for human dialogic interaction (Weigand 1999a). It is simply a myth to believe we could postulate a closed system of rules and exclude problems of understanding from our model. Naturally, we can exclude them, but the model then is not appropriate for our object-infiinction.

Edda Weigand

22

3.4. Not everything is said explicitly The last point I would like to mention contradicts the orthodox view that maintains that in principle everything could be expressed explicitly. If indeed we tried to express everything verbally, dialogue would become as clumsy and never-ending as it would be inefficient. We could not even start talking but would have to reflect on all the points to be mentioned. If at all, only trained linguists could dare to converse. Looking at authentic examples, it becomes immediately evident that even the most important points do not have to be expressed explicitly. This happens not only inadvertently but also deliberately with specific effects. Let us take an authentic example from the journal 'The Economist', an advertising text for the Allianz Group: (5)

Wherever you are. Whatever you do. The Allianz Group is always on your side. For over 75 years we have successfully managed the assets of life insurance policy holders. This, together with the close cooperation of our global partners and the experience of our asset management team leads to improved long-term investment performance. It's no wonder then, that we were recently awarded the prestigious Standard & Poor's AAA rating. Maybe that's why we insure more Fortune 500 companies worldwide than anyone else. Allianz. The Power On Your Side. ('The Economist', March 25th-31st, 2000, p. 3)

Many points relevant to our discussion, which I can only briefly mention, become evident from this example: The verbal text is not an autonomous unit but only a component in the action game. It seems to be a monological text but nevertheless it is part of dialogic interaction with the reader. The action game is a cultural unit. You have to know many things in order to understand the publicity function of the text in the unit of the action game. The main message is not explicitly expressed: 'Join Allianz!' Meaning is persuasion. Syntactic meaning can also be persuasive as can be seen from the heading The Allianz Group is always on your side. The indicative construction does not describe an existing but only a potential or conditional fact insofar as, in a strict sense, something like if you want would have to be added. -

Verbal and cognitive and also perceptual means (a picture is included) are integrated. Word meaning is on the one hand indeterminate, open to negotiation, for instance, in to be on your side, successfully managed, power. On the other hand word meaning is, at least in part, defined, due to the tendency of languages for specific purposes to name things unequivocally, for instance, in life insurance, policy holders, long-term investment performance.

Dialogue Analysis 2000: Towards a Human Linguistics

23

These points demonstrate that it is simply absurd to assume pattern transference would be a useful method for describing texts. Texts have meaning only in the process of negotiation within the action game.

4. Fundamentals of dialogue

Let us now turn to the fundamentals of dialogue to be described in a Theory of the Dialogic Action Game. Three issues have to be tackled: first, the access to the object, second, the issue of understanding the object, and, third, the problem of deriving an adequate methodology from it. The issue of access refers to the problem that there is no empirical evidence as such. We might stress the point that our object is authentic texts but nevertheless I am not of the opinion that the text alone tells us its truth. There is no reality as such, it is always filtered by theoretical questions. We cannot say that we have to start either from empirical texts or from theoretical questions. That is not the issue; it is the integration of empirical data and cognition which is our starting point. Recent research in neurobiology on mirror neurons confirms that at the very outset we have to assume the integration of different abilities such as perception, doing, cognition. The theory itself contains two parts, first, fundamental assumptions on our object, second, the methodology to be derived from them. Having dealt with these issues in detail in my Birmingham paper (Weigand 2000a) and in other articles, I can restrict myself to a few points. Before postulating that there must be rules, we should try to understand

our

object. Our object, human dialogic interaction, is not a homogeneous, clearly separable object. It is the complex ability of human beings to negotiate interactive purposes in the Dialogic Action Game. Human beings, their abilities and the world cannot be separated. We perceive the world as our abilities allow. Human beings are socially purposeful beings and they are different beings. It is their purposes and needs which give us the key concept to guide our analyses. The second part of the Theory addresses the question of methodology.

How do human

beings behave in the complex dialogic world of the action game? There is no pattern predefined in advance, the same for both speaker sides. We are always confronted with different human beings and behave like 'complex adaptive systems', to use Gell-Mann's term (1994), orientating ourselves according to Principles of Probability in a complex mix of order and disorder. On the level of action, which is the level of performance, in the end, everything is dependent on individual human decisions. Principles of Probability can be

Edda Weigand

24

seen as guidelines of behaviour, as guidelines of our competence-in-performance (cf. Weigand 2001). They make use of other methodological techniques, among them rules and conventions. We always try to identify regularities in order to structure the complex. The complex however cannot be captured as a whole by regularities. The Principles of Probability are based on three fundamental principles, the Action Principle (AP), the Dialogic Principle proper (DP), and the Coherence Principle (CohP). The AP refers to the correlation of purposes and means. Purposes are dialogically orientated purposes to be distinguished in a dialogic speech act theory. Means are integrated dialogic means based on the abilities of speaking, thinking, and perceiving: (Fig. 1)

dialogic purposes (state of affairs) Norwegian translations). To take another example, nouns are likely to have higher correspondence values than discourse particles in the compared languages. But we can also expect differences between semantic types or classes within a category. Altenberg (1999: 255) found, for instance, that different types of conjuncts did not have the same mutual correspondence value.

Discourse Particles in Contrast

69

4. Functional analysis

Well has probably attracted more attention than any other discourse particle in English and has been studied from many different perspectives (eg Svartvik 1980, Carlson 1984, Schiffrin 1987, Bolinger 1989, Jucker 1993, Smith and Jucker 2000). The basic meaning (semantic source) of well is acceptance/agreeement (cf. Carlson 1984: 27ff). In the analysis proposed here, well has a number of potential functions constrained by this meaning and the indexical qualities of the particle. Synchronically the meaning of well is reflected in the flexibility of the native speakers' interpretations and the variety of conversational uses compatible with it: Given the meaning of well, one can trace a variety of conversational uses consonant with it, some of which were identified earlier as clues to its meaning. As with any instrument, the uses form a continuum (with varying density), along which one can single out various points for exemplification, few or many depending on how finely one cares to discriminate, (cf. Bolinger 1989: 321)

As a result, it is very difficult to establish a conceptual domain that could be used as a 'tertium comparationis' for a functional contrastive analysis of well. The functions of well may be close to each other or incompatible as when well has both the meaning 'agreement' and 'disagreement' in its semantic package. The norm for the translator must be to preserve functionality. The translations react to the function of well in the source language and pick out the function or pragmatic value that the discourse particle has in the particular context. The difficulties of establishing correspondences can be illustrated by looking at the choices made by the translator. Not surprisingly, Svartvik's attempts to provide translation equivalents for well 'turned out to be only moderately successful': The varying applicability of the Swedish equivalents indicates that there is no directly corresponding Swedish item, and furthermore that well in fact displays a much wider range of functions than those discussed here. (cf. Svartvik 1980: 176)

It is clear that it is not sufficient that a translation preserves meaning. It will only be successful if the item chosen in the target text is nativelike and idiomatic. Pawley and Syder (1983: 193) have drawn attention to the fact "that only a small proportion of the total set of grammatical sentences are nativelike in form - in the sense of being readily acceptable to native informants as ordinary, natural forms of expression, in contrast to expressions that are grammatical but are judged to be 'unidiomatic', odd or foreignisms." The variability and multifunctionality of well is reflected in the large number of different translations it can get: well corresponds, for instance, to jo (turntaking) ja 'yes'

Karin Aijmer

70

(agreement), men 'but' (argument), hör pä 'listen' (attention-getting), da 'then' (inference), näja, tja (deliberation). Moreover, the translator's strategies include the possibility not to translate well at all. The functional contrastive analysis implies that we look for functional classes in which well is translated in one way and other classes where it is translated in another way. However the frequent occurrence of translation equivalents occurring only once and zero correspondences suggest that there are only a limited number of systematic cross-linguistic form-function patterns. An interesting account of how the meaning of well is exploited for different functions is given by Smith and Jucker (2000). According to Smith and Jucker (2000: 217), speakers use well to make their attitudes explicit when common ground needs to be renegotiated because of an actual or perceived discrepancy in the participants' claims.5 For example, in (1) there is a discrepancy between what the speaker believes and what he believes that the speaker already knows. Well does not simply endorse the truth of the preceding statement (like yes) but alerts the listener to the fact that the speaker will embark on a lengthy explanation as the answer to the question "what's wrong with your car": (1)

"No, my car's in the shop." "Your car? What's wrong with it?" "Well. I was driving along and ... you know that little red light on the lefthand side of the dash?" (AT) 6

"Nej.jagharbilenpäverkstad." "P4 verkstad? Vad är det för fei med den?" "Jo, jag satt och körde och den där lilla röda lampan till vänster pä instrumentbrädan, du vet...?"

In a single language we can expect 'good conversational strategies to get conventionalized into immediately recognizable routines' (cf. Carlson 1984: 28). These strategies may also be expected to have systematic translational correspondences. In the following context well serves as a topic shift and seems to go over easily to Swedish jo in the translation, indicating that we have a cross-linguistic link: (2)

5

6

But what has all of this got to do with Greenland? Well. Greenland is just like a huge glass mountain. (RH)

Men vad har dä detta med Grönland att göra? Jo, det är just ett sädant glasberg.

In Bolinger's (1989) analysis, the epistemic well invokes a standard (norm, what is appropriate, etc). Well is, for instance, used when 'one confronts a situation that one cannot quite go along with' and some adjustment is needed. The abbreviation refers to the text from ESPC.

Discourse Particles in Contrast

71

Both ja and jo are frequent as translations of well. The correspondence between Swedish ja and well is the sign of a successful translation since both signal agreement. Agreement is a scalar concept ranging from full agreement to partial agreement (concession, consequence), unwilling or reserved agreement, correction or shift of ground to complete disagreement (objection, rejection, etc.). In (3) where well is used for reassurance or reinforcement, the Swedish translation has ja 'yes': (3)

"When at last his estate was settled, it became apparent that he had a lot more money than anybody had supposed" "Well, of course." said Arthur. "We didn't think he had anything but his chunk of his grandfather's estate" (RDA)

"När boutredningen väl var klar, visade det sig att han hade mycket mer pengar än vad nägon kunde ana" "Ja", sa Arthur, "vi trodde ju inte att han hade nägot annat än de pengar han ärvde av morfadern".

The reinforcing well (of course) is caused by the fact that the first speaker has said something that Arthur already knew. In (4) the speaker has been trying to make up a limerick about Miss Honey. Well serves as a link creating an expectation about what is going to happen: (4)

"I've actually been trying to make up one about you, Miss Honey, while we've been sitting here." "About me!" Miss Honey cried." "Well, we've certainly got to hear that one, haven't we?" (RDA)

"Jag har faktiskt försökt hitta p i en om er, fröken Honung, medan vi har suttit här." "Om mig!" utropade fröken Honung. "Ja, den mäste vi verkligen fä höra, eller hur?"

Well can come as a question (cf. Bolinger 1989: 311). It can be compared with other acknowledging responses {sol, yes?, what?): (5)

"Now - your uncle's birthday fell at a moment when Mercury was the ruling sign of his chart, and Mercury at the uttermost of his power" "So?" "Well?" Maria understands. "Isn't her mother a gifted card-reader?" (RDA)

"Näväl - dinfarbror foddes i ett ögonblick dä Merkurius var den härskande planeten i hans tecken och stod pä höjden av sin kraft." "Ja?" Maria förstär vad jag menar. "Ar inte hennes mor skicklig pä att tyda kort?"

In (6), the sense of agreement is weakened and wellI ja signals a mid-turn pause or hesitation: (6)

The fact was that it seemed... well, foreign. (AT)

Faktum var att det verkade ja, utländskt.

Karin Aijmer

72

Ja and ja dä (corresponding to well then) have the meaning consequence (partial agreement): (7)

"That's ridiculous! She must know." "Well then, you talk to her. I can't get any sense out of her." "I'll talk to her right now!" "Hamish, don't you dare". (RDA)

"Det är ju löjligt! Hon mäste veta." "Ja. dä fär du gä och prata med henne.Jag kan inte fä ett vettigt ord ur henne." "Jag ska genast tala med henne!" "Hamish, du skulle väga".

The young girl does not know who the father of her child is. Well then signals consequence: "Since she must know who the father is, you talk to her". Ja is also used to defer agreement (indicating some reservation): (8)

"And then you went on driving?" "Well, nothing sounded any different and nothing acted any different, so I figured - " (AT)

"Och du fortsatte att köra?" "Ja. ingenting lät ju annorlunda och ingenting bar sig ät annorlunda, sä jag tankte att

When well signals disagreement (correction, challenge, rejection, opposition) it has links with several items in Swedish. It can be translated by men 'but', jo 'but y e s ' , / « 'as you k n o w \ f a s t egentligen 'though actually', faktiskt 'in fact'. In examples (9) and (10), the translation is men 'but' which serves as 'weapon in the verbal conflict'. However the translation results in a loss of politeness. (9)

"I don't know why you feel it's your loss alone." "Well. I just do, sometimes," Sarah said. (AT)

"Jag förstär inte varför du tycker att det bara är du som har lidit en förlust" "Men ibland gör jag det i alia fall."

In (10), men and ju were used when well introduces a correction of what the previous speaker believes: (10)

"Sarah, if I felt we were in the slightest "Sarah, om jag trodde att vi löpte minsta danger I'd have pulled over long ago." risk skulle jag ha stannat för länge sedan." "Well. I don't know that you would have," "Men jag kan ju inte veta vad du skulle ha Sarah said. (AT) gjort", sade Sarah.

73

Discourse Particles in Contrast

The speaker's w i f e has just left him and he reacts by resignation (reluctant acceptance): (11)

He gazed down at the sink, and the warmth from the dishes drifted gently up into his face. Well, you have to carry on. You have to carry on. (AT)

Han säg ner i diskhon, och värmen frän kärlen steg som en mild fläkt upp mot hans ansikte. Näväl. man mäste gä vidare. Man mäste gä vidare.

When well signals deliberation and reluctant hesitation it is particularly difficult to translate. Swedish tja, neej, naja, nja preserve the function o f well but would not be used in authentic dialogue: (12)

"Well, maybe I'll track him down since that's the only lead I have." (SG)

"Tia. dä kanske jap forsöker f l tag i honom eftersom det är enda späret jag har."

Swedish ne-ej signals the speaker's hesitation (cf nej 'no') in answering the question: (13)

"Do you want to sit next to some noisy littie kid?" "Well, no." (AT)

"Vill du sitta bredvid unge?" "Ne-ej."

en liten bräkig

In (14) the translator has used the formal na (hur som heist) to mark a dialogue transition. Well signals the closing o f a topic and a focus on the following discourse: (14)

'"I was scared she'd say no," he said. There was a silence. "Well. Anyhow." Rose said finally. "But I'm getting along!" (AT)

"Jag var rädd att hon skulle säga nej", sade han. Det blev tyst ett litet tag. "Nä. hur som heist", sade Rose till slut. "Men jag klarar mig!"

In (IS), a simple nä signals that the speaker is about to shift ground: (15)

"Well, gentlemen, if you had waited just a little while," Aristotle remembered Plato saying of Socrates speaking to the jury that had just decreed his execution by hemlock, "you would have had your way in the course of nature". (JH)

"Nä. mina herrar, hade ni väntat litet till", kom Aristoteles ihäg att Platon hade berättat om Sokrates tal till domstolen som just hade dömt honom till avrättning medelst odört, "sä skulle ni ha fätt er vilja fram enligt naturens gärig".

Karin Aijmer

74

Swedish hör du 'listen' in the translation of (16) is less formal than nä but shifts the topic more abruptly than well·. (16)

"Well Matilda. I would very much like to hear one of these limericks you say you have written. Could you try to remember one for us?" "Well," Matilda said, hesitating. (RD)

"Hör du. Matilda, jag skulle väldigt gärna vilja höra nägon av dina limerickar. Kommer du ihäg nägon som du kan läsa upp för oss?" "Nja", sa Matilda dröjande.

5. Zero correspondences

In many cases the translator has found it appropriate not to translate well. When is it better not to translate well and why? In (17), where well accompanies a greeting, it disappears in the translation: (17)

"Well, hi there!" she said brightly. "How was your trip?" "Oh, it was... (AT)

"Hej, hej!" sade hon muntert. "Hur har resan varit?" "Jo-o, den var . . .

Well can be omitted since it is an optional element although an element of the meaning is lost (effusiveness). A large part of the zero correspondences are found in indirect or partial answers to wh-questions as in (18): (18)

"Has Sarah been in touch since she left?" "She's come by once or twice. Once, actually," Macon said. "For things she needed." "What kind of things?" "Well, a double boiler. Things like that." (AT)

"Har Sarah hört av sig sedan hon gav sig av?" "Hon har tittat in en eller ett par gänger.En ging ätminstone", sade Macon. "For att hämta ett par saker som hon behövde." "Vad di for saker?" "En ängkokare. Nägra grejor i den stilen."

However the effect of the absence of a translation of well which is "indirect, considered, balanced, diplomatic" is that the Swedish text sounds more abrupt than the English text (cf. Wierzbicka 1976: 365).

Discourse Particles in Contrast

75

6. Conclusion

Multifunctionality and indexicality are a general feature of language: For me, any model which does not accept variability as being at the heart of language will encounter overwhelming problems when set to deal with phenomena like pragmatic particles. Variability works as a counterforce to rule-governedness [...], and gives language its flexibility and allows speakers — through their language — to adapt to and cope with new situations; this in tum allows language change and makes language acquisition possible, (cf. Östman 1995: 104ff.)

Discourse particles are indexical structures that are used variably in the communication situation and have become conventionally associated with certain functions. They are organised around a core meaning but their full meaning only emerges in the dialogue situation. Translations help us to assign a more precise meaning to a particle since the translator is forced to find an appropriate correspondence to the particle in the context where it occurs. The purpose of the analysis of well was to show the distribution of its functional equivalents in the Swedish translations. The results can be summarised as follows: there was a high percentage of types of functional correspondences of well in the Swedish translation many of the functional correspondences occurred only once there was no single item which could be regarded as a prototypical translation equivalent of well zero-correspondences were frequent The investigation can be seen as a complement to other contrastive studies, e.g. that by Bazzanella and Morra of well and the Italian bene. The features of the translations of well into Swedish were equally striking in the Italian translations of well and in Norwegian and German. It would also be interesting to see what changes would occur if the material had been larger. Would the cross-linguistic form-function patterns have been clearer or would the number of types of functional equivalents have increased instead? The problems in translating well are relevant to scholars interested in translation theory. The contrastive approach also provides a method to test hypotheses about the meaning of well and to find cues to its conversational uses. However a discourse particle like well only comes to light in authentic dialogue. Several methods must therefore be combined in the study of discourse particles. The corpora allow us to study 'small* pragmatic features of language and arrive at interesting insights about the compared languages. However when one compares languages one

76

Karin Aijmer

comes up against problems which are language typological and require a deeper explanation. For example, the grammatical and functional properties of well do not necessarily have any direct correspondence in other languages, as has been shown by the present analysis where well did not have a semantic cognate. A 'large' pragmatic feature of English and Swedish (and German; cf. Fillmore 1984: 133) is thus that English has discourse markers like well, which Swedish has not. Why have certain languages developed a discourse particle like well and others not? The explanation of the cross-linguistic differences can benefit from what we know about the historical development of discourse particles. Just as the indexicality of language plays a role in language acquisition, it can explain language change and the grammaticalization of discourse particles. Grammaticalization is the process through which discourse particles (discourse markers) develop on a cline from adverb to discourse marker: clauseinternal adverbials>sentence adverbials>discourse markers. Well is originally a manner adverb ('well and good') which is placed in the 'pre-front field' in a new meaning (cf. Carlson 1984: 27); the initial position is associated with a number of discourse functions and is interesting from a language typological perspective (Auer 1996). Grammaticalization operates according to universal semantic and pragmatic principles resulting in "increased grounding in the speaker's attitude and perspective on what is said" (Traugott 1995). In Swedish the semantic cognate of well is väl. It has developed epistemic meaning (uncertainty) and it is integrated in the proposition (cf. the German cognate wohl which is also a modal particle). The Norwegian vel, on the other hand, has a closer correspondence to English well, formally and functionally, and is less grammaticalized than its Swedish counterpart.

References

Aijmer, Κ., B. Altenberg and M. Johansson (1996): Text-based contrastive studies in English. Presentation of a project. - In: K. Aijmer, B. Altenberg and M. Johansson (eds.) Languages in Contrast. Papers from a symposium on text-based cross-linguistic studies, 73-85. Lund: Lund University Press. Altenberg, Β. (1999): Adverbial connectors in English and Swedish: Semantic and lexical correspondences. - In: H. Hasselg&rd. and S. Oksefjell (eds.) Out of Corpora. Studies in honour of Stig Johansson, 249-268. Amsterdam: Rodopi. Altenberg, Β. and Κ. Aijmer (2000): The English-Swedish Parallel Corpus: A resource for contrastive research and translation studies. - In: C. Mair (ed) Papers from the 18th International Conference on English Language Research on Computerized Corpora. Amsterdam: Rodopi.

Discourse Particles in Contrast

77

Auer, P. (1996): The pre-front field in spoken German and its relevance as a grammaticalization process. - Pragmatics 6(3), 295-322. Bazzanella, C. and L. Morra (2000): Discourse markers and the indeterminacy of translation. - In: Korzen and Marello (eds.) Argomenti per una linguistica delta traduzione, On Linguistic Aspects of Translation, Notes pour une linguistique de la traduction, 149-157. Alessandria: Edizioni dell'Orso. Beijer, F. (1996): Discourse items in contrast. Term paper. - Lund: Lund University. Blass, R. (2000): Particles, propositional attitude and mutual manifestness. - In: G. Andersen, and T. Fretheim, Pragmatic Markers and Propositional Attiudes, 207-37. Amsterdam and Philadelphia: John Benjamins. Bolinger, D. (1989): Intonation and Its Uses. Melody in grammar and discourse. - London: Edward Arnold. Bruti, S. (1999): In fact and in/atti: the same, similar or different. - Pragmatics (9: 4), 519-33. Carlson, L. (1984): "Well" in Dialogue Games: A discourse analysis of the interjection "well" in idealized conversation. - Amsterdam and Philadelphia: John Benjamins. Comrie, B. (1986): Contrastive linguistics and language typology. - In: D. Kastovsky, and A. Szwedwek (eds.) Linguistics Across Historical and Geographical Boundaries. In honour of Jacek Fisiak on the occasion of his fiftieth birthday, Vol. 1, 115-63. Berlin: Mouton de Gruyter. Dickens, A. and R. Salkie (1996): Comparing bilingual dictionaries with a parallel corpus. Euralex' 96, Proceedings I-II, Papers submitted to the Seventh EURALEX International Congress on Lexicography in Göteborg, Sweden, 551-59. Ebeling, J. (1998): The translation corpus aligner: A browser for parallel texts. - In: S. Johansson, and S. Oksefjell (eds.) Corpora and Cross-linguistic Research. Theory, method and case studies, 101-12. Amsterdam/Atlanta: Rodopi. Fillmore, Ch. J. (1984): Remarks on contrastive pragmatics. - In: J. Fisiak, (ed.) Contrastive Linguistics. Prospects and problems, 119-41. Berlin: Mouton Publishers. Johansson, S. (1998): On the role of corpora in cross-linguistic research. - In S. Johansson, and S. Oksefjell (eds.) Corpora and Cross-linguistic Research. Theory, method and case studies, 3-24. Jucker, A. (1993). The discourse marker well. A relevence-theoretical account. - Journal of Pragmatics 19(5), 435-53. Östman, J-O. (1995): Pragmatic particles twenty years after. - In: B. Wärvik, S.-K. Tanskanen and R. Hiltunen, Organization In Discourse. Proceedings from the Turku Conference, 95-108. Turku, Finland: University of Turku. Pawley, A. and F. H. Syder (1983): Two puzzles for linguistic theory: Nativelike selection and nativelike fluency. - In: J. C. Richards and R. W. Schmidt (eds.) Language and Communication, 191-226. London: Longman. Schiffrin, D. (1987): Discourse Markers. - Cambridge: Cambridge University Press. Smith, S. W. and A. H. Jucker (2000): Actually and other markers of an apparent discrepancy between propositional attitudes of conversational partners. - In: G. Andersen and T. Fretheim (eds.) Pragmatic Markers and Propositional Attitudes, 207-37. Amsterdam and Philadelphia: John Benjamins. Svartvik, J. (1980): Well in conversation. - In: S. Greenbaum, G. N. Leech and J. Svartvik (eds.) Studies in English Linguistic, 167-787. London: Longman. Traugott, Ε. C. (1995): The role of the development of discourse markers in a theory of grammaticalization. Paper presented at the 12th International Conference on Historical Linguistics, Manchester, UK. Wierzbicka, A. (1976): Particles and linguistic relativity. - International Review of Slavic Linguistics 1(2-3), 327-367.

78

Karin Aijmer

Appendix

Table 3: Correspondences of well in German (in OMC) also 37 na ia 28 nun 17 nun ia 15 tia 12 na 11 ia 10 na schön 7 eut 2 ach 2 aber 2 nun eut 2 One occurrence: doch, na ... tja, aha, so so, ne gut, und, wohl, sicher, ja also, und nun, grosser Gott, schon, na und 0: 50 occurrences Table 4: Correspondences of well in Norwegian (in OMC) vel 55 ia 28 tia 12 näia 12 men ("but") 11 nei ("no") 9 Nävel 8 Nä 7 Jo 7 Javel 5 da ("then") 4 ia ia 3 altsä ("thus"") 2 iasä ("I see"1 2 "other" 2 One occurrence: likevel, ä, fint, men ηέ, du, her her, ja... jo, egentlig, sannelig 0: 32

John Sinclair

Can We Have a Conversation with a Computer?1

This is a question that has both practical and philosophical interest. The crudity of our present communication with machines does not need to be underlined by me, because for many people it is a daily frustration, and for many more it is the reason why they refuse to engage with machines any more than is strictly necessary. I will concentrate on the ubiquitous personal computer, but the same problems are to be found with mobile phones, automatic exchanges, ticket machines - indeed no machine that tries to interact with a human being makes a very good job of it. Meanwhile we are constantly given glimpses of a rosy future where for example our cars will speed, pollution free, along highways, automatically avoiding each other and reaching precise destinations, while all the time regaling us with information about the history and archaeology of each place we pass, recommendations for local restaurants, parking facilities and advice about refuelling. Having recently travelled in an airport shuttle bus whose announcements were misaligned with its journeys, I await this Utopia with reservations.

1. What is conversation?

To answer the question that is the title of this paper, we have to be clear what we mean by a conversation. Do we have a conversation with a light switch when we press it? Do we have a conversation with a ticket machine as we endlessly select options - apart from occasional curses that the machine does not hear? Do we have a conversation with a web site when we are ordering a purchase or filling in a form? Do we have a conversation with the ghastly comic paperclip that beams itself down in WORD and tries to interfere with your writing? 1

This paper is related to another, "The Deification of Information", which appeared in Scott and Thompson (2001). That paper was written some seven years ago but publication was seriously delayed. The paper dealt with the state of personal computing and the relationship between information science and linguistic description, and given the speed of change it is now in need of updating, and this paper is an update of part of it.

John Sinclair

80

At present the use of the spoken voice is very limited in relation to machines, so most of the "conversations" we have with machines are in writing or through the use of the mouse, touch screens etc.; however the structure of the exchanges at the present levels of sophistication is not much affected by the medium and I will consider all of them possible conversations. If the conclusion is that it is not possible to have conversations with machines, then we must work out other means of communication with them, no longer trying to disguise our activity as if it was like conversation with another person; greater clarity and efficiency should ensue. If on the other hand we decide that it is possible, then we must ask seriously why we are not better at making it work for us.

2. Simulation

There is one influential approach that we can reject at the outset - the simulations and emulations that computer scientists find so attractive. A lot of the daily frustrations and failures come from this source, as we are trapped again and again into thinking that our conversant shares normal human characteristics. The illusion is aided by personalising interfaces, from unattractive humanoids like the paper clip or tasteless anthropoids like Bonzi Buddy through to threats of forthcoming images of attractive members of one's preferred sex, tuned to one's proclivities and devoted to selling some product. The rosy future beckons again. To a linguist, the failure of computers to adapt to the very simple and basic conventions of presenting utterances in a language is an early caution. If they are still having trouble with capital letters and small ones after fifty years of development, their chances of making gaffes and clangers in more complex and delicate matters are high. One of our first investigations must be to assess whether machines can, or can ever, cope with "open text"2. This is at the heart of the problem. Human beings can accept any utterances that come within normal competence, and this includes coping with many that require a bit of guesswork. Humans rarely admit to being unable to maintain one side of a conversation, and struggle on through tricky passages, hoping that they will shortly return to normal levels of understanding. Computational linguists now appreciate that this ability is one of 2

"Open text" is the name given to text that has been produced - as far as can be ascertained - under normal conditions of communication, and has not been restricted or edited in any way to make it more palatable for machines. To avoid a further raft of complications we shall concentrate here on written text.

Can We Have a Conversation with a Computer?

81

the most difficult to replicate in a computer, and the performance of devices that purport to "understand" open text is abysmal. Consider for example the search engine Ask Jeeves (www.ask.com), which claims to process open text questions. I asked "Who runs UK?", and was advised to consult the Community Forum, where other users - human beings - might help me; I was told that people with similar questions have found relevant a number of sites to do with running as an exercise or sport. In other words, the query interpreter could not distinguish different meanings of run, and by default chose the physical exercise meaning, which a corpus analysis would show to be poor tactics. Changing the question to "Who governs Britain?" elicited a list of royals and politicians from 1939 onwards, containing the reasonable answers to my question. Some readers may think that my question was deliberately tricky, and feel that when I rephrased it I got an adequate response - but this only shows how far we must be prepared to vary our normal behaviour in order to achieve sensible communication with a machine. A reasonably educated human being would have no problem understanding my original question, and would not try to sell me a pair of running shoes by way of response. If such a person knew the answers, they would not start off with King George VI, who died nearly half a century ago. I tried another question - a real one this time, a frequent question from my four-yearold, "Why am I so small?" Again, a human would be able to offer a few words of explanation, and perhaps, guessing the underlying motivation, some reassurances that the passage of time would improve the situation. Jeeves, however, cannot cope at all, and says "Sorry, I didn't find any matches." He/it offers by way of consolation ten matches by Yahoo, of which the first is "CNN.com - Bush demands tax refund for Americans." Given this level of relevance, I did not consult the other nine "matches". Again I may be accused of misunderstanding or misusing the search engines, which are not geared to this kind of question, but are really good at "Which US state has the abbreviation MD?" (Actually, even there the response was just a list containing the answer to my question.) The only conclusion to come to is that Jeeves is not able to handle most open text, and to use the site at all the user must second-guess the machine or put up with mindless rubbish, and probably both. Little has changed in thirty-five years; it was about then that the ELIZA program arrived on the scene (written up in Weizenbaum 1984 [1976]), and Artificial Intelligence became a realistic proposition for many. The strength of ELIZA was that it always made a response, no matter what was typed in, and in many cases it appeared to understand what it was told; in fact it held a small vocabulary of words that triggered responses that were likely to be relevant, and it embedded anything else into a "Why do you s a y — " pattern. Weizenbaum says (op.cit. p. 189):

82

John Sinclair

ELIZA created the most remarkable illusion of having understood in the minds of the many people who conversed with it. Alas, only an illusion, and still the stratagem of relying on keywords and not engaging with the structure of a language persists, and it is still only an illusion.

3. Information

Information science has abandoned any attempt to understand texts by means of their linguistic structure, and regards them essentially as unstructured; they have to be given an exoskeleton of a quasi-logical nature in order to be communicated with. Strange concepts like "knowledge mining" have arisen to label the software which attempts to extract information from language text. The image of the sweating labourer swinging a large pickaxe at a solid face of text and managing to dislodge a few irregular fragments of knowledge bears witness to the frustration that has been felt at the resistance of language to yield its secrets to inadequately equipped researchers. It is fair to say that an over-simple model of communication pervades the whole of our attempts to work with machines, and it is unlikely that there will be substantial improvements until it is totally abandoned. The names such as "information science", "the information society", "the information age" lead the way to a view of verbal interaction of a classic kind, popular in the early days of information theory. This model sees individuals as encoding devices. Like a telephone, which converts a sound wave into an electrical signal, a speaker converts an abstract message into a soundwave, which then disturbs the ear of the listener. The signal is decoded by the listener and the original meaning is thus transferred. The language has been simply a carrier. Work on discourse and social interaction has demonstrated that this is an inaccurate picture of what is going on. First, there is no message that is separable from the language that people use to communicate; such phenomena as the ability to paraphrase indicate that people perceive a similarity of meaning in different phrasings, but paraphrase also underlines the point that each phrasing is subtly different in meaning. Second, each person is a unique organism, with a psychological make-up that is impenetrable by another, and so while we share language and sufficient experience to establish common ground with members of our own speech community, we are unable to discover in what precise way an

Can We Have a Conversation with a Computer?

83

utterance is understood. If a conversation continues with superficial smoothness, we do not inquire3. Thirdly, human beings do not simply trade information, in the sense of factual statements about themselves and their environment. There is always another goal being pursued, towards the achievement of which the information may be useful. A purely factual statement, if such exist, plays only a minor role in our communicative structures, so even if it can be recognised, or if some of the factual content of other utterances can be isolated, the researcher is a long way from understanding the conversation, and is heading in the wrong direction. It comes as no surprise that information scientists, armed with such an inadequate theory of discourse, find little structure in ordinary speech and writing, and have to supply a formalism for marshalling what little they can retrieve; why they distrust and even despise language text as a repository, and feel the need to restructure its supposed content and disperse that into containers such as databases. Since these objects constitute the world experience of the computers, the mismatches that we saw above with Ask Jeeves are understandable.

4. Key features of conversation

There are five characteristics of human conversation without which it will be felt to be unnatural, and in all likelihood will quickly be abandoned. 1. 2.

Each participant has a unique personal agenda which he/she brings to the discourse; it is unknowable precisely to the others and only partly revealed in the discourse. Each participant maintains a coherent stance and/or persona throughout a discourse, such that anyone who falls below a reasonable standard of consistency loses trust and respect.

3.

Participants monitor the discourse as it takes shape, and can at any time talk about the discourse itself without causing breakdowns.

4.

Everything a human being says or writes is evaluative, either in itself or because of a systematic connection to an explicit evaluation.

5.

Discourse is simultaneously collaborative and competitive, and participants use a number of interactive devices to exert as much control over a conversation as they wish to achieve, in pursuit of their agendas (see 1. above).

3

This fundamental discontinuity among individuals was given that name by Weinrich (eg. 2000).

84

John Sinclair

In the remainder of this paper I will briefly examine each of these characteristics in order to see if it is also a feature found in conversations with machines.

4.1. Agendas Let us call the collection of plans and priorities that are unique to each participant in a discourse an agenda. Human discourse is then characterised by being two-way, in that each participant is pursing a personal agenda, and tries, within the collaborative conventions that apply, to achieve his or her goals. Each is aware that the others have agendas, and makes assumptions about them that develop during a conversation, but it is all guesswork, and personal agendas may be complex, layered and not available even to the participant who is working them out. Some agendas are very simple - for example a participant may be willing to do just about anything to terminate a conversation as quickly as possible, or they may want to prolong the conversation beyond its natural ending for some obscure reason. Others are highly complex, and some are subject to change as the conversation proceeds, and either overturns some initial assumptions of a participant or reveals new positions or evidence that require adjustment to the agendas. This feature of language used to be called its interactive feature, but in talking about conversations with machines, the word interactive has a more superficial meaning. All early computers handed jobs by means of batch processing, where each job, program and data together, was entered in its entirety, and the results were retrieved perhaps on the next day. Jobs were devised and articulated off-line. When some twenty years ago it became possible to engage the computer in real time work, and get immediate results, this breakthrough was called "interactive computing", and the name has stuck. So, adapting to computer terminology, we can call two-way a discourse which proceeds on the assumption that each party to a conversation has a unique personal agenda that has to be negotiated during the conversation. The alternative, one-way discourse, would be where only one participant's agenda is recognised, lies outside human experience. Cases of extreme authoritarian postures as in courtroom interrogations or highly insensitive writing may get close, but our agendas are never just ignored. The nearest people normally get to giving up their agendas on a temporary basis is in highly routinised interactions like quizzes, or supplying a clerk who is filling in a form with information that he or she needs. The computer has only one-way discourse to offer, where it has the only agenda. The human has to alter or redirect his or her agenda to fit that of the machine, and gradually, with experience, the agenda of the human is eroded to the barest minimum. Ask Jeeves is a

Can We Have a Conversation with a Computer?

85

model of clarity and relevance compared with for example the Help facilities in most software applications. Interactive computing has developed at the expense of the human being, largely because of the inability of the computer to handle open text. If instead of tackling this substantial problem, the software developer can get away with little pull-down menus that offer halfa-dozen alternatives and no others, the packages can tumble off the production lines at high speed. The ideal customer for these systems is the clicker. The entire discourse role of the human being is reduced to a series of clicks. He or she can only have what is clickable, can express no other wants or needs, in fact can only express anything at all by clicking on something already organised and expressed by another. In this world the human being is reduced to an adjunct of the machine, and it comes as no surprise that Utopias of subservience to machines are offered as enticing prospects - if you are satisfied with the clicker role, then the "anthill" model of computer-run society will suit you nicely. You will learn to have no expectations of your machines, to adopt others' agendas cheerfully because you no longer have one of your own, to sublimate your physical needs by accepting virtual provisions. Bugs and glitches will gradually appear like old friends, and you will continue to stumble uncomplainingly upon facilities and options that are useful but not obvious 4 . This is of course the biggest trap ever offered by the simulators - to surrender one's personal agenda. In small ways we do it already, for example when our politeness rules allow a machine to impose its agenda on us, or a fiercely disciplined human clerk/ receptionist/ "consultant" insists that his or her agenda takes priority over yours. No doubt for the sake of efficiency there are times when this is the most convenient option, but it should never be mistaken for a conversation.

4.2. Stance We expect certain consistencies from our fellow-humans. While we may not all be models of rectitude and transparent honesty, there are accepted standards in all societies. W e do not expect that another person will flatly contradict what they said a moment ago, or claim that they do not have something that is sticking out of their pocket, or forget a phone number that tripped off their tongue a little earlier. We do not expect them to change their

4

At the time of writing I was a daily user of Word97. In the last few months I have discovered by chance (a) that the infuriating way in which "(c)" automatically becomes © can be cancelled by an immediate backspace, (b) that two dashes followed by the "greater than" sign, "-->" becomes a long dash ->, (c) that the little down arrow to the right of windows such as "file open", "save as" takes you to the top level of the file system.

86

John Sinclair

attitudes and opinions without warning, to gush about a TV personality and then say they never watch television. If we come across people who are neither senile nor juvenile but who behave like this we usually lose respect for them and avoid them, avoid particularly having to work with them. Both rationally and emotionally people maintain reasonable consistency in matters of fact and attitude that have been expressed in the conversation. They are careful about what they endorse, and what they pass on without endorsement or with a negative tag. Language has devised an elaborate and subtle set of devices for attribution, averral, and disclaiming that are in constant use. By contrast, the computer has no concept of a stance, and merely reports what state it is in when so instructed. So if it allows processing of a file when it is in the appropriate directory and then moves to another directory, it usually loses all awareness of the file and claims that it does not exist. Since most operations by humans in computing involve changing the state of the machine, amnesia is one of the principal features of the operating systems. In some operating systems and packages it is possible to retrace one's previous commands, or the "history" of the interaction step by step, and it would not take a lot of ingenuity for the computer to become aware of at least all the objects and operations of the current session, and to consult its list for likely items when it cannot find them in its immediate state. In this way, and with a few similar crude provisions, some of the more irritating features of a lack of reliable stance could be compensated for, but the chances of the machine eventually being able to adopt a satisfactory persona are fairly remote. This is a fundamental problem; either the human adapts to the machine or the machine adapts to the human. If in a central area it looks unlikely that the machine is capable of adaptation, then its ultimate utility value is threatened and circumscribed. In practical terms it is worrying that modern operating systems, with their millions of lines of code, have not managed even the grossest simulations of some of the indispensable features of humans in interaction. New ailments such as cyber-panic, cyber-agoraphobia and cyber-vertigo are beginning to be identified. Such is the power, speed and flexibility of the technology that we can wander in cyberspace, and frequently lose our way, or find ourselves going round and round the same sector of jungle, with apparently no way out. Here the computer is very little help; the dimensions of information processing are so vast that people often cannot comprehend them, and software which should be immensely beneficial is rejected because the guidance is inadequate, and the lack of conventions is bewildering. The mess that we are in at present has its origins in the encouragement given to emulation, simulation and other kinds of faking, and the blatant misrepresentation of operations in order to sound user-friendly. Our directories and our files are full of hidden

Can We Have a Conversation with a Computer?

87

items, some of which are not retrievable by normal users. The word processor will suddenly alter the typeface or point size of a paragraph, refuse to delete an unwanted line, alter margins without being told to, responding to instructions which are in the text but not visible to the unfortunate user, who gradually works out strategies for dealing with them. In the wider context, mark-up languages such as HTML or XML are hidden but retrievable, and links from one document to another are the high point of the World Wide Web. Again, there is no requirement for the document that is constructed from a ML script to be accurate or even honest, and millions are not, giving the impression sometimes that the web is populated by con artists, twisters, and bumbling incompetents. While the freedom and indeed anarchy of the web is one of its attractive features, well worth preserving, there is a serious need for more disciplined sectors to be established and protected, and within one of these for the stance of the computer to be made reliable, consistent and transparent throughout each session, and even from one session to another.

4.3. Planes o f Discourse To enable consistency of stance, and for many other operations, human beings need to be aware of two levels of organisation of the discourse, called planes (Sinclair 1982). One is the interactive plane, so named before the computing sense of "interactive" became current. This plane describes the ongoing, real-time interactions, or their equivalents in writing, assigning linear structure to the successive moves and looking for expressions that support or disturb the coherence of the emerging discourse. The other plane is the autonomous plane, which receives analysed information from the interactive plane and constructs an interim discourse arena for reference during a conversation; this plane is also in contact with the stored knowledge of the individual speaker, and can bring that to bear when it is made relevant by the activity on the interactive plane. One of the most useful features of ordinary discourse is the option of talking about the discourse itself. It is a major part of the feedback system whereby participants in a conversation can switch from whatever is their current topic to the way in which the discourse itself is taking shape, whether participants are understanding the movement of the talk and the role of the individual contributions to it. It is permanently available and frequently used, and the switch to talking about the discourse itself is called plane change. Unfortunately another term is commonly used for this feature - "metadiscourse" which suggests that the discourse shifts to another plane. Quite the contrary - the discourse continues on the only plane that it relates to - the interactive plane - and when necessary an aspect of the other plane is incorporated, but in the content only.

88

John Sinclair

The ability to talk about the discourse itself is one of its most important structural features. It allows us to clarify the state of the interaction from time to time, to name events in the discourse, and to avoid awkward turnings. The two-level structure of the discourse, where one level monitors the other, has implications that reach far beyond a conversation, and into the current philosophical debate about consciousness itself, which seems to require levels of monitoring. The present configuration of computers offers no clarity of monitoring. The operating system or the applications package can send messages which amount to talking about the discourse, but to do this it has to move to another level and is closer to the idea of "meta-" than human plane-change. It loses knowledge of what it has just been talking about. It is difficult to see how this defect can be remedied because it is in the nature of computing as we have it today that the various levels are hermetically sealed from each other. When a computer is moving a file around it has no knowledge of what is in the file it could be nothing at all or a huge mass of data. The same with directories and all the other items it handles. When it opens a file it loses all awareness of what is outside the file because it has adopted the mode of interpretation that is necessary for that file. A level of monitoring that could maintain awareness of all the means of interpretation that are in use at a time would be a serious problem, and to be able to mix this in with the ongoing job seems on the verge of intractable. Yet without this the computer will not be able to achieve self-reference beyond the unreliable simulations we have at present. It is true that ELIZA was able to mock up something of this kind - when stumped for keywords the program would repeat what had just been typed in with a preface like "Why do you say....?" or "What do you mean by... ?". However, this was just in-file cutting and pasting, and there was no monitoring level nor any concept of planes and plane change.

4.4. Evaluation Everything that a human being writes or says is part of an evaluative process; it is impossible to maintain a neutral or objective stance. This point follows from the personal agendas that participants have and the likelihood that they rarely coincide. We have already castigated the "information" model of discourse, and we must move it even more to the margins of relevance, because the simple exchange of facts either does not take place at all or does so only rarely, and within a structure that ties it to evaluation. People are always trying to do things when they talk, and certainly the acquisition of facts can be a useful stage in the achievement of a goal. But first the need for the fact has to be established, and the particular fact identified; after a fact is supplied the receiver has to evaluate it in various ways - is it the fact that was solicited? If not is it relevant, part of

Can We Have a Conversation with a Computer?

89

the required information? Is it from a reliable source? Does it conform to expectations i.e. does it come within the parameters that one's experience suggests are normal? How does it fit in with any other facts that have recently been acquired or retrieved from storage? Are there any conflicts of fact or inference that need to be resolved? These and many more issues constitute the necessary evaluation of the fact before it can become useful as part of the current strategy. Treating facts with deep suspicion is in line with major intellectual movements such as post-structuralism, and findings like those of Hunston (1994), who shows that even dry and apparently objective scientific writing has plenty of concealed evaluations, perhaps just in the juxtaposition of facts. For the purposes of discourse description it is best to assume that "pure" facts do not occupy either initial or final positions in structures from the exchange above, and that any which apparently do are probably evaluations in disguise. As we attribute more and more evaluative meaning to human discourse, we see it as moving farther and farther from what a machine can cope with in its present conceptualisation. It cannot evaluate at all - it can only simulate evaluation by a process of measurement. A human being has no problem deciding what constitutes "a long time" in relation to various activities, and the relation of this concept to the passage of seconds, minutes, hours etc. will vary according to the activity. The perception of the human, and the consequent actions will be based on the notion of "long" and not on the calculation of hours and minutes. The computer has no problem adding and subtracting, and can handle hours and minutes without having the faintest idea what they are - but to get it to create the human concept of a long time is probably a pseudoprocedure 5 . And the same for "short", "hot", "cold" and all the other simple evaluations we use all the time, before we even consider the complex ones. We are here again faced with worlds which hardly overlap; it is unlikely that either party will completely adopt the others' perspective, so there will always be mismatches and misunderstandings. Perhaps different concepts of the architecture and operation of computers in the future will make it possible for them to come closer to a human perspective, but their role is severely limited by their inability to perform simple human operations like guessing.

'

I refer here to Abercrombie's brilliant little paper of that name (1965), which should be required reading for all concerned with linguistic computing.

90

John Sinclair

4.5. Control Participants in a discourse are constantly striving to control it. Their relative social positions and roles give them different opportunities to gain and preserve the initiative or to move the discourse in their direction by more subtle means. They do this within an overall attitude of co-operation, because discourse structures are mutually made. So there are many types of utterance and utterance sequence that have been established in human conversation which preserve the basic co-operation but allow competitive manoeuvring. The central feature of prospection

(Sinclair 1992) shows a continuous battle to pre-empt

the future, and some conversations are like chess games, where the participants are thinking several moves ahead, but in real time. By contrast, the computer has neither a co-operative nor a competitive stance and no alternative agenda but to respond to the signals from the user. However, as we have seen, it is not by any means merely a subservient amanuensis - once begun a process, it is in charge and allows no deviation from its procedures, punishing mistakes ruthlessly. It has no idea of any plans beyond the immediate operation, nor any means of comparing that with what has gone before. So a human being gets no support from the machine, of the kind that he or she would expect from another participant; the computer takes control as soon as it is consulted, and from then on everything has to be in its terms or the job is aborted, distorted or generally becomes a nightmare.

5. Conclusion

The results of this brief survey of the characteristics of human conversation as compared with interaction with a computer are somewhat depressing; there seems to be no hope of ever accepting computers as reasonably natural conversants, playing an understandable role and co-operating in getting things done. Any deeper probes of this nature merely uncover further problems and incompatibilities. Structurally they just do not match, and in their present form there is no prospect of adapting them to be more useful in a genuine, not simulated sense. Unfortunately the designers of software and computer architectures are concentrating on improving simulations and adding further opportunities for clickers. On track record these extensions are likely to make matters worse rather than better. Computers are not people

Can We Have a Conversation with a Computer?

91

with communication problems - they are totally other, and we will understand them better if we do not ascribe human characteristics to them. The main argument of this paper is that quite a lot can be learned from the study of the way humans converse, and some of this can be used to improve the design of interfaces. Again there is little evidence of a coming together of discourse analysts and software designers; although "HMI" - human-machine interaction - is always said to be extremely important it does not have a prominent place in the conferences and publications of the language engineers, knowledge miners and knowledge managers. Perhaps this situation will improve. I attach as an appendix to this paper the current announcement of the recently-formed Special Interest Group of the powerful Association of Computational Linguistics. There are traces there of many of the points addressed in this paper, though little sign of a move to establish an adequate theoretical model and descriptive apparatus. It mostly looks to be at the level of practicality, fixing and simulating except for some of the topics listed under pragmatics. Other forums have failed to attract a sufficient diversity of interest and experience to influence the field, or even to come to the attention of those who have large investments in continuing to mould people into satisfied clickers.

References

Abercrombie, D. (1965): Pseudo-procedures in linguistics. - In: Studies in Phonetics and Linguistics. London: OUP Hunston, S. (1994): Evaluation and organization in a sample of written academic discourse. - In M. Coulthard (ed.) Advances in written text analysis. London: Routledge. Sinclair, J. (1982): Planes of Discourse. - In S. N. A. Rizvil (ed.) The Two-fold Voice: Essays in honour of Ramesh Mohan, 70-91. India: Pitambu Publishing Co. Sinclair, J. (1992): Trust the text. - In: M. Davies & L. Ravelli (eds.) Advances in Systemic Linguistics: Recent Theory and Practice, 5-19. London: Pinter. Sinclair, J. (2001): The deification of information. - In: G. Thompson and M. Scott (eds.) Patterns of Text: In Honour of Michael Hoey, 287-314. Amsterdam/Philadelphia: John Benjamins. Weinrich, H. (2000): Von der Lieblichkeit der Sprache. - In: Sprache, das heisst Sprachen. Tübingen: Gunter Narr. Weizenbaum, J. (1984): Computer Power and Human Reason. - Harmondsworth: Penguin Books.

92

John Sinclair

Appendix

2nd ACL SIGdial Workshop on Discourse and Dialogue - Aalborg, Denmark, September 1-2,2001 Description Following up on the successful 1st Workshop in Hong Kong in October 2000, this will be the next in a series of workshops spanning the ACL SIGdial interest area of discourse and dialogue. While there has been a lot of activity in this area, and fairly frequent "specialty" workshops on various subtopics, until this series there has not been a regular place for such research to be presented in a forum to receive attention from the larger SIGdial community and researchers outside this community. Topics of Interest We welcome formal, corpus-based, implementational and analytical work on discourse and dialogue, with a focus on the following three themes: (i) Dialogue Systems Spoken, multi-modal, and text/web based dialogue systems including topics such as: * dialogue management models (specific subproblems or general modeling, in particular models for mixed initiative and user-adaptive dialogue); * speech, text, and graphics integration (for understanding or generation); * context-based interpretation and/or response planning,in particular how this contributes to natural interaction; * strategies for handling or preventing miscommuncation (repair and correction types, clarification and underspecificity, grounding and feedback strategies); * utilizing prosodic information for various types of disambiguation; * task-driven versus conversational dialogue; * evaluation of dialogue systems including task complexity measurements. (ii) Corpora and Corpus Tools Coipus-based work on discourse and spoken, text-based and multi-modal dialogue including its support, in particular: * issues and problems in discourse and dialogue annotation; * techniques (including machine learning), tools, coding schemes and data resources for discourse and dialogue studies; * XML-based tools for dialogue access to internet information. (iii) Pragmatic and/or Semantic Modeling The pragmatics and/or semantics of discourse and dialogue (i.e., beyond a single sentence) including the following issues: * the semantics/pragmatics of dialogue acts (including those which are less studied in the semantics/pragmatics framework); * incremental (plan-based,topic-based, etc.) models of discourse/dialogue structure integrating referential and relational structure; * modeling genre-specific aspects of discourse and dialogue structure, including the specific structural aspects of (interactive) digital media; * prosody in discourse and dialogue; * modeling politeness and non-recursive parts of discourse and dialogue; * models of presupposition and accommodation; * operational models of conversational implicatiire.

Malcolm Coulthard What Did They Actually Say? A Forensic Linguist's Evaluation of Police Evidence

1. Introduction

One major type of evidence presented by the police at trial is their record of what the accused said. This evidence comes in two forms - records of interviews with the accused and of statements made by the accused. The police in England, since the Police and Criminal Evidence Act (PACE, 1984) which was introduced following a spate of claims that those accused of crimes were frequently 'verballed', are now required, whenever possible, to make a contemporaneous audio or video recording of verbal evidence. However, the three cases I will discuss in this paper pre-date PACE and come from a time when it was customary to record what was said in interviews contemporaneously and laboriously in longhand. At that time police officers were very much in control of the verbal evidence: not only did they control the topics covered in interviews with suspects, they also controlled what was recorded and what not recorded and, even more significantly, the form in which it was recorded. Of course, it must be said that anyone who has tried to make a verbatim record, whether in real time, or with the assistance of repeated listenings to a tape-recording, knows that, with the best will in the world, it is literally impossible. Police officers were never given any explicit guidelines as to what did and did not constitute a 'verbatim' record, nor what they could and could not legitimately omit; but equally the courts never seriously questioned whether their records were indeed verbatim. It was apparently generally accepted unchallenged that police officers could record contemporaneously what was said with sufficient accuracy for the actual wording of utterances, rather than just their content, to be used in evidence. Indeed, apparently it was even believed that officers could remember substantial chunks of a conversation and then write them down verbatim several hours later. These pre-PACE records are fascinating, because they are, on the one hand, factual records of interaction, but, on the other, texts whose function is to re-present this interaction at a later time to a different audience for a different purpose. So, they have not only an evidential, but also a persuasive function. Indeed, the police participants were certainly aware, at the time of the primary interaction, that the record was intended for, and

Malcolm Coulthard

94

therefore could be designed for, another audience - certainly some records appear to be consciously and specifically constructed with the future audience in mind. In an earlier paper (Coulthard 1996), I specifically drew attention to a number of these features.

2. A case of persuasive dialogue

The first case I will discuss is that of Ashley King, who was accused and convicted of battering an old lady to death. The police presented in evidence records of ten interviews. In the early interviews King admits to having known the deceased and to having visited her to ask for money on the day she was murdered, while in a later interview he admits to having hit her on the head, although this admission is immediately withdrawn in the following interview and never made again. Text la below is the beginning of the first interview in exactly the form in which it was presented in Court - except that I have highlighted some words in bold in order to be able to refer back to them in the discussion which follows. Text la - Extract from the record of an interview with Ashley King KING was then transported to Houghton-le-Spring Police Office where at 11.35 am that day he was interviewed in the following manner by myself and D C Simpson 1. Q. "When we saw you last Saturday you told us that you didn't go out at all on the evening of Monday 4th November. Are you still saying that's true?" 2. A. "Yes, I'm sure that's right." 3. Q. "It can be difficult thinking back Ash, so think very carefully, it's very important." 4. A. "Aye, I'm sure about that. I never left the house." 5. Q. "We've interviewed some people who say you were out that night?" 6. A. "I don't think so." 7. Q. "Think very carefully about it, it's important." 8. A. "I think that was the night we got the tyres, is that what you mean." 9. Q "Tell us about it." 10. A. "Billy, Berti and Melley called for us and we went up the bank and got some tyres." 11. Q. "What time was this?" 12. A. "They came for us about six o'clock." 13. Q. "Who's Berti?" 14. A. "I don't know his second name but he's from Shiney." 15. Q. "When you say you got some tyres, do you mean you stole them?" 16. A. "Aye, I think so" I then cautioned KING. 17. Q. "Where did you get the tyres from?" 18. A. "It's a place next to the garage, its near the pit." 19. Q. "Do you know the name of the premises?"

20. A.

"Tyre Services I think, but I'm not sure." (The interview record continues for another 3 pages)

A Forensic Linguist's Evaluation of Police Evidence

95

There are several points which immediately strike the reader, and which would have similarly struck the members of the jury, because this would have been their first contact with King. Firstly, following the prose introduction we are presented with a verbatim account of what was said by King and the interviewing officer. Secondly, King is seen not tell the truth when first questioned - the interview begins with three denials that he had been out on the night of November 4 th , utterances 2, 4 and 6, followed by a grudging admission, 8. "I think that was the night we got the tyres, is that what you mean". Shortly afterwards he admits, in utterance 16, that, although he used the word 'got', he knew he was in fact stealing the tyres. Thirdly, the police officer is apparently very patient and good-natured in his questioning - he offers a let-out after the first denial "It can be difficult thinking back Ash, so think very carefully," and doesn't directly contradict King but, rather, coaxes the truth out of him. Although cases are decided on evidence, often the value of that evidence depends on the credibility of the witness and juries are repeatedly reminded to evaluate each witness when considering their evidence. In these early exchanges we can see King being constructed as untrustworthy, as someone who needs to have the truth gradually drawn out of him, whereas the police officer is presented as a kindly patient character. This is of tremendous value to the prosecution, because their major problem is that King only confessed once and then only in the eighth interview and the defence will deny the confession and argue that there was undue police pressure. In this context the importance of the beginning of Interview 1 is that it shows a very different picture - a sympathetic interrogator and a shifty suspect. It is therefore something of a surprise to discover that Text la is in reality not a verbatim record at all, but rather a persuasive dramatization based on minimal

trigger

notes, notes which were only written up several hours later. The actual trigger notes are presented below in their original layout: Text l b - Police Trigger Notes 6pm Left house called for Billy Waugh &Melvin Waugh + (Berty) from Shiney Row TO TYRES Place BY PIT (TYRE SERVICES?) (The notes continue for another 15 lines.) A quick glance at the words I have highlighted in bold in Text l a above, the only parts to be based directly on the trigger notes, shows just how little there was in the trigger notes and therefore how much has been (re)created by the police from either memory or imagination. A reading of the monologue, Text lc below, which I offer as an alternative way of presenting the trigger note information to the jury, emphasizes just how much of the persuasiveness of the Text la comes from the police dramatization into dialogue.

Malcolm Coulthard

96

Text lc - A Possible Monologue Expansion King said he left home at 6pm and called for Billy Waugh, Melvin Waugh and Berty from Shiney Row. They went to the tyres place near the pit, which he thought was called Tyre Services.

In this version, gone are the two contrasting characters, the shifty King and the kindly policeman and all that remains is an uninteresting short narrative of events, with none of the evaluation of character and action which makes the dramatization so persuasive.

3. A case of supportive dialogue

This second case concerns the death of a teenage boy who was delivering newspapers to a remote farmhouse. Four men were arrested in connection with the murder and one of them, Patrick Molloy, confessed to having been involved in the murder and incriminated the other three. He later withdrew his confession, claiming that in the period leading up to making it he had been physically and verbally abused. He did not deny making the 'confession', but said he had been merely repeating what one of the police officers had spoken into his ear. In other words, his claim was that he was merely the mouthpiece and not the author of the confession. Nevertheless, on the basis of this confession all four were convicted of murder. After a campaign lasting some 16 years the men were eventually cleared by the Appeal Court, although Molloy had by then died in prison. One of the reasons that the confession was believed by the jury was that it was supported by a record of an interview, timed as occurring immediately before the confession, which contained essentially the same material. Molloy insisted that this interview had never occurred and that the text was a total fabrication. I present below as Text 2a the first two thirds of the confession statement - in what follows I will concentrate on sentences 17-25 which constitute the incriminating part. Text 2a - Extract from Molloy's Statement ( I ) 1 need to tell you the truth, you wouldn't understand the pressure that I have been under. (2) I am terrified of the others, I have been threatened with personal injury. (3) I know they will do it but even so I must tell you the truth. (4) I was at the farm when that lad, the paper boy was killed. (5) I was upstairs searching for something of value, anything, money or coins. (6) Four of us had gone to the farm. (7) There were two motors, a blue Cortina Estate, which I went in with Vinny Hickey who was driving and his relation Micky. (8) I sat in the back. (9) Jimmy Robinson drove a Van I think it had a white top, it belonged to someone in the Dog and Partridge, he borrowed it for the job. (10) We arrived at the farm first and waited for Jimmy who arrived shortly afterwards. ( I I ) We parked both motors away from the farm and walked down to the farm. (12) We didn't all go together, me and Vinny walked down first. (13) We waited and the others joined us. (14)

A Forensic Linguist's Evaluation of Police Evidence

97

Jimmy broke in through a window and loosed us in. (15) They went downstairs and I went upstairs by myself. (16) I searched the bedrooms I remember taking the drawers from some furniture and after searching them I stacked them one on top of the other. (17) I had been drinking and cannot remember the exact time I was there but whilst I was upstairs I heard someone downstairs say be careful someone is coming. (18) I hid for a while and after a while I heard a bang come from downstairs. (19) I knew that it was a gun being fired. (20) I went downstairs and the three of them were still in the room. (21) They all looked shocked and were shouting at each other. (22) I heard Jimmy say, "It went off by accident". (23) I looked and on the settee I saw the body of the boy. (24) He had been shot in the head. (25) I was appalled and felt sick. Below is the corresponding part of the disputed interview. What is immediately evident is that not only are the content and sequencing remarkably similar, but so also is the wording of the utterances attributed to Molloy - I have highlighted in bold sequences that are identical to phrases and clauses in the statement: Text 2b - Extract from Police Interview with Molloy P. How long were you in there Pat? (17) I had been drinking and cannot remember the exact time that I was there, but whilst I was upstairs I heard someone downstairs say 'be careful someone is coming'. P. Did you hide? (18) Yes I hid for a while and then I heard the bang I have told you about. P. Carry on Pat? (19) I ran out. P. What were the others doing? (20) The three of them were still in the room. P. What were they doing? (21) They all looked shocked and were shouting at each other. P. Who said what? (22) I heard Jimmy say "it went off by accident". P. Pat, I know this is upsetting but you appreciate that we must get to the bottom of this. Did you see the boy's body? (23) Yes sir, he was on the settee. P. Did you see any injury to him? (24) Yes sir, he had been shot in the head. P. What happened then? (25) I was appalled and felt sick. Apparently, the persuasive effect in Court of the close textual similarity was to reinforce the credibility of both texts, whereas the same phenomenon would have had just the opposite effect on linguists, who, very conscious of the uniqueness of utterance, are suspicious of such closeness. In arguing against the independent production of the two texts it was necessary not simply to assert the linguist's view on uniqueness, but also to demonstrate how one text had indeed been derived from the other.

98

Malcolm Coulthard

Obviously what is not shared by the two texts is firstly, the overt presence of the police interviewer as interlocutor and secondly, some of the content contained in the police utterances. If Molloy was right in asserting that it was the interview record which was a fabrication, then one would expect to discover some clues to this because, in essence, what we have is a policeman/dramatist constructing a dialogue from a monologue text. This is not as simple a task as it might seem, because it involves first dividing up the monologue text into candidate answers and then devising questions for these pre-existing answers. One would expect, and we do indeed find, problems with cohesion: 1. a tendency to link forward to the already known answer rather than backward to the preceding answer. We can see this in the pair of exchanges (21/22) (21) They all looked shocked and were shouting at each other.

P.

Who said what?

(22)

I heard Jimmy say "it went off by accident".

Following (21) one would expect "Who shouted what" or "What were they shouting?"; there is no reason to use the verb 'say' - except that it is already contained in the phrasing of the answer which follows. 2. an inability to create the appropriate question form P. What were the others doing? (20) The three of them were still in the room. P. What were they doing? (21) They all looked shocked and were shouting at each other

Here we see two successive examples of a 'doing' question apparently receiving a 'being' response 3. problems with grammatical cohesion P.

Did you see the boy's body?

(23)

Yes sir, he was on the settee.

Here we would expect to find the pronoun 'it' referring back to the body, but instead we find 'he'. It looks as if the 'he' has been borrowed from the statement where the sequence is I saw the body of the boy. (24) He had been shot

and there 'he' is an acceptable referent.

A Forensic Linguist's Evaluation of Police Evidence

99

Thus, the linguistic evidence suggests, as Molloy had always asserted, that a complete interview record was created by dialogising the original monologue statement, in order to use the persuasive power of an apparently contemporaneously recorded dialogue to help convict.

4. A case of concealed dialogue

One November evening in 1952 two teenagers, Derek Bentley aged 19 and Chris Craig aged 16 tried to break into a warehouse. They were seen, as they climbed up onto the roof, by a woman who was putting her daughter to bed. She called the police, who arrived soon afterwards and surrounded the building. Three unarmed officers, two in uniform, the other in plain clothes, went up onto the roof to arrest the boys. Bentley immediately gave himself up. Craig drew a gun, started shooting and eventually killed a police officer. Bentley was jointly charged with murder, even though he had been under arrest for some considerable time when the officer was shot. At the trial, which only lasted two days, both boys were found guilty. Craig, because he was legally a minor, was sentenced to life imprisonment; Bentley was sentenced to death and executed shortly afterwards. Bentley's family fought for a generation to overturn the guilty verdict and were eventually successful 46 years later in the summer of 1998. The evidence which was the basis for both Bentley's conviction and the successful appeal was in large part linguistic and will be the topic of the rest of this paper. In the original trial the problem for the Prosecution, in making the case against Bentley, was to demonstrate that he could indeed be guilty despite being under arrest when the murder was committed. At this point it would be useful if you first read through the statement presented below as Text 3 which, it was claimed, Bentley had dictated shortly after his arrest - the only change I have introduced is to number the sentences for ease of reference. Text 3 - Derek Bentley's Statement (1) I have known Craig since I went to school. (2) We were stopped by our parents going out together, but we still continued going out with each other - I mean we have not gone out together until tonight. (3) I was watching television tonight (2 November 1952) and between 8 p.m. and 9 p.m. Craig called for me. (4) My mother answered the door and I heard her say that I was out. (5) I had been out earlier to the pictures and got home just after 7 p.m. (6) A little later Norman Parsley and Frank Fasey called. (7) I did not answer the door or speak to them. (8) My mother told me that they had called and I then ran out after them. (9) I walked up the road with them to the paper shop where I saw Craig standing. (10) We all talked together and then Norman Parsley and Frank Fazey left. (11) Chris Craig and I then caught a bus to Croydon. (12) We got off at

Malcolm Coulthard

100

West Croydon and then walked down the road where the toilets are - 1 think it is Tamworth Road. (13) When we came to the place where you found me, Chris looked in the window. (14) There was a little iron gate at the side. (15) Chris then jumped over and I followed. (16) Chris then climbed up the drainpipe to the roof and I followed. (17) Up to then Chris had not said anything. (18) We both got out on to the flat roof at the top. (19) Then someone in a garden on the opposite side shone a torch up towards us. (20) Chris said: 'It's a copper, hide behind here.' (21) We hid behind a shelter arrangement on the roof. (22) We were there waiting for about ten minutes. (23) I did not know he was going to use the gun. (24) A plain clothes man climbed up the drainpipe and on to the roof. (25) The man said: Ί am a police officer - the place is surrounded.' (26) He caught hold of me and as we walked away Chris fired. (27) There was nobody else there at the time. (28) The policeman and I then went round a corner by a door. (29) A little later the door opened and a policeman in uniform came out. (30) Chris fired again then and this policeman fell down. (31) I could see that he was hurt as a lot of blood came from his forehead just above his nose. (32) The policeman dragged him round the corner behind the brickwork entrance to the door. (33) I remember I shouted something but I forgot what it was. (34) I could not see Chris when I shouted to him - he was behind a wall. (35) I heard some more policemen behind the door and the policeman with me said: Ί don't think he has many more bullets left.' (36) Chris shouted 'Oh yes I have' and he fired again. (37) I think I heard him fire three times altogether. (38) The policeman then pushed me down the stairs and I did not see any more. (39) I knew we were going to break into the place. (40) I did not know what we were going to get - just anything that was going. (41) I did not have a gun and I did not know Chris had one until he shot. (42) I now know that the policeman in uniform that was shot is dead. (43) I should have mentioned that after the plain clothes policeman got up the drainpipe and arrested me, another policeman in uniform followed and I heard someone call him 'Mac'. (44) He was with us when the other policeman was killed.

Bentley's barrister spelled out for the jury the two necessary pre-conditions for them to convict: they must be "satisfied and sure": 1.

that [Bentley] knew Craig had a gun and

2.

that he instigated or incited Craig to use it. (cf. Trow 1992: 179)

The evidence adduced by the Prosecution to satisfy the jury on both points was essentially linguistic. To support point 1) it was observed that in his statement, which purported to give his unaided account of the night's events, Bentley had said "I did not know he was going to use the gun", (sentence 23). In his summing up, the judge, Lord Chief Justice Goddard, made great play with this sentence, arguing both that its positioning in the narrative of events, before the time when there were any policemen on the roof and also the choice of "the gun" (as opposed to "a gun") must imply that Bentley knew that Craig had a gun well before it was used - in other words "the gun" at that position in the statement must be taken to mean "the gun I already knew at this point in the narrative that Craig had". In addition, the judge argued, this sentence also showed Bentley to be an unreliable witness, because he contradicted himself later, in sentence 41, by saying "I did not know Chris had [a gun] until he shot".

A Forensic Linguist's Evaluation of Police

Evidence

101

The evidence to support point 2), that Bentley had instigated Craig to shoot, was that the police officers in their statements and in their evidence given in court, asserted that Bentley had uttered the words "Let him have it, Chris" immediately before Craig had shot and killed the policeman. Bentley, supported by Craig, had denied uttering these words - a claim supported much later by a fourth policeman, who was never called to give evidence at the trial. Bentley's defence, in the words of the judge in his summing-up, was: I didn't know he had a gun and I deny I said "Let him have it, Chris". I never knew he was going to shoot and I didn't think he would.' Against that denial, (which, of course, is the denial of a man in grievous peril), you will consider the evidence of the three main officers who have swom to you positively that those words were said {Ibid.·. 109).

So, as the judge emphasized, the strength of the linguistic evidence depended essentially on the credibility of the police officers who had collected it and sworn to its accuracy. When the case came to Appeal in 1998, one of the Defence strategies was to challenge the reliability of the statement. If they could throw doubt on the veracity of the police they could mitigate the incriminating force of both the statement and the phrase "Let him have it" which Bentley, supported by Craig, had denied uttering - a claim supported much later by a fourth policeman, who was never called to give evidence at the trial.

4.1. The Linguistic Evidence At trial three police officers swore on oath that Bentley's statement was the product of unaided monologue dictation, whereas Bentley asserted that it was, in part at least, the product of dialogue - in other words that the police questions and his replies had been converted into monologue. Q A Q A

Did you in fact dictate that statement as it is written down No Sir. How was it taken from you In questions sir. (Court transcript pp. 100-1)

There is no doubt that this was one procedure used at the time for producing statements - a senior officer in another murder case a year later explained to the Court how he had elicited a statement from another accused in exactly this way: I would say "Do you say on that Sunday you wore your shoes?" and he would say "Yes" and it would go down as "On that Sunday I wore my shoes" (Hannam transcript: 156)

102

Malcolm Coulthard

There are, in fact, many linguistic features which suggest that Bentley's statement is not, as claimed by the police, a verbatim record of what he dictated, and I have written about these elsewhere (Coulthard 1993); here I will confine myself simply to evidence that the statement was indeed, at least in part, dialogue converted into monologue. One of the first things that strikes one on reading the statement is that the narrative of events is fairly coherent up to utterance (38) "The policeman then pushed me down the stairs and I did not see any more". What follows in sentences 39-42 appears to be some kind of meta-narrative whose presence and form are most easily explained as the result of a series of clarificatory questions about Bentley's knowledge at particular points in the narrative, information which the police knew would be very important later at the trial: (39) I knew we were going to break into the place. (40) I did not know what we were going to get - j u s t anything that was going. (41) I did not have a gun and I did not know Chris had one until he shot. (42) I now know that the policeman in uniform that was shot is dead.

At first one might not attach too much importance to these post-narrative questions - they certainly do not materially change the narrative which Bentley has already told - on the contrary, they allow him to clarify what he knew and did not know and also give him a chance to assert his lack of any pre-knowledge of Craig having a gun. Indeed the fact that these may have been elicited rather than spontaneously offered did not seem to trouble the Prosecution. At trial, when he was asked specifically about the two sentences (39-40) Bentley replied "That was an answer, sir, to a question", to which the barrister replied "I daresay it was in reply to a question" and then moved on immediately to a new topic. However, this passing acknowledgement does reinforce Bentley's claim and prompts us to look for evidence of multiple voices elsewhere in the statement. We do so, of course, in the knowledge that there will always be some transformations of Q-A which will be indistinguishable from authentic dictated monologue. In the example quoted above, had we not been told that "On that Sunday I wore my shoes" was a reduction from a Q-A we would have had some difficulty in deducing it, although the pre-posed adverbial 'On that Sunday' is certainly a little odd. We can begin our search with the initial observation that narratives, particularly narratives of murder, are essentially accounts of what happened and to a lesser extent of what was known or perceived and thus reports of what did not happen or was not known are rare and special - there is an infinite number of things that did not happen and thus the teller needs to have some special justification for reporting any of them to the listener, i.e. there must be some evident or stated reason for them being newsworthy. We can see a typical example of the 'normal' usage of a negative early in Bentley's statement:

A Forensic Linguist's Evaluation of Police Evidence

103

(6) A little later Norman Parsley and Frank Fasey called. (7) I did not answer the door or speak to them When Bentley reported that his friends had called the listener would reasonably expect him to at least have talked to them and therefore here there is a quite natural denial of a reasonable expectation. However, there are, in Bentley's statement, some negatives which have no such narrative justification, like sentence (17) below: (16) Chris then climbed up the drainpipe to the roof and I followed. (17) Up to then Chris had not said anything. (18) We both got out on to the flat roof at the top. Chris is not reported as beginning to talk once they have got out onto the roof, nor is his silence contrasted with anyone else's talking, nor is it made significant in any other way later in the narrative. A similarly unwarranted example is: (26) He caught hold of me and as we walked away Chris fired. (27) There was nobody else there at the time. (28) The policeman and I then went round a comer by a door. None of the possible inferences from the denial seem to make narrative sense here - i.e. "that as a result of there being no one else there..." 1.

it must be the policeman that Craig was firing at,

2.

that it must be Craig who was doing the firing,

3.

that immediately afterwards there would be more people on the roof.

So the most reasonable conclusion is that at this point in the statement-taking a policeman, trying to clarify what happened, asked a question to which the answer was negative and the whole sequence was then recorded as if Bentley had made a negative statement. The fact that some of the statement may have been elicited in this way becomes particularly important in relation to sentence (23): (23) I did not know he was going to use the gun which is the one singled out by the judge as incriminating. This sentence would only make narrative sense if it were linked backwards or forwards to the use of a gun - in other words if it was placed immediately following or preceding the report of a shot. However, the actual context is:

104

Malcolm Coulthard

(22) We were there waiting for about ten minutes.

(23) I did not know he was going to use the gun. (24) A plain clothes man climbed up the drainpipe and on to the roof.

If it is accepted that there were question/answer sequences underlying Bentley's statement, then its logic and the sequencing of information were not under his direct control. Thus, the placing of the reporting of some of the events must depend on a decision by the police questioner to ask his question at that point, rather than on Bentley's reconstruction of the narrative sequence and this, crucially, means that the inference drawn by the judge in his summing up is unjustified. If one were able to assume that the statement was Bentley's unaided narrative then the positioning of "I didn't know he was going to use the gun" would be significant, because there is no timing adverbial and so the observation would have to take its time from its position in the narrative. In that view Bentley appears to be reporting that, well before the gun was used, he knew of its existence. However, if this sentence is in fact the product of a response to a question with its placing determined by the interrogating police officers, then there is no longer any conflict with his later denial "I did not know Chris had one [a gun] until he shot". Nor is there any significance either in Bentley being reported as saying "the gun" - all interaction uses language loosely and co-operatively. If the policeman had asked Bentley about "the gun" Bentley would assume they both knew which gun they were talking about and the sensible interpretation would be "the gun that had been used earlier that evening" and not "the gun that was going to be used later in the sequence of events that made up Bentley's own narrative". Being able to demonstrate that the statement was indeed a jointly produced document, in part authored by the police officers who recorded it, both removed the incriminating value of the phrase "I didn't know he was going to use the gun" and undermined the credibility of the police officers on whose word depended the evidential value of the other incriminating utterance "Let him have it Chris". In August 1998, 46 years after the event, the Lord Chief Justice, sitting with two senior colleagues, criticized his predecessor's summing-up and allowed the Appeal against conviction.

5. Conclusion

In focusing in detail on these three cases of invented and concealed dialogue I hope that I have shown how the medium is part of the message. In other words, by simply being in control of the written form in which the original interaction was to be subsequently pre-

A Forensic Linguist's Evaluation of Police Evidence

105

sented to the court, the police were able to play down their own 'voice' and to choose a mis-representation which would significantly influence the outcome of the case.

References Coulthard, R. M. (1993): On beginning the study of forensic texts: corpus, concordance, collocation. - In: M. Hoey (ed.) Data, Description, Discourse, 86-114. London: HarperCollins. Coulthard, R. M. (1996): The official version: audience manipulation in police reports of interviews with suspects. - In: C. R. Caldas-Coulthard and R. M. Coulthard (eds.) Texts and Practices: Readings in Critical Discourse Analysis, 164-176. London: Routledge. Hannam - extract from the examination of Chief Inspector Hannam in the trial of Alfred Charles Whiteway in 1953. Trow, M. J. (1992): Let Him Have it Chris. - London: HarperCollins.

Svetla Cmejrkovä Media Dialogue as a Genre of Public Oral Discourse

1. Confrontational pattern of public oral discourse

In modern society, public figures seldom get the opportunity to express their ideas monologically. More often, they find themselves in the crosshairs of journalists' questions, regularly confronting their opponents in fierce controversies, discussions, and debates. They act as the protagonists of various media programmes that are designed not only to inform but also to entertain their audience. When Edda Weigand (1999: 38) described such programmes in one of her recent IADA papers, entitled "Dialogue in the grip of the media", she identified them as a modern variant of the ancient gladiatorial spectacle. An analogous image is used by John Heritage (forthcoming), in his conversational analysis of the TV interview: In the Roman arena, gladiators fought with trident and net against shield and short sword. In the modern news interview, the ground rules equip the protagonists with similarly asymmetric resources for attack and defence.

This conception of dialogue as a fierce battle between the interviewer and an invited guest intended for public consumption and entertainment seems to be very remote from the prototypical ideal of dialogue, in which, according to Plato, Sokrates encouraged others to speak, using the maieutic method of eliciting and clarifying the ideas of one's counterparts to reveal new and unexplored aspects of the topic under discussion. Does media dialogue have anything in common with real dialogue, or is it rather a travesty of real debate? As Edda Weigand (1999: 36) states, in the grip of the media, dialogue loses some characteristics of its natural form and gains other potentials. To what extent does mediation transform ordinary, private dialogue and private conversation?

108

Svetla Cmejrkovä

2. Dialogue as an encounter of two people and the format of the TV interview

In the course of the last ten years, the model of the broadcast news interview has been examined and described within the methodology of conversational analysis, particularly by John Heritage (1985, 1991, forthcoming), Stephen d a y m a n (1991, 1993), and David Greatbatch (1998). These scholars have provided valuable insights into how the genre of media talk operates for its participants. In their examination into how the news interview is structured around questions and answers, CA researchers stress how it diverges from ordinary conversation. In interviews - which are viewed as an institutionalised variant of ordinary conversation - the dialogical interaction adheres to a certain format and uses a certain routine broadcast technology: the interviewer asks questions and the interviewee answers them. If these constraints are not adhered to, the interviewer sanctions the digression and retrieves the interview so that it once again operates within the constraints (Bell/Garett 1998: 1-20, Greatbatch 1998: 164-165). The methodological value of the model is immense. However, upon applying the model we discover, together with CA researchers, that an interview is above all an encounter of two people, two individuals, whose discursive practices transgress those anticipated by their institutional roles and whose dialogical interaction shares many elements in common with other types of discourse, including casual conversation, ordinary dialogue, discussion, and controversy.

3. Communication and technology

To explain my approach to interviews, and possibly to shed some light on the terms I have quoted here, I will use a metaphor of communication in its relation to technology that is introduced in Frederick Erickson's (1986) article "Listening and Speaking". Erickson quotes an eminent scholar in artificial intelligence, who commented on the theory of dialogical interaction. The quote goes as follows: "What you have been telling me [about dialogical interaction] reminds me that in my own work we've discovered that it's impossible for humans to build a machine that can climb a tree". Erickson claims that the impasse that results when a machine with in-built general intelligence is confronted by the unique particularity of an actual tree is akin to the problems that are

Media Dialogue as a Genre of Public Oral Discourse

109

continually encountered and solved by participants engaged in face-to-face interaction. Erickson concludes: 'Talking with another person is even more tricky than climbing an actual tree. It is like climbing a tree that climbs back."

Erickson's metaphor seems to be very pertinent to all forms of dialogue. It can be applied not only to ordinary dialogue, but also, in my view, to media dialogues and interviews. The metaphor illustrates what we often find in interviews: the interviewer's effort to skilfully climb the interviewee and to make him or her a subject to this effort, and on the other hand, the interviewee's evasiveness and resistance. This led me to recall Erickson's metaphor when analysing interviews broadcast on Czech TV. Though interviews represent an institutionalised form of ordinary dialogue with a macrostructure that is governed by the institutional principles mentioned above - their distance from ordinary dialogue, as it seems to me, is not as great as is often assumed. We find here those acts that feature dialogue in its authentic form. Private face-to-face conversation can be seen as a pretext for the development of the genre of media dialogue, and as a source of its constant revival and regeneration. We cannot deny that a great deal of the authenticity, spontaneity, originality and unpredictability of casual dialogue is inevitably lost in its mediated variety. Though authentic dialogue, including casual talk, also has its conventions, the participants, nevertheless, are much less constrained, and they always have the possibility of choice. They can choose to either follow the conventions or to diverge from them and take steps neither tested nor known in advance, or they can even give up and jump down from the tree. Meaning is constructed by the participants in the course of the dialogue, always with respect to a given situation. The parallel to climbing a unique tree that climbs back is obvious. In media dialogue, the construction of meaning by the dialogue participants is not only negotiated during the dialogue, but it is also determined in advance by a set of frameworks and expectations. Dialogue takes on an institutionalised form that is based on routine discourse practices. We could imagine here - in the place of a skilled and routine interviewer - a machine with in-built intelligence that is supposed to manage the interview.

4. The construction of meaning in interviews

It cannot be denied that the asymmetry of opportunities available to the participants in order to achieve their goals in the interview is very distinctive. The rules that structure the questioning

110

Svetla Cmejrkovä

and answering in the news interview provide the main protagonists - interviewer and interviewee - with different resources for dealing with one another. An interview resembles a legal cross-examination.

4.1. Asymmetry of opportunities and a tendency towards reciprocal discourse Interviews, however, have at least a tendency to become a reciprocal discourse. Occasionally, the interviewee himself/lierself starts asking questions: I-ee: May I also ask you a question? I-ee: Do you really think that's a good topic for discussion? I-ee: Would it not be better to discuss this instead of that? I-ee: Will the viewer understand? Will it be interesting for the viewer?

The tendency towards a reciprocal discourse can be observed especially in media programmes where there is more than one interviewee. In such news interviews, an interviewer confronts two or more experts whose opinions, as a rule, present opposing approaches to a particular issue. Though the speakers do not necessarily engage in an overt dialogue with each other (instead they merely answer the interviewer's questions), a dialogue inevitably develops between them (Cmejrkovä 2000). The invited guests do not necessarily address each other, but they are alert to each other's turns and sometimes they may even initiate their own reciprocal discourse, addressing each other and thereby challenging the role of the interviewer, who becomes a mere witness to their polemics. The interviewer's role may also be straightforwardly challenged by a guest posing an opening question that threatens to take the initiative from the interviewer's hands. For example, the Czech prime minister once addressed an interviewer, asking her: I-ee: How did you enjoy your holiday?

It appears that those speech acts which are forbidden in interviews, or should be avoided within their format, try to reassert their rights, creeping into interviews through the back door. Interviews are not an exception to the general tendency established by Norman Fairclough (1995) and Andrew Tolson (1991) in their analyses of media discourse: they combine the language of face-to-face communication with the language of mass communication, conversationalising media discourse and accommodating the discursive practices of the private domain.

Media Dialogue as a Genre of Public Oral Discourse

111

4.2. Negotiations of meaning The tendency towards a reciprocal discourse as manifested in the posing of questions not only by the interviewer but by the interviewee as well is just a phenomenological aspect of the interview's internal dynamics. The reciprocity of discourse consists above all in the fact that the participants in an interview enter it with different approaches to the issue under discussion and they negotiate - through the given format - the way in which it will be discussed. The interviewee evaluates the interviewer's questions, comments on the appropriateness of asking, and discusses the interviewer's right to pose a given question, negotiating its topic and formulation. At the same time, the interviewer evaluates the interviewee's answers, their exactness or ambiguity, directness or evasiveness, often with regard to the audience. I will begin by quoting examples of the interviewees' evaluating reactions: I-ee: That's a good question, I think it's one of the basic problems we have to solve. I-ee: I appreciate your question and I hope you will give me an opportunity to answer it. I-ee: That's a good question. No comment. I-ee: It's not a question of isolation I repeat it's a question of the speed of the process. I-ee: I don't think it's about that. That isn't how I would put it. I-ee: That's not the best way to put it. I'd prefer to put it this way.

What the interviewer asks often does not coincide with what the interviewee not only wants to say, but also what he or she thinks, what he or she finds true to nature of the problem under discussion. Very often he or she finds the interviewer's questions to be simplified, banal, inappropriate, or provocative: A young journalist interviews a famous Czech opera singer living abroad and asks her: I-er: As far as I know you have never sung at the National Theatre. Is there a role you would accept?

The opera singer avoids the question charmingly: I-ee: Well we'll put the question aside and go on.

What we see behind the format of questioning and answering is a process in which the necessity of answering the question, and especially the appropriateness of the question, are constantly negotiated by the participants, often in consideration regard to the audience, as I have already mentioned: I-ee: I think viewers would first of all be interested in..., may not know... I-ee: The viewer may not remember..., may not understand... I-er: We should not underestimate the viewer. I-ee: I would not dare to underestimate the viewer.

112

Svetla Cmejrkovä

At the same time, the interviewer evaluates the interviewee's answers, either assessing their definiteness or uncovering their evasiveness and ambiguity. I-er: Do you think that the hostile troops will agree to that? I-ee: I think it's a very probable scenario. I-er: I asked: yes or no? I-ee: Yes. I-er: Do you think that the hostile troops will agree to that? I-ee: Yes. I-er: How will they agree to that?

However, what is mainly negotiated in interviews is not only the correctness and the appropriateness of questions and answers, but the construction of meaning in the course of an interview. As has been said, both participants may enter an interview with completely different approaches and ideas and negotiate whose discourse will be led, whose view of things will prevail. This process accounts for two strategies that are very typical of interviews: the interviewer's reformulation of the interviewee's words and the interviewee's reformulation of the interviewer's questions.

4.3. Reformulation o f statements and questions The interviewer very often reformulates the interviewee's statements and adapts them to his or her interpretation of things, e.g.: I-er: You want to say that..., You are saying that..., Are you saying that... I-er: So you're saying that..., What you're saying means that..., etc.

The interviewee may agree: I-er: So there is a clear conflict of interests according to you. I-ee: According to my mind there is a clear conflict of interests, yes.

The interview may agree only partially: I-er: You want to say that there are sufficient funds for health insurance objectives. I-ee: And you know we are at the point now. I want to say that there are sufficient funds in principle,

but... Or the interviewee may disagree with the interviewer's reformulations: I-ee: That's not what I said. I said..., No, I didn't say that, I said..., Wait I did not say that, don't distort my words...

Media Dialogue as a Genre ofPublic Oral Discourse

113

The interviewer's reformulation of the interviewee's answers is just one side of the coin. There is an analogous process on the side of the interviewee, who reformulates the questions posed by the interviewer (dayman 1993). Sometimes the interviewee answers questions straightforwardly, as in: I-er: Who should be elected, in your opinion? I-ee: In my opinion, Väclav Havel should be elected.

Answers that would keep to a syntactic framework and the question's wording are, however, rare. In many cases, before answering, interviewees attempt to paraphrase the question, to reformulate it and to make it match their own structuring of reality: I-er: Is it a regular election campaign? Don't you find the billboard aggressive? I-ee: Well, the billboard, I do not want to deal with that. In this respect, I am interested in one thing.

4.4. The interviewee's fight for space and the interviewer's constraints While with some questions it is easy to identify what a straightforward answer to a direct question would be, with many others this becomes difficult or nearly impossible. There is a distinction to be made between yes/no questions and wh- questions and, particularly, between so called closed questions and those which are open and give the interviewee space to expand on the topic. As answers to open questions (such as What do you think about this and that?, e.g. What is your personal view of globalisation) can be less controlled, there is a favourite questioning strategy frequently used by interviewees, consisting in the formulation of alternative questions. There is a general assumption that things are either white or black (hence, they can always be put into one of two groups) and that ideas and attitudes form dichotomies. Alternative questions force an interviewee to choose one of the two alternatives formulated in the interviewer's question: I-er: Are you of this opinion or not? Do you agree or not?

Alternative questions not only ignore the possibility of gradation and fuzzy boarders between phenomena, but they also create the impression that the underlying assumption is legitimate, that the underlying structuring of reality created by an interviewer is correct, and what remains is only to answer yes or no to the dichotomy they have presented. Alternative questions, and especially the interviewer's insisting on their acceptance, seems to contradict the very basis of the Sokratic method, whereby there is an eliciting of new, unanticipated aspects of the

114

Svetla Cmejrkovä

phenomenon under discussion. It is the interviewee's task, then, to differentiate between the different aspects of the underlying assumption: I-ee: Globalisation cannot be considered either bad or good. It is difficult to discuss something as complex as globalisation as being either bad or good. I-er: And your personal opinion, let's say we have a scale of good and bad, where is your position on the scale, where is your place? I-ee: I try to be everywhere and if there are whole numbers on the scale I try to see those decimal ones in-between. As I see it, the interview presents a confrontation of different points of view and their respective duelling discourses, in which the interviewer's effort to manage the interview's question-answer format within the constraints under which it is expected to operate clashes with the interviewee's effort to resist the interviewer and to climb back. The Interview is a striking example of manifested otherness, of the effort to express difference. Whereas the interviewer often insists on whole numbers and a rather rough structuring of reality, the interviewee's vision of reality is usually more subtle and differentiated, which is why he or she often tries to make further distinctions. In this respect, the interview is a striking instance of climbing the unique tree that climbs back, revealing a plurality of possibilities. The less auspicious image of an interview is due to the fact that the interviewer often forces his or her interviewee to manoeuvre in a limited space, trying to constrain him or her. The insistence on yes/no questions and the requirement of yes/no answers is one of the elements that makes the interview parallel to a gladiatorial spectacle. Alongside this type of interview that resembles a legal cross-examination, there is a very different one, which can be called a "partnership" type of interview. The interviewer genuinely gives the interviewee room to express his or her ideas freely, encouraging him or her to elaborate on the view he or she expresses. The interviewer defers to the position of interviewee and expresses understanding. Only exceptionally would the interviewer try to express a different opinion which, if it did not fit the interviewee's view, would be withdrawn. Such a type of an interview meets the requirements laid out by Emmanuel Levinas (1961), a French philosopher who postulates that the other one in a dialogue should not be taken as an object of our communicative efforts, thereby becoming an object of our violence. The other is not to be viewed even in terms of ontology but in terms of ethics - his/her position should resemble not that of an object but that of an ethic dative. It is not without interest that interviews resembling such an ideal type of dialogical encounter of two people are highly appreciated in Czech cultural context (Hoffmannova 1997), while in other countries, they are considered to be incompatible with the format of TV discussion or interviewing.

Media Dialogue as a Genre of Public Oral Discourse

115

5. Conclusion

When speaking about the genre of the interview, we should take into account that its norms and conventions are not to be reproduced mechanically. No rigid generic schema can account for the actual diversity of output. The genre also offers a challenge by provoking a free spirit that can transcend the limitations of previous examples, offering the possibility of choice (Fowler 1982). Discursive practices of genres are located at the point where the genre's requirements meet the social situation and a speaker's personality. Interviewing a person is a highly individualised practice. As one scholar said, there is no single hunting method. It depends whether you are hunting a lion or catching a butterfly. And it is the same with people.

Acknowledgements

The research was supported by the Grant Agency of the Czech Republic (405/96/K096 - Dialogue in the world of people and machines).

References

Bell, Α., P. Garrett (eds.) (1998): Approaches to Media Discourse. - Oxford: Blackwell Publishers. Bucher, H.-J. (1994): Dialoganalyse und Medienkommunikation. - In: G. Fritz, F. Hundsnurscher (eds.) Handbuch der Dialoganalyse, 471-493. Tübingen: Niemeyer. Clayman, S. E. (1991): News interview openings: Aspects of sequential organization. - In: P. Scanned (ed.) Broadcast Talk, 48-75. London: Sage Publications. Clayman, S. E. (1993): Reformulating the question: A device for answering/not answering questions in news interviews and press conferences. - Text 13, No. 2. Berlin/New York: Walter de Gruyter, 159188.

Cmejrkovä, S. (2000): Analysis of a TV polemical discourse. - Linguistica Pragensia X, 1,1-15. Dascal, M. (1989): Controversies as quasi-dialogues. - In: E. Weigand, F. Hundsnurscher (eds.), Dialoganalyse II, Band 1,147-159. Niemeyer: Tübingen. Dascal, M. (1998): Types of polemics and types of polemical moves. - In: S. Cmejrkovi, J. Hoffmannovä, Ο. Müllerovä, J. Svätlä (eds.) Dialoganalyse VI. (Beiträge zur Dialogforschung 16), 15-34. Tübingen: Niemeyer. Erickson, F. (1986): Listening and speaking. - In: D. Tannen, J. E. Alatis (eds.) Languages and Linguistics. Washington: Georgetown University Press. Fairclough, N. (1995): Media Discourse. - London: Edward Arnold. Fowler, Α. (1982): Kinds of Literature. - Oxford: Oxford University Press.

116

Svitla

Cmejrkovä

Greatbatch, D. (1998): Conversation analysis: Neutralism in British news interviews. - In: A. Bell, P. Garrett (eds.) Approaches to Media Discourse, 163-185. Oxford: Blackwell. Harris, S. (1991): Evasive action: How politicians respond to questions in political interviews. - In: P. Scannell (ed.) Broadcast Talk, 76-99. London: Sage Publication. Heritage, J. (1985): Analyzing news interviews: Aspects of the production of talk for an 'overhearing' audience. - In: T. A. van Dijk (ed.), Handbook of Discourse Analysis, vol. 3. London: Academic Press. Heritage, J., D. Greatbatch (1991): On the institutional character of institutional talk: The case of news interviews. - In: D. Boden, D. H. Zimmermann (eds.) Talk and Social Structure, 93-137. Cambridge: Polity Press. Heritage, J. (forthcoming): Basic ground rules: Taking turns and 'doing' news interview talk. - In: J. Heritage, S. d a y m a n , D. Greatbatch (forthcoming) The Political News Interview. London: Sage. Hoflmannovä, J. (1997): Linguistics, style, and dialogue stylistics. - In: E. Pietri (ed.) Dialoganalyse V. (Beiträge zur Dialogforschung 15), 257-262. Tübingen: Niemeyer. Levinas, E. (1961): Totalite et infini. Essai sur l'exteriorite. - Haag: Nijhoff. Scannell, P. (ed.) (1991): Broadcast Talk. - London: Sage Publications. Tolson, A. (1991): Televised chat and the synthetic personality. - In: P. Scannell (ed.) Broadcast Talk, 176-200. London: Sage Publication. Ueda, M. (1998): A conversational analysis of a debate. - In: Brown Slavic Contributions, Vol. 11, 142157. Brown University. Weigand, E. (1999): Dialogue in the grip of the media. - In: B. Naumann (ed.) Dialogue Analysis and the Mass Media (Beiträge zur Dialogforschung 20), 35-54. Tiibingen: Niemeyer. Weizman, Ε. (1998): Individual intentions and collective purpose. - In: S. Cmejrkovä, J. Hoffinannova, O. Müllerovä, J. Svfitli (eds.) Dialoganalyse VI. (Beiträge zur Dialogforschung 17), 281-290. Tübingen: Niemeyer.

Part Two Papers

Carla Bazzanella and Alberto Baracco

Misunderstanding in IRC (Internet Relay Chat)

1. Introduction

The aim of this paper is to analyze when misunderstanding occurs, and how it is managed in IRC, in comparison to FTFC (Face to Face Communication). The triggers of misunderstanding, the levels of misunderstanding, and the negotiation cycle - which have been proposed on the basis of a corpus of natural conversations between humans (cf. Bazzanella and Damiano 1999) - will be applied to IRC. The structural properties of IRC of being "the first form of remote, synchronous communication providing the many-to-many mode" (cf. Schulze 1999: 73), its 'amusement' feature, and the particular kind of personal deixis will be analyzed in relation to the occurrence and management of misunderstanding in two corpora.

2. Misunderstanding in everyday conversation

2.1. Definition and dynamic nature of misunderstanding Before we enter the subject of misunderstanding in IRC, an introduction about misunderstanding in everyday conversations will be of some use, so that we can both refer to some analytical tools which have already been proposed (see, e.g., levels, triggers, and the negotiation cycle of misunderstanding), and contrast the two different kinds of interaction which in a sense constrain different kinds of misunderstanding. Let us start with an example of misunderstanding quoted in West and Frankel (1991: 180), where the physician's query about a "history of cardiac arrest" is interpreted by the patient as related to the act of being arrested/having trouble with the police: (1)

Physician: Patient:

Have you ever had a history of cardiac arrest in your family? We never had no trouble with the police.

120

Carla Bazzanella and Alberto Baracco

We do not know anything more about the context, but it seems to us that it can be considered a clear case of misunderstanding where the non-coincidence between the speaker's meaning and the interlocutor's interpretation (which is one of the possible definitions of misunderstanding) arises on a semantic level, partly due to the physician's technical jargon. Unfortunately, we do not know how the conversational exchange develops either; strangely enough, many authors who deal with misunderstanding seem not to be interested in what follows the misunderstanding, that is, its interactional handling, or the negotiation cycle of misunderstanding. In general, the dynamic nature of misunderstanding has been stressed recently (cf., e.g., the special issue of Journal of Pragmatics edited by Marcelo Dascal, 1999: 31); in this perspective, in contrast to the traditional dichotomic view where understanding was opposed to non-understanding, misunderstanding, together with other linguistic and conversational resources (such as clarification sub-dialogues and repair structures) may help in "coming to understanding", an analytical key proposed by Weigand (1999: 769): "Dialogic action games are always interactive action games negotiating on an interactive level the general purpose of coming to understanding". In the complex process of understanding which takes place in everyday conversation between humans, and which is characterized by negotiation 1 , misunderstanding commonly occurs, and should be considered in a scalar manner: [...] in so far as misunderstanding is subject to 'negotiation' as part of an ongoing process of 'coming to an understanding' [...], it would seem rather to be viewed as a continuum. [...] The gradation view is further supported by the fact that, from a logical point of view, misunderstanding and understanding are contrary, rather than contradictory, i.e., the propositions Ά misunderstands B's utterance' and Ά understands B's utterance' may be both simultaneously true (regarding the same utterance), (cf. Dascal 1999: 756)

2.2. Triggers of misunderstanding In Bazzanella and Damiano (1999) we proposed labelling as triggers (rather than causes) the factors that can facilitate the occurrence of misunderstanding, in order to underline the non-deterministic role they play in predisposing communication to a negative outcome. For example, many scholars have stressed the relevance of ambiguity to misunderstanding,

1

The positive potential of human dialogue, even in difficult or conflictual interactions, lies precisely in negotiation; as Ringle and Bruce state (1982, 204): "The reason that dialogue is such an effective means of communication is not because the thoughts of the participants are in such perfect harmony, but rather because the lack of harmony can be discovered and addressed when it is necessary."

Misunderstanding

in IRC (Internet Relay Chat)

121

but, although it plays a major role in generating misunderstanding 2 , one should not concentrate on a single element as the only cause; ambiguity can be easily solved in the specific context, and in many cases is not responsible at all for misunderstanding. Both in human/human, and human/computer/human dialogues, the conflicts or mismatches which arise from the respective beliefs or knowledge also play a significant part. The role played both by the implicit and by the inferences which are to be drawn is essential in any form of interaction and can trigger misunderstanding, as has been underlined by several authors (cf., among others, Weigand 19993, and Dascal 19994). With regard to HCl, Joshi et al. (1984) state that: "The aim of the system should be to prevent the user from drawing false inferences from responses [...] the system should try to uncover the plans of the user in order to provide an optimally helpful response." cf. fragment (2), as an example of a 'blocking response' (with regard to the assumption that most associate professors are tenured): "If the system knows that Sam does not have tenure, and also knows that the user is unaware of this, then a simple answer of 'Yes' by the system could lead the user to assume falsely by default reasoning that Sam was tenured." {Ibid.) (2)

User. Is Sam an associate professor? System. Yes, but he doesn't have any tenure.

Triggers may be related not only to the responsibility of the interlocutor and of her/his interpretation choices (as has been underlined in the literature), but to all the components of the interaction: the structural components of communication (i.e.: disturbances along the communicative channel, similarities between elements of the linguistic code, problems caused by the use of a foreign language, lexical or syntactic ambiguities); triggers related to the speaker (i.e.: 'local' factors, such as the speaker's slips of the tongue, misconceptions, use of ambiguous forms, and 'global' factors concerning the structuring of information both on the pragmatic and on the syntactic level - e.g., politeness, indeterminacy, and anacolutha); triggers related to the interlocutor (i.e.: knowledge problems, such as false beliefs, lexical incompetence, gaps in encyclopedic knowledge, and cognitive processes, such as wrong inferences, cognitive load and its effects on the interlocutor's production);

2

3

4

Ambiguity, in its various forms, turns to be the major source of misunderstanding (cf. Zaefferer 1977, Blum-Kulka and Weizman 1988). "When relying on their knowledge of habits, interlocutors may rely on different cognitive means and therefore draw false inferences." Weigand (1999: 777). "The possibility of misunderstanding is ever present because not everything can be explicitly said, and the interlocutor must therefore rely on inferences based on fallible presumptions." Dascal (1999: 75).

Carla Bazzanella and Alberto Baracco

122

-

triggers related to the interaction between the participants (i.e.: non-shared knowledge, problems of topic organization, and focusing).

As Dascal (1999) rightly states: "[...] the various causes [or 'triggers'] of misunderstanding (and understanding) interact, either reinforcing or canceling each other." As in other linguistic and pragmatic phenomena, the best approach is a multidimensional one, which takes the different components of the ongoing interaction into account. In fact, if we consider the various possible triggers we can more effectively cope with different forms of interactions in their relation with misunderstanding. Every detailed analysis needs, in our opinion, a general and complex framework to be put in.

2.3. Levels of misunderstanding In Bazzanella and Damiano (1999) we proposed five levels at which misunderstanding can arise: phonetic, lexical, syntactic, semantic, and pragmatic. Other studies focused more on one or other of these levels, but it seems useful to have a complete taxonomy of the possible levels in order to analyze more clearly the misunderstanding which actually occurs, and, if one refers to a corpus in a quantitative manner (as is desirable), the relative incidence of the various levels should be calculated. Here, instead of the term phonetic level, we will adopt the term perceptive, in order to take into account the graphic features too, which, as we shall see further on, are 'exploited' in human/computer interaction. The lexical and the syntactic levels are often relevant to ambiguity (see 1.2); within the semantic level (which proved the most 'productive' with regard to misunderstanding in Bazzanella and Damiano 1999's corpus), the prepositional content and the reference expressions should be distinguished. At the pragmatic level at least six components have to be taken into account, given their possible relevance to misunderstanding: illocutionary force and indirect speech acts, nonliteral uses (e.g.: implicatures, irony, metaphor), relevance, topic, plans, prototype. Of course more than one level can be involved in a given misunderstanding (as we said when dealing with triggers, see 1.2), and the phases of negotiation can concern several levels at the same time; but usually, in a given interactional setting, one level will be more significant, and 'prevail' over the others, at least on the analytical level.

Misunderstanding in IRC (Internet Relay Chat)

123

2.4. Author, collocation, and negotiation cycle We will not discuss in depth here the topics relating to the author and the collocation of the repair: as for the author, note that the utterer of the turn which is the source of the misunderstanding is not necessarily the same person as the speaker who triggers the repair process. The most common collocation of repair is the so-called "third turn repair" (Schegloff 1992): the interlocutor's incorrect interpretation is immediately detected and corrected by the prior speaker, but the interlocutor her/himself may realize in subsequent turns that she/he has misunderstood and may correct her/himself. What has been labelled negotiation cycle refers to the handling of misunderstanding when it is detected5, either by the speaker of the source turn of misunderstanding itself, or by the interlocutor of that turn: After a misunderstanding has been recognized [...] the participant who has realized it, irrespective of her/his responsibility in the misunderstanding, has to choose between two possibilities: s/he can point it out and make a repair in order to understand or to make her/himself understood [...], or she/he can decide to disregard it [...]. In the latter case, which is more common in manmachine interaction [...], either a topic shift or a communication breakdown can follow. When the participant who has realized the occurrence of misunderstanding tries to make a repair, possibly by proposing a diagnosis and/or a solution, her/his partner can confirm or disconfirm it. [...] If the acceptance is just partial and there is no explicit refusal, either the speaker of the previous repair or her/his interlocutor can make a new repair, activating another negotiation cycle, (cf. Bazzanella and Damiano 1999: 823)

3. Some remarks on human/computer interaction: a collaborative negotiation

In order to analyze why misunderstanding occurs and how it is managed when humans and machines are involved, "We ought to consider the pragmatics of human-computer dialogue", that is, in Booth's words (1989: 47): "We should look, not only at the literal meaning of the symbols within the dialogue, but also what the user understands from the dialogue. What is communicated in computer dialogue may depend, not only upon the content of the dialogue, but also its style and structure. The context of any particular message is also important."

3

It is plausible to claim that absence of negotiation and complete breakdowns are more likely to occur in human/computer interaction, where the 'flexibility' of the system, though improved recently, is still limited in comparison to CMC and FTFC.

Carla Bazzanella and Alberto Baracco

124

This kind of "dialogue", if we consider it in a "prototype" framework (cf. Bazzanella 2002), can take various forms: HCl (Human-Computer Interaction), CMC (ComputerMediated Communication), and CSCW (Computer-Supported Cooperative Work). Some features of "cooperation" in different forms of human/computer interactions have been pointed out (cf. e.g. Dix et al. 1998), and the "flexibility" of the system has increased in recent research in Computer Science: we are thinking in particular of the development of "intelligent agents". Chu-Carroll and Carberry (1995) pointed out three distinctive features of

collaborative

negotiation between agents, that is, in collaborative planning activities in HCl: 1) A collaborative agent does not insist on winning an argument, and may change his beliefs if another agent presents convincing justification for an opposite belief. This differentiates collaborative negotiation from argumentation [...] 2) Agents involved in collaborative negotiation are open and honest with one another; they will not deliberately present false information to other agents, present information in such a way as to mislead the other agents, or strategically hold back information from other agents for later use. [...] 3) Collaborative agents are interested in others' beliefs in order to decide whether to revise their own beliefs so as to come to agreement.

In our present paper we will mainly limit ourselves to IRC, where misunderstanding and negotiation are parallel, in a sense, to human/human interaction, even if the dialogue in IRC is necessarily marked, as we shall see, by the general features of the computer as a medium and by its specific features.

4. Misunderstanding in IRC

4.1. C o r p o r a We have referred to two existing corpora of IRC for our analysis: -

The one (mainly) in English collected by Markus Schulze 21/07/1998, and available on the web: eäK äme®i'©än. ϊ §J>eäK äme®i©äfl. spook, u r spooking me with all these strange signs, what language r u writing? K4ime®i©äfl. american.

In general, "not all ill-formed input will lead to miscommunication. Misspellings and mispronunciations, for example, can often be processed without any disruption of the dialu

11

"[...] participants have evolved a number of innovative linguistic strategies which function to both compensate for and adapt to these constraints [= a complex set of temporal, spatial contextual and social constraints]." Werry (1996: 61) "IRC communication today is almost exclusively a leisure activity" Schulze (1999: 69).

Misunderstanding in IRC (Internet Relay Chat)

127

logue." (cf. Reilly 1991: 284) 12 . If we compare CMC in general and FTFC, the persistence of the message on the screen which we have in CMC should in principle reduce the possibility of perceptive misunderstandings during its processing, but in IRC the necessary rapidity (see feature 5.) makes the interaction more similar to FTFC. On the other hand, the actual technological limits (both hardware and software) strongly differentiate IRC from FTFC, so that we can positively state that it is 'easier' to speak than to send an IRC message. Why are the general features proposed above relevant to misunderstanding in IRC? The lack of a shared physical context (see feature I.) triggers several misunderstandings in IRC, both on the temporal and on the local level: the temporal collocations of the participants can vary greatly (as we see in fragment (4)), and local deixis is not necessarily known, thus inducing other users to draw a mistaken inference about the geographical localization of the interlocutor (cf. fragment (5)): (4)

T1 Virvart: T2 spook: spook: spook: T3 Virvart: T4 spook: T5 Virvart: T6 Virvart:

hallo! guten abend bonjour hello hmm.. a bit early .. isn't it? what is? vor eine guten abend.. I mean .. it practical afternoon here! and where in the world is spook!?

(5)

T1 T2 T3 T4

Hiya FrankNL. How's the Netherlands tonight then? Wull: I would not know about the netherlands.... Sorry Frank - I just assumed that NL stood for Netherlands :) it does, only I live in Scotland

Wull: FrankNL: Wull: FrankNL:

The lack of co-presence of the participants in the situation of utterance in IRC (see feature 1.) and the characteristic many-to-many mode (see feature 3.) sometimes makes it difficult to identify correctly who is speaking to whom. Therefore, the self-selection as an addressee13 can be wrongly inferred, or the greeting to a newly-arrived user may not be recognized, as in (6), where Lilstar -who was chatting mainly with Nedd, and presumably not glancing at others' turns (a sort of preferred 'coupling' takes place in IRC), cannot understand Nedd's T5. In this case we have a request for clarification:

13

The same is true for e-mail: "[..]e-mail writing is often characterized by a certain 'sloppiness' in the correctness of the spelling and the precision of typing which goes together, on the reader's side, with a very degree of tolerance for spelling and typing errors. What would be unacceptable in a regular typed letter, does not require correction in the e-mail format, increasing the informal nature of its style." Violi (1998: 268). The 'addressivity' is a particularly significant topic in IRC, as it was rightly suggested by Bays during Pragma99, held in Tel Aviv, Israel; cf. also Werry (1996: 52-53).

Carla Bazzanella and Alberto Baracco

128 T1 T2 T3 T4 T5 T6 T7

Mirage7: DaMo: DaMo: LilStar: nedd: LilStar: DaMo:

Hi DaMo! Hey mirage re all this has been a busy place always DaMoMmmmmmM DaMo? NEDD ! ! ! ! ! ! !

The 'invisibility' of the participants, which is related to feature /., makes it difficult to address and correctly identify participants, who usually resort to nicknames (see feature 4.), and only sometimes introduce themselves by an acronym "a/s/1" (age/sex/location). In the absence of these data, even their sex seems to become subject to negotiation, as in (7): T1 T2 T3 T4 T9 T13 T16 T20 T24 T25

CuteOne: spook: lippy: CuteOne: lippy: CuteOne: spook: Lippy: spook: CuteOne:

what r u talking about, today? anything at all cutey is he/her cute? i'm male I like males r u a male that likes males? or a female that likes males? lippy = male 20 I'm n o t " " " lippy is really not a male don't be angry, it is ok to be gay. don't be a shame.

The lack of extra-linguistic and prosodic cues (only partly replaced by transcription conventions, smileys, etc., as we mentioned before, in relation to the particular kind of written mode, see feature 2.), together with the rapidity of the 'scrolling' (see feature 5.), sometimes impedes a correct understanding of the pragmatic aspects (illocutionary force, implicatures, indirect speech acts, etc.), and in general the interpretation of an utterance, as in (8), where the intended meaning is sought (lilstar no kidding?), and proposed by the interlocutor, nedd (I'll take it as a compliment), LilStar (it was a T1 T2 T3 T4 T5 T6

LilStar: nedd: LilStar: nedd: LilStar: nedd:

and eventually agreed upon by the speaker

compliment). you're just too funny lilstar no kidding? if u say that u r funny too :) me funny? no. lilstar I'll take it as a compliment it was a compliment lilstar thnx:)

Furthermore, the topic organization, i.e. establishing and managing the topic (which we listed as a possible trigger related to the interaction between the participants,

see 1.2) is

complicated by the many-to-many mode (see feature 3.), and by the amusement function (see feature 6.) which hinder both a standard sequence of adjacency pairs and a clear de-

Misunderstanding in IRC (Internet Relay Chat)

129

velopment of dialogue (since often several discourses are intermingled, as in (3a) above, and beginnings and ends are not necessarily fixed and made explicit). The amusement function (see function 6.) also plays a role in the exploitation of the written mode (see feature 2.), together with the maximal reduction of graphemes, or 'abbreviation', which is prompted by the temporal constraints (see function 5.), as in fragment (9), where the number 6 (pronounce: set) is used instead of sei= you are, xo instead of pero=b\A, mo instead of Modena, chan instead of channel, ei instead of ehi. We could speak of a general 'linguistic contraction'14, where syntactical ellipsis is also widespread (e.g.: mo, tu? anni? which means: I'm in Modena, and where are you? How old are you?). The difficulty of coming to an understanding in (9) is also evident from the numerous requests for clarification, and from the misunderstanding about channel/city: (9)



< A somma A >

< A somma A >

ciao!? ci 6? si! dove, xo? mo, tu? anni? cosa vuol dire mo? modena ah, ma io pensavo ai chan, dove 6 tu? cosa sono i chan? ei ci 6 ancora?

Finally, what is more striking in IRC is the apparent desire to participate and to be connected in a phatic virtual community (see feature 6.) more than the intention of giving/getting information; hence the continuous checking for connection (ei ci 6 ancora?=are you still there?), made necessary also by the lack of physical contact, and by technical difficulties.

4.3. Provisional conclusions The data on misunderstandings in IRC that we found in the two corpora considered lead us to the following - provisional - conclusions: The occurrence of misunderstanding in IRC seems to be commonly located on the semantic and pragmatic levels, mainly due to the specific structural features and to the absence of a completely shared context; the collocation of repair in IRC is constrained by its specific structure, where 'intended' adjacency pairs are interspersed with others' turns (see fragment 3a);

14

"[...] messages in the extracts I examined averaged around six words in length." Werry (1996: 53).

Carla Bazzanella and Alberto Baracco

130

-

the negotiation cycle that we referred to above seems to be applicable in IRC too. Let us briefly analyze e.g. fragment (5) following the suggested lines:

Misunderstanding. Source (Tl) TlWull: Hiya FrankNL. How's the Netherlands tonight then? Detection (T2) T2FrankNL: Wull: I would not know about the netherlands.... Repair (T3) T3 Wull: Sorry Frank - 1 just assumed that NL stood for Netherlands :) Acceptance (T4) T4FrankNL: it does, only I live in Scotland To conclude our first step in the analysis of misunderstanding in IRC, we propose the following rough hypothesis - to be verified on the basis of significant quantitative data - : while the occurrence of misunderstanding in IRC seems to be more commonly located in relation to its specific features (see the above-suggested list), and the phatic and amusement functions induce IRC participants to neglect misunderstanding more often than in everyday conversation, the structural differences between IRC and FTTC seem not to influence the cycle of negotiation, which follows similar paths in the two forms of interaction.

References

Akman, V. and C. Bazzanella (forth.): The complexity of context. - In: V. Akman, C. Bazzanella (eds.) On Context. Journal ofPragmatics, special issue. Bazzanella, C. (1998): On context and dialogue. - In: S. Cmejrkova et al (eds.) Dialogue in the Heart of Europe, 407-416. Tübingen: Niemeyer. Bazzanella, C. (2002) (ed.): Sul dialogo. Contesti e forme di interazione verbale. - Milano: Guerini. - and R. Damiano (1999): The interactional handling of misunderstanding in everyday conversations. Journal of Pragmatics 31,817-836. Blum-Kulka S. and E. Weizman (1988): The inevitability of misunderstandings: Discourse ambiguities. - Text, 8 (3), 219-241. Booth, P. (1989): An Introduction to Human-Computer Interaction. - London: Lawrence Erlbaum Associates.

Misunderstanding in IRC (Internet Relay Chat)

131

Chu-Cairoll, J. and S. Carberry (1995): Conflict detection and resolution in collaborative planning. 1JCAI Workshop on agent theories, architectures, and dialogues. Dascal, M. (1999): Introduction: Some questions about misunderstanding. - Journal of Pragmatics 31,753-762. Dix, A. et al. (1998): Human-Computer Interaction. - Hemel Hempstead: Prentice Hall. Joshi Aravind, B., R. Webberand and M. Weischedel (1984): Living up to expectations: Computing expert responses. - In: Proceedings of the Fourth National Conference on Artificial Intelligence, 169-175. Naumann, Β. (1995): Mailbox chats: dialogues in electronic communication. - In: F. Hundsnurscher, E. Weigand (eds.) Future Perspectives of Dialogue Analysis, 162-184. Tübingen: Niemeyer. Pistolesi, Ε. (1997): II visibile parlare di IRC. - Quaderni del Dipartimento di linguistica dell'Universitä diFirenze 8, 213-246. Pistolesi, E. (2002): Flame e coinvolgimento in IRC. - C. Bazzanella, P. Kobau (eds.) Passioni, emozioni, affetti, 261-277. Milano: McGraw Hill. Reilly, R. G. (1987): Communication Failure in Dialogue and Discourse. - Amsterdam: North Holland. Ringle, Μ. Η. and Β. C. Bruce (1982): Conversation failure. - In: W. G. Lehnert, Μ. H. Ringle (eds.) Strategies for Natural Language Processing. Hillsdale, New Jersey: Lawrence Erlbaum Associate Publishers. Schegloff, E. A. (1992): Repair after next turn: The last structurally provided defense of intersubjectivity in conversation. American Journal of Sociology 97 (5), 1295-1345. Schulze, Μ. (1999): Substitution of paraverbal and nonverbal cues in the written medium of IRC. In: B. Naumann (ed.) Dialogue Analysis and the Mass Media, 65-82. Tübingen: Niemeyer. Violi, P. (1998): Electronic dialogue between orality and literacy: A semiotic approach. - In: S. Cmejrkovä et al (eds.) Dialogue in the Heart of Europe, 263-270. Tübingen: Niemeyer. Weigand, Ε. (1999): Misunderstanding: The standard case. - Journal of Pragmatics 31,763-785. Werry, C. C. (1996): Linguistic and interactional features of Internet Relay Chat. - In: S. Herring (ed.) Computer-Mediated Communication: Linguistic, social and cross-cultural perspectives, 4763. Amsterdam: Benjamins. West, C. and R. M. Frankel (1991): Miscommunication in medicine. - In: N. Coupland, H. Giles, J. M. Wiemann (eds.) Miscommunication and Problematic Talk, 166-194. Newbury Park, California: Sage Publications. Zaeflferer D. (1977): Understanding misunderstanding: A proposal for an explanation of reading choices. - Journal of Pragmatics, 329-346.

Massimo Α. Bonfantini, Augusto Ponzio, Susan Petrilli Trialogo sui dialoghi

- MASSIMO: Cominciamo col dire che il dialogo, nell'attuale forma sociale, e in pericolo. Questo pericolo deriva dalla smaterializzazione della comunicazione che la globalizzazione comporta. - SUSAN: D'accordo. All'idea di partire da questa considerazione arriveremo dialogicamente nel corso di questo dialogo stesso. Ma e giusto, come tu suggerisci, anticiparla qui per ragioni di dispositio. - MASSIMO: II dialogo che nella nostra tipologia abbiamo chiamato dialogo divertente e quello che abbiamo denominato dialogo di cooperazione, contrapponendoli a quello di ottenimento, si distinguono da quest'ultimo perche in essi si e disposti ο si e trascinati a mettersi molto in gioco. Sia pure in qualche maniera egotistica, H dove c'e il trarre piacere dal divertimento, ο in qualche maniera egotistico-altruistica, ü dove c'e il piacere della cooperazione fino a quello della creazione di un'opera insieme, non ci sono gli schermi e le difese del dialogo di ottenimento; e ciö comporta un rapporto di coinvolgimento corporeo con l'altro. - SUSAN: Potremmo dire che il corporeo e ciö che eccede dalla rappresentazione: la rappresentazione come messa in scena di raoli, che come tale comporta il rinvio proprio dell'alibi, e dunque il coinvolgimento limitato all'identitä che si recita, limitato al ruolo, dunque una responsabilitä con scappatoie, da cui l'alteritä del corpo e il coinvolgimento intercorporeo sono estromessi. - MASSIMO: Rispetto alle analisi che noi abbiamo fatto e che adesso qui riprendiamo e proprio il corporeo come legato alio spazio reale e alio spazio vissuto, alia Lebenswelt, ciö che emerge come caratteristico del dialogo. In effetti c'e un senso per cui anche il pur importante dialogo dell'arricchimento sentimentale della metatestualiti, il dialogo a distanza della lettera, ο persino quello della telefonata vecchio Stile, quindi molto preparata, mostravano, pur non essendo dialoghi faccia a faccia, un coinvolgimento corporeo. Ε tuttavia dei dubbi a questo proposito potrebbero essere espressi in quanto si tratta pur sempre di dialoghi a distanza. Immaginiamoci un po' se possono farci credere che una comunicazione cosi stereotipizzata, schermata, difesa come quella dei siti di conversazione di un mezzo che invece funziona, nei limiti in cui funziona, come strumento di lavoro ο di curiositä ο di catalogazione, possa surrogare il fascino del dialogo in presenza, in cui posso sentire la presenza dell'altro...

Massimo Α. Bonfantini, Augusto Ponzio, Susan Petrilli

134

- AUGUSTO: ...ed anche la sua assenza: "sei qui e gia mi manchi!", detto alia persona amata. - SUSAN: II dialogo significa intercorporeitä, coinvolgimento, responsabilitä senza alibi nei confront! dell'altro. Ciö a prescindere dalla effettiva presenza diretta del corpo. Dialogo come intercorporeitä nel senso di un collegamento con l'altro, di una esposizione all'altro, anche al di lä della presenza fisica del corpo. Certo, la presenza fisica puö contribuire a far sentire maggiorxnente il coinvolgimento e la responsabilitä e dunque a rafforzare Γ intercorporeitä dialogica, ma non e ciö che la costituisce nelle conversazioni via internet, a cui Massimo faceva riferimento. Ciö che qui viene a mancare non e owiamente l'intercoiporeitä fisica, ma Γ intercorporeitä dialogica. - AUGUSTO: A questo proposito, si potrebbe ricordare la distinzione evidenziata da Bachtin, nel suo libro del 1929, fra il dialogo di Dostoevskij e il dialogo di Piatone. Quest'ultimo e un dialogo di idee, di posizioni, invece quello in Dostoevskij e un dialogo di voci, intendendo per voce l'essere incarnato, Pessere posizionato... - MASSIMO: Chi presenta cosi i dialoghi di Piatone, compie una operazione riduttiva, ο per lo meno generalizza alcune caratteristiche di alcuni suoi dialoghi facendone la caratteristica generale. II dialogo di Piatone e molto piü corporeo di quanto non appaia in questa schematizzazione, soprattutto nei dialoghi a cui noi stessi abbiamo fatto riferimento nella nostra tipologia del dialogo. Si pensi al discorso di Alcibiade innamorato di Socrate nel

Simposio....

- SUSAN: Si, ma io vorrei tornare a sottolineare che la presenza fisica non e la condizione necessaria delF intercorporeitä dialogica: proprio tu Massimo dicevi, poco prima che iniziassimo questo dialogo, che anche nella conversazione con gli amici, dove c'e certamente il piacere dello stare insieme, se la conversazione verte su cose che non ci interessano, non vediamo Tora di tornare a casa, magari a leggerci un bei libro, realizzando cosi una situazione senz'altro piü dialogica e piü coinvolgente.. - MASSIMO: Certo, ma anche in questo caso c'e il dialogo con se stesso ο il dialogo metatestuale, perche il testo έ Ii con la sua materialitä e concretezza...Questa situazione di solitudine e in effetti una situazione che consente il piacere del testo ma anche il distanziamento critico nei suoi confronti. II "libero esame" prima di diventare la parola d'ordine del protestantesimo, era giä di fatto praticato nella solitudine della cella monastica, nel rapporto a tu per tu con la Bibbia. Questa solitudine e ben diversa da quella del navigatore in Internet, che non approda in nessun luogo che sia effettivamente da esplorare e che rischia veramente di non approdare da nessuna parte: navigare in senso proprio significa che ο vai in malora ο arrivi all'isola. - AUGUSTO: ...che non e ne l'isola del giomo prima, ne l'isola del giorno dopo..; - MASSIMO: Si, non e l'isola ideale, non e Pinformazione che non hai beccato, sicche vai a fondo ο sprofondi nelle sabbie mobili di questa awentura. C ' e dunque un corpo a corpo

Trialogo sui dialoghi

135

con se stesso, in questo dialogo, in questa organizzazione dialogica della riflessione in solitudine, di cui sussiste ormai tutta una serie di surrogati, fra cui quella dei videogiochi, anche se in essi c'e un po' di corpo a corpo con lo schema che sai che uno ha preparato, e puoi anche costruirti tu un percorso, e puoi costruire tu stesso un ipertesto. C'e la sostituzione artificiale del corpo a corpo, invece della costruzione dialogica, spontanea si, ma che risulta proprio in base ad un approfondimento della dialettica e della dialogicita, non come qualcosa di innato ma come abito. II quale puö essere un abito ο un soprabito leggero, oppure un kafetano, e potrei dire che talvolta i personaggi di Dostoevskij indossano kafetani..., - AUGUSTO: Propongo di seguire uno schema in questa nostra discussione rivolta un po' a fare il punto sulla nostra ricerca sul dialogo. Partiamo dalla domanda: che cosa abbiamo detto di specifico noi circa il dialogo? - MASSIMO: Ma perchd in questa nostra conversazione dovremmo fare la storia del giä detto? Certo, l'abbiamo voluta chiamare "Trialogo sui trialoghi" perche volevamo anche fare un po* il punto sul giä detto - ma per andare avanti. - AUGUSTO: Mi sono espresso male: che cosa diciamo noi, che cosa caratterizza la nostra posizione? Vorrei pero anche soffermarmi un momento sul termine "trialogo". Thomas Sebeok, quando Susan ed io, mentre era a Bari per tenere delle lezioni nel nostra dipartimento, abbiamo avuto occasione di parlargli, in una conversazione privata, di questo nostra "trialogo", fece notare giustamente che il termine e etimologicamente errato. Pietro Ispano, che non conosceva il greco, nel Tractatus ο Summule logicales, fa significare a dia di "dialettica" e di "dialogo" duo. Ma noi sappiamo bene che questa interpretazione e sbagliata. Si dovrebbe dire percio "dialogo a tre" e non "trialogo". Tuttavia a noi questo termine suona bene, anche perche ormai lo abbiamo usato parecchio. Dunque propongo di mantenerlo, e chissä che non entri in qualche edizione successiva del Dizionario dell 'uso del De Mauro. - MASSIMO: D'accordo su "trialogo" e d'accordo sul fatto di riprendere le cose gia dette, ma come cid che richiamiamo, a cui facciamo en passant riferimento. Le cose che stavamo dicendo adesso sono giä abbastanza diverse, stavamo introducendo degli aspetti nuovi nella nostra riflessione. Mi sembra che quest'oggi noi tendiamo molto al "corporeo". Forse a mano a mano che si invecchia...o dopo che uno e stato un po' squartato... - AUGUSTO: Certo, il rapporto tra dialogo e corpo...Io perö vorrei adesso proporre di seguire questo schema. Anch'io ho fatto un grafo...Faccio concorrenza a Massimo che ci ha oggi portato una copia del suo libra Breve corso di semiotica contenente ben 45 grafi. Ε uno schema a... "lisca di pesce". - MASSIMO: Ce l'ho, ce l'ho anch'io. Ho anch'io uno schema del genere nel mio libra. Ecco vedi (p. 141): "L'albero delle scelte secondo Poe".

136

Massimo Α. Bonfantini, Augusto Ponzio, Susan Petrilli

- AUGUSTO: Cominciamo a evidenziare il fatto che per noi il dialogo e gia presente a livello di io: e quando dico "noi" mi riferisco anche a tutta una serie di studiosi a cui ci richiamiamo, in primo luogo Peirce, Welby, Bachtin, Levinas... Giä l'io e dialogico. Questo e un punto fermo che ci distingue... - MASSIMO: Che distingue la nostra ditta dalle altre. Tu influenzato da Internet vuoi mettere subito la pubblicitä: La dialogicita dell'io! La ditta dei trialoganti! - AUGUSTO: Ε questa dialogicita e intesa non come dialogicita tra degli io che sono in un rapporto paritario: si tratta del dialogo tra io interpretato e io interpretante e dunque di un rapporto di alteritä, essendo Γ interpretante sempre piü ο meno diverso dall'interpretato... - MASSIMO: Quindi uno stesso io ha dentro il dialogo, la differenza, la dialettica... - AUGUSTO: Un secondo punto di questo schema a lisca di pesce e rappresentato quindi dalla logica, la logica come dia-logica. L'inferenza, il cui motore principale e l'abduzione, e dialogica. Ε questo e un altro punto di differenziazione che "abbiamo dato", volendo dire con questo verbo al passato che tipo di contributo abbiamo dato in questi...ormai quasi vent'anni di dialogo. - MASSIMO: Ma come si intitola questa cosa, questo schema: "il dialogo vent'anni dopo"? - AUGUSTO: La logica e dialogo: e siamo ancora nella testa... - MASSIMO: ...della lisca del pesce. -AUGUSTO: ...di uno solo, dell'io... - MASSIMO: Ε chi l'ha detto, puö darsi che quest'io rimanga fisso nell'induzione e non sappia andare oltre, e ci vuole l'altro che gli dica: parti da questa legge ο ipotizza questa regola come premessa..., insomma che lo introduca alle tre inferenze (vedi i miei grafi, nel libro citato, VIII. 1, VIII.2, VIII.3). - AUGUSTO: Certo! Infatti nel mio schema segue subito il rapporto io-tu, dove il dialogo non e riducibile ad accoglienza, rispetto, intesa, ma b soprattutto il trovarsi a dover rispondere, a dover prendere posizione, trovarsi in rapporto, non scelto, in cui non puoi essere indifferente, che έ una cosa molto diversa dal dialogo Ιο-Tu di Buber. - MASSIMO: Si potrebbe dire allora che mentre nella nostra tipologia del dialogo, dialogo di intrattenimento, di ottenimento, di.... - AUGUSTO: ...di ricerca, di cooperazione... - MASSIMO: Si va bene, ma avevo trovato tre nomi che finiscono tutti in "imento", intrattenimento, ottenimento...Ecco, si: intendimento. Avevamo sviluppato un po' un'analisi del senso pragmatico, da un lato, e della qualita di maggiore ο minore dialetticitä dei dialoghi, dall'altra. Invece adesso passiamo dalla logica, dall'abduzione, che e l'inferenza piü dialogica, in cui quindi puo essere utile un certo dialogo di cooperazione, al dialogo di scontro. Dialogo, che certamente uno puo avere anche in se stesso perche delle tre facce dell'a priori il desiderio e quella, come giä diceva Catullo, "odi et amo", in cui c'e una unita di

Trialogo sui dialoghi

137

opposizioni, si puö anche parlare di istinto di vita e di istinto di morte, di principio di piacere e di principio di realtä. Ma altra cosa e trovarsi al cospetto dell'altro, in un rapporto di confronto con Valtro, che anche quando e oggetto di desiderio lo e in un rapporto piü ο meno conflittuale, e in questo senso dialogico. - SUSAN: Oggi έ molto diffusa l'idea che la comunicazione con l'altro, soprattutto sul piano sentimentale, si basi sull'avere qualcosa in comune, gusti e interessi in comune, affinita, sino a considerare realizzato l'incontro con l'altro nella fusione. Ma anche le comunitä vengono concepite in termini di comunione, di unitä, di identita. In effetti, senza alteritä, senza estraneitä, senza assenza in presenza, non c't comunicazione perche non c'e dialogo. La comunicazione e anche scontro, conflitto, opposizione, e il dialogo έ coinvolgimento con l'altro non sulla base di un rispetto deciso nei suoi confronti, ma spesso a dispetto di ogni tentativo di ignorarlo e di estrometterlo. Le persone generalmente credono di doversi incontrare sulla base di un terzo termine, esterno ad esse che le accomuna, che le rende simili, in cui ritrovano una comune identita. Questa ricerca del terzo, l'etnia, la religione, la lingua, il territorio, e oggi molto diffusa e anche, come gli awenimenti della storia odiema attestano, particolarmente nefasta. Nei rapporti d'amore, l'"odio e amo" di Catullo e soppiantato dall'idea dell'amore come intesa totale, senza scontro, senza sospetto, senza ricerca, senza distanza, senza assenza, senza alteritä. Levinas fa osservare nei suo scritto "L'altro in Proust", riferendosi ai personaggi della Recherche, che Marcel non ha amato Albertine, se l'amore e fusione con l'altro, estasi di un essere di fronte alia perfezione dell'altro, assenza di conflitti, sicurezza nella presenza dell'altro, pace del possesso. Ma questo non-amore e proprio l'amore; la solitudine e comunicazione; la lotta con l'inafferrabile e movimento verso l'altro; l'assenza dell'altro e la sua presenza di altro. - AUGUSTO: Tomiamo al mio schema. Poi si passa in esso dall'io-tu all'io-voi dove entra in gioco l'argomentazione, il convincimento, la persuasione, il discorso pubblicitario, propagandistico, il comizio politico, i vari generi di discorso dialogico in cui c'e un voi, un uditorio collettivo a cui ci si rivolge ο ci si appella. Qui entra in gioco, quando si considerano gli oggetti coinvolti in questo rapporto, il mercato, lo scambio, il rapporto tra cose e il rapporto di cose tra persone. Un ruolo importante svolge in questo scambio dialogico la menzogna. - MASSIMO: Insomma tu vuoi ripercorrere proprio tutti gli argomenti di cui ci siamo occupati nei nostri dialoghi, insomma una Enciclopedia del dialogo in compendio. - AUGUSTO: Ο una Fenomenologia dello Spirito... del dialogo. - SUSAN: ...del Corpo del dialogo. - AUGUSTO: Poi si passa, nello schema, al dialogo all'interno della comunitä e dunque qui rientra tutto cio che abbiamo detto e che stavamo dicendo anche adesso a proposito dei mass media, ma anche delVideologia sociale. Sicche si passa all'aspetto politico del dialogo. Ecco, quindi, questa specie di attraversamento ci permetterebbe di mostrare alcuni

138

Massimo Α. Bonfantini, Augusto Ponzio, Susan Petrilli

caratteri specifici delle cose che noi andiamo sostenendo a proposito del dialogo: cioe che l'io e dialogico, che il rapporto io-tu non e un rapporto tra due soggetti separat! che si incontrano, si salutano e si dicono: "prego, cominci lei", "se permette, comincio io", ma che invece si trovano giä in un rapporto di coinvolgimento tale che preferirebbero non parlarsi piuttosto che parlarsi: parlano insomma per costrizione. In considerazione di ci6 risultano un po' campate in aria e forse anche un po' ridicole le analisi del dialogo dei "conversazionisti"... - MASSIMO:... anche se ci sono alcune cose nelle analisi di questo tipo che in un manuale di analisi del dialogo vanno riprese e possono essere utili. - AUGUSTO:... Ε poi segue il tema dell'argomentazione con tutti i nessi e connessi fra cui il tema della menzogna, come dicevo. Si giunge quindi all'aspetto "comunitario" del dialogo, ivi compresa la comunicazione massmediale, con le considerazioni critiche concernenti internet, la televisione, con le proposte che abbiamo fatto e che possiamo riprendere, quella di Massimo del video-saggio, ecc. Tutto questo e collegato con la critica dell'ideologia e con la riproposizione del dialogo nella politica; che non e quello delle diverse nuove versioni televisive, sul video, e delle riedizioni televisizzate, spettacolarizzate, nella realtä, di "tribuna politica"; ma e invece un convogliamento diverso del politico che sia superamento delle alternative - il dialogo come sbandieramento di varie alternative

-

che fanno parte della stessa realtä, che sono funzionali alia riproduzione dell' attuale forma di comunicazione-produzione, che fanno parte della stessa minestra ο della stessa brodaglia e che non presentano nessuna alteritä. (La differenza, proprio in questi termini, fra alternativa e alteritä si trova chiaramente enunciata, denunziata, in Pasolini). Si pensi alla sequenza di facce sui manifesti durante la campagna elettorale, tante alternative... - MASSIMO: ...senza mai un culo, come potrebbe dire Toto. - AUGUSTO: Ma anche questa non sarebbe che un'alternativa, e per altro giä compresa nella sequenza... Senza un volto, direi, senza alteritä. - MASSIMO: Bene hai fatto la scaletta. Adesso, come direbbe Wittgenstein, gettiamo la scaletta. Sperando di reggerci da qualche parte, ο almeno di restare appesi, come Wittgenstein... - AUGUSTO: ...che per altro disegnava edifici senza cornicioni. - MASSIMO: Mi pare troppo denso il tuo schema. Quanti punti sono? - AUGUSTO: Beh, il mio schema comprende: l'io (che prima di mettersi in dialogo e giä dialogico), l'io-tu, il voi, la comunitä... - MASSIMO: Lui. - SUSAN: Lui, il grande fratello. Oppure esso ο ess α: il territorio, la storia comune, l'origine, la tradizione, la lingua ο la famiglia linguistica, la cultura, il lavoro. - AUGUSTO: la comunitä dicevo, e quindi l'ideologia (soprattutto quella mass-mediale) e la politica.

Trialogo sui dialoghi

139

- MASSIMO: Ti sei dimenticato il corpo! II corpo! - AUGUSTO: II corpo attraversa tutto. Perciö questo percorso si chiama "Fenomenologia del corpo del dialogo"... - SUSAN: ...del dialogo incorporate, incarnato. - MASSIMO: Ma non e meglio partire dall'attualita? Potremmo mantenere questo schema ma incorniciandolo, il che sarebbe molto hegeliano, nel presente, nella situazione attuale del dialogo, che e quella della comunicazione globale. La caratteristica principale del dialogo nella comunicazione globale e quella della sua smaterializzazione. Da una parte la corporeitä del dialogo, a cui abbiamo fatto riferimento sopra, dall'altra il culto della smaterializzazione della attuale forma sociale... - AUGUSTO: ...della comunicazione globale fiinzionale alia riproduzione di questa forma di produzione basata sulla comunicazione e che quindi non comunica nient'altro che se stessa. - MASSIMO: "Culto della smaterializzazione". Noi potremmo partire dalla smaterializzione internet del dialogo versus la corporeitä del dialogo in presenza. - AUGUSTO: Va bene. II che significa che partendo cosi, il punto di awio dello schema che ho proposto si contestualizza; nel senso che viene a stabilirsi una contrapposizione fra un io che, come dice Peirce, ο Bachtin, ο Levinas, έ giä in rapporto con l'altro, senza che abbia bisogno di aprire la porta e di rivolgersi al vicino di casa, ma anche restandosene chiuso in casa, in solitudine - come risulta evidente nei personaggi delle Memorie del sottosuolo di Dostoevskij, un io solitario e meditativo, ο rimuginativo, e con cio stesso dialogico - e dall'altra un io aperto alla comunicazione globale, collegato in internet e che nello stesso tempo si dedialogizza, per il carattere monologico, unidimensionale, smaterializzato del sistema di comunicazione cui aderisce. - MASSIMO: Appunto. Tu potresti sviluppare queste cose, mentre... - SUSAN: ...mentre a me interessa riprendere quelle considerazioni che facevo prima sul carattere materiale, corporeo, anche del dialogo non in presenza, quando in esso si senta la resistenza dell'altro, la sua assenza e la sua distanza, non semplicemente fisiche; si riveli l'alteritä dei dialoganti, la loro refrattarieta alla totalizzazione, all'accomunamento, la loro estraneita, emerga la reciproca irriducibilitä tra interpretato e interpretante, anche nella lettura solitaria di un testo scritto. - AUGUSTO: C'e una distanza che impedisce la comunicazione, ma e anche la vicinanza eccessiva a impedirla. La comunicazione in presenza non e un fatto di awicinamento. Nella comunicazione globale c'e un awicinamento eccessivo, una aderenza addirittura, aderenza che e anche adesione alia sua logica. Una comunicazione in presenza, quella in cui sono consapevolmente e criticamente coinvolto e pienamente responsabile al di lä dei ruoli e dei "giochi comunicativi", richiede perciö una presa di distanza, come condizione dell'alterita e del dialogo.

140

Massimo Α. Bonfantini. Augusto Ponzio, Susan Petrilli

- SUSAN: La situazione di intercorporeita - l'essere collegati non solo all'interno della sfera dell'antroposemiosi ma nell'intera biosemiosi con ogni altro essere vivente a livello planetario come la biosemiotica ci mostra - e qualcosa di ben diverso dal collegamento fornito dai computer e dai telefonini in cui la vicinanza con l'altro e il collegamento realizzano una comunicazione apparente, ο meglio funzionale alia riproduzione di se stessa, alla comunicazione della comunicazione stessa, mentre gli individui restano sempre piü isolati, separati e incapaci di dialogare, ma anche per nulla collegati con il resto del mondo vivente. - AUGUSTO: Io vorrei insistere sul fatto che la comunicazione globale ha una funzione dedialogizzante, se per dialogo si intende il rapporto con l'alteritä, per il fatto che ci pone nella situazione di identici, ci omologa... - MASSIMO: Appunto. Perciö dicevo che dobbiamo partire dall'attuale. Noi dobbiamo cominciare col dire che il dialogo e in serio pericolo. Ε questo un aspetto piuttosto inquietante che dipende dalla contrazione in generale degli spazi della Lebenswelt, per esempio dalla attuale penuria ο assenza delle piazze, perche le piazze sono poco vivibili, perche sono quasi sempre invase dalle automobili, e quindi scarseggiano ο non ci sono proprio i mercatini, fra le altre cose. I bambini non vi possono giocare a pallone, come invece avviene a Trieste, in piazza Unitä d'Italia - forse Tunica piazza di citta dove i ragazzini giocano a pallone - a forma di quadrilatero di cui il quarto lato da sul mare, perche non vi sono macchine, vi sono pochi caffe e solo due fontane, piccole. Non ci sono le piazze e la gente si consola con le finte piazze di Internet dove puoi "ciattare" con (mistero, enigma!) chi sa chi. - SUSAN: Artefatto mistero, artefatto enigma: perche nel dialogo effettivo, corporeo, quello pericoloso, c'e il segreto dell'alterita; qui invece, in questo "ciattare", c'e un mascheramento non dissimile dalle forme decadute del camevalesco, c'e l'incognita di finte maschere, rispetto alle quali tuttavia per nessuno sarebbe difficile poter dire, come Pinocchio nei confronti del Gatto e la Volpe, quando da burattino di legno si a w i a ormai ad essere corpo, persona in carne ed ossa: "Vi conosco mascherine!" - MASSIMO: Bene. Ma adesso torniamo alia mia proposta che riguarda la dispositio di questo nostra dialogo. Noi abbiamo fatto Yinventio, ma adesso mi pare che siamo d'accordo che, per quanto riguarda la dispositio, dobbiamo cominciare dalla situazione attuale della smaterializzazione del dialogo. Ε non e casuale, che quando abbiamo cominciato a conversare, abbiamo iniziato proprio dalla riflessione su tale situazione, che e quella evidentemente che ci interessa di piü e direttamente, cosi come puo costituire l'interesse di partenza di chi ascolta ο legge il nostra dialogo. Pur investendo soltanto una minoranza, soprattutto la comunicazione tramite internet evidenzia bene il pericolo di smaterializzazione del dialogo e di de-dialogizzazione della comunicazione globale. C'e rispetto a ciö un diffuso atteggiamento di accettazione ο di rassegnazione, non solo da

Trialogo sui dialoghi

141

parte dei filosofi e dei semiotici e dei sociologi della comunicazione, ma anche da parte della gente in generale, come di fronte a qualcosa di ineluttabile. Anche Fultimo Rifkin, che generalmente e molto critico nei confronti dell'attuale forma sociale, considera ineluttabile il fatto della compra-vendita e della omologazione delle "esperienze di vita". Adesso, egli dice non si scambiano piü le propriety e i beni, ma "pacchetti" di esperienze di vita. Se e cosi il dialogo va a rotoli, perche esso fanziona all'insegna dell'alterita e del cambiamento e non quando la comunicazione non ha altro scopo che quello di comprare da un altro un pacchetto confezionato. - AUGUSTO: Alia luce di questa riorganizzazione del nostro dialogo possiamo ritornare ai temi della mia scaletta privilegiandone alcuni, lasciando altri semplicemente menzionati a titolo rimemorativo per quanto riguarda i nostri dialoghi precedent!, e soprattutto rimpostandone altri in connessione a questa problematica della corporeita dialogica versus la smaterializzazione. Per esempio, sotto questo aspetto tomerei alia questione della dialogicitä strutturale dell'io. La struttura egologica e dialogica. II dialogo non sta in mezzo ad un emittente e ricevente ma e costitutivo delPemittente e del ricevente. La loro materialita semiotica e data dalla loro costitutiva dialogicitä. II se, che non coincide con l'io e che fa da interpretato nel lavoro di interpretazione dell'io, quando l'io sente, percepisce, prova stati d'animo, esprime volizioni, bisogni, desideri, e la corporeitaι fondamentalmente. Ε la corporeita del se, delΓinteφretato, nel processo di presa di coscienza, a impedire la coincidenza fra questo se e l'io che lo interpreta e dunque a determinare un rapporto effettivamente dialogico fra di essi, un dialogo inesauribile, che costituisce l'aspetto dinamico della semiosi in cui l'io consiste. - SUSAN: Questo distanziamento, questa fermata del segno, questa possibilita di metasemiosi, per il quale l'io sussiste, e la caratteristica della semiosi umana, che dunque si presenta come semiotica. L'uomo e un animale semiotico. Se per tutto il mondo vivente, per l'intera biosemiosfera, vita e semiosi coincidono, nell'ambito della sfera dell'antroposemiosi si dä la possibilita di un distanziamento dialogico tra i segni dell'io che prende coscienza di se e quelli di questo se, fatti oggetti, interpretati dai primi secondo un rapporto di alteritä che rende Γ interpretazione sempre rivedibile, rischiosa e aperta. La capacitä semiotica che caratterizza l'essere umano, cio£ la sua capacitä di metasemiosi, non e altro che la dialogicitä. - AUGUSTO: Certo, e ciö comporta che ci sia un distanziamento tra una identita e una alteritä, che possono essere ο l'identitä e l'alteritä di uno stesso soggetto, ο quella di due ο piü soggetti, ma in ogni caso entra in gioco il distanziamento dialogico che non puo essere spiegato se non sulla base dell'alterita, cioe sulla base corporea, materiale, del rapporto fra interpretato e interpretante. - SUSAN: Io tengo a sottolineare ulteriormente la dialogicitä quale caratteristica strutturale della condizione umana. Perche siamo segni, come direbbe Peirce. L'analisi del soggetto

142

Massimo Α. Bonfantini, Augusto Ponzio, Susan Petrilli

come sdoppiato in segni interpretanti e in segni interpretati dialogicamente interagenti e condotta in diversi inediti di Victoria Welby dedicati al problema dell'io. Welby distingue tra Ident (un suo neologismo per dire l'io nella sua identitä) e Self. L'io nella sua identitä, come Ident e fatto di rapporti con i suoi numerosi Selves. Al posto di un io compatto, integro, unitario, monologico vi e un io aperto e in costruzione che vive nel dialogo fra una molteplicita di espressioni diverse del soggetto che non si riduce mai a Ident. Ciö indica non solo il carattere dialogico dell'io, ma anche il carattere materiale, fatto cioe di alteritä, di questo dialogo. A partire da ciö si puö approfondire il rapporto tra dialogo e corpo, stabilendo addirittura un rapporto di sinonimia tra dialogicitä e intercorporeitä. - AUGUSTO: Soprattutto Bachtin tematizza questa connessione, la cui presa di coscienza non e semplicemente un fatto conoscitivo, ma un fatto etico, perche e la presa di coscienza della propria responsabilitä senza alibi, senza paletti e confini tracciati intorno alia propria identitä in modo di circoscrivere la responsabilitä al solo rispondere di se e non di altri. - SUSAN: Questa presa di coscienza della dialogicitä come intercorporeitä ha un carattere etico perche vuol dire riconoscerla come condizione umana di coinvolgimento, di implicazione di ciascuno con il proprio prossimo, non solo il prossimo vicino, come dice Welby, ma anche il prossimo lontano, su scala planetaria. - MASSIMO: Sono molto d'accordo con questa sottolineatura degli aspetti di etosemiotica del dialogo. Ε necessario cogliere la dimensione storica della dialogicitä, collegata con la storia della semiosi, cioe far vedere come la dialogicitä cresce nel vivente trovando nell'uomo la sua manifestazione come capacitä di presa di coscienza e, inseparabilmente da essa, di responsabilitä. Da questo punto di vista, andrebbe ripreso quell'aspetto di Hegel, che potremmo anche dire vichiano ο che oggi chiameremmo etologistico, malgrado l'impostazione criticabile per il carattere chiuso e prowidenzialistico che Hegel gli dä, che consiste, tradotto nei termini nostri, nel vedere giä in nuce in tutta la storia del vivente, la capacitä, poi sviluppata a grado elevato nell'uomo, di fissare la semiosi per distanziarsene sul piano della metasemiosi. Questo sviluppo della dialogicitä andrebbe studiato anche all'interno della stessa storia umana come sviluppo della metasemiosi e quindi della presa di coscienza. Oggi, grazie alio sviluppo tecnologico che permette una sempre maggiore possibilitä di riflessione metatestuale,

c'e un'oggettiva possibilitä di maggiore metase-

miosi, di maggiore distanziamento, dunque di maggiore dialogicitä, riflessione e discussione collettiva. Oggi tutto diventa testo, nel senso che abbiamo una grande capacitä di registrazione, che e certamente un potenziamento delle possibilitä dialogiche e critiche. Puoi analizzare un film e anche un quadro, ma anche una partita di calcio, giocando con vari aspetti, forme e mezzi di registrazione e di riproduzione di testi, ingrandendo, fermando sequenze, ecc. Da un altro lato c'e perö la riduzione a scambio commerciale di qualsiasi aspetto, forma ο contenuto della comunicazione, a cui lo scambio commerciale richiede di

Trialogo sui dialoghi

143

omologarsi, a standardizzarsi, con il conseguente omologarsi e standardizzarsi degli stessi rapporti umani. - SUSAN: La nuova tecnologia certamente offre dei grandi vantaggi, offre la possibilitä del metadiscorso, del metatestuale, e quindi di sviluppare un atteggiamento dialogizzato e dunque critico nei confronti dei comportamenti, dei prodotti, dei rapporti, delle persone, ecc.; ma al tempo stesso, nella comunicazione globale, stanno venendo meno le condizioni per utilizzare questo aspetto positivo, questo vantaggio offerto dallo sviluppo tecnologico. C'e il dominio della logica del mercato la cui tendenza e il monologismo e non certo il dialogo e il plurilogismo. - MASSIMO: A me non interessa tanto l'effettiva standardizzazione che il mercato comporta, quanto il sentimento della positivita e della ineluttabilitä della standardizzazione che e abbastanza diffliso. Non e tanto il fatto in se della standardizzazione che preoccupa, quanto il suo aspetto dialogico, cosa di cui ci occupiamo, che, consiste, a quanto pare, nel fatto che la gente ci provi un gusto perverso e un senso di rassegnazione, di adesione passive a questa logica, che comporta il trasferimento della taylorizzazione dalla produzione al consumo. Ecco perche oggi il dialogo e in pericolo!

Bibliografia

Bachtin, Μ. M. (1997): Problemi dell'opera di Dostoevskij (1929). - Bari: Edizioni del Sud. Bachtin, Μ. Μ. (1998): Per una filosofia dell'azione responsabile. - Lecce: Manni. Bonfantini, M. A. (1984): Semiotica ai media. - Ban: Adriatica. Bonfantini, M. A. (1987): La semiosi e l'abduzione. - Milano: Bompiani. Bonfantini, M. A. (2000): Breve corso di semiotica. - Napoli: Edizioni Scientifiche Italiane. Bonfantini, Μ. Α., C. Caputo, S. Petrilli, A. Ponzio, A. T. Sebeok (1998): Basi. Significare, inventare, dialogare. - Lecce: Manni. Bonfantini, Μ. Α., J. Bernard,, J. Kelemen, A. Ponzio (1994) (a cura di): Reading su Ferruccio Rossi-Landi. Semiosi come pratica sociale. - Napoli: Edizioni Scientifiche Italiane. Bonfantini, Μ. Α., A. Ponzio (1987): Dialogo sui dialoghi. - Ravenna: Longo. Bonfantini, Μ. Α., A. Ponzio (1996): I tre dialoghi della menzogna e della veritä, con la partecipazione e un contribute di S. Petrilli. - Napoli: Edizioni Scientifiche Italiane. Bonfantini, Μ. Α., S. Petrilli, A. Ponzio (2000): Three dialogues on rhetoric, argumentation, and

new media. - Semiotica. Journal of the International Association for Semiotic Studies, 128, 1/2, 69-112. Cascone, G. (1999): Riscrittura I. - Roma: Derive Approdi. Ciccarelli, R. (1999) (a cura di): Inoperositä della politica. - Roma: Derive Approdi. Deleuze, G., F. Guattari (1980): Come farsi un corpo senza organi? Millepiani. Capitalismo e schizofrenia, sez. II, 1996. - Roma: Castelvecchi. Ispano, P. (1986): Tractatus. Summule Logicale. Trad, dal latino di A. Ponzio. - Bari: Adriatica. Levinas, E. (1964): II dialogo in Proust. - In: E. Levinas, Nomi propri. Genova: Marietti.

144

Massimo Α. Bonfantini, Augusto Ponzio, Susan Petrilli

Ldvinas, E. (1999): Filosofia del linguaggio. - Bari: Graphis. Morris, C. (1998): Lineamenti di una teoria dei segni. (1938), a cura di F. Rossi-Landi (1954), nuova ed. a cura di S. Petrilli. - Lecce: Manni. Pasolini, P. (1976) Lettere luterane. - Torino: Einaudi. Pasolini, P. (1990): Scritti corsari. - Milano: Garzanti. Peirce, C. S. (1923 ): Chance. Love and Logic. - New York: Harcourt. Trad. it. (1956) Caso, amore e logica. Torino: Taylor. Peirce, C. S. (1980): Semiotica. I fondamenti della semiotica cognitiva. - Torino: Einaudi. Peirce, C. S. (1984): Le leggi dell'ipotesi. - Milano: Bompiani. Petrilli, S. (1998a): Su Victoria Welby. Signifies e filosofia del linguaggio. - Napoli: Edizioni Scientifiche Italiane. Petrilli, S. (1998b): Teoria dei segni e del linguaggio. - Bari: Graphis. Petrilli, S., A. Ponzio (1999): Fuori campo. I segni del corpo tra rappresentazione ed eccedenza. Milano: Mimesis. Ponzio, A. (1997): Elogio dell'infunzionale. Critica dell'ideologia della produttivitä. - Roma: Castelvecchi. Ponzio, A. (1999): La comunicazione. - Bari: Graphis. Ponzio, A. e S. Petrilli (2000): II sentire della comunicazione globale. - Roma: Meltemi. Rifkin, J. (1995): La fine del lavoro. II declino della forza lavoro globale e Pawento del post-mercato. - Milano: Baldini & Castoldi. Rifkin, J. (1998): II secolo biotech. - Milano: Baldini & Castaldi. Rossi-Landi, F. (1994): Semiotica e ideologia (1972, 1979). - Milano: Bompiani. Rossi-Landi, F. (1998): Significato, comunicazione e parlare comune (1961, 1980). - Venezia: Marsilio. Schaff, A. (1994): Umanesino ecumenico. - Bari, Adriatica. - (1995): II mio ventesimo secolo. - Bari: Adriatica. Sebeok, Τ. A. (1998): The Sign is Just a Sign. La semiotica globale. - Milano: Spirali. Sebeok, T. A. e J. Umiker-Sebeok (1992): Biosemiotics. - In: The Semiotic Web 1991. Berlin; Mouton de Gruyter. Solimini, M. et alii (1995): L'estraneitä che accomuna. - Bari: Edizioni del Sud. Volosinov, V. N. (1980): II linguaggio come pratica sociale, scritti 1926-30. - Bari: Dedalo. Volosinov, V. N. (1999): Marxismo e filosofia del linguaggio (1929). Trad, di M. De Michiel. Lecce: Manni. Welby, V. (1985) Significato, metafora interpretazione. - Bari: Adriatica.

Regine Borderie

Dialogues de groupe: Balzac, Zola

1. Introduction

La notion de "polylogue", utilisee par les linguistes, permet de distinguer le dialogue de groupe (a quatre, cinq, six personnages et plus) du dialogue a deux ou a trois personnages, et done attire l'attention sur lui. On peut penser, en effet, qu'il presente des specificites, en particulier dans sa version litteraire. Je les aborderai ä partir de La Peau de chagrin (1831), du Pere Goriot (1835), et de Pot-Bouille (1882),' ä la suite d'un travail deja effectu6 sur la trilogie balzacienne du Pere Goriot, des Illusions perdues, et de Splendeurs et miseres des courtisanes (Borderie 1999: 55-76). Les interets du dialogue de groupe dans ces ceuvres me paraissent importants sur le plan litteraire, et, dans l'ordre de la representation, sur le plan psychologique, et sociopolitique. Sur le plan litteraire se posent des questions stylistiques (comment exprimer la vie du langage partagö en groupe? comment signifier la pluralite des interlocuteurs, leurs interventions? comment se construit le dialogue? comment progresse-t-il? y a-t-il des formes speeifiques de dialogue de groupe?). Se pose egalement la question des formes litt6raires et de leur genealogie: l'ecriture du dialogue de groupe s'inspire-t-elle de quelques modeles, litteraires ou non litteraires? Sur le plan psychologique (dans l'ordre de la representation toujours), le "polylogue" propose une vision du comportement verbal des hommes en soci6te; sur ce point je m'inspire du travail de Freiich (1991), Au parloir du roman, qui elle-meme renvoie k Freud (Psychologie collective et analyse du Moi 1921, 1948: 76-162) celui-ci s'etant inspire des recherches de Gustave Le Bon sur les comportements humains dans Psychologie des faules (1911). Sur le plan socio-politique, enfin, ce que sont socialement les personnages rassembl6s pour parier ensemble, et ce qu'ils disent, est eclairant quant ä la vision de la societe que vehicule le roman dans ces parties dialoguees. II est remarquable, notamment, que les romanciers franpais du XlXeme reunissent des personnages d'extractions sociales diverses pour parier d'actualite politique.

Pour les ceuvres de Balzac, j'adopte les abreviations suivantes: PG pour Le Pere Goriot, PCh. Pour La Peau de chagrin.

146

Regine Borderie

Dans le cadre de cet article, je partirai de la question des modeles, comme je l'avais fait pour l'etude de la trilogie balzacienne. On verra que s'operent, dans les exemples choisis, un dereglement et une degradation, autant dire une parodie, du modele mondain et du modele philosophique de la conversation de groupe. Du modele philosophique, le Banquet de Piaton pourrait etre la reference; la discussion, qui n'exclut pas le rire, s'y distingue en particulier par le serieux de ses enjeux, par la qualite de ses exigences intellectuelles, chacun exprimant ä son tour son point de vue. Le modele du banquet philosophique n'est pas oublie dans la tradition fran9aise; il est encore illustre en particulier ä la Renaissance. Le modele mondain, quant a lui, s'est forme dans la vie sociale elle-meme, et on en a une representation dans la litterature romanesque des siecles qui precedent le XDteme, dans le roman pastoral, et dans le roman heroi'que: frequemment les personnages sont rassembles pour parier et debattre d'un sujet. Et dans la tradition des recueils de nouvelles, les recits sont encadres par la mise en scene de personnages civilement reunis pour deviser, se raconter des histoires (il en est ainsi dans L'Heptamiron de Marguerite de Navarre, inspire du Decameron). En dehors de la litterature romanesque, contribuent ä formaliser la conversation de groupe des traites, dont un recueil donne l'idee pour les XVIIeme et XVIIIeme siecles, en France, recueil etabli par Hellegouarc'h (1997). De ces textes ressortent l'elaboration de lois de la politesse, la mise en valeur de la bienseance, du respect de l'autre, de ses idees, de son tour de parole. Le modfcle philosophique et le modele mondain peuvent etre consideres comme proposant une vision elevee, ambitieuse, de la societe et du comportement en groupe, ils impliquent la croyance en une fonction civilisatrice de la parole. Or Balzac propose un dereglement, voire une degradation en particulier du modele philosophique de la conversation, et Zola un direglement, une degradation en particulier du modele mondain. Prendre ä rebours ces modeles montrerait un phenomene de "de-civilisation" par la parole, pour reprendre une notion de Norbert Elias; cette prise a rebours mettrait en question la fonction civilisatrive de la conversation ä plusieurs - tout en liberant (c'est le versant positif) la fonction poetique du langage partage en groupe, sa fonction creatrice sur le plan des formes. Pour developper mon propos, j'ai eu recours egalement aux modeles du theatre et de 1'opera, car dans la pratique du dialogue de groupe, ils ont des specificites susceptibles d'inspirer l'ecriture romanesque. Ainsi, ils presentent au moins une caracteristique commune interessante, la pratique du chceur: plusieurs personnages (plus de trois) parlent ou chantent en meme temps pour dire la meme chose, ä l'unisson s'il y a musique. En ce qui concerne Γ opera, au moins deux pratiques propres meritent d'etre soulignees dans la perspective de cette etude : la polyphonie et la predominance du son sur le sens. Dans la polyphonie, plusieurs voix s'expriment en meme temps pour dire, pour chanter des choses dif-

Dialogues de groupe: Balzac, Zola

147

ferentes, par exemple chez Mozart, ou chez Rossini qu'evoque Balzac; il y a des conventions d'ecriture de la polyphonie dans les livrets d'opera: les repliques sont presentees l'une au-dessous de l'autre comme les repliques echangees l'une apres l'autre, mais le librettiste precise "duo", "trio" ou "tercet", ou encore (on se rapproche du "polylogue") "quatuor", "quintette", "sextuor". La polyphonie vise ä l'harmonie musicale dans un opera; ce n'est pas le cas dans nos dialogues, oil eile est plutot cacophonique. En d'autres termes, quand nos auteurs ont recours aux modeles du theatre et de l'opera pour operer leur parodie des modeles philosophiques et mondains, ces moyens eux-memes peuvent faire l'objet d'une parodie. La predominance du son sur le sens, quant a eile, s'impose quand un meme enoncö est repete non pour assurer la qualite de son expression, ou de sa comprehension, mais pour des questions de rythme ou de melodie; ou precisement quand plusieurs s'expriment en meme temps pour chanter des choses differentes et qu'on entend la musique plus que Ton ne comprend le propos. Cette prddominance du son sur le sens on la remarque egalement dans nos dialogue. Elle peut rentrer dans une Strategie parodique; mais aussi, elle peut servir, en particulier chez Balzac, ä faire valoir la fonction poetique du langage.

2. Balzac et le modele philosophique

Balzac parodie le modele philosophique du Banquet dans La Peau de chagrin, lors du diner et de la fameuse orgie censes feter la naissance d'un journal,2 et dans Le Pere Goriot oil plusieurs dialogues de groupe ont lieu a table, dans la pension Vauquer. Balzac n'est pas le premier a parodier le modele philosophique; Petrone Γ a dejä fait dans le Satyricon3, Lucien dans Le Banquet ou les Lapithes (2£me s. ap.J.C.), et, en France, έ la Renaissance, Rabelais dans le Gargantua, Rabelais que Balzac, precisement, dvoque dans La Peau de chagrin. La parodie, dans les deux romans balzaciens, s'effectue par le choix des circonstances et des personnages mis en presence: une orgie, des convives de plus en plus avinös, des femmes corrompues dans La Peau de chagrin·, les clients d'une pension minable dans Le Pere Goriot. En dehors du dialogue lui-meme, indiquent egalement la parodie les

2 3

Sur ce texte, Rosen a ddjä propose une riche analyse (1979: 115-126). Rosen (1979: 116-117), ä la suite de Florence Dupont, parle de festin dans ce cas, plutot que de banquet, le festin se caracterisant precisement par une "parole libre", le "coq ä l'ane", les "paradoxes", les "idees spectaculaires", le desordre, l'orgie...

148

Regine Borderie

commentaires du narrateur, et dans ces commentaires les images, les references. L'image de l'opera en particulier frappe dans La Peau de Chagrin: L'orgie seule deploya sa grande voix, sa voix composee de cent clameurs confuses qui grossissent comme les crescendo de Rossini. (PCh.: 98)

Si l'on en croit "clameurs confuses", la polyphonie operatique n'y est guere harmonieuse, d'autant que plus loin le narrateur, qui a note: "II semblait que chacun eüt deux voix" (Ibid.), evoque une "melee de paroles" (Ibid.). Dans Le Pere Goriot figure aussi une reference a l'opdra, et de meme la polyphonie y est cacophonique: En quelques instants ce fut un tapage ä casser la tete, une conversation pleine de coq-ä-l'äne, un veritable opera que Vautrin conduisait comme un chef d'orchestre [...] (PG: 202).

Si dans ces cas Balzac dit la cacophonie, si encore il la dit, toujours dans Le Pere Goriot, en precisant, avant d'enoncer les repliques, que les personnages parlent et meme "beuglent" "simultanement" (Ibid.), il trouve moyen de la montrer (visuellement, sur la page) en juxtaposant les propos sur la ligne au lieu de les ecrire, comme ä l'accoutume, l'une en dessous de 1'autre: [...] aussitöt huit voix beuglferent simultandment les phrases suivantes: "Ä repasser les couteaux! - Mo-ron pour les p'tits oiseaux! - Voilä le plaisir, mesdames, voilä le plaisir! - A raccommoder la faience! - A la barque, ä la barque! - Battez vos femmes, vos habits!- Vieux habits, vieux galons, vieux chapeaux ä vendre! - A la cerise, ä la douce!" (Ibid.)

A vrai dire, on ne sait s'il faut dans ce passage parier de "chceur" ou de "polyphonie", mais assurement il y a cacophonie, et parodie d'opera. De plus, toujours dans les commentaires, la reference ä Rabelais, a un texte lui-meme parodique, renforce la prise ä rebours des modeles nobles: Entre les tristes plaisanteries dites par ces enfants de la Revolution i la naissance d'un journal, et les propos tenus par de joyeux buveurs ä la naissance de Gargantua, se trouvait tout l'abime qui s6pare le dix-neuvifeme siecle du seiziÄme. Celui-ci appretait une destruction en riant, le notre riait au milieu des mines. (PCh.: 98-99)

Et le texte rabelaisien est lui-meme renverse puisque la joyeuse plaisanterie devient "triste[...] plaisanterie[...]": Balzac pointe un desenchantement ideologique propre au XlXeme siöcle, et le motif des "ruines" ramene a la question de la "de-civilisation". D'autre part, dans le dialogue lui-meme, au discours direct, on note, toujours dans une perspective parodique, un principe de rupture ou de discordance thematique. Ainsi, on parle politique ou philosophie dans La Peau de chagrin, et au milieu d'une replique, l'un des personnages, se coupant lui-meme, dit: "Passez-moi des asperges" (PCh:. 101), ou

Dialogues de groupe: Balzac, Zola

149

"(Passez-moi du canard)" (Ibid.: 106), la parenthese, qui suggere un changement vocal, attenuant ici le decalage. La rupture thematique, ailleurs, se produit entre deux repliques. Par exemple, au milieu de considerations politiques, un convive s'extasie: "Voilä des petits pois delicieusement fantastiques!" (Ibid.: 101). Balzac met en concurrence le corps et Γ esprit, au detriment de Γ esprit. Contribue encore ä la parodie le fait que les propos ne soient pas attribuds; on remarque dejä ce phenomene chez Rabelais, mais c'est un procede qui n'est pas frequent dans la tradition romanesque. La non attribution implique une conception de la parole eloignee de ce que preconisent la rhetorique et la poetique; elle ne sert plus a construire le personnage dont eile est detachee. Dans La Peau de chagrin, la parole est comme une pensee sans penseur, volontiers exprimee au present gnomique: Ah! la gloire, triste denree. Elle se paye eher et ne se garde pas. Ne serait-elle point l'6goi'sme des grands hommes, comme le bonheur est celui des sots? - Monsieur, vous etes bien heureux. Le premier qui inventa les fosses 6tait sans doute un homme faible, car la societe ne profite qu'aux gens chetifs. Plac6s aux deux extrimiti du monde moral, le sauvage et le penseur ont dgalement horreur de la propriite. (PCh.: 101) Qui parle ici? Et plus loin qui dit: Oh! oh! Oui et non, n'est-ce pas l'histoire de toutes les dissertations religieuses, politiques et litteraires? L'homme est un bouffon qui danse sur des precipices! [...] - L'instruction, belle niaiserie! [...] (Ibid.: 102) La parole sert ä exprimer des idees qui ne sont ä personne; et la valeur gnomique du present ne donne pas au propos valeur de verite, au contraire - on est loin de la force de conviction habituellement pretee ä la sentence ou ä la maxime, sans doute ä cause des circonstances orgiaques. Cette parole sans maitre est une parole sans credit; dans ce contexte, moins la parole est attribuee, puis eile parait gratuite, et subjective, non au sens oü eile individualiserait, mais au sens oü eile est discutable parce que relative ä un point de vue bouffon, provocateur, polemique... Plus encore, cette non attribution de la parole temoigne ä son tour du desenchantement ideologique dont parle Balzac au debut de l'orgie; de fait, les propos ne sont pas attribues parce que les locuteurs ne tiennent pas a leur pensee. Vision sombre d'une societe en voie de democratisation: c'est pr6cisement au moment oil on peut parier politique (il s'agit de feter la naissance d'un journal), qu'on ne croit plus ä la politique ni a la morale. Sur le plan psychologique, des comportements propres au dialogue de groupe apparaissent. On remarque, notamment, des effets de spectacle par theätralisation du dialogue, theatralisation qui n'est pas surprenante; quand on est plusieurs, en effet, il est difficile que tous s'expriment, ceux qui se taisent sont done un public tout trouve pour ceux qui parlent, public devant lequel peuvent s'agiter notamment des bouffons, des pitres (c'est

150

Regine Borderie

un röle, comme celui de meneur ou de spectateur). C'est le cas en particulier dans La Peau de chagrin oü Bixiou "fait" le dix-neuvieme siecle, et se livre ä une pantomime bouffonne, la bouffonnerie consommant l'adieu au serieux philosophique (PCh.: 106). On note aussi le phenomene de "la chasse ä l'hotnme" pour reprendre une expression de Freiich; c'est le principe, dans le groupe, du "tous contre un", principe actualise dans Le Pere Goriot: Une reunion semblable devait offrir et offrait en petit les elements d'une societe complete. Parmi les dix-huit convives il se rencontrait, comme dans les colleges, comme dans le monde, une pauvre creature rebutee, un souffre-douleur sur qui pleuvaient les plaisanteries. (PG: 62-63)

L'agressivite collective se dechatne contrec ce "patiras" (Ibid.): les pensionnaires redoublent de mauvaises, et meme de cruelles plaisanteries contre leur tete de turc, leur bouc emissaire, autre role. La psychologie et la psychanalyse le soulignent, le fait d'etre en groupe facilite le debridement des mauvais instincts.

3. Zola et le modele mondain

Le dereglement et la degradation du modele mondain en realite sont dejä ä l'ceuvre chez Balzac, qui a la nostalgie de la conversation d'Ancien Regime comme d'autres auteurs de cette epoque. De cette nostalgie, Balzac se console parfois en operant une resurrection des formes, telles qu'il les imagine, de la conversation d'antan, par exemple dans Autre Etude de femme. Zola, ä la suite de Flaubert, opere une degradation radicale, sans reste, sans nostalgie. Ainsi dans Pot-Bouille, le salon Josserand et le salon Duveyrier se signalent par la mediocrite, voire la vulgarite des lieux, des personnages. Le desordre, la polyphonie cacophonique s'imposent. Iis sont dits par le narrateur ("Tous parlerent ä la fois" (107)), ou encore mis en valeur par une scene comique, a valeur analogique, dans une evocation du salon Duveyrier oü la maitresse de maison fait chanter aux invites un extrait d'un opera de Meyerbeer: Et la scene s'engagea [...]. Ce debut fiit comme une surprise, car les voix s'etouffaient sous le plafond bas, on ne saisissait qu'un bourdonnement, comme un bruit de charettes chargees de paves, dont les vitres tremblaient. [...] Le salon s'echauffait, les seigneurs criaient ä la volee [...] et, chaque fois, c'etait une explosion qui allait frapper chaque invite en pleine poitrine. - lis chantent trop fort, murmura Octave ä l'oreille de Mme Hedouin. (112-113)

Cette scene de chant rate est l'image des conversations. Quand il detaille les propos echanges, Zola devoile les arriere-pensees de ceux qui expriment, dans le meme salon, des

Dialogues de groupe: Balzac, Zola

151

opinions politiques; au contraire de ce que Ton a vu chez Balzac4, il y a attribution et individualisation des 6nonc£s, ce qui amoindrit le rythme du dialogue, et l'impression de vitalite. Mais l'incise ou les commentaires attribuant et expliquant les paroles permettent d'eclairer brutalement l'egocentrisme, la mesquinerie des interets personnels qui, en realite, l'emportent sur l'interet collectif: les personnages adoptent les opinions politiques qui les arrangent: [...] le docteur Juillerat, d'opinion ath6e et rdvolutionnaire, soutenait qu'il fallait donner Rome au roi d'ltalie [...]. - Peut-etre trouverait-on encore un modus vivendi acceptable de part et d'autre, fit remarquer Leon Josserand, qui arrivait. II 6tait alors secretaire d'un avocat c61£bre, deput6 de la gauche. Pendant des annees [...], il avait promen6 sur les trottoirs du Quartier latin une dimagogie feroce. Mais depuis son entree chez les Dambreville [...] il se calmait, il tournait au republicain doctrinaire. [...] - L'Eglise disparaitra, allons done! dit Campardon d'un air furieux, pour faire sa cour au pretre, dont il attendait des travaux. [...] (106)

La encore est montr6e une faille importante dans une soci6t6 en voie de democratisation, societe dont le principe de solidaritö est mis en question. II Test encore des lors que la haine r£gne; e'est bien elle qui federe les bonnes contre leurs maitres. Car Zola evoque leurs dialogues, trouvant l'occasion d'une autre parodie eclatante du modele mondain, parodie par le lieu (les bonnes parlent entre elles par les fenetres de leurs cuisines qui donnent sur l'arriere-cour ecceurante de salete), par les personnages (vulgaires socialement et moralement), par le theme (les bonnes s'occupent a devoiler les frasques sexuelles des bourgeois de l'immeuble, et leur hypocrisie: les propos sont des commerages dont la progression se fait sur le mode questions/reponses, ou par la juxtaposition, chacune y allant du sien a propos de ses maitres)5, et par les formes (le langage est ordurier). L'adieu aux bienseances est complet! Pour caracteriser ces echanges, Zola parle de "bavette" (on pense ä la bave des crapauds...) (127), il evoque l'"egout" (130), la "vidure des cuisines", le "linge sale de la domestique" (288), un "flot d'ordures" (306). Sur le plan de la representation psychologique et sociale, eclate done l'agressivite des bonnes. Cette agressivite est d'abord interne: elles se disputent, s'insultent - on a encore un exemple de "chasse a 1'homme" quand elles rivalisent d'injures provocatrices ä l'egard de l'une d'entre elle, autre "souffre-douleur" (129) qui toutefois trouve une alliee en la

4

5

A vrai dire il faut mettre έ part, dans l'orgie, telle reflexion anti-monarchiste d'un jeune honime "devenu r6publicain faute d'une syllabe devant son nom." (PCh., 100) Pierre Marotte, dans l'ddition de Pot-Bouille utilisee, parle, ä propos des conversations de bonnes, d'une nouvelle version du chceur antique: "C'est une des trouvailles les plus fortes de Pot-Bouille que la mise en seine de ce choeur vdridique et ordurier, parfaitement accorde au projet de devoilement de toute l'ceuvre." (454-455)

152

Rigine Borderie

"nouvelle bonne" (Ibid.), comme Godot a pu trouver un soutien en Rastignac. Mais ici la victime a "Γesprit de corps": Et les voix grandissaient, des mots aigres commenfaient a s'echanger entre Lisa et la nouvelle bonne, Fran9oise, qui prenait parti pour Adele, lorsque cclle-ci, oubliant les injures, cddant ä 1'instinct de l'esprit de corps, cria: -Mefiance! v'lä madame! Un silence de mort tomba. Toutes brusquement avaient replongi dans leur cuisine [...] (129) La reaction d'Adele montre que la solidarite sociale l'emporte sur la discorde au sein d'un meme groupe. L'hostilite entre classes est la plus forte. Mais encore, toujours dans une perspective socio-politique, on remarque en meme temps l'assujettissement aux maitres, ce qui ramene ä la question de la progression du dialogue, au röle sur ce plan des arrivees (de l'ev6nement non verbal), et ä la theatralite du "polylogue": quand un maitre parait, ou quand on croit qu'il va parattre, c'est le silence et le repli, avant le retour aux fenetres - on se croirait au guignol, dans un theatre de marionnettes: Elle n'acheva pas. Le bruit d'une fenetre qui s'ouvrait brusquement, les mit en fuite. II se fit un profond silence. Mais elles se risquerent de nouveau. Hein? quoi? qu'y avait-il? Elles avaient cru que Mme Valerie ou Mme Josserand les surprenait. (128)

4. Fonction poetique du dialogue de groupe

La cacophonie, P"ordure", l'agressivite ne doivent pas faire oublier que le "polylogue" est egalement un lieu de jouissance de la parole, comme chez Rabelais; il favorise Peuphorie collective, le groupe secrete sa propre excitation, c'est une fete (E.Rosen parlait de carnaval, certes destructeur). Cette id6e de fete du langage rappelle le phenomene de theätralisation; si la parole est spectacle, c'est qu'elle est ludique, agreable. Le plaisir, la jubilation sont sensibles chez Balzac dans le fait que la parole ne soit pas toujours attribuee, on y revient. Dans ce cas, la parole compte pour elle-meme. Le rythme, rapide, allegre, entraine. De tres courtes repliques s'enchainent, dans La Peau de chagrin, sans le frein des incises. Par exemple: - Vous etes un sot! - Vous etes un drole! - Oh! Oh! - A h ! Ah! - lis se battront.

Dialogues de groupe: Balzac, Zola

153

-Non. - A demain, monsieur [...] (PCh.: 105).

De plus, le "polylogue" est l'occasion de bouffonneries langagieres chez Balzac, toujours emule de Rabelais. Plaisanteries macabres, provocatrices, images dröles et iconoclastes pour parier de societe, de politique, de religion, s'accumulent dans le banquet de La Peau de chagrin. On trouve aussi beaucoup de jeux sur les sons dans Le Pere Goriot, tout impregn6 du modele musical, operatique: on se rappelle les fameuses plaisanteries en rama", la liste des mots en "cor-" ("Cor-nouille", "Cor-nemuse" etc. (PG: 92-93)). Les mots sont comme des choses avec lesquelles les pensionnaires s'amusent. Chez Zola, dans les conversations de bonnes, la jubilation est dite. On voit que les bonnes s'en donnent a cceur joie dans l'allusion et les jeux de mots obscenes, car ceux-ci les font "rican[er]", les "chatouill[ent]" (128); et pour les "gros mots" injurieux, elles s'expriment "ä plein gosier" (129). La jubilation est aussi montree par l'inventivite dans l'insulte des personnages: - Voilä torchon! [...] - Tiens! elle s'est lavde, ?a se voit! - Tiche encore de jeter tes vidures de poisson dans la cour, que je monte te debarbouiller avec! - Eh! va done manger le bon Dieu, fille a curi!... Vous savez, elle en garde dans ses dents pour se nourrir toute la semaine. (129)

Tous ces exemples permettent de souligner un point d'esthötique bien connu, la predilection pour le bas du realisme, du naturalisme. On voit surtout comment esthetique et ethique se desolidarisent puisque la parole jubilatoire fait mal aux "souffre-douleur" et autres "patiras", puisqu'elle exprime et alimente la haine. Esthetique et democratisation de la parole font alliance, en revanche: la jubilation verbale, voire l'invention verbale sont attribuees, notamment, aux habitues d'une pension minable dans Le Pere Goriot, dans Pot-Bouille a des bonnes. Les deux romanciers montrent la creativite du bas social et du bas moral. Ainsi, de la parodie des modeles mondains et philosophiques ressort bien l'idee d'une "de-civilisation": sur le plan psychologique, le dialogue de groupe permet la liberation de l'agressivite, sur le plan socio-politique, il revele une societe "en ruines", desenchantee chez Balzac, individualiste, hypocrite, haineuse chez Zola. La conclusion est sur ces plans plutot sombre. Mais sur le plan verbal, pour les personnages comme pour le romancier, le "polylogue" est une occasion de jubilation, de libertd, et meme de cröativite langagiöre. Cette ambivalence peut rappeler ce que disait Freud de l'homme en groupe dans "Psychologie collective et analyse du ΜοΓ: la collectivite ce n'est pas seulement le foyer de l'agressivite, une occasion de debordement pour les mauvais instincts, e'est aussi un

154

Regine Borderie

foyer de culture, le lieu d'elaboration de la langue, des chants populaires, du folklore. (Freud 1921, 1948: 91)

References bibliographiques

Balzac, H. (1835, 1976): Le Pfere Goriot. - Paris: Gallimard, La P16iade, Vol. III. - La peau de Chagrin (1831,1979). - Paris: Gallimard, La Pleiade, vol. X. Borderie, R. (1999): "Polylogues" balzaciens. - In: E. Bordas (ed.) "Balzacien". Style des imaginaires. Eidölon 52, 55-76. Bordeaux III: Universite Michel de Montaigne. Freud, S. (1921, tr.fr. 1948): Psychologie collective et analyse du Moi. - In: Essais de psychanalyse, 76-162. Paris: Payot. Freiich, J. (1991): Au parloir du roman de Balzac et de Flaubert. - Oslo: Didier Erudition, Solum Forlag A/S. Hellegouarc'h, J. (ed.) (1997): L'Art de la Conversation. - Paris: Classiques Garnier. Rosen, E. (1979): Le festin de Taillefer ou les saturnales de la Monarchie de Juillet. - In: C. Duchet (ed.) Balzac et "La Peau de Chagin" 115-126. Paris: SEDES. Zola, Emile (1882, 1984): Pot-Bouille. - Paris: Le Livre de poche.

Adelino Cattani

II disputator cortese e il disputator polemico

La maggior parte dei testi di logica ο di teoria dell'argomentazione ci dice come dovrebbe svolgersi un buon dibattito, quali sono le regole e qual e il comportamento giusto con un awersario leale durante una discussione. Noi invece sappiamo che gli awersari ignorano le regole di buona condotta dibattimentale (e a loro volta gli awersari si lamentano di noi per lo stesso motivo) e avremmo bisogno di sapere come comportarci per fare fronte alle mosse, ai trucchi, agli espedienti e ai giochi sporchi. Chi gioca, ο semplicemente assiste, ad una partita di calcio possiede una competenza ed una conoscenza delle regole del gioco che gli consentono di produrre delle prestazioni in campo ο di essere in grado di valutare sensatamente le altrui prestazioni e di giudicare, ad esempio, se il centravanti ha giocato bene ο male, al di lä ο contro i giudizi dei tecnici e dei giornalisti sportivi. Nel gioco del dibattito non esiste nemmeno lontanamente un'analoga conoscenza/competenza. Chi, da noi, discute lo fa in base a proprie capacitä, per lo piü innate, e chi valuta gli argomenti addotti dall'interlocutore ο dai disputanti si deve affidare al proprio senso istintivo del giudizio, non educato, non coltivato. Non ci sono quindi regole codificate di una buona discussione, ma al massimo semplici prassi e consuetudini. In secondo luogo, c'e una tradizione irenistica, per cui si privilegia la cooperazione rispetto al conflitto ed una cultura pseudo-gandhiana per cui e socialmente censurabile andare alio scontro, cercare il conflitto, entrare in polemica aspra. Non e male perö ribadire l'awertenza che il mondo e molto piü malvagio e pieno di insidie di quanta si sia indotti a credere sulla base di alcune presunzioni logico-etiche di trasparenza e cooperazione, tipiche della tradizione europea dell'umanesimo, della razionalita e della sinceritä; per essere in linea con questa tradizione e politicamente corretti, si dovrebbe invitare a tenersi nei pressi di questa corrente; ma per usare una felice immagine di Leeman (1992), questa corrente non e un grande e maestoso flume, ma solo un rivolo intermittente, un rigagnolo (che comunque rende fertili le zone immediatamente circostanti). Manca una cultura del sano conflitto. Si dimentica che quasi ogni attivita umana e di natura competitiva, in grado diverso, che va dal concorrenziale al conflittuale, e che cooperazione e conflitto sono reciprocamente connessi: non e un paradosso, ma una constata-

156

Adeline Cattani

zione, che dietro ad ogni conflitto c'e una componente di cooperazione1 e che "non si puo discutere se non si e d'accordo", d'accordo su una qualche base di partenza. Sembrera paradossale l'idea che non si possa discutere se non si e d'accordo, ma avere almeno un punto di partenza comune, una premessa condivisa e essenziale come l'avere un metro comune se si intende attuare una misurazione. Come esperienza e psicologi insegnano, un buon litigio coniugale e salutare per la coppia; cosi darsele no, ma dirsele di santa ragione fa bene anche all'intern ο della comunitä dei ragionanti. Ebbene, una discussione puo avere natura dialogica ο polemica. Ciascuno di questi due tipi d'interazione discorsiva si dovrebbe svolgere nel rispetto di certe regole, al pari di qualunque gioco. II dibattito processuale e regolamentato da norme esplicite e da procedure codificate nei dettagli per quanta riguarda l'onere della prova, la formazione delle prove, l'ordine degli interventi, le domande ammissibili eccetera. Anche il dibattito parlamentare ha le sue regole previamente stabilite e consensualmente accettate. Ma difficilmente questa codificazione e trasferibile alle normali discussioni, dove al massimo vigono le regole della buona creanza e della buona comunicazione, regole non scritte ne esplicitate, se non in qualche teoria logico-linguistica (come le regole della cortesia di Robin Lakoff (1973), le regole conversazionali di Paul Grice (1975) ο psico-sociologica (dove si rilevano piü delle regolarita statistiche che delle norme). Ne soprattutto vi sono regole le cui violazioni siano sanzionabili. Ne ci sono verdetti defmitivi. Questo pero e anche il bello della discussione. II gioco della discussione, e le relative regole, sono molto speciali, in primo luogo perche chi dialoga ο polemizza e autorizzato a ridiscuterle e a concordarle con l'interlocutore. In secondo luogo, l'accettabilitä delle regole non e standardizzata, ma dipende dall'ambito in cui si discute: in un dibattito parlamentare sono ammesse mosse che sarebbero fuori luogo ο totalmente inaccettabili in sede di commissione bioetica. Chi infatti contrawiene a queste regole viene meno alia "integritä dibattimentale" (fatto trascurabile per uno che intende ostentare la sua "capacita dibattimentale"): gli si puo quindi imputare una delle tante fallacie, che sono in effetti piü pecche comportamentali che logiche, hanno cioe piü a che fare con le regole della buona condotta che con le regole della veritä e della validitä. II mio intervento si propone di abbozzare un codice di condotta per una discussione cooperativamente polemica. A regola di codice - come invitava a fare Paragone, il buffone di corte di Cosi e se vi pare (V, IV, 75 ssgg.) - si puo anche polemizzare, oltre che dialogare:

A sostegno e a documentazione di questa tesi si puo vedere il recente volume di E. Arielli e G. Scotto (1999:1): "pensare al conflitto semplicemente come opposizione fa perdere di vista un elemento essenziale per poterlo capire: la cooperazione".

II disputator cortese e il disputator polemico

157

"Signore, noi litighiamo a regola di codice, come voi avete il codice delle buone maniere". Sette sarebbero i modi di smentire e i gradi della contesa, che vanno dalla Replica Cortese al Sarcasmo Discreto, dalla risposta Villana alla Ferma Riprovazione; dalla Ritorsione Provocatoria alla Smentita Sotto Condizione per finire con la Smentita formale. 2 Dopo di che non rimane che il duello. Le regole prese in considerazione non sono tanto quelle logiche del corretto ragionare, che garantiscono la veritä della conclusione, ma quelle, di carattere insieme etico e pragmatico, che garantiscono la correttezza dibattimentale. Ripeto, regole di natura insieme etica e pragmatica, che di norma si pretende siano rispettate dagli altri e che considereremmo spiacevole se fossimo sorpresi a trasgredirle. Si tratta di regole per una proficua discussione, ricavate con metodo alquanto empirico (nel senso che e la pratica ad attestare che la loro osservanza promuove la discussione e facilita un'eventuale soluzione) ma che trovano un certo riscontro in alcune recenti ricerche sulla teoria deH'argomentazione. Sono poche regole raccomandate ο raccomandabili, abbastanza logiche ed intuitive ο suggerite dalPesperienza, che dovrebbero rendere fattibile un dibattito decente, quello cio6 nel quale si cerca di ridurre al minimo le mosse sleali e di svincolare il piü possibile le conclusioni da pregiudizi e dati irrilevanti. Quelle a cui penso sono regole, empiriche, di almeno tre tipi: in primo luogo, regole metodologiche di organizzazione del discorso; in secondo luogo, regole di comportamento dibattimentale, ed infine regole di valutazione del dibattito, che di fatto e di norma si rivelano efficaci in una discussione; regole che si possono tradurre in una sorta di decalogo dei diritti del libera disputante. Sul dovere di dialogare e di discutere si sono espressi in molti. Si potrebbero citare maree di inviti imperativi a lasciare che gli altri esprimano liberamente le loro idee e che cerchino di convertirci ad esse, come noi facciamo con loro, e sul "dovere di ascoltarli con la stessa buona volontä di comprendere le loro veritä e di farle nostre che noi reclamiamo da loro per le nostre" (Calogero 1950). Sempre sia lodata questa esortazione. Meno parole sono state spese sui diritti. Tutti, dalla foca monaca al lettore, hanno qualche diritto che

W. Shakespeare: Come vi piace, V, IV, 75 ssgg. "Non mi piaceva il taglio della barba di un certo cortigiano. Egli mi fece sapere che, se io affermavo che la sua barba era mal tagliata, lui invece era convinto del contrario: questo va sotto il nome di Replica Cortese. Se io avessi insistito che la barba non era ben tagliata, lui mi avrebbe mandato a dire che la barba se la tagliava come gli pareva: avremmo avuto il Sarcasmo Discreto. Se io ancora a insistere, "No, non e tagliata bene per niente", e lui avesse risposto che non avevo giudizio avremmo avuto la Risposta Villana. Se io, duro, avessi ancora detto che il taglio non andava, ed egli avesse risposto che negavo il vero, eccoci di fronte alla Ferma Riprovazione. Se io, piü duro che mai, "No, non e tagliata bene" e lui avesse detto che mentivo, ecco la Ritorsione Provocatoria. Cosi via fino alla Smentita Condizionale e alla Smentita Formale." (Trad, di A. Calenda e A. Nediani, Milano, Mondadori)

Adeline Cattani

158

qualcun altro si occupa di codificare. Quelli da accordare a chi si trova a discutere sono tutti da definire. Mi permetto di suggerime qualcuno. Quale primo abbozzo, comincerei con dieci, numero regolamentare di ogni tavola che si rispetti. I died diritti del libero disputante 1. Diritto di esprimere il dubbio su tutto, perche niente e "fuori discussione". Ε un diritto che e anche un dovere e talvolta diventa un piacere: il dovere ed il piacere di replicare. 2. Diritto di non dire tutta la veritä. Lo dice persino il giudizioso Quintiliano: una buona ragione puo in certi casi indurre l'uomo onesto, che deve difendere una causa, a nascondere la verita. 3 Ma niente paura. Vi corrisponde il diritto compensativo riconosciuto alia controparte di reclamare tutta la verita dagli altri, tenendo tuttavia presente che il "nient'altro che la verita" si puo pretendere solo in tribunale. 3. Diritto di sottrarsi al gioco dell'awersario e di svincolarsi dalla sua dipendenza. Quando uno parla, l'altro e tenuto a rispondere. La replica potrebbe essere considerata come la reazione ad uno stimolo. Fortunatamente pero il rapporto fra due che discutono non e meccanico come la regola di reattivita fisica vorrebbe, ma e improntato ad una estrema libertä, per cui la risposta che il secondo interlocutore deve comunque dare al primo - se vuole essere cooperativo, se non vuole fare brutta figura, se non intende perdere punti ο perdere la faccia di fronte aH'uditorio - si puo collocare in un continuum che va dalla risposta perfettamente a tono alia risposta completamente fuori luogo. Come ha dettagliatamente indicato Pierre Oleron (1986), fra questi due estremi c'e spazio per tutta una gamma di risposte parzialmente pertinenti ο impertinenti. Esiste dunque nel replicare un prezioso margine di libertä che consente di rispondere in maniera da assecondare Γ interlocutore e da soddisfare l'uditorio, senza subire tuttavia l'iniziativa deH'awersario. Questa inestimabile libertä di movimento permette, anche in fase di replica, di uscire dalla difensiva, e di prendere le distanze dai temi, dagli argomenti e dalla strategia dell'altro; di non farsi irretire nel suo gioco e di assumere il controllo dell'interazione. 4. Diritto di difendere, in maniera attiva ο passiva, le proprie posizioni e se stesso. Come diceva John Stuart Mill, nel suo manifesto libertario, c'e sempre speranza quando si e costretti a sentire tutte e due le parti; e quando ci si attiene ad una sola che gli errori si cristallizzano in pregiudizi e la stessa veritä perde il suo valore di veritä. Quintiliano, Istituzione oratoria, XII, 1, 36.

II disputator cortese e il disputator polemico

159

5. Diritto di poter concludere il proprio discorso. Diritto elementare, ma non sempre accordato. Ricerche ad hoc sembrano attestare che l'aggressivita litigiosa e l'intemperanza verbale (alzare la voce, sovrapporsi, attaccare la persona e non le sue tesi, beffeggiare...) generalmente non pagano e sono meno producenti delle buone maniere, dei toni pacati, degli argomenti ragionevoli. 6. Diritto di aspirare alia vittoria. Con il suo corollario: diritto di non essere cooperativo. L'essere competitivi non έ incompatibile con una sana discussione. "II disaccordo rende piü prezioso l'accordo." 4 Cosa sicuramente buona e giusta e l'accordo, buoni e giusti sono il consenso e l'unanimita, ma solo quando costituiscano un'autentica conciliazione di divergenze, al termine di un dibattito che queste divergenze non abbia mascherato ο annullate, ma ne abbia viceversa favorito l'emergere e il confronto. 7. Diritto di usare ognuno gli argomenti che possiede e che preferisce. Include il diritto, all'occorrenza, di lasciare da parte la logica e di indignarsi; il diritto di non fingere sempre equilibrio, moderazione e compostezza e di usare le unghie e i denti; il diritto di adattarsi alle condizioni specifiche del dibattito in corso. Essendo in due a giocare, la controparte ha a sua volta il diritto di porre il veto sulle mosse giudicate "discutibili". In un dibattito il problema grosso non e tanto la presenza di fallacie, ma la loro identificazione e neutralizzazione. Visto che cosi va il mondo, meglio contemplare pragmaticamente anche questi nella tavola dei diritti inderogabili. Esortare ad un dialogo onesto e un esercizio edificante che e bene continuare a praticare, ma con qualche disincanto e contemperando questo invito con la realistica presa d'atto che, rubando la formula a Torquato Accetto, il disputante deve cercare di vivere tra gli inganni non ingannato.5 Senza la complicitä colpevole del giocato, responsabile per ignoranza, il gioco ha meno probability di riuscire. 8. Diritto di appellarsi ad una terza parte. La terza parte sara l'uditorio (che va trattato come diente, ma che non sempre ha ragione) ο il giudice (che emette un verdetto, ma non stabilisce l'accettabilita di una tesi). Quando due discutono pensano fatalmente a tutti coloro che Ii ascoltano e che potrebbero essere condizionati, influenzati dal dibattito. Ne tiene conto, e lo confessa, anche il chirurgo Carlo Marcelletti quando deve esprimere un parere di fattibilitä di un intervento: "lo ammetto, ho pensato a cosa avrebbe detto l'opinione

5

Publilio Siro, Sententiae, n. 151. "...se questa έ la condizion dell'uomo, basterä conoscerla e vivere tra gli inganni non ingannato." (T. Accetto, Rime, L'autor a chi legge, 1621.)

160

Adeline

Cattani

pubblica davanti al mio stop". 6 Per dirla con le parole di Platone-Socrate nel Protagora: "Una cosa e trovarsi insieme a discutere, un'altra e tenere discorsi in piazza." 7 9. Diritto di essere giudicati per quello che si pensa e che si dice, non per quello che si e fatto. Ne deriva il diritto collateral che non vengano rinfacciati solo comportamenti, quando si sta discutendo di idee e opinioni. A differenza di chi deve perseguire il corrotto e non la corruzione, il logico e bene che persegua la fallacia e non il fallace. 10. Diritto di cambiare regole e diritti della discussione. Ε di farlo, in corso di discussione, d'accordo con Pinterlocutore. Chi partecipa ad un dibattito infatti puo riformulare, rivedere, ridiscuterne le regole, le procedure, le mosse lecite, la conduzione. In altri termini, oggetto di dibattito diventa anche "come dibattere". 8 Dover continuamente argomentare e controargomentare e impegnativo; ma essere in condizione di non poterlo fare sarebbe molto piü gravoso, anche quando, come capita per lo piü, i due che discutono lo fanno "da sordi" e non si spostano di un millimetre dalle loro posizioni. Lo scopo ideale di una buona discussione e quello di trasformare una contrapposizione di argomenti aventi la forma di "un pro e un contro", "un χ vale quanta y" in un modulo selettivo avente la forma "un pro ο un contro", "un χ e meglio di y perche"; piü realisticamente, lo scopo e quello di consentire ad una terza parte, Puditorio, di chiarirsi un po' le idee e di operare una valutazione ponderata proprio grazie al confronto ο alio scontro fra due protagonist! irriducibili.

7 8

Con riferimento alia decisione di procedere al tentativo, poi fallito, di separare due gemelline siamesi. Dichiarazione in "Panorama", 8 giugno 2000, p. 38. Platone, Protagora, 336b. Questa tavola dei diritti del libera disputante, presentata nel 2000 al Convegno Dialogue Analysis, e stata poi ripresa nel volume Botta e risposta edito nel 2001 da II Mulino.

II disputator cortese e il

disputatorpolemico

161

Riferimenti bibliografici Accetto, T. (1987): Rime. L'autor a chi legge-Torino: Einaudi. Arielli, E. e G. Scotto (1999): I conflitti. Introduzione a una teoria generale. - Milano: Mondadori. Calogero, G. (1950): Veritä e libertä. Appendice di Logo e dialogo. Saggio sullo spirito critico e sulla libertä di coscienza, 193-198. - Milano: Comunitä. Cattani, A. (2001): Botta e risposta. L'arte della replica. - Bologna: II Mulino. Grice, P. (1975): Logic and Conversation - In: P. Cole e J. L. Morgan (eds.): Syntax and Semantics. Vol. 3: Speech Acts, 41-54. New York and London: Academic Press; Trad. it. M. Sbisä (a cura di) (1978): Gli atti linguistici, 199-219. Milano: Feltrinelli. Lakoff, R. (1973): The logic of politeness; or, minding your P's and Q's. - In: C. Coram, T.C. Smith-Stark, A. Weiser (eds.) Papers from the Ninth Regional Meeting of Chicago Linguistic Society, 292-305. Chicago: University of Chicago. Trad. it. M. Sbisä (a cura di) (1978): Gli atti linguistici, 220-39. Milano: Feltrinelli. Leeman, A.D. (1992): Rhetoric versus argumentation theory. - In: F. H. van Eemeren et alii, Argumentation Illuminated, 22. Amsterdam: SICSAT. Mill, J.St. (1859): On liberty. Trad, it.: Saggio sulla libertä. Milano: II Saggiatore- Est (19992). Oldron, P. (1986): Organisation and articulation des echanges de parole. - In: M. Meyer (ed.), De la mdtaphysique ä la rh£torique, 57-77. Bruxelles: Editions de l'Universitd de Bruxelles. Platone (1994): Protagora. - Milano: Mondadori. Publilio Siro (1968): Sententiae. - Torino: Giappichelli. Quintiliano (1989): Istituzione oratoria, XII, 1. - Bologna: Zanichelli. Shakespeare, W. (1992): Come vi piace. Trad, di A. Calenda e A. Nediani. - Milano: Mondadori.

Laurentia Dascälu Jinga Metadiscoursive Triads

1. Preliminary remarks

The present paper analyses some examples of metadiscursive sequences typically organized in three-part exchanges, in which the third turn's function is ratifying (or acknowledging) the outcome of the first two ones. After examination of a large spontaneous corpus we found that numerous utterances containing various types of repetitions are structured in three-turn sequences, forming a metalinguistic triad (=MT). We shall investigate here some examples of this type, intending to find new evidence in favour of a recent hypothesis claiming that the conversation structure is "essentially dialectical, consisting of a three-step sequence: utterance - reply response" (cf. Fleisher Feldman/Kalmar 1996:80); to give an example, let us mention here the non-metalinguistic triad analyzed by Schegloff (1972): summons - answer - reply (e.g. John! - What? — I need the hammer). Our pragma-phonetic investigation is based on two distinct types of audio records: (a) spontaneous informal conversations between relatives, friends, colleagues in their normal ambient (about 26 hours) and (b) television talk shows (=TV), about 30 hours. These records are parts of a large corpus of spoken Romanian (COrpus de Romäna Vorbitä = CORV), representative for middle or highly educated native speakers, i.e. persons speaking what has been called - in Sabatini's (1985) terms - Romanian "dell'uso medio" (Dascälu Jinga 2002a: 22). In what follows, the abbreviations used to indicate various examples refer to the number of the cassette, the face on it, as well as the exact point where the sequence occurs on the tape, and the speaker who uttered it; e.g. : 7B-153-VJ, TV 34B-567-HRP; the intonation of some sequences (as revealed by spectrographic analysis) is tentatively suggested by typographical arrangement of the letters ('Bolinger system'), the sentence stress by bold characters, and the place of'other'-interruption by [i]. In its function of gloss on the discourse, the metadiscourse can intervene in order to explain the conversation conditions or the proper code is used.

164

Laurentia Dascälu Jinga

2. MTs oriented to the conversation conditions

2.1. MTs related to the reception MTs related to the deficient reception are used by speaker Β in order to call for the repetition and/or explanation of a word occurring in speaker's A previous turn.

2.1.1. A common triad of this type is the sequence utterance - query - reply, studied by Garvey (1977): dialogues of the ABA pattern in which an initial utterance by a speaker A is followed by a general request for repetition by speaker B, in its tum, followed by some kind of repetition by A of her/his initial utterance. 'Query' is another term for what Bolinger (1957: 141) calls "general reclamatory question"; this category contains the interrogative words what and how, stereotyped combinations of them (What's that?, How's that?) equivalent expressions (I beg your

pardon?,

Come again?), or non-lexical signals (for a pragmatic and phonetic analysis of eh? in Italian, see Dascälu Jinga/Vanelli 1996: 402-403, 421-422). This signal can not be considered a proper question, as it does not ask for information, but only the repetition of (a part of) the previous utterance (which can be a question, an assertion, an exclamation, etc.). It can not be viewed as an answer either. Poggi/Castelfranchi/Parisi (1981: 576-577) call it a 'pre-answer' reply, "whose goal is the fulfilment of the prerequisite conditions for answering". As we shall see, the common feature of all these queries (or 'reclamatory questions' or 'pre-answer replies') is the rising intonation pattern. The most of the MTs containing a query in our corpus refer to unknown words or family names, which, being unpredictable, can cause uncomprehension in Β (the hearer). He can say an ordinary query or (in most informal situations) he can use a non-lexical signal, as the hummed (close-mouthed) interjection, transcribed by odd 'words' such as mhm, mh, mm, etc. (as described for Romanian in Dascälu Jinga 1996: 26-28). These cases are illustrated in the two successive MTs in example (1): (1)

7B-153-VJ: La Sighifoara am avut un foarte distins om cu care Jaceam rusa care se numea Gvozdinski (I). um? (II) LDJ: C VJ: Gvozdinski, dar ftia perfect rusefte, nu [ i ] (III) LDJ: Cu 'c' saucu 'g'? VJ: Cu 'g': Gvozdinski. Era un om foarte cumsecade, extraordinar. (I) hm? (II) LDJ: Μ VJ: Extraordinar de cumsecade. (Ill)

Metadiscursive Triads

165

'VJ: In Sighisoara there was a very distinguished person who taught us Russian and who was called Gvozdinski. (I) LDJ: I beg your pardon? (literally: "How?") (II) VJ: Gvozdinski, but he spoke Russian splendidly, not [i] (III) LDJ: With a 'c' or with a 'g'?

VJ: With a 'g': Gvozdinski. He was a very fine man, exceptional. (I) LDJ: Huh? (II) VJ: Exceptionally fine. (Ill)'

In informal conversations, Β can even repeat only the part which he has clearly understood, then add to it the interrogative word, thereby resulting a sui generis word: (2)

5B-285-VJ: Asia e patognomonic la noi. (I) cum? (II) LDJ: Patogno... VJ: Patognomonic, adicä un semn care e foarte caracterizant. (Ill) ' VJ: This is pathognomonic with us. (I) LDJ: Pathogno...what? (literally: pathogno...how?) (II) VJ: Pathognomonic, that is a sign which is very characterizing. (Ill)'

In this example, one can see that Β asks not (only) for the repetition of the word, but (also) for its explanation; in other words, she requests specifications both on the signifiant and the signifii. The two objects are decoded as such by A, who in the third turn of the triad, first clearly repeats the unknown word to B, then immediately offers its definition (introduced by the reformulation marker adicä 'that is'). 2.1.2. Bolinger (1957: 141-142) remarks that reclamatory questions are a 'gradient class' according to the amount of repetition called for; thus, to the general ones he adds the "special reclamatoiy questions", which call "for a repetition of only part of the stimulus utterance, that signalled by the interrogative word: - He did it fast. — He did it how? These questions usually call for the repetition of a word which was simply not heard or clearly understood. Some of them can have a second goal; in example (3), Β has heard the word clearly and knows its meaning, but she notices that it has been used in a deviant form and tries to suggest a correction or at least to make A to be aware of it. In spite of this, A repeats the word in the same dialectal form (vifleim), instead of the standard one vicleim ('Bethlehem nativity'), then a third participant proposes the synonym irozi in order to 'solve' the apparent misunderstanding between A and B: (3)

20A-199-LDJ: In copiläria dumneavoasträ veneau colindätorii la poartäl CJ: Veneau, veneau! Veneau cu Vifleimul. (I) eel (II) LDJ: Cu CJ: Cu Vifleimul. (Ill)

166

Laurentia Dascälu Jinga

GD: Cu Irozii. 'LDJ: In your childhood would carol singers come to your door? CJ: They did, they did! They came with the Vifleim. (I) LDJ: With what? (II) CJ: With the Vifleim. (Ill) GD: With the Irods.'

As one can see, the second goal of such a reclamatoiy question can not be decoded when one of the presuppositions on which it relies is not actualized: in our case the knowledge of the regional status of the word vifleim is not shared by B's interlocutors.

2.2. MTs related to the production In (4), A is evincing some difficulties (reflected in repetitions and hesitations) in trying to find the proper word, which makes Β intervene to help him by suggesting it; in the third turn, A repeats the suggested term and goes further; schematically: I Stimulus (A):.. ? II Reaction (Β): X III Response (A): X (4) TV 34 B-567-HRP: Pärerea mea este cä in secolul 20 biefii indivizi au fast confruntafi cu ο formula de statalitate färä precedent; acesta este statul totalitär, care a mobilizat impotriva simplului cetäfean mijloace pe care tehnologia modernd le punea la punct. Si asta este propaganda, ο forfä pe care n-avem voie s-o s-o s-o. [i] ( I ) jäm (II) IS: negli jäm, HRP: s-o negli aceasta este politia [... ] (III) 'HP: In my opinion, in the 20th century, the poor individuals were confronted with an unprecedented formula of statehood: this is the totalitarian state, which mustered against the mere citizen the means provided by modern technology. And this is propaganda, a power which we are not allowed to, to, to ... (I) IS: neglect (II) HP: to neglect, this is the police [... ] (III)'

The repetition for accepting a term suggested by the partner is a singular case in our corpus when it has the same intonation pattern as its 'original'. In example (4), Β suggests the word neglijäm with an intonation suitable with A's previous utterance (a slightly rising pitch, specific to enumeration), then, A accepts the term and repeats it with an unmodified intonation, as it is already congruent with his whole discourse. The function of accepting formulation (cf. Norrick 1987: 251; Bazzanella 1992: 446) is due to a cooperative behavior of the partner, who produces thereby an "eulogic" utterance

Metadiscursive Triads

167

(cf. Stati 1982: 51-53). Seemingly, when interlocutors agree on topics, they tend to use similar intonation contour types, as a kind of "convergence of intonation" (cf. Selting 1988: 320).

3. MTs oriented to the code

3.1. Triads in which the definition or explanation of a term is requested 3.1.1. MT calling for the definition of a word might be represented schematically: I Stimulus (A): X II Reaction (B): What does X mean? III Response (A): (X means) Y (5) 17A - 133 - LDJ: Tu ai stat la cäminl GD: La cäminulIOV. (I) LDJ: In Bucurefti? GD: Da, pe Calea §erban Vodä, 24. LDJ: Ce-n seam nä WVl (II) GD: 'Invalizi, orfani, väduve' (III) 'LDJ: Did you live in a hostel? GD: In the IOW hostel. (I) LDJ: In Bucharest? GD: Yes, on 24, Serban Vodä Street. LDJ: What does IOW mean? (II) GD: Invalids, orphans, widows. (Ill)'

The relations between the three turns are not symmetrical: turn III (definition) is placed immediately after II ('what-means' question), whereas II can be placed at some distance from its stimulus (I); therefore the question asking for definition of an unknown word can be delayed; moreover, it can even be completely absent (which actually happens often). on the contrary, after a 'what-means' question, the explicative answer becomes obligatory and urgent. The request (often interrupting A's previuos turn), causes a break moment, as if it "stops the clock within the temporal frame of conversation" (cf. Fleisher Feldman/Kalmar 1996: 82); since the dialogue can not go further if A does not solve the crisis by offering the required explanation, it becomes immediately necessary.

Laurentia Dascälu Jinga

168

We might use here Schegloff s (1972: 388) 'conditional relevance' notion: "by conditional relevance of one item on another we mean: given the first, the second is expectable"; in (5), turn II ('what-means* question) is conditionally relevant on the occurrence of turn III (A's definition), whereas II is free as against turn I. We could add that the conditional relevance of II on III "must be satisfied within a constraint of immediate juxtaposition; [...] an alternative suggested by Harvey Sacks would make reference to 'nextness' [...]" {Ibid.: 390). 3.1.2. In the following type, Β asks for an explication using an 'explicatory echo question' (Dumitrescu 1995) which repeats an item from A's previous turn, preceded by an interrogative word; e.g.: (6)

10 A - 288 - CM: Amßcut ghiveci umplut. (I) GA: C um 'ghi veci urn

plut'? (II) CM: Adicä: douä vinete umplule, doi dovlecei umplufi, douä roi'ii umplute, un cartof umplut, ardei... (Ill) 'CM: I've cooked a stuffed vegetable hotchpotch. (I) GA: How come 'stuffed vegetable hotchpotch'? (II) CM: That is: two stuffed eggplants, two stuffed marrows, two stuffed tomatoes,a stuffed potato, green pepper... (III)'.

As Reactions (= turns II) asking for definition or explication of a term (3.1.) are structured as wh-questions, they are characterized in Romanian by the falling interrogative pattern (cf. Dascälu Jinga 1979: 113-121; 1998: 243-246): pitch peak on the (stressed syllable of the) interrogative word, followed by an abrupt fall up to the end.

3.2. Triads occasioned by B ' s correcting a term o f A ' s previous turn It is well known that self correction is the preferred device in conversational sense (Schegloff/Jefferson/Sacks 1977). The explicit correction made by B, called 'other repair' is much less common, because it infringes the politeness rules (by rare exceptions like the exchanges teacher

-

pupil or native

- non native speaker); as Ionescu-

Ruxändoiu (1991: 135) notes, the 'contravention' in this case is due to the fact that the correction implies "contesting an aspect of the other speaker's competence". The otherrepair triad we refer to might be represented as follows: I

Stimulus (A): X

II Reaction (B): (you mean) Y or: III Response (A): Y

Y you mean

Meladiscursive Triads

169

(7)

4B - 142 - GD: Cu mit de bombe trägeau. Pentru nimic! (I) le LDJ: Ghiu le vrei sä spui. (II') GD: Bombe de-alea de... tunuri, nu? bu VJ: Ο ze. ( II") GD: Obuze. (III) 'GD: They shot thousands of bombs. For nothing! (I) LDJ: You mean cannon balls. (II') GD: Bombs, like those of... cannons, aren't they? VJ: Shells. (II") GD: Shells. (Ill)'

In (7), the term (ghiulele) proposed by Β (in II') to correct A's error {bombe) is not adopted by A; the correction is achieved only after a third speaker's suggestion (obuze, in turn II"), which is finally accepted and repeated by A (turn III). The 'you-mean' turn is usually pronounced with an emphatic intonation pattern, i.e. with a heavy stress on the correcting term (Dascälu Jinga 2002b: 108,121).

3.3. MT in which Β repeats a term in order to ask for its confirmation The basic turn (II) of these sequences, by which Β repeats a word from A's previous turn, is an elliptical echo question. As in most cases the confirmation is a repetition, this triad is characterized by a triple occurrence of the same term: I. Stimulus (A): X II. Reaction (B): X ? III. Response (A): X Usually, Β asks for confirmation to be sure of the correct reception of a word; he can also repeat a term from A's previous turn in order to express his/her surprise related to its deviant form, asking thereby for confirmation in this very form, but also a possible explication of a metalinguistic type (cf. Stati 1993: 369). In example (8), A uses on purpose, with a jocular intention, the word ouälele ('the egs') in a dialectal variant, instead of the standard one ouäle, then offers the explanation of this odd use: (8)

3 A - 018 - MV: Trebuie sä mä due sä vopsesc ouälele. (I) LDJ: Eu trebuie sä le fi cumpär. uälele? (II) AG: Ο MV: Asa se zice la Bran: "ouälele " sau "oorle ". (III) 'MV: I have to go to dye ouälele. (I)

170

Laurentia Dascälu Jinga

LDJ: I have to buy them just. AG: Ouälele ? (II) MV: This is what they call them at Bran: ouälele or oorle (III)'

The echo questions (including the ones asking for confirmation) have in Romanian an 'emphatic' rising interrogative intonation pattern, i.e. with a negative prominence (low pitch) on the stressed syllable of the word (cf. Dascälu 1975; 1985: 300).

4. Conclusions

Our findings point out the essentially interactive role of repetitions (=R). The great majority of them (especially the other-repetitions) are not identical to their original. We do not refer only to the common division between 'partial' and 'total' (or 'exact') Rs, but also to the occurrence of new pragmatic values; as Fleisher Feldman/Kalmar (1996:79) remark, "repetitions address sense rather than reference, pragmatics rather than semantics or syntax". In this process by which an R becomes a new utterance, its position in the dialogue frame is essential. In our opinion, to this factor, another important one is to be added: prosody, by one exception (Rs accepting a term suggested by the interlocutor), our examples of Rs are intonationally different or even opposed to their original. Repetition seems to be one of the favourite device of producing utterances (or exchanges) of a metalinguistic type; As Mönnink (1996, apud Bazzanella 1996: XIV) remarks, due to their monitoring functions, such Rs "should usually be qualified as highquality utterances and as strategies instead of disfluences". Most studies of conversation are oriented to a basic dyadic unit, i.e. one which is constituted of two successive turns forming an "adjacency pair" (cf. Schegloff/Sacks 1973: 295), from which the first turn determines the occurrence of the second; e.g. Question Answer, Offer - Acceptation, Appel - Reply, etc. The present paper is characterized by a different approach; it deals with cases which show "how a two-step succession of units of adjacency pairs can cumulate to the basic dialectic triad of dialogue" (cf. Fleisher Feldman/Kalmar 1996: 81). From these triadic units, we only dealt with 'metalinguistic triads' (MT), comprising various definitions, explications, confirmations, etc., offered by speaker A to the explicit solicitation of B. Some of our MTs show an asymmetrical relation between the three components:

Metadiscursive Triads -

171

turns II and III are closely connected; we mentioned here the 'conditional relevance' of II on III, meaning that the presence of the former causes the necessary and immediate occurrence of the latter;

-

the connection between turns I and II appears to be weaker than the relation between II and III, as B's choice to correct an A ' s error, to express doubt, surprise, etc., or to ask for explanation or confirmation (of a metalinguistic type or not) is strictly optional.

The examination of MTs suggests that the dialogue can be better understood "as a conceptual three-step process rather than a physical subdivision into smaller units" (Markova 1990, apud Fleisher Feldman/Kalmar 1996: 82), which enables us to see the message embedded in its linguistic and social contexts as being "both past and future oriented" {Ibid.). Investigating some examples of classroom exchanges, Tsui (1989:562) argues that "a potential three-part exchange [...] is a more adequate description of a basic unit of conversational organization than an adjacency pair", a hypothesis which has yet to be demonstrated more convincingly, in our opinion. Nevertheless, Jefferson's (1973: passim)

idea

that the general function of the third term is to acknowledge the outcome of the interaction seems to be valid for all of our examples, which entitles us to consider that a three-part unit is the most useful option for analyzing and understanding the functions of some metadiscursive sequences in conversation.

References

Bazzanella, C. (1992): Aspetti pragmatici della ripetizione dialogica. - In: G. Grobber (ed.) La linguistica pragmatica. Atti del XXIV Congresso della Societä di Linguistica Italiana, Milano, 4-6 Settembre 1990,433-454. Roma: Bulzoni. - (ed.) (1996): Repetition in Dialogue. - Tübingen: Niemeyer. - (1996): Introduction. - In: Bazzanella (ed.), VII-XVII. Bolinger, D.L. (1957): Interrogative Structures of American English (The direct question). Alabama: Alabama University Press. Dascälu Jinga, L. (1975): What are you asking about? (On the intonation of emphasis in yes-no questions). - Revue roumaine de linguistique 20, 477-480. - (1979): On the intonation of questions in Romanian: the rising-falling and the falling patterns. Revue roumaine de linguistique 24, 111-121. - (1985): Intrebärile "ecou" In limba romänä intona)ia lor ['Echo questions and their intonation in Romanian']. - Studii cercetäri lingvistice 36, 299-306. - (1996): Prosodic features of some non-lexical signals in Romanian. - Revue roumaine de linguistique 41, 11-31.

172

Laurentia Dascälu Jinga

- (1998): Intonation in Romanian. - In: D. Hirst, A. Di Crista (eds.) Intonation Systems. A survey of twenty languages, 239-260. Cambridge: Cambridge University Press. - and L. Vanelli (1996): Mi raccomando eh! A pragmatic and phonetic analysis of the Italian interjection eh.-Lingua e stile 31, 393-433. - (2002a): Corpus de Romäna Vorbitä (CORV). Ejantioane. - Bucure?ti: Oscar Print. - (2002b): Corectarea ji Autocorectarea in Conversa{ia Spontanä. - Bucure$ti: Editura Academiei Romane. Dumitrescu, D. (1995): On the syntactic structure and discourse function of multiple constituent repetitive and non-repetitive questions in Romanian. - In: C. Lupu, L. Renzi (eds.) Omaggio a Florica e Alexandru Niculescu. Vol. I: Linguistica, Etnografia, Storia Rumena, 86-114. Padova: Unipress. Fleisher Feldman, C. and D. Kalmar (1996): You can't step in the same river twice. Repair and repetition in dialogue. - In: C. Bazzanella (ed.), 78-89. Garvey, C. (1977): The contingent query: A dependent act in conversation. - In: M. Lewis, L. Rosenblum (eds.) Interaction, Conversation, and the Development of Language: The Origins of Behavior. Vol. V, 63-93. New York. Ionescu-Ruxändoiu, L. (1991): Naratiune §i dialog In proza romäneascä. Elemente de pragmaticä a textului literar. - Bucuresti: Editura Academiei Romane. Jefferson, G. (1973): A case of precision timing in ordinary conversation: overlapped tag-positioned address terms in closing sequences. - Semiotica 9,47-96. Mönnink, J. (1996): Repeating as monitoring. - In: Bazzanella (ed.). Norrick, N. R. (1987): Functions of repetition in conversation. - Text 7, 245-264. Poggi, I., C. Castelfranchi, D. Parisi (1981): Answers, replies and reactions. - In: H. Parret, M. Sbisä, J.Verschueren (eds.) Possibilities and Limitations of Pragmatics, S69-591. Amsterdam: Benjamins. Sabatini, F. (1985): L' "italiano dell'uso medio": una realtä tra le varietä linguistiche italiane. - In: G. Holtus, E. Radtke (eds.) Gesprochenes Italienisch in Geschichte und Gegenwart, 154-184. Tübingen: Gunter Narr. Schegloff, Ε. A. (1972^: Sequencing in conversational openings. - In: J. Laver, S. S. Hutcheson (eds.) Communication in Face-to-Face Interaction, 374-405. Harmondsworth: Penguin. - and H. Sacks (1973): Opening up closings. - Semiotica 8, 289-327. Schegloff, Ε. Α., G. Jefferson, Η. Sacks (1977): The preference for self-correction in the organization of repair in conversation. - Language 53, 361-382. Selting, M. (1988): The role of intonation in the organization of repair and problem handling sequences in conversation. - Journal of Pragmatics 12,293-322. Stati, S. (1982): II dialogo. Considerazioni di linguistica pragmatica. -Napoli: Liguori. - (1993): "Authors" and "speaker" of the utterance. - In: H. Löffler (ed.) Dialoganalyse IV. Referate der 4 Arbeitstagung. Basel 1992, vol.l, 367-374. Tübingen: Niemeyer. Tsui, Α. Β. Μ. (1989): Beyond the adjacency pair. -Language in Society 18, 545-564.

Marianne Doury

Argumentation et "mise en voix" Les discours quotidiens sur Γ immigration

1. Introduction

Les postulats methodologiques et les traditions de recherche, respectivement en pragmatique des interactions conversationnelles et en analyse du discours, conduisent les analystes k s'interesser prioritairement aux mecanismes dialogaux des lors qu'ils travaillent sur des corpus d'interactions en face-ä-face, et aux manifestations dialogiques lorsque les donnees sont dcrites. Pourtant, s'il est vrai que la presence d'elements dialogaux ä l'ecrit reste circonscrite ä des genres specifiques (dialogues theätraux ou romanesques, par exemple), les phenomenes dialogiques a l'oral peuvent se reveler extremement frequents dans certaines situations. Iis permettent aux locuteurs de solliciter des voix multiples autour desquelles ils construisent leur propre discours, en s'y opposant ou en s'en reclamant, ou en tissant avec elles des liens beaucoup plus complexes. Nous proposons ici une description du fonctionnement de ce type de mecanismes ä partir d'une des multiples formes que peut prendre le debat public sur l'lmmigration: les discussions Hees a l'lmmigration enregistrees dans un commerce de presse. Nous identifierons quelques unes des fonctions argumentatives que peuvent remplir les enonces rapportes dans un corpus d'oral spontane. Ce travail nous conduira a reflechir sur les differents niveaux d'apprehension des interactions argumentatives, et a insister sur la necessite de distinguer la structure actancielle de Γ argumentation, qui se joue entre un proposant, un opposant et, eventuellement, un tiers, cible de l'activite argumentative, et sa structure interactionnelle (revelee par l'analyse du cadre participatif de la discussion).

1

II s'agit de ce qui relive de la "dialogisation interne au discours d'un seul et meme locuteur, ou se tissent [...] des voix divergentes, voire contradictoires, imputables ä autant d'enonciateurs distincts" (Kerbrat-Orecchioni 1991: 121).

174

Marianne Doury

2. Argumentation et discours rapporte

2.1. Presentation des donnees Ce travail s'inscrit dans une recherche sur "L*argumentation dans l'espace public contemporain: le cas du debat sur l'immigration". 2 Notre contribution a ete de faire la place, ä cöte de donnees fortement institutionnelles ou mediatiques, ä des discours auxquels on serait tente d'appliquer les qualificatifs de "spontanes", "quotidiens", "ordinaires". Notre corpus est ainsi constitue de discussions ayant eu lieu fin 1997 dans un commerce de presse du vingtieme arrondissement parisien au lendemain des discussions au Parlement 3 4 des lois Guigou et Chevenement. Les donnees enregistrees ont έίέ partiellement retranscrites, en fonction de criteres thematiques: ont ete selectionnees les sequences consacrees au ddbat sur rimmigration tel qu'il s'est döroule sur la scene politique (lois Guigou et Chevenement, ou discussion de certains points juridiques independamment de toute allusion a ces lois: double nationalste), mais aussi de sequences portant sur des sujets divers, mais ä propos desquels les locuteurs mentionnent la nationalite reelle ou supposee ("un algerien", "un marocain"), l'origine geographique ("un asiatique") ou culturelle ("un arabe"), la couleur de peau ("un noir", "un black"), l'appartenance religieuse ("des musulmans", "des juifs") ou l'origine des parents ("un beur") des individus mis en scene.

2.2. Discours rapporte et traitement du discours adverse Vincent et Dubois soulignent Γ existence d'un lien privildgie entre argumentation et discours rapporte, affirmant que l'acte sous-jacent au discours rapporte est souvent argumentatif "dans la mesure oü il vient appuyer ou reproduire une these, ou encore declencher une discussion" (1997: 23). Dans le cadre de cet article, nous ne considererons que les cas oü un locuteur presente, dans son propre discours, un discours adverse pour le refuter. La 2

3

4

Laboratoire "Analyse de Corpus Linguistiques, Usages, Traitements" (ENS Fontenay-Saint Cloud), Groupe de Recherche sur les Interactions Communicatives (CNRS, Universitö LumiereLyon 2). Le caractere ponctuel de la collecte de donn6es interdit toute generalisation, en particulier aux "interactions dans les commerces de presse", ou au "debat sur rimmigration dans les discussions ordinaires": les analyses presentees ici ne concernent qu'une des manifestations ponctuelles possibles de ce d6bat, qui prend sens moins par elle-meme que par contraste avec d'autres realisations, au Parlement, ä la television, etc. Les enregistrements ont ete realises a micro cache, afin de preserver la spontaneitd des discussions, mais les clients, des leur sortie du commerce, ont 6te informes de la realisation d'un enregistrement, et ont signe une autorisation ecrite.

Argumentation

et "miseen

voix"

175

convocation du contre-discours ainsi que sa refutation peuvent emprunter differentes modalites. Le cas le plus attendu, mais qui n'est etonnamment pas le plus frequent, est Ie report d'une position antagoniste, suivi de la refutation de cette position: L: X dit que p. Or, q. Done, non p.

La source de cette position antagoniste est souvent non identifiee, designee par "ils": 5 [Une cliente, F2, repond au commerfant C, qui justiße le contraste entre le projet de loi Chevenement et les attentes des electeurs de gauche en invoquant "la realite du pouvoir":] F2: oui: mais enfin la realite du pouvoir e'est quoi e'est qu'i'z'ont la trouille pasqu'i va y avoir des elections cantonales et machin [ (.) et qu'i s'disent (.) et qu'i s'disent (.) a C: [ regionales qui vont arriver ben oui (.) ben oui F2: machin i faut quand meme plaire un peu ä l'electorat d'droite pasque sinon euh:: (.) i faut arreter e'est nul (.) e'est nul (.) e'est pas en on sait tres bien qu'e'est pas en faisant une politique comme 9a qu'i s'en sortent

Dans la mesure oil l'adversaire n'est pas rapporte έ des individus precis, mais fondu dans un collectif ("les socialistes au pouvoir"), et qu'il ne s'agit pas d'une veritable citation mais de l'attribution d'un raisonnement interieur ("i s'disent"), l'enonce rapporte prepare le terrain pour l'utilisation eventuelle d'un proces d'intention ou de la technique de l'homme de paille. De plus, meme si la forme employee est celle du discours direct, il ne fait aucun doute que les termes utilises sont du fait du locuteur rapportant. En effet, la presentation de la position adverse est en elle-meme polemique: l'emploi de "machin" dans Ie discours direct, qui fait echo au "machin" de l'enonce precedent, designe les considerations electoralistes du pouvoir comme

illegitimes ("machin" etant une fa?on bien

peu

valorisante de condenser les reflexions du gouvernement), et previsibles (puisqu'elle laisse le soin au destinataire de "remplir" le discours; dans le meme ordre d'idee, on notera le caractere elliptique de la justification du raisonnement: "pasque sinon euh", qui appelle une suite du type "on va perdre les elections"). La locutrice articule sa position ä revocation de la position adverse en deux temps:

5

Dans les extraits proposes, C designe le commeriant, Η un client, F une cliente. / indique une intonation montante, (.) (..) (...) des pauses intra-repliques relativement breves; deux signes [ superposes indiquent le debut d'un chevauchement. Les elisions sont notees par une apostrophe; X X X signale une production verbale inintelligible. Des majuscules dans la transcription notent l'emphase, un tiret - un mot interrompu.

6

La Strategie de l'homme de paille, ou constructing a strawman, est definie ainsi par Engel (1994): "the main purpose is generally to impute to adversaries opinions a good deal more extreme than those they have set out and are willing to defend." (p. 120)

176

Marianne Doury

1.

evaluation de cette position (sensible, on vient de le voir, ä travers la formulation meme de la thfese adverse, puis explicitee par le jugement definitif et reitere "c'est nul"), qui tient lieu d'expression d'un desaccord; 2. utilisation d'un argument pragmatique ("on sait tres bien qu'c'est pas en faisant une politique comme ?a qu'i s'en sortent"), renforce par la suggestion qu'il fait l'objet d'un consensus ("on sait tres bien"), suggestion qui fait peser le discredit ceux qui adherent au raisonnement incrimine. Ce precede (qui consiste a convoquer le contre-discours et ä le refuter dans le temps meme de l'interaction) est relativement marginal par rapport au precede qui consiste, pour le locuteur, ä evoquer le contre-discours dans le sein d'une situation de parole, reelle ou fictive (Sl), distincte du moment de l'interaction presente chez le marchand de presse (SO), et ä integrer sa rdfutation comme un element de cette situation de discours (Sl). Autrement dit, le locuteur met en scene l'affrontement entre son propre discours et un contre-discours dans le cadre d'un dialogue rapporte presente comme ayant eu lieu, ou comme susceptible d'avoir lieu. La sequence de discours rapporte devient alors plus specifiquement une sequence d'echanges rapportes, et obeit, au moins partiellement, & une logique narrative, puisque "des que l'on rapporte une interaction, on fait un rdcit" (cf. Bres 1997: 130). a) Dialogues preserttes comme ayant eu lieu: C: (.) moi des fois j'discute je- par provocation comme ?a pour faire parier les gens (.) qui aiment pas les juifs (.) (0) y en a un i m'dit oui:: mais i i z'ont tout (1) (.) mais et alors t'as qu'ä bosser (2) (..) i z'ont tout (3) (..) tu vois i magouillent (4) mais t'as qu'ä faire pareil (5) (..) mais c'est pas une raison d'etre (6) (..) tu vois c'est (.) mais ?a c'est primaire (7) moi j'connais ce gars et j'suis vachement dtonnd (..) intellectuellement c'est quelqu'un: qui:: (..) mais (.) mais il est (.) pour moi c'est du (.) j'vais (8) un jour j'lui dis ecoute si tu paries comme 9a c'est pas la peine de m'dire bonjour (9) (.) je t'le montrerai (10)

II s'agit d'un dialogue presente comme un cas particulier d'une serie d'interactions ("moi des fois j'discute"); l'existence du client antisemite est attestee par l'engagement qu'il prend ä le montrer ä H3 (il est meme nomme dans la suite de l'echange). On peut distinguer dans l'intervention de C deux situations de discours: 1. Sl (dialogue anterieur entre C et son client antisemite), ou se confrontent les propos attribuables au client antisemite (1/3/4) et les propos attribuables ä C (2/5/6/9);

7

L'argument pragmatique est defini par Perelman comme "un argument des consiquences qui apprecie un acte, un evinement, une regle ou toute autre chose, en fonction de ses consequences favorables ou defavorables" (1989 p. 19).

Argumentation et "mise en voix " 2.

177

SO (interaction presente entre C et son client et ami H3), oü se succedent les enonces d'introduction (0), devaluation du contre-discours (7), d'identification du tenant du contre-discours (8/10).

Les sequences introductrices etant reduites a portion congrue, d'autres indices permettent de preter les enonces a C ou ä son interlocuteur: 1. l'attribution du premier tour par un syntagme introducteur ("y en a un i m'dit"); - 8 2. le contenu des enonces, mais aussi 9 3. le role demarcateur de mais\ 4. l'intonation (en suspension apres les interventions de Γ interlocuteur, conclusive apres les interventions de C), ce qui est un indice de 5.

la structuration des tours de parole en S1: le contre-discours a une fonction initiative, le discours assume par C a une fonction reactive (l'interlocuteur deraisonne, C le ramene ä la raison), le commer?ant s'attribuant toujours le dernier mot.

Cette derniere remarque illustre le potentiel argumentatif de ce type de dialogue rapporte. Les locuteurs se livrent ä un important travail de textualisation (cf. Bres 1997: 134), qui reflete le soin qu'ils portent ä "menager des effets", et qui se traduit notamment par le fait que celui qui rapporte une interaction, reelle ou fictive, ä laquelle il participe, se donne souvent le "beau röle" dans la discussion, se prete des effets de manche, et tend en particulier ä se reserver un volume de parole superieur, et ä se conferer le dernier mot (Vincent et Dubois 1997). La ligne argumentative prise en charge par le commer9ant se deroule sur deux plans: 1.

celui de l'interaction rapportee, au sein de laquelle le commer9ant se represente avantageusement dans une joute verbale (voir par exemple l'image de pourfendeur intransigeant de l'antisemitisme que C construit en rapportant ses propres paroles: "si tu paries comme 9a c'est pas la peine de m'dire bonjour"),

2.

et celui de l'interaction en cours, qui lui permet de donner son sens argumentatif ä l'interaction rapportee, et a exprimer des jugements difficilement integrables a la situation de dialogue evoquee ("9a c'est primaire").

8

9

La pertinence du critere du contenu des enonces est largement li6e aux connaissances extralinguistiques que l'analyste ou le rdcepteur a des locuteurs en presence (par exemple, le fait que des propos comme "les juifs ont tout" sont difficilement attribuables ä C), mais aussi ä des mecanismes de coherence argumentative (l'enonce "9a c'est primaire", clairement pris en charge par C, est cooriente avec 2, 5, 6 et 9, mais pas avec 1, 3 et 4). Ce critere fait sens articule avec d'autres, mais la polyvalence de mais ne peut en faire un critere decisif pour marquer de fapon univoque l'alternance des repliques.

Marianne

178

Doury

b) dialogues fictifs präsentes comme prototypiques: C'est le meme type de mecanisme qu'on peut observer dans l'echange suivant, portant sur la double nationalste. Le dialogue mis en scene ici est presente comme une synthese de debats anterieurs ("moi j'vois des mecs"), la quintessence des dialogues possibles sur ce theme, voire la reminiscence d'une discussion anterieure, qui peut effectivement avoir eu lieu, mais n'est rattachee ä aucune situation de communication identifiee: C:

H3: C: H3: C: H3: C: H3: C:

moi j'suis pas tu s'ras pas d'accord avec moi (.) c'est que (.) si t'as des gamins (...) bon (.) la la double nationalite y a un risque (...) t'es franfais ou t'es algerien (.) t'es fran^ais ou t'es marocain i faut savoir moi j'suis moi j'suis pas pour la double nationalite moi j'suis contre ouais moi j'vois [ des mecs [ tu fais ton choix hein: tu fais ton choix (..) dans ces cas-lä (.) c'est trop facile [ (.) alors [ ah ouais (.) tu prends le bon d'un cote:: [ et puis tu laisses le [ bon de l'autre [ voi:lä [ exa:ct (..) oui: mais on a nos origines mais et alors tu peux avoir tes origines/ (.) tu peux aller en vacances dans ton pays (.) mais dans ces cas-lä t'as pas ä avoir (.) et c'est et leurs gamins (.) les aines qui font des conneries comme ?a (...) tu les prends (..) casse-toi dans ton pays (..) pasque c'est des parasites de la societe

II apparait tres rapidement que l'apparition d'une discussion argumentee entre les personnes en presence est peu probable, C et H3 campant sur la meme position ("je ne suis pas pour'V'je suis contre"). C'est alors a travers la mise en scene d'une discussion avec un interlocuteur "immigre defendant la double nationalite" que C peut developper ses arguments, avec d'ailleurs une violence difficile a tenir dans une interaction en face-ä-face. On peut distinguer trois phases dans cette sequence: 1.

une phase introductive durant laquelle C introduit le theme de la double nationalite (avec un tu generalisant), "täte le terrain", affirme sa position et s'assure de celle de H3 (avec un tu qui renvoie ä H3).

2.

Une phase de transition, au cours de laquelle C et H3 utilisent un tu plus ou moins generalisant (peut inclure le locuteur dans "tu fais ton choix", designe de fagon plus restrictive ceux qui defendent pour de mauvaises raisons la double nationalite dans "tu prends le bon d'un cote et tu laisses le bon de l'autre").

3.

Une troisieme phase d'opposition dialoguee, l'enonce "oui: mais on a nos origines" etant attribue ä l'adversaire, et les enonces suivants etant pris en charge par C en SI, le tutoiement renvoyant ä cet adversaire fantomatique. La fin de la phase 3 fait apparaitre ä nouveau une alternance des enonces relevant de C en SI, et de ceux relevant de C en SO.

Argumentation et "mise en voix "

179

c) Enfin, on rencontre ce qu'on pourrait appeler des dialogues hypothitiques, ä travers lesquels le locuteur imagine ce qu'aurait pu donner un affrontement avec tel ou tel interlocuteur. Ainsi, dans l'exemple suivant, H8 se plaint de la concentration d'immigres dans certains quartiere parisiens, et denonce Γ attitude de certaines communes chics de la banlieue parisienne, qui s'en debarrassent: H8: et encore comme Sarkozy l'maire de Neuilly (..) i devrait meme pas parier c'mec-lä (...) mais i devrait pas parier de-j'ai vu Sarkozy i discutait d'vant Hue moi le: gars du Parti Communis' (..) mais Hue il avait pas d'couilles au cul moi j'y aurais dit pas qu'est-ce que tu m'emme:rdes toi (..) toi t'es ä Neuilly toi (..) t'as pas d'difficulte: (..) et viens dans l'coin (..) et tu les prendras (...) alors forcement alors i z'ont toujours la pretention d'vous dire ben vous savez j'paye des taxes (.) non (.) tu GA:::rdes ton pognon (.) et tu LES prends (...) et c'est tout:

L'attribution des tours de parole aux uns et aux autres se fait grace aux enonces introducteurs ("moi j ' y aurais dit", "i z'ont toujours la pretention d'vous dire"), et grace ä l'asymetrie des pronoms personnels (tutoiement de H8, vouvoiement des hommes politiques). On notera cette fois encore la violence de ces simulacres d'affrontement, qui permettent aux locuteurs de se construire un ethos sur mesure (ici, celui d'un adversaire qui a "des couilies au cul") sans avoir a payer le prix d'une confrontation reelle. Cet exemple de dialogue hypothetique, rapporte ä une interaction identifiable (entre Hue et Sarkozy), et impliquant le locuteur H8 et Nicolas Sarkozy, tend en fait vers le dialogue presente comme prototypique: si l'interlocuteur est bien le maire de Neuilly, le contre-discours evoque est attribue ä un ils renvoyant plus largement aux maires de communes riches, et est presente comme une generalisation ä partir de multiples enonciations ("i z'ont toujours la pretention d'vous dire"). Enfin, dans cet exemple comme dans les exemples precedents, la voix de l'adversaire est reduite ä sa plus simple expression, sa contribution ä Γ interaction rapportee se limitant i une, rarement deux, interventions.

3. Pour une definition de Γ argumentation comme mecanisme de resistance ä la contestation

Les phenomenes de citation au sens large sont particulierement presents dans notre corpus, et ce, pour differentes raisons, largement liees au choix du site. Pourtant, tout specifiques que soient les interactions etudiees, les observations degagees plus haut amenent ä questionner la definition de l'argumentation comme moyen de persuasion - questionnement qui nous semble necessaire pour tout type de discours argumente.

180

Marianne Doury

3.1. L'argumentation a-t-elle pour fin de construire un discours persuasif? La definition de Γ argumentation comme moyen de persuader un auditoire est predominante dans les recherches qui inscrivent l'etude de l'argumentation dans le champ de la rhitorique ou des sciences de la communication. D'Aristote, qui definit la rhetorique comme "la faculte de decouvrir speculativement ce qui, dans chaque cas, peut etre propre ä persuader" (Rhetorique: 1355b), ä Perelman, pour qui la theorie de l'argumentation traite des "techniques discursives permettant d'accroitre I'adhesion des esprits aux theses qu'on presente ä leur assentiment" (1988: 5), il est classique d'etablir un lien entre argumentation et persuasion. Pourtant, si cette conception de l'argumentation presente l'avantage de l'evidence intuitive, eile ne va pas sans poser problemes. 1.

Elle met sur le devant de la scene la question des effets de l'argumentation: si sa raison d'etre est de persuader, dans quelle mesure y reussit-elle? Cette question fait sortir l'argumentation du champ des sciences du langage pour la confier ä la Psychologie sociale, discipline qui a forge un certain nombre d'outils susceptibles de permettre une evaluation des changements d'attitudes d'individus exposes ä divers stimuli - dont des stimuli argumentatifs. Et le bilan que tire Billig (1987) de ce type de recherches est pour le moins mesure: les recherches experimentales en psychologie sociale, qui cherchaient idealement ä decrire les processus de persuasion par des lois du type "Le stimulus X produit la reponse Y", se limitent aujourd'hui ä identifier des phenomenes de plus en plus modestes, et ä multiplier les qualifications prevoyant des exceptions aux lois ainsi degagees.

2.

Par ailleurs, il n'est pas possible d'attribuer une finalite unique (ni meme principale) ä l'activite argumentative en elle-meme; si but il y a, celui-ci est attache a une interaction (ou ä un type d'interaction) particuliere, et non ä l'argumentation "en general". Cette remarque est fortement etayee par les donnees analysees ici: il y a indeniablement argumentation dans les sequences presentees precedemment. II serait pourtant bien audacieux de poser qu'en argumentant, les locuteurs cherchent ä se persuader mutuellement (puisqu'ils sont souvent sur les memes positions des le depart), ni meme ä se persuader eux-memes de la justesse de leur point de vue. S'il fallait lui assigner des finalites, on pourrait penser que l'argumentation, dans ce cadre, remplit une fonction cognitive (elle aide les locuteurs ä faire le tour de leurs opinions en les verbalisant, puis en les confrontant in absentia ä des contre-discours), une fonction relationnelle (elle renforce la connivence avec l'interlocuteur en rendant explicite la convergence des points de vue), une fonction "identitaire" (elle consolide le sentiment d'appartenance au groupe des "gens de gauche, pas racistes mais justes", contre "les antiracistes nai'fs" aussi bien que contre "les mechants racistes de droite"). Ces fonc-

Argumentation et "miseen voix"

181

tions sont, on le voit, fortement Hees ä des situations de communications particulieres, et peuvent difficilement etre rattachees ä l'activite argumentative en elle-meme.

3.2. L'argumentation comme confrontation de discours contradictoires Α cöte des definitions de 1'argumentation centrees sur l'idee de persuasion, on trouve une autre famille de definitions, basees sur ce qui est generalement pose comme condition minimale pour l'emergence d'une argumentation: l'existence - ou la plausibilite - d'une divergence d'opinion sur un sujet. L'insistance mise sur l'opposition, dans l'argumentation, d'un discours et d'un contre-discours, se rattache ä une tradition ancienne, dont l'origine est classiquement attribuee έ Protagoras, qui place au centre de sa reflexion l'idee que sur tout sujet, il y a au moins deux discours possibles (de Romilly 1988). Aujourd'hui, nombreux sont les auteurs qui insistent sur le caractere central, dans toute argumentation, de la confrontation entre un discours et un contre-discours. On peut alors considerer l'argumentation comme un mode de construction du discours visant ä le rendreplus resistant ä la contestation. Quelques precisions, afin de bloquer les inferences indesirables auxquelles une telle definition pourrait inviter: 1. nous ne souscrivons pas ä la vision, largement repandue, de l'argumentation comme activite belliqueuse. Comme le souligne Plantin (1996: 72), la confrontation entre discours et contre-discours peut se faire sur un mode aussi bien conflictuel que cooperatif. 2.

le discours argumente n'est pas un discours clos; argumenter suppose que Ton cherche non pas ä se soustraire ä la contradiction, mais ä voir jusqu'ou un point de vue peut "tenir" face ä la contestation. Argumenter - meme si ce n'est pas necessairement l'objectif de tout locuteur tenant un discours argumente - permet de developper un point de vue au-dela de l'adhesion immediate, d'en expliciter les implications, d'en mesurer la coherence.

Enfin, au niveau methodologique, la definition de l'argumentation comme mode de construction d'un discours visant ä le rendre plus resistant a la contestation invite ä porter une attention particuliere aux modalites prises par la confrontation du discours et du contre-discours dans un corpus donne. Elle fait echo aux definitions du discours argumentatif comme dialogue, reel ou virtuel, entre les tenants de positions divergentes. Si cette conception de l'argumentation met en relief le caractere central, dans l'argumentation, de l'articulation d'un discours ä un contre-discours, la reference au dialogue reste metaphorique (Reed 1999), et il faut distinguer nettement, dans l'analyse, les acteurs de la communi-

Marianne Doury

182

cation argumentative, qui sont "les individus concrets engages dans la communication", et les actants de la communication argumentative (les positions de Proposant, Opposant, et, eventuellement, le Tiers indecis) (Plantin 2000). Une telle distinction amene a considerer tour ä tour: 1.

les mecanismes discursifs qui permettent Γ articulation du discours au contre-discours pour un meme locuteur;

2.

le caractere monologal ou dialogal du discours argumentatif etudie, et la fason dont les locuteurs engages dans l'argumentation occupent ou parcourent les differents roles actanciels

argumentatifs.

Ces deux niveaux ne sont bien sür pas independants: les donnees analysees ici montrent que les echanges rapportes apparaissent dans des phases oü Γ interaction permet ä des unites monologales (interventions) de taille consequente de se developper. Pourtant, opposer d'une part un discours argumentatif monologal (oral ou ecrit) fortement polyphonique ä des interactions argumentees constituees d'une alternance d'interventions monologiques courtes ne tiendrait pas compte de quantite d'autres situations de discours possibles, corame le suggere le corpus analyse ici. La description attentive des modes d'articulation du discours et du contre-discours ne peut tenir lieu d'analyse argumentative ä eile seule; mais elle en constitue une part importante, souvent negligee au profit de l'identification immediate des types d'arguments presents dans les donnees.

References bibliographiques

Aristote (1932): Rhetorique, t.l. - Paris: Les Belles lettres. Billig, Μ. (1987): Arguing and Thinking. A rhetorical approach to social psychology. - Cambridge/ Paris: Cambridge University Press/ Editions de la maison des sciences de l'homme. Bres, J. (1997): De l'interaction rapportee dans le recit oral. - Modeles linguistiques vol. 35, XVIII1,129-140. Doury, M. (1997): Le Debat immobile. - Paris: Kime. Engel, S. M. (1994): Fallacies and Pitfalls of Language. The language trap. - New York: Dover Publications. Kerbrat-Orecchioni, C. (1991): H6teK^neite enonciative et conversation. - In: H. Parret (ed.) Le sens et ses heterogeneites, 121-138. Paris: Editions du CNRS. Perelman, C. (1989): Rhetoriques. - Bruxelles: Editions de l'Universite de Bruxelles. - et L. Olbrechts-Tyteca (1988): Traite de l'argumentation. - La Nouvelle Rhdtorique. Bruxelles: Editions de l'Universiti de Bruxelles. Plantin, C. (1996): L'argumentation. - Paris: Seuil (Memo).

Argumentation et "mise en voix"

183

- (2000): La "question disputee", une forme argumentative oubliee. Thomas d'Aquin, La guerre peut-elle etre juste? (Papier de travail). Reed, C. (1999): Is it a monologue, a dialogue or a turn in a dialogue? - In: F. van Eemeren, R. Grootendorst, J. A. Blair, C. A. Willard (eds.) Proceedings of the Fourth International Conference of the International Society for the Study of Argumentation (June 16-19 1998), 681-685. Amsterdam: SICSAT. Romilly, J. de (1988): Les grands sophistes dans l'Athenes de Pericles. - Paris: Bernard de Fallois. Vincent, D. et S. Dubois (1997): Le discours rapporte au quotidien. - Quebec: Nuit Blanche Editeur.

Paula Fatur-Santos Interacting with a Frontal Lobe Patient: What about Pragmatics?*

1. Introduction

The traditional view that only some evident neurological diseases restricted to language areas such as aphasia affect communication has been lately challenged as emerging research in Clinical Neurolinguistics and Neuropsychology of Language

pinpoints

communication impairments in a wide range of neurogenic disorders, such as Alzheimer's dementia (Caramelli, Mansur, Nitrini 1998), Parkinson's disease (Cohen 1998), psychosis (Ribeiro 1994, Tracy 1998), Landau-Kleffner Syndrome (Denes 1998), D o w n ' s Syndrome, Williams' Syndrome, Turner's Syndrome (Sarkari, Tan and Molfese 1998), Autism (Happe' 1998, Perkins 1999). As for Frontal Lobe Syndrome (FLS), while many clinical group studies exist in the area, there is a growing call for studies in a qualitative manner, particularly in interactional contexts, as pointed out by Ehrlich (1988: 2), "Almost no data exist on the discursive abilities of the traumatically head injured". The aim of this singlecase study is to look at the pragmatic features of the interaction with a FLS patient and correlate them with a neuropsychological account of the problem.

2. Patient's history and diagnosis

The patient, A.S., is a 45 year-old male, with 4th grade schooling, who was diagnosed with a frontal lobe damage as a result of a TBI, due to a severe fall. The patient was found unconscious at the time of the trauma. According to his neurologist, he was diagnosed with a frontal bilateral subdural hematome reabsorbed. He remained temporo-spatial disoriented for two months, with psychomotor agitation, and amnesic for events. The amnesia is still ongoing, presently combined with a depressive state. He had a previous history of alcoholism.

*

Acknowledgments: I would like to thank Dr. A. Walkoff, from Hospital Florianopolis and Cefaloclinica, as well as the patient's family, for allowing me to study the case.

186

Paula Fatur-Santos

As clinically observed, he shows a lack of initiative in various abilities: motor (he is slow to move, has gait apraxia), communicative (he has to be prompted to talk, speaks little and with reduced utterances, in low pitch and same tone, i.e. isophonic), cognitive (is not self-motivated so as to engage in any cognitive activity such as reading, writing, learning, playing a game) in the absence of any motor or linguistic impairments per se. There is evidence that he remembers past events previous to the onset of the trauma: when he is asked about past events, he is able to recall them, however constrained by the communication problems resulting from the pathology. As to recent events, being amnesic, information is limited to short-term storage. Observations from collateral source (sister) are that A.S. asks the same questions repeatedly because he forgets information, and is not self-motivated to carry out daily activities, such as engaging in any motor, cognitive or communicative action. Besides, his sister points out that memory problems are not only linked to verbal information, but also to spatial orientation: if he goes out by himself, he is unable to find his way back home.

3. Data collection

The data was collected by means of an audio-taped conversation in Brazilian Portuguese (BP), with open-end questions directed to the patient's previous and present life. The conversation was conducted by the researcher at the patient's home, in Florianopolis, Brazil, in the presence of his sister, who takes care of him at the moment, and his neurologist. They did not participate in the conversation, having side talks during the recording.

Interacting with a Frontal Lobe Patient: What about Pragmatics?

187

4. A neuropsychological perspective

4.1. Frontal Lobe Syndrome The frontal lobe is a very important and complex part of the human brain, and one of the reasons for this is its interconnection with the rest of the brain, more specifically, with the posterior parietal cortex, prestriate and temporal cortex. These regions, in turn, are involved in long-term memory of somatosensory, visual and auditory information. Besides, it is connected to the hypothalamus and amygdala, which regulate emotion; it is also connected with the precentral cortex and neostriatum, which are linked to motor systems (Milner and Petrides 1984). Thus the observed problems in memory, and motivation in several domains: bilateral subcortical damage to frontal lobe has been correlated with reduced verbal fluency both in free speech and formal verbal fluency tasks, shorter phrases and perseverations in speech (Cherrier, Mendez, Cummings and Benson 1998). Recent developments regarding the so called "frontal functions" show that the frontal lobe is associated with communication, attention, memory/learning, cognitive/executive functions, besides personality, emotion and self-awareness (Beaumont, Kenealy and Rogers 1996). Selective impairment of these functions depend on the size and locus of the trauma. A.S. exhibits the following impairments linked to frontal lobe functions. 4.1.1. Communication In terms of communication, this patient exhibits what has been described in the literature as a reduced spontaneous initiation of speech, which has been correlated to lesions in the supplementary motor area. This deficit, in turn, is more of an output disorder than any linguistic deficit per se, in the sense that there is no deficit in his procedural linguistic competence, subserved by the implicit memory system, the system responsible for grammatical competence in the first language (Paradis 1994, 1998). A prime example of a procedural linguistic competence deficit would be an aphasia, which is not the case here. 4.1.2. Attention As a FLS patient, A.S. also exhibits hemi-attention, which has been correlated to a primary deficit in exploratory motor behavior, that is, the inability to suppress attention to other stimuli. Other attentional deficits that have been reported are reduced reaction to new stimuli, habituation to repetitive novel stimuli and deficits in keeping attention (distractibility, neglect).

188

Paula

Fatur-Santos

4.1.3. Memory/Learning It has been suggested by cognitive psychologists (Baddeley 1986) that the frontal lobes have a role in working memory (WM), which is controlled by a central executive system. WM acts as a buffer that maintains information temporarily. McDonald (1998) suggests that Traumatic Brain Injury patients, while frontal lobe patients, have a dysfunction of the "executive control" of cognition, which results, among other things, in stimulus-bound behavior. The case of A.S. seems to support that hypothesis. Concerning long-term memory, however, the issue is not so straightforward. There are divergent theories and results among researchers. The very issue of correlating frontal lobe functions with sites/extension of lesions is a complex discussion beyond the scope of this paper. Concerning learning, FL patients have been reported to have problems with temporal order of information {when) and it has been speculated that also with spatial orientation (where), and interestingly, the problems seem to be related to ordering and perhaps orientation, but not to recall of information itself. Overall, this has been reported in the neuropsychological literature as difficulty with "contexts related to specific information" (cf. Beaumont, Kenealy and Rogers 1996: 350). 4.1.4. Cognitive/executive functions Frontal lesions can impair: (i) generation and fluency of output, including word and gesture fluency; (ii) formulation of hypotheses; (iii) the shifting of strategies of established responses; (iv) the programming and selecting of subgoals of responses. Impulsive responses have also been attributed to frontal damaged patients. 4.1.5. Personality, emotion and self-awareness Once again, apathy is a feature that seems to be linked not only to linguistic and cognitive aspects, but also to personality of frontal lobe patients in general. Thus the difficulty of A.S. in initiating behaviors, the social withdrawal, the isophonic and slow speech, in which affective features are absent, and the inability to get involved in actions and to reflect upon them (self-awareness). As to affect, it has been noted that the ability to understand emotion might be dissociated from the ability to express it (Damasio 1994).

Interacting with a Frontal Lobe Patient: What about

Pragmatics?

189

4.2. Frontal syndrome and Traumatic Brain Injury (TBI) McDonald (1998: 485) proposes that TBI often results in psychosocial and sensorimotor residual impairments. This is in tune with this case, in which A.S. shows social withdrawal and gait apraxia, besides slow movement in general. Concerning communicative aspects impaired in FLS (production of socially appropriate discourse and keeping its macrostructure; pragmatic inference) McDonald {Ibid.) poses that they might be related to the following so called executive functions: stimulus-bound behavior, abstraction skills, planning and monitoring, disinhibition, verbal fluency, learning and memory, including WM. More specifically, she notes that one type of TBI patient does not fulfill the needs of the conversational partner, the conversation is slow and the responses are often incomplete, with pauses. Besides, and probably the most striking communicative feature, is that these patients seem not to provide enough information about the topic, what she calls efficiency of discourse, as measured by (1) rate of speech and (2) amount of information within the utterances. Both were found to be shorter in TBI patients. This seems to be confirmed in the case of A.S.' speech.

4.3. Perseveration Perseveration of previous responses occur in FLS, and it has been credited to the patient's not being able to inhibit certain responses when the context does not require a response or in commands where no response is required. Perseveration has been studied in movement, cognitive tasks, such as drawing and in language. Here we find a clear example of linguistic perseveration, namely, the repetition of the same lexical items and sometimes whole sentences by A.S. Brown and Chobor (1989), reviewing the work of Luria on the frontal lobes, point out that such patients may have a dissociation between a memory store and its motor realization, since they might have normal retention but impaired retrieval. They argue that although this might suggest a memory disorder, in fact their memory for events is intact but there is random association in recall, loss of contextual facilitation or a lack of inhibition. That means that there is an inability to select the correct route and inhibit incorrect ones. This notion of dis/inhibition was first proposed by H. Jackson (1894, cited in Brown and Chobor (1989: 69) who described perseveration as a "release from cortical inhibition".

190

Paula Fatur-Santos

5. A discursive perspective

5.1. Interactional and conversational style Because of the reduced initiative for talking spontaneously, typical of the FL syndrome, during most of the conversation, A.S. had to be motivated and prompted by the speaker in order to generate utterances. It seems that it was an enormous effort to talk. However, this effort was definitely not related to production of speech in the sense of forming grammatical sentences, since his procedural linguistic competence is intact. The effort seems to lie in the microgenesis of the utterances, i.e. the very motivation to speak. Besides the prompting, another characteristic of the conversation is that A.S.' utterances are very short, showing no elaboration of the topic. He speaks slowly, with low pitch, almost always with the same tone, pauses, and takes time to answer questions. Those features indicate that his discourse does not follow the typical Brazilian conversational style. Conversational style is socially shared, and speech ethnographers such as Garcez (1996) have described Brazilian interaction as dependent on the fact that there is high conversational involvement in the interaction. As a result, Brazilian speakers expect a lot from the hearer in order to create involvement. The hearer, in turn, fills that expectation by giving much more information than required, such as background information, thus creating involvement. Expecting a lot from the hearer is also a way of creating involvement by making the hearer participate in sense-making creation. Here, due to the pathological constraints, the hearer does not provide more information than required, rather, provides the exact amount and quality required or no new information at all, leaving no room for involvement. As a result, the conversation resembles a medical interview. By comparing this patient's conversation with that of other patients that do not show a frontal damage one can immediately perceive that he is far behind the baseline in terms of filling expectations when compared to a typical interaction in his culture.

5.2. Speech reduction One way in which A.S.' speech is reduced is by means of using yes/no answers. BP allows null subject clauses, since it is a pro-drop language (Haegeman 1994). Therefore, it is very common to use only the verb, which is marked for person in BP, without sim (yes) in positive responses. Thus, A.S. reduces his answers to the verb when answering positively most probably because the verb appears in the question right before and thus becomes a perseverating response. Besides, using sim involves beginning a novel response, which

Interacting with a Frontal Lobe Patient: What about Pragmatics?

191

seems to be a problem for him. This is very common throughout the entire conversation, as shown in the example below 1 : 033

I: Ta' em casa?

034

Ρ: 4-Τδ.

I: Are you staying home? Ρ:

lam.

5.3. Repetition Repetition is a keyword in A.S.' interaction. Within the field of dialogue analysis, repetition has been studied as a complex phenomenon and as a result there is growing literature about dialogic repetition. Repetition here refers to uttering the same item uttered by another speaker, generally the previous speaker. Because it appeared to have different functions in the interaction, a functional classification was found to be useful in the analysis of the speech of patient A.S. Bazzanella's (1993) is a comprehensive functionalist taxonomy, when compared to others (Norrick 1987). In her taxonomy, she proposes that repetition has macro and microfunctions. The typical repetitional pattern of A.S.' speech involves him repeating what I said in the previous question. Following Bazzanella's (1993) taxonomy, this can be characterized in terms of the following micro and macro functions (where R stands for repetition): 1. Cognitive 1.0 features of automaticity: automaticity, pervasiveness 1.1 R facilitates language production

1

Transcription conventions follow the usual pattern, with 4 indicating low pitch. The translation is a rough one; brackets refer to ellipsis that are allowed in BP and not in English. Transcription conventions: final falling intonation , rising intonation, more to come ? rising intonation low pitch pause of half than a second pause of a second ( ) specified seconds in the pause CAPS emphatic stress [ overlapping turns : elongated vowel /words/ uncertain transcription /?/ inaudible utterance (( )) marked movements in the talk

Paula Fatur-Santos

192 5. Conversational 5.1 R makes the conversation advance easily 5.4 R is used to hold the floor, as a filler 6. Interactional 6.1. R marks agreement

6.1.4 a suggested lexical item is repeated in order to adopt it in the speaker's own discourse This is the case in 45.3% of the turns of the conversation. Some examples follow: 021 022

I: (0.2) Ah ta'.. e agora? como e' que tu ta' se sentindo? melhor? I: I see. And how are you feeling now? Better? Ρ: 4-Τό meihor. Ρ: Ifeel better.

027

I: Que mais? I: What else? 028 P: 4 Esque^o muito ... desse ouvido aqui näo 01190 ((pointing to the left ear)) Ρ: Iforget a lot. I cannot hearfrom this ear. 029 I: Näo escuta bem? I: You cannot hear well? 030 P: 4-Näo escuto bem. Ρ: I cannot hear well. 097

I: Mas tu estudou la' mesmo? ... foi la' que tu fez escola? I: Did you go to school there? 098 P: 4-E' foi... estudei mesmo foi em Mafra. P: Yes,I did. I really went to school in Mafra. 099 I: Näo em Curitiba? I: Not in Curitiba? 100 P: 4 - N ä o em Curitiba... em Mafra nos fomos pra' Curitiba P: Not in Curitiba. In Mafra we went to Curitiba. Another mark of A.S.' interaction is his self-repetition, which can be said to have the same cognitive and conversational microfunctions cited previously. These features happen in 28.3% of the conversation, as shown in Table 1. Some illustrations are given below. 023

I: Como e' que ta': cabeca? I: How's your head? 024 P: 4- So* assim umjeito bobo... esque^o muito P: Only a little dumb, you know, Iforget a lot. 025 I: Esquece /: You forget. 026 P: -J-Esque^o muito. Ρ: Iforget a lot.

Interacting with a Frontal Lobe Patient: What about Pragmatics? 027 028

193

I: Que mais? I: What else? P: 4- EsquefO muito ... desse ouvido aqui näo 0U90 ((pointing to the left ear)) Ρ: Iforget a lot. I cannot hearfrom this ear.

016 P: 4-EU dormi assim de um jeito assim ... näo lembro de nada. Ρ: I slept, you know, in such a way, you know, I don V remember anything. 017 I: Näo lembra de nada I: You don't remember anything. 018 P: 4-Näo lembro de nada. Ρ: I don 7 remember anything. Table 1: Percentage of repetition in A.S.' discourse R of interviewer's utterances

Self-repetition

Total

45.3%

28.3%

73.6%

When the word dai (then) is used, besides having the microfunctions of automaticity and pervasiveness (1.0) and of making the conversation advance easily (5.1), it clearly serves to facilitate language production (1.1), probably due to the very function of dai, as an enumerator. Consider the examples that follow: 019 I: Ε depois que tu acordou? I: And what about after you woke up? 020 Ρ: 4- Dai /eu sei que/ eu fui parar no /mauri/.. dai depois que eu tornei assim. P: Then I /know that 1/ ended up at the /mauri/, and after that I became like this. 011 012

I: Dai alguem te socorreu? como e* que foi? I: And then, did anybody help you? What happened then? P: 4-Quando eu fui la'., dai veio ... ligaram pro /?/ daf que me levaram. P: When I went there, then ... came, they called the /?/, then they took me.

5.4. Evasive answers Evasive answers can be seen in the following example, where A.S. does not actually answer what is being asked: 001

002

003

I: Me conta um pouquinho assim como e' que foi .. como e' que foi quando tu ficou doente: e ο que que tu sentiu primeiro I: Tell me a little bit about how you became sick and how it all started. P: 4-Näo ... so' a cabeca assim que näo tava boa, a parte da frente aqui assim. ((pointing to the forehead)) P: No, it was only my head that was not good, you know, the front part, here. I: Ο que que aconteceu? dor?

Paula Fatur-Santos

194

004

005 006

I: What happened, pain? Ρ: i E ' . . . sei la' ... um tr0£0 ruim ... näo era uma dor assim, um tr090 so' ruim aqui assim. ((pointing to the forehead)) P: Yes, I don't know, a bad stuff, it was not pain, you know, only a bad stuff, right here, you know. I: Ο que que era? I: What was it? P: IParece que era aqui. ((pointing to the forehead)) P: It seems that it was here.

Evasiveness can often be realized through the use of unspecific content words such as trogo (stuff), in line 004 presented right above. It can also be realized through fillers, such as assim (you know): 015 016

I: Ε como e' que tu te sentiu la'? no pronto socorro.. quando cuidaram de ti? I: And how did you feel there, in the ER, when they took care of you? P: 4-EU dormi assim de um jeito assim ... näo lembro de nada. Ρ: I slept, you know, in such a way, you know, I don't remember anything.

Or still through deictics without a referent, such as disso (this), eles (they), la' (there): 039 040 041 042 043 044

I: Ε dai.. tu teve que parar? I: And then, did you have to quit? P: 4-E' dai eles me mandaram embora.. depois.. ./disso/. P: Yes, then they fired me. After /this/. I: ELES mesmo? I: They did? P: 4-Eles me mandaram embora P: They fired me. I: Por que? tu näo conseguia mais fazer? J: Why? Were you not able to keep up with yourjob anymore? P: 4-Näo .. eles näo .. nnäo dai eu nnäo iniciei a trabalhar depois que tava... näo fiii la' pra' /?/ trabalhar mais .. Fiquei parado. cheguei la' e disseram nos vamos te dispensar .. e dispensaram P: No, they didn't, no, then I did not start working again, after I became..., I did not go /?/ to work anymore, I quit working. I went there and [they] told me [they] were going to lay me off and [they] laid me off

An example where A.S. does seem to answer the content of a question is: 065 066

I: Ο que que tu mais gostava? I: What did you like about there? P: 4-Ah ... gostava, nossa... safa P: Oh, I liked [it], I used to go out.

However, this answer is characterized by keywords, not much elaboration. Besides, despite the fact that in 1. 066 he does answer the content of the question, the actual reason why he

Interacting with a Frontal Lobe Patient: What about Pragmatics?

195

likes his hometown is more than that, if we consider the whole interaction. Every time the topic is mentioned, he comes up with another reason: 1. 066, saia (used to go out), 1. 084, trabalhava (was working), 1. 074, lugar da gente (my place), 1. 106, muito tempo ali (lived a long time there). Probably the prompting stimulates his memory and activates more ideas. However even with the prompting, his answers are restricted to keywords or very concise, showing a low rate of speech production and elaboration. In dealing with topics such as going back to his hometown, Curitiba, which is emotionally relevant to him, there are some instances in which we can see an attempt to elaborate more on the topic. In such cases, the elaboration is also done through repetition, here used to increase the argumentative force (Bazzanella's microfunction 4.1). This happens in lines 062,086, as shown below: 061 062

085 086

I: Ah... ta':.. e como e' que era em Curitiba? era bom? eramelhorque aqui? /: I see. And how was it in Curitiba, was it better than here? P: 4-Näo eu gosto de la'.. .eu gosto gosto. eu quero ir pra' LA'. P: No, I like it there. I like, like. I want to go back there. I: Ah ta'. tu veio pra' ca' e logo foi despedido. entäo hi tern vontade voltar? I: I see. You came here and soon were fired. So would you like to go back? P: iPois e'... eu quero eu sinto falta tava dizendo ne tenho que ir casa P: That's it, I want [it], I miss [it], I was saying that [I] want to go back home, you know.

6. Discussion

The view often maintained that repetition in FLS is just a trace of perseveration represents a limited one, since it does not encapsulate the very fact that repetition might be a valuable resource as adaptation to turns at talk in conversation, in other words, a compensatory strategy to converse in the absence of the ability to do it spontaneously, allowing the patient to fulfill his turn. Bazzanella (1993) widens this view as she attempts to identify a variety offunctions of repetition, many of which apply to the discourse typical of the FL syndrome. Thus, this apparent conversation or dialogue is actually a use of micro and macro functions of repetition as coping interactional strategies to deal with some specific memory, cognitive and linguistic deficits caused by frontal lobe damage. In speculating on what neuropsychological impairments underlie reduced utterances in the discourse of A.S., we must first consider that a conversation requires more information than what is literally being asked, specially in the Brazilian style. If asked "Tell me about your illness" this means, pragmatically, that one wants to know how/when/where it

196

Paula Fatur-Santos

happened, what/who caused it, and so on, in short, global aspects. A.S. is unable to fulfill the global aim of the conversation, going beyond the literal questions. He is also unable to provide sufficient and relevant information at the local level, in general not being able to activate frames and sub-frames. Secondly, activating frames depends on making inferences. If inferential processes are not working satisfactorily, frames cannot be activated and consequently filled properly. Noteworthy is the loss of contextual facilitation that frontal patients have. Milner and Petrides (1984) report that frontal patients exhibit impaired performance with visual and tactual stylus-maze tasks, i.e., they have difficulties in using "cues or environmental stimuli", i.e., context to regulate their actions. Turning to A.S.' case, as shown in the analysis, he seems to be unable to utilize the contextual clues that are provided to him (by me) to produce more speech or to elaborate on given topics. No matter how much I probe, and this implies giving cues (sometimes specific ones, as in 1. 027, Que maisl) these cues do not trigger more new information, more specific information, or more relevant information. He does not seem to benefit from them at all or does very little. Regarding neuropsychological impairments that underlie repetition, one hypothesis is that because of memory problems, he repeats in order to keep the topic of the question on the focus of his attention (a mnemonic coping strategy, thus). But above all, the central deficit being in the microgenesis of utterances, it seems clear that he repeats in order to fill his turn, a perseverating behavior, to cope with the difficulty in bringing in novel stimuli to the conversation. Why filling his turns, I ask, if he could do the same as the seminal patient Phineas Gage did, when asked why he would not speak. He simply answered, "I really have nothing to say" (Damasio 1994). Like Phineas Gage, A.S. had reasons to do so, after all, his pathology causes a severe social withdrawal, which can be noted as he does not get involved in the interaction at all. He was totally disconnected from the conversation. I had a feeling that if he could, he would not reply at all. However, A.S. was not in a situation in which he could refuse to speak. He probably associated my presence to his neurologist's, thus inferring that this was actually a doctor interview, meant to be used as a measure of his capabilities. And it seems that he was very aware that the nature of the doctor-patient interaction is one of power (Ribeiro 1994). The power the doctor has upon his future life decisions, since he is not able to manage his own life anymore. He inferred that this interview was really about performing a role, that of the capable patient, and not about exchanging information between two people who are interested in each other. And he did perform his role well, by always finding a way of filling his turn, so that he was not labeled incapable. It is interesting that this knowledge of genre of a doctor's interview is intact, for it is to a large extent dependent on inference and context to be activated, abilities taken to be impaired in the FL syndrome. His participation

Interacting with a Frontal Lobe Patient: What about Pragmatics?

in the genre interview

197

here was probably driven by the fact that emotionally A.S. longs to

be recognized as a person who is able to make decisions about his own life, in opposition to a patient with a chronic illness, w h o no longer is. Since the features o f the discourse of this FLS patient are very much in tune with the neuropsychological findings at all other levels o f impairments of the FLS (memory, cognition, language) the conclusions drawn here can largely contribute to a standard communicative profile o f the syndrome, although not to a universal one.

References Baddeley, A. (1986): Working Memory. - Oxford: Oxford University Press. Bazzanella, C. (1993): Dialogic repetition. - In: H. Löffler (ed.) Dialogue analysis IV: Proceedings of the 4 th Conference, Basel 1992. Tübingen: Max Niemeyer Verlag. Beaumont, G., P. Kenealy and M. Rogers (eds.) (1996): The Blackwell Dictionary of Neuropsychology. - Cambridge, Mass.: Blackwell Publishers. Brown, J. and K. Chobor (1989): Frontal lobes and the problem of perseveration. - Journal of Neurolinguistics 4, 65-85. Caramelli, P., L. Mansur and R. Nitrini (1998): Language and communication disorders in dementia. - In: Stemmer and Whitaker (eds.). Cherrier, M., M. Mendez, J. Cummings and F. Benson (1998): Language and communication in nonAlzheimer's dementias. In: Stemmer and Whitaker (eds.). Cohen, H. (1998): Language impairment in Parkinson's disease. - In: Stemmer and Whitaker (eds.). Damasio, A. (1994): Descartes' Error: Emotion, reason and the human brain. - New York: Putnam. Denes, G. (1998): Landau-Kleffner syndrome: Clinical and linguistic aspects. - In: Stemmer and Whitaker (eds.). Ehrlich, J. (1988): Selective characteristics of narrative discourse in head-injured and normal adults. -Journal of Communication Disorders 21, 1-9. Garcez, P. (1996): Brazilian Manufacturers and United States Importers Doing Business: The coconstruction of arguing sequences in negotiation. Doctoral Dissertation, University of Pennsylvania. Haegeman, L. (1994): Introduction to Government and Binding Theory. - Cambridge, Mass: Blackwell. Halliday, M. and R. Hasan (1976): Cohesion in English. - London: Longman. Happi, F. (1998): Language and communication disorders in autism and Asperger's syndrome. - In: Stemmer and Whitaker (eds.). Milner, B. and M. Petrides (1984). Behavioral effects of frontal-lobe lesions in man. - Trends in Neurosciences 7(11), 403-407. McDonald, S. (1998): Communication and language disturbances following Traumatic Brain Injury. - In: Stemmer and Whitaker (eds.). Norrick, N. (1987): Functions of repetition in conversation. - Text 7(3), 245-264. Paradis, M. (1994): Neurolinguistic aspects of implicit and explicit memory: Implications for bilingualism. - In N. Ellis (ed.) Implicit and Explicit Learning of Second Languages, 393-419. London: Academic Press.

198

Paula Fatur-Santos

Paradis, M. (1998): The other side of language: Pragmatic competence. - Journal ofNeurolinguistics 11, 1-10. Perkins, M. (1999): Disentangling syntactic, semantic and pragmatic impairment in a 9-year-old boy with complex high-level communication difficulties. Paper presented at the 7th Annual Conference of the International Clinical Phonetics and Linguistics Association. Montreal, May. Ribeiro, B. (1994): Coherence in Psychotic Discourse: Frame and topic. Doctoral Dissertation, Georgetown University. Sarkari, S., A. Tan and D. Molfese (1998): The development of language in some neurological diseases. - In: Stemmer and Whitaker (eds.). Stemmer, Β. and Η. Whitaker (eds.) (1998): Handbook ofNeurolinguistics. - London: Academic Press. Tracy, J. (1998): Language abnormalities in psychosis: Evidence for the interaction between cognitive and linguistic mechanisms. - In: Stemmer and Whitaker (eds.).

Gerd Fritz Dialogical Structures in 17,h Century Controversies

1. Introductory remarks

This article is a contribution to historical dialogue analysis, a field of research which has gained momentum in recent years (Fritz 1995, 1997, Gloning 1999, and other articles in Jucker/Fritz/Lebsanft 1999). In the present paper, I report some results of ongoing research from a project on the history of controversies from 1600 to 1800, which Marcelo Dascal and I are conducting at the Universities of Tel Aviv, Israel and Gießen, Germany. 1 One of the reasons why controversies form an interesting topic for historical dialogue analysis is the fact that exchanges of this type follow fairly well-defined rules and principles which are often explicitly stated and reflected upon by the controversialists themselves. Therefore we are in the fortunate position of having comparatively well-founded analytical categories for the history of this form of communication. Furthermore, there are highly interesting historical developments in this form of communication, especially in the course of the 17th and 18lh centuries. Aspects of controversies that are involved in these historical changes include the traditional point-by-point procedure of refutation and its decline, the status of certain types of argument (e.g. the argument from authority, especially from biblical and classical authority in philosophical and scientific controversies), and the use of certain media (e.g. from pamphlets to scientific journals).

2. The basic form of a 17th century controversy

The first question I shall address is the following: What are the fundamental characteristics of a typical 17th century controversy? Basically, it consists of a sequence of printed pamphlets on some topic from science, philosophy, theology, etc. These pamphlets will be The project "Controversies in the Republique des Lettres 1600-1800" is supported by the German Israeli Foundation for Scientific Research and Development. This cooperation between philosophers and linguists was inspired by Dascal's ground-breaking work on the history and theory of controversy (cf. Dascal 1989, 1998).

200

Gerd Fritz

closely related to each other by thematic and functional links. In many cases, the titles of the pamphlets will already indicate a dialogical connection between individual contributions, like in the following examples from a famous controversy between the philosopher Thomas Hobbes and the Anglican Bishop John Bramhall on the problem of free will. The first, comparatively short pamphlet (49 pages) by Thomas Hobbes, printed in 1654, has the following title: Of Libertie and Necessitie, A Treatise, Wherein all Controversie concerning Predestination, Election, Free Will, Grace, Merits, Reprobation etc. is fully decided and cleared, in answer to a Treatise written by the Bishop of Derry on the same subject. By Thomas Hobs. London: F. Eaglesfied 1654.

This title is quite a mouthful, but this is characteristic of the baroque titles of the period. Bramhall answered Hobbes's attack with a much longer pamphlet (253 pages) titled: A Defence of True Liberty from Antecedent and Extrinsecal Necessity; Being an Answer to a late Book of Mr. Thomas Hobbes of Malmesbury entitled A Treatise of Liberty and Necessity. Written by the Right Reverend John Bramhall, D.D. and Lord Bishop of Derry. (1655)

Although the respective pamphlets are termed answers to earlier contributions - which suggests a dialogical relationship - , the question remains, in what sense such controversies can be considered genuine dialogues, or, to put it more precisely, in which way the organizing principles of dialogue structure controversies of this kind. The strategy of this inquiry will be to examine different parameters of dialogue in order to find out in which way the course of the controversy and the structure of the individual contributions are determined by these factors. These parameters are: system of address, topic management, and functional sequencing. In addition I shall present some observations on the use of dialogue form in 17lh century pamphlets. As a preliminary remark, it is necessary to point out that quite a number of controversies actually started out as genuine dialogues, i.e. as oral disputations. This is true, for instance, of Galileo's disputation about floating bodies which was held at the house of his patron Salviati in Florence in 1611 (cf. Bagioli 1993: 159ff.) and of the oral discussion of free will which Hobbes and Bramhall had at the house of their patron, the Marquis of Newcastle, during their exile in Paris in the year 1645. It is therefore not surprising that the traditional rules of oral disputation govern much of 17th century printed controversies.

Dialogical Structures in I th Century

Controversies

201

3. System of address

The main difference between oral disputations and printed controversies lies in the fact that printed pamphlets normally aim at a wider public. Therefore one factor becomes dominant which, of course, may also be present in oral disputes, namely the orientation towards a public audience. This is clearly reflected in the system of address used in pamphlets. In the 17"1 century, the authors usually address two addressees directly, namely the patron to whom their pamphlet is dedicated and the reader who is normally addressed in the preface, typical forms of address being gentle reader, Christian reader, to the sober and discreet reader, etc. The opponent, however, is normally not addressed directly at all but only mentioned in the third person. Hobbes, for example, usually refers to Bramhall as "the Bishop" or "his Lordship". The Englishman Clarke, in his controversy with Leibniz (17151716), usually refers to Leibniz as "this learned author" (cf. Correspondence: 190, 195, etc.). There are some notable exceptions to this rule, e.g. Milton's polemical aside directed at an opponent: "Where didst thou learne to be so agueish, so pusillanimous, thou lozel Bachelour of Art [...]?" (cf. Bach 1997: 146). But, generally speaking, if we go by the system of address, there is no direct dialogue between the opponents in most 17th century controversies.

4. Topic management

But, of course, direct address is just one parameter of dialogue. There are other parameters, in respect of which sequences of contributions to a controversy show very strong coherence relations reminding us of exchanges of letters or even of oral talk exchanges. The most obvious case in point is topic management. In a typical case, author A will raise a number of topics. Author Β will take up every single topic raised by author A in order to deal with these topics point by point. The principle of point-by-point topic organization is very characteristic of many 16th to 18"1 century controversies. It is a kind of generative principle for the textual structure of pamphlets. The power of this principle lies in the fact that someone who fails to take up a topic that has been raised will be seen as having lost a point in the disputation game, and of course nobody wants to lose a point. As this principle only fixes a downward boundary, it does not keep authors from elaborating on topics which have been introduced and from introducing topics of their own. So from the point of view of topic management we often find interesting cases of joint production of topics

202

Gerd Fritz

which remind us of everyday dialogue. From the point of view of the textual structure of pamphlets, the application of this principle leads to very long and complex texts, which are often not particularly well-structured internally. The interesting point about this is that in this case what is obviously a disadvantage for the textual structure of the individual pamphlet is a direct consequence of the dialogical structure of the exchange of contributions. In many cases improving the structure of one's own text would mean relaxing the adherence to the point-by-point requirement, which some late 17th century authors did in fact do. For instance, August Hermann Francke, the Pietist theologian, in two cases answered the accusations of his Orthodox opponent Johann Friedrich Mayer by writing open letters to friends in which he concentrated on selected topics which he himself considered essential, thereby avoiding having to deal with a plethora of irrelevant calumnies. This was in 1706. A year later, however, Francke wrote a long pamphlet in which he reverted to the traditional point-by-point procedure in a final effort to stop his opponent's mouth (cf. Francke, Streitschriften: 217-381).

5. Functional sequencing

A second organizing principle of dialogue which creates text structure is what I call functional sequencing, i.e. the use of sequencing patterns like question and answer, assertion and objection or refutation, accusation and counteraccusation. The functional organization at the global level - i.e. the statement of an author's position followed by the opponent's answer in the form of a refutation or castigation etc. - is naturally also reflected at the micro-level. In the individual pamphlets we find typical reactions which are well-known from oral debates, e.g. giving an argument against a proposition held by the opponent, justifying a controversial earlier statement of his own or countering an accusation by a counter-accusation. So these dialogical micro-patterns also create coherence between the contributions to a controversy and they are also responsible for part of the textual structure of the individual contribution. As this is quite straightforward I shall not go into it in more detail.

Dialogical Structures in 1 fh Century Controversies

203

6. The use of dialogue form in pamphlets

I shall however briefly discuss a typical problem of communication for authors of pamphlets and a solution to this problem - which consists in using dialogue form for the purpose of making visible the dialogical structure of the ongoing controversy. One of the problems authors face in writing a pamphlet as part of such a presentation of dialogue is a problem of knowledge-management. In order to understand the point of a certain pamphlet, the reader has to know the history of the controversy, or, at least, he has to know the direct target of the new pamphlet, i.e. the opponent's earlier contribution to which the present pamphlet responds. In many cases, however, the authors could not assume their readers to have available the texts of which the controversy consisted or to remember previous contributions. So they had to take measures in their own text to provide readers with the appropriate knowledge. One frequently used method is to give a survey of the controversy in the preface. Another device is to begin the actual text with a paragraph summarizing the state of the controversy ("status controversiae"). Other means are used in the micro-structure of the text itself, e.g. when the author gives a short indication of the opponent's position before presenting his own reaction. This kind of textual element usually has the form "My opponent says that p, but, of course, that is not true, as I shall now prove". This is already a rudimentary presentation of dialogue. Now, a particularly ingenious method of providing the necessary background is to actually replay the preceding controversy concerning a particular point in the form of a small dialogue. When authors do this the dialogue structure of the controversy surfaces in a particularly visible form. I shall give two examples of this technique, one taken from a controversy between the German astronomer Kepler and his opponent Röslin, who argues in favour of astrology, and the second one taken from the controversy between Hobbes and Bramhall, which I already mentioned.

6.1. Johannes Kepler vs. Helisaeus Röslin In the year 1604, a spectacular comet was closely observed and widely commented on by astrologers and astronomers, among them Johannes Kepler in his book " D e Stella nova" written in 1606. In this book Kepler complained that in one of his earlier prognostic writings the astrologer Helisaeus Röslin had only picked out the two comets of 1556 and 1580 for astrological analysis, leaving out comets which had appeared before 1556 and others which had been observed between 1556 and 1580. Röslin, in his attempted refutation of

204

Gerd Fritz

Keplers arguments, which he wrote in 1609, quotes Keplers Latin and gives an answer in German, presenting this exchange as a dialogue. Kepplerus: Quare imbecillis est connexio illius anno 1556. fulgentis, cum hoc anni 1580. ex solo contrario signo & contrario motu deducta. (Kepler: Therefore the connection between the comet shining in 1556 and the one in 1580 is rather weak, as it is deduced merely from the fact that they have opposing astrological signs and an opposing direction of movement.) Röslinus: Ich hab auch nicht dahin gesehen / vnd dieser contrarietät der Zeichen vnd bewegung nicht meidung gethan / vnd mein Sach nicht aus diesem grund geführt (Röslin: Actually, I did not consider these facts and I did not mention the contrariety of signs or of movements, and I did not base my argument on these facts.)

In the same year, Kepler answered Röslin's pamphlet and used the same technique. Referring to the passage just mentioned, Kepler first quotes Röslin - not strictly verbatim - and then adds his answer, also in dialogue-fashion: D. Röslin: Ich hab nit dahin gesehen / oder mein sach nit auß diesem grund geführt das der Comet Anno 1556 vnd der Anno 1580 auß contrarijs signis vnd contrarijs motibus gelauffen. (Röslin: I did not consider these facts, and I did not base my argument on the fact that the comets of 1556 and 1580 moved in opposing signs and in opposite directions.) Kepplerus: Ich aber hette gemaint / dise vmbstende solten D. Röslins fürhaben in zusammenfeslung baider Cometen ein vil bessers ansehen gemacht haben [...] (Kepler: But I should have thought these facts would have made Dr. Röslin's attempt at tying together these two comets look much more convincing.)

From the point of view of Kepler's argumentative technique it would be worth analysing in detail Kepler's rather subtle manoeuvre of suggesting a possible improvement of his opponent's argument. However, I shall not go into this here. What interests me at this point is the textual strategy which Röslin uses and which Kepler takes up, i.e. presenting their dispute on this particular topic in the form of an actual dialogue. This form of dialogical rhetoric has at least a double function: it adds to the attractiveness of the pamphlet by giving a more vivid picture of the dialogical aspect of the controversy and, at the same time, it provides the reader with the necessary background of knowledge.

6.2. H o b b e s vs. Bramhall As I said, this technique was also used in the Hobbes vs. Bramhall controversy. But there the authors used it even more extensively by stringing together three or even four utterance units and thereby creating three-or four-part dialogue sequences. Bramhall uses this strat-

Dialogical Structures in

Century Controversies

205

egy in a pamphlet which was the third contribution to the controversy. He had written a treatise on the subject "Of liberty and necessity", to which Hobbes answered with a treatise on the same subject, published in 1654. As I already mentioned, Bramhall replied in 1655 with "A defence of true liberty". In his "Defense", Bramhall replays the dialogue by first giving his own original statement under the abbreviation J.D. and then adding Hobbes's reaction under T.H. The dialogue is then continued with Bramhall's reply. Bramhall does this for every one of the 38 points which are covered in this controversy. A year later, in 1656, Hobbes responded again with another pamphlet "The questions concerning liberty, necessity, and chance". In this text, Hobbes tops Bramhall's technique by recording, for every one of the 38 points, the whole three-part dialogue enacted so far and adding, as a fourth part, his own "Animadversions on the Bishop's reply". As an example of this technique, I present a slightly abbreviated version of the exchange for point no. XVII: J.D. "Fifthly, take away liberty and you take away the veiy nature of evil, and the formal reason of sin. [...] Therefore it appears, both from Scripture and from reason, that there is true liberty." T.H. To the fifth argument from reason, which is, that if liberty be taken away, the nature and formal reason of sin is taken away, I answer by denying the consequence. [...] And thus you have my answer to his objections, both out of Scripture and reason. J.D. [...] it seems T.H. thinks it a more compendious way to baulk an argument, than to satisfy it. [...] But it will not serve his turn. And that he may not complain of misunderstanding it [...] I will first reduce mine argument into form, and then weigh what he saith in answer, or rather in opposition to it. [...]

Animadversions upon the bishop's reply no. XVII Whereas he had in his first discourse made this consequence: "If you take away liberty, you take away the very nature of evil, and the formal reason of sin": I denied that consequence. It is true he who taketh away liberty of doing, according to the will, taketh away the nature of sin; but he that denieth the liberty to will, does not so. But he supposing I understand him not, will needs reduce his argument into form, in this manner, (a) "That opinion which takes away the formal reason of sin, and by consequence, sin itself, is not to be approved." This is granted. "But the opinion of necessity doth this. This I deny; [...]" (cf. Hobbes, Questions: 228-233)

Bramhall himself at one point explains why he uses this textual device: for him, the point of this procedure is that it permits the reader to "compare plea with plea and proof with p r o o f ' (cf. Bramhall, Castigations: 506) and to judge for himself what the truth is. It is a certain rule, "contraries being placed one besides another, do appear much more clearly. He who desires to satisfy his judgment in this controversy, must compare our writings one with another without partiality, the arguments and answers and pretended absurdities on

206

Gerd Fritz

both sides" (cf. Bramhall, Castigations: 226). But, of course, such a dialogue-presentation is also a means of presenting a more lively view of the course of the controversy.

7. On the history of the use of dialogue form

As for the history of this type of presentation of the dialogue structure of a controversy, I am not sure when this device was first used. What is obvious is that dialogues between fictional characters were a widely-used didactic genre, from Socratic dialogue to the many fictional dialogues written in the 16th century, e.g. the "Colloquia familiaria" by Erasmus (1518 ff.). This philosophical and literary genre provided a model for authors of how to present their real controversies in dialogue form. In many cases, fictional dialogues were related to real controversies, as in the case of Galileo's fictional "Dialogue concerning the two chief systems of the world", written in 1632, which can be considered a continuation of earlier real life controversies (Spranzi Zuber 1998). As for the use of dialogue sequences to represent the state of the controversy, examples go back at least until the early 16th century. There is, for instance, a beautiful example in the Reuchlin vs. Pfefferkorn controversy of 1525 (cf. Schwitalla 1999: 119). As I am at the moment not dealing with this earlier period, I shall leave it at that. But it would certainly be interesting to know more about the early history of this dialogical device for text construction.

8. Summary

The upshot of this analysis is that typical 17th century controversies are not, strictly speaking, dialogues between the opponents, although they show characteristic properties of genuine dialogue, like dialogical topic management and functional sequencing of a dialogical nature. Dialogue structures to a large extent determine the text structure of pamphlets. The authors involved in such a controversy present a dialogue-like sequence of texts to the reading public, not unlike what we find today in television debates. This is in consonance with Dascal's view of "Controversies as quasi-dialogues" (Dascal 1989). In the second half of the paper I focussed on a particularly interesting textual device, the use of actual dialogue in monological printed texts, which, as far as I can see, serves at least three different purposes: it helps to present necessary running knowledge to the audience,

Dialogical Structures in l f h Century Controversies

207

it provides an overview of the contrasting positions, and it gives the audience a more vivid sense of ongoing dialogue.

Sources

[John Bramhall]: Castigations Of Mr. Hobbes His Last Animadversions In The Case Concerning Liberty And Universal Necessity [...] By John Bramhall, D.D. And Bishop Of Deny. London 1658. Reprinted in: The works of the most Reverend Father in God, John Bramhall, D.D. [...]. Vol. IV. Oxford 1844, 197-506. [August Hermann Francke]: August Hermann Francke. Streitschriften. Herausgegeben von Erhard Peschke. Berlin/New York 1981. [Thomas Hobbes]: The Questions Concerning Liberty, Necessity, And Chance. Clearly Stated And Debated Between Dr. Bramhall, Bishop Of Derry And Thomas Hobbes Of Malmesbury. London 1656. Reprinted in: The English works of Thomas Hobbes of Malmesbury. Now first collected and edited by Sir William Molesworth, Bart. Vol. V. London 1841, 1-455. [Johannes Kepler]: Antwort Joannis Keppleri [...] AuffD. Helisaei Röslini Medici & Philosophi Discurs Von heutiger zeit beschaffenheit [...]. Gedruckt zu Prag bey Pauln Seese. Jm Jahr / 1609. Reprinted in: Johannes Kepler. Gesammelte Werke. Band IV. Kleinere Schriften. [...] Herausgegeben von Max Caspar und Franz Hammer. München 1941, 101-144. [Gottfried Wilhelm Leibniz, Samuel Clarke]: Correspondance Leibniz-Clarke presente d'apres les manuscrits originaux des bibliotheques de Hanovre et de Londres par Andre Robinet. Paris 1957. [Helisaeus Röslin]: Historischer / Politischer vnd Astronomischer naturlicher Diskurs Von heutiger zeit Beschaffenheit / Wesen vnd Standt der Christenheit / vnd wie es inskünfftig in derselben ergehn werde / [...]. Durch Helisaeum Röslin / Med. Doct. [...] Gedruckt zu Straßburg / bey Conrad Scher / Jn Verlegung Paulus Ledertz / Jm Jahr 1609.

References

Bach, U. (1997): Englische Flugtexte im 17. Jahrhundert. Historisch-pragmatische Untersuchungen zur frühen Massenkommunikation. - Heidelberg: Winter. Biagioli, M. (1993): Galileo Courtier. The practice of science in the culture of absolutism. Chicago/London: The University of Chicago Press. Dascal, M. (1989): Controversies as quasi-dialogues. - In: E. Weigand, F. Hundsnurscher (eds.) Dialoganalyse II. Bd. 1, 147-160. Tübingen: Niemeyer. Dascal, Μ. (1998): Types of polemics and types of polemical moves. - In: S. Cmejrkovä et al. (eds.) Dialoganalyse VI. Teil 1, 15-33. Tübingen: Niemeyer. Fritz, G. (1995): Topics in the history of dialogue forms. - In: A. H. Jucker (ed.) Historical Pragmatics, 469-498. Amsterdam/Philadelphia: Benjamins. Fritz, G. (1997): Remarks on the history of dialogue forms. - In: E. Pietri (ed.) Dialoganalyse V. Referate der 5. Arbeitstagung Paris 1994,47-55. Tübingen: Niemeyer.

208

Gerd Fritz

Gloning, T. (1999): The pragmatic form of religious controversies around 1600. A case study in the Osiander vs. Scherer and Rosenbusch controversy. - In: A.H. Jucker, G. Fritz, F. Lebsanft (eds.) Historical Dialogue Analysis, 81-110. Amsterdam/Philadelphia: Benjamins. Jucker, A. H., G. Fritz, and F. Lebsanft (eds.) (1999): Historical Dialogue Analysis. Amsterdam/Philadelphia: Benjamins. Schwitalla, J. (1999): The use of dialogue in Early German pamphlets: On the constitution of public involvement in the Reuchlin-Pfefferkorn controversy. - In: A. H. Jucker, G. Fritz, F. Lebsanft (eds.) Historical Dialogue Analysis, 111-137. Amsterdam/ Philadelphia: Benjamins. Spranzi Zuber, Μ. (1998): Dialectic, dialogue, and controversy: The case of Galileo. - Science in Context 11, 181-203.

Renata Galatolo Les strategies de changement de footing dans le temoignage commun au tribunal: une ressource pour la construction de credibilite

1. Introduction

L'objectif de cette analyse est de ddceler certaines caracteristiques du discours et de l'interaction qui contribuent a la construction de l'effet de credibilite du temoin au tribunal.1 L'effet de credibilite est un aspect de l'effet plus general de vraisemblance du temoignage en tant qu'effet intersubjectivement construit dans et par le discours. Le concept de vraisemblance, en effet, renvoie ä une veritö narrative (Spence 1984) qui remplace le concept de verite en tant qu'adhesion du discours ä la realitd. Le probleme de la vdrite surgit a l'interieur de celle que Pollner (1987) appelle la raison mondaine·, dans le cas spdcifique, il represente les termes dans lesquels le jury se pose le probleme de revaluation des preuves, tandis que l'analyse du fonctionnement de la preuve judiciaire, qui se construit dans et par la parole, ne peut se baser que sur le concept de vraisemblance. Celui-ci impose le discours et l'interdiscours comme cadre de reference en excluant le probleme de la relation entre le discours et le monde ext£rieur. La vraisemblance s'articule ainsi en deux mouvements opposes et complementaires: le faire croire vrai, de la part de ceux qui temoignent, et le croire vrai, de la part de ceux qui doivent exprimer leur jugement (Galatolo et Mizzau 1999). Les manuels pour avocats (Morrill 1971), dans lesquels on trouve des indications precieuses ä propos des criteres devaluation du comportement discursif du temoin, et les travaux scientifiques sur l'argument (Conley et O'Barr 1990, O'Barr 1982) indiquent des criteres gdneraux devaluation de la credibilite du temoin sans pourtant reussir ä etablir des relations systematiques entre les caracteristiques du discours et sa force de persuasion. Les premiers, en effet, finissent souvent par renvoyer ä Γ experience du professionnel pour devaluation du cas specifique, tandis que les deuxiemes sont souvent dementis par

1

Ce travail fait partie d'une recherche plus ample sur la cr6dibilit6 au tribunal men6e en collaboration avec Marina Mizzau. Je la remercie pour ses sugg6stions et pour les discussions & propos de ce tdmoignage dont eile trouvera les traces dans cet 6crit. Je remercie aussi Leonardo Foggi pour avoir r6alis£ la transcription des donnees.

Renata Galatolo

210

I'analyse des effets des

mecanismes discursifs dans le contexte de proces particuliers

(Jacquemet 1996, Bogen et Lynch 1989). En conclusion, les critöres devaluation de la preuve orale semblent etre de nature impressioniste et non-systematique, similaires aux criteres qu'on utilise dans la vie de tous les jours pour etablir si quelqu'un dit la verite ou s'il ment. L'evaluation de la credibilite en general, et de la credibilite du temoin en particulier, se base ainsi sur une confrontation entre la performance du temoin et le modele implicite et socialement partage du recit 2 credible (Dulong 1998, Jackson 1988). Parmi les traits qui ont ete indiques comme elements contribuant ä la construction de l'effet de credibilite du factual

account, j e concentre ici mon analyse sur les modalites

implicites d'autocategorisation du narrateur en tant que membre d'une categorie sociale competente a dire (Hutchby et Wooffitt 1998). L'analyse se basera sur le concept de changement de footing de Goffman (1979).

2. Le proces

L'analyse se base sur la transcription 3 du temoignage de Maria Chiara Lipari au cours du proces pour l'assassinat de Marta Russo, une etudiante tuee d'un coup de feu alors qu'elle marchait le long d'une allee de l'Universite de Rome. Maria Chiara Lipari a participe au proces en tant que temoin de l'accusation et son temoignage, avec celui de Madame Alletto, secretaire du meme departement universitaire oil Madame Lipari travaillait, constituerent les preuves principales qui ont amene a la condamnation des deux accuses en absence de preuves materielles. II s'agit done d'un temoignage qui, non seulement a ete juge credible, mais qui a represente le chef d'accusation principal du proces. II s'agit d'un proces qui a eu en Italie une forte resonance dans les medias et dans l'opinion publique et ce pour plusieurs raisons: le lieu "sacre" ou il a ete commis, l'absence apparente de mobile, sinon celui du jeu et du defi entre deux amis quant a leur capacite d'utiliser une arme, et les personnes concemees, presque tous des experts en droit. Selon la these de l'accusation, qui a ete confirmee par le verdict, le coup a en effet ete tire

2 3

Avec le terme recit je me reßre ä l'acte de narrer et au produit de ce meme acte. Le videoenregistrement du t6moignage coincide avec les extraits du proefes qui ont et6 montrös par la transmission t616visie italienne Un giorno in pretura. Les donnies ont έίέ transeriptes sdlon la m6thode de l'analyse conversationnelle. Les conventions de transcription sont indiqudes ä la page 218.

Les strategies de changement de footing dans le temoignage commun au tribunal

211

d'une fenetre du departement de philosophie de droit et Fhomicide a ete commis par deux chercheurs du meme departement. A partir de la credibilite que le jury a attribuee ä ce temoignage, je propose ici l'analyse de certains de ses traits qui sont directement Iiis ä ce dernier aspect du contexte, la double identite des participants au proces: ä la fois temoins ou accuses, et experts en droit.

3. Changements de footing

Le temoignage de Maria Chiara Lipari prisente des anomalies frappantes si on le compare au temoignage commun standard.4 Madame Lipari a en effet ete convoquee comme temoin oculaire et comme tel, eile aurait du se limiter ä raconter ce qu'elle avait vu et entendu en relation avec l'homicide. L'autre condition qu'elle aurait dü partager avec tous les temoins en general, les communs et les experts, est le respect de l'obligation tres stricte sur la pertinence des reponses qui est gdndralement imposee durant le contre-interrogatoire.5 Or, le temoignage de Maria Chiara Lipari est parsemö de ses impressions, de ses evaluations et d'opinions personnelles a propos des faits et des circonstances qu'elle decrit et, chose frappante, ses opinions personnelles et ses evaluations sont souvent l'objet des questions de ceux qui Pinterrogent, y compris les avocats de la contrepartie. L'autre aspect interessant du temoignage, qui est central pour cette analyse, est la liberie d'action dont le temoin semble jouir. Non seulement eile obtient des espaces de parole inhabituels dans lesquels eile exprime ses opinions, mais eile riposte aux avocats, elle construit une alliance avec le president et refuse la mediation du ministere public. L'hypothese dont je pars est que tous ces comportements represented des modalites de projection, de la part du temoin, de son identite d'expert en droit dans l'interaction. 6 Cette projection est reconnue et soutenue comme telle par les autres interactants qui, non seulement s'abstiennent de la censurer, mais parfois l'encouragent. Suivant la definition de footing de Goffman:

4 5

6

Le t6moignage commun est ici mis en opposition avec le temoignage de 1'expert. Pour l'analyse du Statut de la maxime de la pertinence en tribunal, j e renvoie ä Mizzau (1998). Pour l'analyse du relächement des obligations sur la pertinence dans le timoignage de Maria Chiara Lipari, cfr. Galatolo et Mizzau 2000. Sur la construction de l'identit6 personnelle dans et par le discours, s61on une approche ethnomethodologique, j e renvoie ä Antaki et Widdicombe (1998).

212

Renata Galatolo

the multiple senses in which the self of the speaker can appear, that is, the multiple self-implicated projections discoverable in what is said and done (cf. Gofftnan 1981: 173)

L'analyse se concentrera sur les changements de footing que Maria Chiara Lipari realise entre l'identite du temoin commun et celle d'experte en droit.7

4. Analyse de deux exemples

4.1. Le relächement des obligations sur les roles conversationnels Le fragment qui suit est eclairant comme exemple de liberte interactionnelle dont le temoin jouit grace aussi ä l'attitude du President du tribunal. II s'agit en effet d'un cas oü le temoin prend Γ initiative conversationnelle (lignes 1927, 1929-30) bouleversant l'ordre de l'interaction qui est base sur l'alternance de questions et de reponses, et sur la subdivision entre les deux roles conversationnels correspondants: le representant de l'institution pose les questions et le temoin repond. Dans l'exemple qui suit, le President du tribunal intervient pour permettre au temoin de maintenir la parole (lignes 1940-1942) et il le fait de fa9on qu'elle puisse gerer en pleine liberte l'espace conversationnel qu'il lui concede: L: Maria Chiara Lipari, temoin Pe: Petrelli, avocat de la defense P: President du tribunal LS: La Speranza, ministere public

Exemple 1 (Lipari , lignes 1917-1966) 1917.L: l'ho ripetuto con il 1918.forse anche qui .hh (2.2) 1919.io non so (.) se 1920.veramente qualcuno si 1921. rende conto cosa vuol 1922. dire: ricostruire tan:ti 1923.particola:ri [tan:ti particolari 1924. Pe: [ma έ proprio 1925.questo il problema che

'

Une analyse des modalitös discursives qui permettent έ un t6moin commun d'activer son identitd de medecin et transformer localement sa diposition dans la deposition d'un expert se trouve aussi dans Matoesian 1999.

Les strategies de changement de footing dans le temoignage commun au tribunal

213

1926. ci rendia[mo benissimo= 1927.L: [no awocato= 1928.Pe: =conto 1929. L: =awocato (..) piano 1930. un attimo [io dico la prima volta 1931. Pe: [ma io non voglio parlare 1932.presidente non voglio 1933. parlare e interloquire se 1934. non attraverso le mie 1935.domande con il testimone 1936. (.) non le consenta di 1937. fare domande e di fare 1938.osservazioni su quello 1939. che io sto facendo 1940. P: si rivolga a no:i 1941.e:h dottoressa 1942. si rivolga a noi 1917. L: je Tai repete avec le 1918.peut-etre ici aussi .hh (2.2) 1919.je ne sais pas (.) si 1920.vraiment quelqu'un se 1921. rend compte ce que (a veut dire 1922.de reconstruire tant de details 1923. [tant de details 1924.Pe: [mais c'est justement 1925.c'est ?a le problfeme qu'on 1926.se rend [parfaitement= 1927.L: [non avocat 1928.Pe:=compte 1929.L: avocat (..) doucement 1930.un moment[je dis la premifere fois 1931.Pe: [mais je ne veux pas parier 1932.monsieur le president 1933.je ne veux pas parier et intervenir 1934.si non ä travers mes questions 1935.avec le timoin (.) 1936.ne lui permettez pas 1937.de poser des questions ou de faire 1938.des observations sur ce que je suis 1939.en train de faire 1940.P: adressez- vous & nous 1941.e:h doctoresse 1942.adressez vous & nous La transcription commence par un conflit entre l'avocat et le temoin. Le temoin polemise avec l'avocat a propos de la fa?on dont il Pinterroge (lignes 1919-1923) et au moment oü l'avocat ironise sur les difficultes de memoire du temoin (lignes 1924-1926), eile contreattaque d'une fa9on assez agressive: "Non avocat avocat doucement un moment" (lignes

Renata Galatolo

214

1927, 1929-30). II s'agit d ' u n e insubordination (Leonardi et Viaro 1983) de la part du temoin et c ' e s t pour 9a que l'avocat demande au President d'intervenir (lignes 1936-1939) pour empecher le temoin de poser des questions ou d'exprimer ses commentaires. Le President intervient (lignes 1940-1942), mais son intervention ne suit pas le sens de la sollicitation de l'avocat. II intervient pour garantir au temoin un espace de libre expression de ses pensees en lui disant tout simplement: "adressez-vous a nous doctoresse, adressezvous ä nous". L'insubordination du temoin est ainsi legitimee par le President qui, non seulement s'abstient de la censurer, mais il lui concede un espace metaconversationnel, c'est-ä-dire de commentaire sur Γ interaction en cours, espace normalement reserve aux representants de l'institution. Voilä la continuation de l'echange: 1943.L: 1944. 1945. 1946. 1947. 1948. 1949. 1950. 1951. 1952. 1953. 1954. 1955. 1956. 1957. 1958. 1959. 1960. 1961. 1962. 1963. 1964. 1965. 1966.

presidente ma se io la prima volta: uso una espressione tipo probabilmente il dottor auriemma: poi mi e- mi viene fatto: e:h ripetere lo stesso fatto mille volte allora io .h ripetendolo per la ventesima volta: dimentico di dire probabilmente dico il dottor auriemma: e- e poi mi viene contestato come un fatto (.) e:h rispetto al quale: nel verbale invece lo davo come cer:to lo d6 come certo o:ra non ho piü pensato da allora se fosse il dottor auriemma ad aver a:- la persona a cui per un attimo: (.) ho detto procuri il dizionario

1943.L: 1944. 1945. 1946. 1947. 1948. 1949. 1950. 1951. 1952.

monsieur le president mais si la premiere fois j'utilise une expression genre probablement le docteur auriemma: et apres on me fait e:h ripeter la meme chose mille fois alors moi .h en le repetant pour la vingtieme fois j'oublie de dire probablement

Les strategies de changement defooting dans le temoignage commun au tribunal 1953. 1954. 1955. 1956. 1957. 1958. 1959. 1960. 1961. 1962. 1963. 1964. 1965. 1966.

215

je dis le docteur auriemma: et et apris on me le conteste comme un fait (.) e:h par rapport auquel dans le procis-verbal au contraire je le consid£rais comme certain je le considire comme certain maintenance n'ai plus pense depuis si c'dtait le docteur auriemma qui avait ä:- la personne έ laquelle pour un instant (.) j'ai dit procurez le dictionnaire

Le temoin utilise son tour de parole pour critiquer la faipon dont l'avocat l'interroge (lignes 1947-1948) fa9on qui, έ son avis, explique ses apparentes contradictions (lignes 19491966).

4.2. La construction d ' u n role interactionnel autonome Le comportement de madame Lipari pendant le contre-interrogatoire se caracterise aussi par ses revendications unilat6rales d'autonomie interactionnelle. Parfois eile semble revendiquer l'autogestion de son interrogatoire et s'opposer k la regie selon laquelle le temoin est protege par la partie, l'accusation ou la defense, qui lui a demande de temoigner ou qui le represent. Cette conduite se concretise par un comportement que j'ai defini dissociation du ministere public. Je me refere par lä aux cas oü le comportement du temoin ignore ou s'oppose a celui du ministere public. Ces dissociations ont ete relevöes pendant le contre-interrogatoire, c'est a dire au moment oil le temoin est interrog6 par la partie adverse, les avocats de la defense. Pendant cette phase, les interventions du ministere public ont la fonction d'objecter k la ligne d'interrogation de la defense, dans le but, si l'objection est acceptee par le President, d'eviter au temoin de repondre. Dans ces moments-lä, le but du ministere public est done celui de sauvegarder le plus possible le temoin et son tdmoignage des attaques de la contrepartie. C'est justement pour cette raison que, dans ces occasions, les dissociations ou divergences du temoin prennent une importance particuliere. Elle fait exactement l'oppose de ce qu'il serait raisonnable de faire. Voici un exemple de ce comportement:

Renata Galatolo

216

Exemple 2 (Lipari 2 lignes 2007-2014) 2007 2008 2009 2010 2011 2012 2013 2014

Pe: prosegua L: e::[h LS: [ha giä risfposto L: [c- ci sono tavoli:: (1.5) mi- mi pare di ricordare: il posto in cui era seduto ci sono tavoli messi:: (..) eh

2007 2008 2009 2010 2011 2012 2013 2014 2015

Pe: continuez L: e::[h LS: [eile a deja re[pondu L: [il- il y a des tables:: (1.5) il- il me semble me rappeler l'endroit oü il 6tait assis il y a des tables mises:: (. .)eh

(La description continue jusqu'ä la ligne 2028) A la ligne 2010 le temoin donne sa reponse sans attendre le resultat de l'objection du ministere public (ligne 2009). L'effet de cette reponse, qui signal un alignement different du temoin et du ministere public, est celui d'annuler l'objection de ce dernier. La reponse du temoin, etant exactement ce que le ministere public voulait eviter par son objection, rend la replique du President superflue, que ce soit du reste une acceptation ou une refutation de l'objection. Du point de vue de la creation de l'effet de credibilite du temoin, ces comportements ont une double valeur. D'une part ils peuvent etre assumes comme la preuve de sa volonte de collaborer, qui est un aspect central, avec la capacite de certifier l'experience autobiographique (Dulong 1998, Sacks 1970), de la credibilite du temoin (Caesar-Wolf 1984). Dans le cas de ces comportements, cette volonte est d'autant plus evidente pour leur etrangete. D'autre part, ils sont la preuve d'une certaine competence communicative parce qu'ils demontrent la capacite du temoin de gerer la situation en pleine autonomic sans avoir besoin d'intermediate. Ces dissociations fonctionnent done comme des emblemes de l'identite extra-procedurale d'experte en droit du temoin et en tant que telles, ils constituent des instruments de changement de footing. Comme dans l'exemple precedent, le comportement du temoin n'est ni censure ni critique, d'autant plus que le ministere public a tout interet a eviter une pol6mique avec son temoin pour sauvegarder l'image d'ensemble de l'accusation.

Les strategies de changement de footing dans le temoignage commun au tribunal

217

5. Conclusions

A partir de l'analyse d'un temoignage particulier, j'ai montre certaines strategies grace auxquelles le temoin, parfois soutenu ou simplement non contraste par ceux qui l'interrogent, projette dans l'interaction son identite de personne juridiquement competente. L'analyse du comportement de madame Lipari m ' a permis de relever des strategies de changement de footing, de l'identite du temoin commun a l'identite d'expert en droit, dans la fa9on dont eile obtient le relächement des obligations sur la conduite du t0moin pendant le contre-interrogatoire et dans la fa9on dont eile refuse la mediation du ministere public dans l'interaction avec les autres parties. L'analyse s'est concentree sur les traits interactionnels, prise forcee du tour de parole dans le premier exemple et superposition dans le deuxieme, qui signalent les changements de footing. Les strategies analysees peuvent toutes etre ramenees ä une certaine competence interactionnelle du temoin grace ä laquelle eile controle et parfois force, de fapon evidemment acceptable (Bercelli 1998), le format rigide du contre-interrogatoire. Par rapport ä une analyse sociologique classique, qui aurait explique les "anomalies" de ce temoignage comme l'effet de l'influence des variables sociologiques sur l'interaction, entre autres le status socio-culturel du temoin, j'ai montre comment ce meme status est localement evoque et active par les pratiques communicatives des participants.

References bibliographiques

Antaki, C. and S. Widdicombe (eds.) (1998): Identities in Talk. - London: Sage. Bercelli, F. (1998): La gestione del frame "intende rispondere". - In: R. Galatolo, G. Pallotti (eds.) Di Pietro e il giudice, 21-41. Bologna: Pitagora Editrice. Bogen, D., M. Lynch (1989): Taking account of the hostile native: Plausibile deniability and the production of conversational history in the Iran-contra hearing. - Social Problems 36: 3, 197-224. Caesar-Wolf, B. (1984): The construction of the 'adjudicable' evidence in a West Germany civil hearing. - Text 4:1/3, 193-224. Conley, M. J., M. W. O'Barr (1990): Rules versus Relationships. The ethnography of legal discourse. - Chicago: The University of Chiacago Press. Dulong, R. (1998): Le temoin oculaire. Les conditions sociales de Pattestation personnelle. - Paris: Editions de l'Ecole des Hautes Etudes en Sciences Sociales. Galatolo, R., M. Mizzau (1999): Pertinenza e credibility nell'interazione in tribunale. Communication prdsent£e au Congresso Nazionale della Sezione di Psicologia Sperimentale. Capri Novantanove.

218

Renata Galatolo

- (2000): La metapercezione e la metacognizione quali risorse interazionali per il testimone oculare. Communication pr6sent6e au Congresso Nazionale della Sezione di Psicologia Sperimentale, Alghero Duemila. Goffman, E. (1981): Forms ofTalk. - Philadelphia: University of Pennsylvania Press. Hutchby, I., R. Wooffitt (eds.) (1998): Conversation Analysis. - Cambridge: Polity Press. Jackson, S. B. (1988): Narrative model in legal proof. - International Journal of the Semiotics of Law 1:3, 225-246. Jacquemet, M. (1996): Credibility in Court: Communicative practices in camorra trials. Cambridge: Cambridge University Press. Leonardi, P., M. Viaro (1983): Insubordinazioni. - In: F. Orletti (ed.) Comunicare nella vita quotidiana, 147-174. Bologna: II Mulino. Matoesian, G. (1999): The grammaticalization of participant roles in the constitution of expert identity. - Language in Society 28:4, 491-521. Mizzau, M. (1998): Risposte impertinenti. - In: R. Galatolo, G. Pallotti (eds.) Di Pietro e il giudice, 53-59. Bologna: Pitagora Editrice. Morrill, A. E. (1971): Trial Diplomacy. - Chicago: Court Practice Institute. O'Barr, M. W. (1982): Linguistic Evidence. - San Diego, New York, Boston: Academic Press. Pollner, Μ. (1987): Mundane Reason. Reality in everyday and sociological discourse. - Cambridge: Cambridge University Press. Sacks, H. (1970): Storyteller as 'witness': Entitlement to experience. - In: H. Sacks (1992) Lectures on Conversation. Vol 2, Jefferson, G. (ed.), Lecture 4 (Part IV) - Spring 1970, 242-248. Oxford: Blackwell. Spence, D. (1984): Narrative Truth and Historical Truth: Meaning and interpretation in psychoanalysis. - New York: Norton.

Conventions de transcription

Gras: indique les elements centrals pour 1'analyse MAJUSCULE: Augmentation de volume souligne: emphase particuliere >texte mais (disaccord) > A (argument) > oui, mais (desaccord partiel) > A > oui (accord).

Accords et disaccords: une dialectique communicative

275

A la fin de cette interaction, on entend une toute autre reflexion de Τ, participante la plus jeune, sur sa sensation du bonheur qui refute en quelque sorte tout ce qui a ete dit anterieurement. C'est evidemment un desaccord, sans toutefois qu'il soit exprime expressis verbis. Et puis, c'est un desaccord qui ne provoque pas de conflit, qui ne brise pas la communication, mais, au contraire, lui donne un nouvel elan, un nouvel objet de discussion, ä savoir "jeunesse/vieillesse". Cf.: (1) Τ - et encore le bonheur je pense / pour moi en tout cas ... le bonheur c'est quand c'est int6ressant quand c'est pas interessant alors quel dröle de bonheur c'est / quand tu sais tout [...] et tu t'ennuies mais le bonheur c'est quand c'est interessant: qu'est-ce qu'il y a done lä-bas / qu'est-ce qu'il y a derriere ce coin et pourquoi c'est comme ?a (rire) L - oui... oui quand eile disparait cette curiosit6-la / alors le desir de la vie disparatt aussi (2) A - la vieillesse arrive L - mais non on peut etre meme dans la vieillesse T[etre jeune L[etre jeune / s' interesser ä tout A - la vieillesse pas dans le sens de l'äge physique L - ouais, oui-oui (rire) c'est interessant

Le trajet dialogique de cette interaction differe de celui de l'interaction precedente: elle commence par une negociation d'opinion pour finir par une negociation sur le sens du mot vieillesse, sur sa valeur discursive. En effet, dans la sequence 2, la locutrice L n'est pas d'accord avec A, mais ce n'est pas un desaccord avec le fait que "la vieillesse arrive", c'est un desaccord avec l'idee implicite que la vieillesse et la curiosite sont incompatibles. La locutrice Τ la rejoint dans son desaccord. La participante A precise alors qu'elle avait en vue la vieillesse comme etat d'äme et non pas l'age physique. Cet exemple montre que les negociations sur les signes, aussi frequentes que les negociations des opinions, peuvent igalement tres bien aboutir k un accord entre les interactants. Le deroulement des echanges conversationnels presente une certaine dialectique, ce dernier terme etant compris au sens platonicien, e'est-a-dire comme un art de conduire le dialogue. Or la dialectique communicative implique, entre autres moyens, un enchainement d'accords et de disaccords qui se negocient constamment au cours de l'interaction. La communication s'avere reussie lorsque nous, les communicants, nous arrivons ä surmonter nos differends, ä transformer nos disaccords en accords, k nous rapprocher ensemble de la verite et, par consequent, a nous enrichir par une nouvelle vision du monde.

276

Ludmila Kastler

Reference bibliographiques

Auchlin, A. (1995): Le bonheur conversationnel: 6motion et cognition dans le discours et l'analyse du discours. - In: D.Veronique, R.Vion (eds.) Modeies de l'interaction verbale, 223-233. Aix-enProvence: Publications de l'Universite de Provence. Brown, P. and S. C. Levinson (1987): Politeness. Some universale in language usage. - Cambridge: University Press. Gofflnan, E. (1974): Les rites d'interaction, - Paris: Les Editions de Minuit. Jacques, F. (1979): Dialogiques. Recherches logiques sur le dialogue. - Paris: PUF. Jacques, F. (1988): Trois strategies interactionnelles: conversation, negotiation, dialogue. - In: J. Cosnier, N. Gelas, C. Kerbrat-Orecchioni (eds.) Echanges sur la conversation, 45-68. Paris: Editions du CNRS. Kerbrat-Orecchioni, C. (1984): Les ndgociations conversationnelles. - Verbum VII/2-3, 223-243. Kerbrat-Orecchioni, C. (1996): La conversation. - Paris: Seuil. Zeldin, T. (1999): De la conversation. - Paris: Fayard.

Jiri Kraus

Argumentation in Czech Political Debates

One part of a grant-funded projected organised in the Ustav pro jazyk öesky (Czech Language Institute of the Academy of Sciences in Prague) is the linguistic analysis of television debates involving Czech politicians who express their opinions on topical issues in regular programmes chaired by presenters. My paper focuses on the methodological principles of this analysis and also attempts to make some of the terminology more precise. The different elements of the specific communication situation of the institutional discourse that these debates are generating will be in the forefront of my interest. The language, style and mode of argumentation in these debates have been attracting considerable attention from television viewers, because they reflect fundamental changes that have occurred in Czech political discourse since 1989. The politicians who appear on these programmes have to engage effectively with the views of their opponents on television in the short time allocated for their answers, and must at the same time try to interest and convince a very diverse TV audience. In their mode of speech they combine various different techniques in different styles: academic, journalistic, oratorical, administrative, colloquial, artistic etc. While for the most part it is the literary language that is employed, with the common spoken language or elements of regional dialects appearing among certain speakers and in some situations (usually when there is a strong personal and emotional element), speakers often deliberately or unconsciously shift from one variety of speech to another (Eckert 1993). It is typical of the Czech language situation that the culture of verbal expressions used in official contexts (and especially in the mass media) is usually critically assessed, and so the choice of appropriate language variety is a major factor in the success or persuasiveness of the communication acts. From the point of view of the television public it is possible to distinguish two types of communication situation (cf. Charaudeau 1992, Maingueneau 1996). The participants in open situations take up the permanent roles of competent speaker (in a given case the guest in the television studio and the anchorman), and of viewer, who is expecting a particular kind of information and the reasoned conformation or rejection of particular standpoints. In open situations the authors of the speech acts therefore generally address the auditorium drawing on contrasting communicational habits and experiences. In closed situations, however, those who appear usually become real or potential authors of discourses of the same kind. The field of mass media, academic or creative literature, administrative regulations, legislative texts and such like therefore covers open situations,

278

Jifi

Kraus

characterised by contrasting communicational habits and experiences in speaker and listener, or writer and reader. Conversations in the family and the workplace, or in a wide range of contexts inspiring friendly and unofficial conversations, by contrast, are characterised by the properties of closed language situations. An essential prerequisite of the strategy of television debates and polemics, usually focused on a small circle of precisely defined themes, is therefore the need to respect the anticipated level of knowledge, values and communication habits of the broadest possible audience. One prerequisite of the linguistic analysis of debates is to identify - difficult though the task may be - the communicational intentions of the active participants in the debates and of the television viewers. What is expressed in the intention is the participant's desire for his partner to reach a certain conclusion, take up a certain point of view or act in a particular way. Communicational intentions are also expressed, however, by the receiving side in the form of expectations that may or may not be fulfilled. (As a result the viewer may carry on watching without much enthusiasm, or his attention may sharpen, or he may turn off the TV). As the distance between the author and receiver of a communication increases - in terms of time, space, generation or social position - so too does the likelihood of conflict between their communicational intentions. This conflict can be interpreted as misunderstanding or as either deliberate or unwitting distortion of the point of the discourse. The greater the difference in space and time, the freer - but also the more difficult - the interpretation of the discourse. The definition of the communicational situation includes not only the concrete form of its elements, such as language, form of argumentation, atmosphere in the studio, placing of its participants around the table (proxemic of the partners) etc., but also assessments of its future course. For each participant this becomes an important part of his communicational intention, in accordance with Thomas's theorem that "even unrealistic assessments of communicational situations have real consequences" (Thomas 1965). For example if one of the participants believes, even mistakenly, that the debate is going to become more heated, his replies will begin to take on a sharp and polemic character. This then leads to changes in the real behaviour of his speech partner, for example to a real sharpening of the debate. Analysis of the type of debate under discussion also brings up the need for terminological distinction between the concepts of text and discourse. I define text as a coherent concatenation of sentences, i.e. language beyond sentences), consequently as an abstract linguistic category on the level oilangue. By contrast I regard discourse as text possessing a concretising attribute that predetermines the possibilities of its interpretation. I thus distinguish, for example, between political, ideological, religious, liberal and socialist discourse. In addition there exist discourses of different social groups, such as doctors, economists, financiers, or sportsmen, and discourses of important people, such as writers or politicians.

Argumentation in Czech Political Debates

279

Thatcherian discourse, or the discourses of de Gaulle, Churchill, Gorbatchev or Vaclav Havel have been the subject of many analyses (cf. e.g. Hartman 1995, Kusse 1996). The rhetorical discourse of the latter, for example, who is both Czech president and a renowned playwright, is characterised by sophisticated compositional structure and metaphors using commonplaces from biblical, philosophical and canonical literary texts. The concrete nature of the concept of discourse makes it possible to classify its various types. Hence we can distinguish between monologic and dialogic discourse, written and spoken, prepared and spontaneous, context-bound and context free, discourses according to different genres, functional styles and registers. In addition, in Czech and also French stylistics, we see the classification of discourses by stylistic sequences into the descriptive, narrative, argumentative, explicative and reflective (Mainguenau 1996). Finally, in relation to the philosophic and literary theoretical question of possible worlds, a distinction is made between discourses that have a real or a Active referential framework. Since participation in communication by means of natural language is one of the basic defining attributes of the human being, human life is played out in the sphere of repeating types of discourse that the individual more or less makes his or her own. It follows from this that the process of stylisation and interpretation is often based on higher unities than the word alone. Consequently, both the author and the interpreter are discoverers or rather retrievers of pre-existing de-authorised text segments representing spatial, temporal, social and other contexts shared by common (wo)men. They enable the user to select from various kinds of intertextuality and interdiscursivity enforcing and reinforcing each other (cf. Plett 1999). While literary theory, especially from the time of E.R. Curtius and Mikhail Bakhtin, has devoted systematic attention to these intertextual relationships and the field of commonplaces, analysis of media and argumentational discourse has tended to consider such questions only in the negative terms of language cliches and stereotypes. In fact, however, these repeating text segments, argumentational models and other conventional expressions play a constant role of stabilisers of mutual understanding. Because they confirm the familiar, and offer release from the unexpected and the threatening, they function as open or hidden means of persuasion as well as eristic manipulation. In the last ten years the inventory of sources of intertextuality in Czech media discourse has undergone fundamental transformations. Ever more frequently we encounter allusions to the Bible, classical literature and philosophy, minority cultures, and also to modern science and technology - especially in the field of computers, artificial intelligence, sociology, psychology, economics and financial operations. This naturally increases the requirements necessary for the understanding of the text and for effective replies in polemical debates, and opens up new possibilities for the use of irony, paradox and wordplay. Intertextuality also plays a major role in the dialogic character of any discourse, which we can only perceive as a part of a broader context. For this reason it is useful to distin-

280

Jiri Kraus

guish between real dialogue, which is interlocutive and involves several speakers, and virtual dialogue, which is interactive and based on interactual relationships in the framework of a single text (Catherine Kerbrat-Orecchioni has called the former type of dialogue dialogal and the latter dialogic.) In real dialogue, and above all in dialogue of the argumentational type, each reply is therefore a part of both dialogal and dialogic discourse. Argumentation is a typical means of expression of the intertextuality of the discourse of television debates on political themes. Here I understand argumentation to mean a mode of presenting the statement or set of statements Vi, whose function is to ensure that the partner in communication should regard a certain claim V 2 as acceptable, justified and convincing. The relation between Vj and V 2 assumes a logical inference, the explicit expression of which consists in connectors (i.e. conjunctions, particles etc.). Polemic argumentation takes its methodological apparatus and often its terminology as well from classical rhetoric, as a discipline which taught the consideration of a problem from several different and usually controversial points of view (Mainguenau 1983, Kraus 1998). The essential semantic functions of connectors provide the basis for the accumulation of pragmatic functions. These normally appear at the beginning of replies in the dialogue and express reaction to the previous reply of the communicational partner or to the whole complex of his previous replies. This reaction normally takes the form of agreement, disagreement, concession, correction, clarification (precision), development using examples, etc. In the following sample from a debate between the Czech Prime minister MiloS Zeman and the chairman of parliament Vaclav Klaus. I shall illustrate the semantic and pragmatic functions of connecting expressions with an analysis of the conjunction but (in Czech 'ale'). The frequent use of this conjunction at the beginning of the moves of both partners in the polemic dialogue eloquently characterises the polarity of their attitudes to the question of the level of state intervention desirable in the economy. As a representative of the political right, the chairman of parliament Vaclav Klaus denies that such intervention is effective, while the Prime minister Milos Zeman, a social democrat, wishes to use such intervention as an essential part of government legislation. Within the moves the conjunction but is employed in the adversative sense, and at the beginning of moves it is employed as a signal of what is often very general disagreement with the standpoint of the partner. Many researchers have already noted the multiple meanings and pragmatic functions of the conjunction but. As Lakoff (1971) puts it, "but has many properties of and plus some complexities of its own". Among these complexities we find, for example, the semantic opposition but, the denial-of-expectation but and various others. Milos Zeman: I think that the point of the legislative is not to ensure that civil servants can interfere in the private affairs of the citizen in any situation and in a completely uncontrolled way. Nobody except extremists wants that.

Argumentation in Czech Political Debates

281

Vaclav Klaus: But what concerns me is the authority of the state administration (correction - not P, but Q). Miloä Zeman: But where we differ is that in my view government bodies have to fulfill two basic functions in society, the supervisory and the repression... And here we're discussing where the line between the two functions should be drawn (precision - distinguo, not exactly P, but P). Vaclav Klaus: We're discussing where the line should be drawn in both cases (confirmation correction). Miloä Zeman: But since we're now discussing the repressive function (distinguo, not all P's but only particular case of Ρ), I differ from you in asserting that in the specific case of economic crime the state has the right to use the financial authorities to ascertain whether a particular person has acquired his property legally... In many European countries property tax is connected with tax declaration. But we don't want anything of that kind (adversative), we only want to know whether a person acquire the property honestly or not. Vaclav Klaus: But let's not use populist words (refutation). In every legal system it is the courts that decide to charge a particular person, or to grant a license for his house to be searched. You seem to want house searches to be done by the local house supervisor again (note: as in communist times). Miloä Zeman: But what do you mean, house supervisors? (refutation - exclamation) Vaclav Klaus: Well, almost like house supervisors (concession). MiloS Zeman: Financial offices, Mr. Chairman. Financial offices (confirmation of speaker's previous assertion).

The connectors at the beginning of the moves create the logical axis for the semantic and pragmatic analysis of the polemic discourse. It is from these that one can go on to develop the analysis of the full meaning of the other words employed, for example in this case the value-laden adjectives (honest, populist), the hyperbolic comparisons (e.g. here the comparison of the financial offices to house supervisors as informers under the former communist regime) etc. The follow-up functions of the connectors thus stands at the crossover between the syntax and composition of the text, and their subtle semantics express the relation between the claim and its negation, the thesis and antithesis, the specific statement and its generalisation, between the general statement and its exemplification, between the premises, warrants and conclusion of the argument, between the question and the answer, and so forth. The fact that the connectors often appear at the beginning of the moves contributes to the overall coherence of the dialogue as a single texts created from two or more speakers. By means of the connector or other means of textual coherence, therefore, each move can follow on in part from the previous move of the speaker himself, and in part from the preceding move of his partner in the dialogue. In the former case we can speak of vertical coherence (coherence in the succession of moves given by one speaker), and in the latter of horizontal coherence (i.e. coherence between the contact and distant moves of Speaker A and Speaker B). A higher degree of horizontal coherence also reflects a higher degree of readiness on the part of the participant in the dialogue to listen to his/her partner.

282

Jiri Kraus

References

Charaudeau, Ρ (1989): La conversation entre le situationnel et linguistique. - Connexions, pp. 9-12. Eckert, Ε. (1993) (ed.): Studies in Czech Sociolinguistics - Amsterdam, Atlanta. Hariman, R. (1995): Political Style. The artistry of power. - Chicago, London: University of Chicaga Press. Kerbrat-Orecchioni, C. (1986): L'implicite. - Paris: Armand Colin. Kraus, J. (1998): Retorika ν dejinäch evropske kultury (Rhetoric in the History of European Culture). Praha. Kusse, Η. (1998): Konjunktionale Koordination in Predigten und politischen Reden. Dargestellt an Belegen aus dem Russischen. München. Lakoff, R. (1971): If s and but's. About conjunction. - In: Ch. J. Fillmore, D. T. Langendoen (eds) Studies in Linguistic Semantics, 114-147. New York. Mainguenau, D. (1983): Semantique de lapolemique. Lausanne. Mainguenau, D. (1996): Les termes cles de l'analyse du discours. Paris. Plett, H. F. (1999): Rhetoric and intertextuality. - Rhetorica 17 (3), 313-329. Thomas, W. (1965): Person und Sozialverhalten. - Neuwied: Luchterhand.

Gerda Lauerbach Opting Out of the Media-Politics Contract Discourse Practices in Confrontational Television Interviews

1. Introduction

Although they do not occur frequently, interviews that go off the track are interesting for "deviant case analysis" since the constraints of a genre, a discoursal domain and a social context are shown up particularly clearly in their breach. The aim of this paper is to analyse some of the discourse practices employed by interviewers and interviewees in such encounters where the dynamic of escalation culminates in the interviewee's opting out. I will start by clarifying the terms "opting out" and "media politics contract" used in the title and then present an analysis of some of the practices employed. The analysis will focus on the properties of escalation sequences, on the practices employed by interviewees in opting out and on the ways in which the interviewers deal with this.

2. Opting out of the media-politics contract In liberal democracies, when politicians are interviewed live on news or current affairs programmes in order to present, justify and defend their policies, they are bound to be confronted with questions that they would rather not answer on topics that they would rather avoid. This is so for two reasons: first, there are no safe issues to be interviewed on: newsworthy issues are controversial issues'; second, the people with whom they are controversial - a divided electorate - are part of the mass media audience to which the interviews are directed. This again has two consequences: first, since interviewers ask the questions that the audience expects them to ask (Hall 1972), at least some adversarial questions are to be expected. And second, since the audience will be divided on the issue, there will be no single good answer that will satisfy every one. Some answers may even be damaging to the politicians themselves or to their parties. So the situation has all the fea1

The degree to which this is true may be subject to cultural variation. I am focussing here on Great Britain.

284

Gerda Lauerbach

tures of a dilemma, for not only is there no solution to the problem that does not also have severe drawbacks, but also there is no way out, no possibility to withdraw: in our media democracies politicians have to be visible in the public sphere in order to be able to stay in or compete for political power. Moreover, a refusal to be interviewed would also have quite specific negative consequences: it would be reported as a newsworthy event in its own right and would invite the inference that the politician concerned has something to hide. However, contrary to everyday opinion, dilemmas are not the exception but a regular feature of social life, and social practices have evolved to provide routine compromise solutions to them. For the modern media interview, this means that politicians regularly and in predictable ways equivocate, evade direct answers, try to change the interviewer's agenda, communicate implicitly (cf. Bavelas et al. 1988, Harris 1991, Greatbatch 1986, Lauerbach 2000b). Interviewers, for their part, have regular procedures to press for more explicitness, to repeat and reformulate questions (cf. Heritage/Watson 1979; Heritage/Roth 1995, Blum-Kulka 1983). This may be met by more evasion, etc., which may in tum evoke more pressure from the interviewer, and so on. It is at the end of such sequences that what is referred to here as "opting out" sometimes occurs. Opting out is of course a Gricean term - within Grice's framework of conversational logic it is one of the four ways in which conversationalists may infringe the Cooperative Principle (CP) and/or the four maxims of quality, quantity, relevance and manner: they may quietly VIOLATE a maxim (which will lead to deception), they may be faced with a CLASH (which will lead to compromise solutions that act as inference triggers), they may blatantly FLOUT a maxim (which will indicate non-literal speech - irony, metaphor and so on), or they may OPT OUT of both the maxims and the CP: a conversationalist may say, indicate or allow it to become plain that he is unwilling to cooperate in the way the maxim requires. He may say, for example, "I cannot say more; my lips are sealed" (Grice 1975:49). While in compromising between clashing maxims and in blatant floutings, the CP is still being observed, it is being surreptitiously circumvented in deception and explicitly or implicitly suspended in opting out. What precisely is it that interviewed politicians opt out of? It may be, at the micro-level of the interview, the response to a specific question, it may be the contribution to a topical sequence, or it may be the continued cooperation with the interview as a whole. The approach to discourse analysis followed here is a constructivist one - that the micro-level of discursive interaction is constitutive of macro-level domains of social organization and vice versa. The frequency with which opting out occurs at the discourse level and the degree of seriousness and aggravation with which it is done will therefore say something about power relations at higher levels of social organization. In the case of the political interview, it will point to the quality of the media-politics contract.

Opting Out of the Media-Politics

Contract

285

This "contract" can, somewhat flippantly, be described as a tacit agreement according to which politics provides the media with actors and a plot, and the media provide politics with a stage, a cast to perform with, a director and an audience. In public service television, the audience pays for the performance in the form of licence fees (whether they watch the performance or not), in commercial television it looks as if the audience gets to watch for free. However, they have been sold to the advertisers, and the price they pay is to have to watch the commercials. The question is whether the tendency towards entertainment and trivialization, ascribed to the commercial imperative but affecting private and public television alike, also affects the media-politics contract. Since this tendency can be discerned also in serious formats like news and current affairs reporting, there is no reason to assume that political programmes will be spared. But at least the media-politics contract is expected to serve both sides equally well: by having political programmes, the public service media fulfil part of the public remit to broadcast a programme-mix of information, entertainment and education, which private media, at least up to the arrival of theme channels (news, sports, music, etc.), have kept to as well. For both public and private broadcasting media newsworthy and prominent political interviewees mean prestige and high audience ratings. Also, for all types of channels, politicians as media actors mean low production costs because they perform for free. For politics, on the other hand, the media provide the only opportunity of public visibility of its actors that counts, also the opportunity to publicly present and argue for policies, or to counter the arguments of the opposition, in short to enter into the kind of staged and mediated dialogue that counts as the discourse of the public sphere. But to say that both sides profit from this tacit agreement does not necessarily mean that the relations of power between them are unproblematic. Looking at the political interview, it is quite obvious that the party whose institutional role it is to ask the questions, to set the agenda and to determine the time-place coordinates is in a position of discoursal dominance. But also on the level of social formations, the "watchdog"-function of the media over politics implies a relation of legitimate control, whereas the suspicion of ideological bias addresses the possibility of the media manipulation. Control is however also exerted in the other direction - by politics granting access to information unevenly to favoured journalists, by carefully selecting politicians for interview, and by imposing a "spin" on the information that does get disseminated. So while there is indeed an area where the interests of politics and the media intersect to the point of mutual dependency, the same site is also an arena of struggle where conflicting interests are contested over. Normally, the orientation to mutual interests dominates. It is only when that orientation is momentarily suspended that opting out will occur.

286

Gerda Lauerbach

3. Discourse practices in confrontational political interviews

In 1989, the journal Research on Language and Social Interaction devoted a special issue to the analysis of an interview the CBS moderator Dan Rather did with the then American Vice President and presidential candidate George Bush, in the run-up to the presidential elections. The interview became itself news for the way in which it turned into a heated argument that, in the ensuing discussion, was commonly described as a shouting match. The analyses showed how the stepwise relaxation of the participants' orientation to the interview framework regarding turn-taking, turn-type and turn-space constraints turned the encounter from interview to confrontation (Schegloff, Clayman/Whalen), how devices for constructing audience scepticism were used to confront the interviewee with damaging contradictions and inconsistencies (Pomerantz), and how interviewer and interviewee contended over what the interview's talk should be about (Nofsinger). While the data I will present below are not as dramatic as the Bush/Rather interview, they exhibit a similar dynamic of escalation that progressively leads up to one or both partners opting out of the interview framework. The following excerpts are taken from the television coverage of the British election night 1997, specifically from interviews done by star presenters of the BBC and ITV with prominent politicians of the Conservative and the Labour Party. This was the election in which New Labour under the leadership of Tony Blair won a landslide victory over the Tories under the leadership of John Major. In excerpt one, Jeremy Paxman of the BBC interviews Malcolm Rifkind, Foreign Secretary in the outgoing Tory government, a good hour after the polls have closed. The exit polls predict a landslide for Labour, but the "real results", i.e. the ballot counts from the constituencies, are not yet in. Before the excerpt starts, the interviewer (IR) had asked the interviewee (IE): "How bad is it for YOU, Mr. Rifkind? ", and had followed this up with a question about the causes for the defeat. Both questions had been responded to with routine indirectness. Excerpt 1:

Jeremy Paxman (BBC) interviews Malcolm Rifkind,Conservatives, Foreign Secretary2

IR 3QU IE 3R

Are you confident you'll hang on to your own seat? I'm extremely hopeful, some (.) quite encouraging signs but I (.) frankly don't know any more than you do at this stage, I would be guessing=

IR=interviewer, IE=interviewee; QU= IR question; R=IE response; CH=IR challenging aspects of IE's response. For a description of the obligatory and optional moves, exchanges and sequences of the generic structure potential of news interviews, cf. Lauerbach 2000a-c.

Opting Out of the Media-Politics Contract IR 4CH IE 4R

IR 5QU

287

Being hopeful isn't quite the same as being confident, is it? Well, it's because you're asking me to predict, I mean are you just wanting me to use the usual optimism that all candidates use or are you trying to find out if I have some (.) PRIVILEGED access to information? I haven't counted the ballots yet, so I don't ACTUALLY know any more than you do, but I'm very happy to sound optimistic. How long do you think John Major will stay as leader?

The excerpt starts with the question Are you confident you ΊΙ hang on to your own seat? which opens the third exchange. This question is not answered with a straight yes or no but with a toning down of the predicate of the question: from confident to extremely

hopeful.

The subsequent 4CH is a classical IR move that asks IE to made an inference explicit (cf. Blum-Kulka 1983). The inference concerned is a generalized conversational inference: if items can be ranged on a scale, then to use an item lower down the scale implicates that the negation of the ones higher on the scale holds (cf. Atlas/Levinson 1981, Levinson 1983). The scale here is something like hopeful - confident - certain, so that hopeful implicates not confident, and it is this inference that IR is trying to get IE to confirm. A cooperative answer to 4CH would be, no, it isn 't, I'm not confident. This is strategically highly undesirable. The agreed strategy of the conservative politicians during the election night was: avoid talk about the results until they are definite, above all don't admit defeat. IE cannot do both - follow that strategy and observe the social norm of answering the question. In 4R he opts out by offering an account for not answering the question, an account that focuses on the interview itself and takes the form of mildly ironical criticism of the sort of empty exchange that goes on air before the facts are in, and also of the stereotypical optimism that is habitually exuded by candidates that are about to lose. This criticism is couched in a rhetorical question that does not expect an answer and it invokes norms of media interaction and of political communication that normally implicitly guide political interviews. I R ' s reaction to this is to ignore it and go on to the next question, leaving the audience to draw the inference that IE is not confident that he will keep his seat (in fact, he lost it). In excerpt 2, Jonathan Dimbleby of ITV interviews Gordon Brown, Labour, Shadow Chancellor about to become Chancellor of the Exchequer in the new government. Before the excerpt starts, IR has asked IE about his party's victory: is he pleased about it, is he surprised, does it instil as sense of fear in him that Labour now has to deliver all the things they promised in the election campaign. After an interruption by footage of Tony Blair leaving his house for his count, the IR focuses on the budget: Excerpt 2:

Jonathan Dimbleby (ITV) interviews Gordon Brown, Labour, Shadow Chancellor

IR 5QU

Uhm you heard us describing the uhm/Tony Blair's journeys, while we pause for a moment, let me please ask you this (.) uhm we've got a pretty good time-table there

288

IE 5R

IR 6F0 3

IE 6aR IR 7aCH IE 6bR IR 7bCH IE 7R

IR 8CL4 IE 8CL

Gerda Lauerbach from George Robertson for devolution, when are you going to have your budget, we know it is going to be quick, when is it going to be, can you tell us? Well, I'd say it will be within two months, and uh it is a budget to equip this country for the future, the welfare to work budget uh that will get young people and the long-term unemployed back to work, we fully intend to keep all our promises (.) and we will move very quickly on that budget And does that mean, when you say keep all your promises, you definitely won't be putting up uhm the rate of income tax, you definitely will introduce the windfall tax and you'll leave open the option of introducing other forms of taxation as well in this budget, is this correct? Look, what I did there Jonathan was I made four commitments, one... ((4 lines omitted)) and fourth/and fourthly we'll introduce there'll be/if there'll be any increases they'll be elsewhere yes? and fourthly we'll introduce the windfall levy on the utilities now these are promises I will keep HANG ON/HANG ON A SECOND/hang on a second, Mr. Brown, any increases therefore will be elsewhere other than there if you have to introduce them No, I'm NOT writing a budget on the programme, but I'm telling you these are the commitments we've made, and we fully intend to keep them, I/I think the most important thing you know is building a new trust (.) between the government and the people ... ((5 lines omitted)) Mr.Brown, soon to be Chancellor Thank you

The opting out utterance to focus on here is 7R No, I'm not writing a budget on the programme. This utterance has an interesting sequential history: even before the excerpt starts, IR had brought the question of keeping election promises into play. Exchange 5 then starts harmlessly enough with a question about the timing of the budget, but is followed up by a formulation 6FO that tries to clarify if the promises mean that there will be no tax increases. IE responds to this in 6aR by shifting the agenda and reiterating at length the commitments made during the campaign. It is helpful here to recapitulate Greatbatch's (1986) analysis of Ies' agenda-shifting procedures. He distinguishes three types: (1) postanswer agenda-shifting, (2) pre-answer agenda-shifting and (3) shifting the agenda instead of answering. (3) is a serious departure from the interview framework and is most frequently sanctioned by IRs, while (1) is sanctioned least frequently. Returning to excerpt (2), by the time IR interrupts and overlaps IE with 7aCH, it is unclear whether a number (3) or a number (2) procedure is in progress, so the intrusion of IR into IE's turn-space cannot be considered a violation of the framework but is an implicit sanctioning move designed to cut IE short in order to keep the framework in operation. 6FO does not receive a

4

FO= IR move checking meaning, or summing up gist or upshot so far, cf. Heritage/Watson (1979). CL= closing move, initiated by IR, optionally reciprocated by IE. Stretches of talk produced in overlap are in italics.

Opting Out of the Media-Politics

Contract

289

direct answer for even when IE sums up at the end of 6bR: these are the promises I will keep, a possible answer can only be inferred. 7bCH is now doubly a move towards explicitness: not only does IR's emphatic HANG ON/HANG ON explicitly sanction IE's agenda shift and prolonged occupation of the turn, but the second half of the move also makes explicit the inference that arises from not answering 6FO: any increases will therefore be elsewhere etc. The inference arises on the basis of two interacting systems of conversational organization: conditional relevance and preference organization. Conditional relevance provides for the normal expectation of an answer after a question, preference organization favours a contiguous and brief POSITIVE answer to a yes-no question like 6FO. If no such answer is forthcoming, an inference to the negative answer is invited. In this case the negative answer would mean that Labour will put up the rate of income tax etc. This is the inference that IR tries to get IE to confirm in 7bCH, and by doing so confronts him with a dilemma: For again IR cannot do both: provide the answer required by the norms of the interview framework (i.e. confirm that there will be tax increases - for to do so would discredit the election promises), nor can he commit himself to not increasing taxes because he cannot be sure that he will be able to keep such a promise. He solves the problem by opting out: His No is an explicit refusal to answer what turns out to be the final question in the budget sequence; and his I'm NOT writing a budget on the programme is the account for that refusal, in which he invokes a norm of political practice and also implicitly criticises IR. Let us briefly summarize the sequential build-up to this refusal: A question is responded to by an agenda shift, this is first implicitly sanctioned by the IR's cutting in, then explicitly by IR asking IE to stop, the original question is then pursued by a procedure through which IR asks IE to confirm an inference that followed from his non-answer - and IE opts out. It is also interesting to see that of the norms that normally implicitly underlie the genre of political interview, IR and IE each explicitly invoke those that govern their own professional domain: IR relies on norms of interactional logic, IE on norms of a logic of political action. Finally, as in excerpt 1, IR lets IE's response stand for the audience to draw their own inferences, and gets on with the business of the interview, here by initiating the closing exchange. The last example presents a more dramatic instance of opting out. It is taken from an interview that BBC's Jeremy Paxman did with Michael Portillo, then Secretary of Defence in the outgoing Tory government, minutes after Portillo had lost his seat. This defeat caused great excitement since Portillo had been a strong contender for the leadership of the Conservative Party after the expected resignation of John Major. So the development in his constituency had been followed with great media interest, and Paxman has in fact done two interviews with Portillo already before this final interview starts. Before excerpt 3 begins, there is an opening sequence and a non-controversial pre-sequence in which Paxman

290

Gerda Lauerbach

asks how Portillo feels about losing his seat and what he attributes the swing to Labour to. The core-sequence o f the interview begins then with 4 Q U about the leadership: Excerpt 3.

Jeremy Paxman, BBC, interviews Michael Portillo, Conservatives, Secretary of Defence, shortly after he has lost his seat

IR 4QU IE 4R

Do you regret that you didn't stand for the leadership yourself?, Uh no, I don't actually, I/I thought John Major was the right man to carry the leadership of the party, (.5) um (1.0) and (now it isuh/it's come to what we see today (smiling)) uh and I will do what I can to work from the outside to(.) to rebuild what is a great party IR 5CH How CAN the party unite around the present policy on the single currency? IE 5R Well I'm now a man outside the House of Commons=I don't have to bother with questions like that IR 6CH I thought you just told the people in your declaration that you were going to work in any capacity you could to further the Party! IE 6R ((smiling, shaking head)) Jeremy, I'm taking an evening off. Listen, I will do what I can to bring unity to the party (.5) but I'm gonna do it by talking to the party, not by talking to you. IR 7CH Sure but you/it follows from that, doesn't it, that you believe you CANnot have unity around the PREsent policy. IE 7R Uh it follows from that that the party needs to reflect very carefully about how it's gonna move forward etc.. IR 8CH Is the present policy the right policy? IE 8R Oh JeremyNDo stop this nonsense! IR 9PC ((laughs)) (well (.5) I think we'll have to stop the interview (is smiling, shakes head)) unless you have anything else you particularly want to tell us this evening IE 9PC (It was you who wanted the interview (smiling broadly)) IR 10PC You do s/this is your last (opportunity, Mr. Portillo (chuckling)) IE 1OPC (Oh, not necessarily, we shall see (smiling)) IR 11 PC You seem very good-humoured for a chap who has lost his seat and his chance of leading the Party IE 1 IPC Well, I think it's uh/it's inherent in politics that you have to be prepared for ups and downs IR 12CL Okay. Mr. Portillo, thanks very much for joining us (.5) IE 12CL Thanks very much. IR 13TR5 David! PRES TR13 We go straight to Exeter for the declaration. This core sequence and the following pre-closing sequence are the locus o f three optings out by IE (in R5, R6 and R8) and one by IR (PC9). Obviously, since the interview keeps going after the first and the second and the third IE opt-out, it is a matter o f interest to look not only at the procedures which precede the IE's opt-outs, but also at how IR manages to contain IE within the interview framework - until he finally sees fit to suspend it himself

TR= transitional move. PRES= presenter, anchor in the studio.

Opting Out of the Media-Politics

Contract

291

in PC9. A detailed analysis of this interview can be found in Lauerbach (2000c), so I will concentrate here on particularly interesting aspects of opting out. Let us begin with R5. In contrast to what was said above about the sequential build-up to opting out moves, this IR refusal comes veiy early in the interview's history and its effect is extremely strong and abrasive not only because of what is said, and of how it is said, but also because of its early positioning. The only trigger that can be found in the history of interview itself is CH5: How CAN the Party unite around the present policy on the single currency? This is a rhetorical question the import of which is that the party cannot unite around the present policy. While this is a severe criticism of the Party, it acquires provocative force for IR only if we make a number of assumptions: that the Tories' present policy on the single currency is inherently divisive, that IR is partly responsible for it and that it was through being divided that the Conservative Party lost the election. Anyone who has watched the BBC's election night programme knows these things because they were controversially discussed in the first Paxman-Portillo interview. So the trigger and sequential build-up to this early opting out lies outside the speech event in which it occurs. This is presumably a much more frequent contextual constellation than one commonly thinks and attests to the essential intertextuality of media discourse. Again, IE is in a dilemma: he cannot give the preferred answer to 5CH: no, the Party cannot unite around the present policy, because thereby he would discredit his previous position and take partial responsibility for the defeat, nor can he say, yes it can unite around the present policy, because he would have to do the impossible and explain how it could now if it could not before. So it is best to opt out. R5 is interesting for another reason: like 7R in excerpt 2, it invokes a norm of political discourse in the media as an account for opting out, one referring to what it takes to be a properly legitimated interviewee (be a Member of Parliament). IR challenges this in 6CH by pointing to IE's inconsistency between refusing to talk on policy and just having publicly announced (in R4 and in his declaration speech a few minutes ago) that he would do everything to help the Party from the outside. This forces IE back into the interview framework. While in the first part of R6 he insists on "staying out" (taking an evening o f f ) , the account he offers for this in the second part deals with the inconsistency. Again this is done by invoking a norm of political strategy as an account: if in trouble, talk to the Party first, not to the media. The following CH7 is interesting because in it IR invokes, against IE's norm of the logic of political action, the norms of interactional logic: He makes explicit an inference that again arises from the joint operation of conditional relevance and preference organization: the inference that IR does not believe that you can have unity around the present policy. This is countered by IE in R7 with a further appeal to political logic, from which different conclusions follow - but which leaves the question unanswered.

292

Gerda Lauerbach

So the question first asked in 5CH is asked once more in CH8, and this provokes the final - half-serious, half joking - IE opt-out in R8. This time, IR does not attempt to reign IE back in - on the contrary, he starts his own opting-out sequence in PC9, which is a hybrid pre-closing move that in its first part exerts media power by threatening to close the interview and in its second part yields this power by offering IE the floor to bring up a topic of his own. So the interview framework has been left explicitly in the first half of the IR turn by meta-talk, and implicitly in the second by veering towards the kind of organization typical for conversational closings, and the one who initiates this is IR himself. IE implicitly declines the offer of the floor by continuing meta-talk on the interview: It was you who wanted the interview implies: not me, and implicitly readjusts the power relation. After some interpersonal work by IR the interview is closed by IR and IE by a return to formal interview procedures that keep the channel open for possible future encounters and reaffirm the media-politics contract. The dynamic in this sequence is one in which there is not only a move towards more explicitness, as in the other two excerpts, but also one towards conversation in which the constraints of the interview framework regarding turn-taking and turn-type are briefly suspended. A corollary of this movement is that facework becomes much more relevant. While in the political interview, adversarial question are to be expected and do not necessarily call for mitigation, this changes if the interview moves towards conversation. Judged on the verbal record alone, this interview sounds very abrasive and aggressive. On the nonverbal channel however, there is a lot of smiling and laughter that indicates non-serious banter and softens and even counteracts what is happening on the verbal level.

4. Concluding remarks

Opting out is a complex phenomenon that can range over a scale of domains: turn, topical sequence, the speech event as a whole. It can also be graded regarding its degree of seriousness and aggravation. It is a conversational practice that in political interviews is predominantly employed by the interviewees, as a last resort when confronted with a communicative dilemma. Opting out thus occurs at the end of a sequential build-up or escalation. This build-up can have its source and pre-history also in intertextual relations outside of the speech event. Two dynamics are involved: one, a stepwise relaxation of the constraints of the interview framework, occasioning a movement towards conversation. Two, a movement from implicitness to explicitness. This can be directed to the negotiation of the topics or to the organization of the framework itself. In topical talk, one kind of movement

Opting Out of the Media-Politics

Contract

293

towards explicitness is driven by the interviewer, directed at the clarification of meaning and typically takes the form of asking the interviewee to make inferences explicit. In this, IR relies on and often makes explicit the norms of an interactional logic. This same logic is also employed when the IR tries to coax or coerce IE back into the interview framework after IE has opted out. When IEs opt out, this is typically mitigated by accounts that explicitly invoke norms o f political practice, strategy and logic. This has a theoretically important consequence: since genre-specific constraints determine what can be advanced as an account for infringing them, it follows that the norms o f a practical logic o f political action must be considered just as much a constitutive part o f the political interview genre as the norms o f conversational or interactional logic. This is not something that is commonly appreciated. In talk directed to the maintenance of the interview framework itself, the movement from the implicit to the explicit can be seen for example in the progression from implicit to explicit IR sanction o f IE violations of the framework (excerpt 2), and also the other way round. The most explicit form o f this is meta-talk about the norms and power relations of the speech event (excerpt 3).

References

Atlas, J. D. and S. Levinson (1981): It-clefts, informativeness and logical form: Radical pragmatics (Revised standard version). - In: P. Cole (ed.) Radical Pragmatics, 1-61. New York: Academic Press. Bavelas, J. B. et al. (1988): Political equivocation: A situational explanation. - Journal of Language and Social Psychology 7 (2), 137-145. Blum-Kulka, S. (1983): The dynamics of political interviews. - Text 3, 131-153. Clayman, S. E. and J. Whalen (1989): A sequential analysis of the Rather-Bush "interview". Research on Language and Social Interaction 22, 248-272. Greatbatch, D. (1986): Aspects of topical organization in news interviews: The use of agendashifting procedures by interviewees. - Media, Culture and Society 8, 441- 455. Grice, H. P. (1975): Logic and conversation. - In: P. Cole and J. Morgan (eds.) Syntax and Semantics 3: Speech Acts, 41-58. New York: Academic Press. Hall, S. (1972): The external-internal dialectic in broadcasting: Television's double-bind. - In: Fourth Symposium on Broadcasting Policy. Manchester: Department of Extra-mural Studies, 91105. Halliday, Μ. A. K. and R. Hasan (1985): Language, Context and Text: Aspects of language in a social-semiotic perspective. - Victoria: Deakin University Press. Harris, S. (1991): Evasive action: How politicians respond to questions in political interviews. - In: P. Scannell (ed.) Broadcast Talk, 76-99. London: Sage. Heritage, J. and D. R. Watson (1979): Formulations as Conversational Objects. - In: G. Psathas (ed.) Everyday Language. Studies in Ethnomethodology, 123-162. New York: Irvington.

294

Gerda Lauerbach

Heritage, J. and A. L. Roth (1995): Grammar and institution. Questions and questioning in the broadcast news interview. - Research on Language and Social Interaction 28 (1), 1-60. Lauerbach, G. (2000a): Negotiating defeat in election night TV coverage. - In: M. Coulthard, J. Cotterill and F. Rock (eds.) Selected Papers of the IADA Conference, Birmingham 1999, 1 ΟΤΙ 20. Tübingen: Niemeyer. - (2000b): Implicit communication in political interviews: negotiating the agenda. - In: E. Weigand, and M. Dascal (eds.) Negotiation: The dialogic question. Amsterdam: Benjamins. - (2000c): Discourse analysis at the interface of politics and the media: TV election night coverage. In: B. Reitz, and S. Rieuwerts (eds.) Anglistentag 1999. Proceedings. Trier: Wissenschaftlicher Verlag. Levinson, S. C. (1983): Pragmatics. - London: Cambridge University Press. Nofsinger (1989): Let's talk about the record: contending over topic redirection in the Rather/Bush interview. - Research on Language and Social Interaction 22,273-291. Pomerantz, A. (1989): Constructing skepticism: four devices used to engender the audience's skepticism. - Research on Language and Social Interaction 22,293-313. Schegloff, E. (1989): From interview to confrontation: observations of the Bush/Rather encounter. Research on Language and Social Interaction 22, 215-240.

Margareta Manu-Magda Ein Sonderfall des sozialen Dialogs: der interethnische Dialog

1. Vorbemerkung

Nach dem Zusammenbruch des Kommunismus haben liberale Konzepte wie zum Beispiel "Marktwirtschaft" oder institutionelle Reformen, deren praktische Anwendung kurzfristig Armut und Unordnung verursachen, weniger Enthusiasmus ausgelöst als die nationalen Angelegenheiten. Menschen leiten ihre persönliche Identität insbesondere aus der kollektiven Identität ab und bleiben unbeweglich wenn es sich darum handelt, zwischen komplexe Alternativen zu wählen, also bleiben sie kognitiv arm. Hinzu kommt die erschrekende Impersonalisierung der kommunistischen Gesellschaften, man muß sich deshalb nicht wundern, daß das Grundbedürfnis der Menschen in Osteuropa nicht die Freiheit, sondern die Identität war, und die Tatsache, daß die beiden in einigen Stellen übereinstimmt haben war einfach eine historische Chance. Dort wo sie nicht übereinstimmt haben, war das Bedürfnis nach Identität stärker (Mungiu-Pippidi, 2000: 15-16). Die zeitgenossischen Theorien über Nationalismus, zahlenmässig weit entwickelt (insbesondere die bezüglich des osteuropäischen Nationalismus), wollten die über die ethnische Identität und Stereotypen der untersuchten Gruppen erhobenen Daten beschreiben und interpretieren, die sozialen Vorstellungen der Geschichte als Basis des nationalistischen Betragens beschreiben, die besonderen sozialen Vorstellungen bezüglich des interethnischen Zusammenlebens analysieren, die ethnischen Konflikte und deren Mechanismus definieren und die möglichen Lösungen solcher Konflikte diskutieren. Überall redet man über die Aufklärung von Divergenzen durch Dialog, aber überraschender Weise muß man feststellen wie niedrig das Interesse einiger Arbeiten zur Kommunikation fiir die linguistische Komponente der besprochenen Aspekte ist (cf. Van Cuilenburg 2000: 167). In den vorliegenden Seiten nehmen wir uns vor, von dem linguistischen Beispiel Siebenbürgens (Rumänien) ausgehend, die wichtigsten Charakteristiken des interethnischen Dialogs, ein Sonderfall des sozialen Dialogs, zu definieren.

296

Margareta Manu-Magda

2. Das theoretische Problem: sozialer Dialog / interethnischer Dialog

Welche sind die Elemente die den interethnischen Dialog im Bereich des Konzepts von sozialem Dialog kennzeichnen?1 Der interethnische Dialog ist eine interaktionelle Beziehung, die sich zwischen Gruppen, die unterschiedlichen, zusammenlebenden Ethnien angehören, entwickelt, im allgemeinen in geographischer Proximität, im Rahmen desselben Staates oder an der zwischenstaatlichen Grenze. Diese Beziehung ist dazu bestimmt, gemeinsame Probleme des Zusammenlebens zu lösen. Was die Ethnie im Verhältnis zu anderen Gruppen charakterisiert, ist die ethnische Identität. In der soziologischen Definition des Konzepts von "Ethnizität" (siehe, unter anderen Smith/Hutchinson, 1996 und Mungiu-Pippidi, 2000: 27) sind die subjektiven Elemente dominant. Die Ethnizität wird von folgenden Zügen charakterisiert: einem gemeinsamen, eigenen Namen, um die "Essenz" einer Gemeinschaft zu identifizieren und auszudrücken; einem Mythos der gemeinsamen Abstammung, welcher die Idee einer in Zeit und Raum gemeinsamen Abstammung miteinbezieht und einer Ethnie das Gefühl einer fiktiven Verwandschaft ("Suprafamilie") verleiht; einem gemeinsamen historischen Gedächtnis, kollektive Erinnerungen einer gemeinsamen Geschichte, welche Helden, wichtige Ereignisse und deren Andenken einschließt;

1

Wir zählen die Besonderheiten des sozialen Dialogs, 1991 von Gutu-Romalo definiert, auf: Der Gruppencharakter der Teilnehmer; das Individuum wird in der dialogalen Tätigkeit durch eine Gemeinschaft ersetzt; jede Einzelperson, die der erwähnten Gruppe angehört, verwandelt sich in einen Repräsentanten der Gruppe, einen Wortträger. Die Existenz einer Hierarchie auf drei Ebenen, welche ausser des Diskursuniversum des Individuums auch zwei Varietäten von gemeinschaftlichen Universen miteinbezieht - das von der Gruppe dargestellte Universum und das der Teilnehmer. Die Dominanz des Lösungsdialogs ("solving problem dialogue") und des Klärungsdialogs ("clarification of assumed or actual misunderstandings dialogue") in sozialen Debatten. Hinsichtlich der linguistischen Organisation charakterisiert sich der soziale Dialog durch: Das Auftreten von Repliken von beträchtichem Volumen (sehr ausgedehnt), welche häufig konvergente Meinungen ausdrücken. Somit können Stagnationsmomente in der diskursiven Tätigkeit erscheinen, ohne daß die Regeln des Dialogs gebrochen werden. Die Abwesenheit von Phänomenen von dialogaler Syntax. Die Notwendigkeit des Gebrauchs von expliziten Indikatoren für die Sicherung des kongruenten Charakters der Repliken. Die Seltenheit von Interferenzphänomenen zwischen gleichzeitigen Repliken. Der Beitrag von schriftlichen und mündlichen Texten zur Abwicklung des Dialogs.

Ein Sonderfall des sozialen Dialogs: der interethnische Dialog -

-

297

einem o d e r mehreren Elementen einer gemeinsamen Kultur (Religion, Sitten, Sprache); einer B e z i e h u n g zu einer Heimat, nicht unbedingt physisch von der Ethnie bewohnt, aber als symbolische Zuneigung zum Territorium der Vorfahren anwesend, w i e im Falle der V ö l k e r aus der Diaspora; einem G e f ü h l der Solidarität, wenigstens von einigen Segmenten der Ethnie ausgehend.

Die ethnischen Gruppen definieren sich selbst im Verhältnis zu anderen Gruppen. Die Mitglieder einer "in-group" haben die Tendenz die G r u p p e zu favorisieren und somit Stereotypien u n d Diskriminierungen gegen die Mitglieder einer "out-group" zu produzieren. An den ethnischen Grenzen erscheint der Nationalismus. Die Soziopsychologen haben die Tendenz die Sprache als Hauptmarker einer Kultur anzusehen. Die Sprache ist das wichtigste Verbindungsglied zwischen der persönlichen und der kollektiven Identität. Von dieser Bemerkung ausgehend, haben die Staaten immer versucht gegenüber den ethnischen Minoritäten ihre linguistische Politik durchzusetzen, u m die erwähnten Minoritäten in gute Untertanen oder Staatsbürger umzuwandeln. In den Situationen v o n linguistischem Kontakt manifestiert sich mit erhöhter Intensität das Phänomen v o n linguistischer Loyalität. Auf der E b e n e des linguistischen Phänomens wird die Problematik der linguistischen Identität unter soziolinguistischem Aspekt relevant. In erster Linie sind die symbolischen Aspekte der V e r w e n d u n g der Sprache in verschiedenen sozio-kulturellen Kontexten angedeutet, und weniger die instrumenteilen Aspekte. Somit definiert man die soziolinguistische Identität einer ethnischen Gruppe durch folgende Elemente: -

eine Sprache (die Weise durch welche eine nationale G r u p p e ihre Stimme zur Geltung bringt) mit einem Prestigestatus (von offizieller Sprache - die einzige oder eine der nationalen Sprachen eines multinationalen Staates); eine Konversationsgeschichte, der ethnischen G r u p p e eigen, also die Gesamtheit der konversationnellen Interaktionen, die in einem gegebenen Augenblick zwischen den Mitgliedern der Gruppe stattgefunden haben; ein stark entwickeltes pragmatisches Bewusstsein der gesamten ethnischen Gruppe, welches die Kapazität der Individuen miteinbezieht, unterschiedliche pragmatische Kompetenzen abzugrenzen (also die Unterschiede zwischen dem linguistischen Verhalten verschiedener ethnischen Gruppen zu bemerken); ein Patrimonium von spezifischen illokutiven Ressourcen, die die kulturellen Werte der G r u p p e ausdrücken, in deren Rahmen die in Routinenformeln ausgedrückten M y then und Metaphern von großer Wichtigkeit in der dialogalen Tätigkeit sind.

298

Margareta Manu-Magda

Die essentielle Bedingung der Kommunikation in den mehrsprachigen Gebieten ist die Respektierung der ethnischen Identität (aller aufgezählten Komponenten, insbesondere der soziolinguistischen Identität) der Gruppen die Kontakt aufnehmen. Die Respektierung dieser Forderung entwickelt die Prämissen des Dialogs (kooperative Kommunikation); das Überschreiten dieser Bedingung löst den interethnischen Konflikt aus (Nonkommunikation, Konflikt). Die Lösung der Grundprobleme der ethnischen Minderheiten wird heutzutage immer mehr in die politische Sphäre versetzt. Da sich der Schweipunkt des politischen Raumes zunehmend aus dem Parlament zu den Medien verlagert, ist die Rolle des mediatisierten politischen Diskurses (dessen prototypische Form die Debatte darstellt) in einem ständigen Wachstum. Die in der Kommunikation miteinbezogenen Partner empfinden das Bedürfnis einen starken Verbündeten für die Lösung ihrer spezifischen Probleme zu gewinnen: die öffentliche Meinung. Im Folgenden werden wir versuchen einige Aspekte der interethnischen Kommunikation auf dem linguistischen Territorium Siebenbürgens zu veranschaulichen.

3. Das historische Problem: die Konversationsgeschichte der Ethnien in Siebenbürgen

Das Hauptproblem Siebenbürgens ist die Verteilung seines Territoriums zwischen mehreren ethnischen Gruppen. Die ethnisch homogenen Dörfer ausschliessend (rumänisch, ungarisch oder deutsch) konstatiert man ein Modell von Zusammenleben, im Rahmen dessen sich Rumänen, Ungarn und Deutsche innerhalb von Dörfern, Stadtvierteln, Straßen und Wohnblöcken vermischen. Die in den letzten Jahren des Ceausescu Regims und nach der Revolution (durch Auswanderung nach Deutschland) stattgefundene massive Zurückziehung der nationalen zahlmässig geringeren Gruppe - die Sachsen - hat an einigen Stellen des Territoriums einen frappanten Charakter. Die Einwanderung einiger Zigeunerfamilien in die sächsischen Dörfer scheint keinen Erfolg gehabt zu haben. Jedenfalls hat die Zurückziehung der Deutschen Siebenbürgen fast biethnisch hinterlassen - ein Feld des ethnischen Wettkampfs zwischen Rumänen und Ungarn, Wettkampf der immer existiert hat, aber in einer viel milderen Form. Die Einteilung des Raumes, der Wettkampfund das Zusammenleben nehmen symbolische und pragmatische Formen an. Wir werden uns auf einige Aspekte dieses Wettkampfes beziehen, der sich auf sozio- und pragmalinguistischer Ebene äußert.

Ein Sonderfall des sozialen Dialogs: der interethnische Dialog

299

3.1. Der Kampf um die Sprache Das Hauptelement der ethnischen Selbstidentifizierung ist die Sprache. Der Kampf, welcher in den letzten Jahren von den ungarischen Führern für die Verwendung der Muttersprache in der Verwaltung und der Justiz geführt wird, ist nicht unbedingt funktionell, sondern viel mehr symbolisch. Die ungarischen Führer haben in erster Reihe symbolische Einwendungen zum Unterrichtsgesetz formuliert, Einwendungen die sich auf die Ausschliessung der Vormacht oder des speziellen legalen Status der offiziellen Sprache (rumänisch) beziehen, welche die ungarische Sprache und deren Benützer erniedrigen (siehe Formeln wie 'Sprache zweiten Ranges', 'Bürger zweiten Ranges'). Die Rumänen fürchten sich vor Zweisprachigkeit und widersetzen sich dieser Idee, indem sie meinen, daß die Ungarn demnächst ablehnen werden, Rumänisch zu lernen und somit die Rumänen zwingen werden, ungarisch zu lernen (siehe Formeln wie 'Fremde im eigenen Land zu werden'). Die Unterrichtssprache in Territorien mit einer zwischen zwei Staaten bestrittenen Geschichte verwandelt sich in ein politisches Subjekt wegen des Pflichtstudiums der rumänischen Sprache, die, laut Verfassung, die einzige offizielle Sprache darstellt. Dieses Pflichtstudium wird sowohl von denen verneint, die die Sprache lernen müssen, als auch von den ungarischen politischen Führern. Der beste gefundene Kompromiss, nach zahlreichen Debatten in den Fachkommissionen des Parlaments, was die Ordonanz nr.36 angeht, die die Abänderungen zum Unterrichtsgesetz enthält, war die Annahme der Variante des UDMR (die ungarische ethnische Partei, Mitglied der Regierung) - Geographie und Geschichte in Ungarisch zu lehren, aber mit der Präzisierung, daß auch alle rumänischen Namen gelernt werden müssen.

3.2. Die Konversationsgeschichte Die Konversationsgeschichte der Rumänen und Ungarn in Siebenbürgen hat unterschiedliche Repräsentierungen im kollektiven Gedächtnis der zwei Ethnien (subjektives Wahrnehmen), aufgrund der Existenz von zwei historischen Versionen und den Folgen dieser Tatsache auf der Ebene der Selbstidentifizierung (der sozialen Repräsentierungen über die eigene Gruppe) und des Zuschreibens (der sozialen Repräsentierungen über andere). Den Eindruck, den uns das Verhältnis von Interaktion zwischen den zwei ethnischen Gruppen hinterlässt ist ein Eindruck von Nonkommunikation. Eine der wichtigsten Ursachen der Nonkommunikation ist die wechselnde Autoritätsposition der einen oder anderen Ethnie, im Verhältnis zur staatlichen Macht im Laufe der Geschichte, sowie auch des Status von offizieller (nationaler) Sprache/Minderheitssprache die die Idiome der beiden Ethnien alternativ besessen haben.

300

Margareta

Manu-Magda

Auf der Ebene des linguistischen Ausdrucks bemerkt man die Existenz von parallelen Diskursen, von den Repräsentanten der einen oder der anderen linguistischen Gruppe, entweder an eine obere, entscheidungsfahige Instanz (normalerweise eine nationale oder internationale politische Behörde) oder an einen Zeugen in einer symbolischen Instanz gerichtet (die öffentliche Meinung). Die interethnische Kommunikation wird oft vermittelt und somit nimmt die Mehrheit der konversationeilen Episoden die Form der Deklarationen (eine autoritär-kriegsfuhrende Position der Teile ausdrückend - siehe die in Kontakt kommen die Formel "der Krieg der Deklarationen") oder der Verhandlung an (die somit ein Verhältnis Übergeordneter/Untergeordneter veranschaulicht). Aus den aufgezählten Gründen nehmen die konversationeilen Episoden die die konversationelle Geschichte zusammenstellen nicht die Form des Dialogs an, mit der Ausnahme einer sehr kleinen Anzahl, aus den letzten vier Jahren der rumänisch-ungarischen Regierung, in welchen die Gesprächspartner von gleichen Positionen aus kooperativ interaktioniert haben.

3.3. Das pragmatische Bewußtsein Das pragmatische Bewußtsein wird von dem metapragmatischen Diskurs veranschaulicht, welcher sich auf Werte bezieht, die ein bestimmtes linguistisches Verhalten bei den der rumänischen oder ungarischen Gruppe angehörigen Sprecher erzeugt. Beispiele: Ungarn aus Rumänien über Ungarn aus Ungarn: 'Umsonst sprechen wir dieselbe Sprache, wenn die Ideen nicht gemeinsam sind'. 'Wir sind kommunikativer als sie, das haben wir von den Rumänen angenommen, wir sind empfanglicher und freundlicher, mehr lateinisch, sie sind distanzierter als wir*. Ungarn aus Klausenburg über Rumänen aus Bukarest: 'Ich höre zwei Menschen in Bukarest auf der Straße streiten und ich wundere mich: "Um Himmels Willen, warum streiten diese Menschen?", und nachher fällt es mir ein - sie streiten nicht, diese ist ihre Art zu sprechen*.

Ein Sonderfall des sozialen Dialogs: der interethnische

Dialog

301

4. Daten der Kommunikation: Illokutive Ressourcen

Das Spezifische des interethnischen Dialogs, als argumentatives Vorgehen, ergibt sich aus der gerade erwähnten essentiellen Bedingung: die Unabsetzbarkeit der ethnischen Identität der Gesprächspartner. Demzufolge ruht diese Art von Diskurs, im Vergleich zu anderen, hauptsächlich auf "Indirektheit" und einem hohen Grad von "Implizität". Bei der Konstruktion des interethnischen Dialogs verlässt man sich auf das komplexe Inventar der diskursiven Manipulationen, auf das Miteinbeziehen der intersubjektiven Dimension (Argumentation, sozio-kulturelle und rethorische Normen, Kommunikationsakte). Die lineare, telegraphische Auffassung des Diskurses (von Empfänger zu Empfänger ) wird durch eine poliphonische Logik des Diskurses ersetzt, in welcher Sprechen eigentlich Voraussehen der interpretativen Erwartungen des Gesprächspartners bedeutet. Demzufolge erhalten in der Interaktion zwischen den zwei erforschten Gruppen eine wichtige Rolle die standardisierten illokutiven Ressourcen (eine spezifische Form des "langage de bois" - der nationalistische "langage de bois"), konzentrierte, vereinfachte, und zu festen Formen des Ausdrucks reduzierte Inhalte (Effekt der Stereotypie) - leicht aufzunehmen, wiederzugeben und zu verwenden in der Manipulation der Massen. Als pathologisches Element des gegenwärtigen politischen Lebens gesehen, ohne selbstverständlich in den vergangenen Epochen zu fehlen, überfällt die "langage de bois" (hölzerne Sprache) hauptsächlich die politischen Diskurse der nationalistischen Parteien (in Siebenbürgen UDMR - die ethnische ungarische Partei und Romania Mare, Grossrumänien - extremistische, nationalistische, rumänische Partei). Es gibt eine wiederholte Thematik und Terminologie des nationalistischen Diskurses, was zum Beispiel folgende Aspekte betrifft: -

-

die Umwandlung der Geschichte in eine Legende: für die Rumänen ist Siebenbürgen 'die Wiege der dakisch-lateinischen Population'; für beide Gruppen ist es die 'urväterliche Heimat'; die Ungarn halten sich hier für 'die Wiege der Zivilisation1; die Wiederherstellung des geschichtlichen Ansehens in der Gegenwart - die Ziele der zwei Lager sind in dieser Hinsicht antagonistisch.

Die Grossrumänienpartei hat eine assimilitionistische Vision: 'In diesem Land sind wir alle Rumänen'; 'Wir wollen uns nicht fremd in unserem eigenen Land fühlen'; 'Die Rumänen sollen Herren bei uns zu Hause sein, sie sollen nicht mehr von den Fremden beleidigt werden'; die orthodoxe Kirche ist 'die nationale Kirche'; für Grossrumänien ist das Hauptziel 'die Verteidigung des Rumänischen, des nationalen Wesen, der nationalen Sache, der Interessen des rumänischen Volkes, der Souveränität und nationalen Integrität'. Ein Teil des UDMR ist Anhänger des Revisionismus:

302

Margareta

Manu-Magda

'Warum soll ich nach Ungarn gehen wenn meine Vorfahren in dieser Erde begraben sind?'; 'Für uns gibt es keinen Trianon Vertrag sondern ein Trianon Diktat'; Die Zeitspanne des Wienerdiktats ist die 'goldene Periode'; Die Gläubigen singen in der Kirche 'Gott, beschütze die ungarische Nation etc'. -

die Vernichtung des inländischen und ausländischen Feindes (von der gegnerischen ethnischen Gruppe dargestellt): in der Vorstellung der Partei "Grossrumänien", wird Rumänien im Rahmen seiner aktuellen Grenzen vom Inneren aus bedroht; die UDMR ist 'eine segregationistische Formation die an der Zerspaltung des rumänischen nationalen Einheitsstaates arbeitet'; die Verteidigung des rumänischen Geistes konkretisiert sich durch 'die Verteidigung der Rumänen in den Gebieten in welchen die Ungarn in der Mehrheit sind (Covasna, Harghita), die von der UDMR und den Agenten Ungarn kolonisiert werden'; viele der Aktionen der ungarischen Bevölkerung sind 'Provokationen gegenüber den Rumänen, unter diesen auch die ethnische Diskriminierung der Rumänen in Harghita und Covasna', usw.

-

Für die Ungarn sind die Rumänen 'ein primitives und unzivilisiertes Volk', etc.

-

die gemässigten politischen Gruppen reden über Dialog, Konsens, Zusammenarbeit, friedliches Zusammenleben, Verständnis, Toleranz.

Debatte,

Man spricht heute von der Wiederherstellung einer gemeinsamen Basis des interkulturellen Dialogs.

5. Schlußfolgerungen

In den vorliegenden Seiten haben wir versucht, vom soziolinguistischen Beispiel Siebenbürgens ausgehend, einige Züge der interethnischen Kommunikation zu veranschaulichen. 1.

Die interethnische Kommunikation ist eine Gruppenaktion, deren Spezifisches aus der Identifizierung der Teilnehmer an unterschiedlichen Ethnien ausgeht. Den Unterschieden von ethnischer Identität entsprechen Unterschiede von soziolinguistischer Identität.

2.

Die Kommunikationsteilnehmer interaktionieren von ungleichen Positionen aus: Übergeordneter/Untergeordneter; Mehrheit/Minderheit.

3.

Die interethnische Kommunikation nimmt nur selten die Form des Dialogs an (kooperative Interaktion); die Repliken sind normalerweise parallele politische Diskurse in Form von Deklarationen oder Verhandlungen.

Ein Sonderfall des sozialen Dialogs: der interethnische

4.

Im

Rahmen

des

demokratischen

Dialog

303

postkommunistischen

Staates

nimmt

die

Kommunikation die prototypische Form der Debatte, des "simulierten Dialoges" an. (der Diskurs

stellt sich in

dialogaler

Form vor,

um

somit

den

essentiellen

monologischen Charakter des verbalen Ereignisses zu verstecken). 5.

In der Lösung der Probleme des Zusammenlebens muß man sich oft an eine Vermitt-

6.

Auf illokutiver Ebene werden als Argumente "Mythen" und "Methapern" (Stereoty-

lungsinstanz wenden. pien) verwendet, in Routineformeln ausgedrückt. 7.

Das Ziel

der interethnischen Kommunikation auf pragmatischer Ebene ist der

Konsens, dessen erste Etappe vom Klärungsdialog dargestellt wird, der in einer gemeinsamen Anstrengung zur "Demythisierung" der Mythen und zur Explizitierung der Methapern besteht.

Literatur

Alvarez-Perreyre, F. (ed.) (1981): Ethnolinguistique. Contributions Th6oriques et Methodologiques. Actes de la Reunion Internationale "Th6orie en Ethnolinguistique" (Ivry 29 mai-ler juin 1979) Paris: SELAF. Gallagher, Τ. (1999): Democrafie nationalism in Romania. - Bucure$ti: All. Golopen(ia, S. (1985): Interaction et histoire conversationnelles. - In: Actes du colloque "Interactions conversationnelles". Urbino, Italy: (pre-print). Guju-Romalo, V. (1991): Considerations sur le "dialogue social" - In: S. Stati, E. Weigand, F. Hundsnurscher (eds.) Dialoganalyse III. Referate der 3. Arbeitstagung, Bologna 1990. Teil 2, 433-439. Tübingen: Niemeyer. Larrue, J. und A. Trognon (1994): Introduction & la pragmatique du discours politique. - Paris: Armand Colin. Mungiu-Pippidi, A. (2000): Transilvania subiectivä. - Bucure?ti: Humanitas. Ple?u, A.(1996): Chipuri $i mäjti ale tranzi{iei. - Bucure$ti: Humanitas. Slama-Cazacu, T. (2000): Stratageme comunica{ionale $i manipularea. - Ia$i: Polirom. Smith, A. D. und J. Hutchinson (eds.) (1996): Ethnicity. - Oxford: University Press. §tefan-Scalat, L. (1999): Partidul Romania Mare. Un profil doctrinar. - Sfera Politicii 67, 16-20. Thom, F. (1993): Limba de lemn. - Bucure§ti: Humanitas. Van Cuilenburg, J. J. et alii (2000): Stiinfa comunicärii. - Bucurejti: Humanitas.

Mirka Maraldi et Anna Orlandini

Exclamatives, interrogatives-exclamatives et le jeu de la negation

Nous enqueterons ici sur la nature semantique des propositions exclamatives, sur la difference argumentative1 entre propositions enonciatives et exclamatives, ainsi que sur le lien qui permet de rapprocher ces propositions du mecanisme de la negation. Au sujet des exclamatives, Culioli (1974: 6) ecrit: "Elles ont un Statut mal defini. II s'agit bien d'assertives, mais avec un quelque chose en plus (c'est nous qui le soulignons) qui se marque par des precedes divers, lexicaux, prosodiques, syntaxiques". II nous semble que ce "quelque chose en plus" releve de la nature argumentative des exclamatives. Comme l'6crit Milner (1978: 253): "C'est I'interpretation qui rassemble les exclamatives: toute phrase que Ton designe de ce nom implique l'expression d'un haut degre soit dans l'ordre de la qualite (quand il s'agit d'un adjectif) soit dans l'ordre de la quantite". La presence d'une exclamative entralne, comme le soutient Le Querler (1994), l'expression d'une modalite (notamment d'une modalite appreciative lorsqu'il s'agit de l'expression du haut degre). Selon nous, ce "haut degre" possede aussi une nature argumentative: l'exclamative signale alors le debordement d'un seuil attendu, autrement dit le depassement de la zone au-dessous de laquelle une norme est applicable. Selon Martin (1987: 106), la caractdristique principale des propositions exclamatives est de presenter, au niveau pragmatique, une tension contradictoire qui "oppose l'univers actuel de celui qui parle ä une image d'univers oü ρ est possiblement faux". Nous croyons que l'on peut rapprocher la notion de tension contradictoire de celle de depassement d'un topos, d'une norme. L'idee sous-jacente d'une negation implicite dans l'exclamation se rencontre aussi dans les explications faisant appel ä Γ existence d'un preconstruit auquel l'exclamation s'oppose de fa9on violente, en le forcluant (cf. Danon-Boileau 1995: 244). Nous nous proposons de verifier, dans les contextes latins, la justesse de cette hypothese qui nous permettra d'expliquer Papparition du pronom indefini pragmatiquement negatif quisquam dans certaines exclamatives. On distinguera d'abord les exclamatives prononcees dans une situation de contraste externe (engendrees par les paroles d'autrui explicites ou implicitement evoquees par polyphonie) et celles qui

Rappeions que dans les Traites de rhetorique ancienne on reconnait ä l'exclamation une certaine force argumentative (cf. Rhet. ad Her. 4,22), puisque eile y est repertoriee en tant que figure rhetorique destinde ä produire un certain effet sur l'auditoire; qu'elle soit confue comme une figure du discours, selon Cicdron (de or. 3,277), ou comme une figure de la pensee, selon Quintilien (inst. 9,3,97).

Mirka Maraldi et Anna Orlandini

306

relevent d'un contraste interne (entre les attentes du locuteur et Γ etat du monde actuel). Ensuite, nous classerons les propositions exclamatives des contextes de contraste interne selon que Γarriere-plan du locuteur (pour employer une expression chere ä Claude Muller) ou Γ image d'univers (pour reprendre un terme de Robert Martin) soit nigatif ou positif. La tension contradictoire qui nalt de l'opposition entre les attentes du locuteur et le monde actuel explique l'effet d'etonnement, d'indignation ou simplement de stupeur, et parfois meme la surprise vis-ä-vis d'un bonheur inattendu, qui connotent pragmatiquement ce type de phrases.2

1. Exclamatives dans une situation de contraste externe: evaluation depreciative de la part du locuteur de la parole d'autrui

Une modalite evaluative explicite est exprimee lorsque, par une expression depreciative realisee par un SN pivot de la phrase exclamative ("Sornettes!", "Balivernes!",

"Baga-

telles!"), le locuteur commente negativement une assertion effectivement prononcee par son interlocuteur, comme dans le passage suivant: (1) TO. Ne hie tibi dies inluxit lucrificabilis. / Nam non emisti banc, uerum fecisti lucri. / DO. Ille quidem iam seit quid negoti gesserit, / qui mihi furtiuam meo periclo uendidit. / Argentum aceepit, abiit. Qui ego nunc scio / an iam adseratur haec manu? Quo ilium sequar? / In Persas? Nugas! (Plaut. Pers. 717 sq.) ("Certes, le jour qui s'est Ιενέ ce matin est un jour de profit pour toi. Tu ne l'as pas achetee, tu l'as eue pour rien!":: "Lui, en tout cas, il sait ce qu'il a fait, en vendant, sans garantie, une fille νοίέε. II a reipu Γ argent, il est parti. Comment puis-je savoir si on ne va pas venir la reclamer? Ou irais-je le chercher? En Perse? Plaisanteries!") L'expression Nugas! est un accusatif exclamatif qui pourrait etre explique par un uerbum dicendi implicite (Dordale commente negativement l'enonce entier de Toxile: - "Tu dis des bagatelles!").3 Parfois le uerbum dicendi est exprime, comme dans le passage suivant,

Selon la definition des rheteurs anciens, Vexclamatio est une figura par laquelle le locuteur manifeste son indignation et peut la susciter chez son auditoire: Exclamatio est, quae conficit significationem doloris aut indignationis alicuius per hominis aut urbis aut loci aut rei cuiuspiam conpellationem (Rhet. Her. 4,22 - "L'exclamation permet d'exprimer un sentiment de douleur ou d'indignation par une apostrophe ä un homme, ä une ville, ä un lieu, ä un objet quelconque"). D'ailleurs, le fran;ais aussi a une expression qui, dans l'une de ses interpretations possibles (celle d'opposition) est proche de ces exclamatives et est realisee par un uerbum dicendi - "Tu paries!".

Exclamatives, interrogatives-exclamatives et le jeu de la negation

307

qui presente toute une serie d'expressions constituant le complement d'objet direct du predicat fabulare: (2) LA. Nummos trecentos. GR. Tricas. / LA. Quadringentos. GR. Tramas putidas. LA. Quingentos. GR. Cassam glandem. / LA. Sexcentos. GR. Curculiunculos minutos fabulare (Plaut. Rud. 1323-1325) ("Trois cents 6cus.:: Plaisanterie! : : Quatre cents.: : Des chiffons pourris!:: Cinq cents.:: Un gland creux!:: Six cents!:: Tu paries de petits charanpons?") Dans ce passage, Gripus utilise des quantifieurs minimaux (Tricas; Tramas putidas; sam glandem; Curculiunculos)

Cas-

pour indiquer des choses d'une infime valeur dans le but

argumentatif de depräcier l'offre de Labrax, son interlocuteur. Une dissociation vis-a-vis d'une assertion de l'interlocuteur pourrait aussi rendre compte de certains subjonctifs d'indignation qui se placent entre Γ interrogation et l'exclamation et qui presentent la reprise polemique d'un meme enonce: (3a) Bonus est hie uir. Hie sit4 bonus? (Ter. Andr. 915) ("II est un homme de bien" :: "Celui-ci serait un homme de bien?") (3b) ER. Tibi dedi equidem illam [...] ME. Mihi tu ut dederis pallam et spinier? (Plaut. Men. 681-683) ("Mais je te l'ai donnde [...]" :: "Toi, tu m'as remis la mante et un bracelet ä moi?")

2. Exclamatives dans une situation de contraste interne: contextes de tension contradictoire exprimant le jugement du locuteur a propos d'un fait

La caracterisation que nous proposons pour ce type de propositions exclamatives est tres generale et adequate ä n'importe quel contexte d'exclamation (interne ou externe au dialogue), puisque Ton pourrait penser, comme le fait Goffman (1987: 119), que toute exclamation releve, en derniere instance, du soliloque. On postulera, pour ces contextes, non pas un simple uerbum dicendi ou cogitandi implicite ou εχρππιέ (comme e'est le cas pour les contextes dialogaux), mais un predicat performatif implicite du type: moleste fero, crucior ou plus generalement, un verbe de subordination

miror,

critique (cf. Muller

1991). La nature argumentative d'une exclamative dans un contexte de contraste interne ne releve done pas d'une simple assertion, mais d'un jugement du locuteur qui denonce, a un autre niveau illocutoire, une contradiction interne. Cela justifierait la contradictoire

tension

dont parle Martin, entre les attentes du locuteur et le monde actuel, oü une

Pour cet emploi du subjonctif dans les questions ichoi'ques ayant une nature refutative, cf. A.R. Anderson (1913).

308

Mirka Maraldi et Anna Orlandini

situation inverse s'est, de fait, realisee. Selon Ernout-Thomas (1964: 23) une exclamation telle que - Me miserum! releve de "l'idee latente d'une constatation ou d'un jugement"; par. ex., - "Que (= ä quel haut degre) je me trouve malheureux!", autrement dit - "Je suis malheureux ä un trop haut degri!". Cette constatation et ce jugement dönoncent, selon nous, une discordance (une tension contradictoire) entre les attentes, les souhaits du locuteur et Γ etat de choses du monde actuel; elles signalent aussi qu'un seuil, une norme ont ete depasses. Nous analyserons ici deux types de contextes qui admettent tous les deux, en franfais, la paraphrase "Ce n'est pas possible!", mais dont Tun exprime l'indignation (§ 2.1.), alors que l'autre Signale l'incredulite (§ 2.2.).

2.1. Contextes exprimant l'indignation Une negation implicite peut etre sous-entendue dans les contextes exclamatifs exprimant l'indignation du locuteur. Ä propos de l'enonce: - Ν on! non! c'est pas possible ςα! II a encore renverse de l'eau! Bres (1995: 85) explique le fonctionnement deictique de l'exclamation "Non!" comme signalant que "Γoperation de negation vise ä forclore fantasmatiquement de la realite un element qui y est indubitablement inscrit". Cette operation negative peut aussi etre justifiee par une discordance entre le monde actuel et le monde des attentes du locuteur, qui engendre une tension contradictoire. Par rapport aux conditions du monde actuel, nous pouvons envisager deux situations relevant toutes les deux d'une discordance selon qu'il s'agisse d'attentes positives (§ 2.1.1.) ou d'attentes negatives (§ 2.1.2.). La negation est le moyen syntaxique par lequel se differencient ces deux types de propositions; elle joue un role fondamental pour faire le clivage entre les deux exclamatives. 2.1.1. Contextes syntaxiquement negatifs, attentes positives: la negation est un marqueur illocutoire d'orientation positive Lorsqu'une negation syntaxique est exprimee dans une proposition exclamative, cela implique un jugement de nature deontique; le locuteur juge que la proposition positive ρ devrait ou aurait dü se r6aliser ( - "Je m'attendais a ce que /?"); tel est le cas du passage suivant: (4) Hoc non uidere, cuius generis onus nauis uehat, id ad gubernatoris artem nil pertinere! (Cie. fin.4,16) ("Comment ne pas voir que la nature du chargement transporte n'a aucun rapport avec l'art du pilote!")

Exclamatives, interrogatives-exclamatives

et le jeu de la negation

309

Dans ce passage, l'infinitif exclamatif nait d'un jugement d'indignation du locuteur: "Comment nepas voir...!", qui releve d'une attente positive: -"Vous auriez dü voir!". De la meme maniere, on pourrait expliquer la ndgation qui η'est pas une negation sömantique, mais un marqueur d'orientation positive dans certains subjonctifs d'indignation ou d'6tonnement, en prisence de l'enclitique -ne et parfois aussi de la toumure dite du ut indignantis: (5a) DI. Egone illam ut non amem? egone illi ut non bene uelim? (Plaut. True. 441) ("Et moi, que je ne l'aime pas? Que je ne veuille pas son bien?") (5b) HE. Ne fle. ERG. Egone ilium non fleam? egon non defleam / talem adulescentem? (Plaut. Capf. 139 sq.) (/ "Ne pleure pas" :: "Moi ne pas le pleurer? Ne pas döplorer la perte d'un si bon jeune homme?")

Un pronom negatif tel que nemo, nihil peut apparaitre dans de tels contextes positivement Orientes: (5c) Nihil excogitem, quamobrem Oppianico damnari necesse sit? (Cie. Cluent. 70) ("Ne saurais-je rien imaginer qui rende inevitable la condemnation d'Oppianicus?")

Dans ce passage, marque par un subjonctif d'indignation, le pronom sdmantiquement negatif nihil ('rien') sert a orienter vers une attente positive, comme si le locuteur Ciceron, dans un mouvement d'indignation vis-a-vis d'une situation donnee, voulait transmettre le message suivant: "Bien evidemment que j'arriverai ä imaginer quelque chose!". Le pronom semantiquement negatif, tout comme la negation semantique, est, en revanche, exclu des contextes exprimant des attentes negatives; dans ce cas, comme nous le verrons tout de suite, seules les expressions pragmatiquement negatives (par ex., le pronom quisquam) sont admises. 2.1.2. Contextes syntaxiquement positifs, attentes negatives: la negation non est exclue II arrive fröquemment en latin que les propositions exclamatives signalant Γ indignation du locuteur prdsentent des occurrences du pronom quisquam qui est pragmatiquement negatif en l'absence d'une n£gation syntaxique: (6a) Hie mihi quisquam mansuetudinem et misericordiam nominal! (Sail. Catil. 52,11) ("Et quelqu'un vient me parier de douceur et de pitie!") (6b) Heu, cadit in quemquam tantum scelus? (Verg. eel. 9,17) ("Ah! Est-il possible pour quelqu'un d'etre ä ce point scelerat!")

Comme on peut aisement le verifier, les attentes du locuteur, dans ces passages, sont de signe negatif: dans le texte de Salluste, le locuteur s'indigne que quelqu'un puisse venir lui parier de douceur et de piti6; de meme, dans Pinterrogation d'etonnement du passage virgilien - "Est-il possible pour quelqu 'un d'etre a ce point scelerat!". Dans les deux cas, il

Mirka Maraldi et Anna Orlandini

310

s'agit de "quelqu'un" qui, selon le jugement du locuteur, ne devrait pas exister, ce qui explique le recours au pronom quisquam. Un defx polemique est present aussi dans les structures realisees par le -ne enclitique : (7) Tune impune haec facias? (Ter. Andr. 910) ("Toi, devrais-tu faire cela impunement!") qui souligne le jugement du locuteur que les choses devraient se passer bien autrement.

2.2. Contextes exprimant l'incredulite Comme le signale Vairel-Carron (1975: 126), "entre les phrases qui denotent la surprise et celles qui denotent l'indignation, il n'existe pas de limites tranchees". Analysons un passage que Bennett (1914: 425) considere comme exceptionnel parce qu'il exprime non l'indignation, mais plutot le contentement du locuteur procedant d'une agreable surprise: (8) LE. edepol senem Demaenetum lepidum fuisse nobis! (Plaut. Asin. 580) ("Parbleu, le vieux Demenete a-t-il ete charmant pour nous!")

On pourrait expliquer ce passage egalement comme relevant d'une tension contradictoire; la satisfaction de Leonide pourrait se justifier par un mouvement qui d£c61e une implication negative de ce type: "Ah! Tres bien! Je ne m 'attendais pas ä ce que le vieux Demenete soit si charmant envers nous ". On contrevient ä une attente du locuteur. 2.2.1. Expressions du debordement fictif, rhetorique, dans des contextes exprimant la surprise, l'incredulite Les attentes negatives permettent aussi d'expliquer le sentiment d'incredulite dans les exclamations: "Ce n'est pas possible!", "Ce n'est pas vrai!". Dans ce cas, comme le dit Martin (1987: 100): "L'attente que la situation donne pour contrefactuelle se prolonge fictivement jusque dans le monde actuel". Une fois de plus, la tension contradictoire justifie l'exclamation qui n'exprime pas, cette fois-ci, un blame ou un regret, mais la stupeur, l'etonnement, et qui prolonge dans le monde actuel l'implication n£gative: "Je ne m'attendais pas ä ce que p".5 La meme valeur d'incredulite (= prolongation d'une image d'univers negative) est exprimee en latin par rexclamation-commentaire: Somnium! ("C'est un reve!"). La situation peut aussi bien etre celle d'un dialogue, mais l'expression

Une interprttation de l'exclamation comme rapport ä l'inattendu se rencontre aussi chez E. Goffman (1987: 107): "Nous y voyons une 'expression', le debordement naturel d'un sentiment auparavant contenu, le bris des barrieres ordinaires par une personne soudain prise έ l'improviste".

Exclamatives, interrogatives-exclamatives et le jeu de la nigation

311

exclamative, n'ayant pas d'efficacite illocutoire dans l'enchainement dialogal, presente plutot les caracteristiques metalinguistiques d'un commentaire a parte du locuteur: (9a) PH. Crede mihi: gaudebis facto; uerum hercle hoc est. DO. Somnium! (Ter.Phorm. 494) ("Fais-moi credit: tu seras content de 1'avoir fait; par Hercule, c'est la verite!" : : "Un reve!") (9b) AN. Quod? GE. Quodnam arbitrare? AN. Nescio. GE. Atqui mirificissimum:/ Patruos tuos est pater inuentus Phanio uxori tuae! AN. Hem! / Quid ais? GE. Cum eius consueuit olim matte in Lemno clanculum. / PH. Somnium! (Ter. Phorm. 871-874) ("Quelle histoire?" : : "Laquelie crois-tu ?" : : "Je ne sais pas" : : "Mais la plus mirifique! On a dicouvert que ton oncle est le pfere de Phanium, ta femme." : : "Hein! Qu'est-ce que tu dis?" : : "II a eu jadis en secret des relations avec sa mere ä Lemnos" :: "C'est un reve!") Rappelons done la specificite de ee type d'expressions: 6 elles ne peuvent pas etre niees7, tout en vehiculant une idee sous-jacente negative. La discordance par rapport ä la realite est toujours liee au fait que la realite devance les attentes, par un 6cart provoquant une tension contradictoire. Toutefois, dans l'expression de la surprise, les attentes peuvent etre negatives aussi bien que positives; mais, dans ce dernier cas, elles sont sürement reductrices par rapport a la realite. Par l'expression Somnium! ("C'est un reve!") on signale toujours l'appreciation d'une realite positive ou fictivement positive, comme dans (Ter. Phorm. 494) qui est une r£ponse ironique. Le meme procede d'apprdciation d'un debordement positif de la realite par rapport ä nos attentes se manifeste dans l'emploi emphatique et non inverseur de Padverbe trop des parier jeunes en frangais et en italien. Ce n'est pas un hasard si l'exclamation - Trop beau! peut etre con?ue comme synonyme de "Que c'est beau!" (="A quel haut degre (depassant une attente) c'est beau!"). Un chainon dans le proces qui degage ce developpement semantique peut etre represente, ä notre avis, par les phrases de politesse, oü trop en fran9ais et nimis en latin s'emploient avec le sens non inverseur de "beaucoup, fort"; par ex.: — Je suis trop heureux de vous voir (Acad.); - Vous etes trop aimable. Ces phrases trouvent des correspondants en latin surtout dans la langue des Comiques: (10) LE. Nimis bella es atque amabilis (Plaut. Asin. 674) ("Tu es trop gentille et trop aimable!") Rappelons qu'en effet, en latin, l'adjectif nimius ainsi que les adverbes nimis, nimium sont polysemiques ("trop" et "tres") et peuvent parfois avoir la meme valeur intensifiante de trop dans l'exclamative; cf. par ex.: (1 la) PE. Nimis doctum dolum (Plaut. Mil. 248 ) ("Voilä une ruse trop habile! (= bien habile!)")

7

Cet emploi de somnium! est tout ä fait rare; il n'est attesti que chez Tdrence et notamment dans les deux seules occurrences du Phormion que nous avons citees. Cf. aussi H. Reichenbach (1947: 343 ss.) ä propos d' 'exclamatory terms'.

Mirka Maraldi et Anna Orlandini

312

(lib) SIMO Nimis illic mortalis doctus, nimis uorsutus, nimis malus (Plaut. Pseud. 1243) ("C'est un etre vraiment bien habile, bien ruse, bien malin")

Le meme emploi se prete bien ä des contextes ironiques (dont le propre est d'exploiter l'emphase, l'hyperbole, etc.); tels que, par ex.: (11c) LY. Oh, nimium scite scitus es! (Plaut. Cas. 522) ("Ah! merveilleux, tu es d'une ingeniosite merveilleuse")

Lorsque 1'interpretation "trop" entraine une inversion argumentative, l'adverbe n'est plus polysemique: la substitution trop par tres n'est plus possible parce qu'elle n'est pas semantiquement equivalente, et le topos evoque ne peut plus etre applique. En effet, selon Moeschier et Reboul (1994: 318), "l'adverbe trop indique justement que la zone au-dessous de laquelle le topos est applicable est dipassee, et que le topos ne peut plus etre applique".8 Or l'exclamative presente la propriete suivante: l'adverbe trop en fran5ais, nimis, nimium en latin, en union avec un adjectif semantiquement favorable (ou contextuellement favorable) ne fonctionne pas comme un inverseur des conclusions argumentatives (ce qui, en revanche, est normal dans une phrase non exclamative).9 Autrement dit, dans les exclamatives, trop (nimis, nimium) avec un adjectif favorable a toujours la meme valeur intensifiante que trop {nimis, nimium) avec un adjectif defavorable, tel que pretiosa dans le passage suivant: (12) PI. Ah, nimium pretiosa es operaria (Plaut. Bacch. 74) ("Non, non! Tes services me coüteraient trop eher!")

Dans ce cas, Pistoclerus se defend d'accepter les services de Bacchides (une prostituee) en soutenant qu'elle coüte trop eher pour lui. Nimium pretiosa signale ici le depassement d'un seuil, mais sans qu'il y ait inversion des conclusions argumentatives, en revanche, avec renforcement de l'argument qui s'impose comme decisif. Le "trop" intensifiant du debordement rhetorique ne fonctionne pas par inversion, mais par augmentation de la force argumentative du faisceau des topoi intrinseques lies au sens lexical du terme focalise.10 Cela est plus frequent dans les contextes exprimant Petonnement, l'incredulitd, bien qu'il ne soit pas exclu dans les contextes exprimant l'indignation. Par ex., le passage suivant: L'analyse qui envisage pour trop le depassement d'un seuil bloquant certains enchainements argumentatifs a etd proposee pour la premiere fois par J. Jayez (1985); en revanche, une analyse de l'operateur trop en termes de "bloc semantique", visant a exclure le recours ä la notion de depassement d'un seuil, a ite ddveloppie par M. Carel (1993). Cf. a ce propos, le dialogue suivant propose par M. Carel (1993: 178): (X) - C'est bon marche; je vais l'acheter (Y) - Non. C'est trop bon marche: il doit y avoir un probleme. Pour la notion de topos extrinseque et intrinseque cf. J.-C. Anscombre (1995).

Exclamatives, interrogatives-exclamatives et le jeu de la negation

313

(13) PS. Sed nimis sum stultus, nimis fui (Plaut. Pseud. 205) ("Mais je suis trop sot, je l'ai toujours 6ti" (ou "bien sot"))

releve d'un mouvement d'indignation ("Je suis trop sot" = "Que je suis sot!" = "Je suis bien sot!") et exprime de cette maniere un haut degre du terme stultus en augmentant de maniere scalaire la force argumentative des topoi intrinseques lies ä l'adjectif. Le passage suivant: (14) DAE. Nimis paene inepta atque odiosa eius amatiost (Plaut. Rud. 1204) ("Elle est presque stupide et hai'ssable, avec ses faijons de t6moigner son affection!")

est difficile ä rendre et aussi ä interpreter ä cause du jeu contradictoire de nimis ("trop" signalant un debordement) et paene ("presque" signalant une limite qui n'a pas et£ franchie)." Le jugement de Demones est depreciatif ("une fa?on de temoigner son affection qui est trop ennuyeuse pour etre appröciee"). II est, ä notre avis, un autre exemple de la valeur intensifiante liee ä des adjectifs defavorables (inepta et odiosa), mais la force argumentative est ici mitigee par l'adverbe paene qui empeche aux adjectifs d'atteindre le sommet, le tres haut degre.

3. En guise de conclusion

Le latin presente des marqueurs syntaxiques caracteristiques pour les phrases exclamatives (tels que, par ex., l'accusatif d'exclamation, Pinfinitif exclamatif, le subjonctif d'indignation). Cette langue dispose aussi de marqueurs semantico-pragmatiques: par ex., le fonctionnement de la negation comme marqueur illocutoire d'orientation positive dans les contextes d'indignation. Enfin, on a souligne la specificite du fonctionnement des adverbes nimis, nimium ainsi que de l'adjectif nimius qui, dans les exclamatives, neutralisent la distinction adjectif favorable - röle inverseur, adjectif defavorable - role intensifiant au seul profit du role intensifiant.

Pour une analyse detaillee de l'adverbe latin paene, cf. A. Bertocchi (1996).

314

Mirka Maraldi et Anna Orlandini

References bibliographiques

Anderson, A. R. (1913): Repudiative questions in Greek drama, and in Plautus and Terence. TAPhA 44, 43-64. Anscombre, J. (1995): Topique or not topique: formes topiques intrinseques et formes topiques extrinseques. - Journal of Pragmatics 24, 115-141. Bennett, C. (1914): Syntax of Early Latin, I The Verb. Boston: Allyn & Bacon. (1966) Hildesheim: Olms. Bertocchi, A. (1996): Some semantic and pragmatic properties of paene. - In A. Bammesberger, F. Eberlein (eds.) Akten des VIII. internationalen Kolloquium zur lateinischen Linguistik, 457-472. Heidelberg: C. Winter. Bres, J. (1995): Höu ! Haa ! Yrrää : interjection, exclamation, actualisation. - Faits de Langues 6, 81-92. Caret, Μ. (1993): Trop: argumentation interne, argumentation externe et positivite. - Dans J.-Cl. Anscombre (ed.): Theorie des topo'i, 177-206. Paris: Kime. Culioli, A. (1974): A propos des enonces exclamatifs. - Langue Frangaise 22, 6-15. Danon-Boileau, L. (1995): Discussion. - Faits de Langues 6, 239-251. Emout, A. et F. Thomas (1964): Syntaxe Latine. - Paris: Klincksieck. Goffman, E. (1987): Faxons de parier. - Paris: Editions de Minuit. (1981) Forms of talk. Philadelphia: University of Pennsylvania. Jayez, J. (1985): Trop: l'exces par defaut. - Le Franfais Moderne 53, 22-48. Le Querler, Ν. (1994): Formes et interpretations des enonces exclamatifs dans Le Lys dans la vallee de Balzac. - L 'Information grammaticale 61, 33-36. Martin, R. (1987): Langage et croyance. Les "univers de croyance" dans la theorie semantique. Bruxelles: Mardaga. Milner, J. (1978): De la syntaxe ä l'interpretation. - Paris: Seuil. Moeschler, J. et A. Reboul (1994): Dictionnaire encyclopedique de pragmatique. - Paris: Seuil. Muller, C. (1991): La negation en franfais. - Geneve: Droz. Orlandini, A. (2001): Negation et argumentation en latin. Grammaire fondamentale du latin. Tome VIII. Bibliotheque d'Etudes Classiques. - Louvain-Paris: Peeters. Reichenbach, Η. (1947): Elements of Symbolic Logic. - London-New York: MacMillan. Vairel-Carron, Helen (1975): Exclamation, ordre et defense. Analyse de deux systemes syntaxiques en latin. - Paris: Les Belles Lettres.

Michel Marcoccia

La politesse dans les forums de discussion sur Γ Internet Regies externes, manifestations discursives et commentaires metacommunicatifs

1. Introduction

Les echanges discursifs sur le reseau internet (courrier electronique, forums de discussion, messageries synchrones) constituent un nouveau champ d'etude pour Panalyse des dialogues authentiques. De recentes recherches francophones (Guegen et Tobin, eds, 1998; Anis, ed., 1999; Flichy, ed., 1999) s'inscrivent dans ce champ, dejä constitue en sous-discipline dans les pays anglo-saxons (les Computer-Mediated Communication Studies) depuis la parution des ouvrages pionniers diriges par Lea (1992) ou Herring (1996). Parmi les travaux se situant dans les paradigmes de la linguistique pragmatique et de Panalyse conversationnelle, certains s'Interessent plus particulierement ä la question de la politesse (Harrison 2000, Maricic 2000) ou, plus largement, aux normes de comportement instituees sur le reseau internet ou Usenet (Mc Laughlin, Osborne et Smith 1995, Marcoccia 1998). On sait bien que la politesse fait partie des objets d'etude privilegies de l'analyse conversationnelle, depuis la Politeness Theory de Brown et Levinson. Ce n'est pourtant pas seulement par reflexe "conditionne" que nous nous interesserons dans cet article ä cette question, mais bei et bien parce qu'elle est centrale si Ton veut comprendre la dynamique des echanges mediatises par ordinateur et identifier les caracteristiques du cyberespace comme "lieu" de dialogues. Pour decrire les regies de politesse ou, plus globalement, les normes de comportement έ respecter dans une situation de communication donnee, on dispose en fait de trois objets d'etude differents: Les regies de politesse "externes", telles qu'elles peuvent etre formalisees de maniere prescriptive dans des traites de savoir-vivre ou de maniere descriptive dans les theories de la politesse. Les regies "internes" qui constituent le fonctionnement effectif de la politesse, l'ensemble des strategies mises en ceuvre par les interactants au cours des echanges. Les divers commentaires metacommunicatifs produits par les interactants eux-memes au cours des echanges, par lesquels ils explicitent les regies et signalent leur appropriation.

316

Michel Marcoccia

Nous etudierons ces trois manifestations de la politesse sur Internet et presenterons alors l'hypothese selon laquelle la politesse sur Internet n'a pas seulement pour but de menager la relation interpersonnelle, mais a aussi une fonction sociale (creer et consolider la "communaute virtuelle"), taxömique (instaurer un rapport de places), et argumentative. Le corpus sur lequel nous baserons notre analyse est constitue de messages adresses ä quatre forums de discussion francophones tres divers'. Les forums sur lesquels porte notre analyse sont des forums non moderes: il n'y a pas d'animateur qui selectionne les messages avant de les publier. lis n'ont pas non plus de chartes specifiques definissant leurs regies.

2. Les regies externes: la netiquette

2.1. Qu'est-ce que la netiquette? Le principe meme de Internet est de mettre en relation des interlocuteurs qui le plus souvent ne se connaissent pas mutuellement et peuvent appartenir a des cultures tres differentes. Pour preserver la "qualite" des echanges, les acteurs centraux du reseau (les "pionniers", les animateurs de forum de discussion) ont elabore un code de savoircommuniquer afin de susciter une attitude d'auto-regulation chez les interactants. Ce code est appele la netiquette: l'etiquette du net. La netiquette se presente comme un systeme normatif et axiologise. C'est un ensemble de regies fixes determinant "ce qu'il faut faire". Ces regies sont presentees dans divers textes disponibles sur le web qui peuvent etre rapproches des traites de savoir-vivre anciens ou modernes, aussi bien du point de vue de leur fonction ideologique que de leur aspect formel (Marcoccia 2000b). Pour decrire les regies de la n6tiquette, nous avons procede a une synthese des trois textes apparaissant comme les "netiquettes de reference": The Net: User Guidelines and Netiquette, de Arlene Rinaldi, Netiquette Guidelines, de Susan Hambridge, Netiquette, de Virginia Shea. Ces sont les textes les plus cites; on les trouve dans un tres grand nombre de sites web.

Nous avons analyse 300 messages dans chacun des forums suivants: fr.comp.sys.mac.: discussion entre utilisateurs de Macintosh, fr.rec.arts.bd: discussion entre amateurs de bände dessinee. fr.rec.boissons.vins: discussion entre amateurs de vins, echanges. fr.rec.sports.football: discussion entre amateurs de football.

La politesse dans les forums de discussion sur l 'Internet

317

La nötiquette est basee sur six categories de regies, chacune etant supposee resoudre un Probleme specifique pose par la communication mediatisie par ordinateur: la politesse au sens strict (par exemple, le menagement des faces), la cooperation dans la production des messages, le contenu des messages, Γ identification des scripteurs et des destinataires, Γ adaptation aux lois de la vie "reelle", l'adaptation aux contraintes techniques du reseau. Nous ne nous interesserons qu'aux quatre premieres categories de regies, que nous considererons comme appartenant ä la politesse et au savoir-vivre dans un sens large.

2.2. Politesse et menagement des faces Le principe de menagement des faces, tel qu'il a ete systematise par Brown et Levinson (1987), fait partie des regies de la netiquette. Trois regies sont instituees, correspondant aux trois strategies de menagement des faces: L'evitement. "Vous n'enverrez pas de messages haineux (on les appelle des 'flammes') meme si on vous provoque" (Hambridge). L'adoucissement. "If you want to express a strong opinion, cute euphemisms and made-up expletives are usually acceptable" (Shea). La reparation. Si vous envoyez un message menafant, excusez-vous: "a gracious apology is almost always appreciated" (Shea). La netiquette rappelle aussi deux regies de politesse, qui prennent un sens particulier dans un contexte de reseau informatique: respectez le territoire spatial et temporel d'autrui. Soyez conscient de la longueur des messages que vous envoyez. Annexer de grands fichiers, tels que des documents en Postscript ou des programmes, peut rendre vos messages si grands qu'ils peuvent ne pas etre transmis ou au moins consommer une part exagiree de ressources (Hambridge). When you send email or post to a discussion group, you're taking up other people's time (or hoping to). It's your responsibility to ensure that the time they spend reading your posting isn't wasted (Shea).

La netiquette preconise aussi de produire des messages dont le contenu emotionnel ou humoristique doit etre mesure. Cette regie peut etre vue comme une regle de politesse, car eile revient ä conseiller de limiter le caractere potentiellement mena?ant des messages:

318

Michel Marcoccia

Attendez d'avoir dormi avant d'envoyer des reponses chargees d'emotion (Hambridge). Soyez prudent quand vous utilisez des sarcasmes et de l'humour. Sans une communication face ä face votre blague pouirait etre vue comme une critique (Rinaldi).

Differents espaces de discussion, pouvant avoir chacun des usages specifiques, coexistent sur Internet. La netiquette tient compte de cette particularite et institue une regle d'adaptation des enonces au "site" dans lequel ils sont produits, regle proche de ce que la tradition appelle la bienseance. Netiquette varies from domain to domain. What's perfectly acceptable in one area may be dreadfully rude in another (Shea).

2.3. Principe de cooperation La netiquette ne se contente pas d'instituer, ou de rappeler, des regies de politesse, eile propose aussi quelques principes de composition des messages envoyes sur le reseau. Ces principes de composition correspondent a ce que Grice (1979) appelle le principe de cooperation·. que votre message corresponde ä ce qui est exige de vous. La netiquette reprend trois regies du principe de cooperation, correspondant a deux maximes (relation et modalite): soyez pertinent, clair et bref. Gardez vos questions et vos commentaires pertinents quant au point d'interet du groupe de discussion (Rinaldi). Limitez la longueur des lignes (Rinaldi). Make sure your notes are clear and logical. It's perfectly possible to write a paragraph that contains no errors in grammar or spelling, but still makes no sense whatsoever. (Shea).

Comme Leech (1983), nous considerons que le principe de cooperation releve de la politesse.

2.4. Le contenu des messages Alors que les principes de menagement des faces et de cooperation portent sur la maniere de produire un message, d'autres regies moins nombreuses et plus specifiques portent sur le contenu des enonces. Sont declares indesirables les messages ä contenu commercial, trop personnels, racistes ou sexistes. N'utilisez pas le r£seau pour des activites commerciales ou priv6es (Rinaldi). N'importunez pas les autres utilisateurs avec des informations personnelles telles que sexe, age ou situation (Hambridge).

La politesse dans les forums de discussion sur l 'Internet

319

Des messages tenant du harcelement sexuel ou racial peuvent aussi avoir des implications legales (Rinaldi).

2.5. L'identification des intemautes La netiquette accorde aussi une large place ä la question de l'identification des utilisateurs du reseau. Puisque l'environnement technique ne permet pas un acces fiable ä l'identite des interactants (meme l'adresse electronique est falsifiable), la netiquette incite les intemautes a apporter dans leurs messages les informations manquantes et, surtout, ä ne pas tirer profit du dispositif technique pour endosser une fausse identitd (ce qu'on peut considerer comme une regle de politesse ou, plus largement, de morale). Incluez votre signature au bas des courriers electroniques. Votre signature doit inclure votre nom, votre position, votre affiliation et votre adresse Internet (Rinaldi) La falsification et la mystification ne sont pas admises comme comportement (Hambridge).

3. La netiquette en pratique

3.1. Quelques manifestations discursives de la netiquette L'analyse de forums de discussion francophones permet d'observer diverses manifestations discursives des regies de la netiquette. Entre autres exemples, le menagement des faces pourra se manifester par le remplacement d'un acte de langage mena?ant par un autre acte de langage moins mena?ant, une critique par une question, comme dans le message suivant: (1) Echange le forum fr.rec.boissons.vins Prenez note: Cotteaux du Languedoc - Cepage Syrah & Grenache Chateau de Puech Haut: cuvee Folie de Courbessac 1996. Medaille d'argent concours gendral Paris 1998. Non merci... je ne digere pas les medailles:-) Reponse: et pourquoi n'aimes-tu pas les medailles? il me semble qu'il faut gouter avant de juger. Amities

320

Michel Marcoccia

On peut observer aussi des procedes d'adoucissement, par exemple l'adoucissement d'une correction par un enonce flatteur et un smiley souriant (sur le röle des smileys dans la politesse: Marcoccia 2000a: 259), comme dans le message suivant: (2) Echange sur le forum fr.rec.arts.bd Ce n'est pas dans la collection Mimolette, mais c'est quand meme chez l'Association: "Nadia et les autres" de l'excellent Guy Delisle. Un bouquin sans paroles, au dessin minimaliste, beau, dröle, cruel, ironique et un peu effrayant. Un delice pour psychanalyste sans doute. Riponse: Appollo, tu causes vachement bien de ce tres beau livre, qui ne s'appelle pourtant pas "Nadia et les autres" mais "Aline et les autres". Mais je suis d'accord avec tout, sinon, hein:-)

3.2. L'importance du metalangage L'analyse des forums de discussion nous permet aussi d'observer de nombreux messages ne respectant pas les regies de la netiquette mais comprenant un enonce metalangagier sur cette violation. Ces enonces metalangagiers peuvent tenir lieu de reparation: une excuse pour ne pas avoir repondu a des messages, comme dans l'exemple suivant: (3) Message adressd au forum fr.rec.boissons.vins: Salut ä tous, Le petit coup de gueule bien nicessaire m'a gentiment rappete que j'avais re?u des reponses trfes instructives ä mes questions sur le boise des vins. Disposant de plusieurs BAL, je n'avais pas pris la peine de rdpondre. Recevez ici toutes mes excuses et mes remerciements.

Pour montrer qu'il est conscient de la regie selon laquelle il ne faut pas encombrer 1'espace d'autrui, un utilisateur pourra indiquer dans la zone "subject" que son message est long et, ainsi, s'excuser par avance, comme dans l'exemple suivant: (4) Zone titre d'un message adresse au forum fr.rec.boissons.vins. From: Newsgroups: fr.rec.boissons.vins Subject: Re: L'hyper specialise dans la boisson [long]

La violation de la "regle de mesure" pourra se traduire par des excuses (exemple 5) ou par l'utilisation de smileys indiquant que le message est ironique (exemple 6): (5) Extrait d'un message adresse au forum fr.rec.boissons.vins. Merci pour cette precision et desole d'avoir donne l'impression que j e m'emportais. (6) Extrait d'un echange de messages postes au forum fr.comp.sys.mac Vu que Quartz semble deja bien debute, c'est possible.

La politesse dans les forums de discussion sur l 'Internet

321

Reponse: Disons que Quartz, 9a fait 15 ans qu'il a d6ja bien ddbute;-) La violation de la regle de pertinence peut aussi donner lieu a un commentaire metacommunicatif: (7) Extrait d'un message adresse au forumfr.rec.boissons.vins.Excusez mon intrusion, y a t'il un groupe de news ds lequel je pourrais trouver la recette de la tarte au citron? Merci et VIVE le VIN de BELLET Marcel Maurel L'exemple suivant contient un cas d'explicitation ironique qui permet au scripteur de montrer qu'il respecte (ou, du moins, n'ignore pas) l'interdiction des messages ä caractere commercial: (8) Extrait d'un message adress6 au forumfr.rec.boissons.vinsPour decouvrir les muscadets, on peut aller faire un tour du cote de la cave de Longchamp, rue Georges-Laffont, a Nantes. Attention: ceci n'est pas de la pub:-) Nous pouvons qualifier ces precedes discursifs de commentaires metacommunicatifs (ou metacommunicationnels) dans le sens oü ils consistent en une objectivation de l'activite verbale et interactionnelle et referent ä la bonne conduite de l'interaction (Gaulmyn 1987). Adoptant une position habituelle dans l'ecole franijaise d'analyse de discours, nous considerons que la presence d'un commentaire metacommunicatif Signale qu'un enjeu particulier pese sur ce qui fait l'objet du commentaire. Comme le souligne Maingueneau, "loin d'etre un procede pour corriger les rates de la communication, le metadiscours constitue un Symptome et doit etre apprehende a travers ce Statut" (1991: 145). Ici, le metalangage peut constituer un moyen d'exhiber les regies jugees vraiment importantes par les internautes, une Strategie pour construire une image d'internaute maitrisant parfaitement les regies, une maniere d'affirmer son appartenance au groupe par l'exhibition des regies supposees s'exercer sur ses comportements.

4. Violation de la netiquette et rappels ä Γ ordre

4.1. Violations de la netiquette: generalites et specificites Les exemples qui precedent montrent que la netiquette n'est pas toujours respectee mais que, quand eile ne Test pas, ses violations donnent frequemment lieu a des commentaires

Michel Marcoccia

322

metacommunicatifs; ce qui montre que ses regies sont neanmoins interiorisees par les utilisateurs. Une observation complete du corpus montre d'ailleurs que la netiquette est plus souvent invoquie que reellement respectee (et qu'elle a done un röle un peu paradoxal). Toutefois, l'analyse des forums de discussion nous permet d'observer egalement des violations des regies de la N6tiquette ne donnant pas lieu a des commentaires metacommunicatifs produits par les auteurs de ces violations. On peut distinguer deux types de violation des regies: -

des violations qu'on observe dans n'importe quel forum, par exemple des messages trop longs, des entorses ä la regle de pertinence, des absences de signature explicite.

-

des violations liees a la specificite du forum dans lequel elles apparaissent. Par exemple, des messages ä caractere commercial dans le forum sur le vin ou des messages agressifs dans le forum sur le football.

Cette distinction renvoie au probleme de la coexistence d'espaces de dialogues specifiques sur Internet. Ainsi, une violation de la netiquette peut-elle etre un comportement acceptable - voire attendu - sur un forum donne et justifier la manque d'excuse. Toutefois, lorsque la violation de regies de la netiquette n'est pas accompagnee d'un commentaire d'excuse, eile suscite souvent un commentaire metacommunicatif produit par un autre utilisateur du forum, sous la forme d'un rappel a l'ordre.

4.2. Rappels ä l'ordre Lorsqu'on analyse un forum de discussion, on est etonne par le nombre de commentaires sur le comportement d'autrui: des rappels ä l'ordre adresses aux internautes n'ayant pas respecte une "regle de comportement". La regie la plus souvent rappelee est la regle de pertinence, comme dans 1'exemple suivant: (9) Forum fr.comp.sys.mac Ce qui est regrettable e'est quand un detenu est condamne sur des presomptions, mais quand il y a preuve... Reponse: 11 y a beaucoup, beaucoup, beaucoup de choses ä dire sur ce sujet. Comme nous sommes sur fesm et non sur fr.soc.politique, je suggfere d'arreter le debat ici et maintenant tant que le climat est ä peu pres calme.

On peut aussi observer une mise en avant du principe de menagement des faces et plus particulierement de la regle de mesure. Cette regle est rappelee dans le forum sur les Ma-

La politesse dans les forums de discussion sur l 'Internet

323

cintosh, oü eile donne lieu ä un commentaire metacommunicatif sur les risques de l'ironie, dans Pexemple 10. (10) Forum fr.comp.sys.mac > > > Resultat: 15 secondes soit 707 Ko/secondes. > >

> > En SCSI ou IDE? > > Puisque la difference est tellement importante, tu devrais pourvoir le deviner rien qu'avec les > specs:-> Serge, laisse done l'ironie a ceux qui savent la manier avec dexterite.

L'interdiction d'envoyer des messages ä caractöre commercial est aussi souvent rappelee, particulierement dans le forum consacre aux vins: (11) forum fr.rec.boissons.vins Joe a ecrit: >surement les meilleurs prix du marche >http://www.vins-fr.com Puis, il a ajoute: > je suis desole de ne pouvoir satisfaire tout le monde. > N'hesitez pas a communiquer les sites qui offrent mieux.

Reponse: Surement pas ! Ce forum n'est la pour parier de tarifs, mais pour parier de vins. C'est clair, non?

On voit done que les participants des forums de discussion construisent un ensemble de normes internes de comportement, qui ne reprennent pas toutes les preconisations de la netiquette et n'en valorisent reellement que quatre: -

II faut etre pertinent. Les forums ne se definissent que par leur contenu thematique; il est done normal que cette regle soit mise en avant.

-

II faut etre clair. II faut menager la face de l'autre: il faut assurer la finalite cooperative des echanges.

-

II faut eviter les messages ä caractere commercial et defendre ainsi un des fondements ideologiques de la cyberculture d'origine, mise a mal par la "marchandisation" de rinternet.

324

Michel Marcoccia

5. Conclusions et perspectives

Le rappel de certaines regies et l'absence de rappel pour d'autres ont une fonction plus ideologique que communicationnelle. La distance que prennent les intemautes avec la netiquette leur permet d'afficher les valeurs qui leur semblent reellement importantes, pour eux-memes et eventuellement pour le groupe constitue par les utilisateurs du forum auquel ils participent. Ainsi, les manifestations discursives de la netiquette, les commentaires metacommunicatifs et les rappels a l'ordre contribuent a construire un

sentiment

d'appartenance ä un collectif specifique, qui est moins l'ensemble des intemautes que la communaute virtuelle formee par les participants ä un forum de discussion particulier. La netiquette ne correspond qu'ä un cadre general: la maniere de se comporter dans un forum de discussion in abstracto. Ce cadre subit alors plusieurs restrictions ä partir desquelles s'elaborent des modeles d'interaction specifique, souvent inspires par des modeles interactionnels sous-jacents. On peut alors definir chaque forum en le mettant en relation avec une forme prototypique d'interaction en face έ face: le forum sur le football a comme modele la discussion entre supporters et assure une fonction evidente d'exutoire, le forum sur les Macintosh reactualise les normes de comportement en usage dans des discussion techniques ä finalite cooperative (interaction de travail entre Ingenieurs, par exemple). De plus, les regies qui font l'objet de rappel ä l'ordre dans un forum donne renvoient souvent ä un aspect du cadre du forum qui est mal defini, polemique ou paradoxal. Ainsi, la regle qui pose le plus de problfemes dans le forum sur le vin est la regle interdisant les messages ä caractere commercial. Elle est assez souvent violee, et donne lieu ä des rappels k l'ordre ou ä des justifications de participants assurant qu'ils respectent la regle ou s'excusant de ne pas la respecter. L'enjeu qui pese sur cette regle est sans doute lie au fait que la finalite du forum n'est justement pas tres claire quant ä son aspect commercial. Donner le nom d'un proprietaire recoltant, est-ce ou non de la publicite? Demander le prix d'un bon vin, est-ce ou non initier un echange έ caractere commercial? Par ailleurs, il est evident que la netiquette n ' a pas seulement pour fonction d'instaurer des regies de comportement supposees assurer la qualite des echanges. En effet, la netiquette et ses diverses manifestations dans les discours font partie des ressources communicatives permettant ä un participant de se construire un role au cours de l'interaction. On peut, par exemple, faire l'hypothese que des utilisateurs appartenant au premier cercle des abonnes ä l'Internet produisent des commentaires metacommunicatifs pour mettre en avant leur maitrise de la netiquette ou des regies speeifiques du forum et afficher ainsi qu'ils sont des "anciens". Ainsi, les commentaires metacommunicatifs jouent-ils un röle important dans la mise en place (Kerbrat-Orecchioni: 1987); ils permettent έ un

La politesse dans les forums de discussion sur l 'Internet

325

participant d'afficher sa maitrise des protocoles de l'interaction, ce qui est un marqueur de position elevee. Enfin, les commentaires metacommunicatifs peuvent avoir une fonction argumentative: un rappel a l'ordre peut tenir lieu de procedö argumentatif. Plantin ( 1 9 9 5 ) appelle "argument du paralogisme", le precede qui consiste ä argumenter en denonpant un argument de son opposant c o m m e ne repondant pas aux regies de la "bonne argumentation". D e maniere similaire, rappeler ä l'ordre un participant, en indiquant que son message viole une regle de la netiquette ou une regle specifique d'un forum, peut etre un moyen de refuter ce message.

References bibliographiques

Anis, J. (ed.) (1999): Internet, communication et langue franfaise. - Paris: Hermes Science Publications. Brown, P., S. Levinson (1987): Politeness. Some universals in Language Use. - Cambridge: Cambridge University Press. Flichy, P. (ed.) (1999): Internet: un nouveau mode de communication? - Numero special de la revue Riseaux, 17-97. Gaulmyn, M.-M. de (1987): Reformulations et planification metadiscursives. - In: J. Cosnier, C. Kerbrat-Orecchioni (eds.) Decrire la conversation, 167-198. Lyon: Presses Universitaires de Lyon. Grice, H. P. (1979): Logique et conversation. - Communications 30, 57-72. (1975) Logic and conversation. Guegen, N., L. Tobin (eds.) (1998): Communication, soci6t£ et Internet. - Paris: L'Harmattan. Hambridge, S., (1995): Netiquette Guidelines. - Version franijaise disponible sous le titre "Les regies de la netiquette" http://www.sri.ucl.ac.be/SRI/rfcl855.fr.html. Harrison, S. (2000): Maintaining the virtual community: use of politeness strategies in an email discussion group. - In L. Pemberton, S. Shurville (eds.) Words on the Web. Computer Mediated Communication, 69-78. Exeter: Intellect Books. Herring, S. C. (ed.) (1996): Computer-Mediated Communication. Linguistic, Social and Cross-Cultural Perspectives. - Amsterdam/ Philadelphia: Benjamins. Kerbrat-Orecchioni, C. (1987): La mise en places. - In J. Cosnier, C. Kerbrat-Orecchioni (eds.) Decrire la conversation, 319-352. Lyon: Presss Universitaires de Lyon. Lea, M. (ed.) (1992): Contexts of Computer-Mediated Communication. - Hemel Hempstead: Harvester Wheatsheaf. Leech, G. (1983): Principles of Pragmatics. - London/New York: Longman. Maingueneau, D. (1991): L*Analyse du discours. Introduction aux lectures de l'archive. - Paris: Hachette. Marcoccia, M. (1998): La normalisation des comportements communicatifs sur Internet: 6tude sociopragmatique de la Netiquette. - In N. Guegen, L. Tobin (eds.) Communication, Soci£te et Internet, 15-32. Paris: L'Harmattan.

326

Michel Marcoccia

Marcoccia, M. (2000a): Les smileys: une reprisentation iconique des emotions dans la communication m e d i a t e par ordinateur. - In C. Plantin, Μ. Doury, V. Traverso (eds.) Les imotions dans les interactions, 249-263. Lyon: ARCI-Presses Universitaires de Lyon. Marcoccia, M. (2000b): La sociabilite sur Internet: reflexions sur une "revolution informationnelle". - In D. Bourg, J.M. Besnier (eds.) Peut-on encore croire au progrfes? Paris: Presses Universitaires de France (a paraitre). Maricic, I. (2000): Cyberpoliteness: Requesting strategies on the linguist list. - Paper presented at the 7th International Pragmatics Conference (9-14 July 2000, Budapest, Hungary). Mc Laughlin, Μ., K.K. Osborne, C.B. Smith (1995): Standards of conduct in Usenet. - In S.G. Jones: Cybersociety. Computer-Mediated Communication and Community, 90-111. Thousand Oaks: Sage Publications. Plantin, C. (1995): L'Argument du paralogisme.-Hermes 15, 245-262. Rinaldi, A. (1995): The Net: User guidelines and netiquette. - http://www.fau.edu/netiquette/net Shea, V. (1994): Netiquette. - San Francisco: Albion Books. Version "en ligne": http://www.albion.com/catNetiquette.html

Liana Pop

Actes vs operations Vers de nouveaux outils dans l'analyse du dialogue

Nous proposons, pour cette annee 2000 et pour les 10 ans d'analyse du dialogue, un petit bilan des categories les plus importantes utilisees jusqu'a present dans l'analyse du discours et du dialogue. II s'agira de categories apportees par differents types d'analyse, ceux, notamment, qui se sont le plus impliques dans la description du discours, telles la theorie des actes de langage (TAL), les analyses du discours (AD), l'analyse grammaticale (AG), la theorie de la refirence (TR) et la theorie des espaces discursifs (TED). Nous adopterons ici une perspective cognitiviste et demontrerons comment la categorie d'acte de langage, s'averant inadequate et insuffisante, a reclame ä tour de role d'autres categories plus larges, comrae celles, principalement, d'acte discursif et d'operation discursive. La demarche theorique que nous proposons sera appuyee sur un extrait de discours oü nous avons pris les quelques exemples necessaires a une demonstration minimale.

1. La theorie des actes de langage (TAL) et l'analyse du discours (AD)

La theorie des actes de langage, nee comme reaction ä l'illusion descriptive, ne tardera pas ä creer a son tour un autre type d'illusion, qu'on pourrait appeler illocutionnaire, car elle soutient que tout acte de langage est porteur d'une force illocutionnaire repertoire dans la langue. Or, des que le besoin se presenta de prendre pour niveau d'analyse non plus un seul acte mais plusieurs tels qu'ils se presentent normalement dans le discours, la contrainte de l'illocutoire ainsi imposee par la TAL, s'est averee trop severe et les inventaires d'actes plutöt insuffisants.

1.1. Definition prototypique des actes de langage (les meilleurs representants) On trouve en effet dans la definition prototypique des actes de langage une fonction dite illocutoire (ou illocutionnaire) FI qui accompagne un contenu propositionnel CP donne.

Liana Pop

328

Ces deux composants feraient comme des conditions necessaires et süffisantes ä la definition de cette catigorie: en effet, tandis que le contenu propositionnel fait r6f6rence au monde, la force illocutionnaire refere ä l'acte dont le locuteur a voulu investir ce contenu et peut etre denommöe ä l'aide d'un verbe naturel, dit illocutionnaire. L'existence de ces "etiquettes verbales" (noms pour des categories d'actes: promettre, ordonner, feliciter, demander, refuser, accepter, s 'excuser, etc.) est la preuve d'une tres forte categorisation de certains comportements linguistiques, dans la conscience des locuteurs. Ainsi, dans la sequence discursive ci-dessous - prise d'une interview et ού nous avons delimite et numerate chaque acte de 1 a 25 - pourrait-on appliquer comme "etiquettes illocutionnaires": -

pour les actes 8 et 10 - aveux; pour l'acte 13 - demande d'excuse·, pour l'acte 22 - rappel·, etc.

En dehors des marqueurs denominatifs (comme les verbes illocutionnaires ou performatifs: avouer/j'avoue que..., s'excuser/je m'excuse...), il existe, en regle generale, d'autres marqueurs segmentaux qui peuvent etre utilises, moins forts, mais tres souvent conventionnalises; certains ont ete appeles marqueurs indicatifs (par ex. l'imperatif pour l'ordre ou la demande), d'autres sont appeles potentiels (comme la forme declarative qui peut etre ou marqueur de description, ou de compliment, ou d'aveu, etc.) (cf. Etudes de linguistique appliquee n° 44). Au type de marqueurs segmentaux s'ajoutent, dans la chaine verbale des marqueurs prosodiques (pauses, decrochements intonatifs, contours melodiques) qui indiquent ce que Ton a appele des "frontieres d'acte" (cf. Ferrari et Auchlin 1995), c'est-ä-dire des signaux qui indiquent que dans le flux discursif il faut indiquer et percevoir des unites. A l'ecrit, ces pauses d'actes portent des marques graphiques plus ou moins fortes, et si gdneralement la virgule est indicative d'acte, le point est un signe prototypique de phrase (au sens discursif de ce terme, de mouvement discursif, cf. Roulet 1994). BP bon'/vous etes romantique2/vous etes na'ifVvous etes timide 4 /vous etes bon 5 / tout 9a 6 / mais n'empeche 7 / G S je ne suis pas bon^/je suis 9 /j'ai mes defauts comme tout le monde 1 ® BP oui mais enfin"/vous etes' 2 /excusez-moi I' expression 1 3/vous etes un 14 /vous etes 1 5 /un 16 /un dröle de lascar avec les femmes 1 5 /parce q u e 1 ' / vraiment1*®/ vous etes l'infidele total 1 7 '7et il y a 1 9 ' / tout de meme 2 0 /votre 21 / votre premiere f e m m e 1 9 /vous le racontez la aussi 22 /moi j'aimerais bien 2 3 / G S ma premifere femme m'avait dit qu'elle se suiciderait 24 ' BPvoilä 2 5 / G S si je la trompais 24 (Apostrophes, 1 9 8 1 : B. Pivot - G. Simenon)

Actes vs opirations

329

1.2. Elargissement dans l'inventaire d'actes (representants moins bons et meme mauvais): actes discursifs, semi-actes... L'hypothese des actes illocutionnaires s'est averee trop forte pour l'analyse du discours oü, dans les successions d'actes, on a le plus souvent du mal a trouver "les meilleurs representants" de la categorie classique "acte de langage". Ainsi, ä cöte d'un contenu propositionnel CP facilement reperable, la fonction illocutoire est-elle souvent, et notamment dans les sequences monologales, non identifiable. D'autres types de fonctions semblent s'imposer dans les sequences monologales d'actes qui sont des fonctions relationnelles et auxquelles on a donnd le nom de fonctions interactives (cf. Roulet et al 1985), notees ici fl. Parmi elles, des verbes naturels suffisamment utilises par les locuteurs (argumenter, justifier, reformuler, preciser, resumer...), mais aussi des "categories de linguistes" comme thematiser, cadrer, etc., moins a la portee des locuteurs ordinaires. Ce qui signifierait que, par rapport aux actes de langage pris ä part, qui se presentent comme categories naturelles - ayant des noms dans le langage commun - les categories de l'analyse du discours sont moins perceptibles naturellement. Leurs noms qui ne se retrouvaient plus dans les inventaires classiques d'actes de langage, et le fait de se definir sur des fonctions essentiellement discursives - relations dans le discours - , leur a valu le nom d'actes discursifs (vs actes de langage). Dans le corpus pris pour exemple ci-dessus, les actes de 2 ä 5 (une sequence reformulative), 10 (un argument), 17 (une justification-commentaire) et 19 (une thematisation), non repertories comme "actes de langage", sont classes par les analystes du discours comme "actes discursifs". La plupart concernent le travail de structuration du discours et leur portee principale ne conceme ni les interlocuteurs ni la reference au monde. Du point de vue de la definition, ce seraient alors des representants moins bons et meme mauvais de la categorie d'acte (dans son acception premiere d'acte de langage). Un premier elargissement s'est done produit par rapport ä la categorie de depart. Un deuxieme 61argissement s'ensuit lorsqu'on constate que certaines phrases mettent en scene plus d'un enonciateur (vs un seul enonciateur pour un acte). Ce entere a suffi pour dire qu'une partie de phrase pourrait representer ä eile seule "une sorte d'acte" (pas tout ä fait un acte, car son contenu propositionnel CP etait insuffisant). Et la decision semble prise d'appeler le phenomene un semi-acte (cf. Rubattel 1996), en parfaite concordance d'ailleurs, avec la theorie revisitee du prototype, qui admet des degres et des demimesures pour ces categories. Malheureusement, cette categorie intermediate est, par apres, "retiree du marche" - ä tort, dirions-nous, car elle relevait d'un phenomene flou, typique du discours naturel, d'un faire distinct, polyphonique, ä l'interieur d'une seule et meme phrase grammaticale.

330

Liana Pop

Pour resumer ce chapitre, nous dirions que l'arrivee des deux categories - celle d'acte discursif, d'un cote, et celle de semi-acte, de l'autre - est symptomatique d'une insuffisance de la categorie initiale ä laquelle on a donne une definition trop forte et qui montre un pouvoir explicatif trop faible pour l'analyse du discours.

1.3. Les non-actes (les exclus): rates, faux-departs, modalisateurs, marqueurs, connecteurs, regulateurs, ponctuants, particules... Mais l'analyse du discours a fait plus que reveler ces representants moins bons de la categorie d'acte de lahgage (les actes discursifs ou les semi-actes): eile a egalement defini des phenomenes places en dehors des frontieres de ces deux classes, ces unites notamment qui ne relevent nullement de la categorie d'actes car ne possedant ni contenu propositionnel CP fini, ni fonction de type illocutoire FI ou interactive fi. Nous pensons, par exemple, aux inachevements elimines generalement des analyses mais que 1'etude de l'oral mentionne souvent comme rates ou faux departs. Rien que dans notre corpus, il en existe plusieurs, car il s'agit d'une sequence oü Bernard Pivot est oblige de chercher les mots les moins genants pour rappeler ä Georges Simenon qu'il aimait un peu trop les femmes: 7 mais η 'empeche, 9 je suis, 12 vous etes, 14 vous etes un, 16 un, 21 votre et 23 j'aimerais bien. Le travail de formulation y est tres important. Ces non-actes sont des unites a contenu propositionnel CP flou, difficilement analysable, et qui ont ete classees comme modalisateurs, marqueurs, connecteurs, ponctuants, particules, regulateurs, etc. Dans le texte analyse, nous avons trouve comme modalisateurs: 18 vraiment, 20 tout de mime -

comme marqueurs / connecteurs: 1 bon, 19 ily a, 11 oui mais enfin comme ponctuants: 1 bon, 6 tout ςα comme regulateurs: 11 oui, 25 voilä comme particules: 1 bon

Les "etiquettes" sont ici legion pour dinommer souvent un seul et meme phenomene, car ces categories - meme si eile sont inventees par les linguistes - semblent etre imparfaites: elles se recoupent souvent pour un seul et meme mot {bon, par ex., est classe connecteur, ponctuant et particule ä la fois, en fonction de la perspective d'analyse). Notons egalement que pour les locuteurs ordinaires, des verbes comme modaliser, connecter, marquer, riguler et ponctuer ne sont point evidents ä l'usage, ce que nous considerons comme un Symptome de categorisation naturelle difficile. En plus, les frontieres prosodiques sont plus rares pour "delimiter" ces mots, apparemment de moindre importance dans le discours et tentant de passer, pour la plupart du temps, inaperpus (les linguistent les considerent souvent de simples "traces").

Actes vs operations

331

Plus la question de la segmentation du discours pose des problemes aux chercheurs, plus la quete des unites discursives

minimales

deviendra achamee, et les concessions

faites aux unites plus importantes. Signe de malaise dans un carcan trop etroit: des criteres de plus en plus vagues vont etre poses, afin que les descriptions deviennent plus plausibles, telles: l'autonomie pragmatique -

la prise en charge par un locuteur et

-

une predication (füt-elle relationnelle - le cas des connecteurs - ou non)

Par rapport aux actes, vus comme predications autonomisables, les connecteurs peuvent dorenavant etre perpus comme predicats relationnels non autonomisables (cf. Rossari 1996) et le mot lexical ä lui seul comme possible candidat au Statut d'unite discursive. La notion de predication semble ainsi faire le pont entre ce qui etait per?u comme unite, car autonome, et ce qui etait perfu comme non-unit6, car non autonome. D'autre part, la notion d'acte semble ceder le pas ä la notion, plus generale, d'unite

discursive.

2. La theorie grammaticale (TG)

La theorie des actes de langage identifiait dans tout acte locutionnaire une predication, et, selon cette perspective, l'absence d'une predication retirerait le Statut d'acte ä une expression. Or, justement, les notions de verbes parenthetiques

(recteurs faibles) (cf. Recanati

1984, Blanche-Benveniste 1989) et de predication seconde (Furukawa 1996) viennent appuyer, du cote des grammaires, la perspective de plus en plus floue des categories du discours en general et des actes en particulier. Elles laissent de toute fa?on entendre que la notion de predication est graduelle, que le discours est le lieu de manifestation de "predicats" plus ou moins forts ou, dans une terminologie prdtheorique, de lieu de rencontre de "faires" distincts, plus ou moins explicites. II s'ensuivrait alors qu'une perspective graduelle de la predication ouvre la voie ä une perspective graduelle des actes: il y aurait done des expressions plus ou moins actes (des semi-actes), en fonction du degre predicatif. La theorie grammaticale rencontre ici l'analyse du discours. Le marquage des predications se fait au niveau segmental et au niveau suprasegmental (voir, par ex., ce qu'on a appele frontieres d'actes) et/ou au niveau non verbal (actes non verbaux). Les marquages segmentaux sont de type plus analytique que les deux autres et se manifestent par des expressions predicatives plus ou moins fortes. Les predicats

332

Liana Pop

prototypiques etant des verbes recteurs par excellence, il va de soi que les predications secondes seront attachees ä des expressions verbales inachevees, ä des verbes recteurs faibles (ou parenthetiques) et ä des expressions autres que verbales, mais dites "predicatives" (adverbes, interjections, constructions absolues), etc. Ainsi, dans le texte analyse, peut-on considerer comme predications secondes, done comme quasi-actes: -

les inachevements 7 mais η 'empeche et 23 moij 'aimerais bien les expressions parenthetiques 13 excusez-moi I'expression et 22 vous le racontez la aussi, et les modalisations 18 vraiment et 20 tout de meme.

Au niveau suprasegmental, les contours melodiques et les pauses perpus a l'ecoute - qui indiquent des frontieres pour les segments a Statut de predications secondaires (mentionnes ci-dessus) 7 mais rt'empeche, 13 excusez-moi I'expression, 18 vraiment, 20 tout de meme, 22 vous le racontez la aussi, 23 moij'aimerais bien - ne font qu'appuyer la qualite d'unite distincte que les locuteurs veulent leur assigner. Ces marqueurs sont, par excellence, de type indicatif, procddural. Enfin, ä Γ oral, les actes non verbaux sont frequents et se manifestent par des gestes, des attitudes, des mimiques plus ou moins conventionnalises, qui se greffent sur les expressions verbales ou les relayent. Notre exemple offre deux de ces predications secondes - un haussement d'epaules en 8 et un sourire en 13 - accompagnant les actes verbaux respectifs qu'ils ne font que renforcer.

3. La theorie de la reference (TR)

En associant des types de predicats ä des types de reference, la theorie de la reference apporte plus de nuances ä la theorie grammaticale. Ainsi, suggere-t-elle pour les deux types de reference - la reference au monde et l'autoreference - la distribution suivante des predications dans le discours: predications au sujet du MONDE, qui sont des predications essentiellement descriptives {reference au Monde) predications au sujet de l'ENONCLATION (autoreference) predications au sujet de l'ENONCE (autoreference).

333

Actes vs opirations

La question la plus interessante qui s'ouvre pour l'analyse du discours est des lors celle de savoir lesquelles de ces predications prdferent les positions parenthetiques, secondes, effafables, c'est-ä-dire les positions moins fortes dans le discours - Celles de modalisateurs, de regulateurs, de ponctuants ou connecteurs - , et lesquelles prennent, de preference, l'allure d'actes proprement dits et occupent les positions fortes?

4. La theorie des espaces discursifs (TED)

Afin de repondre a cette question de fa9on satisfaisante pour l'analyse du discours, nous avons considere qu'il serait necessaire d'associer la categorie grammaticale, plus floue, de predication, a une catdgorie de discours plus permissive que l'acte ou l'unit6, d6finis parfois de fapon trop restrictive au niveau des fonctions ou au niveau des contenus propositionnels. II se trouve qu'une categorie dejä en usage dans l'analyse du discours - celle d'operation discursive - pourrait etre mise έ profit a cet effet. Telle que definie ailleurs (cf. Culioli 1990, Pop 2000), cette notion constituerait un elargissement de la notion d'acte, car elle se fonde sur des types de predications superordonnes ä ces demiers (du moins dans Pop 2000), et specifiques pour le travail enonciatif. Nous pensons que l'analyse du discours - afin d'dviter les inconvenients d'une terminologie butant sur les categories heteroclites et imparfaites d'acte, connecteur, modalisateur, regulateur, ponctuant, etc. - pourrait bien profiter de cette categorie moins naturelle, mais homogeneisante, d'operation discursive. Les types d'operations enonciatives seraient specialises sur les types de reference dans le discours. Nous les avons consideres comme des faires distincts, ä gerer simultanement par les locuteurs, et occupant chacun un autre "espace discursif' / referentiel (d.):

ENONCE

iNONCIATION ENONCE ENONCIATION

Id espace interdiscursif Md espace mitadiscursif Ip espace interpersonnel s espace subjectif D espace ref. au monde pp espace presuppositionnel Pd espace paradiscursif

op6rations de reprise discursive operations metadiscursives (reformulations, prdcisions, commentaires) operations d'appel ä l'interlocuteur (apostrophes) operations subjectives (attitudes, evaluations) operations essentiellement descriptives op6rations au niveau du savoir partage (explications, retours, rappels) opirations de formulations et recherche d'expr. (hesitations, rat6s)

334 SSG NV

Liana Pop Pro espace prosodique Is espace intersemiotique

expression suprasegmentale (graphique, orale) non verbal (gestuel, mimique, iconique, etc.)

Nous considerons que tous les types d'operations peuvent se manifester sous des expressions plus ou moins explicites: les formes assertees, descriptives sont les formes les plus marquees et font figure d'acte; les formes indicatives de l'expression ne sont pas considerees comme des actes, mais peuvent, pour autant, effectuer la meme operation dans le discours qu'effectuerait un acte. Ces espaces constitueraient done des "places" discursives stables, ramenant ä elles actes, connecteurs, modalisateurs, particules, etc. Ceci signifie qu'un acte peut parfois faire la meme chose qu'un connecteur, un modalisateur ou un geste. Ainsi, le prefixe et, respectivement l'adverbe, dans les enonces J'accentue

sur le

fait: il s 'est retire de I 'affaire ou II s 'est bien retire de I'affaire, aussi bien qu'une intensification de la voix, une pause significative ou encore un signe de la main suggerant le depart, effectuent-ils tous une seule et meme operation, sous des formes differentes: ils modalisent, "appuient", "accentuent" ou "soulignent" (attention aux metaphores utilisees pour cette operation typiquement discursive) le contenu propositionnel de la phrase. Deux types d'operations

enonciatives

devaluation et une autre, de type

y seraient

impliquees:

une

operation

subjective

metadiscursif.

Un autre exemple pourrait etre partiellement pris dans Pextrait de l'interview entre Bernard Pivot et Georges Simenon: l'enonce inacheve 9 je suis pourrait aussi s'exprimer, de fafon plus explicite par un acte en bonne et due forme - Je cherche le mot juste — ou, encore, de fa?on implicite, par une interjection (euh), une hesitation prosodique (une pause) ou un geste d'impuissance des mains ou des epaules. Toutes ces expressions - certaines equivalant a des actes, certaines a des particules (euh) ou a des signes non verbaux -

representent une seule et meme operation, que nous avons consideree de type

paradiscursif et inherente ä l'enonciation: la formulation

ou la recherche des mots.

Pour repondre maintenant ä la question: Lesquelles de ces operations sont plus sujettes ä faire actes et lesquelles plutot sujettes a faires des connecteurs,

des particules,

des

ponctuants, etc. ? Nous pouvons juste repondre en termes de preferences: 1.

Les operations

de riference

au monde sont davantage candidates ä des predications

fortes, de type descriptif et prennent plus facilement Failure d'actes - illocutionnaires ou discursifs - (ici "figures" au sens cognitif du terme). Ce sont les sens conceptuels par excellence, et des exemples prototypiques dans notre extrait comme: -

les enonces descriptifs 2 vous etes romantique, 3 vous etes naif 4 vous etes timide, 5 vous etes bon, 8 je ne suis pas bon, 9 j'ai mes defauts comme tout le monde, etc. (avec, pour fonction discursive, celle de reprise);

-

l'enonce 15 vous etes un drole de lascar avec les femmes

(description-evaluation);

Actes vs operations -

335

l'enonce 22 vous le racontez lä aussi (rappel) l'enonce 58 ma premiere femme

m'avait dit qu'elle se suiciderait

sije

la

trompais

(recit); etc. 2.

Les operations

autoreferentielles

preferent les predications plus faibles, de forme

moins descriptive ou non descriptive, et faisant rarement "figure" d'actes, comme par exemple: -

13 excusez-moi

-

12 vous etes, 14 vous etes un, 16 un (tentatives de formulation) = non-actes

I 'expression (excuse) = acte

II est interessant de voir que certains "connecteurs" ou "marqueurs",

certaines

"particules", etc. peuvent s'utiliser avec ou sans frontteres prosodiques, ce qui veut dire qu'ils sont parfois susceptibles de se präsenter comme quasi-actes car plus fortement marques. Ainsi, pour enfin, vraiment et autres "mots du discours", ils se presentent comme suit: -

26 enfin (ci-dessous, avec frontiere d'acte) vs 11 oui mais enfin, (sans frontiere d'acte)

-

27 vraiment (ci-dessous, sans frontiere d'acte) vs 18 vraiment (avec frontiere d'acte), 1 bon, 6 tout ςα, 20 tout de me me, 23 moi, 25 voilä (tous marqueurs-connecteurs avec frontiöre d'acte).

Ces expressions qui preferent les formes implicites, non descriptives, ont toutes ce qu'on appelle des sens instructionnels. (26-27) enfin^/c 'est une union vraiment comme je I'imagine GS 26 27 Id Md Md

c'est une.union

comme je Γ imagine

enfin

Ip s

c'est une.union

D

c'est une.union

vraiment

comme je Γ imagine comme je Γ imagine

PP Pd Pro

enfin /

\

336

Liana Pop

Remarquons egalement que les actes (tel 27 c'est une union comme je I'imagine - evaluation), les connecteurs (tel 26 enfin- reformulatif) ou les modalisateurs (tel vraiment en 27 - operation subjective) dans Pexemple ci-dessus, sont des operations simples (un seul espace occupe) ou des configurations stables d'operations (plusieurs espaces occupes). Notre annexe est clair.

5. Conclusions

1. Dire que "tout est operation", reviendrait ä homogeneiser la perspective sur le discours en general et sur le dialogue en particulier. Les operations peuvent se manifester par des configurations d'operations. Les categories "difficiles" (marqueurs, connecteurs, ponctuants, regulateurs, rates, particules) ainsi que les actes sont des (configurations d') operations. 2. Sur cet ensemble d'operations tres heterogenes impliquees dans la production du discours: les actes de langage sont des categories emergentes d'operations / de configurations d'operations (des "figures" fortes au sens cognitif du terme) = FI+CP. lis sont tous naturellement categorises. les actes discursifs sont des categories emergentes d'operations / de configurations d'operations =fi+CP. lis sont en genöral naturellement categorises. les non-actes (semi-actes, mots du discours, marqueurs, connecteurs, regulateurs, rat6s, ponctuants, etc.) coincident plutot aux "traces d'operations" et se constituent rarement en "figures" (sont des operations moins emergentes). Leurs noms represented des categories moins naturelles de "faires". 3. Plusieurs categories (actes, regulateurs, marqueurs, etc.) se rencontrent en un seul et meme espace: sur I'espace interpersonnel, appels, excuses, reprises diaphoniques, regulateurs, RAD, etc.; sur I'espace subjectif, modalisants, certains phatiques, certains actes rituels, reprises monophoniques, etc.;

Actes vs operations

-

337

sur l'espaceparadiscursif,

tous les actes ou marqueurs d'actes de mise en forme / for-

mulation (rates, hdsitations, faux-diparts), certains actes structurants (cadrages, th6matisations, marqueurs d'integration lineaire MIL, etc.); -

sur l'espace

metadiscursif,

actes de reformulation (reprises, commentaires meta,

metaphores), ponctuants, marqueurs de structuration de la conversation (MSC), MIL, etc.; -

sur l'espacepresuppositionnel,

les retours, les rappels, les explications d'arriere-fond,

les thematisations et les cadrages, des marqueurs du type vous savez, nous avons

vu

etc. Afin que des descriptions plus appropriees puissent etre donnees du discours, il semblerait done que la categorie d'operation, catigorie non naturelle, de linguiste, s'impose actuellement dans l'analyse du discours.

References bibliographiques

Blanche-Benveniste, Cl. (1989): Constructions verbales "en incise" et rection faible des verbes. Recherches sur le franfais parle 9, 53-73. Culioli, A. (1990): Pour une linguistique de l'enonciation: operations et reprisentations, Tome 1. Paris: Ophrys. Etudes de linguistique appliquee no 44, Les conversations authentiques. Ferrari, Α., A. Auchlin (1995): Le point: un signe de ponctualisation. - CLF 17, 35-56. Furukawa, N. (1996): Grammaire de la predication seconde. - Duculot. Pop, L. (2000): Espaces discursifs. - Peeters: Paris-Louvain. Recanati, F. (1984): Remarques sur les verbes parenthetiques. - Linguisticae Investigationes Supplementa (LIS) vol 8. De la syntaxe ä la pragmatique, 319-351. Rossari, C. (1996): Identification d'unites discursives: les actes et les connecteurs. - CLF 18, 157177. Roulet, E. et al. (1985): L'articulation du discours en franpais contemporain. - Berne: P. Lang. Roulet, E. (1994): La phrase: unite de langue ou unite de discours? M61anges de philologie et de littörature medievales offerts έ Michel Burger, Publications romanes et franfaises, CCVIII. Droz. Rubattel, C. (1986): La structure de l'enonce minimal comme condition d'accäs aux strategies interpr6tatives. - CLF 5, 135-148.

Anne Salazar Orvig

Elements pour une analyse de la connivence dans le dialogue

II arrive que Ton dise έ propos d'un dialogue ou de la relation qu'entretiennent les interlocuteurs qu'ils manifestent une grande connivence. Or, ce phenomene, somme toute assez ordinaire, ne semble pas avoir ete traite de fa?on directe dans les recherches sur le dialogue 1 (par exemple le terme est absent des index des principaux livres sur l'interaction). Pourtant, s'interroger sur un phenomene comme celui de la connivence revient a poser la question de la relation entre l'apprdhension de la qualite ou de la tonalitd globale d'un dialogue et l'objectivation de cette interprdtation a travers un certain nombre de faits langagiers. Ce qui permet done en retour de saisir, dans leur dynamique meme, les mouvements constitutifs du dialogue. Mais avant d'entrer en matiere essayons tout d'abord de cerner ce qu'on entend par

connivence.

1. Autour de la connivence

II semble difficile de s'accorder sur le sens meme du terme connivence. Le recours au dictionnaire nous aide partiellement. Le Petit Robert propose les acceptions suivantes: 1. Complicite qui consiste ä cacher la faute de quelqu'un 2. Accord tacite => entente, intelligence. Agir, etre de connivence avec quelqu 'un cf. etre de meche, s 'entendre comme larrons en foire Le Dictionnaire Historique de la Langue Fran?aise (Robert) est plus precis: Connivence N.F. est emprunte (1539) au bas latin coniventia "indulgence", form6 sur le participe prtsent de connivere, d'abord "serrer les paupieres", "fermer les yeux" d'ou au figure "etre indulgent", et surtout sous PEmpire "etre d'accord". [...] Connivence a longtemps exprime [...] I'id6e d'"indulgence coupable". II s'est Oriente vers le sens actuel d"'intelligence secrete, accord tacite" (1796) et a en partie perdu sa valeur p6jorative. Tout comme complicite, il peut aujourd'hui indiCette ^flexion a pour nous deux sources: une communication faite avec C. Hudelot ä propos d'un dialogue adulte-enfants (Salazar Orvig et Hudelot 1997) et le travail de reflexion et analyse mend avec des dtudiants de troisidme cycle, et plus particulierement A. Lambert qui travaille sur ces th£mes έ propos de conversations de cafe. Cf. igalement Andri-Larochebouvy (1984).

340

Anne Salazar Orvig

quer une qualite psychologique d'"entente spontande" (etre de connivence, sourire de connivence) Cet article met en avant le deplacement de sens que Ton peut observer tres nettement aujourd'hui: la dimension de l'entente predomine. Un sondage auprfes d'6tudiants a montre que le sens de connivence tel qu'il est d6pos6 chez eux est en leger decalage avec les traits releves dans les dictionnaires. En effet, l'idee d'indulgence coupable n'apparatt jamais; en revanche ils evoquent preferentiellement l'idee de communautö etroite, d'implicite commun, de codes prives, pouvant aboutir ä l'exclusion des tiers. Par ailleurs, la connivence est associee ä une certaine empathie entre les interlocuteurs et ä une jubilation dans l'entente. Cette diversite se retrouve chez les linguistes: certains associent la connivence ä un certain type de savoir partage (Morel et Danon-Boileau ä paraitre), d'autres a une certaine complicite (Andre-Larochebouvy 1984). En ce qui nous concerne (Salazar Orvig et Hudelot 1997), il nous a d'abord semble necessaire de distinguer la "connivence" de P"evidence". L'evidence correspond a ce qui est donne pour acquis, et qui ne necessite pas d'actualisation - sauf bien entendu s'il y a mdprise a son propos. Elle releve ainsi de ce qui identifie les interlocuteurs les uns aux autres, en tant qu'ils appartiennent ä une meme communaute. En revanche la connivence est recherchee par les interlocuteurs et doit etre confirmee. La connivence qui est construite dans le discours presuppose ainsi la difference des interlocuteurs. Quand on parle de connivence dans le dialogue on ne fait pas reference ä un principe organisateur au meme titre, par exemple, que le principe de cooperation ou l'ajustement des perspectives. La connivence ne constitue pas, non plus, un genre ou un type d'interaction. Bien qu'elle se realise de fafon prdferentielle lors d'echanges de type conversationnels, la connivence peut surgir ä tout moment, pour peu qu'une certaine tonalite soit adoptee et que la relation dialogique devienne, ne serait-ce que tres provisoirement, symetrique. Quand on parle de connivence dans un dialogue on fait plutöt röference ä un evenement, dans le sens de Francois (1990), ä la cristallisation d'un ensemble de facteurs en un moment privilegie. Ce moment se caracterise par une forte entente des interlocuteurs, entente dont on dira qu'elle est ä la fois tacite et jubilatoire. On aurait done lä affaire a une forme de ce que Auchlin (1995, 1996) appelle le "bonheur conversationnel".

Elements pour une analyse de la connivence dans le dialogue

341

2. Apprehender la connivence

On peut cependant se demander comment se manifestent et se combinent ces differentes facettes dans la dynamique d'un dialogue particulier. La connivence prösente-t-elle des marqueurs specifiques? ou releve-t-elle plutöt de ce que Francis (1993) appelle signification atmospherique "non localisable et non analysable" "plus ressentie que signification exploitable", resultat d'une configuration particuliere du discours? Certainement, et en accord avec Fran?ois, je dirais que la caracterisation du dialogue comme conniventiel releve en premiere instance de la fafon dont nous le recevons, de l'experience que nous en faisons (Auchlin 1995, Auchlin 1996) en tant qu'observateurs interpretes. Cependant, ces significations qui resonnent en nous sont le produit des mises en mots, des enchainements, des mouvements discursifs effectues par les locuteurs. C'est done ä travers la caracterisation de ces mouvements que Ton peut tenter d'objectiver notre impression. Afin de d6gager quelques pistes d'analyse, j'ai choisi de travailler sur une conversation entre deux jeunes filles, Denise et Jocelyne. A partir d'une premiere evaluation globale de cet echange, il s'agit de d6gager les differents ölements susceptibles de manifester Tun ou l'autre des traits definitoires de la connivence. On abordera ici essentiellement la dimension verbale de 1'echange. Un travail ultörieur devrait traiter du röle des mimiques (tels que les sourires), des regards (par exemple un regard rieur ou petillant), de Γ intonation2 ou de la qualite de la voix3. Les indices verbaux ne se distribuent pas de fa9on symetrique parmi les interlocuteurs. On peut distinguer des indices specifiques au locuteur et d'autres qui relevent des conduites d'ecoute ou d'interpretation, les indices done du recepteur, ou de l'ecouteur (Morel et Danon-Boileau a paraltre). Bien entendu, cette opposition est labile, non seulement en raison de l'interversion des röles dans une conversation mais aussi parce que dans certains cas cette dissymetrie s'estompe et chacun est ä la fois locuteur et ecouteur.

2.1. Les indices du locuteur Le locuteur Oriente pour beaucoup la tonalit£ dans laquelle se developpe Γ echange, il installe une certaine atmosphere et definit la qualite de la relation. Ceci se realise par au

Par exemple, Morel et Danon-Boileau (έ paraltre) ont montrö que la m61odie de {'intervention de l'ecouteur indique comment il se place dans la relation co-6nonciative. Cf. par exemple l'empathie vocale, citee par Kerbrat-Orecchioni (1990).

342

Anne Salazar Orvig

moins trois voies: le choix du theme et du genre, la gestion des savoirs partages et des appels ä la connivence.

2.1.1. Theme, genre et tonalite Le theme definit le terrain dans lequel l'echange va se derouler. Si "les themes abordes peuvent etre mis en correlation avec le type de relation entre les interlocuteurs"(KerbratOrecchioni 1992: 55), ils sont, corollairement, susceptibles de determiner la nature de la relation dialogique. On peut noter ainsi qu'une atmosphere de connivence est susceptible de s'installer quand on se situe dans le domaine de l'intime. Dans le corpus etudie, le fait que les deux jeunes filles parlent de leurs jambes probablement mobilise et accrolt leur proximite. Mais il suffit d'une experience partagee, pour reduite qu'elle soit, ä laquelle on peut faire allusion. Par exemple, le fait d'assister a une algarade peut creer un ilot de connivence entre les temoins. D'une fa9on generale, parier d'autrui permet aux interlocuteurs de se rencontrer, voire de se reconnaitre, aux depends d'un tiers, comme on le voit ci-dessous: Exemple 1 30 - DEN - {...} et heu ++ Anne Marie y est allee de temps en temps au debut et+ eile est tombee chez une chez une estheticienne vraiment pas douee qui lui a dit "oh la la! mais vous avez des jambes de quarante ans" ä peu pres 3 1 - J o e - oh::::! 32- DEN - T'imagines! 33 - Joe - bonjour le tact

En meme temps, ces themes ne sauraient etre dissocies du genre mobilise. Ainsi le genre "medisance et cancans" pourrait etre un bon declencheur de connivence. II en est de meme de la tonalite (Hymes 1974): on peut imaginer de la connivence autour d'un fait dramatique pour peu que celui-ci soit traite avec humour.

2.1.2. Parier ä demi-mots La connivence est souvent identifiee ä l'existence d'un savoir partage important, et plus encore au fait que les interlocuteurs n'ont pas besoin de s'expliciter mutuellement un certain nombre de choses. C'est le cas, bien entendu, de l'utilisation des noms propres ou autres expressions definies. II en est de meme pour d'autres formes de presupposition linguistique:

Elements pour une analyse de la connivence dans le dialogue

343

Exemple 2 8 - DEN - {...} J'repete ce qu'i(l) - heu +++ c(e) que::: le phebologue qu'Anne Marie est allde voir heu+++ avait dit+++ et puis comme on doit pas etre tris loin des memes problemes

Ici l'expression les memes problemes ne sera jamais explicitöe, eile repose de toute evidence sur la presomption de connaissances partagees. On constate igalement des enonces allusifs: Exemple 3 23 - Joe - d'autant plus que moi j'ai toujours eu ces jambes Ιέ et que 9a n'a absolument pas empire depuis deux ans - d(e)puis un an -d(e)puis::: enfin bon

rires 24 - DEN - {inaudible}

rires 25 - Joe -

e'est 9a. quelques mois

La röference temporelle (depuis deux ans - depuis un an) introduite par Jocelyne suffit ä evoquer un evenement particulier, evenement que Denise semble reconstituer parfaitement, puisqu'elle amene Jocelyne a corriger sa datation. On a lä l'une des sources probables de l'impression d'entente tacite dans ce type de dialogue.4 2.1.3. Lesappels Complimentairement ä ces deux premieres dimensions, la connivence resulte egalement de mouvements visant ä declencher chez l'autre un certain type d'adhesion. Dans le corpus etudie se degagent cinq grands types d'appsls: l'enrölement, les indications d'implicite, les evaluations, les plaisanteries et le rire. 1.

l'enr0lement (Salazar Orvig et Hudelot 1997) correspond aux diverses expressions qui visent a impliquer l'autre dans le discours5. C'est le cas du t'imagines (DEN, 32, ex.1) ou les tu vois? de l'exemple 4:

On peut aussi relever des enoncis inachevds qui sugg&rent k l'autre la suite du raisonnement: 64 Joe - d(e) fa?ons eile voit euh plutot des bonnes femmes §enfin§ enfin des bonnes femmes j(e) sais pas έ partir de trente cinq ans, franchement, euh les bonnes femmes elles sont bien habillees mais euh +++§ D'autres auteurs parlent de "des signaux d'appel au consensus"(Andr6-Larochebouvy 1984) ou de "ligateurs" ou "ponctuants d'appel ä l'attention de l'autre" (Morel et Danon-Boileau 1998).

344

Anne Salazar Orvig

Exemple 4 68 - J o e - moi j(e) veux dire c(e) qui m(e) complexe le plus c'est quand vraiment j'suis poilue! j(e) t'assure, alors qu'elle est Ii pour 9a. eile est la pour 9a mais ä chaque fois je m'excuse, tu (v)ois, j(e) (l)ui dis "oui euh fait euh +++ un peu plus longtemps qu(e) prevu qu(e) j(e) suis pas venue" tu vois euh::: 69 - DEN - ouais! ah c'est §rigolo!

2.

Les enonces allusifs, ou elliptiques, peuvent etre accompagnes de particules et connecteurs qui constituent autant d'indices fournis ä l'interlocuteur sur le fait qu'il y a de l'implicite a reconstituer. C'est le cas de enfin bon6 (Joe, 23, ex 3) qui decrit un double mouvement d'ouverture de reformulation et de cloture et qui pourrait etre paraphrase ici comme tu vois ce que je veux dire. 3. Plus qu'une simple assertion, les evaluations constituent des elements auxquels les interlocuteurs sont particulierement attentifs et ä propos desquels les locuteurs attendent une reaction, le plus souvent une evaluation en retour (Goodwin et Goodwin 1992). C'est le cas dans notre corpus, des evaluations d'une tierce personne, comme dans l'exemple 1, ou des auto-evaluations. Ces dernteres sont le plus souvent negatives: Exemple 5 26 - DEN - non mais remarque, moi un true qu'j'ai +++ enfin j'crois qu(e) j'irai jamais chez une +-H- enfin ä moins d(e) changer tout ä fait heu +++ d'aspect chez une heu +++ eh ben voilä une estheticienne 27 - Joe comment $a? 28 - DEN - ben oui pa(r)ce que:::§ 29 - Joe - j 'y vais tous les mois! 30 - DEN - j ' y vais jamais! j(e) t'ai dit j(e) m'epilais moi-meme les jambes, done j'en ai jamais besoin +++ mais alors regarde pas pa(r)ce que lä elles sont ratdes lä +++ enfin, elles sont ratees, 20 times), but may be considered a more general response mechanism. In the Gorgias it was used in assents to statements (4x) and in consents to directives (3x). 9. The final specific 'yes'-form to which I devote some attention is φημι, derived from the lsg. pres. verb Ί say', Ί affirm', which comes to mean Ί say yes' or simply 'yes'. An example of this use of the bare form φημι from the Gorgias is given below. ούκοΰν τά τούτων νόμιμα κατά φΰσιν καλά, κρειττόνων γε όντων;:: φημι.:: SOCRATES: Then [are (copula implicit)] their ordinances naturally noble, since they are those of the more powerful? :: CALLICLES: Yes. (Plato Gorgias 488e6)

Φημι is not very frequent when compared with ναί, with adverbs in general, or with the elliptical repetitions. Yet as a specific fixed form, φημι recurs significantly. Of the 15 instances in which φημι forms a one-word response to a question, 13 do not correspond with a second-person verb in the question itself. Just as adverbs such as ναί or π ά ν υ γε affirm or confirm the truth of the sentence-question, φ η μ ι does so in a fixed, rather than a grammatical manner. But alongside this 'fixed' use of the form φημι, I found three instances in which φημι repeats the verb in the question (in 2s). The example below illustrates this: here φημι is part of a short, but full sentence, Ί say this'. 5 492d5: ...καί μοι λέγε· τάς μεν επιθυμίας φήις ού κολαστέον, εί μέλλει τις οιον δε"ϊ ενναι, έώντα δε αύτάς ώς μεγίστας πλήρωσιν αύταις άμόθεν γέ πόθεν έτοιμάζενν, και τοΰτο είναι την άρετήν;:: φημι ταϋτα εγώ. ::

In another instance, at 500el-2, the form in the question is a compound (σΰμφηις;), in the answer, a simplex (φημι). In this instance, and in the third instance, φημι answers a combination of statement + tag question, reminiscent of the one we saw on p. 354 above.

Yes (and No) in Ancient Literary Greek

357

SOCRATES: ...Tell me: You say not to curb appetites, if one is to be what he should be, but rather should allow them the fullest possible growth and procure satisfaction for them from whatever source; and this [you say] is virtue? :: CALLICLES: I say this. The point here is that φημι can sometimes be used with its referential meaning to the fore, and as a verb realizing its categories6 and complementation (or having the potential to do so). More often it is felt that such category and grammaticalization potential are weakened and the use of this verb form approximates the adverbial 'yes' forms: without the corresponding element in the question, it would seem incomplete if taken literally or referentially. This is an interesting mechanism, but one which is ultimately short-lived. I note that all other one-word responses deriving from verbs and not involving a repetition of the same verb from the initiating move, are impersonal forms. 7 10. Although implicit in the classification and in the examples of responses they give, grammarians such as Kühner-Gerth and all who follow them do not speak in terms of brevity or incompleteness. As we have seen, however, these two are salient features of responses, even in Socratic dialogue — and this tendency is manifest in dramatic dialogue, especially in Comedy. An exception is the treatment in the grammar of Schwyzer and Debrunner, although even they do not go all the way. II.2 p.628: Eine Antwort erfolgte zunächst durch einen ganzen Satz; diese Weise hat sich weitgehend erhalten bei Fragen ohne Interrogativ (auch das Griechische kennt wie andere altindogermanische Sprachen Wörter für 'ja' und 'nein' erst in Anfängen). ...Doch kann auch auf Fragen ohne Interrogativ gegebenfalls nur z.B. mit έγώ, ουκ έγώ oder mit εύ γε, κάρτα, μάλα, πάνυ, ήκιστα, ούτως, καλώς γε geantwortet werden; jünger nähem sich ναι und ου der Bedeutung von nhd. 'ja', 'nein'. Auf den Aufbau von Dialogien (Stichomythien), Rede und Gegenrede kann hier nicht eingegangen werden. p. 631 ε) Antwort: ...1. Durch einen Satz. 2. Abgekürzte Antworten 3. Wiederaufnahme der Frage seitens des Gefragten durch eine indirekte Frage. From the perspective of grammarians such as Schwyzer and Debrunner, when there is no recourse to yes/no words equivalent to their own 'ja/nein', one option is to answer with a E.g. with change of tense, at 467b3-5 (imperfect:: present). E.g. 506e2 with change of verbal mood to optative. It is perhaps worth quoting the passage more fully (from 506d5ff): 'Αλλά μεν δή ή γε αρετή εκάστου και σκεύους και σώματος και ψυχής αύ και ζώιου παντός, ού τώι ε'ικήι κάλλιστα παραγίγνεται άλλα τάξει και όρθότητι και τέχνηι, ήτις εκάστωι άποδέδοται αυτών· άρα 'έστιν ταΰτα; :: Έγώ μεν γάρ φημι. :: Τάξει άρα και κεκοσμημένον έστιν ή αρετή εκάστου; :: φαίην άν εγωγε. (...But surely the goodness of anything, whether implement or body or soul or any living thing, does not best come to it merely by haphazard, but through a certain lightness and order and through the art that is assigned to each of them. Is this so? :: I certainly agree.:: Then the goodness of anything is due to order and arrangement? :: I should agree.) All very rarely used: έστω, δοκεϊ, εστι, έοικεν, δήλον, εικός, κινδυνεύει.

358

Donna Shalev

(full) sentence. Otherwise with what they call 'Abgekürzte Antworten'. This does not fully take into account the distinction between a short but syntactically 'complete' response on the one hand, and a short but syntactically 'incomplete' response on the other. It is not within the scope of this paper to reconcile this difficulty. I can only observe that what may be considered 'incomplete' in a declarative statement may not necessarily be considered 'incomplete' in a reactive turn, because of its tight cohesion with the initiating turn to which it corresponds. It may be better not to expect the same level of autonomy and 'completeness' in responses as one expects in initiating moves. Evidence from Greek dialogue texts - Plato and Drama equally - points to a rich range of cohesion devices which glue together initiating and reactive turns of all persuasions, including questions and answers. In addition to devices such as those we find in English, such as ellipse, there is repetition (less in the form of substitution favored in English); and there are particles such as γε, devices we have seen in our sample. The rich cohesion and low level of autonomy of responses seem more natural to the dialogue analyst than they do to the classical grammarian, because the dialogue analyst does not perhaps only analyze in terms of sentences. Rather, in terms of an exchange, the initiating and the reactive move are components of a shared unit. They may share the grammar between them, and some scholars suggest that the answer should be treated as a distinct speech act. Joseph and Philippaki-Warburton (1987: 13f) do just this in a descriptive analysis of answers in Modern Greek. 11. If I base my assessment of the response mechanism in Classical Greek exclusively on a depouillement of a single Platonic dialogue, I run into trouble. All it can provide is a point of departure. The following are some of the more blatant distortions created by basing a comparative point of reference on such a narrow and idiosyncratic corpus: the nature of exchange in many Socratic dialogues is one of extreme asymmetry between the contributions of Socrates and those of his interlocutors. The questions are often pseudo-questions formed of a long statement with a question formula such as ή γάρ; tagged onto the end, or are rhetorical questions. In these circumstances, the answers are often only token answers; -

aside from this asymmetry and token interaction, most exchanges are of the artificial binary type; rare are the side-sequences, evaluative turns, and avoidances which figure so freely in natural conversation.8

Side sequences, for example, are found only twice among the 240 question-answer and 72 other initiating-response exchanges in this Platonic dialogue: at 462b5 and 497e7.

Yes (and No) in Ancient Literary Greek

359

But these and other features vary in the different dialogues. Plato's output is heterogeneous. To begin with, his dialogues fall into three main types: direct, indirect, and mixed. Even if we limit ourselves to the direct dialogues, the dynamics between the interlocutors range from monologues in disguise to dialogues with very high levels of real interaction and assertiveness of respondents. Extremists would even admit that each dialogue may be seen as an individual experiment in a distinct balance of dialogal features. 12. Drama provides a different picture. On the one hand it is more homogeneous in the sense that direct speech is the mode of the genre. However, the conventions of Classical Drama - e.g. the verse form, or the reluctance to distribute verses among actors - create a text in which very short turns at talk are rare (especially in Tragedy). As a result, one-word responses such as ναί and φημι will not be frequent. In this sense, in Drama too, generic conventions, albeit of a different sort from those of Socratic dialogue, create an artificiality similarly distant from natural exchange. Simple ναί and φημι in answers to yes/no questions in the comic authors Aristophanes and Menander are significantly more frequent9 than in Tragic authors.10 13.1 am not proficient in post-Classical Greek, but for the sake of a fuller picture, I refer the reader to Jannaris (1987: §§ 2057ff), and to Joseph and Philippaki-Warburton (1987: 13ff) for definitive information on these languages. It is interesting that Modern Greek has preserved the device typifying response in Greek, repetition. The pronoun device died out and the use of ναί and οϋ, which gradually gain ground in Classical Greek, develop into the prevalent form in Modern Greek, where the words ναί and όχι are different from English yes and no, but they operate in the same way and their mechanism is the same. 14. To sum up, Ancient Greek does have a form ναί, which is comparable in nature and use to equivalents such as English 'yes', and this form becomes even more closely comparable in Modern Greek. Yet a close study of Plato's Gorgias suggests that in Classical Literary Greek the expressions for responding to a question are varied in form and construction. Four mechanisms discussed more fully were (1) the adverb ναί; (2) the use of the bare pronoun; (3) repetition of the verb (or some other form) in the question; and (4) the fixed form φημι. The form ναί was not the overall leading device used for positive answers to yes-no questions. Moreover, ναί was not exclusively used to answer questions: although it never consented to commands, it was significantly used in assents to statements. The nature of

Φημι 16 times, ναί 13 times. Φημι 3 times (none in Aeschylus), ναί 16 times (+ 5 in longer moves), 11 of Euripides, whose corpus is much larger than the others.

them in

360

Donna Shalev

the questions answered by ναί, it may be added, was frequently closer to that of a statement. The device φημι had similar behaviour patterns; although primarily used in responses to questions, it was also used in assents to statements, and like ναί, never in consent to commands. The most versatile of the four devices closely studied was the repetition of the element from the question: This is a more general response mechanism, found also in assents to statements and consents to commands. At the other end of the spectrum, the device used most exclusively in response to questions was the bare pronoun (only once in assent to a statement and never in consent to a command). I have tried to illustrate how the analysis of a literary text in a dead language, through the use of tools from a budding discipline, can inform us of the dialogal elements of this text in a way that is forfeited when resorting to tools such as classic grammatical analysis. The findings on the one hand show how a literary text works differently from and similarly to natural speech, but they also throw light on the nature of Socratic dialogue and lack of dialogue, and show as well how Greek works differently from English. This very difference may suggest some of the recurring features of responses which are more firmly anchored in the general nature of communication than in the nature of any specific language.

References

Beversluis, J. (2000): Cross-examining Socrates: a defense of the interlocutors in Plato's early dialogues. - Cambridge: Cambridge University Press. Halliday, Μ. A. K., R. Hasan (1976): Cohesion in English. - London: Longman. Jannaris, A. N. (1897): An Historical Greek Grammar, chiefly of the Attic dialect as written and spoken from classical antiquity down to the present time. - London: MacMillan. Joseph, B. D., I. Philippaki-Warburton (1987): Modern Greek. - London: Croom Helm. Kühner, R., B. Gerth (1904): Ausfuhrliche Grammatik der Griechischen Sprache. II: Satzlehre, 2. Band. - Hannover, Leipzig. Schwyzer, Ε., A. Debrunner (1966): Griechische Grammatik. - München: Beck. TheslefF, H. (1961): Yes and No in Plautus and Terence. [Commentationes Humanarum Litterarum, 26.3]. - Helsinki: Societas Scientiarum Fennica.

Milena Srpovä Dialogue entre cultures differentes: le role de Γ ethos collectif

1. Hypotheses, croyances et definitions concernant la communication

Nous communiquons au moyen de materiaux semiotiques de nature verbale (V), paraverbale (PV) et non verbale (NV). Les trois types de materiau fonctionnent de la meme maniere, ä savoir: a) ils sont faits de signes ä deux faces (face: forme et face: contenu); b) les signes prennent un sens actualise dans une situation de communication concrete; c) sur un fond collectif fortement partage par une communaute se greffent des codes localement partages, puis des variations individuelles. Les codes semiotiques, des plus formels aux plus informels, observes ä un moment donne, varient d'une communaut6 langagiere ä l'autre (contrastivite externe), mais aussi ä l'interieur d'une "meme" communaute (contrastivite interne).

2. Ethos (styles) collectifs

De la vaste litterature existant au sujet des fonds communs culturellement determines, que Ton nomme, depuis peu, "ethos (styles) collectifs" ou "profils communicatifs", ou encore "ethnolectes", on pourrait degager des typologies en utilisant des criteres de classement de nature notamment paraverbale et non verbale - voir la colonne gauche du tableau suivant (tableau elabord par M. Srpova (1999) ä partir des informations transmises par C. KerbratOrecchioni, (1994)).

362

Milena Srpovä CRITERES (OU AXES) DE CLASSEMENT

TYPE D'ETHOS

COLLECTIF (ETHNOLECTE)

ethos fortement "communicatif' (verbiositö forte) ex. France, USA, Italie, Espagne,...

ethos faiblement "communicatif' (verbiosite faible) ex. Laponie, Finlande, Japon, ...

1. LA "PAROLE"

a) Γ importance quantitative de la parole (la "verbiosite")

b) aspects "qualitatifs" de la parole b l ) la "beaute"

b2) la "verite" vs Γ "efficacite"

b3) la fonction privilegiee

b4) degre de dependance du contexte

ethos a parole "belle" (le "beau" parier) ex. societes africaines ethos ä parole "vraie" ex. Indiens, Allemands, Tcheques

ethos ä parole "efficace" ex. l'oralite africaine traditionnelle

la parole "pleine" (referentielle)

la parole "phatique" (sollicitation et maintien du contact)

la parole "directe" (explicite, relativement independante du contexte) ex. discours scientifique en Europe ou aux USA

la parole "indirecte" (implicite, solidaire du contexte) ex. conversation en famille en Europe

363

Dialogue entre cultures differentes: le role de I 'ethos collectif

1 A RELATION INTERPERSONNELLE

a) horizontale (distance)

ithos de proximite (societes "ä contact") ex. Grecs, Slaves

ethos de distance (societes "sans contact") ex.Anglais, Americains, Scandinaves

b)verticale (pouvoir)

6thos hidrarchique ex. Asie, Indon6sie, Afrique

ethos egalitaire ex. soci6t£s occidentales

c) par rapport au "conflit"

6thos consensuel (recherche permanente du consensus, evitement du conflit) ex. Japon, Chine, Madagascar, Bali, pays nordiques et anglo-saxons

ethos conflictuel (grande tolerance au conflit) ex. pays mediterraneens (Italiens ou Fran^ais en contraste avec les Americains)

ethos a politesse "negative" (par evitement, "ne pas deranger") ex. Finlande, Japon en situation formelle

ethos ä politesse "positive" (par "intervention": questions, cadeaux, compliments,...) ex. Venezuela, Grece, Japon en situation informelle

ethos ä face nigative (le "territoire") ex. Occident

6thos ä face positive (la "dignite", Γ "honneur") ex. Orient, lies Pacifiques, Afrique Noire

3. A CONCEPTION DE LA POLITESSE

a) type de politesse

b) "face" concernee

364

c) interets des deux partenaires de Pinteraction

Milena Srpovä

L2 est "valorise" par LI, qui se "devalorise" ex. Japon, Coree, Thailande, Madagascar

les interets des deux partenaires sont relativement equilibres ex. Occident "germanique", France

ethos 6motionnel ex. Portugal, Espagne, Italie, pays slaves

censure des emotions ex. Bali, Java, Hawaii, Malaisie, Rwanda

ethos a haut degre de ritualisation (regies fortement contraignantes) ex. Asie

ethos a faible degre de ritualisation (une marge importante est laissee ä l'improvisation individuelle et a la negotiation; codes flous) ex. Occident

ethos individualiste (l'individu a sa personnalite propre; la personne est un individu) ex. Europe

ethos solidariste (l'individu n'est pas dissociable du groupe; la personne est une / la collectivite) ex. societes amerindiennes, Afrique Noire, Extreme-Orient, le Pacifique

4. I/EXPRESSION DES "SENTIMENTS" (AFFECTS)

5. LE DEGRE DE RITUALISATION

6. LA CONCEPTION DE L'INDIVIDU

Dialogue entre cultures differentes: le role de I 'ethos collectif

365

Le style communicatif d'une societe serait forme d'un ensemble de traits appartenant, k des degres fort varies, aux differents cas de 1. ä 6. Cet ensemble connait des variations intracommunautaires (et, bien entendu, individuelles).

3. Problemes et problematiques

3.1. Les codes non verbaux ne sont pas systematiquement decrits, ni vraiment enseignds. La consequence en est que nous arrivons a prendre plus facilement du recul par rapport au verbal, nous pouvons le soumettre ä l'analyse, alors que le non verbal git dans notre inconscient et se montre rebelle a la prise de conscience, et plus encore ä l'analyse. A tel point que meme les specialistes du domaine depuis longue date, comme Hymes ou Gumperz, ont fini par baisser les bras et par dire que la competence communicative ne peut finalement s'acquerir que sur le terrain. 3.2. Les codes non verbaux relevent ä la fois des savoirs et des savoir-faire referentiels et des jugements de valeur. L'interpretant attribue aux formes V, PV et NV l'expressivite qu'elles ont dans le code culturel de sa communaute, mais pas necessairement dans celui de la communautd de son interlocuteur. Un locuteur appartenant ä une communaute ä verbiosite forte va percevoir un locuteur appartenant ä une communaute ä verbiosite faible comme "reserve", "ferm6", "froid", "tacitume", etc. Un locuteur appartenant ä une communaute ä verbiosite faible va percevoir un locuteur appartenant ä une communaute ä verbiosite forte comme quelqu'un qui parle "trop", etc. 3.3. La non-identite des codes est ä l'origine des malentendus "interculturels". N'etant pas traites "scientifiquement", de fafon desinteressee, et ä froid, ceux-ci se repercutent sur la relation entre personnes concernees, de sorte que la relation finit par l'emporter. Le probl6me n'est pas seulement academique. Chaque fois que 1'evaluation est negative, eile met en jeu "le territoire" ou "l'amour propre" (Brown et Levinson (1987), voir C. KerbratOrecchioni (1994)) des interesses, qu'ils se trouvent chez eux ou ä l'etranger. S'ils n'etaient pas entraines par un mouvement de type "reflexe conditionne", dans un affect negatif, ils pourraient peut-etre, comme il est dit dans des approches optimistes, negocier. Mais meme quand ils se mettent ä negocier, ils negocient en appliquant des codes

366

Milena Srpovä

culturellement determines. Le probleme reste done le meme concernant la comprehension et l'expression.

4. Une approche empirique et une description analytique selon les deux perspectives de la communication

Lorsque deux cultures sont en contact, l'experience empirique montre assez vite que les problfemes sont surtout de deux ordres. Ces problemes se d6gagent dans les deux perspectives de la communication, qui, elles, sont li6es ä la double face des signes: la perspective de la comprehension des formes et la perspective de l'expression des contenus. Concernant la perspective de la comprehension, on est deconcerte quand des formes identiques, ou perijues corame telles, existent dans les deux cultures en contact, mais que, dans une situation comparable, elles expriment un sens different et / ou ont une expressivite differente. C'est ce que j'ai propose d'appeler homonymie interculturelle. Prenons l'exemple d'un objet faisant partie d'une situation ritualisee (Srpovä 1994: Rodriguez, 1993, p. 6-7, "Jus d'orange pressee"). Situation: On est dans un cafe ou dans un bar et on demande au serveur un jus d'orange pressee. Forme: On nous apporte un verre de jus ä moitie rempli, accompagne d'une carafe d'eau et de sucre. Sens et Expressivite: En France: usage courant. En Espagne: ordre marque des choses. Le client pensera qu 'il s 'agitd'un malentendu et demandera qu 'on remplisse son verre convenablement. Concernant la perspective de / 'expression, oü, dans une situation comparable, il s'agit d'exprimer tel sens, avec telle expressivite, le probleme est pose par la non identite des formes culturellement localisees. J'ai propose d'appeler ce cas synonymie interculturelle. Prenons l'exemple d'un objet faisant partie d'une situation ritualisee (Srpovä 1994: Rodriguez, 1993, p. 6-7, "Jus d'orange pressee"). Situation: On est dans un cafe ou dans un bar et on demande au serveur un jus d'orange pressee. Sens et Expressivite: Usage courant. Forme: En France: on nous apporte a peu pres la moitie d'un verre de jus, une petite carafe d'eau, du sucre et une cuillere. En Espagne: on nous apporte un verre rempli de jus. Si le cafe η 'est pas tres "sirieux", on pourra ajouter un peu d'eau dans le jus sans que nous ne nous rendions compte de la "tricherie". Le jus, sauf si on le demande, η'est accompagne ni de carafe d'eau, ni de sucre. Quel que soit le type et quelle que soit la complexite des signes (V, PV, NV), on peut faire une analyse comparative des systemes culturels en contact selon ces deux perspectives - de l'expression et de la comprehension. Une presentation analytique, exposee dans la perspective de la comprehension, permet egalement de rendre compte de Phomonymie in-

367

Dialogue entre cultures differeittes: le role de I 'ethos collectif

terculturelle des formes identiques (II) qui n'ont pas le meme sens ou la meme expressivite (III) dans les cultures en contact dans une situation de communication comparable (I). COMPREHENSION I SITUATION X est re9u par Y negativement

II FORME

III EXPRESSIVITE

Y propose verbalement sous forme de question ä X quelque chose έ boire

Ouganda: marquee (Y met en embarras X, qui peut finir par refuser) France·, non-marquee Boheme: non-marquee

Y prepare έ boire ä X sans le lui proposer verbalement

Ouganda: non-marquee France·, marquee negativement Boheme: non-marquee

Y fait visiter sa maison ä

Ouganda: marquee (Y est perpu par X comme vaniteux) France: non-marquee ou polie Boheme: non-marquee ou polie

Une presentation analytique, exposee dans la perspective de l'expression permet, ensuite de rendre compte de la realisation d'une expressivite comparable (II) par les formes stereotypies (codees) dans chacune des cultures (III) en contact dans une situation de communication comparable (I). EXPRESSION I SITUATION

II EXPRESSIVITE

X est repu par Y

non-marquee

III FORME

Ouganda: X n'apporte rien ä boire ni ä manger έ Y France: X apporte ä Y une bouteille ou un dessert Ouganda: Y ne propose pas verbalement de boisson a X ; il le sert sans lui demander ce qu'il veut consommer France: Y propose ä X verbalement un choix de boissons Ouganda: Y ne presente pas sa maison ä X France: Y peut presenter sa maison a X si X ne la connait pas encore

368

Milena Srpovd

5. Conclusion: les universaux

Les concepts ne sont pas utilisables si l'on ne distingue pas rigoureusement leur face forme et leur face contenu. La distinction de ces deux faces fonde la distinction des deux perspectives d'analyse: la perspective de la comprehension des formes et la perspective de l'expression des contenus. Dans la perspective de la comprehension, on peut considerer comme universaux, relativement aux communautes culturelles observees, des formes cummunes a ces communautes et dont le contenu est identique. Dans la perspective de l'expression, on peut considerer comme universaux, relativement aux communautes culturelles observdes, les contenus communs k ces communaut6s (leurs formes peuvent etre identiques ou differentes). A la question: qu'est-ce que "etre genereux"?, on ne peut pas repondre. On ne peut repondre qu'ä la question qu'est-ce que "etre g6nereux" dans telle situation de communication, localisee dans telle communaute, a tel moment, avec tels interactants? La reponse sera l'inventaire des formes de comportement V, PV et NV statistiquement les plus frequentes (la perspective de l'expression). A la question "quel est le contenu "universel" du "concept" "etre genereux" "?, on ne peut pas repondre. On ne peut que chercher ä savoir la part de l'universel dans deux, trois, quatre, communaut£s x, dans l'expression d'un "concept" comportemental comparable (la perspective de l'expression). Ayant repertorie des formes, on peut ensuite chercher a savoir, lesquelles parmi les formes qui apparaissent dans deux, trois ou communautds x, ont le meme contenu dans toutes ces communautes (perspective de la comprehension): ces formes seront des universaux des deux, trois ou communautes χ concernees, ä Pexclusion des autres, que l'on n'a pas etudiees.

References bibliographiques

Gumperz, J. (1989): Engager la conversation. Introduction ä la sociolinguistique interactionnelle. Paris: Minuit. Hymes, D. (1984): Vers la competence de communication. - Paris: Hatier-Cridif (LAL). Kerbrat-Orecchioni, C. (1944): Les interactions verbales, t. III. - Paris: A. Colin. Srpova, M. (1999): Le non verbal dans la communication interculturelle. - Actes du Colloque international de la Federation internationale des professeurs de fran9aisl999, CIEP, Saint-Cloud, 2000, 17-24. Ce texte est un prolongement des contributions precedentes de l'auteur aux travaux de l'IADA, ä savoir:

Dialogue entre cultures differentes: le role de I 'ethos collectif

369

(1991): L'exp6rience ethnolinguale et ses consequences pour le dialogue interlingual. Pour une approche pragmatique des contenus lexicaux en situation interlinguale, Dialoganalyse, III, Teil 2, Tübingen: Max Niemeyer Verlag, 377-388. (1992): Differences extralinguistiques dans le dialogue entre locuteurs issus de cultures differentes (esquisse methodologique), Methodologie der Dialoganalyse, Tübingen: Max Niemeyer Verlag, 97-101. (1994): Accidents de la comprehension et de l'expression dans le dialogue interculturel, Actes du colloque international,"Le dialogue en question", Lagrasse, octobre 1993, Presses de l'universite Toulouse-le-Mirail: Cahiers du Centre interdisciplinaire des Sciences du langage, n° 10, 441-449. (1997): Approche contrastive dans l'apprentissage des langues et des cultures, Actes du 5®me colloque international de 1'I.A.D.A., Paris III (CRELIC et IADA), 1994, Dialoganalyse V, Tübingen: Max Niemeyer Verlag, 341-349. (1997): L'expressivitd et la relation entre participants de communication interlinguale et interculturelle, Actes de la Table ronde de 1' I.A.D.A. (Universite de Bologne, mars-avril 1995): Dialogue analysis: Units, relations and strategies beyond the sentence, ed. by Edda Weigand, Tübingen: Max Niemeyer Verlag, 215-222. (1997): Le calcul des proc6des de traduction, colloque international de 1'I.A.D.A, Universit6 de Lugano (1997), Paris: PUF, La linguistique, 33/1, 13-22.

Grazieila Tonfoni Information Transport through Dialogue Α System for Reducing Fuzziness due to Culture and Context Shift

1.

Introduction

Research in dialogue analysis is flourishing today. A whole set of the most relevant approaches demonstrate the need for accurate exploration of a variety of aspects involved (Dascal 1994 and Weigand 1998, 1999, 2000). The present approach is meant to support strategic conversations management and dialogue for decision making with a set of conceptual tools for analyzing sequences at a very high level of accuracy. Naturally occurring instability and sudden shifts of context and role between and among participants, due to continuous change, needs to be compensated by a flexible interpretive model for dialogue management. This allows changes and transitional states to be identified, reported, traced back and triggered with their own originating context through the use of a consensually agreed upon and shared system for interpretation. To summarize: each contributor's action carried on through an enhanced annotation system will significantly gain out of consistent visualization of the conditions in which the conversational interaction was first generated and then maybe even radically transformed, due to new conditions of satisfaction having occurred. Even if interpreted as challenging, there is no doubt that change and unstable conditions, during a conversational intercourse, are perceived as a major problem which needs to be carefully considered. Evidently individuals involved with a certain conversation perceive and resent instability either consciously or unconsciously, even if encouraged to feel committed to what they have themselves declared. Dialogical communication today is subject to particularly strong waves of external and multiple context resonance due to communicative context shifts bound to different media involved and multiple participants taking turns. Further stages of transformation need therefore to be accurately reported and contextualized consistently. As a consequence of instability, dialogue management, subject to continuous change, is characterized by the following features: increasingly wide amount of information coming in fuzzy, both synchronously and asynchronously; increasingly wide amount of interferences to be checked and verified continuously;

372 -

Grazieila Tonfoni increasingly frequent change of conditions of satisfaction and continuously occurring communicative context shifts.

Individuals participating in the conversation may be seen as dialogue managers, in charge of meeting the conditions for consistent interaction to occur. If dialogue managers are not able to visualize relevant information and consequently to envision a consistent model for taking action (Tonfoni 1998), they will not be able to make reasonable projections and predictions and they will be inclined, as a consequence, to proceed toward attribution of inconsistent meaning to dialogue utterances. Dialogue management, if subject to change of roles played by single individuals and teams, turns out to be a very delicate operation. If individuals involved do not foresee how long and to what extent they will actually be in charge of and effectively responsible for a certain conversation they have initiated, they are more likely to fail preserving a consistent context for interpretation. Dialogue management is strictly bound to accuracy, motivation and responsibility shown by individuals involved in the process of meaning creation and further development throughout a whole set of transitional moves during the communicative intercourse. The framework proposed here is aimed toward creating and keeping optimized conditions for positively coping with dialogue occurring under unstable conditions by reinforcing the participants' role and responsibility, and providing both conceptual and practical tools meant to establish most favourable conditions for taking action based upon consistent attribution of context bound interpretative clues.

2.

Dialogue management as a learning opportunity

Dialogue comes from Greek "dialogos", which means back and forth exchange, and is the output result of whole sets of interactions performed by individuals collaboratively, therefore creating context for meaning attribution and sharing. Current interactions, occasional conversations as well as planned communicative intercourses establish the conditions for information to flow throughout different media and in many and various ways and create the context for further understanding. Information derived from a dialogical interaction may therefore be viewed as coming in various flows and waves, to be filtered, categorized and organized as to be accessible and reusable for different purposes at different times within the same culture. As soon as some dialogue derived information is found to be of relevance so to become stabilized and be

Information Transport through Dialogue

373

turned into a piece of knowledge, then the need to store information in ways which may be made fully transparent, becomes a very important issue. Information derived from a dialogical intercourse and passed around needs to be based upon both topical continuity and contextual consistency: this is in fact a most fundamental process upon which accurate and timely decision making resides. Availability and accessibility of contextualized packages of information generated out of conversations is to be supported by an enhanced labelling system, which may help speed up effective retrieval of strategic knowledge. Each dialogue sequence - where by dialogue sequence a consistent topic and context completed section of a conversation is indicated, which is recognized as a unit by itself, ready to be linked up to other dialogue sequences, according to topic continuity and context consistency - will be labelled according to qualitative reasoning criteria upon the nature of information exchange occurring throughout verbal interaction, and context and culture shifts. Statistical methods for identifying topical words occurring in conversation in the form of keywords are not at all adequate for supporting qualitative reasoning. Qualitative reasoning upon dialogue derived information is the result of a whole set of complex operations. By showing explicitly which kind of information each dialogue sequence actually contains, the originating context in which the conversation was first initiated becomes visible. By showing which kind of progressive revisions have in fact produced various and some time even contradictory transitional states, it will also become possible to trace back individuals' contributions all throughout the different context shifts which have occurred at different times. Not only is a very specific "context sensitivity" required, but also a consensually shared framework for interpretation needs to be available and referred to, in order to label sequences of dialogue consistently, based upon a common understanding and naming of the different operations performed, which may be referred to as a whole set of pragmatic categories. As a consequence of explicit illustration of a variety of actions taken upon dialogue derived information, the same overall information will become a learning opportunity for those individuals involved, who may then feel the need to modify the conversation by reformulating it either locally or globally at different times. According to such perspective, each conversation will carry its own history attached while undergoing changes of various kinds and may therefore be productively viewed as transitioning throughout different states of context attribution, which affect its originating communicative value more or less radically. Transitional states in a conversation are therefore to be considered those temporarily defined and stabilized states of information which provide evidence and support for a certain set of interpretive processes which have occurred or are going to occur next.

Grazieila Tonfoni

374

Context change obviously will affect conditions of satisfaction any communicative occurrence entails as well as priorities and roles played by individuals during the conversational interaction. Temporarily stabilized contextual conditions will provide evidence and visibility for each individual who is actively involved within the process of interpretation. In the complex environment of information we live in today, conversations occur both synchronously and asynchronously throughout a variety of channels. They may be initiated through one channel like a telephone conversation and be passed to an e-mail system to be then reinforced by a real person to person conversation. It is most important to keep track of discourse sequences generated in such a disperse communicative set up. This is why the concept of transport is productively applied to information management: sequences of conversation may in fact build up very important knowledge, which has to be preserved to be able to proceed further. Transport may cause some of the originating information to become lost or modified in the course of passing throughout media and various cultural set ups or just passing from one context of conversation into another. There is therefore a need for specific context reassessment or assessment of a new context as to make sure that unwished meaning concretions do not accumulate on top of the originating dialogue sequence once extracted and transferred to be further used. This is why just like for documentation management (Tonfoni 1999) attachment to each dialogue sequence containing a linearized and enriched transcription of topical and contextual information is found to be of relevance.

3.

Enhanced encoding procedures for labelling dialogue sequences

The process of labelling dialogue sequences may be made visible throughout a consistent interpretive system meant to describe and define different kinds of communicative actions taken by individuals in the course of a conversation, and finally stored electronically. Such operations are here categorized and analytically defined in the process of progressive organization, each sequence is likely to undergo in the course of a conversation as to be finally complemented by a consistent attachment. Once a consistently interpreted and appropriately selected dialogue sequence is labelled and recognized in its own originating context, it may then be reconfigured by its participants and more or less radically transformed if that was to be found appropriate. Such a process may be activated after relevant clues have been extracted and visually represented by specific icons, supported by the Context Transport Mark up Language also

Information Transport through Dialogue

375

called CTML (Tonfoni 1996, 1999), which are of four kinds, and are precisely the following ones: -

mark up signs: meant to indicate the communicative function or type of a dialogue sequence;

-

mark up symbols: meant to indicate the communicative style of a dialogue sequence;

-

mark up turn taking symbols: meant to indicate the role and interplay between the dialogue sequence producer and the dialogue sequence receiver;

-

mark up amplifier symbols: meant to coordinate wide sets of dialogue sequences, which are characterized by topical continuity and context consistency.

CTML is a system for controlling sequence by sequence the level of topical consistency and topical continuity of communicative interactions, which are of strategic relevance for decision making. It is therefore most important that each sequence is attributed a consistent context so that fiizziness or distortion are radically reduced. By topical continuity, we mean to indicate dialogue sequences focusing on the same topic, which may be either literally extracted as linear sequences out of the dialogue or abstracted as a result of accurate interpretation and further adaptation at a more conceptual level. By contextual consistency, we mean to indicate dialogue sequences showing the same communicative context, which may be explicitly declared as to be easily retrieved. Mark up signs, which represent the various communicative functions a dialogue sequence may convey, are the following ones: Square: for an informative dialogue sequence, which carries information about a specific event or fact, to be linked up with another dialogue sequence made available in order to extend topical continuity and context consistency. Square within the Square: for a summary of a certain dialogue sequence, which has been produced to reinforce contextual consistency between an original dialogue sequence and its own abstract. Frame: for a dialogue sequence, which is found to be analogous in content to other dialogue sequences; it is meant to reinforce contextual consistency between and among different dialogue sequences.

L

Triangle: for a memory and history generated out of a certain dialogue sequence, meant to —establish topical continuity with background information, which is found to be of direct relevance.

Ο

Circle: for a main concept conveyed by a certain dialogue sequence, which has been abstracted and linked to other dialogue sequences, showing topical continuity. It is meant to reinforce topical identification and to effectively link together dialogue sequences which show the same topical word.

Graziella Tonfoni

376

Grouped Semicircles: for main concepts, which are abstracted out of an originating dialogue sequence and meant to establish both topical continuity and context consistency between the originating dialogue sequence and a set of topical words .

D

A

Semicircle: for a locally identified concept abstracted out of a dialogue sequence and meant to reinforce context consistency by establishing further links to other dialogue sequences, which show the same topical word. Inscribed Arcs: for indicating the need for an upgrade and update of a certain dialogue sequence; it indicates that a revision process is likely to occur, though it does not declare if such revision will be a major or a minor one.

f Iν / \

Opened Text Space: for indicating that an upgrade and update has indeed occurred within a certain dialogue sequence; it indicates that such dialogue sequence has now reached a new revision state as a consequence, though it does not declare if the revision has been a major or a minor one.

Ζ

Right Triangle: for a comment made to a certain dialogue sequence, where more contextual information is needed, which has to be derived from other external sources, not previously available, based on topical continuity.

Mark up symbols are meant to indicate communicative intentions and styles more locally within a certain conversation, which means dialogue sequence by dialogue sequence. They are particularly useful to show contributions made by individuals in the process of creation of meaning and may be easily incorporated within the final output result to provide further interpretative clues which may significantly add to clarity and visibility. Mark up symbols, which represent different modes of information rendering activated at different times or at the same time, may be combined and used dynamically, because they effectively indicate conversational transitional states, by declaring explicitly the nature of those changes, which have occurred or are likely to occur next. Symbols are the following ones: Describe: from Latin describo: write around. It means complementing the original dialogue sequence with as much information as may be found interesting to add without any specific constraints. It is represented by a spiral, which starts from a central point - the middle point of the spiral indicating the original topic - and proceeds toward expanding the topic at various degrees, linking it with other information coming in from different sources and found to be relevant to facilitate the originating and intended interpretation. In this specific mode, there is actually no need at all to follow any chronological order; the spiral may be smaller or larger, depending on how much information the dialogue manager and producer an individual or a team - may find relevant to add. Define: from Latin deflnio: put limits. It means complementing a dialogue sequence with limited information about a defined t0 P'c, which has been previously selected and identified as the most relevant one, which is represented by the middle point of the square.

Information Transport through

Dialogue

377

It indicates that there is a specific need to incorporate specific information about a relevant dialogue sequence which is made available. Define means actually describe under specific constraints and implies accurate and most selective focusing on a very limited quantity of highly specific information. i"m V V "'•

Narrate: from Latin narro: tell the story. It means complementing a dialogue sequence with various facts and events, which have been referred to in the originating context, by following a logical and chronological order. It indicates a set of major points or facts representing different diachronic stages, which are strictly linked up together in a longer sequence. The longer the sequence is, the more narrative points are actually added according to the dialogue participants' decision making. Point out: take a point out of a story chain. It means isolating a specific event or fact among those reported within a single dialogue sequence and focusing on just that one and adding more detailed information by expanding it, significantly linking it with other dialogue sequences which have been found to be of relevance to the point made. It represents a specific topic chosen and obviously entails the need to look for more extended information provided and made available from other sources. r~l •I I 1 ι I [ ι 1 I I

Explain: from Latin expiano: unwrap, open up. It means that facts and reasons are given to support interpretation of a certain event within a certain dialogue sequence. Dialogue participants may start by indicating the originating cause and proceed toward showing the effects or start with effects and go back to the cause, according to what is found to be more significant.

Regress: from Latin regredier, go back. It means that more information about a certain topic presented within the sequence I is absolutely needed to gain a deeper understanding. I 1 * 1 ^ it represents a specific topic focusing process and an in depth information expansion, which is activated only for that precise topic. Dialogue participants may want to consider if further information is needed on that and ask for availability of further resources.

21

^

^

Inform: from Latin informo: put into shape, shape up. It means that any dialogue sequence is the result of some information packaging and that the specific dialogue sequence indicated is organized in the most unconstrained way and thus subject to many and various kinds of reformulation. It usually leads toward two different kinds of specification which are respectively conveyed by the "inform synthetically" and the "inform analytically" indications:

"O

CC^

inform synthetically" means departing from a larger dialogue sequence and proceed toward a summary related to a specific topic, identified as being the most relevant one emerging from the originating dialogue.

"inform analytically" means departing from a given dialogue sequence as to expand toward further information.

Grazieila Tonfoni

378 ^—x

Reformulate: from Latin reformo/reformulo: change shape and shape again. It means changing the kind of style which was adopted before and substituting it with a different one, still related to the same dialogue sequence. It may turn into a more or less radical transformation of the originating sequence, according to a precisely defined topic. x

— E x p r e s s : from Latin exprimo: push out and press out. /' It means adding personal opinions and individual feelings related to facts and events """" within a dialogue sequence; it indicates the most subjective mode of information organization, which is openly recognized as bound to personal evaluations, judgements and emotional states. Mark up turn taking symbols are meant to define the mode of accessing a given dialogue sequence requested at each given time; they are the following ones: Major Scale: it shows that literal interpretation is needed and that those dialogue sequences indicated and marked off should be extracted and quoted literally, the way they were first presented. Minor Scale: it shows that accurate interpretation may need a further process of abstraction and that a dialogue sequence indicated and marked off may undergo significant reformulation processes up to high level conceptualization. Open or Unsaturated Rhythm: it shows that just accessing a dialogue sequence may lead toward incomplete interpretation of those facts and events which are presented. It is meant to suggest access to more dialogue sequences and various kinds of sources which are made available. II

III

TTTTT

Tight or Saturated Rhythm: it shows that accessing a dialogue sequence will lead toward complete interpretation of those facts and events which are presented. It is meant to suggest to stick to the interpretation provided, though access to other sources is available, as to support more evidence.

Mark up amplifier symbols come last and may be used only after the previously illustrated ones have been applied; they are meant to complement a whole set of dialogue sequences to indicate specific operations, which are to be performed to connect those sequences to previous ones, which have been already annotated and accurately stored. They are the following ones:

Choose: it is meant to represent the dynamic process of first identifying and then deciding between optional contexts for interpretation, which are mutually exclusive, given a certain set of dialogue sequences.

Information Transport through Dialogue

379

Identify: it is meant to represent definition of a more specific context, within a broader context for interpretation of a set of dialogue sequences; it naturally occurs before "search" and "select".

/ \

Search: it is meant to represent the dynamic process of choosing among different contexts for interpretation of a set of dialogue sequences, which are many and compatible as to find the most appropriate one.

Ill}

ύ

Select: it is meant to represent multiple contexts which may evolve either synchronously or asynchronously and may be modified once a certain decision making process based upon a certain dialogue has been performed, as to be stored and kept.

Copy/Replicate: it is meant to represent the dynamic process of duplication and repetition of a certain context, which, if lost, would radically jeopardize understanding and accurate interpretation of a set of dialogue sequences. Ahead: it is meant to represent the progression of a certain set of dialogue sequences which are linked together by context consistency or harmoniously evolving contexts. Back: it is meant to represent the need to go back to delete and replace the originating context which has radically shifted in the course of various transition states, such that, if not eliminated, would indeed affect consistent interpretation of a whole set of dialogue sequences.

->

Conflict: it is meant to represent an emerging inconsistency and incompatibility between various context attributions to a set of dialogue sequences, which need to be cleared as to proceed toward any further interpretation.

The notational system here illustrated in its various components may be applied at different layers and at various levels o f complexity and is meant to underline the fundamental role and responsibility o f individuals participating in the conversation. If a conversational interaction meant to produce an agreement or be part o f a collaborative project between and among its participants could not be completed because the partners involved were disconnected at s o m e point, still relevant parts o f the conversation may be triggered back as to be continued and completed with further interaction. According to such perspective, individual initiative may be harmoniously incorporated as part o f the enhanced mark up process, which will then reflect directly upon the history o f each dialogue occurring in a longer time span and throughout a whole diversity o f media. Technology may therefore serve as a powerful carrier, transporting through time and space dialogue sequences viewed as the output result o f individuals and teams, w h o s e contributions are made visible according to what they decide and agree should be made

380

Graziella Tonfoni

visible and explicit, as representing the intended context in which the conversation did actually occur. The whole concept of enhancement in dialogue sequences analysis naturally leads toward the concept of harmonious integration of different information systems and technologies in most difficult conditions, as those characterized by continuing change of context, time and space.

4.

Conclusions

Attaching specific culturally bound meaning to each dialogue sequence may result in a very positive process, especially if interpretation by each contributor within a specific culture is made available all the time and it may constitute an effective repository of knowledge to be building upon. Consolidating background assumptions as well as knowledge is a guarantee against useless dispersion and dissipation of relevant information, which is most likely to have been accumulating in the course of the years through the individual and collective effort of interpreting day by day events to turn them into expertise to be made available. To conclude: accuracy of interpretation within any conversational interaction is significantly bound to deeper understanding of an already existing culture and of the reasons why such culture evolved in certain ways. New meaning creation may only be conceived and designed as a consequence of observation, analysis and most accurate interpretation of previous models of reasoning and acting.

Acknowledgment

The author wants to thank Dr. Pete Daniel, Curator in the Division of the History of Technology at the National Museum of American History at the Smithsonian Institution in Washington D.C., for accurate reading of this paper.

Information Transport through Dialogue

381

References

Dascal, Μ. (1994): Speech act theory and Gricean pragmatics: Some differences of detail that make a difference. - In: S. L. Tsohatzidis (ed.) Foundations of Speech Act Theory: Philosophical and linguistic perspectives, 323-334. London, New York: Routledge. Tonfoni, G. (1996): Communication Patterns and Textual Forms. - Exeter, U.K: Intellect. Tonfoni, G. (1998): Information Design: The knowledge architect's toolkit. Lanham, U.S.: Scarecrow Press. Tonfoni, G. (1999): On augmenting documentation reliability through communicative context transport. - In: "The Proceedings of the 1999 Symposium on Document Image Understanding Technology", 283-286. SDIUT99, Annapolis MD, April 1999. Weigand, E. (1998): Emotions in dialogue. - In: S. Cmejrkovä, J. Hoffmannova, O. Müllerovä, J. SvStlä (eds.) Dialoganalyse VI. Proceedings of the 6 th International Congress on Dialogue Analysis, Prague 1996, Bd. I, 35-48. Tübingen: Niemeyer (Beiträge zur Dialogforschung 16). Weigand, E. (1999): Rhetoric and argumentation in a dialogic perspective. - In: E. Rigotti (ed.) (in collaboration with S. Cigada) Rhetoric and Argumentation, 53-69. Tübingen: Niemeyer (Beiträge zur Dialogforschung 19). Weigand, E. (2000): The dialogic action game. - In: Μ. Coulthard, J. Cotterill, F. Rock (eds.) Dialogue Analysis VII: Working with Dialogue. Selected papers from the 7th IADA conference, Birmingham 1999. Tübingen: Niemeyer (Beiträge zur Dialogforschung 22).

Elda Weizman

News Interviews on Israeli Television: Normative Expectations and Discourse Norms

1. Asymmetry in news interviews

News interviews have been widely described as a case of highly structured institutional discourse, whereby deviations from the norms of everyday talk are reflected, among other things, in an asymmetrical division of labor between the participants: "Incidences of powerful language may be found in any kind of talk interaction, but such speech is especially likely to be found where power is asymmetrically distributed: e.g., bossemployee, doctor-patient, and interviewer-interviewee exchanges" (Osley/Scotton 1984: 262). In an overview of research on institutional discourse, Drew and Sorjonen point out that "perhaps the most evident constraint lies in turn-taking systems which depart substantially from the way in which turn-taking is managed in conversation [...] These turn-taking systems involve the differential allocation of turn types among the participants; notably, the interactions are organized in terms of question-answer sequences, in which questioning is allocated to the professional (attorney, interviewer, teacher) and answering to the 'client' (witness, interviewee, pupil)" (1997: 102). The extensive work done on news interviews within the framework of conversation-analysis further indicates that deviations from these routines are usually repaired, acknowledged, or sanctioned. For example, statement components produced by interviewers normally precede the production of questioning components. Interviewees, for their part, "routinely treat IRs' statement turn components as preliminaries to questioning turn components.[...] Departures from the standard question-answer format are frequently attended to as accountable and are characteristically repaired" (Greatbatch 1988: 404). d a y m a n (1988) further elaborates on the types of interviewer's evaluative statements, shows that in his data interviewers' statements are either preceded or followed by questions, and concludes: "In sum, IRs frequently produce evaluative statements, many of which disagree with, criticize, or otherwise challenge an IE [...] Yet because of the various discourse practices outlined above, the participants - as well as the audience - can see such statements as an integral part of a larger questioning action occupying the entire turn at talk. Accordingly, regardless of how opinionated or argumentative these turns might also appear to be, their

Elda Weizman

384

primary or "first-order" business remains that of questioning" (Clayman 1988: 479). Heritage and Greatbatch (1991) elaborate on the manifestations of this division of roles, delineating, among other constraints, the length of the interviewee's turns and the expectation that the interviewer conduct the interview. Deviations from these constraints are shown to be acknowledged, repaired or sanctioned. Interviewees' comments, for example, when occurring, are often directed at their co-interviewees. Interviewers' departures from the expected question constraint are comparatively rare, and often engendered by the interviewees or sanctioned by them (Heritage/Greatbatch

1991,

Heritage 1998). In the Australian context, the relatively high incidence of interviewers' declaratives has been accounted for in terms of gender-specific styles (Winter 1993). This paper explores patterns of division of labor in news interviews on Israeli television. Specifically, the analysis posits a discrepancy between interlocutors' normative expectations, which presuppose an asymmetrical division of labor in terms of turn taking between them, on the one hand, and real-life practice, which exhibits relative symmetry in speakers' selection of discourse patterns and reciprocity in the use of challenge strategies, on the other. By symmetry I mean that all the parties, be they interviewers or interviewees, have access to the same set of discourse strategies. In other words, it is not the case that any of the examined strategies is reserved for only one party, and is therefore avoided by the other. Reciprocity means that the use of challenge strategies is habitually triggered by the selection of a similarly challenging strategy by the other party. To illustrate these claims, I will analyze some of the findings of a study on news interviews on Israeli television, based on two sets of data: (a) an open corpus of metacomments made by interviewers and interviewees in television shows, on the radio and in the written press in Israel; (b) a 24-hour videotaped corpus of news interviews, i.e. 48 halfan-hour shows of the daily program Erev Xadash (literally "new evening"), broadcast daily at 17.00 p.m. on Israeli national television, channel one. 1

In the transcribed interviews, the punctuation stands for transcription signs, as follows: , = a brief pause, the number of commas represents the length of the pause . = utterance falling intonation ? = utterance rising intonation (.) (?) = utterance mixed intonation, with a tendendency towards a falling or a rising intonation, respectively { } = overlap of two utterances [laughs] = comments on paralinguistic features

News Interviews on Israeli Television

385

2. Interlocutors' meta-comments: Normative expectations for asymmetrical division of labor

Basically, explicit meta-pragmatic comments on interviewing strategies reflect a simplified version o f the expectations for asymmetry described above. The meta-comments in examples 1-4 below indicate that interviewers are expected to ask questions (ex.1, 2, 4), and to refrain from giving explanations (ex.1, 3) or from making assertions, even if they are interpretable as reformulations (ex. 4). Interviewees, on the other hand, should express opinions (ex. 3), and should refrain from asking questions (ex. 2): (1) 01 lee: Won't you please explain what the problem is? 02 ler: Our job is not to explain. Our job is to ask questions. (liana Dayan, Musaf Hamusafim, 5.5.89) (2) 01

The question is whether you will do everything you deem necessary [to ensure corrective discrimination in favor of women-politicians] 02 lee: Well, what do you think? 03 ler: 1 am asking the questions here 04 lee: Sorry? 05 ler: I am asking the questions here, (giggles). I am the interviewer. You are the interviewee. (Shelly Yechimovitz and leader of the Ma'arach opposition party Ehud Barak, Hakol Diburim, IBA, channel 2, 7.12.97, 1145). ler:

(3) 01 02

lee: ler:

What do you think, that we'll go on crying and do nothing? My job, Member of Parliament Dalya Itsik, is not to make suggestions. It is your job to do so. (Goni Mardor Biran and Member of Parliament Dalya Itsik, discussing the reaction of the Ma'arach party to the accusations made by the Minister of Justice against former Chief of Staff. IBA radio, channel 2, 25.6.97).

(4) 01

lee:

In other words, as far as you are concerned this business with Moshe Vardi is over now. 02 ler: I didn't say so. Don't put words in my mouth. 03 lee: No, I am only asking. (LioraNir and Miriam Mozes, Hakol Diburim, IBA radio, channel 2, 15.2.96). These normative expectations for asymmetry account for the host's refusal to express his own opinion (5) below, although both he and the interviewee are prestigious hosts in the Israeli media, and although the interview is managed as a conversation between peers:

Elda Weizman

386

(5) 01 Ier: This was another proof that power corrupts. 02 lee: Oh well 03 Ier: Go ahead, I'd like to hear what you have to say. 04 lee: No, you are the interviewee. (Yitshak Livni, Interviewer, and Dan Margalit, Interviewee, Sixa Bishnayim, IBA radio, channel 2,27.6.97, 2300)

3. Discourse norms: Symmetry in discourse patterns

To what extent does real-life practice fulfill the expectations explicated so far? The corpus analyzed here indicates variation in discourse patterns depending on the identity of the interviewee. When interviewing ordinary persons or professional experts, asymmetry is preserved in every respect. The extract in example 6 is indicative of the asymmetrical type: (6) 01 02 03

How many children are still at home?, Right now,, twelve., Can you tell us their names quickly to let us have an idea of the sort of inspiration you have when there are so many? 04 lee: All of them? Ier: You can try. 05 lee: lee: My eldest is Gila, Shula, and Shimon, Miri, Eli, Hanan, Yardena, 06 Naomi„uh David, Moshe, and Michal.uh Hagit, Meytal, Avi, Shay, Zachi, Yoni and Reout. (Rafi Reshef and Jacqueline Elharar, who raises 18 children in Kiryat Shmona, Erev Xadash, Israeli television, channel one, 9.12.91) Ier: lee: Ier:

When public figures are interviewed, on the other hand, the corpus features a different state of affairs. As expected, interviewers use more questions than assertions; however, they do not use them exclusively. More important, when assertions are made, they are not necessarily accompanied by questions, and are not always treated as requests for information by the interviewees. Obviously, given the role relationships between the participants in news interviews, interviewers' assertions are potentially interpretable as indirect questions employed to elicit information (Jucker 1986, d a y m a n 1988, Harris 1991, among others). However, interviewers' assertions on Israeli television are often evaluative and judgmental, and their assertive force is seldom masked by means of accompanying interrogatives. Furthermore, they are rarely repaired or sanctioned in any way. Interviewees, for their part, often actively negotiate the legitimacy or relevance of the

News Interviews

on Israeli

387

Television

issues raised by the interviewer. In more than one respect, this format seems to be rather similar to the one described for the Czech data (Cmejrkova, this volume). The following fragment is a typical example of the symmetrical type. Here, the interviewer makes judgmental statements, and interviewer's and interviewee's assertions alternate in a fairly long stretch of talk: (7) 01

lee:

[...] To my regret we, did not reach such an agreement and furthermore we, who started off from a much more positive starting point, [...] we find ourselves in a conflict lately 02 Ier: But perhaps 03 lee: And I think one of the reasons is, our refusal to share our basic positions with the US and I want to remind you that throughout the negotiations, when, we uh conducted negotiations with the mediation of the Secretary of State, he himself refused at times to convey proposals to the other side [...] 04 Ier: Perhaps in your time there was more trust then say in that of Zalman Shoval who succeeded you in office, since at that time Israel accepted all the Americans' terms more or less while today it is insistent! 05 lee: No! I don't think we accepted all the Americans' Terms 06 Ier: Most of them!,, Anyway we didn't clash with them on major issue!, 07 lee: Oh of course we did!, it's a fact that during that time we were undergoing a reassessment of our relations with the U S 08 Ier: That's during the 70's, that's something else., (Dan Margalit and former Israeli Ambassador to the US, Moshe Arad, on the tension between Israel and the US, Erev Xadash, Israeli television, channel one, 8.12.91)

Following an unsuccessful attempt to interrupt the interviewee (02) while he is presenting his viewpoint on the tension between Israel and the United States (01), the interviewer suggests an explanation of his own (04). This explanation, formulated as a statement, might be interpreted as an indirect request for the interviewee's opinion, mostly due to the use of 'perhaps'. The interviewer's next statement (06), on the other hand, is explicitly evaluative: in response to the interviewee's attempt to refute his alternative explanation ('No! I don't think we accepted all the Americans'

terms', 05), the

interviewer

reformulates his own view in a statement which hardly differs from an expression of opinion of the type expected from the interviewee ( ' M o s t of them!, Anyway we didn't clash with them on any major issue!', 06), and proceeds to discard the interviewee's interpretation ('That's during the 70's, that's something else.,' 08). The subjectivity of the interviewer's statement seems to be enhanced by the use of the solidarity pronoun ' w e ' (06) as a substitute for the referential naming 'Israel' (04), and by the emotional opposition between ' w e ' and 'them' ( ' W e did not clash with them', 06). The fact that interviewers make unacknowledged judgmental assertions without being sanctioned is interesting in itself, mainly because it diverges with the

normative

388

Elda Weizman

expectation that they ask questions. I would further like to suggest, however, that this breach of expectations be viewed as a constituent of a larger phenomenon, i.e. that it be studied in its challenge-value.

4. Challenge

The notion of challenge in news interviews, previously postulated in Weizman (1998, 1999), draws heavily on Labov and Fanshel's (1977) notion of challenge, underlined by Goffman's theory of role (e.g. Goffman 1959, 1969, 1971). For Labov and Fanshel (1977: 97), "a challenge is a speech act that asserts or implies a state of affairs that, if true, would weaken a person's claim to be competent in filling the role associated with a valued status". Accordingly, I suggest that discourse strategies be viewed as potentially challenging whenever they are interpretable as saying or implying that the hearer has not fulfilled with full competence his or her social role, or any component thereof. In news interviews, each speaker fulfills simultaneously two roles: the interaction and the institutional. The interactional role consists of the speakers' obligations as participants in the discourse event, e.g. the interviewer's obligation to manage the interview, and the interviewee's obligation to provide information and to speak his mind. The institutional role, on the other hand, consists of the speakers' social obligations in the relevant extratextual script, for example the interviewer's credibility as journalist, and the interviewee's obligations in his or her role as judge, member of parliament, minister, physician, sportsman, etc. Thus, Winter's (1993) comments on the complex relations between speakers' gender and their roles as interviewers and interviewees are accountable in terms of the conflict between institutional and interactional role, respectively. The distinction between these two types of roles is directly relevant for the analysis of challenge and of its implications, since each of the speakers' roles may be challenged separately. Thus, for example, it might be the case that the interviewer challenges the interviewee's competence in her role as chairperson of some world-wide corporation, by saying or implying that she has made a wrong decision; or, he might challenge her role as interviewee, by saying or implying that her answers are irrelevant. Obviously, it may well be the case that both roles are simultaneously challenged (Weizman 1998, 1999).

News Interviews on Israeli Television

389

4.1. Symmetry in challenge strategies Departing from the notion of challenge thus perceived, I suggest taking the analysis of norms and expectations one step further. In the corpus analyzed here, not only do the interlocutors have access to the same discourse patterns; they also engage in a series of mutually challenging turns, whereby interactional as well as institutional roles are being challenged. Challenge potential resides either in the content of the utterance, be it explicit or implicit, or in the choice of discourse patterns, in which case it is acquired by the violation of the 'common-sense' expectations for asymmetry. From this viewpoint, interviewers' judgmental assertions carry a high challenge potential, precisely because they violate the interlocutors' role expectations, as explicated in their meta-pragmatic comments. Thus, in (8) below, the interviewer challenges both the interviewee's institutional and interactional roles. Challenge potential resides in the content of the interviewer's utterances: (8) 01

Ier:

Chaim Ramon is saying however something clear!, he is saying that within four eight twelve years 02 lee: He is saying! 03 Ier: What happened to the Avoda party in 77 will happen to the Histadrut!, 04 lee: {Listen} 05 Ier: {There is no} other way if {if if} the Avoda party becomes a satellie of the Likud., 06 lee: {Listen} he said the same thing before the elections[...] 07 Ier: {You're saying this is his natural place(.)} (Rafi Reshef and Israel Keysar of the Avoda (=Labor) party, former chair of the trade-union Histadrut, discussing the policy of the latter's succesor Chaim Ramon, 20.11.91).

Here, the interviewer hosts the former chair of the trade union Histadrut, Israel Keysar. By going in some detail into his successor's comments, which imply harsh criticism of Keysar's policy (01, 03), and by adopting the same criticizing attitude (05), the interviewer challenges the

interviewee's

institutional role as former chair. Furthermore,

the

interviewer chooses to make three successive statements (01, 03, 05), and does not let the interviewee regain the floor, despite the latter's numerous attempts to do so (02, 04, 06). Only the last t u m (07) may be interpreted as an indirect request for information. Since "the goal of political interviews suggests a further expectation that the object of the interview from the outset is the interviewee's answers" (Winter 1993: 120), by actively competing for the floor the interviewer challenges the interviewee's interactional role.

Elda Weizman

390

In (9) below, on the other hand, it is the interviewee, Justice Binyamin Cohen, w h o challenges the interviewer's interactional role. The issue on the agenda is the William Kennedy S m i t h ' s rape trial, and the invasion of the media into people's privacy. At this point of the interview, the interviewee has rejected the interviewer's attempts to compare the trial in question with the trials of Eichmann and of Demjaniuk in Israel, and, consequently, the interviewer tries to get back to the Smith case. T h e interviewee constantly interrupts him, insisting on the irrelevance of the comparison (02) and on its unpopularity (04), and thus, paradoxically, interferes with the interviewer's attempts to comply with his own request to change the subject. B y so doing, he challenges the interviewer's attempts to manage the interview and to set u p its agenda: (9) 01 Ier: Now if you'll allow me let's go back {to the trial against uh there} 02 lee: {But this is truly a different issue} 03 Ier: So go back to the issue of {the} 04 lee: {We}'11 be stoned let's leave it be 05 Ier: {Let's mention} that here uh 06 lee: {I didn't say anything you're my witness,,} (Rafi Reshef and Justice Binyamin Cohen, on the complex relations between the court and the media, 11.12.91)

Whereas in this example the interviewee's challenge is wrapped up in a non-serious, joking tone (04, 06), in the next one it is more serious. Here, Justice Cohen is being asked about the legitimacy of broadcasting trials on television. In his answer, he questions the legitimacy of the question in a condescending tone ( ' I ' m a bit surprised', 01), which brings to the fore the high status of his institutional role, thus challenging the interviewer's interactional power: (10) 01

lee:

So look I told you that, as we see it as I see it„uh there is something destructive in letting television into court but I belong to a dying generation, and you this is your job I'm a bit surprised that you're looking {at me expecting expecting} 02 Ier: {that I'm encouraging such a thing} 03 lee: me to tell you it's not right, maybe in the twenty-first century, this is the natural medium [...] (Rafi Reshef and former Chief Justice Cohen, on the complex relations between the court and the media, 11.12.91)

Note also that the interviewee does not cede the floor w h e n interrupted, ignores altogether the interviewee's words (02), and proceeds to complete his own line of thought (03) without digressing in any way, neither at the level of coherence, nor at the grammatical level ('expecting [,,.]me to tell you it's not right', 02, 04). The refusal to relinquish the

News Interviews on Israeli Television

391

floor plays an important role in framing an exchange as "confrontational" (Hutchby 1996: 82). In this specific context it seems also to play an important role in conveying the impression that the interviewee has at least as much control over the interaction as the interviewer.

4.2. Reciprocal challenge The last example illustrates not only the symmetrical use of challenge strategies; it is also indicative of their reciprocity: as noted above (section 1), in the Hebrew data, the use of challenge strategies is habitually triggered by the selection of a similarly challenging strategy by the other party. In (10), the interviewer's intrusive interruption (02) follows the interviewee's challenging comment on the illegitimacy of the interviewer's viewpoint, and is followed, in turn, by the interviewee's refusal to cede the floor, while ignoring the interviewer's words. Thus, challenges to interactional and to institutional roles are intertwined. Reciprocal challenge features also in (11) below. Here, the interviewee explicitly challenges the interviewer's competence in selecting topics for discussion, by bluntly pointing out that the interviewer's question is insignificant (02). In response, the interviewer ironically challenges the interviewee's former institutional role as ambassador (03), referring to his previous experience with insignificant questions:

(ID 01

Ier:

Do you see any point in Israel insisting that, the talks or meetings with the Jordanian Palestinian delegation shouldn't be split at this stage, when Israel has agreed that at some stage,they will be split., 02 lee: Dan as I see it this is an insignificant question, and within a week from now we won't remember having discussed it. 03 Ier: But when you were ambassador you answered many insignificant questions, think for a moment that you're an ambassador (tell me) do you think we should in, insist or refer to or take a firm stand on this issue? (Dan Margalit and former Israeli Ambassador to the US Moshe Arad, discussing the relations between Israel and the United States, and their effects on the peace talks, 8.12.91)

Reciprocity is exceptionally transparent in this case, since it involves the repetitive ironic echoing of the word 'insignificant' by both interlocutors (02, 03) (Weizman, in press). Typically, the use of reciprocal challenge is not as marked as it is here, but is nonetheless the norm in the context of Israeli news interviews.

392

Elda Weizman

5. Concluding remarks

In the following extract, liana Dayan, a leading host on Israeli television, talks about her preferred interviewing strategies: Interviewer: Do you enjoy aggressive interviews, confrontations? liana Dayan: I enjoy myself whenever I share with my interviewee an intellectual challenge, whenever I can exchange with him intelligent blows. (Gaffi Amir and liana Dayan, Yediot Aharonot, Shiv'a Yamim, 11.10.96). Unlike the meta-comments examined earlier, Mrs. Dayan does not seem to expect an asymmetrical division o f labor, nor does she expect her interviewee to be led by the interviewer. It seems to me that the findings described in this paper are in line with her wishful thinking. In the Israeli context, when interviewing public figures, both parties have access to challenging strategies, and use them reciprocally. A s w e have seen, real life practice diverges with normative attitudes, which presuppose asymmetrical

power

relations between interviewers and interviewees.

Acknowledgement

The research has been supported by the Basic Research Foundation Administered by the Israeli Academy of Sciences and Humanities.

References

dayman, S. E. (1988): Displaying neutrality in television news interviews. - Social Problems 35, 474-492. Cmejrkova, S. (this volume): Media dialogue as a genre of public oral discourse. - (1991): News interview openings: Aspects of sequential organization. - In: P. Scannel (ed.) Broadcast Talk, 48-75. London: Sage. Drew, P. and J. Heritage (eds.) (1992): Talk at Work: Interaction in institutional settings. Cambridge: Cambridge University Press. Goffman, E. (1959): The Presentation of Self in Everyday Life. - New York: Doubleday Anchor Books.

News Interviews on Israeli Television

393

- (1969): Role distance. - In: A. Lane (ed.) Where the Action Is: Three Essays, 40-103. London: Penguin Press. - (1971): Relations in Public. - New York: Basic Books. Greatbatch, D. (1988): A turn-taking system for british news interviews. - Language in Society 17, 401-430. Harris, S. (1991): Evasive action: How politicians respond to questions in political interviews. - In: P. Scannel (ed.) Broadcast Talk, 76-99. London: Sage. Heritage, J. (1998): Conversation analysis and talk: Analyzing distinctive tum-taking systems. - In: S. tmejrkova, J. Hoffmannova, O. Müllerova, J. Sv6tlä (eds.) Dialogue Analysis VI, 3-17. Tübingen: Max Niemeyer. Heritage, J. and D. Greatbatch (1991): On the institutional character of institutional talk: The case of news interviews. - In: D. Boden, D. H. Zimmerman (eds.) Talk and Social Structure, 95-129. Cambridge: Polity Press. Jucker, A. (1986): News Interviews: A pragmalinguistic analysis. - Philadelphia: John Benjamins. Labov, W. and D. Fanshel (1977): Therapeutic Discourse. - London: Academic Press. Owsley, H. and H. Scotton, C. Myers (1984): The conversational expression of power by television interviewers. - Journal of Social Psychology 123 (2), 261-271. Weizman, E. (1998): Individual intentions and collective purpose: The case of news interviews. - In: S. Cmejrkovä, J. Hoffmannova, O. Müllerova, J. SvStla (eds.) Dialogue Analysis VI, 269-280. Tübingen: Max Niemeyer. - (1999): News interviews on Israeli television. - In: G. Toury, R. Ben Shachar (eds.) Hebrew. A L. Haifa: Haifa University, (in Hebrew). - (2002): Addresser, addressee and target: Negotiating roles through ironic criticism. - In: E. Weigand, M. Dascal (eds.) Negotiation: The Dialogic Question. Amsterdam/Philadelphia: Benjamins. Winter, J. (1993): Gender and the political interview in an Australian context. - Journal of Pragmatics 20, 117-139.

Larissa Wunderlich

Treaties: A Comparative Analysis of a Complex Dialogic Action Game

1. Introduction

Assuming that language is used for communicative purposes and therefore is always dialogic, this paper will deal with complex dialogic action games, namely with declarative action games as they occur in treaties. I will explain below why I consider treaties as being of declarative nature and not of a directive one, as assumed by Hoffmann (1998: 534a) and others. The communicative unit to be analysed is, according to Weigand (2000), neither the single sentence nor the separate utterance, but the dialogic action game. There are different types of dialogic action games, for example,

explorative,

directive

or

representative action games, and the declarative action game, which is characterized by the fact that a separate verbal reaction to the action is not necessary (Weigand 1989). In the texts regarded here, the expected reaction, i.e. "confirmation", here "agreement", can be taken for granted, because the signatories represent their states and act in the name of their people, by whom they have been elected.

2. The corpus

As the treaties and declarations of the European Union are available in all its official languages, they offer a good basis for a comparative analysis of the various ways in which declarative action games are expressed in different languages. However, the languages considered here will be only German, English, French and Italian. The texts for the corpus were taken from the Internet, 1 and analysed with the help of the xlex-programme. This tool set makes our task easier, but we have to know quite exactly what we are looking for. As one might criticize the fact that these texts are constructed, I would like to argue that

1

Only the Internet addresses of the English versions are mentioned in the references, because it is possible to reach the other versions from there using links.

396

Larissa Wunderlich

nowadays it is a characteristic feature of contracts that they are put down in writing and are formulated in legal terminology, and that the texts of international treaties are translated by very competent translators, so that the exact reproduction of the original text is guaranteed. Therefore we can say that the texts of this corpus do represent actual language-in-use and that they are suitable for our purposes. There is to say that in this study the corpus is analysed from the linguistic point of view. The treaties and declarations are not judged by political, legal or historical criteria.

3. The functional structure of a declarative action game

Declarative as a speech act is not to be mistaken with the declarative clause of the traditional grammar. 2 Declarative action function means that by performing a declarative speech act a certain valid status is created, whether someone answers to this speech act or not. In institutional declarative action games like those regarded here, it is of course required that the speaker is authorized and able to create this valid status (cf. Searle 1979: 7). As we can assume that the contracting parties within the European Union fulfil these conditions, we do not need to discuss here the cases in which the speaker is not authorized or the conditions to fulfil the contract are not given (cf. Austin 1962). A declarative action game can be characterized by the speaker's purpose of creating validity by declaring it and of the hearer's confirmation, which does not have to be verbally expressed in general. Still, in some cases we find relics of the confirmation verbally expressed, e.g. in apologies: "I apologize for my bad behaviour." - "Accepted." There are different types of declarative action games, caused by different propositions (see below), and it is the types that create a certain status or that establish an obligation which we are interested in here, because they are characteristic of treaties. According to Searle (1969, 1979) the functional structure of an illocutionary speech act can be represented by the formula F(p), where F stands for the illocutionary force and ρ for the proposition. The illocutionary force can be seen here as the concept of "create by declaring".3 This action function can be considered to be universal and valid for all texts of 2

It is not to be mistaken with Searle's category of declarations either, because I do not use Searle's (1979) terminology of assertives, directives, commissives, expressives, and declarations, but Weigand's (1989) terminology of representatives, directives, exploratives, and declaratives.

3

DECLARE is to be distinguished from PROMISE, as the first is usually initiative, while the latter can generally be seen as a complex reactive speech act with a declarative component, an undertaking to do something (cf. Weigand 1989: 94ff.).

Treaties: A Comparative Analysis of a Complex Dialogic Action Game

397

contracts.4 The proposition, which is subdivided into referring and predicating, contains information about the state of affairs to which F is correlated. The predication varies in our case, i.e., we find predications that refer to a certain status at the moment of the utterance here the signing - , see example (1)

By this treaty, the HIGH CONTRACTING PARTIES establish among themselves a EUROPEAN COAL AND STEEL COMMUNITY, founded upon a common market, common objectives and common institutions. (ECSC Title I Art. 1)

where the European Coal and Steel Community exists at the time of the utterance or the signing respectively, and we find predications that refer to a future state of affairs, see example (2)

The Court of Auditors shall examine the accounts of all revenue and expenditure of the Community. [...] (CS Art. 45c, 1)

The validity of the taking over of an obligation exists at the moment of the verbal or written agreement, but, for obvious reasons, the obligation itself is fulfilled in the future. As a treaty by its legal nature is binding, the contracting parties are obliged to follow the wording of its declarations. I think that this required fulfilment of an obligation is the reason why contracts are often classified as directive text types, but in my opinion this directive character is of minor importance compared with the declarative character of the illocutionary force. Even if it is hardly ever explicitly mentioned, an utterance like "Hereby the signatories/the contracting parties declare that the following shall be valid:" mentally precedes the texts of the contracts or their single articles and paragraphs, making them declarative speech acts. Another type of declarative action game, the one creating a certain status of a person, does not appear directly in the texts of the treaties. We only find mentioned at the beginning of the texts, in the preambles, that the sovereigns and heads of state have designated certain people as their plenipotentiaries, but the corresponding declarative speech acts with which this status is conferred on the plenipotentiaries have already taken place, see example (3)

4

HIS MAJESTY THE KING OF THE BELGIANS, THE PRESIDENT OF THE FEDERAL REPUBLIC OF GERMANY, THE PRESIDENT OF THE FRENCH REPUBLIC, THE PRESIDENT OF THE ITALIAN REPUBLIC, HER ROYAL HIGHNESS THE GRAND DUCHESS OF LUXEMBOURG, HER MAJESTY THE QUEEN OF THE NETHERLANDS, DETERMINED to lay the foundations of an ever closer union among the peoples of Europe,

As treaties are legal texts, whose compliance is legally required and whose disregard would have certain sanctions as a consequence, they have this declaring force. For further information about the legal basis of contracts see e.g. Hoffmann 1998.

398

Larissa Wunderlich

[...] HAVE DECIDED to create a EUROPEAN COMMUNITY and to this end have designated as their Plenipotentiaries: [...] (EC)

where there is referred to the designation by a speech act report using present perfect (italicized by me). It is, of course, reasonable that the designation of the plenipotentiaries is not a declarative component of the treaty itself, because it is actually the plenipotentiaries who work out the agreements of the treaty, therefore they have to be designated first.

4. Principles of interaction

According to Weigand (2000) dialogic action games are governed by basic principles and corollary principles, which are of different importance depending on the type of action game. As not all of them can be mentioned here, I will only deal with those which are important for my point. In a declarative action game like the conclusion of a political treaty, the setting down in writing plays a decisive role, because the treaty will be valid only at the moment of signing. Without being signed, the text of a contract is nothing more than a sheet of paper. In the signing of the text, which stands for its uttering, the plenipotentiaries, acting on behalf of their nations, declare its validity. The putting down in writing is absolutely essential in the case of such momentous contracts, on the one hand to make sure that the agreements are always available and understandable, on the other hand because the texts are not only addressed to the contracting parties themselves but also meant to inform the general public, whose agreement is presupposed in the moment of the signing, which both argues for the dialogic nature of declarative speech acts mentioned at the beginning. A declarative action game like the conclusion of a contract of course does not appear out of nothing. It is preceded by the - often longstanding - action game of negotiation. In our case, the plenipotentiaries of the states involved certainly have had differing interests and positions, which they had to negotiate, because they necessarily had to reach a settlement on one single text. During these negotiations each party to the contract has had the opportunity to make their positions clear and to protect their interests, and in order to achieve their goals they followed Rhetorical Principles, which make an effective language use possible. These Rhetorical Principles used in the discourse of negotiation include, e.g., indirect 'persuasive' speaking or the use of politeness not with the intention of showing respect, but in order to win someone over and thus to gain his or her support.

Treaties: A Comparative Analysis of a Complex Dialogic Action Game

399

As far as the declarative action game is concerned I want to show the contrast between the corollary Principles of Politeness and of Unambiguousness only. The Principle of Politeness is extremely important in directive action games, where the speaker in the initiative speech act wants the hearer to do something for him or her, or the speaker in the reactive speech act might not want to hurt the interlocutor in case of a negative answer from his or her side. Polite utterances are often formulated indirectly, using the subjunctive and certain stereotyped phrases, which are inconsistent with the Principle of Unambiguousness, the most important principle in a declarative action game. The text of the treaty can be formulated without polite phrases, as it is here exclusively of importance that each party to the contract can clearly and unmistakably understand what they have agreed to and what obligations they have taken over. No room should be left for individual interpretations, which would enable the parties to construe the articles to their own advantage. Thus, the formulation of article 17 of the Treaty establishing the European Coal and Steel Community (4)

The Commission shall publish annually, not later than one month before the opening of the session of the European Parliament, a general report on the activities of the Community.

is a clear declarative statement that leaves no doubt of what has to happen. Polite phrases causing only lack of clarity are superfluous here. Signs of respect and politeness are only to be found in the references to the signatories: (5)

By this Treaty, the HIGH CONTRACTING PARTIES establish among themselves a EUROPEAN UNION, hereinafter called ,the Union*. (EU Title I, Art. A)

where they are called the "high contracting parties", written in uppercase letters, showing the reverence towards the signatories. In the cases analysed here, this showing reverence is caused by the especially high status of the signatories, i.e. heads of state or sovereigns and their plenipotentiaries respectively, and it is very evident in the speech act report in example (3). In a usual contract for sale or lease such a formulation would be exaggerated. The use of uppercase letters, however, is but a visual emphasis. Finally, we can say that the principle of unambiguousness is by far predominant over the principle of politeness in declarative action games establishing an obligation.

5. Verbal means

Using language we integrate several communicative means, i.e. verbal, cognitive and perceptive ones. When we analyse texts put down in writing from a pure observer's perspec-

400

Larissa Wunderlich

tive, the inclusion of facial play, gestures, undertones etc., however, is not possible. All we have which is empirically ascertainable is the pure wording. Therefore, in this chapter, we will deal with verbal means only. Cognitive means concerning the cultural framework of the action game are presupposed for the understanding of the texts. What is striking here in comparison to usual declarative utterances like "I declare the Olympic games open" or "I hereby designate Mr. X as my plenipotentiary", which correspond to the explicit performative formula [first person (I/we) + speech act verb + present tense indicative], is the use of the third person with reference to the speakers, see examples (1) and (5). This shows that a change from first to third person does not automatically cause a change from speech act to speech act report, as has often been assumed. In integrating our cognitive abilities, i.e. our cultural knowledge, we can recognize that the utterance is not representative, but declarative here. According to Searle's formula F(p), the speech act consists of three fundamental functional types, in our specific case the declarative action function and the propositional functions of reference and predication. The means indicating reference shall be mentioned only briefly. As it can be seen in table 1, the use or non-use of a definite article or a possessive pronoun (italicized by me) varies in the different languages, depending on the complex multiword expression, which consists of lexical and grammatical units. (Table l) 5 Aufgabe der Atomgemeinschaft ist es, [...] beizutragen.

It shall be the task of the Community to contribute to [...] (EAEC) The Community shall Aufgabe der Gemeinhave as its task [...] to schaft ist es, [...] zu promote [...] (EC) fördern. The European Coal Die Europäische Geand Steel Community meinschaft für Kohle shall have as its task to und Stahl ist dazu berufen, [...] zur Auswei- contribute [...] to economic expansion [...] tung [...] beizutragen. (ECSC) Die Kommission hat It shall be the duty of die Aufgabe, für die the Commission to enErreichung [...] zu sor- sure that the objectives gen. [...] are attained [...] (ECSC)

5

La Communauti a pour mission de contribuer [...] ä [...]

La Comunitä ha il compito di contribute [·..]»[...]

La Communaut6 a pour mission [...] de promouvoir [...] La Communaute europeenne du charbon et de l'acier a pour mission de contribuer, [...], ä l'expansion economique [...] La Commission est chargee d'assurer la realisation [...]

La Comunitä ha il compito di promuovere [...] La Comunitä europea del carbone e dell'acciaio ha la missione di contribute, [...], all'espansione economica [...] La Commissione ha I'ufficio d'assicurare l'attuazione [...]

The expressions in each horizontal section of the table are always taken from the same treaty, which is mentioned behind the English version.

401

Treaties: A Comparative Analysis of a Complex Dialogic Action Game

What is equally interesting about table 1 is that it illustrates that "identity of the expression predicated is not a necessary condition of identity of predication. Different but synonymous expressions can be used to express the same predication" (Searle 1969: 26 f n . l ) . This also explains why in these cases we will not find any rules (cf. Weigand 1997: 167) telling us that when the German version is like A, the English version is always like B, the French always like C, and so on. It hardly matters whether hat die Aufgabe,

ist dazu berufen, It

shall be the duty of, shall have as its task, a pour mission, est chargee de, ha il compito or ha I'ufficio is used, as the differences of the predication are so minute that it is almost impossible to make them explicit on the level of semantic representation. But although the predications are usually the same in the single texts, we find exceptions where they actually are slightly different, as shown in example (6), where the German expression verurteilen aufs schärfste is - in my opinion - noticeably stronger than the others: (6)

[...] verurteilen6 aufs schärfste alle Äußerungen von Intoleranz und Feindseligkeit sowie die Anwendung von Gewalt gegenüber einer Person oder einer Personengruppe wegen rassischer, religiöser, kultureller, sozialer oder nationaler Unterschiede; [...] vigorously condemn all forms of intolerance, hostility and use of force against persons or groups of persons on the grounds of racial, religious, cultural, social or national differences; (DRX 1) [...] condamnent avec vigueur toutes les manifestations d'intolerance, d'hostilite et d'utilisation de force ä l'egard d'une personne ou d'un groupe de personnes en raison de differences raciale, religieuse, culturelle, sociale ou nationale; [...] condannano con vigore tutte le manifestazioni di intolleranza, di ostilitä e di uso della forza nei confronti di una persona ο di un gruppo di persone a motivo di differenze di ordine razziale, religioso, culturale, sociale ο nazionale;

Of course, the other expressions - containing a graduation of the verb with a form of vigour - are quite strong, but the use of the superlative seems to me a bit stronger, especially because it would have been possible to use an expression like verurteilen verurteilen energisch level as vigorously

or verurteilen

mit Nachdruck,

scharf,

which in my view are on the same

condemn etc. The reason for this choice might be that the Germans

want to show that they take this delicate subject really seriously. Finally, I want to come back to the distinction between predications that refer to a certain status at the moment of the signing and those which refer to an obligation for the future (see 3). In the first case, the grammatical structure as shown in example (1) usually is used in all four languages. In the second case, however, some differences can be noticed. The forward-looking perspective is not always verbally expressed. We know about it by cognitive means. The German texts continuously show present tense, although the use of

6

In this case, italics are used in the original.

402

Larissa

Wunderlich

future tense would be an equal possibility. In the English texts the forward-looking perspective is expressed by the word shall and in a few cases may. In the French and Italian versions, however, we find a few articles where this forward-looking perspective is explicitly expressed by the use of a future tense (stressed by me): (7)

Gli obiettivi dell'Unione saranno perseguiti conformemente alle disposizioni del presente trattato, [...] Les objectifs de l'union sont atteints conformement aux dispositions du present traite, [...] The objectives of the Union shall be achieved as provided in this Treaty [...] (EU Tit. I, Art. B) Die Ziele der Union werden nach Maßgabe dieses Vertrags [...] verwirklicht.

(8)

II Parlamente europeo elaborerä progetti intesi a permettere l'elezione a suffragio universale diretto, secondo una procedura uniforme in tutti gli Stati membri. Le Parlement europeen elaborera des projets en vue de permettre 1'election au suffrage universel direct selon une procedure uniforme dans tous les etats membres. The European Parliament shall draw up proposals for elections by direct universal suffrage in accordance with a uniform procedure in all Member States. (CS Art. 2 1 , 3 ) Das Europäische Parlament arbeitet Entwürfe für allgemeine unmittelbare Wahlen nach einem einheitlichen Verfahren in allen Mitgliedstaaten aus.

We will, however, not find a rule telling us when to use future tense. In case (7), the Italian version is the only one containing future tense, and in case (8) the French and the Italian version contain future tense, so that we cannot say that these two Romance languages always use the same construction. In addition, there are other articles of the same nature where it would also be possible to use future tense, but where present tense is chosen. This leads to the conclusion that the rare cases showing future tense in this corpus are due to the individual style of the translator, as both present tense and future tense can be regarded as equal conventional possibilities in these cases. So far, I found only one example, where future tense is used in all four texts: (9)

Nell'esercizio dei loro poteri e perseguendo gli obiettivi delle Comunitä europee, essi rispettano e continueranno a rispettare tali diritti. Dans l'exercice de leurs pouvoirs et en poursuivant les objectifs des Communautis europeennes, ils respectent et continueront ä respecter ces droits. In the exercise of their powers and in pursuance of the aims of the European Communities they respect and will continue to respect these rights. (DFR) Bei der Ausübung ihrer Befugnisse und bei der Verfolgung der Ziele der Europäischen Gemeinschaften beachten sie diese Rechte und werden dies auch in Zukunft tun.

In this example, a clear distinction between the present and the future is necessary in all the texts. The German version shows a somewhat different construction avoiding the repetition of the same verb beachten, presumably because repetitions are considered as stylistically bad in German.

Treaties: A Comparative Analysis of a Complex Dialogic Action Game

403

6. Conclusion and further perspectives

I tried to make clear that we have to consider the cultural unit of the dialogic action game as the smallest autonomous unit of communication, because without the embedding of the text in the cultural framework of the action game and without the inclusion of our cognitive abilities it would not always be obvious that a speech act is, e.g., declarative and not representative. In contrastive studies, we have to concentrate on multiword lexical units instead of sticking to linguistic methods that are restricted to the analysis of single words. Therefore, we should plead for pragmatic dictionaries based on words-in-use. Although dictionaries are continuously improved, they still have shortcomings concerning the variety of phrases recorded. We can absolutely say that there are always several ways of formulating the declarative speech act in the four different languages, even in legal texts. The formulations are not completely fixed, so that the translators have some latitude concerning the choice. As the translators employed by the European Union are highly competent, we can presuppose that they know to which degree they can vary the formulations without noticeably changing the meaning of the text. It is necessary to continue the analysis in order to carry out a better systematisation, i.e. to differentiate further sub-types of the action game and their action structure, to describe the ways multiword lexical units are used for predicating, and to structure these phrases according to thematic and syntactic criteria (cf. Weigand forthcoming). This was a contrastive study, but not a critical study of translations. Therefore I did not take into consideration which of the texts was the original one, which are translations, and whether they are authoritative translations or not (cf. Labrie 1993). It might also be interesting to analyse the differences of translations caused by different original languages.

404

Larissa Wunderlich

References

Corpus: Treaty establishing the European Coal and Steel Community, 18.4.1951, Paris, (ECSC) Treaty establishing the European Community, 25.3.1957, Rome, (EC) Treaty establishing the European Atomic Energy Community, 25.3.1957, Rome, (EAEC) Joint Declaration by the European Parliament, the Council and the Commission on fundamental rights, of 5 April 1977, Luxembourg, (DFR) Joint Declaration by the European Parliament, the Council, the representatives of the Member States, meeting within the Council, and the Commission against racism and xenophobia, of 11 June 1986, Strasbourg, (DRX) Treaty establishing the European Union, 7.2.1992, Maastricht, (EU)

Literature: Austin, J.L. (1962, 1975): How to Do Things with Words. The William James Lectures delivered at Harvard University in 1955. 2nd ed. by J.O. Urmson and M. Sbisä. - Oxford, New York: Oxford University Press. Blumenthal, P. (1997): Sprachvergleich deutsch-französisch. 2., neu bearbeitete und ergänzte Aufl. Tübingen: Niemeyer. Hoffmann, L. (1998): Fachtextsorten der Institutionensprachen III: Verträge. - In: L. Hoffmann, H. Kalverkämper, H.E. Wiegand (eds.) Fachsprachen/ Languages for Special Purposes. Vol. 1, 533539. Berlin, New York: Walter de Gruyter. Labrie, N. (1993): La construction linguistique de la Communaute europeenne. - Paris: Honor6 Champion Editeur. Searle, J. R. (1969): Speech Acts. An essay in the philosophy of language. - Cambridge: Cambridge University Press. - (1979): A taxonomy of illocutionary acts. - In: J.R. Searle, Expression and Meaning. Studies in the Theory of Speech Acts, 1-29. Cambridge: Cambridge University Press. Weigand, E. (1989): Sprache als Dialog: Sprechakttaxonomie und kommunikative Grammatik. Tübingen: Niemeyer. - (1997): Was macht eine Übersetzung zu einer guten Übersetzung? - In: E. Pietri (ed.) Dialogue Analysis V, 155-173. Tübingen: Niemeyer. - (1998a): Contrastive Lexical Semantics. - In: Ε. Weigand (ed.) Contrastive Lexical Semantics, 2544. Amsterdam, Philadelphia: Benjamins. - (1998b): The Vocabulary of Emotion. Α contrastive analysis of ANGER in German, English, and Italian. - In: E. Weigand (ed.) Contrastive Lexical Semantics, 45-66. Amsterdam, Philadelphia: Benjamins. - (2000): The dialogic action game. - In: Μ. Coulthard, J. Cotterrill, F. Rock (eds.) Dialogue Analysis VII: Working with Dialogue, 1-18. Tübingen: Niemeyer. - (2002): Lexical units and syntactic structures: Words, phrases, and utterances considered from a comparative viewpoint. - In: C. Gruaz (ed.) Quand le mot fait signe. Pour une semiotique de l'ecrit, 129-147. Rouen: Publications de l'Universite de Rouen.