Dialogue Analysis VII: Working with Dialogue: Selected Papers from the 7th IADA Conference, Birmingham 1999 [Reprint 2017 ed.] 9783110941265

Citation preview

Beiträge zur Dialogforschung

Band 22

Herausgegeben von Franz Hundsnurscher und Edda Weigand

Dialogue Analysis VII: Working with Dialogue Selected Papers from the 7th IADA Conference Birmingham 1999

Edited by Malcolm Coulthard, Janet Cotterill and Frances Rock

Max Niemeyer Verlag Tübingen 2000

Die Deutsche Bibliothek - CIP-Einheitsaufnahme Dialoganalyse: Referate der ... Arbeitstagung = Dialogue analysis. - 1 (1986)-...-Tübingen : Niemeyer, 1986 - . . . Erscheint unregelmäßig. Bibliographische Deskription nach 7 (2000) 7 (2000) u.d.T.: Dialogue analysis NE: International Association for Dialogue Analysis 7. Working with dialogue : selected papers from the 7. IADA conference Birmingham 1999. - 2000 (Beiträge zur Dialogforschung; Bd. 22) ISBN 3-484-75022-7

ISSN 0940-5992

© Max Niemeyer Verlag GmbH, Tübingen 2000 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: Guide-Druck GmbH, Tübingen Einband: Nädele Verlags- und Industriebuchbinderei, Nehren

Table of Contents Preface Edda Weigand The Dialogic Action Game

ix

1

Jim Martin Factoring Out Exchange: Types of Structure

19

Michael Toolan Towards a Simple Schema of Speech Moves

41

Frances Rock Exploring Contemporary Method Through Historical Data: Pragma-Discoursal Dialogue Annotation

53

Peter White Dialogue and Inter-Subjectivity: Reinterpreting the Semantics of Modality and Hedging 67 Barry Natusch Talking for a Living on Air

81

Arja Piirainen-Marsh Interjections and the Institutionality of Broadcast Talk

93

Gerda Lauerbach Negotiating Defeat in TV Election Night Coverage

107

Martin Montgomery Televised Talk: Face Work, Politeness and Laughter in The Mrs Merton Show

121

Birgit Meerholz-Haerle "Listen to this one...": Teachers Sharing Stories and Shaping Policies

137

Polly Walsh Investigating Interactional Variation in Interview-Type Speaking Tests

147

vi

Julia Bamford Question and Answer Sequencing in Academic Lectures

159

Luis Pérez-Gonzàlez Agenda Negotiation in Emergency Calls: Modelling the Dynamics of Frictional Encounters

171

Boris Pritchard and Damir Kalogjera On Some Features of Conversation in Maritime VHF Communication

185

Almut Koester Getting Things Done and Getting Along in the Office

197

Agnès Witko et Michèle Grosjean Les Réunions en Contexte Institutionnel et Leur Formalité

209

Joanna Channell Working on the Telephone - How Telephone Receptionists Work With Language

221

Julia Gillen "It's not really time for going home": Three- and Four-Year Old Children Learning to Talk on the Telephone 231 Christian Hudelot et Christine Berger Les Voix du Dialogue: Usage de la 'Communication Facilitée' avec des Enfants Présentant des Troubles Autistiques

247

Michèle Grossen Therapist-Child Dialogues in Clinical Interviews

257

Alison Ferguson Understanding Paragrammatism: Contributions from Conversation Analysis and Systemic Functional Linguistics

271

Branca Telles Ribeiro Listening to Narratives in Psychiatric Interviews

283

Klaus P. Schneider Diminutives in Discourse: Sequential Aspects of Diminutive Use in Spoken Interaction

293

vii

Carol Marley Trading Talk? Interaction in Written Dating Advertisements

301

Fathi Dali Code-Switching and Dialogism in Lolita

315

Murray Knowles 'Suffering is Good for the Soul': Dialogue and Ideology in Traditional Juvenile Fiction

325

Dorota Pacek Wordplay and Nonsense in Polish Translations of Dialogues in Alice's Adventures in Wonderland.

335

Josiane Boutonnet Verbal Humour as the Ultimate Level in Language Learning: Accessing the Contextual Properties of Humour in Everyday Conversation 347 Susan Mandala Talk in the Mind: Scripted Dialogues and Mental Scripts

357

Rosamund Moon Lexical Norms in Spoken English

371

Charles Owen Naturally-Occurring and Artefactual Dialogue

387

Janet Cotterill Multiple Voices: Monologue and Dialogue in the O. J. Simpson Criminal Courtroom

403

Malcolm Coulthard Suppressed Dialogue in a Confession Statement

417

Ron Scollon Hidden Dialogicality: When Infelicity Becomes Fear of Infringement

425

List of Contributors

441

Preface The papers in this volume are a refereed and representative selection from almost one hundred which were presented at the 7th Conference of the International Association for Dialogue Analysis, (IADA), held at the University of Birmingham, England, in April 1999. The resulting publication is a set of papers which are both interesting in their own right and stimulating as a collection. The volume includes both theoretical papers on the analysis of dialogue and studies of dialogue in work-settings, a binary focus encapsulated in the title, Working with Dialogue. It brings together contributors from 30 universities in 22 countries and the papers cover a wide range of theoretical and descriptive standpoints. The settings from which the spoken data has been collected are diverse: media and maritime discourse, the courtroom, the classroom, the home and the clinic. In addition, several papers present new methods for analysing dialogue, while some focus on interaction as represented in written texts. The book is ordered in such a way that each paper links either theoretically, methodologically or topically with those on either side of it. Thus, papers may be read individually, or considered in the context of those which precede and follow. Readers are therefore encouraged to dip into papers they find of immediate interest and then to move on to adjacent papers for coverage of related issues. We hope you find this book as enjoyable to read and use as we did to compile.

Malcolm Coulthard Janet Cotterill Frances Rock

Birmingham November 1999

Edda Weigand The Dialogic Action Game 1. Dialogue Analysis in the New Millennium After nearly a century of modern linguistics the time has come for us to reflect on what we are doing, to think about the state of the art in dialogue analysis. What is striking is the multiplicity and variety of models which surround us. It seems as if every possibility has already been tested. There is the totally rule-governed model which considers dialogue to be a calculable activity, and there is the opposite, the approach which rejects completely rules and units. There are approaches which take empirical data as their starting point and others which are based on deduction. We have different terms at our disposal for identifying our subject matter: dialogue, conversation, discourse, language use, social interaction, etc. On the one hand, this ramification of models which address, roughly speaking, the same question, is to be appreciated because every model takes a different perspective into consideration and thus helps to enlighten the whole. On the other hand, we have been studying the details of different models long enough now to be in a position to have an overview of this complex area and to be able to check the relevant steps and elements. I have called the approach I am proposing 'the dialogic action game'. Since Wittgenstein (1958) the term action game has had some attraction for linguists because it seems to cover relevant features of language use. The usual practice when using the term however has been to do so but without giving a precise justification. All is left in a fascinatingly vague domain. On the other hand, it is not only formal analysts who have tried to fix the term by comparing the action game of language use with a chess game. Even Clark (1996) in his multi-faceted study on 'Using language' repeatedly refers to the chess game when trying to elaborate essential features of language use as a joint activity. This comparison however marks, in my opinion, the crucial point we have to think about as we enter the new millennium: Are we really playing chess when we use language? Before basing our approach on such a hypothesis we should ask what the conditions of playing chess are and whether they obtain in language use. Chess is a game which is totally rule-governed and a game in which white has the same rules and possibilities of moves as black and differs only in the right to make the first move. The object 'chess' can be considered to be identical with its rules. In language use, however, we clearly have to distinguish between object and methodology. If we consider language use to be analogous to a game of chess, we

2

Edda Weigand

are starting with rule-governed methodology without knowing what our object is. The object, whatever it is, is forced to fit the methodology and is thus transformed into an artificial one, whereas Martinet (1975: 10), among others, has already told us 'not to sacrifice the integrity of our object to methodological exigencies'. The title Working with Dialogue reminds us of the necessity to be cautious in not restricting our object by our methodology. I would therefore like to open up the discussion and take the question of methodology even beyond Aristotelian limits of clear-cut closed systems and to recognise a long overdue turning point in linguistic methodology. Even if it has already occasionally been emphasized, it still needs to be brought to general attention that our starting point is our object of study and that an adequate methodology has to be derived from it. The question Are we playing chess in dialogue? might seem provocative for scholars of systematic linguistics and for deductive pragmaticians. It is however a question which every linguist who has not lost their sound and natural feeling for language is concerned about. In this way, Martinet (1975: 9) even if he is still rooted in sign theoretical linguistics, criticizes generative linguistics for giving priority to logico-mathematical methodology. In a similar way, Moore/Carling (1982) criticize the generative approach for having lost sight of its object. When we start to develop a new methodology we should, however, be cautious and not exchange one extreme view for another by substituting total rule-governedness for chaos. What is necessary for the new millennium has already been requested by Austin (1962: 147) we have to look in a holistic way at the whole of our complex object. In Austin's day this was the speech act. It is now the dialogic action game. We have to take up the claim of complexity and to try to cope with it.

2. Progress in Dialogue Analysis The history of Dialogue Analysis is characterized not only by a multitude of models but even more so by sudden leaps and by the juxtaposition of contradictory hypotheses. It is impossible here to give a detailed picture of the state of the art; instead I will take these leaps as guidelines, because it is they which demonstrate, as I see it, how progress has been made in our research process.

2.1. Empirical Evidence There is first the important step of shifting our interest towards the authentic text which we owe to Conversational Analysis, as represented, for example, in the work of Sacks/Schegloff/Jefferson (1978). Focusing on our empirical basis was indeed a necessary step after the artificial caprices of the various models of generative grammar. Perhaps it is due to this jump from generative supposition to conversational reality that the role of empirical data has been somewhat exaggerated in two regards. First, we have to ask whether there is really something

The Dialogic Action Game

3

like empirical evidence. Evidence results from applying a model to empirical data. In the end, 'we do not know what reality is independent of a theory' (Hawking 1993: 44). Second, it is not even the authentic text, not even the corpus which can be considered our object. Language use as a human activity comprehends more than the elements which are registered on the empirical level. Everyone trying to find authentic examples in a corpus knows the difficulty of understanding as an observer what is going on in these discourses. Starting from the authentic text, Conversational Analysis tried to find rules underlying the empirical level. Even if Sacks/Schegloff/Jefferson offer a groundbreaking account for rules of turn taking, they are restricted by their own program to the continuously varying data on the empirical level and cannot tackle the problem of the relation between turn taking and action sequences. 2.2. Structure on an Empirical Base The Birmingham School of Discourse Analysis keeps to the performance line of the authentic text but there is a dominant intention to include the level of meaning and to build up functional structures of discourse. Sinclair and Coulthard in their pioneering work Towards an Analysis of Discourse. The English used by teachers and pupils (1975), very early on pose the right questions about the action functions of utterances and their appropriate sequencing. The methodology used however is still a structural one which segments and classifies the sequence of utterances in a rank system of categories. What we take with us as an important step of progress is the fact that the necessity and possibility of finding structures for so-called 'real examples' with all their performance features is convincingly demonstrated. 2.3. Priority to the Model The model of Dialogue Grammar programmatically developed by Hundsnurscher (1980) in his paper 'Konversationsanalyse versus Dialoggrammatik' changes the direction of research and gives priority to the model. In Dialogue Grammar, the level of performance and the level of communicative competence are clearly separated. The object of study, communicative competence, is restricted to the verbal level of well-formed utterance forms. The fact that we can come to an understanding in dialogue has led to the assumption that common rules exist. Coming to an understanding in language use however is much more complex than being exclusively based on rules. We make efforts in negotiating our positions in dialogue, because we cannot presuppose an equal communicative competence on the part of both speaker and hearer. What first was considered fascinating and a great merit of Dialogue Grammar, that generative methodological principles were re-used in dialogue analysis, has turned out to be the real obstacle we now have to

4

Edda Weigand

overcome on entering the next millennium. With the pattern view of Dialogue Grammar which considers language use as a kind of chess game we deconstruct our object in order to adapt it to the requirements of a pre-given methodology. This pre-given methodology in the end turns out to be a structural one of segmenting sequences and classifying moves. When trying to understand first and foremost our object, language, we have to free ourselves from traditional assumptions which have dominated our view of language over two millennia, assumptions which arise from rational or conventional order in closed systems. How could we assume that our communicative competence would be guided exclusively by rules and conventions, or that we could separate our abilities, a linguistic-verbal competence from other competences, such as to think and to perceive? How could we neglect the fact that it is always different human beings co-operating and negotiating in dialogue? What we take from Dialogue Grammar or from similar models such as Stati's 'II dialogo' is the emphasis on communicative purposes as a key concept and on the general communicative purpose of coming to an understanding, even if we now recognize that there are more ways than the one of presupposed understanding referred to in Dialogue Grammar. 2.4. Methodology at the Cross-Roads Despite some important insights gained by focusing on rules and well-formedness, obstacles appeared which could not be overcome but only excluded. We find the same problem with all rule-governed methodologies, not only in pragmatics. There are in principle three ways of addressing this problem: first, to insist on the pattern methodology and to exclude disturbing phenomena such as problematic understanding and emotion; second, to totally reject the rules-and-units model; and third, to open up the model. The first possibility, to insist on rule-governed methodology, demonstrates the limits of the model and changes the object inevitably into an artificial one. Natural language use does not coincide with the well-defined objects of rulegoverned models. Generative models, not only in syntax and semantics, severely damage the integrity of their object, language and languages. In speech act theory, we find the same insistence on abstract models which have little to do with language use as a natural phenomenon (e.g. Tsohatzidis 1994, Weigand 1996). In the same way orthodox Dialogue Grammar is in danger of losing its object. The second possibility, to totally reject rules and units in dialogue analysis, is followed, for instance, by Taylor/Cameron (1987). It has to be considered as going from one extreme to the other and is, in its categorical negation, not acceptable. Nevertheless, it represents an important step towards problematizing the closed system of patterns which is supposed to underlie our communicative

The Dialogic Action Game

5

competence. Substantial constructive proposals for how to deal with the problem are however missing. We find the same situation in semantics with Baker/Hacker's (1984) excellent study on 'Language: sense and nonsense' which makes a devastating critique of all mathematical, calculable, compositional theories; again, any positive indication of how to overcome the bankruptcy in semantics is missing. Searle (1992), in his book on conversation, gives up his attempt at finding rules and units of conversation, resignedly admitting that this is an unsatisfactory result. Thus we are faced with the third possibility, the necessity of opening up the model. It has become manifest that the direction cannot be from methodology to the object but has to be the reverse: we must try to understand our object and to develop a model which is appropriate for it. Some steps have already been taken and we can build on them. I can refer only to a few, for instance, to the study by Moore/Carling 'Understanding language: towards a post-Chomskyan linguistics' which appeared in 1982 and emphasized the point that our starting point is language use and not generative methods. Moore/Carling bring in the interesting concept of 'emergent meaning' which corresponds to the similar concept of 'emergent grammar' by Hopper (1987). We find it again in the so-called 'emergent or opportunistic view' by Clark (1996) in his study on 'Using language'. The 'emergent view' is opposed to the 'goals-and-plans view' and holds that conversations are 'purposive but unplanned' (319). The study by Clark embraces social interaction as a multitude of variables offering the term 'ensemble' as an alternative to the term 'action game' (4). This multitude of variables is dealt with as a variety of different aspects which, however, are not brought into a consistent whole. Describing language use mainly as 'accumulation of the common ground', Clark remains too much on an informational level and neglects the action-theoretic point of negotiating different positions. An approach similar to the emergent view is taken by Coulthard (1985: 145) in the second edition of his 'Introduction to Discourse Analysis' in so far as in the end he leaves everything to the individual choice of the speaker. The 'gap between speaker and hearer' is stressed in an interesting recent study by Brown (1995: 24) which problematizes the notion of understanding although Brown also considers dialogue mainly as the exchange of information. Harris (1997: 253) focuses once again on the crucial point by characterizing "Chomsky and his fellow-travellers" as "doing cloud-cuckoo-land linguistics" and demanding an integrational approach which requires the integration of verbal and non-verbal activities. Finally, it was Grice (1975) and Dascal (1994), who emphasized the essential point of nonconventional meaning and of the system's open-endedness. As a consequence, inferences and suggestions have to be considered as communicative means.

6

Edda Weigand

2.5. The Model of the Action Game When we focus on communicative aspects of this kind, in the final analysis we go beyond the type of methodology that has come down to us from antiquity. We must be prepared to accept our object in all its complexity and to develop a new methodology which does not damage the object's integrity. Such a methodology will accept rules, conventions and units but will take into account that they are only of partial and restricted relevance. In language use we behave and interact in cultural units which I call 'action games'. They correspond to a certain degree to similar concepts such as that of 'Lebenswelten/life-worlds' by Habermas (1981) or of 'Lebensformen/forms of life' by Wittgenstein (1958). At the centre of these units there are human beings with human characteristics, i.e. in principle different human beings with different cognitive backgrounds and different personal experiences, which inevitably imply different understandings. Difficulties in understanding, misunderstanding and non-understanding occur and have to be dealt with in the theory. Meaning and understanding cannot always be matched appropriately, inevitably open questions will remain, and, consequently, the theory has to be designed as an open system. The action game therefore cannot be compared with a chess game, but nevertheless it is a game, a game with a purpose. There are not only rule-governed games, there are also creative games, games which are in part rule-governed but open as a whole. I use the term action game in this open, variable sense in which Wittgenstein also introduced his term 'language games'. However I do not agree with him in assuming infinite games, only infinite ways of playing the game.

3. How to Describe the Action Game Having discovered essential features of our object, language use, we now have to tackle the question of how to describe the Action Game. The key to developing an adequate methodology lies in the statement that our starting point has to be human beings and it is their abilities which determine the ways they interact with each other. From this central premise we can derive other fundamental premises which constitute the basis of the theory. These fundamental premises on the nature of our object of study will then confront us with the question of how to behave under these conditions, i.e. with the question of the methodological techniques to be used. Here I will introduce besides rules and conventions the methodological technique of principles which determine social interaction as methods of orientation. The theory of the dialogic action game in its action-theoretical part thus consists of a set of fundamental premises on our object and a set of principles about an appropriate methodology.

The Dialogic Action Game

7

3.1. Fundamental Premises on the Nature of the Phenomenon Systematizing the points already mentioned, we arrive at a set of ten fundamental premises which characterize our object: 1. 2. 3. 4.

5.

6.

7.

8. 9.

10.

Language is used by human beings and cannot be separated from them. Human beings are oriented towards purposes or needs. Purposes therefore are a key concept for explaining human behaviour. Human beings are always different human beings and do not only take different roles as speaker and interlocutor. Human beings are social beings. They use their language or communicate in order to come to an understanding with other human beings, i.e. they have to correlate and to negotiate their positions, tasks and interests. Language use accepts misunderstanding. There are different ways of coming to an understanding, not only by understanding but also by correcting misunderstanding and clarifying difficulties in understanding and non-understanding. For human beings there is no independent world, no reality as such, only a world as perceived by them. From the start, human beings are within the world, not opposed to the world. The minimal communicatively autonomous unit is the action game, a unit of our cultural world which comprises the different communicative worlds of the interlocutors. It is the unit in which the components co-operate and function. The action game is not constituted as a type of situation but determined by its interactive purpose. The authentic text represents a component in the action game. Human beings use different abilities together as communicative means. They produce verbal texts and simultaneously, and not separably, they draw inferences and rely on what can be perceived in the speech situation. Language use therefore can in most cases be described only incompletely from the observer perspective and is only in part represented by corpora of authentic texts. Not everything is said explicitly, indeed, not everything can be said explicitly because of the complexity of meaning to be negotiated.

3.2. Basic Methodological Principles of H o w to Behave in Dialogic Action Consequently, communicative competence cannot be considered a closed system of rules and patterns. We negotiate via conventions and rational expectations as well as via moment-by-moment judgments, ad hoc associations and presuppositions, non-conventional suggestions and presumptions. We must therefore think of our action competence as an open system mainly based on principles as techniques of orientation. Before specifying more concretely basic Principles of Dialogue I would like to distinguish between the methodological techniques of rules, conventions and principles as I am using these terms. In order to come to an understanding in language use we need some techniques which allow intersubjective negotiation and convergence. In my opinion, different possibilities are used for different

8

Edda Weigand

cases: we can structure the multiplicity of verbal means according to rules as we do in morphology, or we can use conventions as we do in communicative grammar with lexical expressions. Rules are a technique which is speaker-independent and presupposes more than one case, i.e. more rule-governed cases than exceptions. In contrast to this, conventions are dependent on speaker groups and can also refer to single cases. Principles on the other hand refer to complex holistic processes which are in need of techniques of orientation and cannot be dealt with only by rules and conventions. The meaning to be negotiated in dialogic action comprises infinite reference points and is neither completely nor explicitly expressed. We have to give up those favourite ideals which we have cherished so long and have to tackle the problem of an open, continually changing communicative universe. Principles or maxims of behaviour are the best means of orientation in such a complex whole. They are flexible enough to guarantee highly effective standards of language use. The term 'Principle' is used in the literature in various ways ranging from mathematical axioms to maxims of orientation. I will use it as a technique of orientation in complex and open surroundings. There are various Principles of Dialogic Action which we know in part already from the literature. These principles are inter-connected, i.e. some of them are basic: the Action Principle, the Dialogic Principle proper, and the Coherence Principle, the others are derived or corollary ones. Human beings have developed language because they need a means to talk to each other and to co-ordinate their points of view. In contrast to other actions, communicative actions are oriented towards another interlocutor, real or fictitious, or towards the speaker themself. Single utterances therefore are not autonomous, they are either initiative or reactive. It is not just the position in the sequence that makes an utterance either initiative or reactive, it is the claim of the utterance itself. With an initiative utterance the speaker makes a claim which relates to the interlocutor, with the reactive utterance the interlocutor is expected to take up this very claim and to fulfil it by accepting or rejecting it or by negotiating the conditions. The communicative purposes are based on these pragmatic claims which are essentially a claim to truth and a claim to volition corresponding to the two basic mental states of human beings, belief and desire. Dialogic actions are thus rooted in our cognitive demands from which all types of action can be derived as dialogically oriented actions (Weigand 1989, 1991) and all types of action games can be distinguished according to their interactive purposes. It was a specific type of conversation which prevented us from recognizing the general validity of this principle: small talk. Small talk, however, is not started just by mood, it is started because we feel somehow obliged to talk with the other. Establishing or confirming social relationships represents the communicative

9

The Dialogic Action Game

purpose of small talk. We can thus generalize for every action game and for every utterance that there is some dialogic purpose to be negotiated. The Action Principle states that taking communicative actions means pursuing specific dialogic purposes with specific dialogic means. (Fig.l)

dialogic purposes (state o f affairs) < — > making and fulfilling pragmatic claims

dialogic means verbal, perceptible, cognitive means

As already mentioned, dialogic means are verbal, perceptible and cognitive means. Speaking is always accompanied by thinking which goes far beyond what is expressed in the verbal utterance. We have to presuppose many things which we cannot express explicitly, otherwise we would never come to the end of our conversation. We draw conclusions of various types, rational and conventional ones as well as moment-by-moment judgments which inherently imply the risk of problems of understanding. Now why are we taking the concept of action as a principle, i.e. interpreting the arrow in Figure 1 not as what is definitely the case but as a principle of orientation? Actions are not given as such on the basis of fixed rules, they are negotiated by the interlocutors in specific circumstances. First, we have to take utterances as directed to someone and not only as locutionary exercises. Second, verbal means, the means we produce when speaking and we register when listening are only part of what constitutes action. As long as we neglect cognitive means, as we do in Dialogue Grammar and in Corpus Analysis, we cannot fully understand our competence in language use. To give an example: A mother entering the room in which her daughter is playing the piano says: (1)

You are playing the piano again.

From the stance as an observer which we take in corpus analysis, it is quite impossible for us to decide whether this utterance is a pure statement or an evaluative speech act of reproach or praise. In so far as intonation is usually not decisive, not even the daughter might know exactly what her mother has meant. It depends on the one hand on general cultural evaluations of piano playing, and on the other, on particular conditions of the individual situation. The mother might in general be pleased that her daughter is practising, but at this moment she might feel disturbed. So in the end, only the speaker knows how the utterance is meant. Consequently, a theory of using language cannot be restricted to generalizing rules only. Having clarified what action means, we now have to ask what dialogue means. The crucial point has already been emphasized by Austin: the utterance has to be taken up. What 'take up' means exactly, Austin does not tell us. In my

10

Edda

Weigand

opinion, it cannot be restricted to meaning and understanding one and the same utterance. Having understood the utterance the interlocutor is meant to react. It is action and reaction which constitute a dialogue. In so far as the interlocutors are different human beings, understanding what the speaker meant cannot be presupposed. Dialogue in general represents a process of negotiating between different positions (cf. Adamzik 1995). Even if - as we will see - the dialogic sequence is not calculable in advance but emerges, there is some rational basis which determines the internal interdependence of the initiative and the expected reactive action which constitutes the Dialogic Principle proper. This rational basis lies in the very characteristics of the initiative speech act itself, in its pragmatic claim which is to be negotiated in dialogue. It is, for instance, rationally defined by the characteristics of the Representative itself as a speech act which makes a claim to truth that a reaction is expected which will take up this claim, accept it or reject it, or clarify its conditions. In the same way, it is rationally defined by the characteristics of the Directive as a speech act which makes a claim to volition that a reaction of Consent is expected. This internal interdependence of action and reaction is constitutive for every type of action and results in a dialogic speech act taxonomy which considers speech acts to be components of minimal dialogues (cf. Weigand 1989, 1991). Thus we arrive at the following global speech act types: (Fig. 2)

Representative Directive Explorative Declarative





± ± ± [+

Acceptance Consent Response Confirmation]

It is, in my opinion, important to recognize that the internal interdependence of action and reaction represents a functional principle and is not just gained in a structural way by segmenting sequences and classifying moves. Accepting the Dialogic Principle proper for language use, we have to abandon the philosophical perspective of speech acts as autonomous acts. There is not only one type of action function, the illocutionary. Speech acts are either initiative, when making a claim, or reactive, when taking up this very claim. By differentiating the fundamental claims to truth and volition we achieve multiple derived sub-types of these global types of speech acts. Recently, Clark (1996) also used a dialogic speech act typology which, however, was not theoretically based but gained by generalizing intuitively the principle of adjacency pairs. In so far as the DP proper is based on rationality one might consider the interdependence of initiative and reactive speech acts to be a rule. However the relationships are not given, they represent expectancies and are dependent on how the individual interlocutors take them. The crucial point consists in the fact that new ways of interaction can always emerge. If we try to describe human behaviour

11

The Dialogic Action Game

systematically, we have to accept open points in the system, points that are not calculable, that necessarily emerge from time to time, mainly for the two reasons mentioned in our premises: first, speaker's world does not coincide with hearer's world; second, not everything can be explicitly expressed. Let us consider the following example: (2.1) (2.2) (2.3)

Tony Doris Tony

I want to hear the news. Keep the door open. That's not possible. Our neighbour is at home. Keep the door open. I am coming.

If it were contained in a corpus, it would be difficult for an observer to understand what is really going on because much more is going on than is expressed. Examples from corpora are very simple ones, like the exchange analysed in detail by Clark (1996: 221) which can be completely understood on the verbal level: (3.1) (3.2) (3.3) (3.4)

Roger Nina Roger Nina

now, - um do you and your husband have aj-car - have a car? yeah no -

Our example (2) however is quite different and can be understood only from inside the action game. What is needed is a precise description of what the interlocutors perceive and of what they think. Referring only to what is said clearly misses the point. Tony is preparing a meal in the kitchen. Doris is in the living room next to it. Both rooms are connected by a door which is open in this moment. Doris refers to Tony's habit of listening the television news from the kitchen through the open door. The television then has to be very loud, which is possible only if the neighbour is not at home. Habits however are not valid for every case. In our example, however, Tony wants to carry the meal into the living room without having to open the door and to listen to the news in there. What becomes clear from this example is that not everything is explicitly said. In order to avoid the risk of misunderstanding, Tony would have had to explain why he wants the door open. Because of the complexity of meaning which surrounds us, it is impossible always to think of every piece of information that is necessary for clear understanding. Also for time-economical reasons not everything can be expressed. The risk of misunderstanding therefore is inevitably included in our way of using language. By accepting it, language use becomes as highly effective as is needed. It can be accepted because misunderstanding will mostly be immediately corrected (cf. Weigand 1999). The type of misunderstanding we are confronted with in example (2) arises not because of the fact that the Dialogic Principle has been misunderstood but rather because both interlocutors refer to different habits and the habit actually meant is not explicitly expressed. There is no ambiguity, only points which are

12

Edda

Weigand

open already in the system and can be closed differently by the interlocutors using different cognitive means. Let us take another example: (4.1) (4.2) (4.3) (4.4)

Mother Child Mother Child

Can you hold the ball? {gives the ball to the girl) And you keep the flower. Do you want it? No, you keep it.

Mother and child are on a walk. The mother carries both the ball the child wanted to take along and the flower the child picked up during the walk. The mother misunderstands the child's utterance (4.2) as a reproach because she is thinking about the child's preference for flowers. In this case however the preference is not valid, as the child's paying attention to other things demonstrates (4.4). The point in which example (4) differs from example (2) lies in the fact that utterance (4.2) can be seen as ambiguous if it is taken only on the verbal level. As was the case with example (1), without decisive intonation, it is not clear whether it is only a statement or an evaluative speech act. Preferences, like habits, are concepts which inevitably carry the risk of misunderstanding. The interlocutors have to negotiate meaning and understanding not in the sense of a 'joint construal of what the speaker is to be taken to mean', as Clark (1996: 21) assumes, but in order to arrive at what the speaker originally meant. From this view of Dialogic Action arises a third basic principle, the Coherence Principle, which covers a complex phenomenon expressed by different terms in the literature: coherence, relevance, co-operation. I will try to demonstrate that in language use coherence is not a relation between different pieces of text, but a principle used by the interlocutors, namely the joint attempt by the interlocutors to understand and to give sense to what is going on. This can be seen quite clearly with the well-known example from Brown/Yule (1983: 196): (5.1) (5.2)

There's the doorbell. I'm in the bath.

Nothing in the text relates the two utterances; nevertheless they are related. Addressing the interlocutor with an utterance means presupposing relevance and co-operation. The initiative utterance can be understood as a question: who is going to open the door, as well as a directive speech act: could you please open the door. There is no definite decision about what is right and this is what characterizes speaking indirectly. The same way of speaking indirectly can be seen in the utterance of the interlocutor. It is not a clear rejection of the claim to open the door but a statement that it would be difficult to do so, which can be concluded on the basis of our knowledge of everyday habits. So both interlocutors are

13

The Dialogic Action Game

communicating mainly on the basis of the DP, making claims and fulfilling them, without explicitly expressing their purpose but relying on cognitive means. Coherence is constituted by the joint attempt to understand and to negotiate what is to be done in the current situation. Let us take another authentic example: (6.1) (6.2)

Mother Father

One for her. Eve, too.

Again the talk exchange on the verbal level is not complete enough to allow a nonparticipant to understand what is going on in the action game. Contained in a corpus we would not understand the exchange as observer, even if it was neither indirect like our example analysed above nor ambiguous. On the contrary, what is expressed is very clear. However one has to know the situation and to perceive the actions accompanying the exchange in order to know what is meant. A family father and mother and two young children, a boy and a girl - is out for a walk around a lake with willows on the bank. The father cuts a willow branch and gives it to the boy. In utterance (6.1) the mother asks for one for the girl. In (6.2) the father agrees and cuts a branch for the girl, too. Two points become very clear, which have also been elaborated in recent research, for instance, by Givon (1993): First, coherence is established beyond the rules of grammar, according to which the personal pronoun would have to refer back to a preceding noun. Second, we notice again that coherence is not a feature of the verbal text. In contrast to the basic assumption of Corpus Analysis (cf. Sinclair 1994: 19), the text will not reveal itself to us. Coherence is established by the interlocutors in the joint effort to understand what is said in the complex framework of the action game. 3.3. Corollary Principles of Dialogic Action Beside these three basic principles there are various other dialogic principles which can be considered corollary ones. In cases of difficulties of understanding, we might try to understand what the speaker meant by adhering to the Principle of Rationality, of Suggestion or of Convention. Thus we have seen that it is in the end by way of rational conclusion that we expect a specific reaction to follow a specific initiative speech act. It is by way of suggestion that we try to find the right habit or preference on which an utterance may be based, and it is by way of convention that we use certain phrases such as how are you? in small talk. Longer dialogues are developed via Sequencing Principles by which the conditions for taking up or modifying a claim are negotiated. As I see it, there are three main types of Sequencing Principles: principles of insisting, of clarifying

14

Edda Weigand

open or unclear conditions and of preparing the situation by so-called presequences before making the claim. In connection with Sequencing Principles, there is the old and always new question whether the sequence is structured according to two-part or three-part chains (cf. e.g. Sinclair 1992: 85). In my view, the Dialogic Principle is the basic one and determines a structure of two-part chains, i.e. every three-part chain can be re-structured as two two-part chains. On the one hand, this is also taken for granted by Clark (1996: 207): 'The second part of one adjacency pair is almost invariably the first part of a second one.' On the other hand, however, Clark stresses the point that 'every joint action has three parts' (p. 201) which he does not only understand in the trivial sense that every joint action has an entry, a body, and an exit (p. 331), but which he also uses for his analyses of speech act sequences. I take two of his examples (p. 208): (7) (8)

Here's your bag. - Thanks. - No problem. I'll have cake. - Here.-Thanks.

Example (7) starts with an utterance which hardly can be considered as initiative. It might be a response to a question like Where is my bag? or a consent to a request like Could I have my bag? which heed not be verbally expressed: (7')

Where is ray bag? - Here it is. - Thanks. - No problem.

If we also add no problem to (8) we have two sequences which are structured in parallel according to two-part chains: (Fig. 3)

I

1

1

EXPLORATIVE

RESPONSE

DECLARATIVE

CONFIRMATION

Where is my bag?

Here it is.

Thanks.

No problem.

DIRECTIVE

CONSENT

DECLARATIVE

[CONFIRMATION]

I'll have cake.

Here.

Thanks.

[No problem.]

In this way the basic Dialogic Principle of action and reaction is realized in sequences of two-part chains. Every utterance in the sequence which is not in the first nor the last position can have two functions, primarily a reactive one and additionally and subordinately an initiative one. Finally, there is the important field of Rhetorical and Emotional Principles which are an intrinsic part of our use of communicative means (cf. Weigand 1998,

The Dialogic Action Game

15

forthcoming). If we consider texts as verbal components in the action game, we have to leave behind us the traditional view which separates texts into rhetorical and non-rhetorical ones. Even explicating information so that it is easily comprehensible is based on rhetorical principles such as, for instance, to proceed from the general to the particular (cf. Coulthard 1994: 7). In the action game, rhetoric is always present; wherever there is 'meaning', there is 'persuasion' as Burke (1950: 172) told us. We always try more or less to be effective in the process of dialogic negotiation. As was the case with coherence, rhetoric is no longer a feature of texts but Rhetorical Principles are used by the interlocutors in pursuing their goals. These principles are based on different ideological guidelines. Some of us might be convinced by the effectiveness of the Gricean principles (1975) which tell us 'to be perspicuous' and 'to make our contribution as informative as is required'. Others might be of the opinion that we can achieve our goals more easily by indirect 'persuasive' speaking such as in the following example: (9) Daughter Mother

We have to talk about America. I thought I could save, I could earn money by giving lessons and I would contribute the money which I have in my account. Well, I'll think about it. Be quiet, I tend to agree.

The verbal text alone, taken as a piece of a corpus, does not tell us what the object of this negotiation is. Again as observer we need a detailed description of the situation: The daughter is trying to persuade her mother to allow her to take part in a very attractive but also expensive school exchange programme with America. The mother has not yet decided and is more prone to decide for a cheaper programme. The dialogic purpose brought in by the daughter is not explicitly expressed, there is no directive speech act; nevertheless, daughter and mother understand what is going on because they presuppose the same cognitive background which is used as part of the communicative means in this dialogic interaction. The Rhetorical Principle chosen by the daughter is very effective because it is based on positive motivation and collaboration and does not blurt things out too directly. As with Rhetorical Principles, emotions are also an integral part of the action game. Starting with the assumption that we as human beings use different abilities as communicative means and that these abilities cannot be separated, we have to include emotions, too. Emotions have a strong influence on our goals and the way we deal with them in dialogue. In dialogic action we always use, mostly unconsciously, Emotional Principles, for instance, the culturally dependent principle 'Hide your emotions in public speech'. Sometimes, however, emotions are so strong that they lead us to over-react and to turn rational behaviour into the opposite, as in the following example:

Edda

16 (10)

Weigand

He When will you come? She You know, at Easter. He Well, then we will see each other at Easter. I will phone you before then, if I can still remember your number.

The last sarcastic turn from him does not at all coincide with his intention. He loses self-control and behaves in a way which hurts her, because Easter is a long time away, and he is deeply disappointed that she will not visit him sooner, although he knows that it will scarcely be possible for her.

4. Conclusion: Towards a Human Linguistics Having thus described the Dialogic Action Game in two parts: fundamental theoretical Premises and a set of methodological Principles, I think it has become manifest that we are not playing chess in language use, and that we cannot start from a corpus as observers when trying to understand what is going on in the action game. We have to start from human beings and their abilities, from their way of perceiving the world, of having goals, of being oriented towards each other, of always negotiating meaning and understanding. The world is complex and infinite from the eye of different perceivers. There must be some relatively simple principles at our disposal which we can use as guidelines in our dialogic behaviour. It is these principles we are trying to discover and verify using authentic examples. As we begin the new millennium, I hope linguists will finally recognize that their object is neither a logico-mathematical apparatus nor restricted to speaking and listening alone. We have proceeded far enough to leave behind us questions like: Does this belong to linguistics? Isn't it rather an object for psychology? Do linguists have to worry about emotions? Language as a natural phenomenon is used by human beings in dialogic action games. In this sense, language is a kind of human behaviour, not an object of philology nor of natural science. It is not even a corpus which registers only physically given data. The corpus is a highly valuable instrument for verifying a description which remains within the verbal domain, for instance, lexical description. Dialogic actions, however, are a complex phenomenon combining different areas and abilities: language and cognition, meaning and understanding, rules and suggestions, experience and imagination, the rational and the irrational. After the era of philology and rule-governed patterns, at the verge of a new century and even millennium, linguistics consequently has to be defined as a human science which describes and explains what human beings are doing when they try to negotiate their positions in social communities.

The Dialogic Action Game

17

References Adamzik, Kirsten (1995): Dialoganalyse: eine Disziplin auf der Suche nach ihrer Identität. In: Hundsnurscher, Franz/Weigand, Edda (eds.) Future Perspectives of Dialogue Analysis. Tübingen: Niemeyer, 35-78. Austin, John L. (1962): How to Do Things with Words. The William James Lectures Delivered at Harvard University in 1955. London, Oxford, New York: Oxford University Press. Baker, Gordon P./Hacker, Peter M. S. (1984): Language, Sense and Nonsense: a critical investigation into modern theories of language. Oxford: Blackwell. Brown, Gillian (1995): Speakers, Listeners and Communication: explorations in discourse analysis. Cambridge: Cambridge University Press. Brown, Gillian/Yule, George (1983): Discourse Analysis. Cambridge: Cambridge University Press. Burke, Kenneth (1950): A Rhetoric of Motives. New York: Prentice-Hall. Clark, Herbert H. (1996): Using Language. Cambridge: Cambridge University Press. Coulthard, Malcolm (1985): Ar Introduction to Discourse Analysis. London, New York: Longman, 2nd edition. - (1994): On Analysing and Evaluating Written Text. In: Coulthard, Malcolm (ed.) Advances in Written Text Analysis. London, New York: Routledge, 1-11. Dascal, Marcelo (1994): SpeecH Act Theory and Gricean Pragmatics: some differences of detail that make a difference. In: Tsohatzidis, Savas L. (ed.) Foundations of Speech Act Theory: philosophical and linguistic perspectives. London, New York: Academic Press, 323-334. Givón, Talmy (1993): Coherence in Text, Coherence in Mind. Pragmatics and Cognition 1. 171-227. Grice, Herbert Paul (1975): Logic and Conversation. In: Cole, Peter/Morgan, Jerry L. (eds.): Syntax and Semantics. Vol. 3: Speech Acts. New York etc.: Academic Press, 41-58. Habermas, Jürgen (1981): Theorie des kommunikativen Handelns. 2 Vols. Frankfurt/M.: Suhrkamp. Harris, Roy (1997): From an Integrational Point of View. In: Wolf, George/Nigel, Love (eds.) Linguistics Inside Out, Roy Harris and his critics. Amsterdam/Philadelphia: Benjamins, 229310. Hawking, Stephen (1993): Black Holes and Baby Universes and Other Essays. New York etc.: Bantam Books. Hopper, Paul (1987): Emergent Grammar. Berkeley Linguistics Society 13. 139-157. Hundsnurscher, Franz (1980): Konversationsanalyse versus Dialoggrammatik. In: Rupp, Heinz/Roloff, Hans-Gert (eds.) Akten des VI. Internationalen Germanisten-Kongresses. Basel (1980) Part 2. Bern, Frankfurt/M., Las Vegas: Peter Lang, 89-95. Martinet, André (1975): Studies in Functional Syntax. Études de syntaxe fonctionnelle. München: Fink. Moore, Terence/Carling, Christine (1982): Understanding Language: towards a Post-Chomskyan Linguistics. London etc.: Macmillan. Sacks, Harvey/Schegloff, Emanuel A./Jefferson, Gail (1978): A Simplest Systematics for the Organization of Turn Taking for Conversation. In: Schenkein, Jim (ed.): Studies in the Organization of Conversational Interaction. New York, San Francisco, London: Academic Press, 7-55. Searle, John R. et al. (1992): (On) Searle on Conversation. Amsterdam, Philadelphia: Benjamins. Sinclair, John (1992): Priorities in Discourse Analysis. In: Coulthard, Malcolm (ed.) Advances in Spoken Discourse Analysis. London, New York: Routledge, 79-88. - (1994): Trust the Text. In: Coulthard, Malcolm (ed.) Advances in Written Text Analysis. London, New York: Routledge, 12-25. Sinclair, John McH./CouIthard, R. Malcolm (1975): Towards an Analysis of Discourse: the English used by teachers and pupils. Oxford: Oxford University Press. Stati, Sorin (1982): Il dialogo. Considerazioni di linguistica pragmatica. Napoli: Liguori.

18

Edda Weigand

Taylor, Talbot J./Cameron, Deborah (1987): Analysing Conversation: rules and units in the structure of talk. Oxford etc.: Pergamon Press. Tsohatzidis, Savas L. (ed.) (1994): Foundations of Speech Act Theory: philosophical and linguistic perspectives. London, New York: Academic Press. Weigand, Edda (1989): Sprache als Dialog. Sprechakttaxonomie und kommunikative Grammatik. Tübingen: Niemeyer. - (1991): The Dialogic Principles Revisited: speech acts and mental states. In: Stati, Sorin/Weigand, Edda/Hundsnurscher, Franz (eds.): Dialoganalyse III. Vol. I. Tübingen: Niemeyer, 75-104. - (1996): The State of the Art in Speech Act Theory: review article on: Tsohatzidis, Savas L. (ed.): Foundations of Speech Act Theory. Pragmatics and Cognition 4, 367-405. - (1998): Emotions in Dialogue. In: Cmejrkovà, Svétla et. al. (eds.): Dialoganalyse VI. Vol. I. Tübingen: Niemeyer, 35-48. -(1999): Misunderstanding: the Standard Case. Journal of Pragmatics, 31, 763-785. - (1999): Rhetoric and Argumentation in a Dialogic Perspective. In: Rigotti, Eddo (ed.): Rhetoric and Argumentation. Tübingen: Niemeyer, 53-69. Wittgenstein, Ludwig (1958): Philosophical Investigations. Oxford: Blackwell.

Jim Martin

Factoring out Exchange: Types of Structure 1. Orientation In this paper I will address two disjunctions which have arisen in Australian work on dialogue within the framework of systemic functional linguistics (SFL). One is the conative/expressive disjunction which has distinguished studies of interaction in dialogue (move and exchange) from studies of evaluative meaning (appraisal) Ventola (1987) vs Coffin (1997) for example. The other is the pragmatic/casual disjunction which has differentiated studies of interaction in relatively formal institutionalised settings (e.g. service encounters, classroom discourse) from studies of casual conversation (e.g. chat, gossip) - Hasan (1977) vs Eggins and Slade (1997) for example. Here I will try to rework these disjunctions as complementarities, drawing on Halliday's (1979) association of metafunctions with types of structure to do so.

2. Types of Structure Basically Halliday's suggestion is that language has evolved functional resources to construe three orders of reality. Ideational resources construe the world as we think we see it; interpersonal resources construe social relations as we enact them; and textual resources map these construals onto one another as digestible bites of information. According to Halliday, ideational resources are associated with particulate forms of realisation - they naturalise reality as bits and pieces. The particles may be organised orbitally, into configurations consisting of a nucleus, margin and periphery (experiential meaning); or they may be organised serially, into chains of interdependent steps (logical meaning). This orbital/serial complementarity is exemplified below for transitivity and projection respectively (following Halliday (1994), Matthiessen (1995); examples throughout the paper are adapted from the film Educating Rita unless otherwise noted): orbital structure You would have thrown it across the room... MARGIN (AGENT)

| NUCLEUS (PROCESS)...

You

I would have thrown

....(MEDIUM)

I PERIPHERY (LOCATION)

i across the room

Jim Martin

20 serial structure Willie Russell wrote that Rita said that Trish thought the poetry was brilliant. VERBAL PROCESS

" LOCUTION

" LOCUTION

' IDEA

Willie Russell wrote

that Rita said

that Trish thought

the poetry was brilliant

Interpersonal resources on the other hand are associated with prosodic forms of realisation - they enact social reality as splashes of engagement, which saturate their domain. This is exemplified below for English polarity, which establishes the arguability of its interact and at the same time conditions the realisation of any indefinite deixis within its scope (perhaps more strikingly so in the stigmatised non-standard realisation I can'/ bear nothing from no-one no longer)-. prosodic structure I can't bear anything from anyone any longer. ! NEGATIVE...

I

! can't

bear

...NEG.

! ...NEG.

...NEG.

anything

| from anyone

any longer

Finally, textual resources are associated with periodic forms of realisation - they organise semiotic reality as waves of information (the rhythm of discourse). This culminative patterning is exemplified below for Theme and New; Theme choices construct the speaker's angle on his field (in this case someone's poetry), while choices for New elaborate the point of the discourse (in this case the value of the Themes): periodic structure It is brilliant. It's it it's witty; it's profound, full of style. THEME

It It it (it)

1 NEW

i is I's !'s 1 's

! brilliant ! witty 1 profound ! full of style

A summary of these types of structure and their association with modes of meaning (Halliday's metafunctions) is presented as Fig. 1. For further discussion of this reading of Halliday in relation to grammar and discourse see Martin (1995a, 1998).

21

Factoring Out Exchange: Types of Structure

Mode of meaning

Types of structure

particulate

ideational meaning

- orbital [mono-nuclear]

- experiential

- serial [multi-nuclear]

66b

- logical

prosodie

interpersonal meaning

periodic

textual meaning

Fig. 1 : Types of structure in relation to modes of meaning

3. Exchange Structure Berry (1981a, b) proposes a model of exchange structure which draws on Halliday's metafunctions and is organised around layers of interpersonal, textual and ideational structure. Her interpersonal tier, as developed in Ventola (1987) and Martin (1992), will be taken as point of departure here. Ventola's structure potential for exchanges negotiating goods and services is outlined below. Minimally it consists of an obligatory A1 move by the primary actor (who is responsible for giving goods, performing a service, or promising to do so). Additional moves allow for a demanding A2 move by the secondary actor (who is receiving goods or benefiting from services), and if that is present, an additional Dal move by the primary actor delaying the proffer of goods and services until a go-ahead is given by the secondary actor. In addition the structure potential allows for optional follow-up moves by the secondary and primary actors. An example of a five move exchange, initiated by the primary actor, is presented below: ((Dal)

A2)

[D 'delay';

A 'actor';

Dal [Rita]:

Do you want to lend it?

A

Al

A

1 'primary';

(A2f

(Alf))

A

2'secondary'; offer

f'follow-up']

22

Jim Martin A2 A1 A2f Alf

[Frank]: [Rita]: [Frank]: [Rita]:

Ah yes. Here. Yes. Well, thank-you very much. That's okay.

accept proffer thank acknowledge thanks

The structure potential for exchanges negotiating information is parallel, although involving knowers instead of actors. The primary knower is the interactant responsible for adjudicating the polarity and modality of the information at stake; the secondary knower is the receiver of that authority. An example of a five move exchange, initiated by the primary knower, is presented below (this is the 'text' question sequencing of pedagogic discourse and quiz shows, in which primary knowers ask questions they must adjudicate the answer to): A

((Dkl) A K2) A K1 A (K2f (Klf)) [D 'delay'; K 'knower'; 1 'primary'; 2 'secondary'; f'follow-up'] Dkl K2 K1 K2f Klf

[Frank]: [Rita]: [Frank]: [Rita]: [Frank]:

What's assonance then? A form of rhyme? Right. Yeah. Yeah.

'test' question suggest answer validate confirm reconfirm

Developing Berry (1981a, b) and Burton (1981), Ventola and Martin propose two types of potential interruption to the formula presented above. One type involves tracking moves which attempt to clarify the ideational content of a preceding move before developing the exchange; these may occur at any point in the exchange where such clarification is warranted: Dal cl rcl A2 Alf

[Frank]: [Rita]: [Frank]: [Rita]: [Frank]:

Would you like a drink? What of? Whisky. Oh yeah. [NV Frank gets drink]

K2 cl rcl K1

[Frank]: [Rita]: [Frank]: [Rita]:

What is your name? Me first name? Yes.1 Rita.

request clarification clarify

request clarification clarify

The other type involves challenging moves which resist resolution of the exchange, because one of the knowers or actors is uncomfortable with the way the exchange is interpersonally positioning them:

1

Frank actually says, sarcastically, "Well that would at least constitute some sort of start, wouldn't it?"

23

Factoring Out Exchange: Types of Structure

A2 ch

[Rita]: Aren't you supposed to be interviewing me? [Frank]: Do I need to? [Rita]: Oh, I talk too much, don't I? Yes, I know, I talk a lot...

K2 ch rch rrch K1

[Rita]: What does assonance mean? [Frank]: [laughing] What? [Rita]: Don't laugh at me. [Frank]: Ah no. Assonance, it's a form of rhyme.

challenge response to challenge response to rch

Ventola and Martin propose a form of constituency representation for moves generated by the structure potentials outlined above, alongside a form of dependency representation for the more contingent tracking and challenging moves which may frustrate these potentials. For reasons of space, neither this complementarity of structural representations, nor Berry's suggestions for additional ideational and textual layers of structure will be considered further here.

4. Types of Structure in the Exchange - Another Look What is of interest here is the presence of particulate, periodic and prosodic motifs in exchanges such as those outlined above. The presence of a single obligatory move in Berry's formula is indicative of orbital structure, with the K1 or A1 move as nucleus; the optional D/Akl, K/A2, K/A2f and K / A l f moves could then be treated as dependent satellites. A crude representation for this kind of structural configuration is offered below for a Dkl K2 K1 K2f exchange: Dkl K2 K1 K2f

[Frank]: [Rita]: [Frank]: [Rita]:

What's assonance then? A form of rhyme? Right. Yeah.

'test'question suggest answer validate confirm

What's assonance? • A form of rhyme. • Right. - Yeah. Alternatively, this time from the perspective of turn taking, the same exchange could be viewed as a serial chaining structure involving interdependent turns at talk - with interlocutors responding to preceding moves (a structure more akin to Berry's a i A b i A a i i A b i i ... a11 A b n formula).

24

Jim Martin

What's assonance? A form of rhyme. Right. ^

Yeah.

At the same time, a similar exchange might be viewed as a wave of information, with all but one missing piece of information provided in the opening move; the missing information is then supplied in the second move, and confirmed as what the teacher was looking for in the third. Informationally, the amount of ideational content in the moves declines throughout the exchange2, ultimately resolving perhaps in paralinguistic interaction (gaze, facial expression, body language) which has not been included in the analysis. Typically, the next burst of information in pedagogic registers indicates the initiation of another exchange (i.e. a new questioning triad), establishing the periodicity of the discourse: Dkl [Teacher]: K2 [Student]: K1 [Teacher]:

You can see very clearly that she has indicated...? a paragraph? Right.

You can see very clearly that she has indicated - a paragraph - Right Finally, the telos, (the end, purpose or ultimate objective) of the exchange needs to be considered, with respect to the way it is designed to culminate purposefully. Once an exchange is initiated, we know how it is expected to finish - what it's goal is. It is on this basis that we recognise the tracking and challenging moves noted above as interruptions. Thus Frank's K1 move has staying power, scoping over Rita's clarifying move and Frank's response; its purpose stays alive until Rita responds. Because it maps over several moves this teleological aspect of the exchange can be treated as prosodic, establishing the domain through which the exchange unfolds:

2

This pattern is related to Berry's proposition base A proposition completion structure (pb A pc A ps).

A

proposition support

25

Factoring Out Exchange: Types of Structure

K1 cl rcl K2f

[Frank]: [Rita]: [Frank]: [Rita]:

...I don't think I can bear it any longer, Oh, can't bear what Frank? You, my dear. You. Yeah, Yeah.

In summary, it seems possible to factor out 4 complementary structural motifs contributing to Ventola's exchange formula - the orbital motif of obligatory and optional moves, the serial motif of turn taking, the periodic motif of an initial burst then wane of information, and the prosodic motif of telos consummating resolution. Recognition of these motifs raises questions about the adequacy of Ventola's formula, and of simpler formulae - for example the variations on Initiation A Response A (Feedback) structure associated with Birmingham research, or the adjacency pair of conversational analysis. Perhaps one factor that makes these formulae more plausible than they deserve is the way in which particular, periodic and prosodic structures recurrently map onto one another in certain registers. Canonical adjacency pairs for example map a single move onto a single turn, and map an initiating informational peak onto an initiating prosodic launchpad. This kind of harmonious mapping is outlined in Fig. 2 for the following exchange: [Frank]: Are you a good ladies hairdresser Rita? [Rita]: Yeah, lam. information flow

turns

resolving telos Fig. 2: Factoring out types of structure (canonical adjacency pair)

1

26

Jim Martin

Alongside mapping of this kind, however, we need to consider exchanges in which structural motifs pull in different directions. With tracking and challenging moves, for example, serialisation falls out of step with prosody, since the second turn initiates a new adjacency pair instead of culminating one already under negotiation: [Frank]: [Rita]: [Frank]: [Rita]:

Would you like a drink? What of? Whisky. Oh yeah.

To take another example, information flow may fall out of step with orbital structure. In classroom discourse, the K1 modally responsible for the proposition under negotiation contains no new information; its content seems too low: Dkl K2 K1

[T]: [Vu]: [T]:

...you can see very clearly what she has indicated. Vu? A paragraph. Right.

Alternatively, information flow may fall out of alignment with prosody. In the clarification sequence interrupting culmination of the following exchange, the content of Frank's move is overwhelming - one indication that it should be heard as sarcasm: K2 cl rcl K1

[F]: [R]: [F] : [R]:

What is your name? Me first name? Well, that would at least constitute some sort of start, wouldn't it? Rita.

Other examples would include more than one move per turn (in so-called monologue), more than one turn per move (for jointly constructed interacts), serial re-initiations of Dkl moves as teachers move from student to student in search of a valid exchange culminating response, and so on. Historically, data of this kind has tended to frustrate the easy application of structural formulae to dialogue. But once we factor out exchange structure into structural motifs, it is easier to ask questions about why the formulae fail. Instead of giving up or complexifying formulae to the point where generalising insight is lost, we can reason about how orbital, serial, periodic and prosodic structure may pull at times in different directions, in relation to the register and speakers' local needs. Perhaps we can work towards proposing styles of harmonious and dissonant mapping of these

27

Factoring Out Exchange: Types of Structure

structural tiers, proceeding with caution on a register by register basis to see if generalised 3 patterns of mapping become clear.

5. Interpersonal Telos Above, prosodic structure was associated with the telos of the exchange - its global purpose in the interaction at hand. The domain of this telos defines the boundaries of an exchange, and so the way it is established grammatically across languages demands close scrutiny. We'll concentrate on English here; for work on other languages see Caffarel et al. (in press), Martin (in press). 5.1. M o o d Telos For English, Halliday (1994) anchors one source of interpersonal telos in what he calls the Mood function, including the Subject, Finite, and Modal Adjuncts. The presence and sequence of the Subject and Finite functions establish basic MOOD options:

declarative interrogative imperative

Mood You're Are you -

Residue coming in coming in Come in

Mood tag aren't you? fare you? 4 ] won't you?

The Finite and Modal Adjuncts (realising tense or modality, and polarity) establish the arguability of the clause. Negotiating information involves scales of probability and usuality (modalisation of propositions): value [positive] high median low

risi must certainly would probably might possibly

probability

[negative]

[isn't]

usuality [does] must always would usually might sometimes [doesn't]

Negotiating goods and services involves scales of inclination and obligation (modulation of proposals):

3

4

As Suzanne Eggins has stressed to me, we need to be prepared for the fact that harmonious mapping might mean quite different things in different registers. In pragmatic registers, for example, the goal is to close exchanges in order to resolve the purpose of the genre; in casual registers on the other hand the goal is to keep talking, to avoid embarrassing silences - so prolonging exchanges is critical. A grammatical tag in Australian English, and some in British dialects I am told.

28

Jim Martin

value [positive] high median low [negative]

inclination [will] must, be determined to will, be keen to may, be willing to [won't]

obligation Idol must, be required to will, be supposed to may, be allowed to [don't]

In addition Halliday notes the elaboration of this meaning potential through grammatical metaphors of mood (indirect speech acts) and of modality. The mood metaphors allow for the alternative realisations of the K2 move in the following K2 A K1 exchange (grammaticalised congruently as whinterrogative, metaphorically as declarative or interrogative): Information question

Unmarked realisation

wh- interrogative: declarative imperative

What is your name? - Rita.

Mood metaphor

And you are? - Rita. Tell me your name. - Rita.

Modality metaphors allow for explicitly subjective and objective realisations of modalisation and modulation, illustrated below for probability and obligation: Probability modalise: subjectify objectify

Unmarked realisation I can't bear it, can I?

Obligation modulate: subjectify objectify

Unmarked realisation You should get it fixed.

Modalisation metaphor I don't think I can bear it. can I? It's not likelv I can bear it, can I?

Modulation metaphor I'd like vou to get it fixed. It's advisable you get it fixed.

The grammar of MOOD, MODALITY and interpersonal metaphor combine with KEY (intonation), giving rise to a vast resource for launching Mood telos and thus defining a vast range of culminations for an exchange (as exemplified in Halliday (1982) and Martin (1995c)). The following exchanges compliantly resolve as the selections from these systems would predict: [Frank]: [Rita]:

Are you a good ladies hairdresser Rita? Yeah, I am.

[Frank]: [Rita]:

Are vou sure you're serious about wanting to learn? I'm dead serious.

Factoring Out Exchange: Types of Structure [Rita: [Frank]:

29

That's a nice picture, isn't it Frank? Uh yes. I suppose it is.

Note however that the modality metaphors underlined in the two preceding examples do introduce an 'extra' Mood function (Subject and Finite), by way of elaborating the meaning potential of the proposition under negotiation. As metaphor, this Mood function retains something of its negotiating potential allowing for Mood play, which puts the intended trajectory of the negotiation at risk: [Frank]:

What do you mean am I determined to go to the pub? determination to get me into a pub. (Russell 1985: 2)

[Frank]: [Rita]:

...but, uh, yes, it is, I suppose so. Well, there's no suppose about it.

[Watson]: [Holmes]:

I'm inclined to think... I should do so. (Doyle 1981: 769)

[colleague]: [server]:

I was wondering if I could have... (in bakeshop) Why do you wonder; it's right there in front of you.

I don't need

Mood telos is even more at risk where non-metaphorical Mood functions are introduced into the domain of an exchange. Ventola (1987) for example allows for the expansion of moves via clause complexing (clause serialisation). Expansions of this kind are underlined in the following examples: K1

[R]:

K2f [F]:

A2 A1

[F]: [R]:

It's very erotic. 1

Actually I don't think I've looked at that picture in 10 years.

x

2

but, uh, yes, it is, I suppose so.

1

Would you like to borrow it? (Rita is holding a book) Yeah, all right.

+ x

2

I'll look after it for you. x

3( ba)

If I pack the course in. I'll post it back, (putting book in bag)

With the second example, Frank in fact responds as if the expansion were better treated as initiating a new exchange (thus the conflated A l / K l move analysis below): [Frank]: [Rita]:

[Frank]:

Would you like to borrow it? Yeah, all right. I'll look after it for you. If I pack the course in, I'll post it back. Pack it in. You haven't even started yet.

[Al/Kl]

Jim Martin

30

[Rita]:

Why would you pack it in? Well, I just might, you know. Might think it was a soft idea.

In a similar way, Rita responds to Frank's challenge below as if it were a genuine question, negotiating a new proposition (the conflated ch/K2 move below): [Rita]: [Frank]: [Rita]: [Frank]:

Well, you know, when do you actually start teaching me like? What can I teach you? [ch/K2] Everything. You want a lot and I can't give it.

What these interpersonal metaphors, move expansions and challenging moves underline is the power of Mood functions to shape the domain of an exchange, whether they were actually intended to generate a new direction of negotiation or not. In spite of this, where a Mood element functions in the service of another exchange, it seems a marked option to take it up and negotiate it in its own right. The prosodic domain of the exchange is in some sense derailed, with new telos replacing old. Reasoning along these lines, Mood telos can be read as projecting closure for the exchange; it grammaticalises a culmination - defines an end.

5.2. Appraisal Telos For scholars researching pragmatic registers (classroom discourse, service encounters, appointment making, interrogation, business meetings etc.), focussing on Mood governed closure seems a worthwhile goal. Exchanges in discourse of this kind are designed to get somewhere - to resolve the genre at hand. With casual registers on the other hand, where keeping talking and avoiding embarrassing silence is an important task, focussing on closure is a less appropriate, and ultimately frustrating task. Consider the following dinner party data from Eggins, featuring friends in their 20s (Eggins and Slade 1997: 171): 9

Fay

10

Nick

11 12 13 14

David Fay Liz Fay

15

David

16 17

Fay David

You met his sister that night we were doing the cutting and pasting up. D'you remember? Oh yea. = You met Jill. = Oh yea. That's David's sister. Oh right. J ill. Jill's very bright actually. she's very good. She's extremely = bright = Academ-academically she's probably brighter than David... David's always precocious with his...

Factoring Out Exchange: Types of Structure

18 19 20

Nick Fay David

31

The only sixteen year old superstar () arrives in Sydney to () and straight into the mandies Straight into the what? Mandies. [laughs] He was a good boy but just no tolerance for the alcohol. I've pulled him out of so many fights it's ridiculous.

Eggins divides this dialogue into two exchanges (9-14 and 15-20), partly on the basis of audience configuration. The first exchange is oriented to bringing Liz into the picture, after which David, Fay and Nick carry on with their discussion of their absent friend David Allenby. Obviously this means including several distinct Mood functions in each exchange; but for Eggins, if each of these were taken as initiating a new exchange, the interactional dynamics of part of the conversation would be obscured. Once we turn from addressing the need to close exchanges to the need to keep going, our focus on interpersonal telos needs to shift as well. Instead of asking what it is about an exchange that leads to culmination, we need to ask what it is that encourages more talk. Consider now the following conversation between parents and their son, sitting in their car on a street next to the son's university (Eggins and Slade 1997: 67-68): 1

Brad

2 3

Fran Brad

NV Dave Fran 4 5 6

Brad Dave

7

Brad

8 9 10 11 12 13

Dave Brad Dave Brad Fran Brad

14

Fran

Look. [Mum Fran 54, Dad Dave 57, son Brad 27] See that guy. He plays the double bass. Does he? In the orchestra. He's a funny bastard and his wife's a German and she's insane. [coughs] He's funny = and she's insane? = All Germans are in=sane. = You know... You know a lot of funny people don't you Brad? Yeah, everyone at Uni is = = They're ALL mad = = They're all FREAKS Except you. Yeah. And they're all coming home now. Waddyamean? Coming, oh Like, they're coming up the hill are they?

32

Jim Martin 15

Brad

no, this... For General Studies we've got this... tutor and he's German and he's insane.

16

Fran

17

Brad

18 19 20 21 22

Fran Brad Dave Brad Dave

I didn't know you had to do General Studies. Yeah, I I got an exemption from = [noise of passing bus] Bastards! Last year. From half of it. When are you gonna do... all you odds 'n' sods subjects? Waddvamean "odds 'n' sods subjects"? Well, y'know, you can't just do languages can you?

In presenting this text I've highlighted the explicit evaluations. The text shifts from Brad's judgements about people to his father's evaluation of general subjects: He's a funny bastard When are you gonna do... all you odds 'n' sods subjects?

And where judging people, the text moves from considering individuals, to groups, and back to individuals again: and she's insane (that guy's wife) All Germans are insane. everyone.at Unj is ... They're all F R E A K S and he's insane. (German tutor)

It would appear from examples of this kind that one major source of propulsion in casual conversation is evaluation. How Brad feels about people keeps the first phase of this dialogue going; and the next phase then takes off around the value of his courses. Space precludes consideration of an explicit framework for analysing evaluation here (for outlines of appraisal systems see Eggins and Slade (1997), Martin (1997, 2000)). The three key lexical systems are illustrated below AFFECT, JUDGEMENT and APPRECIATION (akin to emotion, ethics and aesthetics in more traditional parlance): AFFECT (emotions; reacting to behaviour, text/process, phenomena) Rita: I love this room. I love the view from this window. Do you like it?

Factoring Out Exchange: Types of Structure Frank:

33

I don't often consider it actually.

JUDGEMENT (ethics; evaluating behaviour) Frank: You want a lot and I can't give it. Between you, and me, and the walls, actually I am an appalling teacher. That's all right most of the time. Appalling teaching is quite in order for most of my appalling students. But it is not good enough for you young woman. APPRECIATION (aesthetics; evaluating text/process, phenomena) Rita: Rita Mae Brown, who wrote 'Rubyfruit Jungle'. Haven't...haven't you read it? Frank: No. Rita: It's a fantastic book, you know. Do you want to lend it? Frank: Ah yes. Rita: Here. Frank: Yes. Well, thank-you very much. Rita: That's okay.

In these terms, Brad kept the first phase of conversation going by judging people as not normal in some respect {funny, insane, mad, freaks). The same strategy is used by David, Fay and Nick to propel dialogue as they elaborate on the social behaviour of their absent friend (who talks a lot, perhaps too much, is naughty, anti-social and has a drinking problem): David Fay Nick

Fay

Nick 6 7

David Nick

8

Fay

This conversation needs Allenbv. [capacity] Oh he's in London so what can we do? We don't want - [capacity] we don't need Allenbv in the bloody conversation, [capacity] 'Cause all you'd get is him bloody raving on. [tenacity] [to Liz] He's a bridge player, a naughty bridge player, [propriety] He gets banned from everywhere [tenacity] because of this antisocial or drunken behaviour, [propriety] And he just yap yap yaps all the time, [tenacity] S'pose he gives vou a hard time Nick? [propriety] Oh, I like David a lot Still but He has a very short fuse with alcohol, [propriety]

In another of Eggins' texts, workmates (aged 35-55) use a different set of resources to keep talking about a woman one of them has met. Whereas the men above were evaluated in terms of judgements about their character, these blokes

5

For the subcategories of judgement and appraisal used in this paper see Eggins and Slade 1997, Martin 1997, 2000.

Jim Martin

34

objectify the woman in terms of appreciations of her physical attributes (in a ritual gendering of appraisal behaviour - see Eggins and Slade for discussion). 11

John

12 13

Steve John

NV All

Well I went there and this eh this pretty girl come in. [reaction] She's beautiful, [reaction] [eating] What she said? She said "come in." Started to talk, you know? She's Italian. Only this big -[reaction] she had beautiful eyes, mate, [reaction] My wife next me, she's only talking to me. [laughter]

A fuller version of the Allenby text is presented below to illustrate something of the rhetoric of more extended evaluative sequences (with explicit evaluation highlighted). Taking Eggins' exchange boundaries as a guideline, the text opens by judging Allenby's conversational skills and drinking, shifts to judging Nick's sensitivity, moves on to judgements of Allenby's sister's intelligence, bridging back to Allenby's intelligence and drug-taking, and then moves on to judgements about Allenby's skill as a semi-professional bridge player. In a sense, the dialogue unfolds ideationally as a kind of scaffolding for evaluation; aligning opinions and exploring differences around judgements of behaviour is what the conversation is about: 1 2

David Fay

3

Nick

4

Fay

5

Nick

6 7

David Nick

8 9

This conversation needs Allenbv. Oh he's in London so what can we do? We don't want we don't need Allenbv in the bloody conversation. 'Cause all you'd get is = him bloody raving on. [to Liz] = He's a bridge player, a naughty bridge player. He gets banned form everywhere because of this antisocial or drunken behaviour. And he just yap yap yaps all the time.

S'pose he gives vou a hard time Nick? Oh, I like David a lot Still but Fay He has a very short fuse with alcohol. [10 sec pause...] Fay You met his sister that night we were doing the cutting and pasting up. D'you remember?

Factoring Out Exchange: Types of Structure 10

Nick

11 12 13 14

David Fay Liz Fay

15

David

16 17

Fay David

18

Nick Fay David

19

20

21

22 23

Nick David

24

Nick

35

Oh yea. = You met Jill. == Oh yea. That's David's sister. Oh right. Jill. Jill's very bright actually, she's very good. She's extremely = bright = Academ - academically she's probably brighter than David... David's always precocious with his... The only sixteen year old superstar () arrives in Sydney to () and straight into the mandies Straight into the what? Mandies. [laughs] He was a good boy but just no tolerance for the alcohol. I've pulled him out of so many fights it's ridiculous. At least he's doing well at least he's doing well in London. He's cleaning them up () Well, he rang Roman he rang Roman a week ago Did he? I didn't know that.

5.3. Mood vs Appraisal Telos Drawing the discussion together at this point, perhaps what we are looking at here is a complementarity of interpersonal resources for closure and expansion. Mood telos grammaticalises closure, projecting culmination across the domain of the exchange: [Frank]: [Rita]:

K2 K1

Are you sure you're serious about wanting to learn? I'm dead serious.

Appraisal telos on the other hand lexicalises the potential for expansion, seeding the proliferation of evaluation which might ensue: [Frank]: [Rita]:

Are you sure you're serious about wanting to learn? I'm dead serious. Yeah, huh look I know I take the piss and that. But that's only because I'm not... well you know like confident like.

Jim Martin

36 But I mean I want to be (confident! Honest (I want to be confident).

This complementarity is configured as an image in Fig. 3, which attempts to outline the way in which an exchange engenders closure with respect to Mood at the same time as it enables expansion around evaluation.

Fig. 3: Complementarity of Mood and appraisal telos (judging confidence)

Here's another example from the Allenby text. In a sense, Fay allows for closure of David's initiation by replaying his Mood function and reinforcing his evaluation: David Fay

K1 K2f

Jill's very bright actually; she's very good. She's extremely bright

But David carries on, comparing Jill to David and then moving on to further evaluate David's strengths and weaknesses: David Fay David

Jill's very bright actually. she's very good. She's extremely bright Academ - academically she's probably brighter than David

just outlined, what might well ask about their continuity - in what senses are these two faces of a more generalised telos we might want to recognise as interpersonal? Halliday's work on MOOD, MODALITY and interpersonal metaphor

Factoring

Out Exchange:

37

Types of Structure

provide the insight here (Halliday (1994), Martin (1992b, 1995b); see also Lemke (1998)). Beginning with propositions, we can easily construct a series of realisations for both probability and usuality which begins with congruent realisations and pushes through metaphorical ones towards lexis which is clearly appraising in nature. In this way modalisations of probability in Mood can be related to lexicalised judgements of veracity: probability H e ' s naughty. H e ' s certainly naughty. It's certain he's naughty. It's true h e ' s naughty. It's true, honest, credible, authentic, bogus etc.

[judgement: veracity]

Similarly, modalities of usuality can be related to judgements of normality: usuality H e ' s naughty. H e ' s often naughty. It's usual for him to be naughty. It's normal for him to be naughty. It's normal, average, fashionable, peculiar, odd etc.

[judgement: normality]

For proposals, modulations of inclination can be related to lexicalised affect: inclination I'll go. I ' m willing to go. I'd be delighted to go. I'd be rapt to go. I'd be excited, sad, proud, comfortable, uneasy etc.

[affect]

And modulations of obligation can be related to lexicalised judgements and appreciations: obligation Go. You should go. Y o u ' r e supposed to go. It's expected you'll go. It'd be clever for you to go. It'd be brave, honest, rash; sensitive, unfair etc. ... It'd be innovative, challenging, thought provoking etc

(judgement: capacity] [judgement: various] [appreciation]

Reasoning along these lines, we can perhaps position Mood telos and appraisal telos on a cline, with grammaticalised realisations at one end and lexicalised

38

Jim Martin

realisations at the other - and with Halliday's modality metaphors construing meaning in between.

6. Reconciliation In orienting this paper, I noted my concern with two disjunctions - the conative/expressive and the pragmatic/casual. In response, I've explored exchange structure from the perspectives of types of structure, and appraisal. Extending Berry, I've suggested that the exchange can usefully be factored out as tiers of orbital, serial, periodic and prosodic structure, mapped onto one another in various ways. To the extent that this array of readings proves productive, we should perhaps be careful of theories that prescribe one kind of structure for dialogue analysis (e.g. the preoccupation with seriality in conversational analysis) - preferring theories with bi/tri/multinocular vision. And, we should also be careful of analyses that privilege specific mappings of tiers by 'reducing' the exchange to a single layer of structural functions (e.g. Birmingham research and their I A (R/I) A R A (F n ) structure potential). It may well prove more constructive to explore types of structure and their interdependencies as a way of understanding differences among dialogic registers. Extending Eggins, I've suggested that two relatively distinct pulses of telos can be posited for an exchange - Mood telos and appraisal telos. Mood telos precipitates closure, with respect to Halliday's Mood function; appraisal telos on the other hand anticipates expansion, with respect to the attitudes available for negotiation (as speakers align and individuate by way of negotiating solidarity relations). More generally, we can interpret Mood telos as the grammatical face of appraisal telos, or alternatively, appraisal telos as the lexical face of Mood. Seen in this light, Mood telos suits pragmatic registers, where hegemonic institutional pressures favour non-negotiable solidarity; whereas appraisal telos suits casual registers, where solidarity relations are put at risk. On the basis of these suggestions, I'd like to see dialogue analysis move into a much more fertile dialectic with both lexicogrammar and social context, moving beyond forms, and on to meaning as function in context - so that we can better understand language as a resource for negotiating social relations. It seems to me that this is especially important in relation to casual registers, which ideologically speaking are anything but 'casual' as they coercively dissemble the many prejudices tearing apart our world at the same time as they generate and consolidate our energy for social change!

References Berry, M. (1981a): Systemic Linguistics and Discourse Analysis: A Multi-Layered Approach to Exchange Structure. In: Coulthard and Montgomery. 120-145.

Factoring Out Exchange: Types of Structure

39

Berry, M. (1981b): Towards Layers of Exchange Structure for Directive Exchanges. Network 2. 2 3 32. Burton, D. (1980): Dialogue and Discourse. London: Routledge and Kegan Paul. Burton, D. (1981): Analysing Spoken Discourse. In: Coulthard and Montgomery. 146-157. Caffarel, A., J. R. Martin and C. M. I. M. Matthiessen (in press): Language Typology: A Functional Perspective. Amsterdam: Benjamins (Current Issues in Linguistic Theory). Christie, F. and J. R. Martin (1997): Genre and Institutions: Social Processes in the Workplace and School. London: Cassell (Open Linguistics Series). Coffin, C. (1997): Constructing and Giving Value to the Past: an Investigation into Secondary School History. In: Christie and Martin. 196-230. Coulthard, M. (ed.), (1992): Advances in Spoken Discourse Analysis. London: Routledge. Coulthard, M. and M. Montgomery (eds.), (1981): Studies in Discourse Analysis. London: Routledge and Kegan Paul. Doyle, A. C. (1981): The Valley of Fear. The Penguin Complete Sherlock Holmes. Harmondsworth: Penguin. Eggins, S. and D. Slade (1997): Analysing Casual Conversation. London: Cassell. Fuller, G. (1998): Cultivating Science: Negotiating Discourse in the Popular texts of Stephen Jay Gould. In: J. R. Martin and R. Veel, Reading Science: Critical and Functional Perspectives on Discourses of Science. London: Routledge. 35-62. Fuller, G. and A. Lee (1997): Textual Collusions Discourse: Studies in the Cultural Politics of Education, 18.2. 409-423. Halliday, M. A. K. (1979): Modes of Meaning and Modes of Expression: Types of Grammatical Structure, and their Determination by Different Semantic Functions. In: D. J. Allerton, E. Carney and D. Holdcroft (eds.), Function and Context in Linguistic Analysis: Essays Offers to William Haas. Cambridge: Cambridge University Press. 57-79. Halliday, M. A. K. (1982): The De-automatization of Grammar: from Priestley's 'An Inspector Calls'. In: J. M. Anderson (ed.), Language Form and Linguistic Variation: Papers Dedicated to Angus Macintosh. Amsterdam: Benjamins. 129-159. Halliday, M. A. K. (1994): An Introduction to Functional Grammar. London: Edward Arnold. Hasan, R. (1977): Text in the Systemic-Functional Model. In: W. Dressier, Current Trends in Textlinguistics. Berlin: Walter de Gruyter. 228-246. Hutchby, I. and R. Wooffitt (1998): Conversation Analysis. Cambridge: Polity. Iedema, R. (1995): Literacy of Administration (Write it Right Literacy in Industry Research Project Stage 3). Sydney: Metropolitan East Disadvantaged Schools Program (Bridge and Swanson St, Erskineville, NSW, Australia). Iedema, R., S. Feez and P. White (1994): Media Literacy (Write it Right Literacy in Industry Research Project - Stage 2). Sydney: Metropolitan East Disadvantaged Schools Program. (Bridge and Swanson St., Erskineville, NSW, Australia) Lemke, J. L. (1992): Interpersonal Meaning in Discourse: Value Orientations. In: M. Davies and L. Ravelli (eds.), Advances in Systemic Linguistics: Recent Theory and Practice. London: Pinter (Open Linguistics Series). 82-194. Lemke, J. L. (1998): Resources for Attitudinal Meaning: Evaluative Orientation in Text Semantics. Functions of Language 5.1. 33-56. Martin, J. R. (1992a): English Text: System and Structure. Amsterdam: Benjamins. Martin, J. R. (1992b): Macro-Proposals: Meaning by Degree. In: W. C. Mann and S. Thompson (eds.), Discourse Description: Diverse Analyses of a Fund Raising Text. Amsterdam: Benjamins. 359-395. Martin, J. R. (1995a): Text and Clause: Fractal Resonance. Text 15.1. 5-^2. Martin, J. R. (1995b): Interpersonal Meaning, Persuasion and Public Discourse: Packing Semiotic Punch. Australian Journal of Linguistics 15.1. 33-67.

40

Jim Martin

Martin, J. R. (1995c): Reading Positions/Positioning Readers: JUDGEMENT in English. Prospect: a journal of Australian TESOL 10.2. 27-37. Martin, J. R. (1996): Evaluating Disruption: Symbolising Theme in Junior Secondary Narrative. In: R. Hasan and G. Williams (eds.), Literacy in Society. London: Longman (Applied Linguistics and Language Study) 124—171. Martin, J. R. (1997): Analysing Genre: Functional Parameters. In: Christie and Martin, 3-39. Martin, J. R. (to appear): Prosodic structure: grammar for negotiation. In: S. Cummings (ed.), Constituency... Cambridge: Cambridge University Press. Martin, J. R. and G. Plum (1997): Construing Experience: Some Story Genres. Journal of Narrative and Life History 7. l ^ t . (Special Issue: Oral Versions of Personal Experience: Three Decades of Narrative Analysis; M. Bamberg Guest Editor). 299-308. Martin, J. R. (1998): Practice into Theory: Catalysing Change. In: S. Hunston (ed.), Language at Work. Clevedon: Multilingual Matters (British Studies in Applied Linguistics 13 (BAAL 13)). 151-167. Martin, J. R. (2000): Beyond Exchange: Appraisal Systems in English. In: S. Hunston and G. Thompson (eds.), Evaluation in Text: Authorial Stance and the Construction of Discourse. Oxford: Oxford UP, 143-175. Matthiessen, C. M. I. M. (1988): Representational Issues in Systemic Functional Grammar. In: J. D. Benson and W. S. Greaves (eds.), Systemic Functional Approaches to Discourse. Norwood, N. J.: Ablex. 136-175. Matthiessen, C. M. I. M. (1995): Lexicogrammatical Cartography: English Systems. Tokyo: International Language Sciences Publishers. Russell, W. (1985): Educating Rita. London: Longman (Longman Study texts). Tsui, A. B. M. (1994): English Conversation. Oxford: OUP (Describing English Language). Ventola, E. (1987): The Structure of Social Interaction: A Systemic Approach to the Semiotics of Service Encounters. London: Pinter (Open Linguistics Series). White, P. (1997): Death, Disruption and the Moral Order: the Narrative Impulse in Mass 'Hard News' Reporting. In: Christie and Martin. 101-133.

Michael Toolan Towards a Simple Schema of Speech Moves 1. Introduction Many discourse analysts are deeply sceptical of speech-act or speech-move schemas, regarding them as unreliable, partial, context-insensitive, disproportionately 'top-down', oriented to speaker-intention but unable to track revised intentions, and so on (see, as representative of discourse linguists criticisms of speech act theory, those of Eggins and Slade 1997 and Lesser and Milroy 1993). My present proposals are at variance with the speech-act tradition which aimed to specify the constitutive rules for producing, say, a promise. At the same time categorising some things as questions, others as requests, and so on, seems a necessity at some level of processing - and if it could never be done, with any degree of reliability, we could never elicit information from each other, or get others to do things. But we do; so requests and informs surely exist, and can usually reliably be recognised and oriented to by integrated participants in real contexts of engagement. What we may argue about, however, is the degree to which participants' reliable recognition of requests, informs, etc., draws on, is guided by, or even is 'governed' by, the presence of this signifying formal device or element or that confirmatory sequential factor. The Integrationist and reflexivity-attentive tradition (Harris 1998; Harris and Wolf 1998) will also draw one to ask: what are we doing when we classify certain moves as questions, undertakings, and so on? I will give three simple examples why a workable basic categorisation of communicational moves, and further to that a workable of move-types, is necessary. Though not a random sample, each of these seems representative of recent work which both depends to a degree on act-categorisation, and may be undermined by the uncertainty of the very categorisation it would hope to be able to invoke.

Example 1: Identifying Mitigated Directives in Doctor-Patient Discourse Why the reliable identification of speech-act or speech-move categories remains in my view a pressing requirement in linguistically-minded discourse analysis is I think indirectly highlighted by an interesting recent paper (Skelton and Hobbs 1999) that appeared in the February 1999 issue of the British Medical Journal.

42

Michael Toolan

The paper is entitled "A descriptive study of co-operative language in primary care consultations by male and female doctors". As it happens, both authors are in the Department of General Practice Medicine of Birmingham University. In passing it is worth noting that a growing number of language-focused papers like Skelton and Hobbs' is appearing - papers that analyse and critique the language of one group of specialist users or another and reporting their findings back to those specialists in their own professional journals. The authors have taped and transcribed something like sixty hours of general practice consultations, involving both male and female general practitioners, and their present paper uses a very simple operational measure of cooperative language to probe whether doctors differ in the cooperative language they use, depending on their gender. They find no significant gender contrast. As an operational quantification of 'cooperative language', the authors search in the consultation transcripts for three measures which they postulate may be indexical: 1. 2. 3.

sheer number of words used by female doctors in comparison to male doctors; frequency of use, by the male and by the female doctors, of facilitative tag questions with falling tone such as It's a bit sore there, isn 't it?. No significant difference found. Frequency of use, by male vs. female doctors, of mitigated directives, such as You could try going for a 15-minute walk before going to bed [invented example-MT].

This third diagnostic of cooperative language is arguably the most discoursally interesting of those the authors opted to use, but it also turned out to be the most problematic to identify and isolate: There was no clear evidence that either male or female doctors used a greater number of mitigated directives, but this was in part because there was no clear distinction between mitigated and other directives. For example, If you don't see any good improvement with it then come back and I'll move you up a scale of sort of erm potency really, give you something a bit stronger. Is this a straightforward directive (it has an imperative), or a mitigated directive (the tentative if and sort of soften it), or simply a statement about what might happen? As the authors note, this kind of difficulty in interpretation of language in context makes quantitative claims - such as that male doctors are dogmatic and female doctors cooperative in speech dubious. (Skelton and Hobbs 1999: 577)

One might add that such difficulties, if general and unresolvable, render any discourse analysis that invokes and relies on speech act categories like directive etc. unsustainable. And yet, as the paper itself shows, some medical trainers are keen to use such categories. The other obvious thing to note is that, presumably most of the time the doctors who say things like if you don't see any good improvement with it then come back and I'll give you something stronger etc., along with the patients who are addressed in this way, do not seem to have the interpretive difficulties that the counting - and categorising - analyst has. There

Towards a Simple Schema of Speech Moves

43

is nothing wrong with the data, the stimuli to our metalinguistic categorisations, nor indeed any malfunctioning among the integrated users, here doctors and patients, engaged in using the language in the course of meeting particular needs. The data are not impoverished or degenerate; but our descriptive categorisations may be. Nor should it be thought that the researchers' problem lies only with isolating mitigated directives, as if what they call straightforward directives are unambiguous - because, perhaps, they are coterminous with use of imperative mood. We know the latter notion is false, and that the category of bare directives is as in need of careful characterisation as any putative subordinate categories.

Example 2: Warnings that are Not Warning Enough This example is just one of the numerous legal disputes we can find which hinge on the court's assessment of the nature or intensity or scope of a particular communicative act. This case I will mention is the early medical negligence case of Clarke v Adams (1950), Judge Slade presiding. A physiotherapist was treating a patient who suffered from a fibrositic heel, and the treatment involved subjecting the heel to heat provided by a machine; unfortunately the heel was subjected to too much heat, was further damaged, and an amputation was necessary. At trial the defendant physiotherapist and his employers were held liable for the resultant damage to the patient. The trained and experienced physiotherapist had warned the patient, using the wording he had been taught to use, as follows: When I turn on the machine I want you to experience a comfortable warmth and nothing more: if you do, I want you to tell me. Clarke v Adams (1950)

But the judge held that this warning was inadequate in the circumstances, in which what he called a 'warning of danger' was required: "The warning must be couched in terms which make it abundantly clear that it was a warning of danger." In short, the courts were of the view that there are such things as warnings, and also a more particular category of 'warnings of danger'. If anyone ought to be able to characterise and explain such 'warnings of danger', discourse analysts should.

Example 3: Users' Alterations to Recalcitrant Coding Systems The third example is offered simply as a recently published representative of the many articles published over the years, that report 'difficulty' with applying extant move or act schemas, and which they then feel little embarrassment about adopting selectively, to devise their own customised system. The article is by Deirdre Bishop and others, appeared in 1998, and is titled 'When a nod is as good as a word: form-function relationships between questions and their responses'. A crucial stage in Bishop et al's study is the 'Coding of communicative functions' (419ff.). They write:

44

Michael Toolan The system for coding communicative functions was developed over a period of five years and was influenced by a range of sources [including Dore, Coulthard and Fey], No existing system proved entirely suitable for our purposes. Most of the coding schemes developed for handling child language were targeted at relatively young children engaged in toy play, and, as Chapman (1981) pointed out, little information is available on the reliability of existing coding systems.

In the literature, many researchers applying discourse analytical categories can be found making declarations of this kind. The level of divergent understandings and classifications that this seems to indicate is I suggest regrettable and damaging.

2. A Simple Schema of Speech Moves I briefly outline my speech-move schema below. But it is a moot point whether the modifier speech should be retained, in the phrase 'speech move', since one underlying goal is to reconnect spoken and non-speech moves, in the way that these are subtly interwoven in many interactions. Ultimately the fact that particular acts are moves in an interactional encounter should trump whether they are verbal or not. In the longer run what I think we should be characterising are better called communicational moves rather than speech moves. My warrant for this line of reasoning is the Integrational Linguistic principle that verbal and nonverbal behaviour should be re-integrated in our analyses to match their interwovenness in real life (see, e.g., Harris 1990: 43). Clearly, not all the moves (gestures, actions, etc.) in a spoken encounter need be speech moves; nevertheless, since many are (particularly in plays), and to reflect the interest here in the linguistic exponents of moves, the complex phrase is retained here. Devising the scheme began with the general points made in chapter 3 of Halliday (1994) when individuals talk to each other, they are enacting exchanges (although below I will substitute the term engage(ment) for that of exchange), and these exchanged phenomena can be thought of as predominantly either mental or physical, and the grammar of English reflects this. If the enacted exchange is chiefly mental, the conversational contribution amounts to a giving of information or a seeking of information; if the exchange is chiefly physical, the contribution amounts to a giving or seeking of goods and services. Four core conversational moves, or acts, thus amount to the giving, or seeking, of either information or goods and services. These can be represented as in the grid below, with example utterances likely to perform each of the acts:

45

Towards a Simple Schema of Speech Moves goods and services

information

speaker is giving to addressee

Can I give you a hand?

I mustn't do any heavy lifting

speaker is seeking from addressee

Will you give me a hand?

Have you got a good hold?

Familiar labels for the typical kind of conversational act performed in these four core categories can be proposed:

giving seeking

PROPOSALS

PROPOSITIONS

goods and services (typically, non-verbal action)

information (typically, non-verbal action)

Undertaking Request

Inform Question

To a very considerable extent these are Halliday's categories, if not quite his labels (he opts for Offer, Command, Statement, and Question, respectively). For reasons I cannot elaborate here, given space constraints, I much prefer the move labels Request and Inform to those of Command and Statement; and the Undertaking category is arguably more comprehensive than that suggested by Offer (offers are one type of Undertaking in my schema). We can also group Undertakings and Requests together, and Informs and Questions: since Undertakings and Requests both concern future proposed action by one interactant or the other, they are called Proposals; since Informs and Questions provide or seek information, they are called Propositions. The future action that a Proposal specifies is normally nonverbal (washing the dishes, closing the door, paying for a purchase item, etc.) although occasionally it can involve a verbal performance (where the teacher requests: Billy, recite the present tense conjugation of 'donner', please). The information sought or given in response to a Proposition is normally verbal, but replies to Propositions can be performed nonverbally (A: Where's the oilcan? B: [points to far corner of garage]). Undertakings seem to be the least extensively used. But they are a relatively well-defined group: they are proposed future actions or services on the part of the speaker, ostensibly to the benefit of the addressee, the undertaking of which is, significantly, made contingent upon the addressee's express or implied consent. This last point is arguably crucial, and distinguishes Undertakings from announcements {I'm going to reorganise your bookshelves into some sort of order), which are a kind of Inform. Grammatically, Undertakings often involve a first person pronominal and one of the modal verbs shall, can, or may, where these can be interpreted as contributing to a proposal meaning 'Do you consent to me

Michael Toolan

46

doing x for you?'. Alternatively they may use a second person pronominal subject in a projecting structure using a verb such as want or like: Do you want me to take the dog for a walk? Would you like some tea? Semantically, Undertakings seem united in at least implicitly foregrounding the addressee's postulated modality of willingness or inclination: what they wish or want. A typical and expectable response to an Undertaking is a reply expressing consent - OK; alright - or a declination with a reason for so declining - No thanks, I've just had one. Each of these category types includes various subtypes. Thus Requests range in force from orders and directives to mild suggestions. One of the most extensive recent discussions of requests in English comes in Tsui (1994: 90-115). But there are divergences between Tsui's category of Requestives, which covers five rather distinct subclasses, and the Request category adopted here. Tsui summarises Requestives in the following chart: speaker action

speaker benefit addressee benefit

Requestives

addressee action

speaker and addressee action

request for permission offer

speaker benefit

request for action

addressee benefit

invitation

speaker benefit or speaker + addressee benefit

proposal

Relative to my scheme, Tsui's requests for action and invitations (such as a host's saying to a guest "Please sit down") are unambiguously Requests, while request for permission (granted by addressee for speaker's benefit - e.g., Can I get myself a drink from your cabinet?), is essentially a request, getting the addressee to consent to a course of action which is to the speaker's benefit. But Requestives involving speaker action for addressee's benefit - Can I get you a drink? - are a kind of Undertaking. It is interestingly doubtful whether proposals can be established as a distinct category, since they are quite intermediate: Shall we share a bottle of Friuli? implies shared action and mutual benefit, and by those considerations is both a Request and an Undertaking. Such borderline cases, linguistically non-congruent in Hallidayan terms, with politeness and acknowledge uncertainty of sequel subtly grammaticalised, may well be the most interesting to study in the long run. Undertakings include promises, vows, some invitations, and offers of help. Informs include claims, warnings and compliments; they entail the imparting - at one level or another - of verbalisable information. Whether or not the addressee

Towards a Simple Schema of Speech Moves

47

finds an Inform informative is a separate matter. And Questions are acts designed to obtain the kind of information that Informs supply. As is well recognised, the matching of types of speech move with particular linguistic forms is nothing if not problematic. Nevertheless, situated functionalist interpretation can be usefully underpinned, often with reference to formal and grammatical evidence. Confirmatory criteria are ordinarily available, which help ensure that disputes where analyst A says "That's an Inform" and analyst B says "No it's not, it's an Offer" are actually quite rare (and interesting). Similarly, in actual interaction, we are infrequently forced to pause and wonder "Is he asking me or telling me?", "Was that an offer or a request?"; and similarly the occasions when we do hesitate in this metapragmatic way are worthy of attention for that reason alone. Two long-recognised confirmatory criteria of Request function are (1) '/?/ease-insertability' and (2) prospection. With Requests, unlike Informs and Questions, you can usually insert the word please (or the slightly old-fashioned kindly) before the verb denoting the action to be performed, even if the Request is indirect as in I want you to kindly eat your spinach this minute! Typical prospections of the four moves are as follows: move type

possible prospection:

Undertaking Request Inform Question

'Thanks ' / 'No thanks ' 'OK ' + action/ 'No! ' + action 'Oh ' 'Yes '/ 7 don't know '

The items listed here as prospections are not cited as necessarily the normal or most usual response to the given act, but rather as entirely possible responses for that act which are in addition highly implausible as responses for any of the other acts. Consider Oh, a legitimate response to Informs. While this would be an awkward response to some kinds of Informs it remains a coherent one; but Oh as a complete and freestanding response to an Undertaking {I'llput the garbage out) or a Request {Put the garbage out, would you?) or a Question {Did you put the garbage out?) would be decidedly odd. While Undertakings, Requests, Informs and Questions are argued here to be central to discourse, they are not the only acts involved. But they are treated here as the core ones on the hypothesised grounds that just one among these four can occur as the nucleus of a first move in an interaction. But interlocutors' responses also merit classification. And in fact each of the canonical initiating acts strongly specifies a particular kind of response:

48

Michael Toolan Undertake Request Inform Question

Acknowledge (accept/decline) (Acknowledgement +) Non-verbal Performance Acknowledgement Inform

This picture introduces just one broad new kind of move, Acknowledgements. Typical examples of verbal Acknowledgements are: thanks, ok, very well, oh, and really?', and nonverbal equivalents are often used instead. Acceptances and acknowledgements are secondary in that they are semantically attenuated, as the above examples suggest, and in that they are contingent upon some prior, exchange-driving act from among the set of four described above: an Acknowledgement cannot normally initiate a speech exchange. Relatedly, while talk is often characterised as kinds of exchange, very little is given (back) when an acceptance or acknowledgement alone is made; and a bare positive acceptance of an Undertaking is not very different from a bare negative one, e.g., a declination of an offer, even though the interactional implications and consequences may be great. Nor, taken as whole groups, are Acknowledgements profoundly different from each other in form or function. For these various reasons, all verbal Acknowledgements are treated here as members of a single secondary class of move. One revision here of my previous accounts (Toolan 1997, 1998) is the adoption of the term Undertaking to denote the fourth core move, and the demotion of offer, here treated as just one kind of Undertaking. Offers, too, can be informally characterised as 'proposals by addresser to do something, for or with the consent of the addressee', but Undertaking seems the more appropriate general term, confirming that the class includes promises as well as offers (that is, undertakings that are typically 'binding' upon acceptance, as well as those that may remain indefinite). Undertakings comprise all speaker-proposals where the addressee's acceptance or consent is at least implied, if not express. Where the declared speaker undertaking is detrimental to the addressee - I'm going to have you fired - then it will typically constitute a threat, with implied if not overt lack of addressee consent; it thus falls out of the Undertaking category and into that of Informs. The alternative approach, which I have rejected, would be to say that Undertakings are a yet broader category, covering all addresser proposals to or for the addressee, with the latter's express or implied consent irrelevant. Intriguingly, that approach would somewhat parallel recent judicial reasoning concerning the crime of theft, where it has been controversially argued that the victim's consent or lack of it - cannot be determinative (see, e.g., the House of Lords ruling in R v Gomez). As implied earlier, any proposal (Request or Undertaking) involves a speaker intent on some act of doing, while a proposition (Inform or Question) involves a speaker intent on some act of knowing. And the relations of dependence or

Towards a Simple Schema of Speech Moves

49

obligation between speaker and addressee are often quite different in the four basic cases. A rather different dependence-relation is implied when a speaker Undertakes: Can I give you a hand with that? Do you want a hand with that?

than when she Requests: Will you give me a hand with this?

In an Undertaking, as the grammar reflects {...I give you...; ...you want...), the speech act is focused on the needs of the Other, the addressee - or at least on what the speaker thinks are the Other's needs. Other is cast as beneficiary, speaker-Self is presented as the giver who, as in any genuine act of giving, is likely to incur some costs. Additionally, although less overtly, the speaker-Self who makes an Undertaking may be adopting a stance of deference or subordination to whoever the addressed Other is (but not in all cases: cf. threats). In a Request, on the other hand, the relations are broadly reversed: the utterance is Self-oriented, imposes some cost on the addressee, and may cast that Other in a subordinate stance. Similar characterisations of Informs and Questions are possible. On the assumption that knowledge is desirable, Informs involve a knowing Self going to the trouble or cost of informing an Other, primarily for the latter's benefit; and Questions typically involve a not-knowing Self imposing on an Other-addressee, to Self s benefit. So, unlike the cases in Proposals, in which the subordinatesuperior roles match up with those of cost-incurrer and beneficiary, in Propositions the two pairs of roles diverge: Undertaking', Request: Inform: Question:

speaker cast as subordinate and incurs cost, addressee cast as superior and beneficiary speaker cast as superior and beneficiary, addressee cast as subordinate and incurs cost speaker cast as superior (the Knower) but incurs cost, addressee cast as subordinate but beneficiary speaker cast as subordinate but beneficiary, addressee cast as superior (Knower) but incurs cost

But it must be emphasised again that these characterisations are tendencies, fitting canonical instances fairly well and other instances more loosely or not at all. Thus in a canonical inform, the speaker is a cost-incurring Knower, and the addressee is informationally subordinate but the beneficiary. But in an atypical Inform such as a compliment, these characterisations are obviously modulated or tweaked: the superior/subordinate contrast is minimised, and speaker's cost is slight, and the extent to which the speaker is telling the addressee something they do not already know may range from great to small. Propositions and Proposals contrast in other

Michael Toolan

50

interesting ways besides. Undertakings and Requests specify actions scheduled to occur within a timespan that extends from the speaker's present into the future; their temporal reference is delimited to the non-past. Informs and Questions, by contrast, are quite unrestricted in their potential temporal reference: a Proposition can refer to a state of affairs sited in the distant past as easily as in the distant future. Critics, with reason, sometimes ask 'why just four primary moves?'; the answer is simply that this level of classification seems to be the one that strikes the best balance, at least as a starting-point for linguistically-supported analysis of communicational moves, between generality and specificity. Four moves is only one category fewer than speech act theory's five types. And these primary distinctions are the crucial ones to get right. These are the categories that need to be robust and enduring; in this they contrast with the more controversial or opaque subclassifications that have been proposed, notwithstanding the need to develop such more delicate descriptions (e.g., that of Tsui 1994, or that of Eggins and Slade 1997). Interestingly, Eggins and Slade's elaboration of move types is almost entirely focused upon the second or non-initial moves in exchanges (Eggins and Slade 1997: 169-226). Their inventory of kinds of 'sustaining' moves, if I have counted correctly, numbers 38 sub-types. But their opening moves (and for them all moves are either opening or sustaining), following Halliday as I have done, essentially number just four, with quite secondary distinctions between fact and opinion and openness vs. closedness (Eggins and Slade: 193^1). Thus, for example Is Al Gore 'Presidential'? would be classed as 'question:closed:opinion'. By way of encapsulation, all four core moves can be represented, not so much as kinds of Exchange, but as kinds of interactional Engagement: Engagement

Proposition

Inform

Question

Proposal

Undertaking

Request

How valid and valuable can such a scheme be, with its abrupt categorisation of initiating moves into four broad types - Inform, Question, Undertaking and Request? Numerous caveats need to be entered. To begin with, the scheme is notional and functional: it suggests that in functional terms, these four moves constitute the four most fundamental, and fundamentally contrasting, possible first steps in interactions. And while I seek to underwrite the validity of the types by making reference to formal and sequential and interactive criteria, ultimately these can never be anything more absolute than an appeal to typical or normal codified

Towards a Simple Schema of Speech Moves

51

practices. Consider /?/ease-insertability, prior to the lexical verb, in Requests. Noting such a tendency, in congruent or non-indirect requests, does not amount to saying that p/ease-insertability is defining of or possible in all requests, as any number of utterances, justifiably received as indirect requests will confirm: e.g., It's awfully stuffy in here with your window rolled all the way up like that. Pleaseinsertability is only a useful (not definitive) diagnostic with interrogative mood utterances that might otherwise appear quite Question-like or Undertaking-like: compare Would you please phone me when you get there with ? Would you please like to live there? and 1 Would you please like me to do that for you?. We cannot establish any set of criteria which will either a) guarantee production of a Request move or b) guarantee production of moves which are not Requests; and mutatis mutandis, the same applies for the other three moves equally; only the situated language user, not any grammatical system, controls these matters. In particular unfolding and unpredictable circumstances, any kind of imaginable verbal or nonverbal material, appropriately deployed, may be interpreted as intended to convey a Request. Nevertheless there is copious evidence to support the contention that users of English routinely perform each of the core moves in patterned and usually contrasting linguistic ways, to the point that those standard formats serve as cues or 'tilts' in the course of situated sensemaking. Accordingly it is useful to pay a good deal of attention to such language patterns, proceeding in a similar spirit to Aijmer, who elects to focus on "the set of strategies which have turned out to be functionally most appropriate and therefore [have] been codified linguistically", or what Geis (1997: 23) terms "the conventionalisation of form for function" which is "a general characteristic of colloquial language". Such attention to language patterns does not compromise the recognition that no linguistic or semiological feature is a necessary or sufficient condition for any of the core engagement moves. Finally, it should be emphasised that the core schema of four moves proposed above is intended to be at a remove from speech-act models as conventionally designed. There is no assumption or requirement here that language, as traditionally understood, need be involved in any of these - all four moves can be performed by gesture or gaze, suitably integrated with circumstances. It follows that the moves are not language-specific (they may not even be species-specific), but universals. Ultimately, the moves are to be conceived of as core resources of people, and not of 'a language'.

Michael Toolan

52

References Adams, J. and Bishop, D. (1992): Conversational characteristics of children with semantic-pragmatic disorder. I. Exchange structure, and cohesion. British Journal of Disorders of Communication, 24,211-39. Aijmer, K. (1996): Conversational Routines in English: Convention and Creativity. London: Longman. Bishop, D., J. Chan, J. Hartley, and F. Weir. (1998): When a nod is as good as a word: formfunction relationships between questions and their responses. Applied Psycholinguistics. 19, 415-432. Chapman, R. (1981): Exploring children's communicative intents. In: J. Miller (ed.), Assessing language production in children, 111-36. Baltimore: University Park Press. Coulthard, M. (ed.) (1992): Advances in Spoken Discourse Analysis. London: Routledge. Culpeper, J., M. Short, and P. Verdonk (eds.) (1998): Exploring the Language of Drama: from Text to Context. London: Routledge. Eggins, S. and D. Slade (1997): Analysing Casual Conversation. London: Cassell. Fey, M. (1986): Language intervention with young children. Boston: College-Hill Press. Geis, M. (1998): Speech Acts and Conversational Interaction. Cambridge: CUP. Halliday, M. A. K. (1994): An Introduction to Functional Grammar. 2nd edition. London: Arnold. Harris, R. (1990): On redefining linguistics. In: H. Davis and T. Taylor (eds.), Redefining Linguistics. London: Routledge, 18-52. Harris, R. (1996): Signs, Language and Communication. London: Routledge. Harris, R. (1998): An Introduction to Integrational Linguistics. Oxford: Pergamon. Harris, R. and G. Wolf (eds.) (1998): A First Reader in Integrational Linguistics. Oxford: Pergamon. Lesser, R. and L. Milroy (1993): Linguistics and Aphasia: Psycholinguistic and Pragmatic Aspects of Intervention. London: Longman. Skelton, J. and J. Hobbs (1999): A descriptive study of co-operative language in primary care consultations by male and female doctors. British Medical Journal. February 26. Toolan, M. (1997): Language in Literature: An Introduction to Stylistics. London: Arnold. Toolan, M. (1998): The give and take of talk, and Caryl Churchill's Cloud Nine. In: J. Culpeper et al. (eds.), Exploring the Language of Drama: from text to context. London: Routledge, 142-

160. Tsui, A. B. M. (1994): English Conversation. Oxford: Oxford University Press.

Cases cited Clarke v Adams (1950) 94 Sol Jo 599. R. v Gomez. House of Lords [1992] 3 W.L.R. 1067.

Frances Rock

Exploring Contemporary Method Through Historical Data: Pragma-Discoursal Dialogue Annotation 1. Introduction The Early Modern English (EmodE) courtroom is an awesome place, full of contradiction and mystery, rule, ritual and wrangle, so where better to begin to investigate courtroom dialogue? The texts under investigation here reflect language use during a turbulent and progressive period of English history. Concomitant influences included the development of the printing press and of popular education, both of which led to increased standardisation, increased internal communication and increased language awareness (Baugh and Cable, 1993: 195ff), circumstances which would no doubt have impacted on courts, as defendants became more informed and literate. Yet, despite these changes, courtrooms remained somewhat barbaric and the EmodE conception of justice was not one with which we would now hope to associate the word. This paper offers a glimpse into the EmodE courtroom through the language practices of those who could be said to control courtroom dialogue. Data under examination here is drawn from the trials sub-corpus of Culpeper and Kyto's Corpus of Dialogues, 1550-1750 (hereafter 'the corpus') (Culpeper and Kyto 1996 and 1997: 60f) and we begin our journey into the past by scrutinising the integrity of the data itself. Kopytko (1995: 516) examined Shakespearean texts, claiming that dramatic writing provided "the best information on the colloquial speech of the period". However I contend that these courtroom transcripts, recorded by court officials can be seen as more reliable sources as they were created to record, rather than imitate, naturally occurring data. Therefore this corpus potentially reveals something both about the historical courtroom and about EmodE language itself. Unfortunately the quality and detail of historical legal transcripts is not ideal for the linguist, especially the modern linguist, accustomed to the luxury of audioand even video-recorded data. Many theorists repeatedly stress the importance of intonation in all aspects of utterance interpretation (Quirk and Greenbaum 1972: 191ff and Coulthard 1992: 35). Sadly, intonational cues are, of course, not recorded in the transcripts, therefore considerable meaning potential is obscured. Similarly, the discoursal value of minimal units of talk are omitted. The transcripts do not demonstrate how "people establish the 'traffic rules' of talk" (Mey 1993: 216). Lakoff (1990: 91) comments that even in American courts of the late

54

Frances Rock

twentieth century "the court reporter normally makes note only of verbal expression, not of back channels, such as grunts, nods, or even whispered "Mmhms"". Such omissions affect our understanding of data; despite the position of power of the speaker in the excerpt below, an interruption would not be atypical, so we wonder how he holds the floor after his first question: Judge Puck:

You say, you seek no unlawful Means: What can be meant but unlawful Means in the words of your Book? If it come in by that Means that will make all your Hearts to ake1, blame your selves. What good Means can be meant by these Words?

Limited detail in EmodE trial transcripts is potentially not their only flaw. Whilst Atkinson and Heritage (1984: 3) stress the value of audio-recordings over recollection Jardine (1832: 37) reveals that "before the full introduction of shorthand, the reports of trials even by eye witnesses of the proceedings were necessarily defective". Most English shorthand systems of the period were orthographic or alphabetic, while Pitman's phonetic scheme did not appear until 1837 so we can assume that the trials examined here did not benefit from efficient shorthand. Court-transcribers frequently filled blanks in their transcriptions with "vague recollection....or their own imaginations" (Jardine 1832: 37ff). This might not trouble the linguist, who may not be concerned with the precise details of a trial's events, rather with the representation it offers of 'realistic' contemporary usage which may be illustrated through recollections and imaginings. The data should, however be considered as an artefact of these processes and influences not as transcripts as we understand the term today.

2. Courtroom Dialogue, Past and Present Many theorists describe what they perceive as special about contemporary courtroom interaction. Woodbury (1984: 198) isolates both a "game-like quality", which specifies participants' goals, and a "rather rigid and explicitly stated structure", which specifies rights, obligations and acceptable interaction patterns and topics. She focuses on use of questions and answers and the different participant roles attached to each. Atkinson and Drew (1979: 34ff) contrast legal examination with conversation and similarly highlight the relative importance of the "environment of questions and answers" in the courtroom. Luchjenbroers (1997: 481) further indicates that questions do have the same function as in "ordinary discourse settings"; a "summons to reply", but may also function as fundamental '"weapons' to test or challenge claims made by witnesses, or as 'vehicles' for accusation" (1997: 478). 1

All spellings and names in examples are as in the original documents.

Pragma-Discoursal

Dialogue

Annotation

55

Courtroom language of the EmodE period is interestingly compared and contrasted with current judicial activity as presented above. EmodE courts had very different constraints from those pertaining in today's 'western' legal settings and therefore some different linguistic practices. Questions could be said to be subject to less constraint and used with more variable function. Below, for example, we see very typical discussion and negotiation of the organisation or practice of the trial. Here the Attorney General seeks to clarify and negotiate the physical location set aside for witnesses: Lord High Steward: Attorney General:

.... and let me know who you begin with. My Lords, We do not know what Place is designed for the Witnesses; but if they are to be on the other Side, and this Croud do interpose between us and them, it will be impossible for us, who are the King's Council, to hear what they say, or know how to proceed.

The Attorney General opens his turn with a polite address-term "My Lords". This shows deference which may be particularly important, as he is intending to go on to violate the maxim of relevance by failing to reply to the question posed by the previous speaker, instead imposing on the court. This discourse marker also signals his upcoming turn. The interactional structure of EmodE courtrooms' seems less rigid, rules and 'acceptable' linguistic behaviour are less established and elicitation devices more variable in terms of their uses and users than those of the current courtroom. Lakoff (1990: 85f) observes that today's "adversarial encounters are thrilling". The adversarial English legal system at the time of the Corpus was given extra spice by such factors as "dependence upon unpaid amateur local officials" (Sharpe 1984: 21) and the "deliberate theatricality of criminal courts" (Beattie 1986: 316). I will now consider the role of dialogic participants in the EmodE courtroom to give a little context to the ensuing discussion. 2.1. The Legal Profession In contemporary Anglo-American courtrooms roles are clearly delineated, judges are "largely silent", and "division in testimony phases according to barrister objectives" are ratified by restrictions on what each set of barristers may say (Luchjenbroers 1997: 479). In contrast the EmodE courtroom saw "Few prosecuting counsel ... engaged in ordinary cases before 1714" (Beattie 1986: 353f). Practicing attorneys were "regarded with suspicion" and were likely to have a skill "analogous to that of contemporary cunning men in medicine" (Sharpe 1984: 146). Beattie (1986: 353f) summarises "the intrusion of lawyers on both the prosecution and defence sides" in recent times has removed judges and defendants from the action of modern-day trials. EmodE judges were colourful characters, apparently highly involved in the 'action'. Pike (1876: 212) notes of Judge

Frances Rock

56

Jeffreys, who appears in the corpus, that "his system of browbeating and bullying terrified....allhis audience except the counsel for the prosecution". Here Jeffreys intimidates, demonstrating an enterprising awareness of the third person: Jeffreys:

(1) That is all one whether you have his Examination or not; (2) I expect he should tell me the Truth, and the whole Truth here. (3) Friend, (4) you know your Soul is at pawn for the Truth of what you testify

In (1) and (2), he ostensibly addresses Mr. Pollexsen, a lawyer or cleric, however his comments are clearly also intended for the overhearing audience, the next witness, who he then addresses in (3) and (4). Jardine (1832: 9f) notes that "Judges and King's counsel were accustomed to display their ingenuity by perplexing the prisoner with questions and endeavouring to extract his confession from his own mouth". Wodak (1985: 181) describes the goals of today's courts as centring around finding "the truth or a "truth" that can be believed in and defined in legal terms" it would appear that the EmodE courtroom shared this goal, favouring truths established by the powerful. 2.2. Defendants and Witnesses Defendants were given "every opportunity to prove their innocence", but were allowed little help from anyone in the courtroom (Beattie 1984: 347). Their active role can be compared to that allotted to defendants in contemporary Zimbabwe. Disturbingly, EmodE defendants probably endured considerable hardship, including solitary confinement and physical torture prior to trial; defendants during the reign of Elizabeth I would reach court with "powers of body and mind wasted by confinement". Defendants may have felt their impending fate was a foregone conclusion and that their trials were a mere charade; "to be accused of a crime against the state, and to be convicted (and executed) were almost the same thing" (Jardine 1832: 9f). Such pressures would no doubt have affected the goal orientations of defendants and probably of witnesses too. Luchjenbroers (1997: 501) finds that, in Australian courts, "witnesses provide little of the informational input to jury" and have "little scope....in influencing a barrister's line of reasoning". Her investigation of witnesses finds them powerless, manipulated and having "no voice". We could predict that this would be the situation in these EmodE trials, however, both defendants and witnesses frequently assert speaking rights and exercise their voice fully. Lilburne battles throughout, to question the legitimacy of his trial. Udall too, sets much of the agenda of his trial himself, particularly by making requests for information and favours. Here, he requests the opportunity to petition, forcing the judge to either reject his request completely or ask about the petitions:

Pragma-Discoursal Udall: Judge:

Dialogue

Annotation

57

Then I humbly crave of your Lordships to grant me two Petitions What are your Petitions?

Below, we see a witness commanding his own evidence. The judge, rather than setting the questioning agenda, conforms to the elicitee's narrative and merely seeks clarification within it: Wakeman: Judge: Wakeman:

While I was there, one Mr. Holt came in, a Man that I never saw in all my Life, and said, The Business cannot be done this Day. What Day of the Week was this? Thursday: he appointed me to call upon him at Nine next Morning....

2.3. Juries Many contemporary theorists comment on the jury's role; Drew (1985: 134) notes their influence on talk, even when they are not the direct addressee, in terms reminiscent of Bell's (1991: 126) discussion of audience design in the news media. He similarly notes the "lack of access in actual trials to jurors' assessments of particular items of evidence". As a historical phenomenon jury deliberation processes are inaccessible, but the role and influence of the jury remains fascinating; Beattie (1986: 314) cites trial by jury as "among the most cherished" rights of the English throughout history; indeed several defendants in the corpus comment on the reliability of their jury-members. Yet juries were frequently punished and the prisoner returned to the tower following an acquittal (Jardine, 1832: 9f).

3. History, Intuition and Investigative Annotation Schegloff (1984: 50) identifies an "overhearer's problem" whereby a researcher "getting a snatch of conversation....can hear ambiguities in the talk that are not there for the ratified participants". This aspect of Goffman's notion of ratified participants is particularly relevant to historical studies of meaning, when, not only is the researcher 'outside' the discourse situation, but is unable even to imagine being inside the practices and events described. This makes a clear challenge for the linguist who typically denies any influence from intuition and here is truly intuition-less. Historical analysis using contemporary theory is valuable, enabling us to question the extent to which 'contemporary' theory is only truly relevant to contemporary data in its dependence on contemporary frames of reference, understanding and practice. Analysis of historical data ideally illuminates contemporary theory and its application to current data. What below exemplifies ambiguity which we can possibly not resolve neutrally:

Frances Rock

58 Witness:

as I came over the Way thro' the Court, this Gentlewoman look'd out of the Window.

Judge:

What, before you went in?

What could be interpreted as: (i) (ii)

a request for a repeat (the elicitor could not hear); a challenge to the truth or validity o f the elicitee's utterance (the elicitor does not believe the elicitee); (iii) a request for clarification (the elicitor thinks he does not understand); (iv) a request for clarification (the elicitor needs more precise details).

I have informed my analysis by reference to context and cotext and explored it by devising and applying an extensive, systematic annotation system to the corpus. Over 24,000 codes have been used to give a sense of interrelationships, similarities and differences between acts and the patterns they create, therefore providing a detailed data exploration. The scheme divides the dialogues, not using established unit divisions, such as 'utterance' or 'clause' (see Slobin 1993: 210) but using divisions into meaning units (exemplified below) (after Stiles 1992 and McGaughey and Stiles 1983: 78f). It then analyses each unit along seven dimensions, the most important of which are briefly summarised and exemplified below.

3.1. Fundamental Orientation This field is the foundation of the scheme, it explores force. In ambiguous cases, where for example the force of offer, order, advice and request could all be identified in a unit, annotation is assigned on the basis of felicity conditions (Austin 1962: 39f) in the absence of paralinguistic cues. For example, in several trials speakers were 'unable' to give orders because they lacked necessary status (Searle 1969: 63f and 1976: 5). Therefore, coding decisions here draw on Wodak's (1985: 185) "implicit rules of the court", such as who has authority to perform certain activities, "explicitly stated rules voiced by the judge and defendant", such as who is permitted to speak and when, and on interactional norms, informed by talk from beyond the courtroom.

59

Pragma-Discoursal Dialogue Annotation CODING OPTIONS REQUESTIVE TALK Is it a request for any verbal or non verbal response?

EXAMPLES • Who was at that meeting? • W h a t do you know? • But to whom do you bear Malice? • Ask him • will any man say that he will not believe these were right read • Was there ever such a damnable blasphemous heresie as this is

•

FACTIVE TALK Does it provide factual information or opinion based on fact?

•

•

DISCOURSE MARKERS Does it mark a new addressee, new activity or link or divide?

• • • • • • •

RHETORICAL TALK Is it a question which does not expect a literal answer?

ADDRESS TERMS Is it used to nominate a participant or participant group? PARAPHRASING Does it paraphrase something which has been said or done prior to or during the trial? Does it paraphrase: (i) a prior turn or turns within the trial; (ii) testimony taken before the trial or (iii) events which form part of the crime narrative? ADVICE-GIVING TALK Does the unit convey something which might help the addressee or give some favour?

•

• • •

• • •

DIRECTIVE TALK Is it an order? OFFERING Is it an offer?

•

ORGANISATIONAL TALK Is it concerned with the organisation of the interaction?

• •

•

by the lawes of the Land you are to plead to your charge it is no accusing of your selfe to say guilty, or not guilty But And Then, my Lord Come My Lord Mohun Mr. Powell Mrs. Bracegirdle You say both their Swords were drawn when you came down? she said, They threatened to be revenged this that you say concerning your Memory, which you say is not able to carry away, or containe the particulars

Heare what is said to you never make a flourish of what is in you You can easily answer, whether you are married or unmarried you must be silent and hear the Court it may be, that we, reporting your Submission unto her Majesty, may procure her Pardon for you I will ask you one Question we shall go on to call some Witnesses that were not present at the Murder

• • •

go on be patient I desire to be resolv'd by the Judge

•

Will you say your heart is lifted up unto God, when your understanding is in your heels Have we dealt so fairely with you all this while he had three messengers sent unto him I know, by information, that this Adams is a hired witnesse Next Well then therefore Mr. Lilburn

• • • • • • •

My Lords Mr. Hudson

•

you sayd at first, you would speak as a rationall and moderate man You speak great words of your own judgement Mr. Attorney hath made processe against you but I tell you the opinion of the Court

• • •

• •

• • •

• • •

those that are quotations there, are not for your purpose The ceremony is for your advantage more then you are aware of speak out Hear my Question take your Oath, and swear you did it not

for first... and afterwards I must needs tell you take away his Chaire, for I cannot see the Bar

60

Frances Rock

3.2.Topical Orientation CRIME NARRATIVE Is it concerned with or have direct relevance to the 'story' of the crime? Does it reference characters, places and events which become relevant to the crime in the context of each trial. NON-CRIME NARRATIVE Does it not develop the crime story or provide or discuss information about the crime? Within such units, elicitors commonly advise, request cooperation and make procedural explanations.

• • • • • • • •

When did you see the Body of Sir Edmund-bury Godfrey? What Time of the Day was it? What did he say? What do you say to that, Dunne? I desire she may consider well Here is a great Circumstance is not that a plain Question? to tell the Truth, would rob thee of none of thy Senses

3.3. Level Orientation Wodak (1985: 185) mentions metalanguage, isolating "implicit rules of the court" and "explicitly stated rules voiced by the judge and defendant", but focuses on how witnesses may exploit these rules. She examines the way witnesses appropriate such devices as metacommunicative frameworks to maintain role ambiguity and show identity. Metalanguage also seems to relate to the development of the trial as a whole, to the judge's role and to other talk within the trial. Searle (1975: 5) comments on "relations to the rest of the discourse". He presents metalinguistic or metadiscoursal lexicalisations such as: I reply, I conclude which "relate utterances to other utterances and to the surrounding context". He claims that these are mostly associated with 'statements', however, in this data they have a special role within questions and organisational nonquestions: METALINGUISTIC TALK Does it mention the physical form of talk, for example volume, grammar or the very presence or absence of speech? METADISCOURSAL TALK Does it request, state or query topic, speakers, audience or turn order? METACOMMUNICATIVE TALK Does it describe, suggest, paraphrase or evaluate content of either language or activity in court or outside?

• • • • • • • • • • • • • • •

that matter will be very short Go on, be silent you hear what is charged upon you you cannot now be heard any more Hark you Pray give my Lord an Account and so we shall shew your Lordship and the Jury We ask you but as to one particular thing we will hear you till midnight You have given a great slander a word is a great deale It was a very unlucky thing that he had it then you had Order to let none come in it is a plain Question thou had'st better tell the Truth

Pragma-Discoursal

Dialogue

61

Annotation

METAPRAGMATIC TALK Does it: (i) mention utterance meaning; (ii) state, negotiate or query force; (iii) Exhibit performativity? NON-METATALK Is it not 'meta' on any plane? Does it stand alone, in not containing metareference to other talk inside or outside the current interaction?

• • • • • • • • •

I charge thee Then prithee let me persuade thee I advise you for to plad I promise you she cannot swear Have you seen him at the Plow at any Time? Was Mr. Green there? we will take care of that for your Witnesses you should have brought them with you look to thy own Soul that is in great peril of everlasting Ruin and Destruction by these means

•

3.4. Cohesive Orientation Annotation here allows quantitative assessment of reference to events either inside or outside the trial. For example, investigation of the proportion of units-per-trial devoted to discussing the crime (the extrinsic past tense), in relation to total questions asked. EXTRINSIC ANAPHORIC REFERENCE Does it refer to previous states or activities beyond the current interaction? EXTRINSIC 'ONGOING' REFERENCE Does it refer to contributions, states or activities which were in operation before the current interaction (trial) and will continue once the trial is over? EXTRINSIC CATAPHORIC REFERENCE Does it prospect states or activities beyond the current interaction? INTRINSIC ANAPHORIC REFERENCE Does it refer anaphorically to previous contributions, states or activities within or throughout this interaction, that is, within the trial? INTRINSIC 'ONGOING' REFERENCE Does it refer to current contributions, states or activities within the trial?

• • • • • • • • • • • • • • • • • • • • •

you knew of your comming long agoe Was Mr. Green there? Do you remember any other Company was with him? Have you seen them twice together? they are Richard Cary, and William Evans Then you do not know Girald the Priest? Who is your Uncle? dost thou call this Religion? I know my own Heart So then you are the man if we doe take it away contrary to law, we are liable to smart for it thy Conscience will bear witness to thy Accusation But that is not my Question That is all Nonsense You said you would be rationall and this you have said now we have heard it you were here and stood divers hours yesterday We have been as long as you have the Jury stand here charged dost thou imagine that any Man hereabouts is so weak as to believe thee?

62

Frances Rock

INTRINSIC CATAPHORIC REFERENCE Does it prospect, anticipate or expect contributions, states or activities within this interaction?

• • • • •

therefore I will say no more to thee you must immediatly go on Dispatch Sir I desire to ask Barter one Question Say the Truth, Man

3.5. Propositional Content Orientation This category draws on Lyon's (1972: 50f) distinction between the descriptive, social and affective functions of language. F ACTIVE UNITS Does it derive from, contain or concern only facts? AFFECTIVE UNITS Does it contain any indication of, or reference to the feelings or emotions of speaker, hearer(s) or third parties.

• • • • • • • • •

did you tell this to Col. Penruddock? We will not deny a tittle of the Law Was Dunne taken in the Malt-house? When was that? Have we dealt so fairely with you all this while This is a mighty improbable Business Thou art a very innocent Soul Alack-a-day! How long do you believe he might be dead before you saw him?

4. Using the Annotated Data The resulting annotated texts offer a route into quantitative and qualitative insights, allowing analysts to search the corpus for specific annotations or sequences of annotations rather than searching by word(s) or sequences of words, the annotation allows researchers to identify and systematically examine: • • • • • • •

the extent of similarities and differences between speakers' strategies, such as requesttechniques, across trials; meaning; meaning patterns within and between trials; relationships between meaning and form; speaker meanings which are somehow related, despite dissimilar surface form; particularly frequent or infrequent activities; particularly frequent or infrequent combinations of activities.

We can use the annotations to explore our expectations about courtroom language. It would be reasonable to expect high frequencies of 'Requestive', 'Crime Narrative' units making 'Extrinsic Anaphoric Reference'. Such results would demonstrate that much talk in the trials was predictably questions about the crime itself. The combined results of all trials violate this expectation however. Quantitative analysis shows that trials which contain few requests, also contain

Pragma-Discoursal

Dialogue

Annotation

63

less extrinsic anaphoric reference and less crime narrative, while those which were high in one of these 'anticipated categories', were high in all of them. These findings indicate the diversity of activity to be found in the EmodE courtroom and the heterogeneity of its concern with accounts of crimes. Analysts are able to use the annotation to identify trials which call witnesses to describe their knowledge of the crime and its circumstances and to focus on the questioning strategies in those trials. These can then be compared to trials in which questions about the crime narrative do not proliferate to investigate what is actually happening in those trials. I will now expand with some more qualitative discussion of elicitation in EmodE courtrooms.

5. Form and Function The scheme identifies items which have identical realisations but different functions. For example the address term Sirrah used to nominate an addressee is coded differently to Sirrah used to address without specifically nominating. Therefore such items become searchable by meaning rather than form. The coding system reflects small differences in function, each code changing slightly to reflect exact usage. For example, differences in requests for information about specific past events and ongoing states which began in the past. In several trials, elicitors requested that witnesses might be asked a question rather than asking directly. Two styles of this question-requesting activity emerged: In the trial of Mohun, for example, Mohun himself says; "I desire this witness may be asked" while in the trial of Love, the Attorney General just says "ask him". Annotation therefore reflects the similarity in function of these two formally different units, the only difference in the annotation assigned to each unit reflects the contrasting concern with 'affective' meaning about "desire" in the first unit as opposed to the 'factive' activity of asking. Mohun's technique could be seen as face-saving to the judge whom he asks for permission. Status differences between the Attorney General and the judge and between Mohun and the judge could also be the cause of Mohun's deference.

6. Idiosyncratic Elicitation Strategies Annotation finds the Attorney General in Mohun's trial repeatedly using a distinctive type of eliciting move to begin witness examination: Attorney General:

(1) Mrs. Sandys, (2) give my Lords an Account, (3) whether you were present with my Lord Mohun and Mr. Hill at the Tavern that Day that Mr. Mountford was killed, (4) and what passed there.

This move typically contains several key elements:

64

Frances Rock (i) An initial address-term (1), can be seen as multifunctional in nominating an addressee and introducing them to the court; (ii) The 'request' unit itself, (2) has two features: (a) it could be said to have the force of an order; and (b) it uses a metadiscoursal request for "an account". (iii) Units (3) and (4) specify details for the elicitee to focus on. Elicitees typically respond to these moves with long narratives. Below is the only exception to the Attorney's pattern, even here the format remains identifiable: Attorney General: Dixon.: Attorney General:

Mr. Dixon, do you remember the Day when Mr. Mountford was killed? Yes, Sir. Pray acquaint my Lords, what Persons you carried in your Coach that Day; whether you carried my Lord Mohun and Mr. Hill, and to what Place, and where you waited for them.

The Attorney General seems anxious about this witness' potential contribution, and indeed begins his examination by checking that Dixon knows which day he has been called to testify about. However, he maintains the initial address-term and once his check has accomplished a desired outcome, returns to the format outlined above. Daulton, the Attorney General in the trial of John Udall, also has a fairly idiosyncratic style of questioning. Rather than requesting information from Udall, he states information in apparent attempts to provoke responses. We could say that he uses information units like a variant on the declarative question: Daulton:

It proveth this, that you had a purpose to write this Book; and those things were Collections from your Friends, and Preparations thereunto.

Below, Daulton presents circumstantial evidence for jury consideration. This declaration to the jury also functions as an accusation for Udall to answer: Daulton:

You of the Jury consider this, that Thompkins was Mrs. Crane's Man, and one that was privy to all the Printing that was at her House; and Mr. Udall used to go often thither.

When this Attorney General does ask questions, they frequently take the form of declarative challenges: Daulton:

(1) But you cannot deny the second Point, (2) that you had a pretended Malice, (3) for it is extant in your own Confession

(1), for example, is interpreted as a rhetorical question on the grounds that it could be a contraction of you cannot deny the second point can youl. This excerpt is very typical of Daulton's speech, in that he supplements his question with a subquestion and information unit, using provocative, value-laden, affective judgements.

Pragma-Discoursal

Dialogue

Annotation

65

Unluckily for Udall, the judge in his trial adopts a similar strategy. Below, he uses affective vocabulary such as clear and well enough and metacommunicative comments at (1) and (2), as a prelude to a metadiscoursal move, requesting a response (3): Judge:

(1) The Matter is clear, (2) and we see what you can say to it well enough; proceed, Mr. Daulton, to the Proof of the second Point.

(3)

Lady Lisle's judge uses a similar tactic. He appears to offer a compliment in (1), however annotation suggests that the judge is actually offering his opinion or a sarcastic jibe in an attempt to perhaps unnerve: Lord Chief Justice: (1) You are all very free, kind-hearted People, (2) I say that for you

This paper overviews the potential usefulness of pragmatic and discoursal annotation to dialogue analysis. Full and systematic annotation cleans the spectacles of analysis, allowing the researcher to see macro and micro structures and activities which might otherwise have been viewed differently or overlooked completely.

References Atkinson, J. and Drew, P. (1979): Order in Court The Macmillan Press Limited: London. Atkinson, J. and Heritage, J. (1984): Structures of Social Action: Studies In Conversation Analysis Cambridge University Press: Cambridge. Austin, J. (1962): How to Do Things with Words Harvard University Press: Massachusetts. Baugh, A. and Cable, T. (1993): A History of the English Language Routledge: London. Beattie, J. (1986): Crime and the Courts in England, 1660-1800 Clarendon Press: Oxford. Bell, A. (1991): The Language of News Media Blackwell: Oxford. Coulthard, M. (1992): The Significance of Intonation in Discourse. In: Coulthard, M. (ed.) 1992 Advances In Spoken Discourse Analysis, Routledge: London, 35-49. Culpeper, J. and Kyto, M. (1996): Corpus Manual a Corpus of Dialogues, 1590-1720 Unpublished. Culpeper, J. and Kyto, M. (1997): Towards a corpus of dialogues, 1550-1750. In: Language in Time and Space. Studies in Honour of Wolfgang Viereck on the Occasion of His 60th Birthday (Zeitschrift fur Dialektologie und Linguistik - Beihefte 97) Ramisch, H. and Wynne, K. (eds.), Franz Steiner Verlag: Stuttgart, 60-73. Drew, P. (1985): Analysing the Use of Language in Courtroom Interaction In: Van Dijk, T. (ed.) Handbook of Discourse Analysis: Volume 3 Academic Press: London, 133-149. Jardine, D. (1832): Criminal Trials: Volume One Charles Knight: London. Kopytko, R. (1995): Linguistic Politeness Strategies in Shakespeare's Plays In: Jucker, A. (ed.) (1995) Historical Pragmatics John Benjamins Company: Amsterdam, 515-540. Lakoif, R. (1990): Talking Power: The Politics of Language Basic Books: USA. Luchjenbroers, J. (1997): In Your Own Words: Questions and Answers in a Supreme Court Trial In: Journal of Pragmatics, 27, 477-503. Lyons, J. (1977): Semantics: Volume 1 Cambridge University Press: Cambridge. McGaughey, K. and Stiles, W. (1983): Courtroom Interrogation of Rape Victims: Verbal response Modes Used by Attorneys and Witnesses during Direct Examination vs. Cross Examination Journal of Applied Social Psychology, 13, 1, 78-87.

66

Frances Rock

Mey, J. (1994): Pragmatics: An Introduction Blackwell Publishers Limited: Oxford. Pike, L. (1968): A History of Crime in England: Illustrating the Laws in the Progress of Civilisation Paterson Smith: New Jersey. Quirk, R. and Greenbaum, S. (1972): A University Grammar of English Longman Group UK Limited: Harlow. Searle, J. (1969): Speech Acts: An Essay on the Philosophy of Language Cambridge University Press: Cambridge. Sharpe, J. (1984): Crime in Early Modern England, 1550-1750 Longman Group UK Limited: Harlow. Slobin, D. (1993): Coding Child Language data for Crosslinguistic Analysis In: Edwards, J. and Lampert, M. (eds.), Talking Data: Transaction and Coding in Discourse Research Lawrence Erlbaum Associates: London, 207-219. Stiles, W. (1992): Describing Talk: A Taxonomy of Verbal Response Modes Sage Publications: London. Wodak, R. (1995): The Interaction Between Judge and Defendant In: Van Dijk, T. (ed.), Handbook of Discourse Analysis: Volume 4 Academic Press: London, 181-191. Woodbury, H. (1984): The Strategic Use of Questions in Court Semiotica 48 - 3/4, 197-228.

Acknowledgements This piece is based on a thesis submitted as part fulfilment of an MA in Language Studies at Lancaster University. The author gratefully acknowledges the valuable assistance of all who have helped in the preparation of this version, particularly Jonathan Culpeper, Malcolm Coulthard and Janet Cotterill.

Peter White

Dialogue and Inter-Subjectivity: Reinterpreting the Semantics of Modality and Hedging 1. Introduction One of perhaps the most rudimentary divisions which can be made among texts is that between monologue and dialogue. The taxonomic criterion at stake here seems obvious, turning as it does on whether the one voice speaks or writes alone or whether the texts involve the intersection of multiple voices in immediate interaction. In this paper, I will, in a sense, argue against the distinction, contending that all texts are dialogic. In this I rely upon the now widely influential insights of Bakhtin and Voloshinov1 into what they term dialogism and heteroglossia (see, for example, Bakhtin 1973, Bakhtin 1981 and Voloshinov 1995). My core purpose, however, is not to rehearse Bakhtin/Voloshinov's arguments as such, but to show how their framework can provide a novel, more insightful understanding of the linguistic resources which have traditionally been analysed under headings such as modality, evidentiality and hedging. My purpose, then, is to show how a dialogic perspective on these resources provides us with a clearer sense of their rhetorical potential and enables us to provide better models of the way texts use these meanings to position themselves and their audiences evaluatively and inter-subjectively. One of my key claims will be that traditional analyses of these resources are, in effect, insufficiently dialogic, in the Bakhtinian/Voloshinovian sense, in their orientation. That is to say, they construe the communicative act as essentially deriving from individualised consciousness and as an act of monologic self expression, rather than as intrinsically and fundamentally interactional, as coming into being in the context of dialogic exchange and being thereby conditioned. I will argue that as a consequence, the traditional accounts often misunderstand what is at stake rhetorically when these resources are brought into play. This framework for dealing with evaluation and intersubjective positioning arises from research within the Hallidayan systemic functional framework

1

This is obviously not the place to enter the debate on the respective identities of Bakhtin and Voloshinov and I take no position here on whether or not the Voloshinov texts were in, fact, authored by Bakhtin. For the sake of simplicity I will acknowledge the various authorships of texts cited but in generally treat them as constituting a single philosophical/linguistic framework. For more on this debate, see, for example, Dentith 1995)

68

Peter

White

conducted in Australia over the past 15 years. The work was primarily directed at literacy curriculum development for the New South Wales Department of School Education. (For an overview of the research see Christie and Martin 1997, Martin 2000 or reference the Web site at www.grammatics.com/Appraisal/ for an introduction and bibliography of further readings.) The research team developed the term 'Appraisal' as a label for the network of interconnected linguistic resources for adopting and managing evaluative positions.

2. The Language of Inter-Subjective Positioning - A Textual Illustration Let me illustrate in general terms the linguistic resources, semantic values and rhetorical issues at stake here. The following text extract, from the opening of a paper exploring the relevance of Marx for contemporary anthropological theory, makes significant use of values which would be central to any traditional discussion of evidentiality and hedging - values concerned with epistemic modality, vague attribution and intensification/down-toning. (Values of most obvious interest have been underlined). This consideration of Marx as a precursor, though a largely unacknowledged one, of the modern anthropological theory of culture is situated on somewhat improbable terrain. It lies in a no-man's-land between two rather unlikely propositions: first, that there can be anything much new to be said about Marx; and second, that, having been enthusiastically cited now for a century by those who would entirely conflate human history with natural history and culture, into its occasioning circumstances, Marx had anything at all of value to say to his contemporaries - still less has anything to offer us about culture. Yet such a consideration is neither absurd nor untimely, as Raymond Williams' recent discussion cited above demonstrates. (Kessler 1987: 35)

We see that the author's proposition that Marx has something to contribute to modern cultural theory is presented as only a 'consideration', as in itself 'improbable' ('situated on somewhat improbable terrain') and associated with other 'unlikely' propositions ([lying]... between two rather unlikely propositions). We notice also that these modal qualifications are themselves qualified - the metaphorical terrain of uncertainty is 'somewhat' improbable and the associated propositions are 'rather' unlikely. We further observe that the scope of the claim about Marx's contribution is constrained - the author's proposition is not that there 'is' something new to be said about Marx but that there 'can' be 'anything much' to be said about Marx. My concern in this paper is with the rhetorical potential of such modal and evidential values, with how we construe the inter-subjective positioning such resources may serve to negotiate. In order to argue that we need to go beyond the explanatory frameworks provided by the traditional literature, I will firstly explore

Dialogue and Inter-Subjectivity: Reinterpreting the Semantics of Modality and Hedging

69

some of the key assumptions which sustain the traditional framework and then turn to the Bakhtin/Voloshinov-inspired approach, which I propose as an alternative.

3. Evidentiality and Hedging - Traditional Approaches We can say that traditional approaches (see, for example, Lyons 1977, Palmer 1986 or Chafe and Nichols 1986) have construed these resources as either being concerned with the speaker's commitment to the truth-value of their utterances or with allowing the speaker to characterise their utterances as less than factual or as less than certain 'knowledge', as having diminished or doubtful epistemic reliability. Thus Lyons contrasts what he terms the 'subjectivity' of the modal meaning with the 'objectivity' of 'bare assertions'. Modalised utterances he characterises as 'non-factive', in contrast to what he terms 'factive' utterances, which are 'straightforward statements of fact [which] may be described as 'epistemically non-modal' because the speaker commits himself to the truth of what he asserts' (Lyons 1977: 794). In a similar vein, Chafe observes with respect to the use of evidentials, People are aware, though not necessarily consciously aware, that some things they know are surer bets for being truer than others, that not all knowledge is equally reliable. Thus one way in which knowledge may be qualified is with an expression indicating the speaker's assessment of its degree of reliability. (Chafe 1986: 264)

Under these formulations, therefore, the semantics at issue is represented as emerging from meaning making in which individual speakers apply a 'subjective' coloration or slant to the propositional content of their utterances in order to hedge the truth value of that content or to indicate doubts about its reliability. The semantics are construed as turning on whether individual speakers present themselves as able or unable, or as willing or unwilling to commit to the truth of what they assert. Frequently the choice is construed as one between objective 'facts' and the subjective uncertainty of the modal or the evidential value - hence Lyons' contrast between 'non-factive' and tellingly what he terms the 'straightforward statements o f f a c t ' (Lyons 1977: 794, emphases mine). There is an implication, therefore, in the various formulations that the overriding purpose of communication is for the speaker to offer 'truth' or certain knowledge and that these modal, evidential or hedging values are introduced only in communicatively non-optimal circumstances. The speaker is represented as inserting modal values and hence adopting an interpersonal position when they have failed to achieve an absolute, and hence 'straightforward' (following Lyon's citation above), commitment to the truth of their utterances. These are thus values to be used, so to speak, when facts fail you. The term 'hedge', I believe, reflects

70

Peter White

this perspective, suggesting as it does some form of evasion or even deceit, some sense of improperly 'having it both ways'. As we have seen, the text extract above is rich in various modal and evidential meanings - what Lyons and others would term 'subjective', and 'nonfactive' values. Accordingly, under traditional approaches, the text would presumably be seen as 'heavily hedged', with the various values discussed above supposedly indicating uncertainty or low commitment by the author to the truth or epistemic reliability of his propositions. The text would thus be construed as potentially equivocal, indirect, uncommitted, even evasive, as in some fundamental sense non-optimal communicatively in that it fails Lyon's test of representing a 'straightforward statement of fact'. Such a characterisation is, of course, out of step with the text's actual interpersonal and rhetorical functionality. It fundamentally misrepresents the way the authorial voice is positioned inter-subjectively in the sense that, according to the conventions of argumentation which operate in such academic contexts, the authorial voice here would be read as authoritative and strongly committed. There is nothing evasive, indirect or uncertain about the writer's communicative stance, a point which is made clear by the strength with which the writer presents his position at the conclusion of the paragraph. We also note that notions of truthvalue and epistemic reliability, so central to the traditional accounts, are largely irrelevant in the argumentative context in which such a text extract operates. Brown and Levinson's (1987) widely influential theory of politeness and positive/negative face provides for an approach to these values which is much more compatible with the dialogic framework I advocate below. They construe such values as providing a socially-motivated mode of negotiation, rather than being seen as simple reflexes of truth-value or epistemological reliability. They state, for example, that their theory, 'argues for a shift in emphasis from the current preoccupation with speaker-identity, to a focus on dyadic patterns of verbal interaction as the expression of social relationships' (1987: 2). In the current context, it is not possible to explore their framework so as to outline the specific points at which it might be compatible with the perspective I set out below and the points at which it might be less so. I do believe, however, that the dialogic account I propose can be seen as providing a descriptive grammatical apparatus by which to develop a systematic linguistic exploration of the types of social positionings with which Brown and Levinson are concerned.

4. Bakhtin and Voloshinov on Language as Dialogue The approach I propose for understanding the rhetorical functionality of these resources follows suggestions from Lemke (1992), Fairclough (1992), Thibault (1997) and Fuller (1998), all of whom make reference to Bakhtin and

Dialogue and Inter-Subjectivity:

Reinterpreting the Semantics of Modality and Hedging

71

Voloshinov's inter-connected notions of heteroglossia and dialogism (Bahktin 1973, 1978, 1981, 1986; Voloshinov 1995). My argument, as previewed above, is that we need to understand these meanings as resources for interpersonal negotiation and positioning, rather than by reference to truth value and epistemology and that we need to understand them in social rather than individualistic terms. That is to say, rather than seeing these meanings as necessarily reflecting individual, inner mental states of certainty or commitment, we need to see them as resources by which speakers, as representatives of social positions, negotiate a space for particular attitudes and points of view within the diversity of value positions operating in any speech community. Here I am specifically influenced by the insights of Bakhtin and Voloshinov when they insist upon the intertextual and dialogic nature of all texts, observing that all texts necessarily reference, respond to, interact with, anticipate and to greater or lesser extents incorporate other texts, both actual and prospective. The desire to make one's speech understood is only an abstract aspect of the speaker's concrete and total speech plan. Moreover, any speaker is himself a respondent to a greater or lesser degree. He is not, after all, the first speaker, the one who disturbs the eternal silence of the universe. And he presupposes not only the existence of the language system he is using, but also the existence of preceding utterances - his own and others' - with which his given utterance enters into one kind of relation or another (builds on them, polemicises with them, or simply presumes that they are already known to the listener). Any utterance is a link in a very complexly organised chain of other utterances. (Bakhtin 1986: 69)

Or, as Voloshinov observes, The actual reality of language-speech is not the abstract system of linguistic forms, not the isolated monologic utterance, and not the psychological act of its implementation, but the social event of verbal interaction implemented in an utterance or utterances. Thus, verbal interaction is the basic reality of language. Dialogue, in the narrow sense of the word, is of course only one of the forms - a very important form, to be sure - of verbal interaction. But dialogue can also be understood in a broader sense, meaning not only direct, face-to-face, vocalised verbal communication between persons, but also verbal communication of any type whatsoever. A book, i.e. a verbal performance in print, is also an element of verbal communication. ... [it] inevitably orients itself with respect to previous performances in the same sphere... Thus the printed verbal performance engages, as it were, in ideological colloquy of a large scale: it responds to something, affirms something, anticipates possible responses and objections, seeks support, and soon. (Voloshinov 1995: 139)

Additionally, the Bakhtinian/Voloshinovian framework stresses that the dialogue is not so much an interaction between self-determining individuals as between broader social positions. Bakhtin stresses the role of utterances in positioning speakers and their texts within the heterogeneity of social positions and world

72

Peter White

views which operate in any culture. All texts reflect a particular social reality or ideological position and therefore enter into relationships of greater or lesser alignment with a set of more or less convergent/divergent social positions put at risk by the current social context. Thus every meaning within a text occurs in a social context where a number of alternative or contrary meanings could have been made, and derives its social meaning and significance from the relationships of divergence or convergence into which it enters with those alternative meanings.

5. Inter-Subjective Positioning - A Dialogic View of Modality and Evidentiality When informed by this view of text as dialogic and heteroglossic, our approach to the rhetorical potential of values of modality and evidentiality is, predictably, different from that arising from the truth-functional, individualistic approach exemplified, for example, by Lyons' account. Rather than seeing these values as necessarily oriented to a speaker's internal state of mind, we will see them as operating to reflect the process of negotiation within a text between alternative socio-semiotic positions - we will see them as setting up different modes of dialogic interaction. Let's look at what insights this perspective provides on the rhetorical functionality of the modal and evidential values employed in the cited text extract. We can see that they are used in this particular text in order to construct an intersubjective stance which is highly conventionalised in such contexts of academic argumentation. Here we notice, most tellingly, that the use of modals and evidentials is carefully staged. There is a rhetorical transition by which the proposition at issue is introduced as 'improbable' and 'largely unacknowledged' but which, by the end of the paragraph, is represented as 'neither absurd' nor 'untimely' and as endorsed by the authority of the eminent Raymond Williams. This movement is clearly not from actual doubt, vagueness or epistemological unreliability to certainty and commitment - there is no shift in the stance the text adopts towards the central proposition. This movement, of course, is interpersonally and rhetorically motivated - a strategic response by the writer to the sense that his proposition will be seen as novel and contentious within the discourse community in which the text operates. The characterisation of the proposition as 'a consideration', as 'largely unacknowledged' and as 'occupying somewhat improbable terrain' is dialogic and intertextual. By such means, the writer constructs this text as coming after prior texts into which it enters into a relationship of tension or divergence. But the 'improbability' is not simply a product of the intertextual past. It also gestures towards the dialogic present and future - that is, towards anticipated responses to the text which, by the use of such meanings, the writer acknowledges as likely to be sceptical, questioning,

Dialogue and Inter-Subjectivity:

Reinterpreting

the Semantics of Modality and

Hedging

73

oppositional and so on. By such resources the writer acknowledges the heteroglossic diversity out of which the proposition's contentiousness emerges and within which he would negotiate a position for his current point of view. By doing so, the text acts to maintain the possibility of exchange and interaction with these divergent positions, thereby negotiating interpersonally in order to maintain the possibility of solidarity between the writer and those who would question, challenge or reject the key proposition. Here, then, is an instance where the use of modals and evidentials clearly has little to do with truth-functionality and much to do with arguability, interpersonal positioning and the negotiation of solidarity. These are the terms, under the 'dialogic' orientation adopted within 'Appraisal' 2 theory, in which these and related meanings are explicated. In general, then, we see these as 'heteroglossic' resources. They act to acknowledge in some way, and thereby to introduce into the text, the more or less divergent positions which the utterance in question responds to, or anticipates dialogically. I have, obviously, used just the one text to illustrate the dialogic. In the context of the current paper, not much more is possible by way of illustration and explication. The following sections will be taken up with a brief exploration of some other key theoretical insights which follow from the adoption of the dialogic orientation. I will return, by way of conclusion, to the original text extract to demonstrate how these insights enable us to better analyse its rhetorical potential.

6. No More Facts - The Inter-Subjectivity of the 'Objective' One final consequence of this dialogic approach needs to be noted. As indicated above, following Bakhtin/Voloshinov, we see all utterances as operating in an intertextual and heteroglossic context where they enter into relationships of greater or lesser tension with more or less divergent utterances. This perspective accordingly provides a potent counter to the common-sense notion that certain utterances are interpersonally neutral and hence 'factual' or 'objective' while others are interpersonally charged and hence 'opinionated' or 'attitudinaP. Under the heteroglossic orientation, we are reminded that even the most 'factual' utterances - those which are structured so as to background interpersonal values are nevertheless interpersonally charged in that they enter into relationships of tension with a related set of alternative and contradictory utterances. The degree of

2

It must be stressed that Appraisal theory is concerned broadly with the language of evaluation and inter-subjective positioning. The values that I am considering in this paper - those related to modality and evidentiality - constitute therefore but one sub-system of Appraisal. The theory seeks also to model the rhetorical potential of meanings by which speakers make moral and aesthetic judgements, by which they indicate emotional responses and by which they intensify their utterances.

74

Peter White

that tension is socially determined. It is a function of the number and the social status of those alternative socio-semiotic realities under which the utterance at issue would be problematised. Accordingly, the difference between the utterance 'Marx was a precursor of modern anthropological theory' and 'In my view, Marx was a precursor of modern anthropological theory' is not one of 'fact' versus 'opinion' but of the degree to which the utterance acknowledges the dialogic context in which it operates. The distinction, therefore, can be represented in terms of heteroglossic negotiation - the first utterance (the so-called 'fact') ignores or down-plays the possibility of heteroglossic diversity by dint of its grammar and lexis, while the second actively promotes that possibility. Alternatively, we can say the first denies or ignores the dialogic heterogeneity in which it operates while the second asserts it. The key insight here is that the positive declarative is not any less interpersonal and positioning than the array of alternatives by which the heteroglossic diversity is invoked, acknowledged or fended off. Rather, in using the positive declarative, the writer adopts a particular rhetorical strategy towards the possibility of heteroglossic diversity and dialogic interaction, namely of either choosing not to directly acknowledge that possibility, or of assuming a homogeneous rather than a heterogeneous speech community. The fundamental choice, then, in the context of grammatical options for managing inter-subjective position is thus between what we might term the monoglossic option (the so-called 'bare' declarative) and the network of resources by which we directly inscribe the possibility of social heterogeneity into the text.

7. The Dialogic View - Beyond Modality The dialogic approach, adopted within Appraisal theory, represents, therefore, a major departure from that found in the traditional modality, evidential and hedging literature. Its theoretical consequences, however, go well beyond this rather different view of the communicative/rhetorical functionality of these resources. Under this approach we see a significantly larger array or meanings as acting to position the speaker/author with respect to the various propositions and proposals conveyed by a text. The array of resources includes those of attribution/quotation, hearsay, concession, causality and negation, along with some additional semantic subsystems. These are all seen as resources by which the text interacts in some way with the view brought into play by its various propositions and proposals acting to acknowledge, invoke, challenge or suppress the heteroglossic diversity represented by those various divergent positions. Once again, there is not space here for a full account of this semantic domain. I will confine myself, therefore, to a brief discussion of how the dialogic perspective enables us to discover a connection of rhetorical functionality between

Dialogue and Inter-Subjectivity:

Reinterpreting the Semantics of Modality and Hedging

75

the types of modal and evidential values already discussed and the systems of attribution and hearsay. By attribution I refer to the citation of external sources by means of direct or indirect quotation and related mechanism. By hearsay I refer to the citation of non-specified external sources by means of structures such as it's said that.., it's claimed.., I hear..., and reportedly/allegedly. While values of hearsay are typically included within the repertoire of evidentials (see for example Chafe and Nichols 1986), attribution is much less consistently seen in these terms.3 Under the dialogic approach, however, both attribution and hearsay are unproblematically seen as resources for inter-subjective positioning and heteroglossic negotiation, and hence are linked functionally with modals of probability (epistemic modality). Let me briefly explore this point. Under the systemic functional approach, epistemic modality (probability) can take several forms - it is encoded by means of modal verbs (can, may, might, must), by adverbial adjuncts (possibly, probably, certainly, definitely), by adjectival impersonal constructions (it's possible that..., its probable that..., it's certain that...) and by certain mental-verb constructions (I think..., I reckon..., I guess..., I consider that..., I believe that... etc.4). It can also be communicated through a range of nominalised constructions - there's a possibility that...; in all likelihood...; the probability of him... and so on. In all such instances, the modal value acts to acknowledge dialogic interaction, to acknowledge heteroglossic alternation in that the proposition or proposal conveyed by the modalised utterance is thereby characterised as but one of a number of possible propositions or proposals. To state, for example, that 'Marx possibly has something to contribute to contemporary theory of culture' invokes, of course, alternatives such as 'Marx definitely has something to contribute to contemporary theory', or 'Marx doesn't have anything to contribute', or 'Marx possibly doesn't have anything to contribute' and so on. The modal, thereby acts to render the utterance contingent, as located in the subjecthood of a now foregrounded authorial voice. The degree to which the contingent, explicitly subjective authorial voice is foregrounded depends, of course, on the nature of the coding of the modal value. Foregrounding is maximal with what Halliday terms the interpersonal metaphor of the mental-verb constructions - for example, 'I suspect/reckon/consider/believe Marx has a great deal to offer'. Here the authorial source is directly inserted into the text. Foregrounding is progressively less with the alternative codings, though they all act to explicitly subjectivise the utterance, that is to construct the speaker's utterance as but one of a number of utterances which might have been made in the current heteroglossic context. 3 4

Chafe (1986) does very tentatively allow that attribution may have evidential properties. For a full discussion of the modality of such constructions, see Halliday 1994: 354.

Peter White

76

Attribution and hearsay directly acknowledge heteroglossic alternation and diversity by inserting into the text the words or meanings of a more or less precisely identified external voice. As in the case of modals of probability, the utterance in question is rendered contingent, characterised as but one of a number of possible utterances, by being associated with the subjecthood of a particular voice - that of the externally referenced source. Thus both attribution/hearsay and modals of probability introduce voices into the text (they render it heteroglossic) but stand in contrast in that, with probability, the voice is an internal one while, with attribution/hearsay, the voice is more or less external. Attribution and hearsay differ, of course, amongst themselves in the degree to which the voice is externalised. The rhetoric functionality of hearsay turns very much on the ambiguity of that externalisation. Structures such as 'its said that Marx has much to contribute'/'I hear Marx has much to contribute' are indeterminate as to the source and as to where the authorial voice stands with respect to the hearsay utterance. This cline of externalisation/internalisation operating across probability, hearsay and attribution is diagrammed below. [most externalised] (direct attribution) (indirect attribution) (hearsay) (impersonalised probability) (personalised probability) [most internalised]

The writer states that, 'Marx has much to offer contemporary cultural theory'. The literature generally holds that Marx has much to offer cultural theory. It's said that Marx has much to offer cultural theory. It's possible Marx has much to offer cultural theory. I reckon Marx has much to offer cultural theory.

We note also that attribution, hearsay and probability are alike in that they allow for grades or degrees of intensity, from low to high. Thus: low

median

high

probability

possible

probable

certain

hearsay

it's intimated that

it's said that

its asserted/it's been proved that

attribution

he claimed that.

he said that

he demonstrated that

Table 1: the cline of intensification across modality and attribution/hearsay The dialogic framework also offers insights into the rhetorical effects motivating choices amongst these graded options. All values of probability, hearsay and

Dialogue and Inter-Subjectivity:

Reinterpreting the Semantics of Modality and Hedging

11

attribution are dialogic in the sense that they acknowledge or invoke heteroglossic alternation - they all act to render the utterance contingent, to characterise it as but one of a number of alternative utterances currently available. Thus they all open up the text to heteroglossic diversity. The options differ, however, in the degree of that openness or expansiveness. The low values maximally open up the text to alternation - they afford maximal scope, rhetorically, for heteroglossic diversity. Thus 'its possible Marx has much to offer anthropological theory' is maximally expansive in the degree to which it invokes or acknowledges alternative positions. In contrast, the high values are relatively less expansive with respect to heteroglossic alternation. Thus while 'It's certain Marx has much to offer' still invokes alternative positions, the utterance nevertheless acts rhetorically to contract the heteroglossic space, to provide less interpersonal space for dialogic alternation.

8. Conclusion My purpose, then, has to been to argue for an approach to the analysis of modals, evidentials and related values which is informed by the dialogic perspective of Bakhtin/Voloshinov. Let me now return to the original text extract to illustrate, somewhat more comprehensively, the types of insights this approach can provide. I noted above that the central proposition of the extract is typically characterised as 'a consideration': This consideration of Marx as a precursor... of the modern anthropological theory of culture... Yet such a consideration is neither absurd nor untimely.

From the perspective of Appraisal theory, we would say the utterance has been probabilised - that is to say, it has been modalised by reference to some degree of likelihood. Specifically, the construction is capable, at least in one interpretation, of being related to mental-verb projections such as, 'I think/ reckon/believe/consider Marx was a precursor of the modern anthropological theory of culture'. As already discussed, such structures render the proposition contingent by associating it with the foregrounded subjectivity of the authorial voice. They thereby open up rhetorical space for heteroglossic alternatives. Of course, here, the author has exploited the resources of nominalisation to complicate the interpersonal positioning somewhat. By means of the nominalisation, the author suppresses or renders non-specific the source of the mental process of 'considering'. The precise rhetorical functionality of the utterance is thereby made somewhat more ambiguous or under-specified since the contingent subjecthood in which the proposition is now located is neither clearly external or internal to the text. Nevertheless, the use of this probabilising value acts to construct the text as dialogic, as explicitly opening up the text to

78

Peter White

heteroglossic alternatives. As well, we notice that, on the cline of low-to-high values outlined above, 'consideration' is down the lower end of the scale of intensity, contrasting with higher values such as 'the conviction that' or 'the knowledge o f . Thus the utterance is maximally expansive in heteroglossic terms, in the degree to which it invokes alternative positions. Such a choice of wording is, of course, strategic. It makes a space for a contentious and novel proposition by clearly characterising that proposition as one of but a number of possible related propositions, by thereby anticipating a dialogue with those who hold these alternative positions and hence signalling a willingness to maintain academic solidarity with those positions. This strategy of opening up the heteroglossic space is, in general terms, maintained across the opening few sentences of the extract. Accordingly, the proposition is said to be 'a largely unacknowledged one'. The resource employed is that of hearsay, but once again the rhetorical effects are complicated, here by the addition of negation. By characterising the proposition as 'unacknowledged', the text emphasises its minority status and thereby overtly signals a willingness to acknowledge and engage with divergent, majority points of view. Interestingly, the writer is not prepared to leave it at that. He refines the rhetorical strategizing by the addition of 'largely' - he offers, not, 'though an unacknowledged one' but 'though a largely unacknowledged one'. By the use of such a value, the text, of course, acts to preserve some heteroglossic space for its central proposition - the 'non-acknowledgement' is not universal. The various probabilising values which follow serve a similar rhetorical purpose. Thus the characterisation of the proposition as 'improbable' and 'unlikely' acts to maximally expand the dialogic space for alternative positions. It is only in the extract's final clause that this strategy of maximal heteroglossic expansion is adjusted. The proposition is now asserted as 'neither absurd nor untimely' and as 'demonstrated' previously by 'Raymond Williams'. The movement is from expanding the space dialogically for alternative positions to significantly contracting it. How is this contraction achieved? Firstly, the text directly challenges certain alternative positions - that the proposition is absurd or untimely - by directly denying them. Such denial is still heteroglossic in the sense that negation invokes the alternative position that it contradicts. The alternative position is thus inscribed in the text, even though it is then rejected. (The view that the negative is interpersonally charged and that it invokes the positive, is, of course, relatively widespread in the literature. See, for example, Pagano 1994 and Fairclough 1992: 121). What is, perhaps, the strongest assertion of the text extract's central thesis is found in its final clause, where the authority of the eminent Raymond Williams is invoked to 'demonstrate' its validity. Citation is, of course, explicitly

Dialogue and Inter-Subjectivity:

Reinterpreting the Semantics of Modality and Hedging

79

heteroglossic - it introduces an external voice - and hence renders the proposition contingent and explicitly subjectivised. But this particular citation is formulated so as to employ a specific rhetorical strategy. It is formulated so as to significantly contract the dialogic space available for alternative positions. Firstly, the choice of the term 'demonstrate' locates the utterance at the high end of the cline discussed above - it is a 'high' value which serves to contract the heteroglossic space. By choosing 'demonstrate' over, for example, 'state', 'say', 'argue', 'claim' or 'advocate', the role of the authorial voice is foregrounded in the text as endorsing the cited proposition. The invoking of 'Raymond Williams' has a similar rhetorical functionality. Such an authoritative, widely quoted source carries with it a similar 'high' rhetorical force, once again serving to contract the space for heteroglossic alternation. The demonstration of the dialogic framework in action in textual analysis has been necessarily brief, confined as it has been to just the one paragraph of text. I would hope, nevertheless, that it has demonstrated the sorts of insights of which the approach is capable and the degree to which these differ from those which follow from the more traditional frameworks.

References Bakhtin, M. (1973): Problems ofDostoevsky's Poetics. Michigan: Ann Arbor. Bakhtin, M. (1978): Esthétique Et Théorie Du Roman. Paris: Gallimard. Bakhtin, M. (1981): The Dialogical Imagination, M. Holquist, (éd.), C. Emerson and M. Holquist (trans.). Austin: University of Texas Press. Bakhtin, M. (1986): The Problem of Speech Genres. In: Speech Genres and Other Late Essays, McGee, V. W. (trans.). Austin: University of Texas Press: 60-101. Bernstein, B. (1970): Class, Codes, and Control. Volume 1 : Theoretical Studies Towards a Sociology of Language. London: Routledge and Kegan Paul. Brown, P. and Levinson, S. (1987): Politeness: Some Universals in Language Usage. Cambridge, U.K.: Cambridge University Press. Chafe, W. L. and J. Nichols, (eds.), (1986): Evidentiality: The Linguistic Coding of Epistemology. Norwood, N.J.: Ablex. Chafe, W. L. (1982): Integration and Involvement in Speaking, Writing and Oral Literature. In: Spoken and Written Language: Exploring Orality and Literacy, Tannen, D. (ed.), Norwood, N.J.: Ablex: 35-54. Chafe, W. L. (1986): Evidentiality in English Conversation and Academic Writing. In: Evidentiality: the Linguistic Coding of Epistemology Chafe, W. L. and Nichols, J. (eds.), Norwood, New Jersey, Ablex Publishing Corporation. Channell, J. (1994): Vague Language. Oxford: Oxford University Press. Christie, F. and Martin, J. R. (eds.), (1997): Genres and Institutions: Social Processes in the Workplace and School. London: Cassell. Dentith, S. (1995): Bakhtinian Thought, An Introductory Reader. London: Routledge. Fairclough, N. (1992): Discourse and Social Change. Cambridge, UK: Polity Press. Fuller, G. (1998): Cultivating Science: Negotiating Discourse in the Popular Texts of Stephen Jay Gould. In: Reading Science - Critical and Functional Perspectives on Discourses of Science, Martin, J. R. and Veel, R. (eds.), London: Routledge.

80

Peter White

Halliday, M. A. K. (1994): An Introduction to Functional Grammar. London: Edward Arnold. Holmes, J. (1984): Hedging Your Bets and Sitting on the Fence. Some Evidence for Hedges As Support Structures. Te Reo 27: 47-62. Iedema, R., S. Feez, and P. R. R. White (1994): Media Literacy, Sydney, Disadvantaged Schools Program. NSW Department of School Education. Jakobson, R (1957): Shifters, Verbal Categories and the Russian Verb. Russian Language Project. Harvard: Dept of Slavic Languages and Literature. Kessler, C. S. (1987): Marx As Cultural Theorist: the Prehistory of Modern Anthropology. In: Creating Culture-Profiles in the Study of Culture. Austin-Brooks, D. J. (ed.), Sydney, Allen and Unwin: 35-49. Lakoff, G. (1973): Hedges: A Study in Meaning Criteria and the Logic of Fuzzy Concepts. Journal of Philosophical Logic 2: 458-508. Lemke, J. L. (1992): Interpersonal Meaning in Discourse: Value Orientations. In: Advances in Systemic Linguistics. Recent Theory and Practice, Davies, M. and Ravelli, L. (eds.), London: Pinter Publishers. Lyons, J. (1977): Semantics. Cambridge: Cambridge University Press. Markkanen, R. and Schröder, H. (eds.), (1997): Approaches to the Analysis of a Pragmatic Phenomenon in Academic Texts. The Hague: Walter De Gruyter and Co. Martin, J. R. (1997): Analysing Genre: Functional Parameters. In: Genres and Institutions: Social Processes in the Workplace and School. Christie, F. and Martin, J. R. (eds.), London: Cassell: 3-39. Martin, J. R. (2000): Beyond Exchange: APPRAISAL Systems in English. In: Evaluation in Text. Hunston, S. and Thompson, G. (eds.), Oxford: OUP, 143-175. Pagano, A. (1994): Negatives in Written Text In: Advances in Written Text Analysis. Coulthard, R. M. (ed.), London, Routledge. Palmer, F. R. (1986): Mood and Modality. Cambridge: Cambridge University Press. Prince, E. F., Frader, J., and Bosk, C. (1982): On Hedging in Physician-Physician Discourse. In: Linguistics and the Professions, di Pietro, R. J. (ed.), Norwood: Ablex. Thibault, P. (1997): Mood and Eco-Social Dynamics. In: On Subject and Theme: A Discourse Functional Perspective. Hasan, R. and Fries, P. (eds.), The Hague: Benjamins: 51-90. Voloshinov, V. N. (1995): Marxism and the Philosophy of Language, Bakhtinian Thought - an Introductory Reader, S. Dentith, L. Matejka and I. R. Titunik (trans.). London: Routledge. White, P. R. R. (1997): Death, Disruption and the Moral Order: the Narrative Impulse in MassMedia Hard News Reporting. In: Genres and Institutions: Social Processes in the Workplace and School, Christie, F. Martin, J. R. (eds.), London: Cassell. White, P. R. R. (1998): Telling Media Tales: the News Story As Rhetoric. Unpublished PhD dissertation, University of Sydney.

Barry Natusch

Talking for a Living on Air 1. Introduction The people who lead the dialogues analysed in this study are professional media interviewers. Interviews can range from stiffly formal, earnest exchanges (see Rogers 1974, and Wales 1989) to ones which have a chatty conversational style (Wardhaugh 1985). Such chatty, informal dialogues are the focus of this discussion. Dialogues discussed will be "on air" that is publicly broadcast. I was watching an exchange between Larry King and Dr Joy Powers on CNN TV recently, the one an interviewer, the other a radio counsellor. They seem like old friends, sitting there talking about human relationships. A woman caller comes on the line, says she's in a relationship where her partner sometimes beats her. "You've got to leave him," says Joy decisively. "But I LOVE him," says the woman. "Are you dependent on him for money?" asks Joy. "Well, yes," admits the woman. "That's the reason you think you love him. Get a job and LEAVE him," responds Joy. Larry cuts off the caller, asks Joy how she was able to cut through to the heart of a problem so decisively.... Joy replies that it was something she actually learned from Larry himself many years ago, when he urged her to be more assertive and to push her own views forward. It is a satisfying conversation. It seems a very natural exchange, progressing smoothly and economically with some useful advice being given. Larry looks pleased at being praised and Joy is pleased that she's made Larry feel good. Larry has done a good job of managing the interaction. It is also a good example of an interview which falls midway between a formal interview and a chatty conversation. In a formal interview, one person elicits information from the other. It is an unequal, time-limited, dyadic communication in which the interviewer controls the topics discussed. (Rogers, 1974). Silvester (1993) notes that the interviewer may follow different strategies to elicit information, such as by playing the psychoanalyst (being non-judgmental, saying little so the other party feels obliged to open up) or to talk a lot, to disclose facts about oneself so that the other party, obeying the rules of conversation, also feels obliged to offer confidences. In this latter case, the interview approaches the structure of a conversation where there is a balance of talk between the two parties, and in fact the roles of interviewer and interviewee may rapidly switch. By contrast, a conversation is a shared exchange with each partner having the freedom to choose topics, hold the floor, steer the direction, or disclose

Barry Natusch

82

information. The following discussion is concerned with conversations on air, that is where the interactants speak to each other for a roughly equal amount of time. It is less concerned with interviews, although initially an interview will be referred to for comparative purposes. In order to highlight the differences between private and public conversations, the following two differences will be focused on: (1) the satisfaction in engaging in and listening to conversation, and (2) the management of performed conversation on public radio broadcast.

2. Feelings of Satisfaction in Everyday Conversation Gutman: Spade: Gutman:

Well, sir, here's to plain speaking.... You a close-mouthed man? Mnn. I like to talk. Better and better. I distrust a close-mouthed man. He generally picks the wrong time to talk and says the wrong thing. Talking's something you can't do judiciously unless you keep in practice. Now sir, we'll talk if you like. I'll tell you right out I'm a man who likes talking to a man who likes to talk. (John Huston, The Maltese Falcon, 1941)

Table 1 summarises what people feel are the elements of engaging in a satisfying conversation with someone else. Twenty people were surveyed, rather unscientifically, by simply being asked the open-ended question, "What do you consider are the elements of a satisfying conversation with someone else?" These respondents were aged between 21 and 57 with the group comprising 16 Caucasian and 4 Japanese. Topic: what's talked about

Management: how talk is handled and how the thread progresses

Affective-Interpersonal : feelings between conversationalists

useful information other people's behavior funny events future planning common experience or problemmemories of the past two-way exchange 30/70 > 50/50 not too broad, not too detailed understanding aided by examples smooth progression, not too jumpy not knowing how the exchange will turn out witty, creative word play repartee elements of challenge partner listens feedback occurs eye contact, no distractions body language complementary feelings of sympathy mutual flattery

Table 1 : Feelings of satisfaction in everyday conversation

83

Talking for a Living on Air

People may well have differing criteria for judging whether a conversation is satisfying to them or not, clearly, not everyone favors the same way of talking. All respondents liked talk to have content, the feeling of having learned something new, in their exchanges. Two remarked that they judged a conversation to be successful if they were still thinking about the topic a day or two later. All respondents liked to be both a listener and a contributor. Three people liked a structure to the exchange, a feeling that there was a beginning, middle and end, while others said they liked to talk about many different things, to change the topic frequently. One female said she ideally liked a 50-50 split but depending on the person would go as far as 30/70. Two admitted to enjoying conversation when it turned out unpredictably, that they entered conversations with a frisson of anticipation not knowing where talk was going to lead them. One woman phrased it elegantly when she said she liked to risk a relationship by breaking new ground. Several Australasian males said they enjoyed the challenge of sparring, teasing, provoking the other party by trading insults. One North American said he liked having competitive word games in conversation. Relating to affective/interpersonal concerns, intercultural differences may be a factor determining preferences. One U.S. male admitted to enjoying interactions where praise was given and received (in his words, "mutual flattery was exchanged"). This was also a feature of the exchange between Lany King and Joy Powers cited above. A Japanese female noted she liked to talk to someone who shared her views and didn't disagree with her (though this may have been a personal, rather than a cultural distinction). Several commented that they liked to have the undivided attention of the other party: listening, feedback being given, eye contact made (even among the Japanese), that the parties' body language did not differ radically.

3. General Considerations for On Air Conversations Well, well, perhaps I am a bit of a talker. A popular fellow such as I am - my friends get round me - we chaff, we sparkle, we tell witty stories - and somehow my tongue gets wagging. I have the gift of conversation. I've been told I ought to have a salon, whatever that may be. (Kenneth Grahame, Toad, in The Wind in the Willows, 1908)

Some radio and TV presenters are a pleasure to observe demonstrating their conversational virtuosity in talking with others. These presenters may be thought of as running on-air salons (in the sense of a periodic gathering of people for social or intellectual discussion). Such salons are indeed large, being as extensive as a country's nationally broadcast programme. In Table 2 the general considerations for an on-air exchange are described. At the preparation stage, the presenter needs to decide on a suitable conversational partner, has to find out about that person, and decide what to talk about.

84

Barry Natusch

preparation (getting to know you) "The best interviews - like the best biographies should sing the strangeness and variety of the human race." Lynn Barber (1991) overall planning (choreographing the conversation) "I always find plans are useless, but planning is indispensable." Dwight Eisenhower ongoing management (putting out fires) "The opposite of talking isn't listening. The opposite of talking is waiting." Fran Lebowitz (1981) interpersonal (playing the puppet master) "There comes a point in many people's lives when they can no longer play the role they have chosen for themselves. When that happens, we are like actors finding that someone has changed the play." Brian Moore (1990) extras (icing the cake) "Wit is the salt of conversation." William Hazlitt

selecting a partner establishing common ground choosing a topic

initiation agenda items closing dealing with responses monitoring output

role play competing external directing

ornamenting talk body language follow up Table 2: General considerations for On Air Conversations There is likely to be some plan driving the conversation, and an analogy here might be choreographing the steps of a dance. A talk needs a clear initiation phase, and introducing the speaker often involves reading some background text. The more successful interviewers read with what Brazil calls "full engagement" (Brazil 1992) making it sound as if it is not being read. There is a list of points to be covered, often phrased as questions. Moves often follow both two move (initiation-response) and a three move structure (initiation-response-reaction) (Sinclair 1992), with initiations being any utterance that requires a response, not just questions (Tsui 1992). Finally, there is often a formulaic closing once the time has been filled or the informational expectations have been satisfied. While the conversation progresses, because there is an audience listening, talk needs to be actively managed to ensure that it flows smoothly and is interesting. To achieve this, speakers need to listen to their own phrasing and to respond quickly and appropriately to the other person's comments and questions, more so than in everyday conversations. An on-air conversation is a kind of theatrical performance. Certain theatrical artifice is therefore discernible in on air conversations in the form of dramatic form (prologue, intermission, climax, epilogue), characterization (ranging from participants taking on roles to showmanship and even staged histrionics),

85

Talking for a Living on Air

scriptwriting (in the form of notes and quotations), external direction (from the control room). With such preparation it is not surprising that on air conversations are much richer than everyday conversations, that they are actually exchanges planned to maximise dramatic intent.

4. Differences between On Air and Everyday Conversations To make your dialogue sound natural, make it artificial.

(Kress 1994)

Four conversational style interviews broadcast on Radio New Zealand between participants well known to each other were taped and transcribed. The first was an interview conducted semi-formally by Brian Edwards to elicit information from Edward De Bono. Two other informal interviews were conducted between Brian Edwards and a media personality (Gary McCormick) and a successful businessman (Bob Jones). The fourth interview, also informal, was between a female presenter, Kim Hill, and Gary McCormick. Due to limitations of space, full transcripts are not appended here. It has often been observed that the nature of everyday conversation is untidy and unstructured. Transcript data suggests that on-air conversations appear to be more tightly structured and flow more smoothly than everyday conversations, although the end result is nonetheless designed to project the image of being a spontaneous unprepared conversation. This seems to be achieved on several fronts as summarised in Table 3. It is emphasized that these are impressions gained from listening to tapes of both on-air and casual conversations and the observations made are offered here as hypotheses to be statistically tested. Transcript Key 1 2 3 4

Brian Edwards and Edward De Bono Radio New Zealand 17/6/95 Brian Edwards and Gary McCormick Radio New Zealand 11/1/96 Brian Edwards and Bob Jones Radio New Zealand 28/10/95 Kim Hill and Gary McCormick Radio New Zealand 18/1/96 on air conversation

selecting a partner 'What ho!' I said, 'What ho!' said Motty. 'What ho! What ho!' 'What ho! What ho! What ho!'After that it seemed rather difficult to go on with the conversation. (P.G. Wodehouse)

intentional, set up before by others, dependent on performance ability, topicality, fame, knowledge, audience expectations

examples from transcripts

everyday conversation casual, r a n d o m meeting of family, friend, colleague

Barry Natusch

86 common ground "Two may talk together under the same roof for many years, yet we never really meet; and two others at first speech are old friends." (Mary Catherwood)

background reported

choosing a topic

public issues, useful information, so prepared that some facts even read out questions, quotations, comments

initiating "Questioning is not the mode of the conversation among gentlemen." (Samuel Johnson)

handling subtopics "Conversation should touch everything, but should concentrate itself on nothing." (Oscar Wilde)

smooth structured progression, almost narrative style

closing "All farewells should be sudden." (Lord Byron)

usually abrupt

monitoring output "For one word a man is often deemed to be wise,

less evidence of repair

Edwards: Well I first met my next guest on a late night television program I was running TWENTY years ago. Edward de Bono, the man who devised the concept of lateral thinking had COME on the program to explain what lateral thinking WAS...

not usually commented on

unprepared, arise spontaneously Edwards: I'd like to start by talking about humour and intelligence if we CAN, er in my office at HOME, I have a a thing on the wall, it's a QUOTE from Clive James who was reviewing an Arnold Bennett play and my listeners will have heard it before but this is what Clive James said... McCormick: Indeed it is. And I - 1 the reason I's talking about, I'm telling you about this shirt button is that there's a parallel to my woodworking days when... Hill: Thank you Gary. Nice to talk to you AGAIN. Gary McCORmick. It's MIDDAY now. De Bono: So I think there's a pretty good close correction -

fewer formal markers

random, backtracking, sidetracks, false starts, etc

frequently protracted

often more repair

87

Talking for a Living on Air and for one word he is often deemed to be foolish. We need to be careful what we say." (Confucious) external directing

role play (dressing up, putting on a mask) "The real test of a man is not how well he plays the role he has invented for himself, but how well he plays the role that destiny assigned to him." (Jan PatoCka) ornamentation "It is not what we learn in conversation that enriches us. It is the elation that comes of swift contact with tingling currents of thought." (Agnes Repplier)

body language "Never try to look into both eyes at the same t i m e . . . . Switch your gaze from one eye to the other. That signals warmth and sincerity." (Dorothy Sarnoff)

connection.

hidden cues

Hill: Oh, they should have courses, 12 step courses on how to turn nasty, in 20 seconds

little external directing

performance oriented roles: adversary, challenger, inquisitor, devil's advocate

McCormick: You see this is the thing that a lot of people don't understand and probably only fashion designers like mySELF understand...

personal identity generally prevails

professionals performing for a huge unseen audience thus denser incidence of witty phrase expected

Hill: I said I don't like the idea of these collars because cats climb trees and they get you know and WHY don't we see trees fesTOONed with dangling cats cause they've all got collars these days and THEY said cause they've all got these safety catches on them now. McCormick: Did you read the other day in the paper that canTANKerous, WILful people, I don't why I'm looking at you and saying this but canTANKerous, WILful people live LONGER by and large?

wit and cleverness may occur at greater intervals

hidden from audience, sometimes reported

Table 3: Differences between on air and everyday conversations

visible, n o need to report

88

Barry Natusch

The emphasis on structure and organization in the exchange begins from the selection of a conversational partner who will perform well, and have something worthwhile to say, whereas this is less important in a private conversation which may occur at random with anybody. At the beginning of an on air conversation the host will introduce the other speaker. This usually includes a description of their career, their background, their interests and may also make specific mention of the relationship between the two where relevant, a header which is not included in everyday conversations. This introduction is often read out from a prepared script. Initiating the on air exchange is often marked by a clear signal ("I'd like to start by..."), direct questions and sometimes reference to other opinions through quotations ("In my office at home I have a QUOTE from Clive James..."), whereas there is less of this in everyday conversations. The same factors defining a satisfying conversation between friends (Table 1) appear to be present in on air conversations. Stories play a major role, and are generally narrated very smoothly, suggesting that often they may have been rehearsed. Informal on air talks demonstrate a somewhat different pattern of exchange to more formal interviews. Informal exchanges often contain bursts of sparring repartee which border at times on being point-scoring exchanges: McCormick: Edwards: McCormick: Edwards:

I just think what an eNORmous amount of TAlent we have in this country. Well, there's you and me for a START well, indeed... though I'm not enTIREly sure about YOU.

This kind of trading insults questioning abilities is a common feature of conversations between Australian males. In male-female teasing exchanges however, gender attitudes may also be directly attacked: Hill: McCormick: Hill:

I think that's probably because cats are too smart, rather than not smart enough, don't you think? Well, that's possibly true but from terms of satisfaction wouldn't it be BETTER to have an animal that you can teach to roll over. That's very MALE of you Gary.

If the discourse features more story-telling and sparring, however, the density of factual informative reporting tends to diminish. Topic shift is also often clearly marked by signals ("The reason I'm telling you about this shirt is there's a parallel to my woodworking days") whereas casual conversation may often be more abrupt. Punchlines are crisp ("It's the LEG chains they have difficulty with but the collars are fine") and may also signal topic shift.

Talking for a Living on Air

89

Topic shifts are sometimes accompanied by references back to summarise, or perhaps to gain a breather while a new idea germinates ("I mean you're in a house with a RAT and CAT AND a small daughter and you know ... you've got to spend some time on the ironing and vacuuming.") In terms of abruptness, however, it is on air closings which are more sudden than many which occur in casual conversation, due of course to the time constraints in radio or TV programming. Hill curtails her exchange with a typical radio broadcast closing: "Thank you Gary. Nice to talk to you AGAIN. Gary McCORmick. It's MIDDAY now." There is sometimes evidence of external direction, one example being when Hill embeds a hidden message in her talk ("Oh, they should have courses, 12 step courses on how to turn nasty, in 20 seconds"), the reference to 20 seconds actually being a reminder as to how much on air time remains. Since on air talk is a performed drama, it is not surprising to find participants taking on roles more frequently than they do in everyday conversations to heighten the dramatic effect. In the space of 30 seconds the interviewer Edwards adopts three different roles (tease, political interviewer, medical authority). Possibly the awareness that one is performing to a larger audience results in attempts to employ interesting phrases as much as possible (eg "trees fesTOONED with dangling cats..."). Repair of output in the interests of clarifying while it certainly does occur in on air exchanges, seems to be less evident in on-air conversations, as might be expected when two professional speakers exchange views. Another curiosity of radio conversations is that since the audience cannot see body language, this is sometimes reported as well as performed ("Did you read the other day in the paper that canTANKerous, WILful people, I don't why I'm looking at you and saying this but canTANKerous, WILful people live LONGER by and large?"). And because there is an audience a speaker might acknowledge them sometimes (eg "My listeners will have heard this before but...."). Obviously, there are many instances of discourse which occur equally on air and privately between two people conversing. We all know people whose talk is a kind of performance; they take on roles even when they only have an audience of one. Nevertheless, broadcast conversations generally tend to be fast paced, with little waste, the exchanges are lean and sometimes mean. Natural conversation seems to be conducted at leisure, for pleasure, and at times, this is markedly clear when transcribed, it seems rather untidy. On air exchanges are more demanding than just having a casual chat with a friend. There needs to be a higher density of information, chosen and packaged for public consumption, often by means of telling a story. Talk needs to be tightly organized, tidier, with fewer regressions, repetitions, etc. than an ordinary conversation, and performance needs to be tuned up to maintain interest and the sense of suspense. To achieve this broadcast chat incorporates some elements of dramatic form even if it is not scripted. For that is

90

Barry Natusch

what on air talk is essentially: a crafted performance, artificial in order to appear natural.

5. Conclusion The conclusions highlight two main axioms about conversing on radio: that conversationalists are aware (1) that there is an audience, and (2) that their talk is a performance. On air talking is a somewhat externally managed performance driven by the expectations of the participants, their managers, PR people and the audience, rather than an internally driven agenda drawn up and evolving between two people. There is pressure on the conversational participants to perform in a conversation which is not just satisfying to them personally but provides satisfaction to the wider audience's expectations. Because this is a performance art, careful consideration is given to selecting a person who makes a good on air conversationalist. Participants are selected for their ability to perform in terms of informing, storytelling and sparring, not for the same reasons that one might choose a friend. Brian Edwards chooses people who are good at story telling, who don't get bogged down in detail, have energy and enthusiasm, who can give the impression that they are not aware that there is an audience listening in (Edwards, 1999). Transcript data suggests that there are other, dramatic arts related skills too: verbal dexterity, being quick-witted and the ability to take on different roles are important too. On air exchanges differ energetically from natural conversations. To take a sports analogy: a conversation between friends is like a game of croquet, on a lawn, stopping at times for tea, for a chat while leaning on the mallet, who cares if a shot is missed. An on-air dialogue is akin to championship golf, the crowds lining the course, where every shot counts and the greatest economy of strokes being the most valued. The most skilled make it look so effortless that it sounds like unrehearsed spontaneous authentic conversation, while in reality it is a carefully orchestrated performance.

References Barber, L. (1991): London Independent on Sunday (24 Feb. 1991). Brazil, D. (1992): Listening to People Reading. In: Coulthard, M. (ed.) Advances in Spoken Discourse Analysis. London: Routledge. 209-241. Byron, G. (1992): Quoted in Daintith, John and Anne Stibbs (eds.) Quotations for Speeches. London: Bloomsbury 76. Catherwood, M. (1992): Quoted in Daintith, John and Anne Stibbs (eds.) Quotations for Speeches. London: Bloomsbury 53. Confucious: The Analects. Coulthard, M. and D. Brazil (1992): Exchange Structure. 50-78. In: Coulthard, M. (ed.) Advances in Spoken Discourse Analysis. London: Routledge.

Talking for a Living on Air

91

Edwards, B. (1999): personal communication. Grahame, K. (1908): The Wind in the Willows. London: Bloomsbury Books. Huston, J. (dir) (1941): The Maltese Falcon. Johnson, S. (1791): In: Boswell, James: Life of Samuel Johnson, 25 March 1776 (1791). Kress, N. (1994): Say What? Writer's Digest, April 1994, 8. Lebowitz, F. (1981): Social Studies, "People" quoted in The Columbia Dictionary of Quotations. Columbia University Press. Moore, B. (1990): Sunday Times (London, 15 April 1990). Patocka, J. (1986): In: Havel, Vaclav: Disturbing the Peace, ch. 2, 1990. In: Czechoslovakia, 1986. quoted in The Columbia Dictionary of Quotations, 1993 Columbia University Press. Repplier, A. (1904): The Luxury of Conversation. Rogers, D. (1974): The Interview: Special Form of the Dyad. ERIC document. ERIC NO: ED089388 Sarnoff, D. (1988): Times (London, 27 April 1988). Silvester, C. (1993): The Penguin Book of Interviews: An Anthology from 1859 to the Present Day. Harmondsworth: Penguin. Sinclair, J. (1992): Priorités in Discourse Analysis. In: Coulthard, Malcolm (ed.) Advances in Spoken Discourse Analysis. London: Routledge. 79-88. Tsui, A. (1992): A Functional Description of Questions. In: Coulthard, Malcolm (ed.) Advances in Spoken Discourse Analysis. London: Routledge. 89-110. Wales, K. (1989): A Dictionary of Stylistics. London: Longman. Wardhaugh, R. (1985): How Conversation Works. Oxford: Blackwell. Wilde, O. (1891): in Gilbert, The Critic as Artist, pt. 2 (published in Intentions, 1891). Wodehouse, P. G.: Jeeves and the Unbidden Guest, quoted in Cohen J. M. The Penguin Dictionary of Quotations, 2nd edition Harmondsworth: Penguin.

Arja

Piirainen-Marsh

Interjections and the Institutionally of Broadcast Talk1 Research on the social and interactional organization of broadcast talk has drawn attention to different ways in which representatives of the media institution manage the tasks and activities typically found in different broadcast formats. Studies of news interviews, political discussions or debates and talk shows have uncovered a variety of practices through which journalists may elicit, attend to, direct, (re)present, and control the production of meanings in broadcast discourse (see e.g. Scannell 1991, Heritage and Greatbatch 1991, Hutchby 1996). This paper considers one aspect of this kind of interactional conduct in a specific broadcast setting. It examines how interviewers or hosts in multiparty interviews and discussions intervene in the flow of talk in order to formulate a particular understanding of it, or to selectively develop or direct subsequent talk in accordance with relevant institutional purposes or agendas. In line with previous work on institutional talk and discourse (e.g. Boden and Zimmerman 1991, Drew and Heritage 1992, Heritage 1997), the analysis focuses on details of language use at the local level of turn-design and the sequential organization of talk. These are examined with a view to throwing light on the ways in which options taken by speakers show an orientation to the 'institutionality' of talk, particularly as it is reflected in asymmetries of resources and participation rights. In broadcast interaction it is the journalists' task to elicit, produce and mediate talk, which is in some way informative, interesting or entertaining to the target audience. Although forms of talk and discourse may vary across different broadcast formats, these tasks are typically accomplished by employing techniques of questioning and other practices inviting responses from co-present participants (Heritage and Greatbatch 1991, Heritage and Roth 1995). Alternatively, journalists may participate in on-going talk by employing formulations, metastatements (Heritage and Watson 1979, Heritage 1985) or parenthetical comments and remarks, which shape the construction of meaning and guide participation in different ways. In this paper I will examine some of the latter kinds of resources used by hosts in multi-party broadcast settings. I investigate how journalists sometimes interject into a current speaker's turn or intervene in the topical flow of talk in order to draw attention to some aspect of prior talk, formulate a particular understanding of it, or to selectively direct or develop subsequent talk in some way. 1

I am grateful to Paul Drew for his insight and support during the preparation o f this paper at the University of York.

94

Arja Piirainen-Marsh

The data examined here consist of multiparty television discussions and interviews produced by the Finnish broadcasting company and broadcast between 1994 and 1996. The interviews and discussions involved participants from different linguistic and cultural backgrounds and were conducted in English. The settings are thus rather complex: the broadcast discourse events involve 5 - 7 participants, whose backgrounds and linguistic resources may vary considerably. While talk in these interactions is subject to the constraints of timing, editing and other restrictions arising from the media setting, it is also shaped by asymmetries arising from the diversity inherent in multiparty, multicultural settings.

1. Interjections as an Interactional Phenomenon I have adopted the notion of interjections here to refer to a group of practices, which are carried out in turns that in some way intervene in the flow of talk. My use of the term is different from its technical use in linguistics, where the term interjection is associated with exclamations, expressions of emotion and other elements, interjacent in but distinct from syntactic units such as clauses (see e.g. Quirk, Greenbaum, Leech and Svartvik 1985). The term is not used here in any technical sense, but rather in accordance with some dictionary definitions which point to the conversational phenomenon of 'throwing in' remarks, expressions of feeling or parenthetical comments in between or among other (conversational) turns. Although the practices I examine below have connections with the interactional phenomena investigated under interruptions (Hutchby 1996) and other 'incomings', which may be associated with competition for turns (French and Local 1983), they are not inherently competitive or interruptive. Although they frequently involve overlap, they do not necessarily occur in sequentially incursive positions, but may involve more collaborative forms of simultaneous speech and concurrent activities. For the purposes of this paper, interjections are characterised as utterances which may (a) interject into a current speaker's turn and be produced in overlap, (b) come in at points where the prior turn is not grammatically or pragmatically complete, or (c) occur at points where several participants are engaged in talk on a particular topic, often involving overlap. In these data, their starting point is frequently in positions which are not sequentially incursive, but rather indicative of overlap where speakers already project a possible completion point or are able to recognise the content or direction of the turn in progress (cf. Hutchby 1996, Jefferson 1986). Linguistically, interjections may take a variety of forms, ranging from brief parenthetical remarks and reactions consisting of single words or phrases to grammatically more complex utterances, such as clauses which make some observation about or express some stance towards prior talk. Interactionally, interjections are strongly tied - even parasitic - to turns and actions that they follow, and typically emerge when speakers use the resources made available in

Interjections and the Institutionality of Broadcast Talk

95

prior talk to make an observation or interpretation of the content of most recent talk. In this way they are distinct from other activities frequent in broadcast interviews and discussions, e.g. questions. In what follows, I will present some observations about two broad types of interjection used by hosts or interviewers in the present data. While they are not representative of all forms of interjection in the data, they point to some characteristic ways in which hosts in multiparty broadcast talk actively and selectively influence or control the construction of meanings and interpretations in broadcast talk. The first group of interjections described below consists of utterances through which hosts or interviewers formulate understandings of prior talk at specific points in the flow of talk. The second type similarly orient to prior talk and formulate a selective summary or description of it, but are distinct in terms of their consequences: they involve more marked shifts in the thematic direction of talk. Below I will consider some characteristics of these two broad types of utterances, focusing on the linguistic and interactional features typical of their use. In this task the key questions to be addressed are (1) what form do these utterances take and what interactional functions do they serve in specific contexts, and, (2) how are they managed sequentially, in other words, what kinds of actions do they make relevant in next turns and how these are managed? Further, I will discuss how the local management of these actions is related to the institutional arrangements of the broadcast settings.

2. Candidate Understandings The first group consists of interjections, which offer candidate understandings of a prior turn. These are often produced in overlap with current speaker's talk. Alternatively, they may occur in places where the prior turn is not complete grammatically or in terms of its content, for instance when a speaker hesitates or interrupts his or her own turn. Typically interjections which function as candidate understandings do not lead to any major shifts in topic or participation framework. The host or interviewer does not assume the role of primary speaker or actively direct subsequent talk, but rather, takes an active part in the local construction of meaning. Interjections of this type emerge from the immediately preceding turns and use resources made available in them to point to particular meanings or inferences that can be drawn from them. Candidate understandings are constructed in a variety of ways and serve particular purposes in the sequential environment in which they emerge. Sometimes they are used for clarifying or correcting the way that prior utterances should be understood. They may, for example, orient to some problematic aspects of prior turns and lead to brief 'time-outs' from the current focus of talk in which such problems are clarified or repaired. Similarly, they may make explicit something that was not said in so many words in prior talk or formulate an

Arja

96

Piirainen-Marsh

interpretation, which makes the content of prior talk more lively, dramatic or 'punchy' and hence more appealing to the viewing audience (cf. Heritage 1985). Sometimes they may express a particular stance on prior talk, e.g. challenging or criticizing its content. In brief, the uses of candidate understandings are shaped by institutional considerations in so far as they serve the interests of eliciting and producing talk which is suited to the overall aims of the broadcast setting. The first two examples show how a host in a multiparty discussion interjects into a guest's turn to make an affiliative comment. Both examples come from a 'multicultural' group discussion broadcast for a series of language education programmes on Finnish television. A number of observations can be made about the form, interactional function and sequential environment of these utterances. The interjections in examples 1 and 2 take the form of deictic comments, which attend to and affiliate with the prior speakers' turns. In example 1, AR is describing the problems she experienced when she first came to Finland and had to communicate with people without a common language. In the course of her description, she reports how she had to ask her Finnish husband to translate for her and his reaction ('and he goes cra:zy you know'). By referring to her husband, who is present in the studio, and shifting her gaze to him in the course her talk, AR co-implicates him in her telling (cf. Goodwin 1986, 1987). In the next turn (line 7), her husband PR begins to respond to AR's turn, apparently disagreeing with her. However, he hesitates, and before he can continue, the host (SS) in line 9 intervenes with an empathetic comment ('it's very tiring to translate all the time'). The form of her comment suggests that she is affiliating with PR's point of view by showing attention to the tiresome task of translating. Interestingly, however, its non-specific formulation (it's very tiring) seems to address both the previous speakers as 'knowing recipients' (Goodwin 1987) and does not specify one of them as primary addressee: Example 1 1 2 3 4 5 6 7 8 9 10 11

AR

SS AR PR

?

SS AR

=well nowadays most of them do speak English when I first came ((clears throat)) nobody spoke English it was very very difficult so of course I was on the outside. mm-hm= and uh 11 went to Pertsa and (say) you have to translate translate for me (.) and he goes so cra:zy you know no I- er hhh [>it's ve]ry tiring [to translate all the time precedes lines describing non-verbal aspects of interaction or translations. These lines are printed in italics.

Teachers Sharing Stories and Shaping

Policies

141

construction, and that exams should reflect what has been taught in class and that exam items need to be graded. Within these pre-existing policies, however, some room exists for the TAs to decide as a group how exactly to distribute points, what items to include in tests, and how to grade these items. Working towards policy-related decisions, the TAs frequently employ narratives to make a point and to express their own perspective, i.e., their subjective theories, of the issues under consideration. Narratives help them raise particular cases and involve their audiences more than abstract theoretical reasonings would. Giving voice to different characters, it is possible to create "the imagination of alternative, distant, others' worlds by linking them to the sounds and scenes of one's own familiar world" (Tannen 1989: 133). In one of the meetings, for instance, Katja 4 asks her colleagues how they would handle a student's excused absence during an oral practice session. When the group suggests not giving the student any points, she obviously disagrees with this suggestion. She tells a hypothetical tale to illustrate the unfairness to which she fears a student will be exposed. Katja

yeah, but, I mean, actually, that

207 Katja

is very unfair, I mean, if I'm walking on a street, and

Katja

a car, eh, hits me, and I end up in the hospital, and

Katja Andy

then get minus 20 points on it, because I'm in the hospi(that's true)

Katja

tal, it's (unfair)

208

209

210 2ii

Katja in her narrative challenges her colleagues' grading policy and illustrates her subjective theory in THE form of a narrative: Events which lie beyond the students' control should not affect their grade. Katja thus forces her peers to reconsider their judgment, and possibly to redraft their own subjective theories on grading. Positing a hypothetical case which involves an actual person - the teller herself as the protagonist - Katja wants to enable her colleagues to better relate to the situation. The discussion continues to the point where the TAs reach a compromise: a student missing class because of an excused absence will still receive attendance points, although points for participation will be deducted. 4

All names for all participants are pseudonyms.

Birgit

142

Meerholz-Haerle

Based on the present corpus of data, it appears that the TAs across all four groups are willing to make allowances for their students if valid excuses exist. In general, when the TAs exemplify their theories through narratives which support a certain degree of individual variation, i.e., which do not advocate absolute student perfection, these theories are more likely to be accepted and transformed into general policies than those which paint a student's options in black and white. Students are thus given the benefit of the doubt, and TAs seem willing to make allowances for certain student omissions. Narratives centered around grading and assessment policies also allow insights into the TAs' subjective theories on the concepts of communication and accuracy. A look at narratives dealing with the exam item Persönliche Fragen (Personal Questions) illustrates this claim. For this item, students are asked to answer questions about their own situation and are encouraged to express their personal meaning. Since student answers are likely to be very diverse, it is difficult for the TAs to establish concrete criteria for grading. In the following excerpt, Katja explains her grading policy. ¡Katja

no, but if, if it doesn't make any

¡Inge ¡Katja

even if the subsense, eh, I don't give any points, like ( )

¡Inge ¡Katja ¡Andy | >_

ject verb agreement is right?

¡Katja |>

say, ich esse gerne Pizza, and die Fra-, eh, die Frage (.) [I like eating pizza] [the quethe question

¡Katja ]>

war, wann stehen Sie auf? then, then I don't give anywas as when do you get up?]

42

43

44

yeah, but I mean, if they laughs

45

46 ¡Inge ¡Katja ¡Andy ¡Todd 47

so, what do you do thing that's true, that does happen well yeah, that's true

143

Teachers Sharing Stories and Shaping Policies ¡Inge ¡Katja ¡Andy ¡Todd

in that case, take off three points? yeah (.) yeah yeah be-

¡Katja ¡Todd >

they can answer the of what they cause they had no concept of laughter

¡Inge ¡Katja

mhm same (.) sentence to ail the questions

49

50

The subjective theory underlying Katja's policy might be paraphrased as "student answers have to 'make sense' with respect to the questions." While two other members of the group, Todd and Andy, seemingly agree with Katja (48-49), Dorothea raises her objections. ¡Doro

well, I wouldn't

50 51

]Doro

give points if this was the case for all questions, but,

¡Doro

maybe not your example, but if some eh, verb is similar

¡Doro

and they could have misunderstood, I would still give at

¡Doro

least something for the correct grammar, even though it

¡Doro

doesn't answer, if it looks as if they had misunderstood

¡Katja j> ¡Doro

(mhm) looking at Inge, then at Mike when he starts talking this is kind of eh, obvious where they didn't under-

¡Mike ¡Doro

'cause also on communication ( stand, ( )

¡Mike

nication is better than none, so even if they misunder-

52

53

54

55

56 ), some commu-

57 58

Birgit

144 ¡Inge ¡Mike

( ) stood you, (.) and they answered your question, and they

¡Mike ¡Doro |>

answered it wrong, at least (you as a)

¡Mike ¡Doro \>

yeah, zum Frühstück, you could then go, or something for breakfast] laughs

Meerholz-Haerle

59

ah, Sie essen Pizza [you eat pizza

60

61 ¡Inge ¡Mike ¡Doro \> _

¡Inge |> _ ¡Mike

or the crust, for that matter (true), yeah

so then you I(

)

( have a chance to,

yeah laughs

)

(I eat) some others mumbling you know, repair

63 ¡Inge ¡Mike ¡Andy

light crust for breakfast the situation pizza bones

64

Dorothea first mitigates her objection, agreeing with Katja that students do not deserve any points if they answer the same sentence to each question (50-51). However, she then distances herself from Katja, suggesting that there are situations when she would give the students some points (52-56). Mike aligns himself with Dorothea, also making allowances for certain student answers. He explicitly voices the theory that "some communication is better than none" (5758). Dorothea takes up this theory and alters the hypothetical narrative introduced by Katja (60-61) accordingly. In her version, the student's seemingly incoherent reply does not result in a breakdown of the conversation, but merely leads it into a different direction. The discussion sparked by this tale eventually ends with the following decision: ¡Mike

yeah, so two

205 ¡Inge ¡Mike

yeah for Sinn and one for grammar and then that way, this

Teachers Sharing Stories and Shaping

!>

145

Policies

[meaning]

206 ¡Mike

is the risk takers' section, where they can be a little

¡Mike [Doro

bolder with their (.) language and, cause everywhere experiment

¡Mike

else it's either right or wrong pretty much (

207

208 )

209

Thus, Katja's hypothetical tale opened a discussion which lead the TAs to reconsider their subjective theories, to negotiate their grading policies and to make allowances for student answers which might not completely adhere to the principles of accuracy and coherence. The TAs' co-narrations provide insights into their theories about the concepts of language and communication. They apparently value the students' ability to communicate. Although they certainly see the need for focusing on issues of accuracy and coherence, they also take into consideration the communicative nature of a task when evaluating their students' performance. The theory underlying this approach suggests that students should not be tested on issues of right or wrong only, but should also be encouraged to take risks and to express their own meaning. This theory is also advocated by the larger culture of this particular German Studies department, and is apparently accepted and supported by the majority of the TAs.

4. Conclusion In conclusion, when TAs tell narratives such as the ones presented above, they ask their colleagues to realign themselves, rethink and redraft their subjective theories, and to support the tellers' cause. It is interesting to note that the TAs are mainly successful with their motions and manage to work out compromises. It appears that frequently, theories and subsequent policies are worked out which allow for certain student flaws or omissions. The TAs seem willing to grant exceptions if legitimate excuses are provided, and seem reluctant to accept theories and policies which do not allow for individual variation. They also try to protect their students from potentially unfair or problematic grading procedures, and try to encourage them to experiment with the language. As this discussion has shown, a focus on narratives told in professional teacher meetings can help shed light on the subjective theories which teachers

146

Birgit Meerholz-Haerle

hold about their teaching practice. Investigating co-narration among participants can provide insights into which theories are rejected and which are ratified and transformed into more general policies. Studies of this kind might also help investigate the larger culture(s) of teaching which are developed and cultivated in individual language programs.

References Cameron, Deborah (1994): Putting our Practice into Theory. In: D. Graddol and J. Swann (eds.), Evaluating Language (15-23). -Philadelphia: Multilingual Matters. Carter, Kathy (1993): The Place of Story in the Study of Teaching and Teacher Education. Educational Researcher 22. 5-12. Clandinin, D. Jean and Connelly, F. Michael (1992): Teacher as Curriculum Maker. In: Philip Jackson (ed.), Handbook of Research on Curriculum (363-401). - New York: Macmillan. Ehlich, Konrad (1993): HIAT: A Transcription System for Discourse Data. In: Jane Anne Edwards and Martin D. Lampert (eds.), Talking Data. Transcription and Coding in Discourse Research (123-148). - Hillsdale, NJ: Erlbaum. Grotjahn, Rüdiger (1991): The Research Programme Subjective Theories: A New Approach in Second Language Research. - Studies in Second Language Acquisition 13. 187-214. Kallenbach, Christiane (1995): Das Konzept der subjektiven Theorien aus fremdsprachendidaktischer Sicht. In: Lothar Bredeila and Herbert Christ (eds.) Didaktik des Fremdverstehens (81-96). Tübingen: Gunter Narr. Latour, Bruno and Woolgar, Steve (1986): Laboratory Life: The Construction of Scientific Facts. (2 ed.). - Princeton, NJ: Princeton University Press. Lynch, Michael (1985): Art and Artifact in Laboratory Science: A Study of Shop Work and Shop Talk in a Research Laboratory. - London: Routledge and Kegan Paul. Meerholz-Haerle, Birgit (1998): Teachers Talking Shop: A Discourse Study of TA Coordination Meetings. Unpublished Ph. D. Dissertation. - Tucson: University of Arizona. Ochs, Elinor, Smith, Ruth and Taylor, Carolyn (1989): Detective Stories at Dinnertime: ProblemSolving through Co-Narration. - Cultural Dynamics 2. 238-257. Ochs, Elinor and Taylor, Carolyn (1992a): Mother's Role in the Everyday Reconstruction of "Father Knows Best". In: Kira Hall, Mary Bucholtz and Birch Moonwoman (eds.), Locating Power: Proceedings of the Second Berkeley Women and Language Conference (Vol. 2, 447462). -Berkeley: Women and Language Group. Ochs, Elinor and Taylor, Carolyn (1992b): Science at Dinner. In: Claire Kramsch & Sally McConnell-Ginnett (eds.), Text and Context. Cross-Disciplinary Perspectives on Language Study (29-45). - Lexington, MA: Heath. Ochs, Elinor, et al. (1992): Storytelling as Theory-Building Activity. - Discourse Processes 15. 37-72. Sacks, Harvey, Schegloff, Emanuel A. and Jefferson, Gail (1974): A Simplest Systematics for the Organization of Turn-Taking for Conversation. -Language 50. 696-735. Tannen, Deborah (1989): Talking voices: Repetition, Dialogue, and Imagery in Conversational Discourse. - Cambridge: Cambridge University Press. Taylor, Carolyn (1995): Child as Apprentice-Narrator: Socializing Voice, Face, Identity, and SelfEsteem amid the Narrative Politics of Family Dinner. Unpublished Ph.D. Dissertation. - Los Angeles: University of Southern California. White, Jane J. (1991): War stories: Invitations to Reflect on Practice. In: B. Robert Tabachnick and K.enneth M. Zeichner (eds.) Issues and Practices in Inquiry Oriented Teacher Education (226252). - London: Falmer Press.

Polly Walsh Investigating Interactional Variation in Interview-Type Speaking Tests 1. Introduction During the last ten years or so, a considerable amount has been said and written about what goes on in language proficiency interviews (LPIs). Although they have face validity and are easy to administer, they have come under fire, on various counts. On the one hand they have been attacked for failing to satisfy the criterion of reliability: if the same elicitation procedure is not followed with each candidate, the samples of language produced will not be comparable (Lazaraton 1996; Bachman 1990). In an LPI there is inevitably a degree of variation in the interviewer input because interviewers have to take account of the candidate's contributions. In fact they have been found to adjust their behaviour, possibly unconsciously, in response to the perceived level of different candidates (Lazaraton 1996; Ross 1992; Ross and Berwick 1992). If the learner has difficulty understanding, for example, the tester is likely to react by speaking more slowly and articulating more clearly, repeating and/or rephrasing questions, etc. According to Lazaraton (1996: 154), these and other types of interlocutor support add "an element of uncontrolled variability to the assessment picture". On the other hand, the construct validity of LPIs as tests of speaking skills has also been questioned. Briefly, the argument runs as follows: the nature of the interaction prevents candidates from demonstrating skills which are an integral part of the ability supposedly being measured. The model of communicative competence on which speaking test assessment criteria are explicitly or implicitly based, is derived from conversation analysis, but interviews are not conversation. The purpose of the LPI - obtaining a rateable sample of language - determines the behaviour of the participants (van Lier 1989). Whereas topic, turns and meaning are negotiated in natural conversation, in LPIs it is the interviewer who dominates the interaction, moving it through its various stages, in control of topic and able "to take or withhold the floor at any time" (Young and Milanovic 1991: 417). Questions are designed to allow (or push) the candidate to speak at length (Ross 1998). As a result of this, and of the power distance between the participants, candidates have little or no chance to show their ability to "control conversation, to produce topic initiations, or to assume responsibility for the continuance of the

148

Polly

Walsh

discourse" (Perrett 1990: 236). This type of test may therefore "obscure discourse differences between learners" (Young 1995: 37). The problem, from the point of view of fair testing, would appear to be not so much that there is interactional variation in LPIs, but that we do not know enough about the type or range of variation possible in this form of dialogue to take account of it in the assessment process. Our definitions of proficient performance are theory-based and not data-based, as Fulcher (1996) points out. However asymmetrical, an interview is a form of collaborative discourse, and should therefore give us some indication about candidates' relative ability to "take part in the negotiative and interactive process of speaking" (Byrnes 1987: 301) in realworld contexts; more, certainly, than a test consisting of recorded prompts. But, unless the nature of the interactional variation possible in this type of speech event is understood by investigation of real interview data, it will inevitably remain an "element of uncontrolled variability" because interviewer/raters are not trained to be aware of it, or to take account of it consciously and consistently in their evaluation of the candidate's performance.

2. An Approach to Investigating LPI Interaction An approach to investigating interactional variation in LPIs is suggested in Walsh (1999). Since the interviewer in this type of test is generally a native or near-native speaker, the LPI can be seen as a form of Native/Non-Native Speaker, (NS-NNS), discourse. Non-test NS-NNS interaction is reported to be asymmetrical, with the NS speaking more and taking control of topic. This has been accounted for in various ways; either learners are aware of their linguistic limitations and "accept a role which is less desirable than [they] could ordinarily achieve" (Harder 1980: 268), or superior language skills give higher interactional status (Beebe and Giles 1984). Although it has not been systematically studied as a variable which might affect the interactional balance, there is some evidence in the literature that the NNS's level of proficiency also affects the degree of NS dominance. Gaies (1982) reports that although NSs appear to be more disposed to treat NNS interlocutors as conversational equals when they have shared subject knowledge, they nevertheless take control if the NNS has serious linguistic difficulties. Zuengler and Bent (1991: 409) suggest that fluent NNS topic experts are more able to "overturn ethnolinguistic dominance" than less proficient ones. In an early study of foreigner talk, Long (1981, 1982) explained NS dominance in terms of interactional adjustments made to avoid and repair communication breakdown. With an NNS interlocutor, NSs were found to speak more slowly and clearly, stressing key words, repeating, rephrasing, checking comprehension. They made frequent topic changes, generally in the form of

Investigating Interactional Variation in Interview-Type Speaking Tests

149

questions, and used fronting to make topics salient. They also made allowances for productive deficiencies, accepting ambiguity and unintentional topic switches, reformulating or expanding incomplete or unclear contributions from the NNS. Long's findings would seem to imply that native speakers make interactional adjustments to help the NNS contribute. So, for example, they ask a lot of questions because less obvious turn-allocation devices, such as pauses after statements at "transition relevance places" (Sacks et al. 1974: 703) are more difficult for a less fluent speaker to take advantage of. Similarly, they introduce topics which they think will be easy for the NNS to handle, relating mostly to the latter's interests and experience. This analysis of NS-NNS discourse would lead one to expect that the extent of NS dominance will vary with the NNS's linguistic skills. Presumably, the less need there is to provide conversational support, the less the NS will do so and the more balanced the interaction will be.

3. Language Interviews as NS-NNS Discourse A language interview is certainly a contrived and unequal encounter, but it is a form of collaborative NS-NNS discourse. It would seem likely therefore that there will be variations in the interactional behaviour of both participants which are related to the candidate's level of linguistic proficiency. In a small scale study carried out in an Italian university (Walsh 1999), I investigated interviewer and candidate interactional behaviour in a speaking test, comparing the pattern of interaction in interviews with candidates awarded different grades. The following indicators were examined: 3.1.

Interviewer Behaviour: Control and Accommodation

Two measures of interviewer dominance were considered in the study: on the one hand, questions and other direct prompts; on the other, interruptions. a) questions and direct prompts In that they require a particular type of response and cannot be ignored, questions and other kinds of direct prompts were classified as controlling moves. They certainly serve this purpose in LPIs. However, interviewers, like NSs in non-test discourse, use questions at least in part to carry the interaction forward when candidate responses are insufficient, as He (1998) reports. With more fluent candidates, it was hypothesized that fewer prompts would be needed to keep the conversation going.

150

Polly

Walsh

b) interruptions Interruptions have been taken in other studies as disconfirming moves, denoting lack of respect for the speaker (Zuengler 1989; Kennedy/Camden 1983) and were therefore investigated here as indicators of dominance. c) non-directional comments It was hypothesized that with more proficient candidates the interviewer would feel free to use less obvious turn allocation devices, and in particular nondirectional comments. When NSs do this, according to Whyte (1995: 164), it is because they feel that the NNS is "shouldering greater responsibility for maintaining the flow of conversation". d) accommodation It was hypothesized that interviewers would use various forms of accommodation, adjustments of the type Long found in non-test NS-NNS discourse to avoid and repair conversational trouble (cf. Lazaraton 1996). It was expected that the extent of accommodation would vary inversely with the rating given, as reported in Ross (1992) and Ross and Berwick (1992). The following forms of accommodation were considered in the study: grammatical simplification, lexical simplification, over-articulation, slowdown, question reformulation, fronting, clarification requests, other expansions and completions. 3.2.

Candidate Behaviour

a) speaking: topic developments and long turns In spite of overall interviewer control of topic, it was hypothesized that more proficient students would be less passive than weaker ones. It has in fact been found in other LPIs that while weak candidates limit themselves to providing answers to the questions asked, more advanced learners are able and willing to elaborate and expand (Young 1995; He 1994). This has been explained partly in terms of perceptions of the interview context; advanced students see beyond the question to a chance to display their language skills. However it is also clearly linked to the candidate's productive skills. While topic nominations were unlikely in the LPI context, it was hypothesized that more advanced candidates would make topic developments; either elaborating after answering a question or taking the floor after interviewer non-directional comments. It was also hypothesized that more proficient students would take longer turns (cf. Young 1995; He 1994).

Investigating Interactional Variation in Interview-Type Speaking Tests

151

b) listening Looking at the candidate's listening behaviour, it was expected that here too fluent candidates would be more active, and that at higher levels of proficiency we would find backchannelling, overlapping speech, turn claims and even interruptions. These behaviours can be taken to indicate that the listener is following what is being said, and is ready and willing to take the floor (cf. Fulcher

1996).

4. The Study The study involved students who were taking a diploma in business and tourism. The test took the form of a relatively unstructured interview consisting of questions relating to the student's work experience, questions on the topics discussed in class, questions on a topic chosen and prepared by the student. In all three parts the theme was tourism. The candidates' performance was assessed by the interviewer and a second rater. Nine interviews were selected for analysis; four awarded the highest grade (group H), two the lowest grade (group L), three an intermediate grade (group M).

5.

Results

Considerable variation was found between the two extremes, both in interviewer and in candidate behaviour. Interviews with group L candidates were characterized by a large number of interviewer prompts, accompanied by extensive accommodation. There were no non-directional comments. Group L candidates took no long turns and never made topic developments, limiting themselves to answering the questions, often with great difficulty. The interviewer was supportive, providing completions and reformulations, accepting ambiguity and only on one occasion interrupting, when the candidate had completely misunderstood the question. At the other end of the scale, there were considerably fewer prompts, and virtually no accommodation; although their English was far from perfect, the group H candidates had no trouble in understanding and in formulating answers. They were able to hold the floor for extended periods, making in-turn topic developments, and they also appeared to have little trouble taking the floor when the interviewer made non-directional comments. They were active listeners, giving frequent signs that they were following and ready to contribute, even to the point of occasionally interrupting the interviewer. The latter, meanwhile, made fewer concessions with these learners than with either of the other groups, challenging what they said and sometimes even interrupting. In fact the distinguishing feature

152

Polly Walsh

of these interviews is that there are some exchanges in which the participants seem to be competing for the floor on a more or less equal basis. Group M was less homogeneous, but two of the three candidates were similar: they showed that they could take some longer turns and make occasional in-turn topic developments, but only on the topic they had prepared, while they were clearly at a loss on the rare occasions when the interviewer made nondirectional comments. Their receptive and productive difficulties made accommodation necessary, though less so than in group L interviews, and the number of prompts was still quite high. For the most part these candidates were interactionally passive, while the interviewer was supportive, almost never interrupting or challenging.

6. A Case of Evaluative Inconsistency In short, in this very small sample of interview discourse, a fairly clear pattern of proficiency-related interactional variation does emerge. However, even in this tiny sample, where all the candidates were interviewed and assessed by the same raters, one of them does not fit the pattern. Unlike the other two members of group M, this young man has good comprehension, is fluent and interactionally assertive, contriving to keep the floor for long periods. He makes frequent topic developments, both in-turn and after non-directional comments and corrections. There are relatively few prompts and no accommodation. In all, from an interactional point of view, his performance deserved a higher assessment and, as seen above, interactional skills are closely related to linguistic ones (cf. He 1998). Admittedly, it is true that the English he produces is wildly inaccurate, and this must have swayed the raters, but communication takes place, and without undue effort on either side. This certainly cannot be said for the other two group M interviews! It is evident that, so long as interviewer/raters do not have clear guidelines as to how to take account of interactional skills, there is a danger of evaluative inconsistency. That the risk of inter-rater inconsistency is even greater is confirmed by the preliminary results of a study of IELTS speaking tests.

7. The Second Study: Preliminary Results The sample consists of 18 IELTS speaking tests with Indonesian candidates, six assessed as "very good users" (Band 8), six as "competent users" (Band 6), six as "limited users" (Band 4). Although no systematic study of the transcripts has yet been carried out, it would appear that the overall pattern of interactional variation resembles that found in the Italian study, making allowances for the fact that the level of the learners in the two studies is not quite equivalent (group L would

Investigating Interactional

Variation in Interview-Type Speaking Tests

153

correspond approximately to a Band 3, group M to a Band 5 and group H to a Band 7). In Band 4 interviews the interaction is extremely asymmetrical, with frequent interviewer prompts and extensive accommodation. Band 6 candidates have fewer receptive and productive difficulties, speak at greater length and make some topic developments, but interviewers still need to be supportive. In particular, they do not challenge the content of what the candidate is saying, accepting non sequiturs (cf. Ross 1998) and ambiguity, as in the following examples: I: S: I:

have you had a chance to visit any other parts of Australia or have you just stayed in Sydney the whole time? er I don't know it depend if I pass the test maybe I'll win a place at the university of Oolangong yeah

In the second example, the interviewer backchannels encouragingly, although what the candidate is saying does not make much sense: I: S:

so do you think that there's an imbalance between imports and exports in Indonesia at the moment? er yeah now Indonesia has so many exports the imports better than before we got so many import chains from other country that we not do a lot of export

[ ]

I: S:

mm now so I think it's very good

In interviews with Band 8 candidates the interaction is more balanced, with the candidate speaking at greater length and participating more in shaping the discourse. There are fewer questions and more non-directional comments.

8. Uncontrolled Variability in the Assessment Process? While this is the general picture, there are cases in which discourse features of the interaction appear to have been disregarded in the assessment. The first case is that of a male candidate assessed as Band 8. According to IELTS, this should mean that he "has fully operational command of the language with only occasional unsystematic inaccuracies and inappropriacies. Misunderstandings may occur in unfamiliar situations. Handles complex detailed argumentation well." One would not therefore expect to find exchanges of this kind (from the elicitation section): S: I: S:

I have heard from someone that you want to go to visit your friend that's right? yes er what e:::r country er is he from?

[

154

Polly Walsh I: S: I: S: I: S: I: S: I: S: I:

England England ah so have you er been there? well I'm English ah English yeah not Australian yeah have you been to England? yes right have you been to England before? yeah I'm from England I am I'm English oh I see oh I see I was born there

The candidate not only does not understand, but does not signal that he has a problem or what exactly the problem is. Instead, like the learner described in He (1998), he uses "yeah" to cover his failure. He misunderstands again when the interviewer asks him if he has been to England, taking the question mistakenly to be a prompt. At another point, the candidate fails to make himself understood and the interviewer eventually accepts that the confusion cannot be cleared up: I: S:

it says here you've also been a lecturer is that here in Indonesia? no I mean that er in case I have already taken a master's degree so I can have a = [ ]

I: you want to do that? S:= broad I mean that a broader er perspective so I can tell my experience my knowledge to the students I: mmhm you say students er you want to go to er a university and lecture?

[

S: I:

yes yes I'm planning to go to the university but also work at Indosat when you come back so are they financing your study in Australia or are you doing it yourself?

When the candidate misunderstands the question "you want to go to a university and lecture", assuming that he is being asked about his study plans in Australia, the interviewer accepts the unintentional topic switch. Very probably, the interviewer disregards these episodes, which point to considerable receptive and productive failings, because he has formed a favourable impression at the beginning of the interview, when the candidate is talking about his job. Certainly they communicate more successfully here, though this may be partly because the interviewer is doing a lot of the talking and laughing himself! In another case, the candidate, assessed as a Band 4, has considerable linguistic difficulties, but shows some discourse competence, as the following extracts illustrate:

Investigating Interactional Variation in Interview-Type Speaking Tests I: S: I: S: S: I: S: I: S:

155

right so you don't have any spare time spare time? example? I'm sorry erm do you have any time for music or for reading or church?

[

]

oh yes I know I always swimming e::r a::nd fitness wow ((laughs)) I like exercise mm I see I see yes

When she does not understand the question, she asks for clarification (she has already told the interviewer that she is attending various courses in her spare time). She backchannels to indicate that she has got the point and is ready to answer. She then follows up the interviewer's non-directional comment with an elaboration, and closes the exchange with her final yes. Whenever this candidate fails to understand, she indicates where the problem lies, facilitating repair, and on various other occasions makes topic developments after non-directional comments, as in the following exchange: I: S: I: S: I: S: I: S: I: S: I: S:

and what time do you come home? e::r sometimes I go home 7 o'clock er 8 o'clock it's a very long day yes after afte::r e::r after I go university sometimes I go shopping or I have coffee yes with my friends uhuh where do you like to go shopping? in Djakarta? mm mm my university opposite shopping centre oh how wonderful and er cinema ((laughs)) oh I see ((laughs)) so everything is very close yes

The IELTS rating descriptors make no reference to interactional skills; it is true that this candidate "has frequent problems in understanding and expression" and "is not able to use complex language", as the Band 4 descriptor specifies, but that is only part of the picture. One cannot help feeling that the interviewer, given the candidate's discourse competence, might have allowed that she had "partial command of the language, coping with overall meaning in most situations, though likely to make many mistakes" and that she "should be able to handle basic communication in own field" (Band 5). However, without specific guidelines relating to discourse features of the interaction, interviewers can only make their own decisions, and this does indeed introduce "an element of uncontrolled variability to the assessment picture"

156

Polly Walsh

(Lazaraton 1996: 154), as these examples show. What are needed are data-based rating criteria which take account of the type and range of proficiency-related interactional variation possible within the confines of a particular test format, and, of course, trained interviewer/raters following clear guidelines.

References Bachman L. (1990): Fundamental Considerations in Language Testing. - Oxford: Oxford University Press. Beebe L. and Giles, H. (1984): Speech Accommodation Theories: A Discussion in Terms of SecondLanguage Acquisition. - International Journal of the Sociology of Language 46. 5-32. Byrnes, H. (1987): Speech as Process. - Foreign Language Annals 20. 301-310. Fulcher, G. (1996): Does Thick Description Lead to Smart Tests? A Data-based Approach to Rating Scale Construction. - Language Testing 13. 208-238. Gaies, S. J. (1982): "Modifications of discourse between native and non-native speaker peers". Paper presented at the 16th Annual TESOL Conference, Honolulu, Hawai. Harder, P. (1980): Discourse as Self-Expression - on the Reduced Personality of the SecondLanguage Learner - Applied Linguistics 1. 262-270. He, A. W. (1994): "Elicited vs. Volunteered Elaboration: Talk and Task in Language Proficiency Interviews." Paper presented at the annual conference of the American Association for Applied Linguistics, Baltimore, Maryland. - (1998): Answering Questions in LPIs: A Case Study - In: Young, R. and He, A. W. (eds.) Talking and Testing: Discourse Approaches to the Assessment of Oral Proficiency. Amsterdam and Philadelphia: Benjamins. 101-116. Kennedy, C. W. and Camden, C. T. (1983): A New Look at Interruptions. - Western Journal of Speech Communication 47. 45-58. Lazaraton, A. (1996): Interlocutor Support in Oral Proficiency Interviews: the Case of CASE. Language Testing 13. 151-172. Long, M. H. (1981): Questions in Foreigner Talk Discourse. - Language Learning 31. 135-157. - (1983): Native Speaker/Non-Native Speaker Conversation and the Negotiation of Comprehensible Input. - Applied Linguistics 4. 126-141. Perrett, G. (1990): The Language Testing Interview: A Reappraisal. In: Jong, J. and Stevenson, D. K. (eds.) Individualizing the Assessment of Language Abilities. Clevedon: Multilingual Matters. 225-238. Ross, S. (1992): Accommodative Questions in Oral Proficiency Interviews. - Language Testing 9. 173-186. - (1998): Divergent Frame Interpretation in Oral Proficiency Interview Interaction. In: Young, R. and He, A. W. (eds.) Talking and Testing: Discourse Approaches to the Assessment of Oral Proficiency. 333-353. Ross, S. and Berwick, R. (1992): The Discourse of Accommodation in Oral Proficiency Interviews. Studies in Second Language Acquisition 14. 159-176. Sacks, H., Schegloff, E. A. and Jefferson, G. (1974): A Simplest Systematics for the Organization of Turn-Taking in Conversation. - Language 50. 696-735. van Lier, L. (1989): Reeling, Writhing, Drawling, Stretching and Fainting in Coils: Oral Proficiency Interviews as Conversation. - TESOL Quarterly 23. 489-508. Walsh, P. (1999): Can a Language Interview be Used to Measure Interactional Skills? - C.A.L.S. Working Papers. (1-26) Reading: Centre for Applied Language Studies. Whyte, S. (1995): Specialist Knowledge and Interlanguage Development. A Discourse Domain Approach to Test Construction. - Studies in Second Language Acquisition 17. 153-183.

Investigating Interactional

Variation in Interview-Type Speaking Tests

157

Young, R. (1995): Conversational Style in Language Proficiency Interviews. - Language Learning 45. 3-42. Young, R. and Milanovic, M. (1991): Discourse variation in Oral Proficiency Interviews. - Studies in Second Language Acquisition 14. 403-424. Zuengler, J. and Bent, B. (1991): Relative Knowledge of Content Domain: An Influence on NativeNon-Native Conversations. - Applied Linguistics 12. 397—415.

Acknowledgement I would like to thank the University of Cambridge Local Examinations Syndicate, and, in particular, Nick Charge, the IELTS Subject Officer, for kindly giving me access to recorded IELTS speaking tests.

Julia Bamford

Question and Answer Sequencing in Academic Lectures 1. Introduction Academic lectures have always been regarded in the literature on oral/written differences as being something of a hybrid, an oral genre which displays most of the characteristics of those texts at the most written end of a written/oral continuum (Biber 1988). Yet academic lectures are rarely studied as an oral genre, with the notable exception of the studies in Flowerdew (1992). The perspective presented here is slightly different from those studies in that it focuses on the interactive conversational nature of academic lectures and uses the analytical tools typical of conversation analysis.1 In other words this research examines those aspects of lectures which differentiate them from written texts and analyses some textual features whose occurrence in written texts is much less common.2 The specific aspect I will be examining here is question and answer sequences in a small corpus of lectures. The lectures examined are all fairly formal in that there is little or no direct exchange of talk with students, but none of them was read from a written script. The corpus consists of eleven lectures in several sub disciplines of economics - development economics, rational expectations, economic theory, employment - and law - business law, tort law and privacy and emotional distress. Some of the lectures were for undergraduates others for post graduates. The lectures in economics were given by visiting lectures and were recorded at the University of Siena, the law lectures are part of a commercially available series of award winning lectures recorded at the University of Texas.3 All of the lecturers are native speakers. Transcription loosely follows conversation analysis conventions omitting, however, unconventional spelling to denote nonstandard pronunciation (Jefferson 1979). The salient features for this analysis are

1 2

3

Many of the papers in Flowerdew's collection examined academic lectures using the tools of text analysis developed for the written text. This is not to say that questions do not occur in written texts. In the bestselling economics textbook for over forty years - Samuelson - the first chapter contains no less than 23 questions, however none of these are answered in the immediately following text but are dealt with in different chapters of the book. Moreover the questions seem to serve as exemplification presenting concrete examples for which the text later provides a theoretical explanation. In the remaining chapters questions are much less frequent. The lectures are from "Great courses on tape. The Law of Torts" The Teaching Company, Springfield Va.

160

Julia Bamford

the use of a question mark to indicate rising intonation while a full stop denotes falling intonation. The role of questioning in teaching and learning has been recognised since Socrates assigned such fundamental importance to it in his Dialogues. The Socratic method depended on engaging the listener in a dialogue based on the question/answer sequence. Questions and answers have formed part of education systems ever since and have the potential to both seek information and at the same time exercise control and emphasise power asymmetries between teacher and pupil. (Goody 1978) When both questions and answers are written, as in textbooks, or spoken by the same actor as in lectures, both the risk of threatening the face of students and of exercising overt control are very much attenuated. However questions are potentially problematic social actions even in spontaneous conversation and are therefore often prefaced to prepare the ground for the impending trouble as in example 1 below: (1)

(Bamford: Roch. 94) G. Muum? .. uh . I've got to ask you something.

2. Questions and Answers What is a question? In spoken texts it is often difficult to pinpoint what a question is actually doing in the interaction as Levinson (1983: 275) says: ..there are no isolable necessary and sufficient conditions on, for example, questionhood, but rather that the nature of the use to which interrogatives are put can vary subtly with the nature of the language games or contexts in which they are used.

Besides being a call for information, interrogatives can be used with the illocutionary force of exam questions, rhetorical questions, requests, offers, suggestions, threats, invitations, requests for clarification, confirmation, permission etc. Indeed questions in actual usage are just too variable and context dependent to be captured by a closed definition. Answers are also complex and escape precise definition depending, as they do, on sequential location and topical coherence so that what makes an utterance which follows a question constitute an answer i.e. its answerhood, is not only the nature of the utterance itself but also the fact that it occurs after a question within a particular context. In the following example (2) an answer to the question would have been less collaborative than the response actually given. Yet this particular response in fact constitutes an extremely economical and collaborative answer to the question:

Question and Answer Sequencing in Academic (2)

Lectures

161

A. Is John there? B. You can reach him at extension thirty four sixty two (Levinson 1983: 293)

Answers often take the form of action; typically this consists of compliance to a question/request such as "Could you close the window?" In the small corpus analysed here the problem of definition of what constitutes a question does not really arise in these terms since all the questions which occur are relatively straightforward - what Schegloff (1984: 35) has called "the clear case" - both in form - either "wh" questions or with subject operator inversion resulting in a yes/no question and also in their illocutionary force.4 Answers, however, can be more difficult to isolate than in spontaneous conversation because their argumentative nature leads them to be complex, dispersive and speculative.

3. Adjacency Pairs Question and answer pairs are a typical form of what conversation analysts call adjacency pairs. These constitute the basic building blocks of interaction and communication and have a sequential organisation by which the utterance of a first pair part normatively requires the production of an appropriate second pair part in close contiguity (Sacks 1987). When an answer is not forthcoming it is heard as noticeably absent and must in some way be accounted for. Question and answer sequences have been analysed extensively in both spontaneous conversation (Schegloff 1984) and in institutional interaction (Drew and Heritage 1992) but both of these presuppose the active participation of at least two interactants. Question/answer pairs in lectures are, on the other hand, characterised by one interactant holding the floor exclusively and performing both the asking and answering roles. The lecturer already possesses the information the question posits, but nevertheless asks the question and provides his student listeners with the answer. In the case of the lectures in this corpus, in which the speaker utters both question and answer, can this still be considered an adjacency pair? Extract 3 below illustrates an adjacency pair, a question/answer sequence from the corpus, taken from a lecture in law dealing with business torts. If we examine extract 3 we 4

However although some of the questions in the corpus have subject operator inversion they never have a mere "yes" or "no" in the subsequent answer, but are more articulated. In spontaneous conversation this is often the case also since straight "yes" "no" answers to questions are relatively rare compared to longer answers even those prefaced by a "yes" or "no". As a rough check on this a look at the BNC sampler of oral texts showed than only 6% of "yes" answers to questions had a stand alone "yes" answer and the equivalent for "no" answers was predictably even lower. (I am endebted to Guy Aston for this information). In other words actual stand alone yes/no answers are much less common than we might suppose.

Julia Bamford

162

find a question/answer sequence which is typical of academic lectures where, instead of the question being asked by a co-participant, it is uttered by the speaker himself. The lecturer then proceeds to answer the question, which is inserted into the appropriate second pair part, thus satisfying the expectations raised whenever a question is asked: (3)

(Frank Cross 7/ Business Torts) FC: Are you liable? F C : . . Well heck no

Sequences of this kind are a frequently recurring pattern in the lectures in this corpus. The question/answer sequences display the general orientation to conversational norms by the speaker as is evidenced in the next slot placing of the answer and his recognition of accountability as witnessed by the provision of an appropriate answer. In the following sequences we find the speaker orienting his second pair part and pushing the answer into the next slot: (4)

(Pari K. Development Economics 14.5.98) PK. What is science? PK . . . Science is facts . and figures

Extract 4 demonstrates a typical adjacency pair sequence with the repetition of "science is" in the answer paralleling the "is science" of the question. Adjacency pairs have the characteristic of being highly cohesive; in this case cohesion is reinforced by the repetition of lexical items (Hasan 1985 and Halliday and Hasan 1976) Repetition of lexical items in the answer which were originally uttered in the question are typical of the lectures in the corpus; extract 5 below has no fewer than three exact repetitions ("an at will contract") and a partial repetition ("at will"). Repetition serves a cohesive function, but also reinforces the signal to the distracted student listener that this is the answer to the question. Moreover repetition of the same lexical items in both question and answer replicate the same phenomenon in adjacency pairs in conversation reinforcing the impression of interactivity (Schegloff 1984). This sequence displays a double question, the first question serving as an introduction to the topic - 'at will' contracts. It is not answered but reinforced by a second question which focuses more explicitly ("What is an 'at will' contract?"). 5 This second question is then answered in the appropriate slot to preserve adjacency pair contiguity:

5

This is mirrored in conversation where as Sacks (1987: 60) points out "it is a rather general rule that where two questions are produced, and you are going to have two answers, then the order of the answers is the reverse of the order of the questions."

Question and Answer Sequencing in Academic (5)

Lectures

163

(Frank Cross Business Torts) FC. What about an at will contract? . .What is an at will contract? FC: An at will contract is one . that can be terminated by either side at will....

Extract 6 shows a further example of a double question, however in this case the second question ("forever?") acts as a dummy reply to the first question but also functions as an elaboration on it.6 The answer is slotted in the appropriate second pair part space and the question answered is the second question as is emphasised by the repetition of the lexical item "forever". This is a "yes/no" question with the "no" answer slotted in initial position and justification for this "no" provided after a slight pause (See footnote 2): (6)

(Frank Cross Business Torts) FC: So how long does a contract last? . . forever? FC:. No . You have no promise to be employed forever.

4. Insertion Sequences In an adjacency pair sequence, fashioning an appropriate second pair part may require intervening activity, for example gathering information to respond to a request, ordering a complex answer chronologically, ordering the answer logically etc. (Jefferson 1972). In addition, and this is particularly true of academic lectures, the answer itself may be so complex that before it is actually completed much intervening activity is required. In order to maintain accountability the speaker provides an insertion sequence which prepares the other interactants for the projected second pair part. The insertion sequence thus shows the speaker's awareness of the fact that the projected second pair part is still pending: (7)

( Frank Cross Business Law) FC: Some people have a privilege to interfere with contracts . . well. say why do you have a privilege? FC: Well I'll consider this . suppose you are a manager, a vice president. very well paid of course. . of a company an you discover that one of your suppliers . is cheating you . or supplying you with insufficient goods ...

Extract 7 demonstrates an insertion sequence to maintain accountability where the speaker is preparing his listeners for a lengthy answer/explanation to his question, he prefaces it by "well I'll consider this" thus showing his orientation to the fact that the second pair part is still pending and preparing his audience for a longer answer sequence before completion. In fact the answer (extract 8 below) is 6

It is interesting to see that the single lexical item "forever" can act as a question even in a lecture because intonation, context and location induce the listener to perceive it as such.

Julia Bamford

164

preceded by a lengthy explanation which takes up almost a page of transcribed text. The answer itself when it is finally uttered is negative: (8) (Frank Cross Business Law) FC. No . you're deemed to have held a privilege of important advice . that your company needs to hear . . you can give that without worrying about being held liable for wrongful interference with contract

5. Student Answers In the corpus we find very few examples of questions in which students actually try to formulate an answer. One example is extract 9 below which also displays an insertion sequence: (9)

(S.B. Employment, 12.5.98) 1. S.B. Now what's the maximum fee which could be .ah . charged? 2. S.B. Well. let me start with that.. what's the maximum fee that employers could ask for). .ah 3. Stu: (The employment bond. 4. S.B. The entire employment bond of course . . B minus Z (shows this on board)

This is an interesting interaction because there is good reason to suppose that the lecturer did not really intend the question to be answered by a student. The student on the other hand clearly interprets the repetition/reformulation of the first question in which "what's the maximum fee which could be .ah . charged?" where the repetition of the first four lexical items and the reformulation of the rest becomes "what's the maximum fee that employers could ask for" as an invitation to supply the answer. How do we know the speaker did not intend this? In turn 2 he provides an insertion sequence showing his orientation to the fact that the second pair part is still pending but also taking on the responsibility himself for doing this (". let me"). Furthermore the repetition/reformulation of the first question does not have an utterance final rise in intonation and the slight pause before the "ah" indicates he was about to supply the answer when the student began to speak in overlap. The repetition of the student's answer plus the intensifier "entire" and the marker "of course" indicates agreement; however the intersubjectivity marker "of course" also serves to emphasise the asymmetry between the two participants in this dialogue. The audience probably interprets the whole exchange as a student misinterpretation of the lecturer's intentions since no more student/lecturer exchanges take place for the rest of the lecture, indicating perhaps that they perceive the lecturer as not wanting interruptions.

Question and Answer Sequencing in Academic

165

Lectures

6. "Well" and "So" Prefacing Questions and Answers A further commonly recurring pattern in connection with questions is the presence in the immediate proximity of discourse markers such as "so" and "well". Schiffrin (1987). discusses the role of "well" in the accomplishment of conversational coherence in relation to question/answer pairs where both question and answer are needed to complete a proposition. She finds that "well" prefaces answers to "yes/no" questions less frequently than answers to "wh" questions. Although this occurs in the present corpus also, the incidence of "yes/no" questions was low compared to that of "wh" questions making generalisations risky. Thus we can only tentatively assume that the use of "well" as a preface to an answer is influenced by the form of the question. Unlike Schiffrin's analysis, "well" in this corpus is found as a preface both to the question and to the answer as exhibited in extracts 10 and 11 below: (10) (Frank Cross Tort Law) F.C. Well what's improperly F.C. This is difficult because in America.. and the rest of the world

suppose

(11) (Frank Cross Tort Law) F.C. What would be a wrongful means of acquiring a trade secret? F.C.. Well industrial espionage is clearly a wrongful means

As these examples, which are typical of this corpus, demonstrate, "well" can preface both questions and answers but it fulfils a different function when so doing. When "well" prefaces a question, it seems to indicate that the answer is going to be problematic in some way. This problematicity can involve an answer which will require lengthy discussion, sometimes this answer can be slightly less specific and more speculative. This use of "well" as a preface to a question which preludes a complex, speculative, non-specific answer, seems to be peculiar to academic lectures since the speaker has control over both question and answer and can foresee that the answer will have these characteristics. Thus "well" can preface the question but in this case it acts as a signal to the listener about the type of answer to expect. Extracts 12 and 13 below are examples of this: (12) (D. Foley Rational Expectations) D.F. In thinking about this I began to ask myself. well . what would be the grounds for assuming rationality on the part of human beings? D.F. . . One answer would be that human beings have evolved biologically . not sociologically but biologically evolved . in our DNA . or however that works . . to be rational

Julia Bamford

166 (13) (D. Foley Economic Theory)

D.F. I think it would be use:ful . . whenever rationality is used . if we said well . do we include rationality . in the definition of rationality . selfishness?

On the other hand examples of "well" used to preface an answer typically precede a relatively straightforward answer (15): (14) (Frank Cross Business Law) FC: To sue for wrongful interference with prospective business you have to establish a strong possibility that you'd get that business . . . how would you do that? FC:. well. one place where it often comes up is retaining customers

Schiffrin finds that "so" occurs in relation to the organisation of transition in participation frameworks in which the current speaker shifts the responsibility of the interaction onto the hearer. "So" seems to occur at a transition relevance place where there is a potential change in floor holder (Sacks, Schegloff, Jefferson 1978). Although this changeover does not necessarily occur and the offered turn is sometimes not taken up, the potential is often indicated by "so". It marks the opportunity to change speaker. In academic lectures "so" prefaces a question as in the extract 15 below: (15) (D. Foley Economic Theory) DF: So what's the point? DF: . It's not exactly to criticise Gintis and Bowles for making a particular assumption that they do . . but. I suppose it's . ah . a way of asking yourselves whether some form of bilateral negotiation . might be imagined to model the types of effort inducement that they talk about.

In this corpus "so" always prefaces the question and serves the function of marking a shift in focus in the discourse. It is interesting to see the use of "so" in relation to Schiffrin's findings and note that the lecturer uses it when he moves into his questioning mode in which he assumes the role of the student questioner. "So" always comes at the beginning of the sequence before the question and it serves as a sort of preface to the question itself; a signal to the student listener of the switch in frame, of the movement into the question/answer pair in addition to the switch in interactional role from speaker to questioner. Thus there seem to be analogies between the use of "so" in spontaneous conversation and its use in academic lectures.

7. Series of Questions A final question/answer pattern represented in the corpus is the series of questions. This consists of several questions closely linked semantically,

Question and Answer Sequencing in Academic

Lectures

167

grammatically and lexically in that they are often reformulations of the first question and contain many lexical repetitions and grammatical parallels. Sometimes the series of questions receives an answer, sometimes they serve the rhetorical purpose of underlining the problematic nature of the discussion and the fact that many of the questions posed do not have a satisfactory answer. Example 16 discloses this problematic questioning in a post graduate lecture where the speaker proposes several questions which pose two different and opposite hypotheses. These are: i. does rationality help us to get relevant answers; ii. does rationality lead to answers which, however ingenious, don't explain much?

The answer to these questions could be interpreted as a criticism of the theory in question and a potentially face threatening act but by formulating the criticism as a question/answer sequence where most importance is given to the question component, this FTA is deflected, because the speaker is perceived as giving a considerable degree of consideration to the question before making a criticism (Brown and Levinson 1987, Myers 1989): (16) (D.Foley Rational Expectations) DF: Does rationality HELP us? DF . does . i .direct ( ) to ask the right questions . that is questions that have an answer and are relevant to the issue at hand ? DF: . . OR is it tending to lead us to . perhaps ingenious but extremely . complex arguments which in fact don't have much robustness or explanatory power DF. . and . I think . the answer to this question is . just up front.. is that they don't have a general answer . the answer on the specific issue.

8. Conclusions Despite being uttered by the same actor question/answer adjacency pairs in this corpus of academic lectures function interactionally and it is this property of interactivity which makes the academic lecture such a widely used teaching technique. Why are question/answer sequences interactive even in cases like this where only one interactant has the floor and the other participants take a merely passive role? We have seen how adjacency pairs are cohesive because both pair parts are necessary for a proposition to be complete. But the explanation of why a lecturer decides to present his information in the question/answer sequence format lies in his attempt to approach the interactivity of a similar sequence from spontaneous conversation. Questions are used in conversation to force people into social exchanges. Questions are also a way of creating intersubjectivity, the sharing of interpretations and meanings between one who is already a member of the discourse community and those who are being co-opted into it (Swales 1990).

168

Julia Bamford

These functions of questions and answers are transposed from spontaneous conversation to other types of interaction such as academic lectures where they create a dialogic atmosphere between speaker and listeners. When a question is asked in spontaneous conversation, an answer is routinely expected. Even in academic lectures when a question is posed the listener normatively expects an answer to be provided. A question puts him on the alert for the second pair part completion. The listener feels an unanswered question as incomplete and expects the speaker to account in some way for the missing element. Lecturers use this strong expectation as an attention getting device, a rhetorical strategy designed to focus the student's attention on the answer. This study raises the problem of the displacement of question asking from the listener to the speaker because the questions in this corpus are not to be considered rhetorical questions but real questions. Although they are asked and answered by one and the same person they have an answer and are clearly interactional. Obviously they are also an effective rhetorical strategy since they focus the attention of the listener on the answer by providing him with a signal in the form of a question.

References Biber, Douglas (1988): Variation across Speech and Writing. Cambridge: Cambridge University Press. Brown, Penelope and Levinson, Stephen (1987): Politeness. Cambridge: Cambridge University Press. Drew, Paul and Heritage, John (1992): Analysing talk at work: an introduction. In: P. Drew and J. Heritage (eds.) Talk at Work: Interaction in Institutional Settings. Cambridge: Cambridge University Press, 3-65. Flowerdew, John (ed.) (1994): Academic Listening. Cambridge: Cambridge University Press. Goody, Ester (1978) Towards a theory of questions. In: E. Goody (ed.) Questions and Politeness. Cambridge: Cambridge University Press, 17-43. Halliday, Michael and Hasan, Ruqaiya (1976): Cohesion in English. London: Longman. Hasan, Ruqaiya (1985): The texture of a text. In: M. A. K. Halliday and R. Hasan Language, Context and Text: aspects of language in a social-semiotic perspective. Oxford: Oxford University Press, 70-96. Jefferson, Gail (1972): Side sequences. In: D. Sudnow (ed.) Studies in Social Interaction. New York: Free Press, 294-338. - (1979): A technique for inviting laughter and its subsequent acceptance/declination. In: G. Psathas (ed.) Everyday Language: Studies in Ethnomethodology. New York: Irvington, 7 9 96. Levinson, Stephen (1983): Pragmatics. Cambridge: Cambridge University Press. Myers, Greg (1989): The Pragmatics of Politeness. In: Research Articles Applied Linguistics, 10.1.1-35. Sacks, Harvey (1987): The Preferences for Agreement and Contiguity. In: G. Button and J. R. E. Lee (eds.) Talk and Social Organization. Clevedon: Multilingual Matters, 54-69.

Question and Answer Sequencing in Academic

Lectures

169

Sacks Harvey, Schegloff, Emanuel A. and Jefferson, Gail (1974): A Simplest Systematics for the Organization of Turn Taking in Conversation, Language 50, 696-735. Schegloff, Emanuel A. (1984): On Some Questions and Ambiguities in Conversation. In: J. M. Atkinson and J. C. Heritage (eds.) Structures of Social Action. Studies in Conversation Analysis. Cambridge: Cambridge University Press, 28-52. Schiffrin, Deborah (1987): Discourse Markers. Cambridge: Cambridge University Press. Swales, John (1990): Genre Analysis. Cambridge: Cambridge University Press.

Luis Pérez-González Agenda Negotiation in Emergency Calls: Modelling the Dynamics of Frictional Encounters 1. Introduction This paper reports on our dynamic modelling of the conversational interaction which characterises the interrogative series, a negotiation-based constitutive stage of citizens' calls for emergency assistance. Drawing upon the analysis of a corpus of real data, we discuss the relevance of the institutional setting and the citizens' emotional distress to the conversational negotiation in each encounter. Our dynamic modelling is represented by means of flowcharts which are able to formalise how current conversational actions affect later interactional choices. Such flowcharts draw upon a three-layered synoptic description of conversational structures in emergency calls, the units of analysis at each layer being the agendaridden generic stages, exchanges and moves. This paper draws upon the basic insight that frictional encounters will result in idiosyncratic patterns of interactional organisation that can be exposed when compared with the unfolding of conventional encounters. In this respect, our discussion will focus on the implications of such distinctive dynamics for the institutional evaluation of the citizens' incoming reports in terms of credibility.

2. Rationale for the Study of Dialogue in Emergency Calls The increasing access of citizens (C) to the emergency services in modern societies has had serious side effects and the call-takers (CT) have witnessed a remarkable proliferation of hoax or malicious calls in which citizens report nonexistent emergencies. While institutional policies state that emergency assistance cannot be denied on the grounds of CT's doubts about the genuineness of an incoming report, his/her assessment of the latter in real time has an important influence on the number of assistance units dispatched and the priority given to the mobilisation in each case. In this respect, this paper outlines our attempt (Pérez-González 1998a) to come up with a set of discourse-based criteria that could enable CT's categorisation of a given report in terms of credibility. Our study draws upon a corpus of sixty-six hoax 999 calls to the switchboards of a number of British fire brigades, ambulance dispatch-centres and police-stations, divided into three different sections depending on the nature of the incident in question:

172

Luis • •

•

Pérez-González

Section A includes calls seeking the dispatch of fire brigade or ambulance units made with reference to incidents which are currently in progress. Section B consists of a number of citizens' reports, complaints and help-requests to the police. These concern not only current emergencies, but also other incidents which C regards as potential safety hazards. Section C consists of personal and bomb-threats.

The examples included in this study are all excerpts from our corpus samples. The recordings were kindly provided by J.P. French Associates, a York-based consultancy, which had previously used this material in the course of their speaker-profiling work as expert witnesses in the field of forensic linguistics. The recordings were carefully transcribed and later analysed according to the model we will outline later in this paper. This paper reports on our analysis of one specific segment of the canonical structure of 999 calls, i.e. the interrogative section or interrogation series (IS), as it is precisely this stage that provides CT with the information required to process the request for assistance and determine the urgency of the latter. As 999 calls are a form of institutional interaction, the length of the interrogation, the number of issues to be elucidated and the order in which these are raised should depend, in principle, on CT's discretion. But conversation is not always that smooth. Interlocutors may fail to understand each other; they may wish to confirm information items they have obtained under poor hearing conditions, attempt to correct what they regard as a wrong representation of their own words or, more generally, show their disagreement with the overall unfolding of the encounter. In this respect, it should be noted that citizens phoning the 999 services are subject to an obvious emotional distress. Under these circumstances, C is allowed to have a voice in determining the direction and shape of the interaction. Within these parameters, our study draws upon the contention that this "matching [of the] interactants' expectations and goals" (Ventola 1989: 136) would be particularly difficult to achieve in hoax 999 calls. In broad terms, it was hypothesised that: •

• •

During the Interrogative Section (IS), CT attempts to gather enough information to provide C with a fast and efficient response. However, C's malicious agenda will lead him/her to oppose and challenge CT's probes for certain elements of information. As a result of C's evasive agenda, the unfolding of IS will differ according to the genuine/malicious nature of the report. Conversational markers of C's malicious agenda can be exposed by comparing the canonical and deviant unfolding of IS in genuine and hoax encounters, respectively.

In sum, our focus on the modelling of the IS was developed from the basic insight that hoax ISs constitute: deviations from or disruptions to the generic flow. However, the reason that we can recognise these deviations or disruptions as such is that there is an underlying abstract structure to each

Agenda Negotiation in Emergency Calls: Modelling the Dynamics of Frictional

Encounters

173

generic type. To recognise deviations from a pattern is a verification of the facts of the underlying patterns. Because there is a generic expectation, participants are able to negotiate in any instance the actual unfolding of talk. Eggins and Slade 1997: 234

3. Modelling the Interrogative Series One of the key claims of systemic functional linguistics is that the contextual configuration of a given event correlates with certain linguistic patternings. As far as conversational encounters are concerned, the theory of genre helps to establish how texts work through stable unfolding structures, that can be formalised at different levels of conversational organisation, in order to achieve their social purposes (Eggins and Martin 1997). In this respect, 999 calls can be defined as staged, goal-oriented social processes whose typical sequences of realisation, i.e. "generic structure potential" (Hasan 1984) can be formulated as follows: Opening Section A Request Response A Closing Section

for

Assistance A

Interrogative

Section A Organisational

(the caret A sign means that the stage to the left precedes the one to the right).

In this paper, we will advocate the need to implement a dynamic modelling of the IS in 999 calls at different levels of conversational organisation. Putting this issue in perspective, our model consists of three layers of analysis, which exclude nonverbal behaviour, as represented in Figure I below. In the remainder of this paper we will carry out a top-down account of this framework which sees the IS as a sequence of generic stages orienting to the accomplishment of the interactants' agenda: •

•

At the top structural stratum, IS would appear to be organised as a succession of agendaridden, topically-bound generic stages. The theoretical status of such stages resembles that of Swales' (1990) notion of 'moves', i.e. their counterparts in the organisation of written discourse. Accordingly, each generic stage constitutes one step towards the accomplishment of the overall communicative purpose, insofar as it orients to one of the specific requirements on which the dispatch of assistance is contingent. At the intermediate and lower strata on the cline of analytic delicacy, we find the exchange and moves. The exchange may constitute a generic stage on its own, provided that the amount of information transferred within that exchange fulfils the requirements posed by the generic stage at issue. However, a single exchange does not usually map on to a distinct agenda-defined task. That being the case, the agenda which governs the unfolding of each generic stage manifests itself in the initiation of further exchanges until the pertinent information has been gathered and CT finds it pertinent to move on to the next stage. Space restrictions preclude discussing the level of grammatical structures which realise the conversational moves.

Luis

174

GENERIC STAGE I

Exchange 1

K2/A2 Kl/AI (K2f) (A2f)

Exchange x

K2/A2 cf beh elfy - reify eh-rch

Kl /Al (K2f)(A2f)

GENERIC STAGE n

Exchange 1

Kl /Al (K2f) (A2f)

Exchange y

K2/A2

K2/A2 cf beh elfy - rclfy ch-reh

Pérez-González

cf beh elfy - rclfy ch-reh

Kl / A l (K2f)(A2f)

cf beh elfy - rclfy ch-reh

Figure 1: Conversational structure in the modelling of IS

Now it should perhaps be highlighted at this point that CTs dealing with potential hoaxers have to assess the ongoing report while the interaction progresses. Consequently, any attempt to provide the dispatchers with useful filtering tools should ideally orient to model the process of interaction rather than the ensuing text structures. In other words, our modelling task should aim to formalise what may come next at each point of the interaction. Ventola (1989) constitutes probably the most thorough account of different conversation-modelling frameworks and draws a distinction between "synoptic" or "static" and "dynamic" approaches. It is argued that models within the first strand, i.e. Van Dijk's transformational-generative framework, the Birmingham Discourse Analysis school and Hasan's Linear Generic Structure Potential, succeed in formalising the canonical sequence of generic structure elements but fail to account for the process whereby interactants, once they have got involved in the conversation, negotiate with the fellow participant on how to proceed when their respective planned agendas do not match. In this respect, Transition Networks (O'Donnell 1990) and Dynamic Flowcharts (Martin 1985; Ventola 1987) constitute alternative formalisms used for the dynamic modelling of interaction. Broadly speaking, dynamic formalisms represent "the choices which can be made at each point throughout the interaction, and how the choice affects future behaviour choices" (O'Donnell and Sefton 1995: 65). In sum: the three-layered model of analysis provided above represents the types of interaction structures that can be produced in IS interaction at different levels of organisational delicacy. Therefore, our project focused on these strata in turn, shifting from a synoptic to a dynamic perspective. In other words, we advocate the need to move from the analysis of the conversational synoptic structures to the study of their interactional consequentiality, i.e. the structures

Agenda Negotiation

in Emergency Calls: Modelling the Dynamics of Frictional Encounters

175

they activate in progressing towards the accomplishment of the communicative goal.

4. Generic Stages: Modelling at the Top Level of Conversational Organisation Pérez-González (1998a) explored the register choices which apply in 999 calls and found that this encounter-type revolves around a restricted set of tenor, field and mode variables. From the point of view of tenor, calls to the police originate from professional/layman interaction, the actors' performance being constrained by variables such as power, contact, emotional involvement and genuineness. The field choices are oriented to emergency matters and the ensuing informationprocessing activities, whereas their mode counterparts are determined by sophisticated channel-related constraints indicating that language is used in an action-oriented, non-face-to-face domain of interaction. The stability of these system networks of choices across specific emergency encounters is to be held accountable for the fact that IS revolves around a number of stable generic stages that deal with institutionally relevant tasks. Generic stages have been already defined in synoptic terms. IS is thus a succession of generic stages which together work towards the participants' goals. In other words, each generic stage seeks to accomplish one of the tasks required for an appropriate and efficient management of the request. After extensive acquaintance with our transcripts, we were able to identify several generic stages, i.e. tasks: • Focus on the nature of the problem. The fact that most emergency centres use technologicallyenhanced consoles forces CT to secure an accurate insight into the nature of the problem. Since the product of the interrogation has to be transmitted electronically, the results of the interaction between C and CT should be precise enough to allow CT to choose the relevant processing-codes and type them into the computer-aided dispatch form. The length of this stage, namely, the number of exchanges used to accomplish this communicative task, varies according to the accuracy of C's initial report. • Focus on C's identity. This is another of the most widespread generic stages displayed by the samples gathered in our corpus. Again, the number of exchanges devoted to the accomplishment of this task depends on C's willingness to furnish the items of information required. C's identity is required by CT in order to hold somebody responsible for the organisational mobilisation, as "police dispatchers are sensitive to the possibility that Cs may be motivated to settle scores with other individuals by getting them into trouble with the police" (Zimmerman 1992: 466). • Also crucial are the 3 different location-eliciting stages, i.e. Focus on the location of the emergency, Focus on the production of orienting directions and Focus on C's whereabouts. The occurrence of the first one and its length depends on C's ability/willingness to formulate the location in canonical or organisationally relevant terms (cf. Pérez-González 1999). In the second generic stage CT elicits landmark-based formulations of location, i.e. the location of the fire with respect to a recognisable public building, shopping precinct, traffic junction, etc.

176

Luis

Pérez-González

• Section B calls often contain a Focus on the description of the suspect generic stage, a Focus on the suspects' whereabouts and a Focus on the aftermath of the incident. These stages are particularly prominent in section B of our corpus since they rely on the assumption that C is being/has been a direct/indirect victim/witness of the incident for which the suspect is to be held responsible. Consequently, the occurrence of these stages is contingent on the specific features of the incident at issue. • Finally, there is the case of the Shift the focus on... stages, also documented in our corpus. At each point of the interaction, CT may wish to pursue a previous generic stage further, but to do so using a slightly different interrogative strategy. Such loop backs aiming at the achievement of incomplete tasks recur until C's response becomes relevant to the institutional agenda in terms of content and quantity. Putting it in systemic functional terms, this generic strategy is available prosodically throughout the interaction.

So far we have outlined the types of tasks which are available for negotiation at the beginning of IS. Now a dynamic modelling of interaction should be able to indicate that the successful completion of a task and, hence, of a generic stage, modifies the interaction context and leads to a change in the available task potential. In other words, successful tasks are no longer available unless CT considers that the result of the task may be improved by using a Shift the focus on... generic stage. As opposed to the synoptic modelling of the IS in terms of generic stages, its dynamic counterpart should be able to formalise how the decision to begin by initiating a given generic stage affects future choices. Martin (1985) and Ventola (1989) have represented the relationship between each action and the change in the range of options that follow by means of flowcharts which link each action to consequent actions or decisions. Flowcharts "capture how texts [which are] structurally unique, but at the same time generically the same, were generated" (Ventola 1989: 136). The flowchart displayed in Appendix I below intends to represent how the shared construction of IS is achieved dynamically by the interactants. The notational conventions for the interpretation of this figure are summarised in the key section, below the flowchart itself. As the reader will note, the flowchart helps the analyst to represent the whole of IS but, at the same time, to account for each generic stage individually. This dynamic formalism allows for a variable number of generic stages (n) without specifying the goal to which they orient (x/y/z) or the order in which the stages are expected to combine. Furthermore, this flowchart also acknowledges the possibility of iteration, i.e. the recurrence of certain generic stages which become prosodically available at the dispatcher's discretion.

5. Exchanges and Moves: Modelling at the Intermediate and Lower Levels of Conversational Organisation Drawing upon the systemic functional model, I have elsewhere (Pérez-González 1998b) carried out a structural characterisation of IS in terms of its constitutive exchanges. According to this analysis, IS in 999 calls typically consists of a

Agenda Negotiation

in Emergency

Calls: Modelling

the Dynamics

of Frictional

Encounters

177

succession of exchanges selecting B-events, i.e. those in which the speaker A inquires about information which only B has or requests an action which only B can carry out. In more specific terms, C is the primary knower in the knowledgeorientated exchanges within IS, insofar as s/he is asked by CT to provide information regarding the reportable incident following these two exchange combinations: K2 K1 K2f and K2 K l . For his/her part, CT is the primary actor in action-orientated exchanges within IS, inasmuch as s/he is responsible for the dispatch of the assistance units requested by C according to the following formulae: A2 A1 A2fand A2 A1 (see Berry 1981a and 1981b for a full account of these labels and formalisms). Now we have argued in previous sections that the channel of communication in which 999 calls take place and the interactants' uneven degree of familiarity with the organisational processing policies may help to explain why CT is in many cases unable to obtain elements of information within the boundaries of a single exchange or why C is not always able to secure an immediate dispatch of the assistance required. In order to account for such conversational contingencies, Ventola (1987: 105) draws heavily upon Martin and advocates the differentiation between the 'synoptic' and 'dynamic' exchange systems. On the one hand, the former models exchanges according to Berry's formula and thereby caters for the expected trajectory of the encounter. On the other hand, the latter purports to "capture the potential dynamic aspects of genres more efficiently" and thereby caters for any deviations from the typical generic unfolding of the exchange structure in an institutional setting. The differentiation between the synoptic and dynamic systems of moves is conspicuous in the two rightmost columns of our sample analysis in Appendix II. According to Ventola, there are three separate types of moves whose generation does not respond to the synoptic system of conversational structure, but to its dynamic counterpart. To start with, the suspending moves "concentrate on checking and giving assurance about the transmission of knowledge/action" (1987: 105-106). Among the different types of suspending moves, we find the 'confirmation move' (cf.), the 'response to confirmation move' (rcf) and the 'backchannel signals' (bch) - "which give assurance to the speaker that his message is being received" (1987: 106). Second, there is also the case of elucidating moves, i.e. 'clarification-seeking moves' ( c l f y ) followed by responses to the latter (rclfy). The third and last type of dynamic moves, i.e. the aborting one, is realised by 'challenges' (ch) - which may be followed by a 'response to the challenge' (rch) - and relates to Halliday's (1985) concept of 'discretionary speech functions', insofar as challenging moves do not comply with the interactional expectations enacted by the immediately preceding one. Dynamic moves may occur either in isolation (e.g. bch) or combined into dynamic exchanges (e.g. cfrq-rcfrq). According to our corpus data, the use of such moves is

178

Luis Pérez-González

to be explained in terms of both endogenous and exogenous criteria. The first case can be illustrated by [bch, clf/rclf] moves, which are motivated by CT's processing needs and aim at an efficient management of C's report. On the other hand, our transcripts attribute the remaining dynamic moves to exogenous reasons such as the hearability conditions and the quality of the line, which can never be overlooked when analysing phone-mediated conversations. At any rate, both the supply of a discretionary answer and the withholding of a response are inextricably related to the shift from the synoptic to the dynamic system of exchanges. Synoptically speaking, the number of exchanges involved in the realisation of each generic stage is highly variable insofar as CT can initiate as many exchanges as necessary for the successful completion of the current task. A dynamic modelling of these exchanges, however, should be able to formalise the interactional implications of discretionary conversational moves, i.e. incorporate the dynamic systems of exchange generation. In this respect, the foregoing flowchart should undergo a process of refinement and raise the interactants' awareness that the initiation of dynamic exchanges are consequential with regard to later conversational options. The two flowcharts represented in Appendix III and Appendix IV below are precisely intended to account for such interactional contingencies within 999 calls. Accordingly, they constitute a more comprehensive account of the dynamic processes and decisions involved in the shared, real-time construction of a given generic stage. In other words, they are a detailed reproduction of the steps to be followed by each party once a generic stage has been initiated by CT. The flowchart in Appendix III, for instance, reads from right to left and represents CT's interactional choices within each stage. From CT's standpoint, the failure of any genuine informant to supply a satisfactory response can only be explained as a result of channel-related difficulties or his/her lay status - which does not allow him/her to understand the relevance of the current question to the organisational decision. At any rate, both contingencies are repairable by means of further exchanges of information/directions, which consist of dynamic moves and result in the dynamic expansion of the core structure of the current generic stage. The same process is formalised in the flowchart represented in Appendix IV from the opposite perspective. Accordingly, it reads from left to right. These expanded versions of the core structure at the level of generic stages have further advantages. In this respect, note that CT's sequence of steps starts with a task (inquire about X), whereas C's does so with a decision (respond to query about X). In other words, the flowcharts in Appendix III and Appendix IV allow for the possibility of C's deferring his/her reply to CT's question until certain hearability/relevance problems have been solved. In sum, these expanded flowcharts acknowledge the consequentiality of C's own interactional discretion

Agenda Negotiation in Emergency Calls: Modelling the Dynamics of Frictional Encounters

179

and, moreover, represent in detail the process whereby CT attempts to get the trajectory of the generic stage back on the right track. Consequently, these dynamic formalisms would seem to endorse Martin's claim that "dynamic systems are invisible (...) as text is formed, [although] they may in fact appear to repair a process which is breaking down (...) In a sense it is only when something goes wrong that process can be distinguished from text" (1985: 265).

6. Conclusion: The Implications of Dynamic Modelling for the Assessment of Emergency Reports The flowcharts formalise what we may regard as idealised generic expectations on the part of the interactants as they approach the conversational event. In Butler's words, "as members of a particular culture, with a particular social structure, each of us has available to him a range of behavioural options (a 'behaviour potential') for use in certain types of social context" (1985: 59). However, the generic expectations do not always hold in practice. In fact, certain gaps between the institutional conventions and the actual realisation arise, for instance, when the C's performance is determined by a malicious agenda. In more concrete terms, the hoaxer's conversational behaviour seems to be characterised by his/her deliberate attempts to depart from and exploit the normative generic expectations with a view to "achieve private intentions within the framework of socially recognised purposes" (Bhatia 1993: 15). Now our flowchart-based modelling of the ISs in our corpus of hoax emergency calls revealed that the former can enable the institutional evaluation of the citizens' reports in terms of credibility. Although space restrictions preclude discussing these contributions in detail, let us highlight that hoax emergency calls incorporate a set of generic stages which has not been mentioned so far. In all three cases, they are activated when CT suspects that C's refusal to comply with the institutional requirements might respond to his/her malicious agenda. In sum, the possibility of such a tactical approach to genre construction on the part of CT manifests itself in the activation of these stages: • • •

Focus on the explicit realignment of C's role. When C withholds information, CT often opts to remind the former of his/her interactionally relevant role, i.e. that of informant. [Focus on C's claim], which aims to glean C's sources of knowledge or challenge the latter's categorisation of the problem when these appear inconsistent in CT's eyes. [Focus on the means to achieve the communicative purpose]. In activating this stage, CT attempts to confirm the motivation behind C's uncooperativeness by spelling out the relevance of his/her compliance to the achievement of their shared transactional goal.

More generally, the interrogations in hoax emergency calls display a higher rate of monitoring [Shift the focus on...] stages. At the lowest level of conversational organisation, interactants in hoax ISs produce more dynamic conversational moves

180

Luis

Pérez-González

than in their canonical counterparts. While CT uses them to bring C back on the right interactional track, C opts for dynamic moves to express opposition or disagreement with the processing of their request. All in all, the scrutiny of our hoax samples (Pérez-González 1998a) suggests that CT's appraisal of C's veracity depends on three key issues. First, on C's overall degree of compliance or cooperativeness with the organisational requirements, which manifests itself in the topical and sequential (non)-relevance of C's responses to CT's questions. Secondly, on the vulnerability of C's "interactional stance" (Whalen et. al. 1990), i.e. the account of the reasons which have warranted C's categorisation of the ongoing event as an emergency and the circumstantial bases whereby s/he has become actually involved in the incident reported. Finally, on C's (un)-willingness to modify his/her original stance following CT's confirmation-seeking reformulations of the prior statements.

References Berry, M. (1981a): Systemic linguistics and discourse analysis: a multi-layered approach to exchange structure. In: Coulthard, R. M. and M. Montgomery (eds.) Studies in Discourse Analysis. London: Routledge and Kegan Paul; 120-145. Berry, M. (1981b): Polarity, ellipticity and propositional development. Their relevance to the wellformedness of an exchange. Nottingham Linguistic Circular, Special Issue on Text and Discourse, 10(1)36-63. Bhatia, V. (1994): Analysing Genre: Language Use in Professional Settings. London: Longman. Butler, C. S. (1885): Systemic Linguistics: Theory and Applications. London: Batsford. Eggins, S. and J. R. Martin (1997): Genres and registers of discourse. In: T. van Dijk (ed.) Discourse as Structure and Process. Discourse Studies (1) A Multidisciplinary introduction. London: Sage Publications; 230-256. Eggins, S. and D. Slade (1997): Analysing Casual Conversation. London and Washington: Cassell. Halliday, M. A. K. (1985): An Introduction to Functional Grammar. London: Edward Arnold. Hasan (1984): The nursery tale as a genre. Nottingham Linguistic Circular, 13: 71-102. Martin, J. R. (1985): Process and text: two aspects of human semiosis. In: Benson, J. D. and W. S. Greaves (eds.) Systemic Functional Approaches to Discourse. Norwood, N.J.: Ablex; 213-228. O'Donnell, M. (1990): A dynamic model of exchange. Word, 41 (3) 293-327. O'Donnell, M. and P. Sefton (1995): Modelling telephonic interaction: a dynamic approach. Interface: Journal of Applied Linguistics, 10(1) 63-78. Pérez-González, L. (1998a): Towards a dynamic model of discourse: issues of a forensic-oriented study of spoken interaction, Unpublished PhD thesis, University of Valencia. Pérez-González, L. (1998b): The conversational dynamics of interactional dispute in conflictive calls for emergency assistance: a single-case study. In: Sánchez-Macarro, A. and R. Carter (eds.) Language as choice. Amsterdam/Philadelphia: John Benjamins; 265-290. Pérez-González, L. (1999): Interactional implications of computer mediation in citizens' calls for emergency assistance. In: Pemberton, L. (ed.) Words on the web. Computer-mediated Communication. London: Intellect. Ventola, E. (1987): The Structure of Social Interaction: A Systemic Approach to the Semiotics of Service Encounters. London: Frances Pinter Publishers. Ventola, E. (1989): Problems of modelling and applied issues within the framework of genre. Word 40(1/2): 129-161.

Agenda Negotiation in Emergency Calls: Modelling the Dynamics of Frictional

Encounters

181

Whalen, M. and D. H. Zimmerman (1990): Describing trouble: practical epistemology in citizen calls to the police. Language in Society, 19: 465^492. Zimmerman, D. H. (1992): The interactional organisation of calls for emergency assistance. In: P. Drew and I Heritage (eds.) Talk at Work. Cambridge: Cambridge University Press; 418—469.

Appendices Notational Conventions for the interpretation of dynamic flowcharts: (i)

Social process;

(yes/no)

Final decision; Task to be carried out;

C^^)

Structural element;

O

Decision to be taken; j

Skip/repeat an element

182

Luis Pérez-González

Appendix 1: Flowchart representing the dynamic modelling of IS in 999 calls SOCIAL PROCESS COMPLAINT-TAKER (CT)

CALLER (C)

Agenda Negotiation

in Emergency Calls: Modelling the Dynamics of Frictional Encounters

183

Appendix 2: Sample Analysis The following table displays the analysis of a generic stage in the IS of an emergency encounter. For the sake of clarity, the IS has been edited in such a way that the constitutive exchanges of the generic stage in question appear in aggregated form. The first column headed [Ex.] identifies the position of the exchange in this encounter, while the second column [Transcript]

reproduces the lines of the

dialogue. The third column [SysJ represents the discourse function from which the speaker chooses the "synoptic system" on the discourse stratum, whereas the fourth one [Dys] shows the element of the "dynamic system" of discourse which the move realises. The full analysis of this encounter can be found in Pérez-González (1998: 2 8 8 - 9 0 ) .

[Transcript]

[Ex]

[Sys]

[Dys]

c stage: Focus on the C's whereabouts 003 004 007 008 009 010 011

10

012 013 025 026

12

030

13

031 032 033 034 035 036

where are vu:? inne town centreuv birmingham! (10) vou'r in birminaham, whereabouts? = C: = town centre (h) (h). CT: ye- where = alright th 'n. (.) whe//rel abouts in (the) town cen//tre] uv birmingham? = C: b-] C: b-1 C: = by "(prudential) (h) (h) (2.0) caniu hurrv'D Dlea::se? = I'm shittim'mvself here! CT: wudiu answer m v Questions Dlea::se CT: (2.0) ((impatient)) well if you tell me where you aire I'll get somebody down to juh! C: (1-0) uh(1.0) eh... (I'm standing out here in the centreuv birminaham). CT: ((conciliating)) I know you're in birmingham and you're in the town centre. = b't the town centreuv birmingham's quite big. = now tell me (.) WHERE a:re you? C: = okay boss, prudential. CT: prudential? yeah, birmingham, j'st outside prudential. C: (1.0) by the I S B. CT: by the TSB bank. C: yeah. C:

K2 K1 K2 K1 K2F K2

K1 A2 ch K2

K1 K1

K2 K1 cf rcf cf rcf

184

Luis Pérez-Gonzalez

Appendix 3:

Flowchart representing CT's generative choices within each generic stage

Appendix 4:

Flowchart representing C's generative choices within each generic stage

Boris Pritchard and Damir Kalogjera On Some Features of Conversation in Maritime VHF Communication 1. Introduction: Maritime English and the Language of Maritime VHF Communications The language of maritime VHF radiotelephone communications, the subject of this paper, can be regarded a particular subset of English whose lexical component, grammatical and discourse features: (a) (b) (c)

are appropriate to the requirements and restrictions of specific communication situations in navigation, both at sea and in port, meet the communicative needs and linguistic capabilities of users (participants) in conveying and exchanging messages within outboard or inboard communications, and meet the specific technical, i.e. extralinguistic requirements of using VHF radiotelephone in conversation and broadcasting.

English for maritime communications is a specific, narrow-scope realisation of Maritime English, which in turn, following Halliday, Mcintosh, Strevens 1964, Crystal and Davy 1969, and Bhatia 1993, represents a language variety adapted to a wide range of needs appropriate to the use (e.g. registers of navigation, seamanship, marine engineering, communications, maritime law and shipping business) and users of such a language variety in the specific speech community and the maritime-related environment. This highly restricted sub-variety of maritime English is characterised by a limited vocabulary and simple grammatical structure to suit the specific requirements of interpersonal communication and interaction with similar goals as mentioned before. It is also constrained by situational and contextual dimension (Bhatia 1993, Coulthard 1985, Hatch 1992) such as subject-matter or domain (navigation, safety, shipping, maritime law, etc.), field (e.g. distress communications), mode (cf. spoken vs. written varieties; e.g. those to be read or uttered, transcriptions of spoken text, etc.) and tenor or attitude (e.g. formal - informal, stiff - loose, polite - impolite, standard - colloquial styles in maritime discourse and communications).

186

Boris Pritchard and Damir

Kalogjera

2. Maritime English in VHF Communication - General Features Though never obtaining a mandatory status, English has been used as an internationally-agreed language of maritime communications for over half a century. The risks of navigation and the development of modern technology and communication, combined with the multi-national and multi-cultural ship's crew, necessitate unambiguous and effective communication at sea. This has resulted in a gradual normalisation (Riggs 1989) and, in some instances in standardisation of maritime communications, predominantly using VHF radio frequencies as the medium. Such a standardised linguistic form of maritime communications is a specific kind of a restricted language (Crystal 1987, Bhatia 1992), usually labelled 'speak' in the compounds such as PoliceSpeak, MeteoSpeak, SeaSpeak, AirSpeak, EmergencySpeak etc. (Crystal 1995). As a particular form of discourse in English, maritime VHF communications have a specific purpose (safety of navigation) and are characterised by a set of fixed operational routines (Strevens and Johnson 1982: 5) in dealing with a restricted number of topics. The most typical unit is a transaction between two participants on a single topic or two topics or sub-topics. Opening and closing exchanges tend to be compressed into not more than two turns each. Turn yielding and taking is typically signalled by lexical devices carrying performative power (e.g. over, roger) and is entirely controlled by the speaker holding the turn. For safety reasons no (simultaneous) back-channelling or participants' overlapping is possible because of the technical restraints exclusively allowing simplex mode of working. Question - answer acts are the most frequent form of exchange, with Yes, No, Right, and OK used as acknowledgement/compliance-signalling moves and back-channel signals in the response turn. The speakers are normally the carriers of two different roles, e.g. one speaker (usually a coast station, Vessel Traffic Service (VTS) centre, port control station etc.) assuming the superior position in respect of the other speaker/receiver (i.e. a ship station) in a particular communicative situation. Most frequently the participants are non-native speakers of English. As already stated the conversational structure and format of most messages is simple, routine-like and therefore predictable (consisting of three parts: opening, message exchange, closing) and in this respect VHF communications closely resemble ordinary telephone conversations from which they mainly differ in tenor or attitude (e.g. a higher degree of formality) and a special restraint of an extralinguistic nature (i.e. simplex working system). The grammar is comparatively simple (at least in the more formal, regularised instances of dialogues). The vocabulary of the message part is largely topicdependent, whereas that of the opening and closing sections is restricted and subjected to more or less ritualised opening and closing formulae used in maritime

On Some Features of Conversation in Maritime VHF Communication

187

VHF conversation. Since the number of topics is restricted, the register-bound technical (i.e. navigational and safety) terminology tends to be highly regularised. The fact that the speakers or participants in such conversations are predominantly non-native, places special requirements on their interaction with other non-native and, particularly, with native speakers of English who either seem to neglect any standard form of communicating or are even unaware of the standard. Here is an example of a typical maritime VHF communication as introduced above: (1) Setting: Harbour approach channel / fairway Speakers: MV MARLIN (non-native), (DB) Deutsche Bucht Vessel Traffic Service (nonnative) Main topic: First notice of ship's arrival MARLIN: W - 4 - X - N , W - 4 - X - N . Over DB: Marlin. Deutsche Bucht. Good afternoon. MARLIN: Ah. Good afternoon, Sir. I'm just passing Delta Bravo 13. Over DB: Yes. Your maximum draught? MARLIN: My maximum draught is 5.5 metres. Er..., last port Felixtowe. Destination Hamburg. Over DB: Yeah. And ETA Elbe Light Vessel? MARLIN: ETA Elbe 19.30. Over DB: OK, thank you. Please call me again abeam of Delta Bravo 17. MARLIN: Roger, Sir. I will give you a call back on Delta Bravo 17. DB: Thank you. Good voyage.

3.

The Corpus

To substantiate the above description of the nature of maritime VHF communication in English, a corpus of recordings of authentic VHF communications at sea has been analysed. The corpus was compiled over a period of eight years, from 1990 to 1997. It includes over 500 records, each record consisting on average of 60-90 words. The majority of the recordings were made on a two month voyage on board a container ship in the Mediterranean, Red Sea, US East Coast and Northern Europe, and the rest in some of the world's busiest waters. The records include VHF conversations (90%) and broadcasts (10%). No distress communications were recorded. The principal settings in most records are ship approaches to major ports, fairways or straits both by daylight and at night, in fair and adverse weather. The most frequent communication topics in the recordings were: arrival of a ship at a port, pilot arrangements, berthing, anchoring, departure from a port, towage, manoeuvring, collision avoidance, telephone link calls, cargo operations, bunkering, etc.

188

Boris Pritchard and Damir

Kalogjera

As far as the corpus is concerned, the typical conversation in marine VHF communication normally takes place between the ship's master and the harbour pilot or VTS operator (ship-to-shore communication) followed by brief exchanges between the master/pilot and the tug skipper, or between the master and ship's officers at the ship's communication stations (fore, aft, bridge, engine room: onboard communication). Most frequently the participants in the conversation are a non-native shipmaster speaking to another non-native VTS/port control operator/pilot or shipmaster. In all those instances there is always a relationship of subordination of varying degree, but the necessity for co-operation (safety of navigation) between the carriers of the different roles provides a great deal of redundancy and has a felicitous effect on the overall success of the conversation/communication, i.e. conveying the communicational intent and obtaining an appropriate response.

4. Maritime VHF Communication: Corpus Versus Standard The vocabulary, basic sentence structure, and to a certain extent the format and structure of conversation in the language of maritime VHF communications were first standardised by IMO (International Maritime Organisation) in 1977 in the form of 'Standard Marine Navigational Vocabulary'' (SMNV), followed by Seaspeak (1988), and recently in 1997 - Standard Marine Communication Phrases - SMCP. These were laid down by IMO as a set of recommended standards for use first in safety of navigation and subsequently in other spheres of maritime activity. This paper aims to show that actual maritime VHF communications differ significantly from the relative standards laid down by IMO. The difference lies mainly in the format and structure of the conversation (transaction), topical development, the linguistic and other means of signalling or marking turns (taking, holding, yielding and ending turns), the making of moves and performing interaction acts. It will be shown that these also depend on the degree of formality and subject-matter (field) of discourse as opposed to standard VHF communications which are highly formal. The conversational structure in maritime VHF communications has been studied extensively by the authors of the Seaspeak Project (Strevens and Johnson 1982, Johnson 1982, Strevens 1983, Strevens and Weeks 1985). Following Sinclair et al. (1972) and Sinclair and Coulthard (1975) they developed a model of basic international maritime English conversation in 7 and subsequently 9 stages (cf. Weeks et al. 1985, Strevens 1988), i.e.: 1. make initial call, 2. respond to call, 3. indicate working channel, 4. agree working channel, 5. switch over, 6. send message, 7. respond to message, 8. end transmission, 9. end procedure.

On Some Features of Conversation in Maritime VHF Communication

189

This is referred to as an 'ideal procedure' by Johnson 1982. However, in real communications this 'ideal procedure' undergoes significant changes (most frequently shortening of the opening stage and merging of the final turn of the message stage (call-offs) with the closing one, closers and farewells). The study of the corpus for this work has indicated that a more general and simpler (e.g. three stage) model, applicable to everyday (particularly telephone) conversation, can also successfully account for the description of maritime VHF conversation. Therefore, the model of conversation analysis adopted here is an adaptation of the model offered in Sinclair and Coulthard 1975, Coulthard 1985 and Stenstrom 1994: 135. It includes the following major stages: I. OPENING, II. MESSAGE, III. TERMINATE. The following example illustrates the simplified model: (2) Setting: Savannah River Fairway-outbound traffic, U.S. Speakers: US pilot on board an outbound vessel, Savannah Information Service (US) Topic(s): 1. Proceeding down the river - departure from the port, 2. Estimated time of arrival I - Initial (or Opening) Stage P: Vessel outbound on Savannah River giving security call. SIS: Say again, please. II - Message Stage P: This is the container vessel KE, outbound on Savannah River. SIS: Container ship, will you be continuing along on the Savannah River area near the coast? Over. P: Roger, we're bound to sea. SIS: What is your ETA for Pilot Station - in the junction with the Savannah River? Over. III - Closing (or Terminate) Stage P: In about 2 to 3 minutes. SIS: OK. Thanks very much for the information.

In stage I, i.e. the opening and make contact procedure, the usual adjacency pair ADDRESS - RESPOND is left out and, since reception is poor, the vessel S resorts to a repair move in order to maintain contact, marked by: Say again, please. However, full contact is only achieved in the course of the second turn started by the speaker P. In stage II (Message) the response {Roger, we 're bound to sea) doesn't seem to follow the line of the main topic but communication is successful thanks to shared knowledge and participants' awareness of the extralinguistic situation. This, as the corpus shows, is a frequent feature of maritime VHF communication. Overlapping of Stage II and III (shown in the graph above) allows an abrupt winding-up move and termination of the exchange. Note also the ellipsis of the subject and verb and shifting of focus to the prepositional/adverbial phrase (absence of the usual back-channel marker YES, which is substituted, however, by heavy nuclear stress, falling tone and pause) in

Boris Pritchard and Damir

190

Kalogjera

the response to the question on the ship's ETA. Unlike most maritime VHF exchanges the greeting sequence is omitted here because the pilot and the Port Information Service speaker know each other well and perform a routine exchange. There is obviously no place for phatic or small talk in the opening stage (two native speakers). There are no special lexical markers or discourse connectors to smooth out shifting from one (sub)-topic to another in the message stage: instead, direct questions apply and, at the same time, a very strong turngiving marker (Over) ensures the turn holder full command of the conversation. Note also the ever present, though, by IMO standards, optional polite closing and farewell sequence. Here is another example with a different, though very frequent topic: (3) Setting: the Mediterranean, off Port Said Speakers: Port Said Radio, Velar (Greek ship) Topic(s): VTS, traffic control - advice on anchoring on arrival I - Initial (or Opening) Stage PS: This is Port Said Radio, ... Centre, Channel 1-6. V: Velar, Velar. Come in. II - Message Stage PS: OK, my friend, will you pass east o f . . . , will you pass east? V: Because I'm now proceeding to er, I will now alter my course about two miles to the waiting area, anchorage. Over. PS: Correct, my friend, correct. V: I will go to the harbour waiting area, over? PS: Correct, harbour area, half mile, half mile from port, drop the anchor, call me again. V: OK. I will drop my anchor at the anchorage area, waiting pilot. Is that correct? Over (rising intonation) PS: Correct, Sir. Correct, Sir. Anyhow, when you are past Number 2 buoy area, please. V: Yes, Sir. How will I proceed from anchorage? III - Closing (or Terminate) Stage PS: I will give you all information after drop of anchor, my friend. No problem. V: OK. I'll do that. Thank you very much.

In the opening stage the second speaker (MV Velar) does not, as is frequently the case, address the calling station (Port Said Radio) and, by ignoring such a requirement imposed by IMO recommended standards on VHF communications, runs the risk of failing to establish contact. This, however, is counteracted by the participants' mastery of the situation (i.e. shared information) and helped by intonation (falling tone on Velar) and a pause before Come in on the one hand and 'tactics' (linguistic information, cf. Coulthard 1985: 129), back-channelling marker (OK) and the almost phatic insertion made by the calling station (my friend) on the other. Lexical repetition in the response acts also reinforces success of communication (correct).

On Some Features of Conversation in Maritime VHF Communication

191

Within the first topic the interrogative sentence uttered by the holder of the first turn (PS) is here a request closing on a command, and therefore interpreted in maritime VHF communication as ADVICE (used as one of the 'message markers' in Seaspeak, though in fact performing the role of interaction act markers). In the response turn, the vessel Velar first makes comments on advice and then shifts to a different topic, without any linguistic signalling (except the gap filler er and a pause) giving a pragmatic signal of the performance of his next action (altering course). The pragmatics of this sequence (sub-topic), however, is only known to the two participants in the communicative event, and the fact that the communication was successful can only be traced from the lexical and sentential repetition (correct, my friend, correct; harbour area; half mile) and a question I will go to the harbour waiting area?). The conclusion (winding-up) move, just as in the example before, overlaps between the last turn in the message stage and the first turn (of only one or two, in the closing stage). In this respect there is a striking difference from a telephone conversation where such overlapping occurs during a single act, the participants speaking at the same time, which is undesirable and is therefore made technically impossible in safety-bound VHF communications. It could be concluded therefore that the stages observed in the corpus mainly differ from the above model in that the opening stage is much simpler, the exchange being usually reduced to a simple greeting (Good morning/ afternoon/ evening). End transmission procedure, or closing, is abrupt if compared with those in everyday telephone conversation, though farewell and polite remarks are always present and can sometimes be pretty elaborate (e.g. Yes, thank you very much, Sir. We '11 be standing on channel 1-2. Good night.; Thank you very much for your cooperation. Bye, bye. Going back to channel 1-6; Have a good watch; I wish you a safe transit. Good trip to ...; A pleasant watch to you, Sir, and pleasant voyage onwards; Good watch to you, Sir, and pleasant journey on the way to ...) The corpus furthermore shows an interesting regularity as the various records of conversations move down the scale of formality. Thus, as the maritime VHF discourse reduces formality, the markers of such exchanges and the moves and acts will become more similar to those encountered in an every-day telephone exchange. In addition to the universal frames (Coulthard 1985: 123) marking the boundaries of transactions such as OK, well, right, now, good, and No problem (a marker so often used by non-native speakers of English), such markers as anyway, anyhow, you know, actually, sorry, etc.), typical of informal phone conversation also occur in the more informal comparable conversations by VHF. This is even more so when the participants are native speakers, particularly in internal waterways (rivers, lakes etc.):

Boris Pritchard and Damir

192

Kalogjera

(4) Setting: New York: Sailing in the river - barge towing Speakers: tug skipper (US), barge skipper (US) Topic(s): Picking up the barge, turning around - safety of navigation A: B: A: B: A: B: A: B: A:

Asper. Good morning, Sir. I'm coming around now with the load. Yes. I've got the Ellen Bush and I'll pick you up at 59 bridge. The Moran is stemming the current in the road, she's waiting to ... I guess we'll work out something. Yes. OK. There is not any tide here at the gate. So, if you wanna come here that '11 be fine. Well, the problem is turning around, you know. But turning around in the river is pretty difficult, if you wanna do it. Yes, roger on that. OK. I'll kick her ... and try to get her on by you. I'll appreciate that, Sir. Thank you very much.

or if the subject-matter of VHF conversation does not immediately refer to safety of navigation, e.g.: (5) DR: Yeah. OK. OK. That's good. OK. Captain (name). DS: Yeah, and also please inform that the ship is in my point of view not fit for anything but harbour operation until, until, minimum requirements are fulfilled. DR: Yeah, OK, Captain (name). You know, may I suggest one thing, that, you know, while you are at the anchorage. If you can just swing the vessel and try to get some deviation card or something, then, you know, I know it's so ins to be really (interruption), but then we can see, vou know, if we can get somebody from the shore . Anyway, you can send me deviation card and I will have a look in order to do something. And, vou know, what is required with regard whether it is the ... DS: Ya, ya. You know also. I know this is a little bit, er, beside er .., are you there? DR: Yeah. I'm there Captain. DS: OK, we can have a look at it but then, er, we will need some time to do it. Look, anyway. I hope you take this carefully with the Owners. DR: Yeah. I will take this. In the meanwhile, you know. I mean there is no hurry about this, Captain. DS: Yes, this is, by the way. You know, that is, I can start working on this. I assure you. DR: Yap. OK. Captain (name). Thanks very much for the call. DS: Same to you. You're most helpful, and thank, all the best, nice talking with you. DR: OK, Captain. Thanks. Bye. DS: Bye, bye.

As a restricted 'language' English for maritime VHF communications is principally aimed to serve the needs of emergency situations, particularly if the participants in communication are non-native English speakers. 'Standard Marine Navigational Vocabulary' (SMNV 1977/1985), SEASPEAK (Weeks et al. 1985) and 'Standard Marine Communication Phrases' (SMCP), are the current IMO

On Some Features of Conversation in Maritime VHF

Communication

193

official labels for an extreme sub-division (variety) of maritime English, deliberately designed and more or less expressly recommended for use in maritime safety communications. One of the most prominent pragmatic features of both the Vocabulary/SeaSpeak and the Phrases is the recommended/obligatory use of message (i.e. communicative act) markers within a single turn signalling the speaker's intention: QUESTION, INSTRUCTION, ADVICE, WARNING, INFORMATION1 These are always implied and therefore normally left out in real maritime exchanges because of the economy of communication, as shown in all the examples above. In the Supplement, a comparison has been made between the standard as contained in SMNV/SMCP and Seaspeak (the dialogue on the left-hand side has been devised here for the purpose of contrast) and 'real' conversation, i.e. the type of dialogue recorded in the corpus (on the right-hand side). This shows that the main weakness of the standard lies neither in the vocabulary (e.g. lexical ambiguity) nor in the grammar (in spite of some obvious grammatical errors) but rather in the fact that persistent use of some of the forms of expressing communicative acts and purpose as recommended by IMO proves unnecessarily repetitive, uneconomical, at times too formal, and therefore inappropriate to the smooth unfolding of maritime discourse. This often makes the standard unnatural and therefore not sufficiently user-friendly. This is one of the reasons why IMO VHF communication standards in English are difficult to learn and to teach.

5. Conclusion Though the recommended IMO standard (i.e. SMNV) has been in use for more than twenty years now, actual communications at sea differ significantly from the recommended standard and there is no reason to believe that the latest standard (SMCP) will change the situation significantly. Thus there is a sharp distinction between the standard and the actual form of communication (by VHF or any other means of radiotelephony). This is particularly true of the elicitation moves (i.e. repeated use of Question, Instruction marker words, etc.) in contact-making, responding or confirming message receipt, particularly in the communication involved in the ship's arrival at a port. Though regularly studied and drilled in the process of education and training for certification, highly formalised IMO signals of making contact, tend to give way to much shorter forms, often leaving out the whole preliminary or opening exchange (usually consisting of two turns). In this way contact is established more directly and thus more naturally, almost assuming the course of a normal telephone exchange. The same holds in the case of

1

Along with the distress, urgency and safety markers MA YDA Y, PAN PAN, and SECURITE these markers are frequently used to signal the whole transaction both in exchanges and in broadcasts.

Boris Pritchard and Damir

194

Kalogjera

adjacency pairs (cf. Coulthard 1985: 70-73) so typical of safety-related maritime VHF communications. It is therefore claimed by way of conclusion in this paper that: a) b)

the dialogue in real maritime communication differs, sometimes significantly, from the IMO recommended standard (SMNV 1977/1985 and SMCP 1997), and apart from the structure layout, (i.e. division into stages), it does not essentially differ from everyday speech dialogue (i.e. conversation), minor differences being attributable to the various factors operating in a particular context of situation

Moreover, a maritime VHF exchange, particularly the one involved in ship-toshore communications, with topics like arrival at a port, exchange in collision avoidance etc., largely resembles a telephone conversation, showing many lexical, syntactic and pragmatic elements typical of such discourse. The corpus has also shown that, unlike Airspeak, the restricted code in marine VHF communications works less successfully than expected and intended. The reasons are (a) linguistic (unnatural structure, especially the syntactic placement of discourse markers in individual speech acts: adjacency groups with questions, requests, advice etc., which are strange both to the non-native and native speaker) and (b) extralinguistic (impact of formulae and language habits on board, lack of training, etc.). Although originally based on field research (recordings within the Seaspeak and earlier projects) the standard does not entirely meet the needs of the modern seafarer, except in case of distress communications (which today tend to be fully automated, e.g. the GMDSS, Global Maritime Distress and Safety System). The recommended language standards of maritime VHF communications, particularly the discourse and pragmatics elements, need to be further investigated and constantly adapted for everyday use. This also urges for constant adaptation and improvement of the linguistic form of the standard to the needs of both acquiring and teaching it. This paper, therefore, is an attempt at casting more light on the nature of real and standardised maritime communications in English, with a view to reducing the gap between the two.

References Akmajian, A., Demers, R. A., Farmer, A. K., Harnish, R. M. (1995): Linguistics, An Introduction to Language and Communication. Cambridge, Mass.: The MIT Press. Bhatia, V. (1993): Genre Analysis. London: Longman. Coulthard, M. (1985): An Introduction to Discourse Analysis. London: Longman. Crystal, D. and Davy (1969): Investigating English Style. London: Longman. Crystal, D. (1989): The Cambridge Encyclopaedia of Language. Cambridge University Press. Crystal, D. (1995): The Cambridge Encyclopaedia of the English Language. Cambridge University Press. Glover, A, Johnson, E, Strevens, P., Weeks, F. (1983): 'Seaspeak: Computer-assisted language Research'. The IVth European Symposium on Languages for Special Purposes, University of Bordeaux II.

On Some Features of Conversation in Maritime VHF

Communication

195

Goffman, E. (1976): Replies and Responses, Language in Society, 5, 3, 254-313. Halliday, M. A. K., Mcintosh, A. and Strevens, P. (1964): The linguistic Sciences and Language Teaching, Longman. Hatch, E. (1992): Discourse and language Education, Cambridge University Press. IMO Standard Marine Navigational Vocabulary (SMNV), IMO, London, 1977, 1985. IMO Standard Marine Communication Phrases (SMCP), IMO, London, 1997. Johnson, E. (1982): The Seaspeak research strategy and the computer, paper delivered at Wolfson College, Cambridge. Riggs, F. W. (1989): Terminology and lexicography: Their Complementarity, International Journal of lexicography, 2, 89-110. Sinclair, J. McH, Forsyth, I. J., Coulthard, R. M. and Ashby, M. C. (1972): The English used by teachers and pupils, Final report to SSRC, mimeo, Univ. of Birmingham. Sinclair, J. McH and Coulthard, R. M. (1975): Towards and Analysis of Discourse, Oxford University Press. Stenstrom, A. (1994): An Introduction to Spoken Interaction. Longman. Strevens, P. (1983): Language engineering, applied linguistics and special-purpose language teaching: A case history in the construction of International Maritime English, paper presented at the Univ. of Illinois. Strevens, P. (1988): International Maritime English (Seaspeak), Fachsprache, Wien, 2-9. Strevens and Johnson (1982): Seaspeak: A project in applied linguistics, language engineering, and eventually ESP for sailors, TESOL Convention Proceedings, Honolulu. Strevens, P., Weeks, F. (1985): The Creation of a Regularised Subset of English for Mandatory Use in Maritime Communications: SEASPEAK, Institute of Culture and Communication, Honolulu. Weeks, F., Glover, A., Johnson, E., Strevens, P. (1985): Seaspeak, Pergamon Press.

Boris Pritchard and Damir

196

Kalogjera

Supplement vì

SMNV/SMCP - SEASPEAK A

B

A

Lloyd's Signal Station. This is Sea Dog, Juliet-Sierra-Alfa-Alfa; Sea Dog, Juliet-Sierra-Alfa-Alfa-Alfa. On VHF channel one-six. Over Sea Dog, Juliet-Sierra-Alfa-Alfa, This is Lloyd's Signal Station, AlfaSix-Zulu-Zulu. Switch to VHF channel zero-six. Over. Lloyd's Signal Station. This is Sea Dog. Agree VHF channel two-six. Over

B A B A

B

B

A

B

Lloyd Signal Station, Lloyd Signal Station. Motor vessel Sea Dog, Sea Dog, on One-Six. Come in, please. Lloyd Signal Station. Lloyd Signal Station. Sea Dog, Lloyd, Sea Dog. One-Four, Sir. One—four.

A

One - Four.

B

OK.

SWITCHING TO A Lloyd's Signal Station. This is Sea Dog. Over. Sea Dog. This is Lloyd's Signal Station. QUESTION: What is your destination? Lloyd's signal Station. This is Sea Dog. ANSWER: My destination is Buenos Aires. Over Sea Dog. This is Lloyd's Signal Station. UNDERSTOOD, Buenos Aires.

Sea Dog. This is Lloyd's Signal Station. INFORMATION: A delay is expected in the approaches, period: four-eight hours. Over. Lloyd's Signal Station. This is Sea Dog. INFORMATION RECEIVED: A delay is expected in the approaches, period: four-eight hours. Thank you. Over Sea Dog. This is Lloyd's Signal Station. Nothing more. Out.

One-Four.

WORKING CHANNEL A Lloyd Signal Station. This is motor vessel Sea Dog, Sea Dog. B Roger, Sir. Good afternoon. Could you spell me your vessel name, please? A My vessel's name is Sea Dog, Sierra, Echo, Alfa Delta, Oscar, Golf.

B A

A B

A

CORPUS

B

Thank you, Sir. And your next port of call, please, Sir? Er. Next port of call will be Hull. Hull. April 14 daybreak. Over Thank you, Sir. And your travelling state. Are you stopping in Gibraltar at all, Sir?

A

Er. We are going straightly through. Over.

B

Roger, Sir. Thank you very much indeed for your co-operation. Pleasant journey to Hull. Lloyd's Service on channel One Six. Good watch. Good bye, Sir. OK. Good watch, Sir. Thank you very much.

A

Almut Koester

Getting Things Done and Getting Along in the Office 1. Introduction In a corpus of naturally-occurring office conversations, it was found that speakers frequently prefaced their conversations with metastatements such as: (1)

(a) (b) (c)

I have a quick question for you. I hope it's a quick question I don't know if you've heard, but if you haven't heard, it's confidential I got a suggestion, by the way with this

What all these metastatements have in common is that the speakers say something about the upcoming discourse, thereby giving the addressee some idea of what they want to talk about. I have a quick question... shows that the speaker is going to make some kind of query or needs some information; I don't know if you've heard... indicates the speaker has some news to tell, and the comment about the news being confidential is strongly suggestive of gossip. Finally, with I got a suggestion... the speaker signals that what he is going to say should be taken as a suggestion or advice. This paper explores how such metastatements (which will later be defined as a type of discourse frame) relate to the notion of speech genres. It seems that speakers use such openers to signal the genres or types of communicative activities they wish to engage in. The notion of genre is therefore seen as being very closely linked to that of speakers' discourse goals. It will be argued that speakers use such frames or metastatements to signal their transactional as well as relational discourse goals; in other words their goals both in getting things done and getting along in the office.

2. Data and Background to the Study The findings reported in this paper are part of an on-going research project into the nature of relatively informal face to face communication in office settings. The study is based on a small corpus of data, gathered by means of audio-recordings, of naturally-occurring conversation in a variety of office environments in the USA and Britain, including university offices, publishers and sales offices. The corpus of transcribed data currently comprises approximately 30,000 words. One of the main aims of this project is to identify some of the recurring speech genres that occur in such office settings across different work environments.

198

Almut Koester

The type of data targeted can be located on a continuum between very formal workplace situations, such as meetings with a set agenda, or job interviews, and informal 'chit chat' which might take place between colleagues, but does not actually deal with workplace tasks. The study focuses mainly on two-party encounters between co-workers which are task-oriented, but are still fairly informal. These encounters occur either ad hoc or, if they are arranged in advance, do not have an official agenda.

3. Genre and Workplace Talk It would go beyond the scope of this paper to fully develop a theory of genre or to discuss how genre might be differentiated from speech events or text types. Most researchers working with the notion of genre agree that genres comprise classes of socially-established, norm-governed communicative activities that share a communicative purpose. Often the presence of certain obligatory features and their relative staging are seen as defining the genre. This is often referred to as the genre's 'generic structure potential' (cf. Hasan 1984 and Ventola 1987 regarding the generic structure potential of service encounters). The emphasis has been on devising a structural formula that can be applied to any individual instance of the genre, and the approach to any given text therefore tends to be 'top-down'. A great deal of research on institutional talk has taken place within the field of Conversation Analysis (cf. Drew and Heritage 1992), which takes a very different approach from that of genre analysis, as briefly described above. Rather than try to describe the generic structures of different types of institutional talk, conversation analysts take a 'bottom-up' approach and examine the local management of discourse in order to find recurring features, e.g. of the turn-taking system, in any given type of institutional discourse. The approach to genre taken here attempts to take on board such insights from Conversation Analysis. Some genres, such as certain types of service encounter, are quite routinized, and what gets said is fairly constrained by the setting (e.g. a shop) and the roles of the participants (customer and server), although even here a great deal of variation is possible, as research into authentic service encounter data has shown (cf. Ventola 1987 and Ylanne-McEwen 1996). 'Office Talk' does not constitute a genre in the way a service encounter does, where the type of encounter is predetermined by the setting and certain expected linguistic routines and patterns can be predicted. Rather office settings allow a variety of discourse types, as they place fewer constraints on what kind of talk is allowable, except that it should be largely relevant to the work at hand (but even this is not always the case, as small talk among colleagues is frequent). There is a range of topics that typically get talked about and actions that typically are accomplished through talk in an office. A preliminary classification

Getting Things Done and Getting Along in the Office

199

of the types of conversation that recurred across different office settings in the corpus yielded the following broad categories: • • • • •

Information provision Directive discourse and advice-giving Problem-solving and decision-making Planning and arrangements Reporting and evaluating

But except for formal situations like business meetings or job interviews (not dealt with in this corpus), what happens precisely is not determined by the setting. In order to determine what genre or activity type any given conversation constitutes, it is necessary to examine the discourse in detail using a 'bottom-up' approach. By looking at the local management of discourse, it is possible to see how speakers negotiate genre in the discourse and also how genre-mixing and genre-embedding occurs. As McCarthy (1998: 30-31) points out, it is also important to remember that speakers are concerned with pursuing certain goals in discourse, not with replicating genres. By looking at the evidence of speakers' goals in the discourse, one can perhaps get some idea of what activity or genre the speakers see themselves as being engaged in, and not run the danger as analysts of imposing a generic category that has no relevance for the speakers themselves.1

4. Discourse Goals When attempting to examine speakers' discourse goals, it is important first of all to recognize that people usually have multiple goals when speaking (cf. Tracy and Coupland 1990). Broadly speaking, two distinct types of goal can be identified: transactional (or task) goals and relational goals (cf. Ylanne-McEwen 1996). The literature on institutional talk has stressed the predominance of task goals in such talk. According to Drew and Heritage (1992: 22), one of the things that distinguishes institutional interaction from ordinary conversation is that 'it involves an orientation by at least one of the participants to some core goal, task or identity... conventionally associated with the institution'. My own data shows that while speakers do primarily orient to transactional goals in accomplishing workplace tasks, relational goals are also important. Most speakers had regular contact with one another, so maintaining a good working relationship was important. Other research by Iacobucci (1990), McCarthy (forthcoming) and Ylanne-McEwen (1996) has emphasized the important of

1

The goal-orientation approach to genre taken here is based on McCarthy 1998 (26-48).

200

Almut

Koester

relational goals in transactional settings as diverse as telephone conversation regarding billing, hairdressing salons, driving lessons and travel agencies. In an examination of data from a hairdressing salon, McCarthy (ibid: 13) points out that only identifying the transactional elements in this type of encounter would leave a large portion of the data unaccounted for. A comparison of the occurrence and distribution of relational talk at the hairdresser's and during a driving lesson shows that there are important differences between the two genres in this respect. McCarthy (1998 and forthcoming) argues that because of the important role played by relational goals, even in transactional discourse, genre analysts cannot afford to ignore them. Paying attention to speakers' relational goals can add valuable insights to description of spoken genres and can help to account for some of the following features in authentic data: 1) the occurrence of relational talk in transactional discourse 2) the presence of non-obligatory generic elements 3) breaks in transactional discourse frames

Finally, focusing on speakers' goals means recognizing that speakers use genre as a resource to pursue their goals (both transactional and relational). However, trying to identify speakers' goals can be problematic for the discourse analyst, as speakers do not, as a rule, make their goals explicit. How then does one identify these goals? The notion of frames or framing, as described by Goffman (1974) and Tannen (1993) is useful here. Frames are participants' sense of what activity is being engaged in, and thus have to do with underlying expectations and not with verbal phenomena as such, but they may be made explicit through a number of linguistic devices. Frames can therefore provide evidence of speakers' goals or what genre they see themselves as 'doing'.

5. Framing the Discourse As mentioned at the beginning of the paper, in my data frames are sometimes made explicit at the beginning of conversations in the form of metastatements about the conversation, e.g.: (2) I got a suggestion, by the way, with this

Such meta-statements set up expectations about the conversation as a whole. In this particular conversation, the speaker signals that what he is going to say is meant as advice or suggestion, and in fact the rest of the conversation consists of a series of suggestions made by the speaker to his co-worker (regarding changes to a particular document). Overall these suggestions are worded tentatively and

Getting Things Done and Getting Along in the Office

201

indirectly, using modal verbs and adverbs (maybe, might) and modal lexical words (know, think), e.g.: (3)

(a) I was thinking this also... (b) An' maybe just a note at the end here... (c) An' you might include...

The lexical choices made by the speaker throughout the conversation thus are consistent with the initial discourse frame of advice or suggestion. Such frames seem to be evidence of speakers' invoking certain genres for whole conversations or stretches of conversation. That these frames can set up expectations for the whole discourse is partly a result of their conversation-initial or sequence-initial placement. The significance of what comes first in a conversation has been noted by Schegloff and Sacks (1973: 300-301) in what they say about 'first topic': 'to make a topic "first topic" may provide for its analyzability (by co-participants) as "the reason for" the conversation'. More recently, McCarthy (1998: 33-34) also noted that speakers signal upcoming generic activity through frames, which he calls 'expectations'. A more detailed examination of another extract from the corpus also illustrates how frames like the one just described can have implications for the ensuing discourse. This encounter deals with a similar situation as the example of advice-giving discourse above: it is a workplace encounter between a superior and subordinate (as was the previous example) and it also involves the discussion of a task. However, the speakers' communicative goals are different, and the conversation is initiated and developed according to a different frame. The encounter takes place in a small American company that sells and publishes advertising. Chris, the president of the company and Mike, his circulation manager, are discussing setting up a method to evaluate Mike's performance. Mike is supposed to draw up a log, in which he enters how long it takes him to respond to requests for advertising material from customers. The speakers are both quite young (in their thirties) and the office is informal. The transcript of the beginning of this encounter can be found in the Appendix - the numbers in brackets following the examples refer to the turn number in the transcript. After a phatic exchange to open the encounter (Twelve o 'clock already - Isn 't it amazing), Mike makes explicit reference to the transactional goal of the encounter: (4)

You want me to make a log (3)

Mike elicits instructions from his boss and thus frames the upcoming interaction as directive discourse. As in the previous example, this frame then has

202

Almut

Koester

implications for the development of the discourse, as Chris does in fact issue directives using deontic (obligation) modality and a 'you' imperative (items shown in bold): (5) you know you don't have to like write down the minute that you- got the request and the minute that you got- it done, an' you just say well that took me about four hours I to /deliver it/. (6)

Later in the conversation, (not shown in extract in the Appendix), Chris uses other deontic modal verbs: need to, gotta.

6. Relational Goals So far we have looked at evidence of speakers' transactional or task goals in formulating metastatements about the discourse. Let us now turn to some examples that show how speakers also orient to relational goals. In the same conversation about performance evaluation, the task frame is broken in turns 12-20 with a 'joking' sequence: (6) 11) 12) 13) 14) 15) 16) 17) 18) 19) 20)







L S o - so a request at five p.m. an'and an' I- complete it at nine the next day that's one hour Right. Hahahahahahahahah .hh N o I think hehehehehheheheh LRight What? [2] What? Yes correct. That's not a- that's not a: uh twenty-hour waiter. Yes heheh yes [mock whiny voice] I ' m sorry but I had to sleep first and eat breakfast LHeheheh [mock whiny v o i c e ] t l d o - 1 do that almost every day! LHeheheheh (Heheheh)

From a genre point of view, one could view this as an interruption and ignore it. On the other hand, one could ask why it occurs and see what role it plays in the discourse at this point. At this point in the conversation (turns 7-11), Mike is trying to establish what he should enter in his log if he receives a request in the evening and does not deal with it until the next morning - he is trying to make sure he does not need to count the time in-between. The shift in the frame is signaled by laughter and also by the mock-whiny tone of voice adopted by both speakers. The frame switch is initiated by Chris (the boss) in turn 12, and Mike is at first disoriented by this switch (in turn 14 he says 'What? What?'), before he joins in the joke. The reason for Chris' laughter at this point seems to be that he sees as ludicrous the implication in what Mike says in turn 11 {So a request at five p. m. an' and an' I- complete it at nine the next day that's one hour) that, if he does

Getting Things Done and Getting Along in the Office

203

not deal with a request that comes in the evening until the next morning, this could be considered 'a 20-hour waiter'' (meaning a 20-hour waiting period or interval). The notion of alignment is useful here: Gavruseva (1995) uses the term 'alignment' to refer to whether the speakers position themselves in an equal (symmetrical) or unequal (asymmetrical) relationship to one another. In this encounter, Mike, by eliciting directives from Chris, is clearly invoking Chris' institutional role as his boss, and his own less powerful role as his subordinate. By implying in turn 11 that not dealing with a request that came in the evening until the following morning might be considered to be a late response, Mike seems to position Chris as an authoritarian, unreasonable boss. Chris' laughter may be due to his not feeling comfortable with such an identity. By switching to a joking frame, he seems to try to reduce the asymmetry and establish a more equal relationship. The fact that he adopts the same mock-whiny voice as Mike is an indication that he is trying to align himself on a more equal footing with his employee. In any event, this joking sequence shows a shift in the speakers' footing away from their task goal. Whatever its precise function, the joking plays some kind of role on the relationship level, possibly in defusing potential threats to face in the discussion of such a delicate matter as this one (evaluating the performance of an employee). Simply to ignore this sequence as not relevant to the genre of directive discourse would mean to neglect a very significant aspect of this encounter. Such relational sequences may in fact be common features of directive discourse, in which potential face-threat is quite high. This would be an interesting point to pursue for further research.

7. Frames and Relational Goals Relational goals may also be apparent in the types of metastatements examined previously. In the following example, the speaker introduced the final segment of a longer conversation with her boss in the following manner: (7) I've got a couple more queries actually, Mary, then I'll leave you to get on.

The purpose of this frame is not just for the speaker to indicate what her task goals are (getting answers to her queries), but also to show that she does not want to impose too much on her boss's time. She uses hedges (a couple, actually) and indicates that it will not take a long time (then I'll leave you to get on), thus minimizing the imposition and reducing potential face threats. She is in fact making extensive use of negative politeness (cf. Brown and Levinson 1978, 1987). The use of hedges (in bold in examples below) was also common in other such framing metastatements from the corpus:

204

Almut Koester (8)

(a) (b) (c)

Uh, just wanted to tell you about my conversation with Tony. I have a quick question for you. I hope it's a quick question. I was wondering if you an' I could possibly this week, at about eleven o'clock on Thursday morning, reinforce each other half an hour on - just to look through [name of journal] and see where we are. (d) Just wanted to come and chat to you a little about the company.

All these examples show speakers' attending to the relational aspect of interaction through the use of negative politeness whilst pursuing task goals. Frames cannot always be taken at face value. Indeed they may not be real accounts of what is actually going on, but speakers may try to project a certain frame on a conversation, as in the following example: (9) Just wanted to come and chat to you a little about the company.

This is how a supplier began a meeting with a customer he was visiting. Clearly, the supplier had not come 'just to chat', but to tell the customer about new services his company was offering, and ultimately to get more business. But by using this frame, the supplier downplays the actual business nature of the encounter and instead stresses the friendly relationship between himself and the customer. There are two ways of interpreting this: that the speaker is simply acknowledging the importance of good relationships in business, or (more cynically) that the speaker is using nominally relation-oriented talk to achieve a transactional goal.2 In either case, such a projection of a relational frame to initiate a transactional encounter clearly shows the importance of relational goals for speakers even when they are doing business.

8. Conclusion In this paper I have been concerned with one recurring feature of office conversation data: the use of metastatements by speakers to initiate segments of talk; and I have examined how these metastatements or discourse frames might be related to the notion of genre. I have argued that such discourse frames provide fairly clear indications of what genres are being engaged in from the speakers' own perspective. I have also suggested that it makes sense to approach genre from the point of view of speakers' goals, as speakers do not simply replicate genres, but use generic resources in order to pursue their discourse goals. In the examples shown,

2

In a study of service encounter data (telephone conversations regarding billing) Iacobucci (1990) concluded that speakers used talk that was ostensibly relational as a strategy to accomplish a task goal.

Getting Things Done and Getting Along in the Office

205

speakers could be seen to orient both to transactional and relational goals in the way they framed the discourse and maintained or broke current discourse frames. As the transactional purpose of discourse tends to be foregrounded in genre studies and in research on institutional talk, the role of relational goals has frequently been neglected. But generic descriptions that do not take into account the relational aspects of discourse may well be neglecting important features of the genre in question. I have only been able to hint here at the kinds of functions relational talk might perform in different genres, but I hope to have demonstrated that relational goals do play an important role in the type of discourse examined here, even when speakers are primarily engaged in task-oriented workplace talk.

References Brown, P. and Levinson, C. (1978, 1987): Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Drew, P. and J. Heritage (eds.) (1992): Talk at Work: Interaction in Institutional Settings. Cambridge: Cambridge University Press. Gavruseva L. (1995): Positioning and Framing: Constructing Interactional Asymmetry in EmployerEmployee Discourse. Discourse Processes 20, 325-345. Goffman, E. (1974): Frame Analysis. New York: Harper and Row. Hasan, R. (1984): The Structure of a Text. In Halliday, M. A. K. and R. Hasan. Language, Context and Text: Aspects of Language in a Social-Semiotic Perspective. Cambridge: Cambridge University Press. Iacobucci, C. (1990): Accounts, Formulations and Goal Attainment Strategies in Service Encounters. In: Tracy K. and Coupland N. (eds.) Multiple Goals in Discourse: 85-99. Clevedon, Avon: Multilingual Matters. McCarthy, M. J. (1998): Spoken Language and Applied Linguistics. Cambridge: Cambridge University Press. Chapter 2: Spoken language and the notion of genre. McCarthy, M. J. (forthcoming): Captive Audiences: Small Talk and Close Contact Service Encounters. In: Coupland J. (ed.) Small Talk. London: Longman. Schegloff, E. and Sacks, H. (1973): Opening Up Closings. Semiotica 8 (4): 289-327. Tannen, D. (1993): Framing in Discourse. Oxford: Oxford University Press. Tracey, K. and Coupland, N. (eds.) (1990): Multiple Goals in Discourse. Clevedon, Avon: Multilingual Matters. Ventola, E. (1987): The Structure of Social Action. A Systematic Approach to the Semiotics of Service Encounters. London: Frances Pinter. Ylanne-McEwen, V. T. (1996): Relational Processes Within a Transactional Setting: An Investigation of Travel Agency Discourse, unpublished PhD Thesis, University of Wales, Cardiff.

206

Almut Koester

Appendix Performance Evaluation President Circulation Manager Mike comes into Chris's office to discuss his KRA's ("Key Responsibility Areas") - a method of evaluating his performance. Mike has drawn up a proposal and is discussing it with Chris, his boss. 1) 2)



3)

[4] 4) [3] 5) [3] 6)

7) 8) 9) 10) 11) 12) 13) 14) 15) 16) 17) 18) 19) 20) 21 ) 22)

Twelve o'clock already [as he's walking in] Isn't it amazing? [Mike sits down] (Okay). [2] You- you want me to make a log. To- to- it'll measure... how well I'm doing /pre-counting. recounts/ an' getting / the list out to ??/ how do you want that to show up over here. As an average... delivery.

number

of hours...

between...

uh:m..

request

and

(Hm) t An' it's an' it's kind of a- 4- you know you don't have to like write down the minute that you- got the request and the minute that you got- it done, an' you just say well that took me about four hours 4 to /deliver it/. [1] /??? four/ So what if the request comes... uh... by e-mail at the end of a day, on Monday, and I don't even see it until... I don't know how were you- how were you gonna keep this before. (Uh that's a good point) Okay. So I'll... so I'll just- ignore time... between... me leaving and- heheh an' me coming in. [3] iOkay. /??/ LikeSo-yeah./???/. L So- so a request at five p.m. an'and an' I- complete it at nine the next day that's one hour Right. Hahahahahahahahah .hh No I think hehehehehheheheh |_Right What? [2] What? Yes correct. That's not a- that's not a: uh twenty-hour waiter. Yes heheh yes [mock whiny voice] I'm sorry but I had to sleep first and eat breakfast LHeheheh [mock whiny voice]Tl do-1 do that almost every day! LHeheheheh (Heheheh) Okay. An' we decided... this stuff was gonna be- pass? fail? Or no:t applicable? [2] The- [2] maintaining of those two things? Right.

Getting Things Done and Getting Along in the Office

207

Transcription Conventions , ? ! [ ] (the t 4( ) / / /?/ L .hh

slightly rising intonation high rising intonation falling intonation animated intonation noticeable pause or break in rhythm within a turn words in these brackets indicate non-linguistic information, e.g. pauses of more than 1 second number of seconds is indicated), speakers' gestures or actions sound abruptly cut off, e.g. false starts A step up in intonation (higher) A shift down in intonation (lower) parentheses around tone units spoken with "sotto voce" colon following vowel indicates elongated vowel sound words between slashes show uncertain transcription indicates inaudible utterances: one ? for each syllable overlapping or simultaneous speech inhalation (intake of breath)

Agnès Witko et Michèle

Grosjean

Les Réunions en Contexte Institutionnel et Leur Formalité 1. Introduction L'analyse des interactions pluri-locuteurs qui font l'objet de cette contribution s'inscrit dans le cadre des recherches sur le 'polylogue'. La définition de 'polylogue' est issue des travaux sur le 'trilogue' élaborés dans KerbratOrecchioni and Plantin (1995). Nous adoptons pour le polylogue la définition (minimale) suivante: 'situation réunissant trois participants en chair et en os'. Les interactions que nous allons analyser sont issues de réunions à finalité essentiellement décisionnelle ou organisationnelle. Par définition, la parole y est plurielle et la mécanique interactionnelle extrêmement complexe: la co-locution, les mécanismes de sélection et de hiérarchisation des différents destinataires, les sélections du prochain locuteur, tous ces phénomènes sont démultipliés dans les polylogues. Nous les aborderons en retenant les éléments suivants: le nombre de participants, le cadre participatif, le type de focalisation. En ce qui concerne le nombre d'interactants, c'est une question qui devient secondaire et qui dépend du niveau de contraintes et des possibilités de développement d'une interaction. Toutefois, le nombre de locuteurs effectifs 1 n'est pas proportionnel au nombre de participants ratifiés et il semble que plus le nombre de participants est petit et plus l'obligation d'engagement des participants dans l'interaction est forte. La focalisation collective est forte mais néanmoins soumise à des phénomènes de scission qui bouleversent totalement les cadres participatifs et qui sont différents selon la formalité de la réunion. Dans la perspective de l'ethnographie de la communication 2 , nous adoptons pour cette communication une démarche comparative à propos de deux types de réunions: des réunions très conventionnelles, très formalisées, dont le cadre est très rigide comme les réunions du conseil municipal qui relèvent, à notre avis, d'une parole 'protocolisée' et des réunions plus flexibles comme les relèves à 1' hôpital qui, pour certaines d'entre elles, se rapprochent plus d'une parole 'conversationnelle'. L'événement de communication étiqueté 'réunion' pose précisément de façon cruciale la question de la formalité pour laquelle nous faisons l'hypothèse

1 2

Par exemple, au conseil municipal, sur une cinquantaine de participants ratifiés dans la situation, une trentaine sont locuteurs potentiels et une douzaine seront locuteurs effectifs. Référence aux travaux de Hymes ( 1962).

Agnès Witko et Michèle

210

Grosjean

suivante: pour décrire l'objet 'réunion', il existerait un axe graduel de la formalité qui correspondrait à un ensemble de critères. Comment rendre compte de la formalité d'une rencontre? Quels sont les critères qui permettent de repérer le degré de formalité d'une réunion? Ceci nous amène à évoquer deux dimensions de la formalité: •

•

la formalité en tant que prescription qui recouvre d'une part un règlement explicite comprenant un ensemble de règles ou de lois et d'autre part un dispositif physique d'action (lieu réservé à un usage, temporalité repérée, acteurs jouant un rôle précis), les deux types de prescription sont complémentaires. Au conseil municipal, nous sommes dans ce cas de figure puisqu'il existe un arrière-plan institutionnel défini par un ensemble de textes de loi ainsi qu' un dispositif de communication prévu pour la réunion, en terme de lieu, de programmation et d'acteurs. la formalité en tant que normes qui correspond à un ensemble de références spatiales, temporelles, actionnelles partagées par les participants dans un groupe donné. Ce type de formalité3 n'est pas prescrit par des textes. Il s'impose du fait du partage des normes. C'est le cas dans 1* environnement de travail qu'est l'hôpital ; on observe un ensemble d'accords préalables sur le fonctionnement de la relève complètement incorporés, comme par exemple le démarrage de la réunion quand tout le monde est là, l'installation des participants dans l'espace etc...

Explicitement prescrite ou 'normes partagées', il y a de la formalité. Le but de cette communication est précisément d'en situer les différents éléments dans les deux types de réunions étudiées. Nous avons tenté le repérage de la formalité à différents niveaux d'une interaction: au niveau situationnel, au niveau de l'organisation structurale, au niveau langagier.

2. Au Niveau Situationnel 2.1. Les Aspects Institutionnels En ce qui concerne le conseil municipal, l'arrière-plan institutionnel est celui de l'Administration Territoriale et des élus des collectivités locales. Cette double appartenance au milieu administratif d'une part, et politique d'autre part installe une double culture omniprésente, y compris dans les règlements puisque l'interaction fonctionne sous la tutelle de deux textes: le Code Général des Collectivités Territoriales4 qui précise les règles de pouvoir et de compétence au sein des collectivités locales et le règlement intérieur voté en début de mandat par le conseil en place qui fixe le fonctionnement de l'événement 'réunion du conseil municipal'. Le cadre réglementaire mentionné est complémentaire: le Code Général précise l'ensemble des droits et des devoirs au sein des collectivités

3 4

Voir Grosjean et Lacoste (1999) qui étudient la formalité dans les relèves. Le CGCT: Loi N°96-l 42 du 21 Février 1996 (partie législative).

Les Réunions en Contexte Institutionnel et Leur Formalité

211

locales; quant au règlement intérieur, il explicite entre autres le protocole officiel l'interaction, le cadre participatif, la place des écrits que sont le rapport, le compte-rendu, le procès verbal5. L'activité de relève à l'hôpital n'est pas prescrite de la même façon. C' est une activité qui assure la continuité des soins aux malades, la coordination des actions du personnel des différentes équipes et la sécurité des soins. Elle est prévue par l'organisation des horaires de travail: il y a, en effet, des chevauchements entre équipes, chevauchements plus ou moins longs le matin et le soir, beaucoup plus longs entre les équipes du matin et de l'après-midi. Elle comporte deux versants, un versant écrit qui figure dans le dossier de soins et un versant oral qui se manifeste dans les réunions de relève. Aucun texte légal ne l'encadre mais le dossier de soins (relève écrite) peut être versé au dossier lors de procès au pénal ou au tribunal administratif. 2.2. L'Espace et le Temps Le conseil municipal correspond à une communication publique. Il réunit 33 conseillers municipaux dans une salle de la mairie ad hoc: il s'agit de la salle du conseil. L'espace est distribué, les conseillers sont assis autour de la grande table en demi-cercle, deux employés de la mairie sont installés près du maire qui occupe la place centrale, la presse est assise sur une petite table derrière le maire alors que le public est positionné à l'extrémité de la salle, en face du maire, sur des chaises pliantes installées en rangées. Un dispositif d'enregistrement est en place ; des micros fixes ou mobiles circulent de part et d'autre de la table et sont à la disposition de tous les conseillers, le maire disposant d'un micro attitré. Un rétroprojecteur est à la disposition des rapporteurs qui souhaitent utiliser des supports visuels comme les plans. Les séances du conseil municipal ont lieu en soirée, à partir de 20h30, sans prescription de durée puisqu'elles peuvent s'étendre de 3/4 d'heure à 2h30 environ et se terminent quand toutes les questions à l'ordre du jour ont été traitées. La relève correspond à une communication interne à un milieu de travail. Elle ne se déroule pas dans un lieu réservé mais sur le lieu principal d'activité des infirmières, à savoir la salle de soins. En fonction du type de relève (si elle rassemble beaucoup de monde ou non) on regroupe ou non des chaises en cercle qu'on va chercher un peu partout dans le service. Les relèves durent de 15 à 45 minutes et s'arrêtent quand tous les malades ont été passés en revue.

5

En ce qui concerne les écrits au conseil municipal, voir Witko (1998), communication au colloque 'Le langage dans les restructurations du travail Paradoxes et enjeux', CNRS, Paris, 24— 26 septembre 1998 intitulée 'Une réunion mise en forme par des écrit, l'exemple du conseil municipal'.

Agnès Witko et Michèle

212

Grosjean

Les données situationelles mettent en évidence une mise en forme de l'espace à la mairie et une relative liberté proxémique à l'hôpital.

3. L'Organisation Structurale de l'Interaction 3.1. Le Script au Conseil Municipal Pour décrire l'organisation séquentielle des interactions, nous nous référons au concept de script défini par Schank et Abelson (1977: 41): 'a script is a structure that describes appropriate sequences of events in a particular context'. L'organisation séquentielle s'appuie au conseil municipal sur le script prescrit dans le règlement intérieur. Le droit d'accès est matérialisé par une convocation stipulant le statut de membre autorisé ainsi que les dates, heures et lieu de la rencontre. Elle s'accompagne du pouvoir à transmettre en cas d'absence. La convocation est un critère fort pour définir le degré de formalité d'une rencontre, elle prescrit à la fois la participation et la non-participation en gérant l'absence. Le conseil se déroule autour d'un ordre du jour écrit que chaque membre a reçu avec la convocation et qui comprend toutes les questions à traiter ; il se peut que l'ordre du jour soit réactualisé en tout début de séance si une question urgente est à rajouter. Les participants arrivent à l'heure prévue dans la salle et prennent place, le maire s'empare du micro. Il a la présidence de l'assemblée avec pour objectifs: le déroulement de l'ordre du jour que tout le monde a sous les yeux et le vote des questions préalablement étudiées en commissions, et exposées en séance par les rapporteurs. La prise de décision en public s'accompagne d'un temps de débat quand les membres de l'assemblée souhaitent problématiser une question. Les séquences s'enchaînent dans l'ordre suivant: -

Séquence d'ouverture: le maire accueille l'assemblée, déclare la séance ouverte et délègue l'appel nominal au secrétaire de séance qui démarre l'appel en même temps que les personnes s'installent. M: mesdames messieurs mes chers collègues 6 je vous propose qu'on commence cette séance du conseil et si vous n'y voyez pas d'inconvénient je vais demander à monsieur S de bien vouloir assurer le secrétariat de séance (.) dès qu'il se sera assis (.) monsieur S je vous encourage même à faire d'appel S: merci (.) X lucien X: présent [...]

6

Les principales conventions de transcription sont notées à la fin de l'article.

Les Réunions en Contexte Institutionnel et Leur Formalité -

Séquence d'approbation du compte-rendu de la séance précédente: le maire initie les thèmes (l'approbation d'un compte-rendu est généralement le premier thème à l'ordre du jour) et appelle au questionnement collectif. M: vous avez reçu le procès verbal et le compte-rendu de la séance du conseil municipal du 17 novembre (.) dernier y-a-t-il des observations i Y: à la page 5 du compte-rendu (.) dans ma réponse [...]

-

Séquence d'information: le maire a un rôle d'informateur M: dans l'article 2122.22 (.) je dois vous informer (.) que [...] et que d'autre part (.) j'ai signé avec [...]

-

Séquences thématiques en fonction des questions inscrites à l'ordre du jour: le maire distribue la parole aux rapporteurs et annonce les thèmes en suivant l'ordre du jour

-

1ère question à l'ordre du jour M: on passe aux décisions modificatives de l'exercice 1997 (.) monsieur J vous avez la parole (..) je crois que tout le monde a eu les papiers hein (.) ça a été distribué J: oui [...]

-

2ème question à l'ordre du jour M: monsieur J vous gardez la parole pour parler des restes à réaliser (.) toujours pour l'exercice bien sûr 97 (pause 6 sec.) ah c'est en distribution (pause 58 sec.) monsieur J vous avez la parole J: oui je rappelle [...]

-

5ème question à l'ordre du jour M: monsieur E (.) vous allez nous parler de la participation financière de 1 a commune aux frais de colonies de vacances des enfants [...] E: à la demande du Trésorier Principal [...]

-

Séquence de clôture: le maire clôt les échanges M: je crois qu'actuellement sont en train de (.) circuler des documents que vous devez signer (.) donc (.) après je pourrai lever la séance (.) en vous précisant (.) qu 'en principe le prochain conseil municipal aura lieu (.) sans doute autour du 20 janvier (.) puisque y'a une date butoire du 20 janvier (.) qui nous est (.) imposée (.) donc il y aura proch/ probablement un conseil municipal le 19 ou le 20 jan/ le 20 janvier (.) j'ai pas regardé les les dates (.) et puis je me permettrai simplement de vous souhaiter une bonne fin d'année et de bonnes fêtes de Noël (.) à bientôt

Agnès Witko et Michèle

214

Grosjean

Pour le conseil municipal, l'ordre est immuable et s'organise autour d'une procédure qui constitue un critère de formalité. Cependant, à l'intérieur de chaque séquence, on repère des sous-séquences: c'est à ce niveau de l'interaction qu'existe la variation qui investit tous les usages du langage. Il y a du jeu, de la provocation, des alliances, des négociations diverses qui se jouent tant sur le plan structural que sur le plan des contenus d'informations. 3.2. Le Script à la Relève La relève est une rencontre ritualisée, obligatoire dans le cadre de 1' organisation du travail de l'hôpital; l'accès est dépendant du statut professionnel et du rôle à jouer dans la transmission des savoirs locaux. Elle s'organise autour de la liste écrite des malades qui sont passés en revue à tour de rôle comme nous le verrons dans le script. La relève formalise le moment de rencontre entre les deux équipes. Le déclenchement dépend de l'arrivée de l'équipe montante et de la disponibilité de l'équipe partante. Les relèves sont généralement précédées de temps morts durant lesquels les uns attendent les autres tout en s'installant. Quant l'infirmière partante estime que le monde requis est là et assis, elle démarre. La relève présente diverses formes en fonction des services. Mais toutes ces formes se réalisent sur la base d'un script minimal que l'on retrouve dans tous les services. Ce script minimal est lié aux objectifs de la relève qui est la passage en revue systématique de tous les malades du service par l'infirmière partante qui réalise ce travail pour l'infirmière arrivante. Ceci repose sur le fait qu'il ne faut pas oublier de malades. Ce qui guide le script est la liste des malades qui est généralement affichée à la vue de tous. L'infirmière partante parle de chaque malade, elle a l'initiative de la parole et c'est elle qui décide de l'enchaînement d'un malade à un autre. L'infirmière arrivante ou d'autres membres du groupe posent à chaque fois des questions afin de faire expliciter des énoncés, d'éclaircir des situations, de demander des précisions. Le script peut être ainsi décrit: Inf partante (IP): malade A Inf arrivante (IA): questions sur malade A ou sur prescriptions et soins Echanges éventuels Inf partante: malade B etc... Exemple: IP: Mr A sa réa sur cathé (.) rien de particulier sur lui IP: MrB (.) alors c'est un entrant de cet après-midi qui est adressé par la consultation en bas (..) c'est une maladie de Crohn (..) on l'a perfusé parce que ça diffusait (..) on lui a prélevé deux hémocultures il faut que tu lui en prélèves une dernière et que tu délivres les antibiotiques (..) il y a tout dans la chambre (..) il est prévenu

Les Réunions en Contexte Institutionnel et Leur Formalité

215

IA: donc y'a plus que l'hemoc à faire et les antibio quoi t IP: voilà dès que tu as prélevé le CBU (.) tu débutes les antibiotiques Mr T (.) c'est l'installation à poursuivre Mme B (.) elle a eu son lavage de vessie (.) le nursing

Sur la base de ce script minimal, on peut voir de nombreuses variations qui tiennent à la fois au contenu verbal, aux thèmes échangés, aux participants et au cadre participatif dont nous reparlerons plus loin et enfin à la nature des échanges: dans un des services, l'infirmière partante fait d'abord un tableau du malade très circonstancié sans être interrompue. Cette structure en forme de récit n'est pas retrouvée de façon systématique dans les deux autres services. Dans tous les cas, le script définit un certain nombre de rôles participatifs qui nous amènent à observer de plus près les cadres participatifs de chaque interaction. 3.3. Le Cadre Participatif du Conseil Municipal Selon Goffman (1981), le cadre participatif se définit ainsi: 'When a word is spoken all those who happen to be in perceptual range of the event will have some sort of participation status'. Il fait référence à la situation globale dans laquelle se trouvent tous les interactants. Quel est le statut participatif de chaque individu qui a accès à un événement communicatif donné? Comment repérer la position de locuteur ou d'auditeur, le statut d'auditeurs adressés ou non-adressés? Quels types de rôles locutoires peut-on dégager? Au conseil municipal, un moment central de l'interaction est réservé à la constitution du cadre participatif: l'appel nominal est une séquence officielle, repérable en début d'interaction dans la séquence d'ouverture. C'est un moment inaugural qui cadre l'interaction: le maire donne la parole au secrétaire de séance pour faire l'inventaire des participants susceptibles de prendre la parole dans la salle, c'est-à-dire tous les conseillers municipaux. Le cadre participatif désigne à la fois: • les participants ratifiés, c'est-à-dire l'ensemble des destinataires qui se répartissent sur deux cercles: le premier cercle correspond aux auditeurs adressés que sont les conseillers et qui répondent 'présent' lors de l'appel ; le deuxième cercle correspond aux auditeurs nonadressés, silencieux, que sont les employés de la mairie, la presse et le public7. • les locuteurs potentiels: Le maire est le locuteur principal. Il joue plusieurs rôles car il occupe plusieurs 'positions' locutoires. Il est animateur et à ce titre, il distribue la parole aux locuteurs sollicités (les rapporteurs qui sont souvent des adjoints) ou aux locuteurs solliciteurs, tous les conseillers qui souhaitent intervenir pour poser une question, demander une précision, apporter un commentaire. Il est président de séance et remplit cette fonction en régulant l'ensemble du protocole et notamment la procédure de vote. Enfin, il est aussi l'homme politique à la tête du pouvoir exécutif de la commune qui informe les électeurs,

7

Rappelons que l'analyste a le statut d"observateur déclaré' dans le public.

Agnès Witko et Michèle

216

Grosjean

prend position et explicite des décisions. Les statuts institutionnels de maire, adjoints, conseillers sont déterminants dans la gestion des rôles locutoires au sein de l'interaction.

3.4. Le Cadre Participatif de la Relève Dans le cadre de la relève, les rôles locutoires sont simples: l'infirmière partante fait le tour des malades successivement. C'est elle qui initie les thèmes et parle plus ou moins longuement sur le thème; l'infirmière arrivante et/ou éventuellement d'autres membres du groupe, réagissent dans certains cas au thème, mais pas toujours. Dans tous les cas, l'infirmière partante enchaîne ensuite sur un nouveau malade. Dans tous les cas, le cadre participatif dominant et minimal est centré sur l'infirmière partante et l'infirmière arrivante. En effet, même quand le cadre participatif est plus large et que tous les récepteurs sont adressés et peuvent intervenir en fonction de leur connaissance des malades, c'est à l'infirmière arrivante que l'infirmière partante s'adresse essentiellement. La comparaison des deux cadres participatifs met en évidence des modalités de prises de parole très spécifiques. Au conseil, les modalités de prise de parole sont fixes, déterminées par le statut institutionnel et explicitées dans le règlement intérieur. C'est toujours le maire qui, systématiquement, annonce les thèmes inscrits à l'ordre du jour et adresse la parole à un rapporteur prévu d'avance. Alors que dans la relève, la prise de parole est liée à la place dans la journée puisque c'est l'infirmière partante qui possède le savoir et le transmet à l'infirmière arrivante. 4. Les Aspects Langagiers 4.1. Au Conseil Municipal Au conseil municipal, le flux de la parole est rythmé par des formules qui reviennent à des moments précis de l'interaction. L'expression consacrée, reprise à chaque début de séance par le maire, pour l'accueil et le déclenchement de la réunion, est souvent: M:

mes chers collègues je vous propose qu'on ouvre cette séance

Après chaque exposé fait par un rapporteur, le maire fait un appel aux questions et donne la parole au locuteur solliciteur, par exemple: M: L:

bien (.) des questions (.) à monsieur A t monsieur L oui en (.) Commission des Finances [...]

Parfois l'appel aux questions et l'appel au vote sont imbriqués quand il n'y a pas de question de la part de l'assemblée:

Les Réunions en Contexte Institutionnel et Leur Formalité

217

M: merci monsieur E des questions au rapporteur t ou des observations t pas de questions pas d'observations t qui s'abstient t qui vote contre t je vous remercie (5 sec.) monsieur J vous (.) reprenez la parole pour nous parler de la convention [...] J: comme vous le savez [...]

Au conseil municipal, l'acte de remerciement ponctue les échanges et fonctionne très nettement comme un marqueur de structuration polyvalent puisqu'il marque la fin d'un échange, la fin d'une séquence thématique, le préalable et le démarrage d'une intervention. Il y a une ponctuation de l'interaction grâce à ce marqueur de structuration, omniprésent. M: merci monsieur J des questions ou des observations t pas de questions pas d'observatidns t qui s'abstient T qui vote contre î je vous remercie (.) monsieur P maintenant une fois (.) qu'il vous sera distribué le rapport (.) sur tableau sur les effectifs (.) nous parlera des modifications du tableau des effectifs suite à (.) ce que j'appellerai un toilettage qui [...] monsieur P dès que les documents ont été distribués vous (.) pourrez prendre la parole (31 sec.) merci monsieur P t P: merci monsieur le maire (.) comme on vous l'a rappelé plusieurs fois ce soir (.) le conseil municipal en sa séance du 17 novembre a décidé de [...]

Quant à la séquence de clôture, le travail rituel est présent dans les formules de voeu ou les promesses de se revoir comme par exemple M: je: vous signale que la séance de ce conseil est levée (.) je vous rappelle que (.) monsieur G m'a dit qu'il y avait 7 papiers à signer par chacun et je vous dis je pense sans doute à peu près à tous (.) au dimanche 25 et au dimanche 1er juin pour (.) tenir (.) les urnes pour les élections (.) bonne fin de soirée

4.2. A la Relève Contrairement au conseil municipal, les formules rituelles sont peu nombreuses à la relève. Elles sont peu conventionalisées. Le déclenchement de la réunion est généralement précédé de questions concernant la présence des uns et des autres, des appels au téléphone pour battre le rappel. Quand l'infirmière arrivante s'est assurée que tout le monde est là, la prise de réunion est généralement précédée, de la part de l'infirmière partante de connecteur du type: IP: bon bon on y va allez allez en deux coups de cuillères à pot bon on fait le petit secteur bon vous me dites si vous ne connaissez pas

218

Agnès Witko et Michèle

Grosjean

Parfois c'est l'infirmière arrivante qui prend l'initiative: IA: bon allez vas-y je t'écoute

A la fin, la conclusion est généralement: IP:

voilà voilàj'ai fini c'est tout

Signalons néanmoins qu'au-delà des aspects strictement formels et ritualisés sur lesquels cette communication est focalisée, de nombreuses séquences conversationnelles peuvent être repérées, et de manière plus importante dans la relève que dans le conseil municipal.

5. Conclusion Notre travail se situe dans une perspective d'ethnographie des communications telles que Hymes (1962) l'a développée et vise une description permettant les comparaisons. Cette contribution est une tentative de comparaison structurale sur la base de laquelle pourraient se construire d'autres comparaisons de réunions. Il ressort de cette étude que les réunions s'étagent sur un continuum de formalité, repérable à différents niveaux de l'interaction: au niveau institutionnel en fonction de ce qui est prescrit ou non par l'institution (protocoles, procédures), au niveau situationnel en fonction de l'écologie du site, de l'organisation temporelle, du programme, des écrits, au niveau de l'organisation structurale en fonction du script et du cadre participatif et au niveau langagier en fonction des formules consacrées. Nous rejoignons à cet égard Goody (1979: 107) quand il dit que i a structure sociale qui est à l'arrière-plan des actes de communication est souvent de première importance'. La comparaison des deux types de rencontres montre la force des dispositifs institutionnels sur les événements de communication. Les réunions sont plus ou moins contraintes par les dispositifs institutionnels.

Bibliographie Goffinan, E. (1981): Forms of Talk, Philadelphie: Univ. of Pennsylvania Press (trad, française Façons de parler, Paris: Minuit 1987). Goody, J. (1979): La raison graphique. La domestication de la pensée sauvage (trad, de 1977), Paris: Minuit. Grosjean, M. and Lacoste, M. (1998): L'oral et l'écrit dans les communications de travail ou les illusions du 'tout écrit', Sociologie du travail 4/98, p. 439-461. - (1999): Communication et intelligence collective. Le travail à l'hôpital, Paris: PUF Le Travail Humain. Grosjean, M. and Traverso, V. (à paraître): Les polylogues - Typologie, problèmes méthodologiques et théoriques, Interaction et Cognition.

Les Réunions en Contexte Institutionnel et Leur Formalité

219

Hymes, D. H. (1962): The Ethnography of Speaking. In: T. Gladwin and W. C. Sturtevant (eds.), Anthropology and Human Behavior, 13-53). Washington, D. C.: The Anthropology Society of Washington. Kerbrat-Orecchioni, C. and Plantin, C. (éds) (1995): Le trilogue, Lyon: PUL. Schank, R. C. and Abelson, R. P. (1977): Scripts, Plans, Goals, and Understanding. An Inquiry into Human Knowledge Structure, Hillsdale (N.J.): Lawrence Erlbaum Associates. Schegloff, E. (1995): Parties and talking together: two ways in which numbers are significant for talk-in-interaction. In: P. T. Haves and G. Psathas (eds.), Situated Order, Washington, D. C.: International Institute for Ethnomethodology and Conversation Analysis and University Press of America, 31—42. Witko-Commeau, A. (1995): Du trilogue dans le polylogue. In: Kerbrat-Orecchioni, C. and Plantin, C. (eds.), Le trilogue, Lyon: PUL, 284-306.

Conventions de Transcription (.)

pause brève ou pause respiratoire entre 20 et 40 centièmes de seconde

(..)

pause moyenne

[...]

coupure du transcripteur

caractères soulignés: passages retenus pour l'analyse M:

maire

t /

intonation montante coupure à l'intérieur d'un mot

Joanna Channell

Working on the Telephone - How Telephone Receptionists Work with Language 1. Introduction The work reported here came about because I was approached by a local council in the UK in pursuit of its plain language and good customer service policies (for more on 'plain language', see Cutts 1995, Channell 1997).The Council1 sought someone to design training for staff who look after the reception desk and the telephones in the main council office. The paper is an account of that work. It therefore involves 'service encounters' of a specific type. It is also an example of language use within a specific public institutional structure. The fact of being initiated by the institution gives the work its particular characteristics, in contrast to being initiated by the investigator (for examples of similar consultancy work by applied linguists, see Shuy 1998 and Channell and St John 1996). The Council was interested in the way that its receptionists talk to the public and wanted me to devise training that would help them to be more effective. There was thus a specific brief which, after negotiation, involved work in two phases: a first phase where I would visit the Council and spend some time collecting data, talking with the receptionists, observing, and listening to their work, and a second phase where I would return several months later and run a training day for the reception team. This account focuses on three aspects of the work: •

Privileged access to data Because the client, the Council, invited me, I gained privileged access to the internal workings of the Council: to committee papers, to documents, to the way in which people would talk about their work, and to recordings and observations of interactions. Constraints and challenges (= institutional practices) The second aspect is the way in which the interactions took place within a particular set of constraints and challenges. There are obvious, observable and documented institutional practices for local authorities within which the reception service talk had to occur.

•

1

The main interest is therefore how the receptionists meet the challenges, and the skills and expertise they demonstrate in working within their specific context.

All names of people or places have been replaced with invented ones. The co-operation and help of council staff and members of the public is gratefully acknowledged.

Joanna

222

Channell

2. Data There are four kinds of data: 1 2

3

4

Text data from council documents Recorded telephone calls a. those observed while listening to both sides of the conversation at the reception desk; and b. others which I made from home, where I acted as a caller and recorded the work of receptionists at different councils (all the calls were 'real' to the extent that I recorded calls which I needed to make, rather than inventing spurious reasons for calling). Observation and recordings from a day spent with the receptionists These include recordings of personal callers coming to the desk. In these cases, I or one of the receptionists explained to people approaching the desk that we were going to record them and explained why, asking their permission to record before the conversation took place. Participant observation of the training session for receptionists.

Material was obtained from three councils: a major city in the Midlands; a rural borough in the Midlands, and a smaller city in the South of England.

3. Approach This was informed by previous work on telephone discourse of which there is a large amount (usefully, here, Psathas 1991 and 1995), by pragmatic analyses and theories (Grice 1967, 1981, Brown and Levinson 1987); and by my previous work with public institutions in the UK (Channell and St John 1996, Campanelli and Channell 1996). The most important orientation was, however, to the design of appropriate practical training.

4. The Institutional Setting: Constraints and Challenges Since the last Conservative Government (1993: 97), all public sector organisations in the UK have been working within the 'citizens charter'. This requires public organisations to develop good practice guidelines and codes which ensure appropriate service. This in turn has led to a specific reflection in the activities of the UK Audit Commission, whose job is to ascertain whether public sector organisations are doing their jobs properly, more specifically, whether they are giving value for money. Currently the terminology is 'best value' rather than 'value for money'. The Audit Commission, to that end, has attempted to develop performance indicators which will measure how well a public sector organisation is doing its job. There is, for example, an indicator which looks at the proportion of public buildings which are accessible to people with disabilities. In the case of services which involve communication, such as answering letters and giving information, developing suitable indicators is more difficult, and it has been found

Working on the Telephone - How Telephone Receptionists Work with Language

223

that some indicators actually have an adverse effect on the service they are intended to monitor. The Audit Commission's work therefore affects the practice and performance of public sector organisations, and in the present context it has a direct effect, because one of their performance indicators is how quickly an organisation answers the telephone. The following text extracts from council documents show how local councils interpret the task of giving 'best value' in their reception services: Example 1 Welcome to Nunthorpe Borough Council. We are here to help you. What can you expect? As often the first point of contact for enquiries about the Council, the Town Hall Reception and Telephone staff are committed to providing a high quality response for users of our service: • • • • •

we will be courteous, helpful and provide an efficient service; we will provide a comprehensive information service about all Council Departments; we will try to answer your call within 20 seconds; we will train our staff to enable them to deal courteously with enquiries.

The team of four receptionists that I worked with answered between them 1000 calls to the switchboard per day. The following extracts from internal documents show something of the way in which their performance is monitored by their managers and ultimately, by one of the Council's committees. The software which runs the switchboard logs how quickly the receptionists answer calls and how quickly they deal with each call. This provides statistical data which their manager can use to assess their performance and report on it to the Council. Example 2: Internal Quality Standards Review 199Standard agreed

Performance on 31.3.9-

90 percent of calls to be answered in 20 seconds

87.94

80 percent of calls to be answered in 10 seconds

72.28

99 percent of calls to be answered

99.43

Example 3: Report to the relevant council committee Following the outcome of a survey, it is recognised that speed of response is not the only criteria [sic] for measuring performance. Answering calls promptly is still very important but ensuring the call is connected to the correct destination has to be the overriding priority.

Joanna Channell

224

The survey showed that at that time there was a council-wide problem with callers being passed around to different people. Although not the only people involved, the reception service needed to improve its success rate. It became clear, therefore, that a key constraint for the receptionists was how to reconcile the equally important, but completely conflicting, goals of: • •

dealing with calls quickly dealing with calls effectively (sending them to the right place).

5. Telephone Calls to Local Authorities What are telephone calls to a local authority like? We begin with two straightforward examples: Example 4 Caller Receptionist C R C R

Ring ring Milford Council, Good morning Do you want my account number? Could you tell me what it's about, please? Council Tax I'll put you through....

The receptionist begins with her standard response. However you can see that although this is a simple example, the caller makes an assumption that they are already speaking to the department they want, so the receptionist needs to ask a clarification question. Although this one is straightforward, it provides an example of what I came to analyse as the most interesting, or perhaps the most difficult, challenge for the receptionists: where the caller's first turn does not make clear where the call needs to go, it is the receptionist's next question, effectively the second question, that is crucial in ensuring the success of the rest of the encounter. So the 'second question', as I began to call it, became a key aspect. I needed to work on question formulations that would lead quickly to the relevant information. Example 5 In this case I am the caller. This is a helpful receptionist who has the information that is required and is able to communicate it successfully: Caller Receptionist C R C R C

Ring ring Milford Borough Council Hello, can I get some information about the concert in the park, I think it's this Sunday September 6? Is that Sunday? yes, that's rthe concert in Valley Park Yes

Working on the Telephone - How Telephone Receptionists Work with Language R C R

225

C

Yes, What is it that you're wanting? Okay, what time does it start? Just one moment I've got a leaflet over here, I think that the support who's the organist, he starts playing at 5.15 that's Peter Brooks and then the concert itself starts at 6.30 and that's with a synchronised firework display. There's no charge it's just £2 for car parking And where is the car park? Erm I presume its actually on the park, just one moment [Sarah is all the car parking actually on the park? yes yes] yes it is actually on the park, All right? Thank you for your help

R

Okay you're welcome

C

bye

C R

Next, some examples of the kinds of interactions that cause more problem, that cause calls to be misrouted and that give receptionists a challenge in dealing quickly and effectively with the caller.

6. 'Troubled Interactions' Example 6 Caller Receptionist C

Officer C

C O C O C

Ring ring Nunthorpe Borough Council, Good Morning Hello, I wonder i f - could I speak to somebody about recycling household waste, I think that the Council has a scheme for [Ringing tone, caller is being put through] Hello, Environment Unit Hello, I'm phoning about recycling. I think that the Council has a scheme of selling cheaply recycling bins, or giving them away or something like that - have I got this right? I'm not sure, what you mean waste disposal bins, for household waste? Yeah for household Right, I'll just put you through to the recycling officer. Can I ask you, where it says about buying, is it subsidised bins, or - ? Well, I - I remember, I think earlier this year in Fielden, on the council car park, there were signs up saying Oh, we had a compost bin sale there Um, oh, okay, maybe that's what I was thinking of. So you don't have a scheme for recycling? [continues]

In this call, a vague-sounding caller (me) wants to get information about having a free recycling bin. I know from a friend that she has one and I want one too. I have not finished saying what I want before the receptionist cuts me off (it feels like an abrupt interruption) and puts the call through. There are two things wrong here. First I was not allowed to finish my turn, and second, I was passed on without being told where the call was going.

Joanna Channell

226

Example 7 [Caller with strong ESL accent, or perhaps African Caribbean] C Ring ring R Nunthorpe Borough Council? C I've left my bean R I'm sorry, what did you say? C I've lost my bin R I'll put you through to the Cleansing Department...

This is one which I listened in to. Neither the receptionist nor I understood the caller's opening remark because of his non-standard accent, and his realisation of the phoneme in the word bin. He is talking about the bin which the Council provides for his household waste, but we both heard it as bean. He then said I've lost... which cued the receptionist to understand that he was talking about his rubbish bin and therefore she was then able to connect the call to the appropriate department. Example 8 C R C R

Could you help me, regarding my rent What's your address 22 Brown Grove I'll put you through to the Rent Officer

This begins with a helpful-sounding caller, but for the receptionist this call is not straightforward because there is an issue about whether the person is a council tenant or whether they are living in private rented accommodation as to who they need to speak to. The receptionist needs to find out what type of accommodation they live in. Some callers react badly to being asked if they are council tenants. These receptionists know the names of most of the streets where there are council tenants, so asking the address is a good strategy, In this case, for example, the receptionist knows that there is no council accommodation in this particular street and therefore the call goes to the Rent Officer. Example 9 C R C

I've just been speaking to the lady who books the accommodation - could I have another word with her? Which lady was it? I don't know her name

This fourth example of a troubled call is someone who rings back to the Council's main number and does not know the name of the person who they want to speak to. It could be in a number of departments, so the receptionist is in difficulty. And

Working on the Telephone - How Telephone Receptionists Work with Language

227

of course, in this case, her question does not help and what subsequently happened here was that a whole story was told, taking several valuable minutes. Example 10 C R

Can I speak to Caroline? We have several Carolines, could I have a second name please?

Examples 6-10 all demonstrate the challenge of the receptionists' work, and the importance of the 'second question' to what follows.

7. The Receptionists' Perspective The next three sections show some data on, and interpretation of, the receptionists' understanding of their work. This comes from individual talks with them and from their participation in the training session. First, they were aware that people (seen from the receptionists' perspective) don't ask for the 'right' thing, and that makes their job more difficult. One of them said: 'The job is about how to ask the right questions. Sometimes people give you no clue, but it mustn't be an inquisition.' 'If they start telling you the whole story, you have to interrupt.' 'You need a kind of second sight.'

They told me about a category of callers who label themselves at the start of the call, for example 'I'm a pensioner', 'I'm a single parent' as a way of helping to cue their listener into what they might need. They also felt that in dealing with elderly people, this was a way of saying 'give me time to say what I need to say'. Third, there was an interesting phrase they used which I found slightly troubling: Get the call off the board

What that means is, 'get rid of the caller as quickly as possible to somewhere else in the Council so that you can take the next call', and this is coming about because of the Audit Commission's performance indicators. As the receptionist is dealing with one call, a flashing light shows that subsequent calls are ringing, and how many are waiting. Thus there is enormous pressure to put a call through to any extension rather than the one they really need, so as to 'get the call off the board', and move on to the next one.

Joanna Channell

228

8. How to be an Effective Receptionist Let me now move on to the training. In the session I wanted to stimulate reflection on what receptionists thought best practice was (and thus to give value to their expertise), but without asking them to analyse the material I had recorded of them doing their work. I therefore used a piece of published data (Grundy 1995: 112114). It is from a recording of a service encounter at a travel agent, where the customer wants to travel from London to Edinburgh using a flight shuttle service. The travel clerk who deals with the enquiry is not wholly effective. I showed this data to the receptionists and said 'what do you think of this: What is going right? And what is going wrong?' With some prompting, the group produced the following list of best practice guidelines for their role: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Listen to the caller Answer their question Be positive (avoid 'not') Give enough information Get the information you need at the beginning 'Technical terms' and 'jargon' need to be explained Avoid 'abrupt' questions Be honest if you can't help If passing on an enquiry, tell the caller exactly what is happening Keep the enquirer in the conversation (when you bring in a third party) Be prepared for anything 'Help' the enquirer to say what they need to say. Open up the conversation. Use key words to help. Allow the enquirer to start from where they are Check that the caller has what they need Pass the caller on as soon as you are clear what they need Tell the caller information they can use in the future Let the caller finish (within reason) Don't assume what the caller might know.

These show a high level of metalinguistic awareness. The opening comment from one participant, looking at Grundy's data, was 'it's all about power, isn't it'. During a training session, it is my normal practice not to allude directly to any concepts from applied linguistics. I want to go from where participants are rather than imposing my understanding. In this case, these participants themselves introduced a meta-level of analysis.

Working on the Telephone - How Telephone Receptionists Work with Language

229

9. 'Second Questions' This section shows the receptionists' proposals for questions which are effective and some which are not effective. They listed these during the training session. Where did you hear about this?

They told me that this is useful because people often have a council leaflet in front of them or they have seen a poster in an office, or the library, and written a telephone number down, and if the receptionists know what the callers have looked at, that can often tell them which department to send the call to. Can you tell me the nature of your enquiry? Which department do you want? Which section do you want? What is your address? Is this a private (or council) property?

They also suggested two repair strategies. Can you spell that word, please?

Particularly interesting as a sensitively thought-out repair strategy is the following. Ask a question which isn't the right one, and use the response to get more information. So, if you have not heard what department a caller is specifying, or cannot understand what they are saying, say, in order to get a repetition or another formulation from the caller: I'm sorry, is it the Council Tax Department you want?

The caller will then reply, perhaps, 'no, it's Cleansing' and then the receptionist has the information she needs to direct the call. The group also understood that some questions are too open, and therefore unhelpful, and they gave some examples of those: Where do you live? What is it about?

10. The Role of Applied Linguistics To what extent were analysis and theory from applied linguistics and conversation analysis applicable and useful in this consultancy and training task? More precisely formulated, what did the client gain from using an applied linguist? The following seem to me to be the main points:

Joanna Channell

230

•

An approach which gives value to the unfolding nature and sequencing of conversations. The idea of a turn as a 'move'.

I do not think one could understand the nature of reception work without these. •

Detailed analysis of grammar, phonology, vocabulary. The possibility to clearly differentiate medium from message is a valuable resource. A pragmatic framework which I could apply to the data and which I could use to help the receptionists to understand and to articulate their understandings of the context of their work. An understanding of the interaction of context and what occurs linguistically.

•

11. Summary and Conclusions For an applied linguist, working as a consultant gives privileged access to different types of data. The receptionists, as an example of people who work professionally with language in a demanding situation, drew on a high level of metalinguistic understanding of the task and demonstrated skill in meeting the challenges of their work. •

The application of theory and practice from applied linguistics contributed in an important way to the success of the consultancy.

References Brown, P. and Levinson, S. C. (1987): Politeness: some universals in language use. Cambridge: Cambridge University Press. Campanelli, P. and Channell, J. (1996): The conceptualisation of qualifications by individuals. London, HMSO. Channell, J. (1997): How to implement plain language in standards of competence. Competency 5/1:26-34. Channell, J. and St John, M. J. (1996): The Language of Standards. Competence and Assessment Issue 31: 2 - 6 1. Cutts, M. (1995): The Plain English Guide. Oxford: Oxford University Press. Grice, H. P. (1967): Logic and Conversation. In: Cole, P. and Morgan, J. L. (ed.) Syntax and Semantics. New York: Academic Press. Grice, H. P. (1981): Presupposition and conversational implicature. In: Cole, P. (ed.) Radical Pragmatics. New York: Academic Press. Grundy, P. (1995): Doing Pragmatics. London, Edward Arnold. Psathas G. (1991): The structure of Direction-giving in Interaction. In: Boden D. and Zimmerman D. H. Talk and Social Structure. Cambridge: Polity Press, 195-216. Psathas G. (1995): Conversation Analysis, Qualitative Research Methods Series number 35. Thousand Oaks, CA: Sage. Shuy R. W. (1998): Bureaucratic language in government and business. Washington, Georgetown University Press.

Julia Gillen

"It's not really time for going home": Three- and Four-Year-Old Children Learning to Talk on the Telephone 1. Introduction: Study of the Development of a 'Specific Dialogic Sub-Competence' Approaches to child language development which involve the study of actual language used in social contexts have tended to display a focus on the attainment of 'communicative competence' (Saville-Troike 1989). Research on early encounters with books, television and other communicative tools add to our understanding of language development overall. Using the telephone, whether in play or actuality is a fascinating activity for children (Hall et al. 1996). It can symbolise mastery of adult abilities including successful manipulation of the dialling process, communicating with familiar, yet distanced others and even combining this with other simultaneous activities (Spero 1993). The telephone is a common cultural object in our society, to which children are exposed from a very early age. It poses particular problems for them, yet very little study has been made of their telephone interaction. Toy telephones are ubiquitous in homes and day-care settings and even real telephones are used by children from a very early age. Bretherton et al. (1981) suggest that even at thirteen months old many children will demonstrate holding a receiver near the neck and shoulder and making a vocalisation approximating to 'hi' (or whatever is the common greeting in their society). Bretherton (1984) believes that this behaviour is imitation of what they have seen adults doing but that it is obviously a further leap from this to understanding that the telephone is a communication tool. Use of the telephone necessitates a break from the early primacy in dialogue of physically present interlocutors talking about the 'here and now' (Brown and Belugi 1964). This channel presents a challenge through the distancing of the interlocutor and thus loss of a necessarily shared environment. Until the work of the present author and the developmental psychologist Cameron and colleagues in Canada (e.g. Cameron and Lee 1997; Cameron and Wang, in press) the processes by which children come to gain competence on the instrument have been very little studied (cf hundreds of studies of adults, chiefly by social psychologists: see Short et al. 1976 for one review). Research into young children learning to talk on the telephone has mostly been concentrated on studies in experimental settings where the adult researchers have set the agenda, generally

232

Julia

Gillen

for adult-child telephone calls (e.g. Holmes, 1981; Bordeaux and Willbrand 1987; Warren and Tate 1992; Cameron and Lee 1997). A strongly emphasised aspect of findings in these studies is children's deficiencies in attaining many of the essential skills of telephone discourse (as identified in the work of the Conversation Analysts; see e.g. Hopper 1992.) For example Bordeaux and Willbrand (1987) found that three- and four-year-old children cannot introduce first topics and Holmes (1981) that even eight-year-old children cannot produce pre-closings. Recognition of both the active nature of meaning-making as well as its social contextualisation leads to an emphasis on trying to comprehend the subject's perspective on the studied situation. In examining the route to competence in telephone discourse skills Mininni (1985) usefully employed the phrase 'specific dialogic sub-competence' to characterise the level of the demands of the channel. Further, from naturalistic observations he suggested that children, as might perhaps be expected, generally seem to engage first in 'play' telephoning, either with a toy telephone, a 'real' telephone not in use, or a substitute. He suggested that in considering young children's acquisition of telephone discourse, the two modes (pretence and actual dialogic) should be investigated together, while of course distinguishing between the two. Further, he proposed that,, for a time, performance in the two modes exists in a "chiastic" relationship; that is, the child will display more sophisticated linguistic behaviour in one-sided pretence conversations than in actual dialogue where an actual interlocutor, who is uncontrollable, acts as a constraint. Exploration of this interesting hypothesis was a major wellspring for this study.

2. Research Design The research was conducted over a nine month period in a nursery class of a state primary school in a town in Lancashire, North West England, UK. The school caters for Roman Catholics in a white, working-class area. Its nursery is part of the main school buildings although existing mainly as a self-contained unit with its own staff. Children enter after their third birthday, attending either mornings or afternoons. The data reported upon here was gathered from the nineteen members of the afternoon group who were in the nursery throughout the two phases of the study discussed here. Considerable time was also spent observing the children's language in the nursery. The curriculum for the children consisted partly of structured activities, both inside the nursery and in other locations, and partly of free choice of a range of activities within the nursery. In these latter play sessions the opportunity to engage with the telephones was made available, in the conditions described below. All telephone activities were thus essentially spontaneous rather than contrived and

Three-andFour-Year-Old

Children Learning to Talk on the Telephone

233

task-laden. In Phase 1 a telephone box obtained from a commercial supplier was installed immediately next to the 'house' play area. This box was child-sized with a roof but no door, but otherwise looked reasonably realistic. An unobtrusive miniature video camera was installed in a recess on the upper part of the back wall so that it captured all the action in the telephone box. The length of play session varied but was usually about half an hour. In this phase, as in each following, data was recorded from ten sessions. In Phase 2 a second telephone, mounted on a little shelf at child chest height, was also installed in the nursery but around the corner of the L-shaped room. The two telephones were connected to one another and speech into both telephones was recorded simultaneously by the video recorder. However, visual images only were recorded from the telephone box, but I took notes identifying the children on the second telephone. In Phase 1 194 call participations, totalling 84 minutes and 16 seconds, were then available for transcription and analysis. The maximum number of call participations made in Phase 1 was 33 (by the youngest child) and the mean was 10.21 (SD 9.34). Apart from the youngest child, Dennis, older children made more calls than younger. Almost all call participations constituted solo calls. However there were three joint calls, in two of which each child spoke once. 307 call participations were analysed in Phase 2. One child did not make any call participations during Phase 2 and five other children made fewer than 10. The most made by one child was 40 and the mean was 16.06 (SD 10.81).

3. Analysis My overall approach to the data was inductive, owing much to 'grounded theory' (Glaser and Strauss, 1967). I created, or at least modified, definitions and other taxonomic tools. I sought to measure how rich the children's talk was, particularly when considering the initial pretence calls as a kind of base and thus exploring Mininni's hypothesis. In the space of this presentation I have to omit discussion of many important issues in the organisation and transcription of data, e.g. the decision to use the term 'call participation' for a single 'call', either actual or pretence, and even reaching a definition of 'turn' in these circumstances. I have to omit many possibly contentious features of my methods in the analysis that follows, such as the focus upon 'content-bearing' discourse with the exclusion, for example, of monosyllables o f ' f i l l e r ' and 'phatic' function. Conscious then of the inadequacies of this brief presentation, I endeavour to sketch out the broad outlines of my approach. I derived categories for types of utterance, working within each phase/call condition separately. I created a category inductively (that is having perceived a particularly characteristic feature) and then examined each call participation in

234

Julia

Gillen

that same phase/call condition to see if it included an example of that category of utterance, continuously working at refining definitions. I did not 'score' for repetition. Thus, when considering for example 'openings' in Phase 1 (recognising the influence on perceiving this as a category from the work of the ethnomethodologists I had studied) I looked at each of the 194 call participations saw if each had an 'opening'. For the purposes of the exercise in data analysis considered here I disregarded any further occurrences of 'openings' within any particular call participation. 141 of the call participations in that phase included (at least) one call opening; this is 73% of the call participations in this phase. Having arrived at a final list of categories and scores, I calculated these as proportions of the total number of call participations in each phase/condition, arranged this information in rank occurrence and present it in Table 1. It is important to point out that where possible categories utilised similar or identical terms across phases to aid comparison; sometimes though this apparent identity of terms masks certain ambiguities or differences, beyond the scope of discussion in this paper. Having arrived at a final list of categories and scores, I calculated these as proportions of the total number of call participations in each phase/condition to facilitate comparisons. I arranged this information in rank order of occurrence. This is presented in Table 1 below. Table 1 : Categorisation of 'remaining turns' within call participations in order of prominence in each phase Phase 1 (n= 194) Opening: Hello/Hiya Closing: bye Topic (introducing/responding) Other's identification Pre-closing Play theme - emergency Play theme - family Sound/rhymes/music/wordplay Play theme - other Self identification - pretence Greetings/how-are-you Self identification - actual Other's identification (given during aside) Play theme - birthdays/parties

No. 141 104 86 71 36 29 26 14 4 4 2 2 2 1

%

73 54 44 37 19 15 13 7 2 2 1 1 1 1

235

Three-and Four-Year-Old Children Learning to Talk on the Telephone

Phase 2 (n = 307) Opening: Hello/Hiya Minimal response to other's topic Introducing topic to other Other's identification - actual Closing: bye Sound/rhymes/music/wordplay Elaborate/continue other's topic Concrete nursery/phone topics Pre-closing Other's identification - pretence Is someone there/who is it Self identification - actual - given in response Play theme - family Greetings/how-are-you Play theme - other Self identification - actual - initiated Play theme - Blackpool/holidays Self identification - pretence - given in response Play theme - emergency Play theme - birthday/parties Self identification - pretence - initiated

No. 173 114 105 84 81 59 56 47 36 34 34 17 17 13 12 9 8 5 4 2 2

% 57 37 34 27 26 19 18 15 12 11 11 6 6 4 4 3 3 2 1 1 1

4. Variations in Linguistic Behaviour Across Phases/Call Conditions A comparison between the pretence calls of Phase 1 and the two-way child-child calls of Phase 2 is instructive. Firstly, it must be remembered that children had roughly equivalent opportunities to make calls, yet far more call participations were made in Phase 2 (307 as against 194). The number of categories is greater in Phase 2, demonstrating that the effect of the transition was not to reduce or block linguistic behaviour. It is worth looking at comparative variations in detail. In Phase 2 'calls' were in no way generally discrete units in the way that they were in Phase 1. The telephone system used had certain qualities that mitigated against the creation of calls as entities. Both phones were always connected (so 'getting through') was not a problem; there was no ring nor dialling tone. Yet nevertheless some basic linguistic elements of"call structure have held up well; in Phase 2 'openings' appear in 57% of call participations (as against 73% in Phase 1); 'closings' in 26% (as against 54%). Pre-closings have fallen to 12% from 19%; given their supposed 'difficulty' this seems a considerable figure, although it can be argued that with the presence of an actual interlocutor their use as a signal

236

Julia

Gillen

towards a negotiated closing is perhaps more likely to be demanded by the situation. In Phase 1 topic introduction (combined with 'responding'; actually of course in a very real sense also an introduction in pretence mode) featured in 44% of call participations. In Phase 2 the introduction of topic appears in just 34% of calls. The effect of picking up the telephone and hearing somebody else already talking of course gives one the opportunity of doing little more than listening. Indeed, 'minimal response to other' was the second most popular category in Phase 2, at 37%. However, a more sophisticated response to an interlocutor, i.e. 'elaborating or continuing other's topic' was present in 18% of calls. So the ability to react contingently to others' speech, is often demonstrated, in nearly a fifth of calls in a fairly elaborate way. 'Pretence' elements still feature in a fairly high proportion of call participations with a greater diversity of themes than in Phase 1. So the effect of Phase 2 has not been to stifle pretence but indeed even to stimulate creative manifestations. But there are also more manifestations of talk about reality. Giving one's own actual identity has increased from 1% of call participations in Phase 1 to 9% in Phase 2 - mostly in response to being asked for that information. 15% of call participations in Phase 2 contain talk either about using the telephone or concrete nursery topics. The analysis given above is limited in its usefulness; I pass into illustrations of some instances of communicative strategies apparently elicited by the use of the telephone.

5. Passages of Talk Involving Callum In this analysis I will discuss extracts from episodes involving Callum. I will start with perhaps the most developed pretence play of the corpus, created by Megan. Her skills and imagination in the world of sociodramatic play were sometimes effective in encouraging others less articulate than herself towards more elaborate play than they would otherwise be likely to create as suggested by analysis of their speech. Callum did not have the tremendous fluency of Megan. Darren, who also features in the following extract, was a boy with language and learning difficulties who received specialist help and did not proceed into mainstream schooling. He did not use words much beyond the occasional monosyllable. Megan's creation and sustenance of the piece of pretence play from which the following extract is taken is a tour-de-force in the terms of this particular study. It is also important as she is shown to be inspired by her partners and in turn to provide a tremendously supportive environment for the boys she involves.

Three-and Four-Year-Old

Call L40309/17 (extract) Time'

Child's speech in phone

237

Children Learning to Talk on the Telephone

Pretence call by Megan Child's speech not in phone I'm going in

Child's actions

Others' speech and actions

Megan climbs in box and takes phone away from Darren.

I'm ringing somebody up. My dad's at work and I'm going to ring him now. Yeh!

points outside box

D: "My daddy!" D looks to see where she is pointing. Callum approaches box, showing his hand, "Ow"

What, what dad? [to Callum]

Oh 0 2 [to C] dials I'm going to ring the doctor up I'm going to ring the doctor up hangs up doctor up dials

24.25

C approaches and shows M and D his hand C goes away D exits

C approaches, showing finger, "It's broken" We're going to ring lifts phone, dials the doctor up aren't we C, "Who, me?" enters box

24.44 Yes 'cos you've got a broke dials puts phone to ear 24.50 1

2

Hello doctor

Part of my investigation into the development of children's telephone discourse skills was concerned with the salience of turntaking abilities. Making use of the work of Veach (1981) and earlier studies by myself and colleagues, I found that young children were less skilled at rapid turntaking than adults but that this skill was relatively unregarded by them. Long pauses were unmarked, whereas a second's pause in telephone talk is marked for an adult, even when talking with a small child. This issue receives detailed attention in Gillen, (1998, unpub.) for of course it has many consequences in the development of telephone discourse. Here I simply observe that only occasional timings are given in these transcriptions to give some very approximate notion of the fluidity of the talk at some points. () indicates untranscribable utterance up to four syllables.

Julia Gillen

238 She's there now [turning to D and C, sotto voce]

puts phone to ear

Hello, come to my house 'cos my dad's got a broken finger 25.01 Yeah 25.03 He's got a broken finger all well and he's all right and he isn't crying now 'cos he's just all right he's not crying aren't you not [to C] 25.16

holds phone away from head

He's got blu tac in his nails um have to get em out and are you the doctor

nuzzles C with head holds phone to ear

in pause tries to touch C and D also try to touch dial. C: "I ring dial it" C moves back suddenly holding his finger, "Ow" takes phone away puts phone to ear

25.32 25.33

He's hurting now 'cos Darren did his finger he went it like -ow! - an' he said ow and he said ow 25.44 he did said ow

bends finger C pummels D gently

looks at C and D I know you have haven't you

takes phone away

got it haven't you haven't you daddy

examines C's finger

25.52

C looks at finger again C: "I got broken finger" to D

C: "Yeh"

puts phone to ear he's got he he's got a broken finger and he's hurt so and he's OK smiles at D

D bangs side of box with fist D bangs side of box with fist

and he's OK and 26.18

C: "Me ring doctor" D vocalises loudly

Three-and Four-Year-Old

Children Learning to Talk on the Telephone

239

puts finger to mouth

26.28

shush I'm on the phone [to C and D] I'm on

C: "I'm ringing up doctor"

Megan's creation and elaboration of a pretence scenario is a remarkable individual achievement but one in which she makes imaginative use of others' potential contributions. She is most unusually responsive to cues created by the others as at the beginning where she turns from an articulate announcement of her call. Darren has probably not totally taken in the import of what she has said and simply responded to the pretence play context and Megan's mention of 'Daddy'. He brings himself into the pretence by pointing to Callum and saying, "My daddy!" Megan makes the immediate shift of responding explicitly to this and then also incorporates the new content introduced by Callum's first utterance and accompanying action. The overall effect reminds me of the shifts required in a certain TV adult charade game for accomplished comic actors. Her play consequent upon this is neither fragmentary nor lacking in intersubjectivity; working often simultaneously in the two modes of 'telephone talk' to the doctor and talk off the phone she creates and sustains an entire scenario. Megan's pretence telephone talk is spoken with consistency as to what is required by her presumed audience on the telephone and with regard to her actually present audience of the two boys. Her subsequent turns at 24.50, of "Hello doctor" and then "She's there now," spoken sotto voce to Darren and Callum, is a fine example of this. What is rare in the corpus is for a pretence call to be sustained in such a way that its narrative development is tight enough for the children present and involved to remain potentially cognisant of the scenario being developed 'with' as it were, the imaginary interlocutor. Usually the pretence call betrays its essential insubstantiality and fluidity. Megan's pretence call is highly cohesive, with the employment of repetition and elaboration of detail, neither departing from the theme. Her choices of utterances to the doctor are essentially appropriate, being related to the injury she is reporting on. It includes such as the use of reported speech (with a very pardonable confusion of the use of auxiliary in "did said." She is impressively adept at differentiating between the personal pronouns as they are used deictically both in the phone to the imaginary doctor and in her asides. Her sustained talk to the doctor functions in holding the pretence scenario together and enabling the boys to develop their roles. She succeeds in this particularly with Callum. His utterances are far simpler than hers and the part he plays less complex. He responds to her original suggestion that they phone the doctor and listens to her but then in his actions and words shows that he wants to

240

Julia

Gillen

take her up on the first person plural element in which her original formulation is framed, i.e. he decides he too wants to call the doctor. His actions are supported by Darren who also wants to touch the phone. I think that Darren's vocal if nonverbal accompaniment to Callum's grammatically poorly formed, "Me ring doctor," is also a support. In the continuation of play which followed the quoted extract Callum seemingly gives up on getting the phone and leaves the box; Megan reacts to this by hanging up in turn. Perhaps regretting risking the loss of dramatis personae, she immediately encourages Darren to make a call: "You ring the doctor's up OK." Darren lifts the phone up, holds it the wrong way up and without saying anything replaces it and leaves the box. Megan then lifts the phone up again and urges Callum, again in the box, to tiy to make a call. Callum is stimulated by his participation in the pretence scenario and Megan's example to manifestly extend his linguistic capabilities. In comparison with Megan's, his utterances look relatively incoherent but a comparison with his utterances in the two extracts reveals them to have become more complex in structure and content with regard to the imaginary elements that are now actively invoked: Call JA0311/18 Time

Pretence call by Callum

Child's speech in phone

Child's speech not in phone

Child's actions

Others' speech and actions

takes phone from Megan holds phone to ear

27.05 I got broken finger and I - it's going () alright

M leaves 27.14

it's broken f i n g e r - i t ' s dials and it got () one on and it got dad one on doctor and (dad / that) [alternative possibilities] go home and me go home now () dad M enters: "Dad!"

27.46 What?

M: () going up to your bed 27.53

hangs up exit

It is likely that in Callum's last telephone turn he has become burdened by the double load of trying to sustain the 'doctor' identity as the imaginary interlocutor

Three-and Four-Year-Old

Children Learning to Talk on the Telephone

241

and himself as 'dad,' once he tries to move away from the simple statement of, "I got a broken finger." In such situations the response seems to be to drop one's own pretence identity or at least to modify it closer to one's own persona to make it easier to cope with. This interpretation cannot be rendered with any (relative) certainty, but it seems possible that in mentioning 'dad' Callum then moves along a train of association to the phrase 'going home' - perhaps either accompanied by dad or to see him later. The qualities of the pretence scenario Megan has developed so well, incorporating early contributions from her 'junior' partners, has inspired Callum to attempt pretence telephone talk that is explicit and otherwise appropriate to the telephone channel: his first telephone utterance is explicit in a way demanded of the telephone (remember that his earlier face-to-face presentation of a hurt finger was to hold it out and say, 'Ow'); his final telephone clause, ".. and me go home now () dad," also seems to reflect a sense of distance between self and interlocutor. Callum is the telephone interlocutor found by Karl in an example of the kind of borderline interaction between the worlds of pretence calls and actual dialogue that often occurred during Phase 2. Karl here recognises that there is one of his companions on the other end, but nevertheless proceeds to carry out his intention to verbalise his wish to end the nursery session and head for home. It can be elucidated that his statement, "I'm tidying up," which is not and cannot be true at the moment it is uttered, is expression of his wish to hurry things up, because in the routine of the nursery everybody knew that 'tidy-up time' heralds the remaining ritualised activities prior to going home. Similarly the final act of running to the toilets also prefigures this period. IIA 0511

Karl and Callum

Time

speech

phone actions

speech

18.16 18.18 18.19 18.22 18.23

Karl Hello

picks up phone 1 Callum Hello

Mum Yeah I want to go home 'cos i t - w h e n it's time to go home

18.30 18.31 18.38 18.41 18.43

What time is it? It's eight o'clock [coughs] Good bye What are you doing? I'm tidying up

phone actions picks up phone 2

242

Julia Gillen Are you? What time is it?

18.50

Eight o'clock What time is it now? Past o'clock See ya

18.56 Bye bye

Karl!

19.00 Bye

Karl! Yeah?

Bye Bye

takes phone away from head [earpiece is further from head than mouthpiece]

Bye Bye 19.14 19.16

You know when you start going home then you get your shoes on then you go () outside Bye Bye bye hangs up

See ya tomorrow

hangs up C: "Karl." K looks round corner. K: "What?" C: "You go on t'telephone again."

19.22 19.25

19.32

Hello

lifts phone 1

Callum It's not really time for going home

Nearly Bye

hangs up

What? Karl Karl

lifts phone 2

K runs to toilets hangs up

19.36

C runs to toilets

Of course, it is impossible to assert with certainty that any particular feature of the dialogue could not have happened face-to-face. Nevertheless I believe that actually virtually all of it is linked with the use of the telephone and would not otherwise have occurred. The initial casting of roles is not impossible but in face-to-face contact would have been the prelude to actions, very likely accompanied by words. But on the telephone the only actions that can be achieved within the call are those that are accomplished in language. Therefore there is no joint activity

Three-and Four-Year-Old

Children Learning to Talk on the Telephone

243

entailed by the agreement of the roles; rather Karl is free to concentrate on the expression of his wish that the role invocation was made for. Callum seems to recognise that it is more important to engage in some way with the content of what Karl is saying than to focus upon taking on the role of 'Mum.' (It is also possible he has noticed that many children invoke 'Mum' on the phone although Callum himself has not done so prior to these call participations.) The dialogue about time is interesting; of course the children are too young to learn how to tell the time, but they show that they are starting to learn something about the phenomenon from encountering many examples of talk about time conventions in society. Karl comes up with a well-formed time; Callum shows he knows that the time changes constantly (his insertion of 'now' in his request, "What time is it now?" indicates that something more than repetition is going on) and Karl makes a good effort at an appropriately changed response in his "past o'clock." Callum is stimulated by the conversation into more elaborate speech than his norm in his turn beginning, "You know when you start..." The topic is abstract and the syntax appropriately intricate. The boys interact in truly dialogic fashion at many points, utilising adjacency pairs and question and response structures for example. Knowledge of these conversation management strategies has facilitated their joint exploration of the linked topics of their talk.

6. Conclusions Practice with routine beginnings and endings is especially salient as a first element of emerging pretence calls. Formulaic, ritualised routines certainly appear to be particularly associated with the beginnings and endings of calls, with more ostensibly creative utterances associated with the middle part of calls. Early pretence calls show a very strong tendency to be composed of these routines, expanded initially with such phatic monosyllables that serve as signs of apparent turntaking participation. This early preference for emphasising the most salient features of the anatomy of telephone discourse i.e. the most important elements of structure and organisation, is reminiscent of Bakhtin's (1986:78-79) account of how we organise our linguistic resources through mastery of specific speech genres: To learn to speak means to learn to construct utterances (because we speak in utterances and not in individual sentences, and, of course, not in individual words). Speech genres organise our speech in almost the same way as grammatical (syntactical) forms do... If speech genres did not exist and we had not mastered them, if we had to originate them during the speech process and construct each utterance at will for the first time, speech communication would be almost impossible.

244

Julia Gillen

Bakhtin's quotation here is apposite in considering not just how children's early (pretence) calls can be characterised as illustrating the most essential features of the genre, but also the notion that probably telephone discourse is learnt initially partly from a combination of witnessing one side of adults' demonstrations of the genre, with the presumably gradually infused knowledge that in fact the talk is two-way. Mininni's proposition of a close relationship between pretence and dialogic telephone talk receives some support with this study, although further ethnographic investigation is called for. The element of 'pretence' in the development of actual telephone discourse is twofold: firstly in the sense that these pretence calls with toy telephones seem likely to feature largely in early telephone discourse (whether solo or embedded in socio-dramatic play); and secondly that the creation of an imaginary interlocutor is a significant (partial) motivation for many children in these calls and a consequent source for development of the discourse (content, register etc). However, Mininni's notion of a chiastic relationship between pretence and dialogic telephone talk, i.e. a shortterm decrease in richness of linguistic behaviour when the child is introduced to telephone dialogue, is refuted. On the contrary, the growth in utterance categories, including diversity of topic themes, indicated greater complexity. Children's knowledge of telephone discourse is revealed to be far more substantial when studied in a setting of spontaneous play than in certain previous experimental studies with a higher degree of artificiality and adult control. For example the finding of Bordeaux and Willbrand that only at age five is the ability to introduce topics beginning to emerge is contradicted many times in the performance of these three- and four-year-olds. The design of this study enabled the children to demonstrate richness of language within a genre, or 'specific dialogic sub-competence' of everyday conversation that is no less creative for children than it is for adults.

References Bakhtin, M. M. (1986): The Problem of Speech Genres, In: Speech Genres and other late essays. (trans. V. McGee) Austin, Texas: University of Texas Press. Bordeaux, M. A. and Willbrand, M. L. (1987): Pragmatic Development in Children's Telephone Discourse, Discourse Processes 10 (3) 253-266. Bretherton, I. (1984): Representing the social world in symbolic play: reality and fantasy, In: Bretherton, I. (ed.) Symbolic Play - the development of social understanding, London, Academic Press. Bretherton, I., Bates, E., McNew, S., Shore, C., Williamson, C., and Beeghly-Smith, M. (1981): Comprehension and production of symbols in infancy, Developmental Psychology, 17, 728736. Brown, R. and Belugi, U. (1964): Three Processes in the Child's Acquisition of Syntax, Harvard Educational Review (34) 133-151.

Three-and Four-Year-Old

Children Learning to Talk on the Telephone

245

Cameron, C. A. (1998): Telephone Communication Enhances Recontextualized Language, paper presented in symposium The Emergence of Academic Skills: on the Importance of Home Experiences at the Annual Meeting of the Canadian Psychological Association. Edmonton, AB (June). Cameron, C. A., and Wang, M. (1999): Frog, where are you? Children's narrative expression over the telephone. Discourse Processes. Gillen, J. K. (1998): An Investigation into Young Children's Telephone Discourse, unpublished PhD thesis, The Manchester Metropolitan University. Glaser, B. and Strauss, A. (1967): The Discovery of Grounded Theory: strategies for qualitative research, Chicago: Aldine. Hall, N., Gillen, J., and Greenall, R. (1996): Don't cry, I ring the cop shop: young children's pretend telephone behaviours, In: Hall, N., and Martello, J., (eds.) Listening to children think: exploring talk in the early years, London: Hodder and Stoughton. Holmes, J. (1981): Hello-goodbye: An Analysis of Children's Telephone Conversations, Semiotica 37(1-2)91-107. Hopper, R. (1992): Telephone Conversation. Bloomington, Indianapolis: Indiana University Press. Mininni, G. (1985): The Ontogenesis of Telephone Interaction, Rassegna Italiana di Linguistica Applicata 1 7 ( 2 - 3 ) 187-197. Saville-Troike, M. (1989): The Ethnography of Communication 2nd edn. Oxford: Blackwell. Schegloff, E. and Sacks, H. (1973): Opening Up Closings Semiotica 7 289-327. Short, J., Williams, E. and Christie, B. (1976): The Social Psychology of Telecommunications Chichester: Wiley. Spero, M. (1993): Use of the telephone in play therapy, In: Schaefer, Charles E., and Cangelosi, D.M. (eds.) Play therapy techniques, Northvale, NJ, Jason Aronson, 101-108. Veach, S. R. (1981): Children's Telephone Conversations PhD thesis, Stanford University. Warren, A. and Tate, C. (1992): Egocentrism in Children's Telephone Conversations, In: R. Diaz and L. Berk (eds.) Private Speech: from social interaction to self-regulation, 245-264. Hillsdale, N.J: Lawrence Erlbaum Assoc.

Christian Hudelot et Christine Berger Les Voix du Dialogue: Usage de la 'Communication Facilitée' avec des Enfants Présentant des Troubles Autistiques 1. Introduction Depuis 1970, la découverte de la Communication Facilitée (technique pour communiquer avec les personnes atteintes de troubles du langage) soulève une vive polémique. Cette technique, mise au point par l'australienne Rosemary Crossley, implique "l'assitance physique et émotionnelle" d'un "facilitateur", qui soutient le doigt du "facilité", lequel écrit des mots en les tapant lettre par lettre sur le clavier d'un ordinateur (Torjman and Botbol 1996: 44). La polémique, devenue mondiale, concerne notamment l'usage de la CF auprès de jeunes enfants dits autistes. Un rapport de l'Agence Nationale pour le Développement de l'Evaluation Médicale (1994) examine la CF dans le chapitre traitant de l'évaluation des modalités actuelles de prise en charge des enfants autistes et psychotiques. Ce rapport conclut que les réponses obtenues à l'aide de cette technique sont en réalité induites, volontairement ou non, par le facilitateur (ANDEM 1994: 48). La volumineuse littérature polémique a été abordée dans un article de synthèse (Howlin 1997) qui interroge l'enjeu de l'habilitation ou non de cette méthode. Le recours à la CF est souvent employé au détriment d'autres activités éducatives, et son usage systématique n'incite pas les proches ou les éducateurs à être attentifs aux conduites significatives idiosyncrasiques propres à ces enfants. Nous avons examiné à partir de séances de CF enregistrées sur cassette vidéo, le produit (les messages tapés) et le processus de la CF, qui implique des comportements dialogiques complexes entre le facilitateur et l'enfant, voire même avec les personnes présentes (parents, éducateurs).

2. Methode et Problématique Nous avons retranscrit les cassettes vidéo des séances d'apprentissage et des séances poursuivies chez un facilitateur par trois enfants (autistes mutiques) sur deux ans, et cela dans deux axes de réflexion. D'une part, un travail sur les "dialogues" censés se dérouler entre les messages tapés sur clavier par le dispositif des "mains collées" (celle de l'enfant tenue par celle du facilitateur) et les interventions orales du facilitateur. Nous avons fait l'analyse sémanticosyntaxique des interventions attribuées au "facilité", et l'analyse interactive du dialogue facilitateur/facilité. D'autre part, en partant de la conception multicanale

Christian Hudelot et Christine Berger

248

et pluricodale de la communication (initiée par l'école de Palo Alto), nous avons analysé les interactions non-verbales (gestuelle, mimiques, regards, postures corporelles, rires et vocalisation) afin d'observer leur évolution sur une histoire interactive de longue durée. Pour cela, nous nous sommes appuyés sur le travail de Condon (1984) démontrant l'unité qui existe constamment entre la parole et la motricité corporelle normale et le synchronisme intriqué entre discours et mouvement chez un locuteur ordinaire (Condon 1984: 48) ainsi que la synchronie corporelle interactionnelle qui existe entre deux acteurs d'un dialogue. Y-a-t-il en CF un apport sémantique de l'énoncé gestuel à l'énoncé "verbal" (les messages tapés)? Voit-on émerger de vrais dialogues? 3.

Contexte

La polémique montre que les arguments des deux parties reposent sur une conception limitée des usages du langage, réduit à une activité de transmission univoque d'information sur soi et sur le monde. Ceux qui doutent de la paternité du message insistent sur l'influence du facilitateur (qui serait le seul à avoir connaissance du message à transmettre); ceux qui défendent l'idée que le facilité est bien l'auteur insistent sur la spécificité de ces messages (fautes d'orthographe, termes inattendus, idiosyncrasie, poétique particulière, ou révélation de faits ignorés du facilitateur). Le point de vue d'une co-élaboration de la signification reste absent de ces débats. Une analyse tenant compte des approches interactionnelles devrait pouvoir apporter de nouveaux éléments au débat. 3.1 Le Dialogue entre les "Messages" Tapés et les Interventions Orales du Facilitateur A. Une oralisation de l'écrit Le couple facilitateur-facilité tape lettre par lettre les mots: or, il s'avère qu'il n'a qu'une seule voix, celle du facilitateur qui déchiffre oralement ce qui s'inscrit (production matérialisée sur un ruban de papier ou sur un écran). Par moments, le facilité interrompt, accompagne ou suit ce déchiffrage par quelques vocalisations, cris, ou reprises en écho. Le facilitateur, en déchiffrant, commente ou pose des questions à l'enfant ou aux autres participants (parents, éducateurs). Le message lu ne tient pas compte des fautes de frappe, répétition de lettres, mots inachevés, ni des cris ou échos de la part de l'enfant. Ce qui se donne comme message signifiant est davantage cette verbalisation que le texte frappé. De même, certains partisans de la CF affirment que l'on doit davantage faire confiance à ce que les facilités "écrivent" (donc, à ce qui est lu) qu'à ce qu'ils disent (dans le cas d'autistes non mutiques). Soulignons enfin que cette lecture n'est pas neutre: le facilitateur module ses intonations en fonction de son interprétation, ou met en

Usage de la 'Communication Facilitée ' avec des Enfants Présentant des Troubles Autistiques

249

relief certains mots ou tournures censés être typiques de la production infantile comme: Il a mouru moi. L'activité commentative du facilitateur prend souvent l'aspect d'une reformulation: E (Enfant): Joie d'être l'arbre. F (Facilitateur): Ah c 'est toi qui es un arbre. B. Linguistique de la phrase Sur un plan syntaxico-sémantique, les messages sont le plus souvent des énoncés assertifs ou injonctifs, les énoncés interrogatifs étant réservés au facilitateur. Soulignons que la frappe des messages peut apparaître à la suite d'une question posée par le facilitateur, mais pas toujours. Ce ne sont donc pas toujours des "réponses". Les phrases sont simples, peu sont complétives. Nous trouvons peu de relations chronologiques dans les corpora étudiés. Les verbes sont plutôt de type verbes d'état, entraînant donc des prédications équatives: Pour lui la mort est vie. Ces verbes sont souvent au présent, et parfois au passé. Souvent, les phrases apparaissent comme des énoncés de vérités générales, non actualisées dans le ici et maintenant de la communication. S'ajoute à cela l'emploi inégal du "Je", et la fréquence des sujets inanimés: Jour de mort vous tue. Ce qui donne parfois à ces textes un caractère poétique. Sur le plan des usages singuliers du langage, on note des constructions-type comme: fou l'homme, ou tu silence que. Sur le plan lexical, on remarque la présence d'éléments qui signifient le partage d'un monde de symboles avec le facilitateur et entre plusieurs facilités travaillant avec un même facilitateur: dans nos corpora, c'est l'exemple de l'emploi de cave, de bout de vie, de muter. On est frappé par la fréquence de termes abstraits, de concepts comme vie, mort, liberté, amour, joie. Cet ensemble de traits offre l'image de textes opaques: ils ne dessinent pas un univers référentiel auquel on puisse, par exemple, appliquer des décisions vériconditionnelles. Ces énoncés, qui ne sont dès lors ni vrais ni faux, relèvent souvent d'autres genres que le genre descriptivo-narratif. On voit poindre alors les limites d'une analyse strictement linguistique. Seule une prise en compte du contexte discursif des énoncés nous permet de "désopacifier" notamment les articles définis. C. Linguistique de

l'inondation

Les conditions de production des messages rendent difficile l'emploi du terme d'énonciation en ce qui concerne l'enfant: celui-ci ne parle pas, ou ses productions verbales sont écholaliques, ou consistent en de petits bouts de phrases {Pourquoi moi, Je n 'veux pas) ou en l'énonciation de son prénom, mais dénués de tout contexte. Est donc considéré comme "énoncé" le message tapé. La polémique s'appuie sur cet "énoncé" écrit: soit il serait produit par le facilité (d'après ceux qui défendent l'idée que la CF est une aide à l'expression verbale écrite via la

250

Christian Hudelot et Christine Berger

machine, pour des enfants qui ne peuvent parler); soit il serait produit par le facilitateur (d'après ceux qui ne voient pas pourquoi des enfants incapables de parler y parviendraient grâce à l'utilisation d'une machine). Certains observateurs des processus de la CF désignent le dispositif des "mains collées" comme "chimère": cet être à deux têtes et à deux mains serait-il l'auteur des messages? Nous nous interrogeons: Qui parle, quand "je" profère un énoncé, en CF? On peut distinguer, dans un énoncé ordinaire, celui qui profère physiquement l'énoncé (le locuteur) et celui qui en est responsable (l'énonciateur): c'est le cas des discours rapportés (Pierre m'a dit: "Je viendrai demain") (cf Ducrot, 1984). Cette polyphonie, comprise par les acteurs d'un dialogue ordinaire, est plus opaque en CF. Dans certains cas, le message tapé présente une similitude avec l'énonciation ordinaire: E: Moi vois tu (le moi est censé être celui de l'enfant, et le tu celui du facilitateur). Cependant, seul le facilitateur participe de cette polyphonie, en se faisant le "porte-parole" du facilité. Parfois, le facilitateur dénie la "chimère" par un commentaire: E: Jure de tuer moi. F: Ah bah ça Q je peux pas te jurer ça. Or ici, l'interprétation du message par le facilitateur joue sur l'absence de pronom personnel devant le verbe: le facilitateur choisit d'interpréter jure comme un impératif. En d'autres lieux, il choisit autrement: E: Va faire rire tout le monde. F: Ah comment tu vas faire pour faire rire tout le monde?. Le facilitateur lit ce qui s'écrit, mais souvent l'enfant ne regarde pas le clavier: il entend alors ce qui est verbalisé. A qui est destinée la lecture? Si certains témoins de la lecture sont proches (l'enfant, les parents, les éducateurs présents à la séance), d'autres les sont moins (notamment les chercheurs à qui sont destinés les enregistrements). Cette remarque est importante, car les séances ont un double but: offrir à l'enfant un espace d'expression, et démontrer dans l'usage de la technique sa propre validité. Les destinataires sont parfois désignés d'une façon collective (F (commentant le "message" qui vient d'être tapé: c'est très rassurant ce que tu nous dit là). Dans ce cas, l'interaction déborde du cadre facilité/facilitateur, pour prendre une dimension élargie envers des témoins potentiels. D. Linguistique de la variété des voix ou place discursive des locuteurs Les réflexions sur la notion de "place" (Bakhtine 1979; Flahaut 1978; Vion 1995) nous permettent d'étudier ce système des voix produit lors des séances. Le facilitateur, qui occuppe une place institutionnelle, joue sur les places en jouant d'une part sur la voix enfantine (rendue audible par la lecture du message) et sur la sienne propre (lorsqu'il parle en son nom). Ainsi s'élabore un jeu sur les places "énonciatives" qui, comme le note R. Vion (1995: 186) concernent la manière dont le locuteur construit des énonciateurs dans son propre discours et procède ainsi à une mise en scène de lui comme des "autres" (...) L'énonciateur est par là

Usage de la 'Communication Facilitée ' avec des Enfants Présentant des Troubles Autistiques

251

un être intra-discursif', une sorte d'image à l'intérieur du discours, que le locuteur construit de lui-même et de ceux qu 'il fait parler. Le facilitateur semble parfois prendre la place de l'enfant: quelle est la marge entre "le faire parler" et "parler à sa place"? On note, en effet, une adéquation étrange: les messages "tapés" ne réfutent jamais l'interprétation. Les seules conduites de refus (du contenu du message, ou de l'exercice?) que l'enfant manifeste, sont non-verbales, ou consistent en la profération de non, non, qui ne sont pas suivis d' "explications". En outre, le facilitateur se positionne d'une façon subjective, en faisant intervenir son propre ethos dans ses commentaires: C 'est drôlement intéressant, ce que tu nous dit!. L'ambiguité de la CF en matière de places a provoqué la polémique: faire parler l'enfant qui ne peut pas parler, peut être une façon de parler à sa place, l'utiliser comme porte-voix. Du côté du facilité, les énoncés semblent parfois proférés du point de vue d'une sorte d'autiste générique, dont le discours révélerait l'intelligence des autistes, ou la douleur d'être autiste: La vie d'autiste bout de vivre et personne ne le sait (Vexiau 1996: 61), ou sermonnerait les parents d'autistes par l'intermédiaire de l'un d'entre eux: Je veux parler le plus possible pour punir papa et maman d'avoir parlé pour moi papa parle de moi comme un bébé (Vexiau 1996: 62). Ce type d'énonciation offre parfois des tournures prophétiques adressées à la collectivité des autistes, comme les écrits de Birger Sellin: Maintenant seulement je compose un chant sur la joie de pouvoir parler un chant pour les autistes muets à être chanté dans les asiles et les maisons de fous (1994: 7) qui prirent leur source dans l'exercice de la CF. La superposition des voix dans cette interaction complexe, rend difficile la comparaison avec un dialogue ordinaire. Le facilitateur à la fois aide à dire, fait dire, dit à la place, interprète, évalue, commente. L'être qui parle par la bouche du facilitateur est-il cet enfant, l'enfant autiste en général, les handicapés, un modèle d'enfant pour sa mère, etc? Les voix sont-elles celles d'un adulte et d'un enfant autiste en dialogue? Peut-on parler de l'émergence d'une voix mitoyenne, celle de la "chimère", sujet hybride créé par le cadre situationnel de la CF? L'analyse linguistique des textes produits, ainsi que des conditions discursives de leur production, montrent qu'il est urgent de ne pas aborder la CF avec une vision lacunaire de ce que déterminent les processus langagiers, et de la complexité qu'implique le concept même de dialogue. 3.2

Observation des Conduites Non-Verbales au cours de l'Interaction

A partir de l'enregistrement vidéo horlogé de douze séances de CF, prises chacune à six mois d'intervalle pour chaque enfant, nous avons analysé des séquences, sur le modèle des micro-analyses de comportement d'enfants autistes (Etourneau 1985; Adrien and Cie 1988). Dans ce domaine d'observation, L'interprétation des

252

Christian Hudelot et Christine Berger

conduites des enfants et de leur organisation spatio-temporelle dans cette situation d'interaction, doit tenir compte des mécanismes psychiques sous-jacents à la réalisation de ces tâches et du contenu relationnel (Adrien and Cie 1988: 258). Nous avons relevé les éléments de cohérence ou d'incohérence du comportement de l'enfant, en regard du cadrage offert par la situation, et des demandes de coopération à l'action (pointage, et frappe des messages). Que semblent dire, ou signifier, les comportements non-verbaux de l'enfant, et sont-ils entendus par le facilitateur? Nous avons comparé entre elles les séances d'apprentissage, puis comparé entre elles les séances suivantes, et enfin comparé les séances dans leur suivi total, pour rechercher d'éventuelles homologies de structure entre les comportements non-verbaux des enfants. A. Les séances d'apprentissage Le facilitateur commence par inciter l'enfant à des activités de pointage sur des objets: Montres-moi le carton rouge. Puis il dirige cette demande vers des objetstexte, du type étiquette ("boulangerie") ou mots écrits sur une ardoise, en posant des questions comme: Où va-t-on acheter le pain: à la gare ou à la boulangerie? L'enfant, très rétif au début, se laisse progressivement tenir le doigt par le facilitateur. L'enfant, à qui ce dernier a montré l'objet (Regarde cette ardoise! Tu vois, j'écris des mots) s'intéresse aussitôt à l'objet, mais ses gestes montrent que son intérêt n'est pas lié à la tâche proposée. Il veut lécher ou examiner l'objet, empiler un carton l'un sur l'autre, ranger les étiquettes dans un sac, etc... Dès que le facilitateur a pu s'emparer de la main de l'enfant, on voit apparaître un dédoublement postural de l'enfant, en fonction d'une partition gauche-droite de son corps. En général, le facilitateur maintient dans sa main gauche la main droite de l'enfant: la partie droite du corps de l'enfant est donc contrainte par ce dispositif. Cela entraîne des "réponses" hétérogènes: F: Montres-moi le carton rouge! E: pointe son droit doigt contraint sur le carton rouge, prend en même temps le carton noir dans sa main gauche et le tend à sa mère. Les "réponses" sont souvent hétérogènes à la demande (l'enfant sourit au facilitateur, se détourne, veut se lever) ou hétérogènes entre elles (les deux parties du corps de l'enfant agissent différemment). Ou encore, parfois l'enfant montre le bon objet avec la main tenue, parfois avec sa main libre. Le facilitateur glisse ensuite ce qui apparaît comme une autre activité de pointage: celle des lettres sur le clavier. Pour lui, montrer/taper/écrire/dire sont interchangeables. Mais jamais il n'offre d'activité de pointage orientée vers les lettres elles-mêmes (Montre-moi un B, un H, etc...). L'enfant n'a pour toute expression reconnue que ce qui concerne la frappe des messages. Or, le comportement mimo-gestuel de l'enfant semble décalé par rapport au message tapé. Un message comme: Je finis de vivre comme une chose, est reconstitué à la

Usage de la 'Communication Facilitée ' avec des Enfants Présentant des Troubles Autistiques

253

fin de tout un ensemble de péripéties qui ont conduit à son obtention. Pendant la production du message, l'enfant ne cesse de chercher à attrapper le ruban de papier qui sort de la machine, veut saisir la machine elle-même, se tortille, veut se lever, ce qui conduit le facilitateur à barrer cette activité: il cache le ruban, enlève la machine et la remet devant l'enfant, réassied l'enfant. Pour ces séances, nous observons donc un manque de suivi dans la production des messages (au cours de laquelle le facilitateur est sans cesse obligé de rappeller la continuité thématique); une incohérence entre le comportement corporel de l'enfant et l'action désirée par le facilitateur; et une incohérence entre le comportement corporel de l'enfant et la production du message qu'il serait censé accompagner (Condon 1984: 48). Si nous dressons une liste des activités spontanées autonomes de l'enfant (la poursuite de ses propres thématiques) elles concernent la partie gauche libre de son corps (il tape la machine, touche le facilitateur ou les objets environnants, résiste au dispositif) ou des parties périphériques (il regarde ailleurs, sourit, crie, vocalise). Si l'on observe le cours de l'action en termes de dialogue corporel, on voit que le facilitateur refuse à l'enfant tout ce qui ne va pas dans le sens de l'action demandée et les deux corps semblent souvent en lutte. B. Les séances suivantes Le facilitateur oriente désormais l'action à 90% vers la frappe des messages. Dans cette situation contrainte, les enfants se conforment progressivement au modèle de conduite exigé. Les critères qui indiquent cette conformation progressive sont le recentrage corporel et la relative obéissance au processus de frappe. Nous observons un apaisement de la coupure précédente: l'enfant accepte d'avoir la main tenue, et assagit l'autre partie de son corps. Mais nous observons alors un détachement corporel périphérique: l'enfant continue d'une façon plus sage à toucher les objets environnants avec sa main gauche, laisse le ruban s'enrouler autour de son doigt gauche, regarde par la fenêtre ou ailleurs que le clavier, ou se couche sur la table en abandonnant sa main droite au facilitateur. Nous nous interrogeons sur la réalité de la coopération au processus de frappe. Pourtant, les enfants glissent de plus en plus souvent leur main dans celle du facilitateur, comme s'il s'agissait d'une demande de "parole", surtout lorsqu'une question est posée. Quand l'enfant fait-il ce geste? A la suite d'une intonation toujours identique (la question), lorqu'il semble éprouver un désarroi sur ce qu'il doit faire, ou dans ce qui apparaît comme une intention de bien faire (surtout après un épisode de rébellion ou d'agression contre le facilitateur). Indifférent au "co-texte" qui implique le non-verbal (Cosnier and Brossard 1988) le facilitateur ne voit pas que celui-ci n'accompagne pas d'une façon éclairante le "texte". Ce co-texte est comme la voix silencieuse de l'enfant, que le

254

Christian Hudelot et Christine Berger

facilitateur considère alors comme un brouillage, une perturbation du "texte" qui est la seule "voix" autorisée. Parmi les éléments de cette voix refusée, les plus intéressants sont les bruits de bouche, et les reprises en échos du dernier mot, échos encouragées par le facilitateur comme une ébauche de langage. Ils reprennent le rythme et l'intonation des mots lus par le facilitateur. Parfois, ils s'opèrent dans un décalage qui fait penser à une "citation" du même mot entendu dans un autre contexte: F: Tu dois aller E: allez-allez-allez. Nous remarquons parfois de petits coups frappés sur la table, comme une reprise en écho rythmique du dernier mot. L'écho apparaît comme une voix impersonnelle reprenant des morceaux d'énoncé, ainsi vidés de leur contenu sémantique et détachés de leur niche syntaxique: il arrache un mot et en fait une chose. Mais cette production peut apparaître comme la manifestation d'un désir de participation à la communication, en reprenant pour soi le dernier fragment. L'écho ferait du mot une chose "pour jouer"... Mais pas seulement: ces productions d'échos semblent plus nombreuses au cours des séances suivantes, ce qui pourrait aller dans le sens d'une participation interactionnelle, où l'écho peut prendre plusieurs fonctionnalités, y compris l'affirmation ou la protestation (Prizant and Rydell 1983; Loveland, Landry and Cie 1988). L'enfant présente parfois des conduites gestuelles de demandes: il veut un câlin, écouter de la musique, faire dessiner sa mère, etc... La comparaison des émotions exprimées par l'enfant lors de ces demandes, et des émotions exprimées lors de la frappe des messages soulève un problème. Ces dernières correspondent rarement au contexte des messages (souvent tapés dans l'indifférence), et semblent exprimer des préoccupations internes non élucidées en séance. Ainsi, les seuls dialogues cohérents sont ceux qui sont initiés par les demandes gestuelles de l'enfant, auquel les acteurs acceptent de répondre. Sont-ils l'indice d'une amélioration progressive des compétences communicatives de l'enfant dans le suivi des séances? Il semble que la CF offre une sorte de "holding émotionnel" participant à l'amélioration du comportement social de l'enfant (impliqué aussi dans d'autres structures thérapeutiques). Mais nous ne pouvons, dans l'état de nos observations, conclure en faveur du co-pilotage d'un "vrai" dialogue. Néanmoins, il faut insister sur le fait que des échanges signifiants apparaissent, ôn forme de dialogue, entre les "énoncés" gestuels de l'enfant et les énoncés verbaux des autres acteurs. En ce sens, certains chercheurs se penchent aujourd'hui sur les compétences pragmatiques des enfants autistes (Wetherby and Prutting 1984; Loveland, Landry and Cie 1988). C'est l'aspect pragmatique de la communication qui n'est pas pris en compte par les facilitateurs, orientés vers la recherche d'un "dialogue" verbal, au détriment de l'écoute de cette autre "voix" de l'enfant qui s'ex prime.

Usage de la 'Communication Facilitée ' avec des Enfants Présentant des Troubles Autistiques

255

Bibliographie Adrien, J., Barthélémy, C., Etourneau, F., Dansart, P., Lelors, G. (1988): Etudes des troubles de la communication et de la cognition d'enfants autistiques, Neuropsychiatrie de l'enfance, 36(7) 253-260. Bakhtine, M. (1979): Esthétique de la création verbale, trad, du russe, Paris: Gallimard. Biklen, D. (1993): Questions and answers about Facilitated Communication, In: Facilitated Communication Digest, 2(1), November. Condon, W. (1984): une analyse de l'organisation comportementale, In: La communication nonverbale (sous la direction de Cosnier and Brassard), Paris, Neuchâtel: Delachaux and Niestlé. Ducrot, O. (1984): Le dire et le dit, Paris: Minuit. Eberlin, M., Ibel, S., Jacobson, J. W. (1994): The Source of Messages Produced During Facilitated Communication with a Boy with Autism and Severe Mental Retardation: a Case Study, Journal of Pediatric Psychology, vol. 19 n° 6, 657-671. Flahault, F. (1978): La parole intermédiaire, Paris: Seuil. Howlin, P. (1978): Prognosis in Autism: Do Specialist Treatments Affects Long-term outcome? European Child and Adolescent Psychiatry, 6, 55-72. Loveland, K. A. and Landry, S. H. (1988): Speech Acts and the Pragmatic Deficits of Autism, Journal of Speech and Hearing Research, Vol. 31, 593-604. Prizant, B. M. and Rydell, P. J. (1984): Analysis of Functions of Delayed echolalia, In: Autistic Children, Journal of Speech and Hearing Research, 27, 183-192 . Sellin, B. (1994): Une âme prisonnière, Paris: Laffont. Vexiau, A. M. (1996): Je choisis ta main pour parler, Paris: Laffont. Vion, R. (1993): La gestion pluridimentionnelle du dialogue, Cahiers de Linguistique Française, n° 17, 179-203. Wetherby, A. M. and Prutting, C. A. (1984): Profile of Communicative and Cognitive-Social Abilities, In: Autistic Children, Journal of Speech and Hearing Research, vol. 27, 364-377.

Michèle

Grossen

Therapist-Child Dialogues in Clinical Interviews 1. Introduction Clinical interviews are socially and institutionally mediated practices which, at first glance seem to constitute a homogeneous type of interview, or even a genre (Grossen and Salazar Orvig 1998). However, despite a number of common points (Salazar Orvig 1998), there are different types of clinical interviews which vary depending upon the setting, the number of participants and therapists, the scope of the interview (does for example the therapist have a role of gatekeeper?), etc. In this paper, I shall examine a very common type of clinical interview, namely when one or two therapists meet, for the first time, parents (one or both) who are consulting for their child in a counselling centre for children and adolescents. In all the cases which will be examined, the reason for the consultation is the child's learning difficulties at school. The issue I will raise here concerns children's participation in this institutionalised type of interview (see also Grossen 1996; Grossen and Apotheloz 1998). In fact, the children's position in this situation is very particular: 1) 2)

3)

At a macro-social level, their status is subjected to a double asymmetry: the asymmetry opposing an adult to a child; the asymmetry opposing a patient to a therapist. Children, but also most of time adolescents, do not consult on their own initiative, but at their parents' request. Sometimes the situation is even more complex, since in public counselling centres such as those in which the present corpus has been collected, the parents themselves are often sent by their child's teacher. In many cases, even though they formally agree to consult, they may feel obliged to do so and have not made any personal request. The child's social and institutional position may contradict a common assumption in the field of psychotherapy, namely that it is impossible to help somebody who is not willing to be helped and does not actively collaborate in the search for a solution to his or her "problem".

At a micro-social level, an important task which child therapists have thus to carry out is that of favouring the child's participation as much as possible in order firstly to have access to his or her own definition of the problem, secondly to insure the child's collaboration in the therapeutic process. Previous research into this topic has shown that children's participation in multi-party institutional interactions is rather limited and that their contributions are to a variable extent under the adult's control. This has been seen to be

Michèle

258

Grossen

particularly evident in paediatric consultation (Aronsson 1991; Aronsson and Rundstrôm 1988, 1989), but also in clinical interviews (Aronsson and Cederborg 1996; Cederborg 1994, 1997). I shall briefly report results showing that the children's participation, at a quantitative level, is rather scarce. I shall then take a closer look at sequences of adult-child dialogues with the aim of examining whether there are any interactional dynamics which would account for this scarce participation.

2. Corpus The corpus is made up of 6 videotaped interviews which were selected from a larger corpus and fully transcribed (see appendix 1 for the norms of transcription). These interviews were the first sessions of consultations which took place in two counselling centres for adolescents and children and were all prompted by poor school achievement. They were conducted by one or two psychotherapists and lasted for about sixty minutes. In some cases, the whole family was present, while in other cases only one parent, usually the mother and the child participated in the interview. Table 1 : Characteristics of the corpus Gender Age Participants Child's first name and name (pseudonyms) M 6 Julien ALBERT WT, M Julien, researcher 1 M 8 WT, M, Alain Alain GONTRAND 12 F MT, M, Alice Alice HECTOR M 13 MT, F, Bernard Bernard JULES M 13 MT, WT, F, M, Jean, S Jean AMBROISE 14 M MT, F, M, André, SI, S2, researcher 2 André BARNABE WT: woman therapist; MT: man therapist; M: mother; F: father; SI: sister. As can be seen in Table 1, two of the children were between 6 and 8 years old, while the four others were between 12 and 14. Quite obviously, it will be impossible to say whether some of the types of dialogue I shall describe further on may - at least in part - be explained by developmental factors.

3. The Child's Participation: Quantitative Results I will now participation: -

briefly report

the

quantitative

results

concerning

children's

The 6 interviews contained 131 sequences in which the child made at least one intervention. The number of these sequences varied from 13 to 32. Out of these 131 sequences, the major part (78) were opened by an adult (62 by the therapist) who asked

Therapist-Child Dialogues in Clinical Interviews

-

-

259

the child a question. However, out of these 78 sequences, 32 gave way to a dialogue between the adult and the child. Dialogues between a parent and the child, or between the child and his sisters (AMBROISE and BARNABE) were rare, an observation which shows that it is actually the therapist's role to guide and orchestrate the interview. On a strictly quantitative level, the children's speaking time was much lower than the adults', which confirms the results of previous studies (Cederborg 1994, 1997). However, further analysis shows that the number of questions that the therapist(s) asked the parent(s) was not different from that addressed to the child(ren). Consequently, it cannot be concluded that children were less solicited than their parents. Another hypothesis was also that the children's scarce participation might be due to the fact that they did not take the floor without being invited to do so. But here again, the hypothesis was not confirmed: in the whole corpus, the children made 72 non-invited interventions. Hence, the issue is rather that their interventions were rarely taken up by the adults. As a matter of fact, only 9 of them gave way to a dialogue between the adult mostly the therapist - and the child. Moreover, this result was apparently not due to the fact that the children did not make their intervention at the right moment, or at a transition point.

These quantitative results are only first indications which show that in order to gain a better understanding of the children's scarce participation, it is necessary to analyse the interactional dynamics.

4. Analysis of Therapist-Child Dialogues Let me now present some results from the analysis of the 32 sequences which were opened by an adult, namely in all cases by a therapist's question, and then move to the analysis of the 9 sequences which were opened on the child's initiative.

4.1. The Sequences Opened by a Question from the Therapist A close examination of the 32 sequences which were opened by an adult shows that there are three types of dialogues. One type may be observed throughout a whole sequence but can also be only part of it. The first type of dialogue is illustrated in Excerpt 1 and consists of a pattern in which: a)

b) c)

The therapist formulates a closed question, namely a question which, at a literal level, can be answered with a minimal ratification such as "yes" or "no" (see turn 1MT), but could also be elaborated. The child answers with a minimal ratification or a brief response which gives the floor back to the therapist. The therapist acknowledges the child's minimal ratification (see turn 3MT) or repeats the child's response (see turn 9MT), and asks another closed question, as in 5MT, 7MT, 13MT, 17MT.

260

Michèle

Grossen

After one such pattern, or as in excerpt 1 after a repetition of this pattern, the dialogue fades out. In this case, the therapist addresses another participant, or one of the parents takes his or her turn after a pause. This type of dialogue was observed in 17 of the 32 analysed sequences:

Excerpt 1 (AMBROISE, 784-802) 1MT: 2Jean: 3MT:

4Jean: 5MT:

6Jean: 7MT: 8Jean: 9TF: 10 Jean: 11MT: 12TF: 13MT: 14Jean: 15MT: 16 Jean:

17MT: 18Jean: 19MT:

[hmm] ((à Jean)) tu as entendu un peu ce qu'ont dit tes parentsî hmm ouais + tu trouves que c'était des bonnes cho:ses des moins bonnes cho:ses ou pas des bonnes chosesî des bonnes choses des bonnes choses ouais + et pis quand ils disent que tu parles pas beaucoup ou que tu dis pas ce que tu fais ou:: tu es d'accord avec eux làt certaines fois avant beaucoup maintenant moins maintenant moins + tu essaies de dire plusî à l'école je parle beaucoup = = à l'école tu parles beaucoupî ouais xxx [ouais] [hmm] tu as des copains avec qui tu parlest euh oui hmm je suis jamais tout seul maintenant + je sors presque tous les jours jouer au hockey hmm + tu aimes bien le hockeyt ouais le hockey sur glace ou hockey sur terreî

1MT: 2Jean: 3MT:

4Jean: 5MT:

6Jean: 7MT: 8Jean: 9TF: lOJean: 11MT: 12TF: 13MT: 14Jean: 15MT: 16Jean:

17MT: 18Jean: 19MT:

[hmm] ((to Jean)) did you hear a little bit what your parents saidt hmm yeah + do you think there were good things few good things or no good things good things good things yeah + and when they say that you don't talk very much or that you don't say what you're doing or do you agree with them thent sometimes before a lot now less now less + can you try to tell a little bit moreT at school I talk a lot= = do you talk a lot at schoolt yeah xxx [yeah] [hmm] do you have friends you talk withT err yes hmm now I am never all on my own + I go out almost every day to play hockey hmm + do you like hockeyT yeah ice hockey or ground hockeyt

Therapist-Child Dialogues in Clinical

Interviews

261

In this type of dialogue, the therapist clearly assumes the responsibility of guiding the interview. He or she is in a position to continue the dialogue with the child, to let somebody else take the floor, or to give the floor to another participant. He or she is also in charge of introducing new topics or maintaining some coherence between one question and another by taking up the child's answers and putting them in a question form, as in 19 MT for example. Thus, the therapist's contributions have strong initiatory aspects (Linell, 1998) and the child-therapist positions are very asymmetrical. This type of dialogue is often observed at the beginning of an interview, when the therapist collects some information about the situation. Questions about topics such as the name of the school or of the teacher, the school curriculum, the level of achievement, the number and names of siblings, etc. are frequent and are, at least in some cases, pseudo-questions since the therapist already has this information in the child's file. But these types of dialogue may also appear further on in the interview, and seem to include some unavoidable (or routine) questions, such as the child's relationship with his or her friends, classmates or siblings, the child's attitude towards school, or the child's hobbies. This type of dialogue is very close to IRF formats described by Mehan (1979) or, more recently, by Mercer (1995) in teacher-student verbal interactions. However, a difference is that the therapist does not guide the child towards a predetermined (or expected) answer, but uses closed questions as resources to give the child an opportunity to participate in the conversation. But since the child treats the questions at a literal level, the therapist's interrogation, instead of producing a dialogue, leads to its extinction. The paradox which is worth stressing here is that even though the therapist seems to control strongly the child's discourse, the child's non-talkativeness may also be considered as a constraint put upon the therapist's discourse. In fact, instead of providing the therapist with information which could be used in the therapeutic process, the child implicitly obliges the therapist to take the floor and to be the one who produces a discourse and puts him or her in the position of a research interviewer, a journalist or even a policeman interrogating a suspect. The second type of dialogue is illustrated in Excerpt 2 and consists of the following pattern: a) b) c)

The therapist asks an open-ended question (5WT: "How do they [Alain's friends] react when you 're abrupt?") or a closed question. The child gives a minimal ratification or no response at all. The therapist reformulates his or her question by putting it into the form of an alternative. For example in 5WT, the therapist reformulates the open-ended question "How do they [his friends] react when you 're abrupt?" in the following alternative "What does it mean that you 're abrupt? do you kick or punch or or do you swear at them?". This question offers two possible answers: kicking or punching on the one hand, and swearing at his

262

Michèle

d)

Grossen

classmates on the other. Thus, the child's task is framed in such a way that he only has to choose between one of the two terms of the alternative, Following this the pattern may then take on two different forms: in the first, the therapist continues the dialogue by regularly offering alternative questions. This is the case in Excerpt 2. In the second, the therapist exploits the resources offered by the first term of the alternative and, when the dialogue begins to fade out, exploits the second term of the alternative.

This pattern was observed 9 times in the 32 sequences analysed. Excerpt 2 (GONTRAND, 125-151) 1WT: 2M: 3WT: 4M: 5WT:

6Alain: 7WT:

8Alain: 9WT:

lOAlain: 11WT:

12 Alain: 13WT:

alors il est brusque et puis ça provoque des réactions? ouais de la part des autres? ouais ((se tourne vers A)) comment ils réagissent quand tu es brusque? ça veut dire quoi que tu es brusqueî tu donnes des coups de pieds ou des coups de poings ou ou ou tu les engueules? comment ça se passeî des fois ils m'embêtent ils t'embêtent, alors quand ils t'embêtent qu'est-ce que tu faist je me défends mais tu te défends comment avec les poings ou avec la la la bouche? et pis tu dis des des choses qu'ils aiment pas entendre ou plutôt tu cognesî qu'est-ce que tu fais plutôtî euh: je leur dis des vilains mots= =la bouche, et pis tu leur dis des vilains motsî + et pis après? + ils sont fâchés? + ou bien ils arr[êtent de t'embêter] [non après ils continuent] ils continuent^ et pis après comment ça se passet alors ils continuent toi tu continues à

1WT: 2M: 3WT: 4M: 5WT:

6Alain: 7WT: 8Alain: 9WT:

lOAlain: 11WT:

12 Alain: 13WT:

14 Alain: 15WT:

so he is abrupt and it provokes reactions? yeah from the otherst yeah ((turning towards A)) how do they react when you're abrupt? what does is mean you're abrupt? do you kick or punch or or to you swear at them? how does it happen? sometimes they bother me they bother you, so when they bother you, what do you do? I defend myself but how do you defend yourself with your fists or your mouth? and then do you say things they don't like to hear or do you instead fight? what do you do instead? euh: I call them bad names= =the mouth, and then you call them bad names? + and then? what + are they angry? + or do they [stop bothering you?] [no then they carry on] they carry o n l and then what happens next? so they carry on and you do you carry on shouting? well yeah and then how does is end?

Therapist-Child Dialogues in Clinical

14 Alain: 15 WT: 16Alain: 17WT: 18 Alain: 19WT:

20Alain: 21WT:

263

Interviews

crierî ben ouais

16Alain:

well + + it ends with- with m e getting angry y o u get angry and then do y o u throw p u n c h e s t < ® n o I shake them ® > < ® y o u shake them ® T > and

et pis a p r è s t ça finit c o m m e n t ! 17 WT: ben + + ça finit avec - avec que 18Alain: j e m'énerve tu t'énerves et pis après tu 19WT: then when y o u shake them, donnes un coup de p o i n g s î what happens nextT do the < ® n o n j e les s e c o u e ® » et pis girls c r y t and the boys what après quand tu les secoues do they d o t do they shake y o u qu'est-ce qu'il se p a s s e t les toot filles elles pleurentt pis les 20Alain: no they don't shake m e they garçons qu'est-ce qu'ils font ils stop te secouent a u s s i î 21 WT: they stop, hmm (...) non ils m e secouent pas ils arrêtent ils arrêtent^ hmm (...)

The main characteristics of this type of dialogue is that the therapist uses different devices or strategies in order to guide the child's discourse and help him or her provide a chronological account or story. In Excerpt 2, the guiding strategies or devices used by the therapist are the following: -

-

the therapist reformulates her question (see for example in 9WT: "But how do you defend yourself with your fists or with your your your mouth? (...)"); she helps the child to tell a story firstly by focusing her request upon the child's actions (5WT: "How do they react?"-, 7WT: "What do you do?"-, 9WT: "What do you do instead?", etc.); secondly by regularly providing temporal markers which structure the story (9WT and 11WT: "(...) and then? (...)"; 13WT: "(...) and then what happens? (...)"; 15WT: "and then? how does it end?"; 17WT: "you get angry and then do you kick?"-, 19WT: "you shake them and then when you shake them what happens next? (...)"); she suggests some possible answers by taking the child's perspective and mental state, and imagining the continuation of the child's story (UWT: "(...) and then are they angry? or do they stop bothering you?"; 13WT: "(...) and then what happens next? so they carry on, you, you carry on shouting?")-, 19WT: "(...) what happens? do the girls cry? and the boys what do they do? do they shake you too?"). But of course, in so doing, the therapist runs the risk of proposing a formulation or an alternative question which does not correspond to the child's point of view.

In this case, the child-therapist dialogue reminds us of other types of adult-child dialogue, in particular those described by Rogoff (1990) by the term of "guided participation" (see also Rogoff 1995). Even though children are led to be more talkative, and in some cases to take on more responsibility for what they say, the major part of the interactive work is done by the therapist. But the therapist's guidance also has two other effects: the first is implicitly to teach the child how to

264

Michèle

Grossen

produce a certain type of account in this particular context; the second is indirectly to provide the parents with an account or another perspective on the child's 'problem', by using the child's discourse. The third type of dialogue (see Excerpt 3) consists of a pattern in which the therapist's questions elicit elaborated answers upon which the therapist can formulate his or her next question. This type of dialogue is found 9 times in the 32 sequences and concerns two of the adolescents (BARNABE and HECTOR), as well as two of the youngest children of the corpus:

Excerpt 3 1MT:

2André:

3MT: 4André:

5MT: 6André:

(BARNABE, 1174-1186) [quand quand quand tu entends ton père] là qu'est-ce qu'est-ce que tu lui répondsî t'as parfois la trouilleî [pour tes moyennes] [quand j'ap]prends quelque chose je crois que des fois j'apprends TROP j'apprends tellement que j ' e n oublie la moitié après hmmt + + je sais pas comment expliquer ça c' c'est quand je me retrouve devant l'épreuve + j'ai trop de soin pour les choses importantes en tête pis je confonds je mélange hmml + pis quand c'est non appris j'aij'ai r- j'ai rien en mémoire pis euh je fais sur ce que j'ai entendu ou sur ce que j'ai vu je je je mets je mets tout ce qui me reste dans la tête

1MT:

2André:

3MT: 4André:

5MT: 6André:

[when when you listen to your father then] what do you say to h i m t are you sometimes afraid [of your markst] [when I learn] something I think that sometimes I learn TOO MUCH I learn so much that I forget half of it then hmmt + + I don't know how to explain it it's its when I am in front of the exam + I pay too much attention to the important things in my head then I get mixed up hmmi + then when it's not learned I have I forget everything and err I work on what I heard or saw I I put down every thing that's left in my head.

In this type of dialogue, the therapist's question, even though it is an open question, elicits an elaborate answer. Sometimes, as in Excerpt 3, the therapist (1MT) reformulates his or her open-ended question as a closed question, but the child answers the question by adding elements which were not contained or presupposed in the therapist's question. Moreover, as Excerpt 3 shows, a simple "hmm" or a pause (3MT, 5MT) are liable to be interpreted by the child as an invitation to continue his or her discourse. A difference between this type of

Therapist-Child Dialogues in Clinical

265

Interviews

dialogue and the preceding ones is that the therapist-child dialogue seems to be based more upon reciprocation, namely the co-ordination and synchronisation of both parties' actions (Linell 1988); consequently, the child plays a more active part in the construction of the dialogue. The responsive and initiatory aspects (in the sense defined by Linell, 1998) are more balanced and continuity between the partners' discourse seems more important. Of course, this does not mean that the therapist does not guide the child's discourse, but he or she does it in a way which does not differ from adult-therapist dialogues or any other institutional discourse.

4.2. The Sequences Opened by the Child's Non-Invited Intervention Let me now say a few words about the sequences which were opened by a child's non-invited intervention. As I mentioned before, 9 of these interventions gave way to a dialogue between the adult and the child. As a preliminary observation, it is worth noting that some of these sequences, but not all of them, elicit a type of dialogue in which the child is active and takes on responsibility with respect to his or her own discourse. From this vantage point, non-invited interventions seem to offer the adults, and in particular the therapist, an opportunity to take the child's perspective. However, one could ask why the therapist does not take this opportunity more often. In line with previous work on multi-party talk in clinical interviews (see for example Aronsson and Rundstrôm 1989), a possible hypothesis could be that the children's non-invited interventions pose a threat to the adults' face, the parents in particular, and constitute a possible danger for the management of the participants' interpersonal relationships. A close analysis of the corpus suggests a confirmation of this hypothesis, showing that this is especially the case for the children's non-invited interventions which express a conflict or a disagreement with the adult's discourse. The disagreement may concern the parent's discourse, as in Excerpt 4, but sometimes it also concerns the therapist's discourse.

Excerpt 4 (HECTOR, 384-400) 1MT:

2M: 3MT: 4M:

5Alice:

est-ce qu'au début aussi quand ça allait bien il y avait aussi des difficul- des difficultés pour faire les devoirst oh oui oui [il y a] [toujours] =toujours eh heini il y a toujours eu depuis la première armée il y a eu= non [en première année ça allait quand même]

1MT:

2M: 3MT: 4M: 5Alice: 6M:

was it only at the beginning as well when things were going ok or were there also difficuldifficulties doing homeworkT oh yes yes [there was] [always] =always e r r i there were always since the first year there were= no [in the first year it worked though] [oh: well well] there was always

Michèle

266

6M:

[oh: bon bon] il y a toujours plus ou moins 7Alice: en première année ça allait encore 8M: des cris pis des pleurs 9 Alice: xx 10M: pas toujours! mais je dis pas que c'est tous les jours [Alice] 11 Alice: [oui je sais] 12M: j'ai pas dit que c'était tous les jours que tu faisais que tu faisais ta petite crise, quand + mais ça arrive souvent, + d'accordî + hmmî 13MT: donc le travail scolaire a toujours été un peu difficile

Grossen

more or less 7 Alice: in the first year things were still going ok 8M: shouts and crying 9Alice: xx 10M: not always! I'm not saying it's every day [Alice] 11 Alice: [yes I know] I didn't say that it was every day 12M: that you had you had your little crisis, though ((®)) + but it often happens + ok?+ hmmt 13MT: so school work has always been a little bit difficult

In Excerpt 4, Alice expresses her disagreement with her mother's account that since her first year of school she had difficulty doing her homework. Alice seems to interpret the mother's account as a threat posed to her own identity and her intervention has the effect of leading her mother to mitigate her discourse (10M: "not always, but I don't say it was every day"; 12M: "I did not say that it was every day that you did not. ") and to engage in a repair sequence with her daughter. The therapist then reformulates the mother's discourse without directly acknowledging Alice's discourse but by using a mitigator (13MT: "school work was always a bit difficult") that he did not use in his first question (1MT). Thus, the child's intervention seems to be a sort of side-sequence which temporarily interrupts the mother-therapist dialogue, but orientates the adults' formulation of her own "problem". However, it may also threaten the mother's face, since it calls into question the accuracy of her discourse and hence her accountability. Excerpt 5 is another example which illustrates that the child's non-invited intervention may threaten the mother's face. Excerpt 5 (GONTRAND, 197-212) (La mère raconte à la thérapeute comment le (The mother tells the therapist how her children's return after school is organised) retour des enfants de l'école est organisé). 1 Alain: 2M: 3WT:

maintenant j'ose plus aller vers 1 Alain: mes copains parce qu'elle m'interdit maman (sourit) 2M: ouais ben ça c'est autre chose, + elle t'interdit ta maman d'aller 3WT:

now I don't dare to go to my friends because my mother forbids it (smiles) yeah but this is another story, + does your mother forbid you to

267

Therapist-Child Dialogues in Clinical Interviews

4Alain: 5M: 6Alain: 7M: 8WT: 9M:

chez les copainsî [oui] ((se tourne vers Alain)) [avant] + les devoirs oui ((à Alain)) bon ® c'est pas tout à fait la même chose non

go to your friendsT [yes] ((turns towards Alain)) [before] homework 6 Alain: yes 7M: ((to Alain)) well (laughs) 8WT: it's not exactly the same thing 9M: no 4M: 5M:

(suite du dialogue entre la thérapeute et (then the therapist speaks with the child) 1'enfant)

In Excerpt 5, the child addresses the therapist without having been invited to do so (turn 1 Alain). He takes up a theme which has been introduced by the mother, namely the fact that when he or her sister go to friends after school, they give a ring home. His intervention seems thus to be in contradiction with his mother's account. However, it is immediately picked up by his mother who brings a new element (5M: "before homework") which, in this context, appears to be both a counter argument opposed to the child's discourse and an argument defending her own position in front of the therapist. As a result, Alain (turn 6) accepts his mother's counter argument, while the therapist's (8WT) intervention seems to be another argument in favour of the mother's account. The fact that the therapist then enters into a dialogue with the child might be seen either as a sign that she interpreted the child's intervention as an attempt to participate, or as a sign that she tries to repair the threat that this dialogue has posed to the child's face. This example illustrates the complexity of this multi-party talk, since when discussing with one person, the speaker has the delicate task of protecting the addressee's face without threatening the other participant's face. In this perspective, limiting the child's access to the floor or, if one considers the child's side, self-limiting one's own access to the floor, may be considered as ways of having a better control upon interpersonal management and avoiding any potential faux-pas that such a situation might bring about.

5. Conclusions In conclusion, the comparison between the sequences opened by an adult, and those opened by children's non-invited intervention leads to the following paradox: when the therapist invites the child to participate in the conversation, the dialogue is difficult to establish and to maintain, and the child's discourse appears to be determined mostly by the adult's guidance. Contrary to this, when the child makes an attempt to take the floor without being invited to do so, his or her intervention is rarely taken up by the adult, even though there are serious reasons

268

Michèle

Grossen

to think that this is not because the intervention is topically or sequentially out of place. A hypothesis which is also supported by other studies is that the child's scarce participation might be due to the social delicateness of a situation in which, after all, it is not completely clear who, between the child and the parents, are the patients.

References Aronsson, K. (1991): Facework and control in multi-party talk: A paediatric case study. In: I. Markova, and F. Foppa (eds.), Asymmetries in dialogue. - Hemel Hempstead: Harvester Wheatsheaf, 49-74 Aronsson, K., and Cederborg, A-C. (1996): Coming of age in family therapy talk: Perspective setting in multiparty problem formulations. Discourse Processes, 21, 191-212. Aronsson, K. and RundstrOm, B. (1988): Child discourse and parental control in pediatric consultations. Text, 8, 159-189. - (1989): Cats, dogs, and sweets in the clinical negotiation of reality: On politeness and coherence in pediatric discourse. Language and Society, 18, 483-504. Cederborg, A.-C. (1994): Family therapy as collaborative work. - Linkoping: Linkoping Studies in Arts and Science, 106. - (1997): Young children's participation in family talk. The American Journal of Family Therapy, 21,(1), 28-38. Grossen, M. (1996): Counselling and gatekeeping. Definitions of the problem and situation in a first therapeutic interview. Text, 16 (2), 161-198. Grossen, M., and Apotheloz, D. (1996): Communicating about communication in a therapeutic interview. Journal of Language and Social Psychology, 15 (2), 101-132. - (1998): Intelligence as a sensitive topic in clinical interviews prompted by learning difficulties. Pragmatics, 8 (2), 239-254. Grossen, M., and Salazar Orvig, A. (1998): Clinical interviews as verbal interactions: A multidisciplinary outlook. Introduction. Pragmatics, 8 (2), 149-154. Linell, P. (1998): Approaching dialogue. Talk, interaction and contexts in dialogical perspectives. Amsterdam: John Benjamins. Mehan, H. (1979): Learning lessons. - Cambridge: Harvard University Press. - (1995): The guided construction of knowledge. Talk amongst teachers and learners. London: Multilingual Matters. Rogoff, B. (1990): Apprenticeship in thinking. Cognitive development in social context. New York: Oxford University Press. - (1995): Observing sociocultural activity on three planes: Participatory appropriation, guided participation, and apprenticeship. In: J. V. Wertsch, P. del Rio, P. and A. Alvarez (eds.), Sociocultural studies of mind, 139-163. Cambridge: Cambridge University Press. Salazar Orvig, A. (1998): Interpreting and diverging in clinical interviews. Pragmatics, 8 (2), 167184.

Therapist-Child

Dialogues

in Clinical

Interviews

Appendix 1: Norms of transcription Overlaps [

]

[ and

]

[[

]]

[[

]] when two overlaps or more are next one to the other

Telephone rings, whispers, etc.: in brackets Transcriber's comment: ((

))

Stress on a word or a syllable: CAPITAL LETTERS Intonation markers:

T rising intonation (not necessarily a question) 4- falling intonation

® laughter ® Pauses:

+ (1 sec) ++ (2 sec)


< >:

uncertain transcription, two possible versions are proposed.

269

Alison Ferguson

Understanding Paragrammatism: Contributions from Conversation Analysis and Systemic Functional Linguistics 1. Introduction This paper considers some of the insights gained from very different ways of investigating and understanding the effects of one type of aphasic impairment on conversational interaction: Conversation Analysis, and an approach based on Systemic Functional Linguistics. The unique contribution of each perspective changes the ways speech pathologists assess and intervene when working with aphasia resulting from damage to the dominant cerebral hemisphere. More specifically, the contrasting insights from the different perspectives allow us to re-examine key issues in the on-going debate regarding the nature of paragrammatism. The aphasic speaker who participated in the conversations being discussed is JB, a right-handed 78 year old man who had suffered a left cerebrovascular accident following heart surgery 6 months prior to participating in the study. This paper presents the analysis of three 10 to 15 minute conversations with JB that were video recorded and later transcribed. His conversation partners were three females aged between 26 and 36 years: a Familiar Clinician - the researcher, who was his speech pathologist, and had been seeing him once or twice a week since the stroke; an Unfamiliar Clinician, who was a speech pathologist with experience in working with people with aphasia, but who was unknown to JB; and an Unfamiliar Layperson, who was a postgraduate Linguistics student, who knew about aphasia, but who had not interacted with a person with aphasia previously. The conversations were transcribed orthographically, with phonetic transcription of neologisms when required. Detailed analyses were undertaken for the first 5 minutes of each conversation, and the quantitative data presented in this paper are drawn from these sub-samples. To illustrate the findings in this paper, what I propose to do is work from one small excerpt of one of the conversations, in order to bring out and exemplify the patterns and trends observed from the larger samples (see Example below).

2. Background A traditional syndrome approach to aphasia would involve error analysis from the spontaneous language of the aphasic speaker alone, in comparison to psychometric

Alison

272

Ferguson

test results (Goodglass and Kaplan 1983), and would note such features as seen in the following example: normal melodic line, phrase length, and articulatory agility; marked anomia (word-finding difficulty - e.g. t25 "what do they call that..." - the target word was not retrieved despite repeated attempts over the ensuing five minutes conversation); 'jargonaphasia', involving frequent semantic paraphasias (word substitutions - e.g. t l 9 "hail" - target not established), some of which may arise from phonemic paraphasias resulting in other recognisable English words, and occasional neologistic words (words not previously in the lexicon of the language of the speaker - e.g. t l 3 / n w I , n w a e n / - target not established); and 'empty speech' with the lack of clear anaphoric reference, particularly relating to pronoun use (e.g. tl3 "he...she...her"). In contrast to the agrammatic output associated with non-fluent Broca's aphasia, JB shows the type of output referred to as 'paragrammatism' in which the speaker shows the full range of grammatical form, but "confused and erroneous grammatical structures" (Butterworth, Panzeri, Semenza and Ferreri 1990).

Example. JB and FC, Turns 13-27 JB 13. FC 14. JB 15. FC JB FC JB

16. 17. 18. 19.

FC JB FC JB

20. 21. 22. 23.

FC 24. JB 25.

FC 26. JB 27.

So he had he's, he's only forty-one and she had a blooming.. A i w I/, a little / n w ae n/ (INDICATES SIZE) to take her like, Right, so (OVERLAP) And Dulcie's only sixty-two like you know she's younger sister so she brought him down from, Junee, Yes .. and how long is it since you've seen them ? (OVERLAP) Eh? How long since you've seen him? Oh he hasn't been there for about 12 months, he'd be there, he'd stay about, it all depends how he feels, he's real lively for a little bloke and that like you know, you'd think he's a big bloke but he's not, he'll stay about three weeks or a month or, and then he's gotta fix his hot day fixed up otherwise well it's gonna just rain out of the hail and that sort of thing, you've gotta have people to look after their water Of course they can't leave the property for long (OVERLAP) They can't get everyone of them all the time Yeah He's good when he likes the drought like that, we're lucky there's one extra X, that all gave him plenty of room right there's one room for him (FC BC Hmm) and then George came today Yeah Cause his brother, she didn't come just is she had to go for a holiday down the coast down Mulla way, what do they call that new place, new um, the ah ..what do you call blooming ....we used to call it ah ah.. I don't know what the damn things calls. Down the south coast? Yeah but down on the south like you know where the right where the right the, place is I haven't been there lately and we went there a lot you know

Understanding Paragrammatism: Contributions from CA and SFL

273

Such syndrome approaches have been criticised by some as being primarily descriptive, and as failing to provide an explanatory account of aphasia (Schwartz 1987). In contrast, applications of theories of cognitive-neuropsychological processing have highlighted the diversity of underlying deficits within any one syndrome grouping such as 'Broca's' or 'Wernicke's' aphasia, and even within any one symptom complex such as 'agrammatism' or 'paragrammatism'. From a cognitive-neuropsychological perspective, there is a debate emerging between the nature of the underlying deficit (Davis 1993). Some suggest there is a structural deficit (i.e. a selective impairment of syntax where thematic role assignment appears to be differentially impaired in contrast to relative sparing of traces in deep structure left by constituent movement - Saffran, Schwartz, and Linebarger 1998). Others suggest there is a processing deficit (i.e. an impairment in controlled processing, whether selective or more generalised - Blumstein, Milberg, Dworetzky, Rosen and Gershberg 1991; Zurif, Swinney, Prather, Solomon and Bushell 1993). Alongside this debate is the contention that what we are seeing in paragrammatism is the fundamental impairment of Wernicke's aphasia, without any adaptation or strategic compensation (Kolk and Heeschen 1992). Such psycholinguistic approaches lead to intervention which focusses on deficit, for example, semantic representation, phonemic selection/discrimination, thematic role assignment. In these approaches to therapy it is common to establish the therapeutic response in a one-to-one clinical setting, but then to promote carryover to other dialogic settings using other more sociolinguistic approaches. With these issues in mind then, we turn now to look at what alternative insights might arise from two very distinct sociolinguistic perspectives: Conversation Analysis, and Systemic Functional Linguistics.

3. Conversation Analysis The early work of Sacks, Schegloff and Jefferson (1974) was applied by Schienberg and Holland (1980) to the analysis of two speakers, both with severe Wernicke's aphasia, and I recently published a study that replicated their findings (Ferguson, 1998). JB was one of two aphasic subjects in the Ferguson (1998) study, and the findings from the study that are relevant to the present paper are summarised here (i.e. from conversations between JB and FC - Familiar Clinician, JB and UC - Unfamiliar Clinician, and JB and UL - Unfamiliar Layperson). Turn allocation techniques were described as 'you next' (current speaker selects next speaker), or 'me next' (next speaker self-selects), or 'me more' (current speaker self-selects) or the conversation ends. What was striking from the results was the preponderance of 'me more' turn allocation by JB (67% of turn allocations by JB with UL, 71% with UC, 76% with FC), and without analysing

274

Alison Ferguson

the partners' contributions it would be tempting to consider this to reflect his 'press of speech'. However, looking at the partners' contributions, a preponderance of 'you next' turn allocation techniques were identified (86% of turn allocations by UL, 90% by UC, 88% by FC). So, although the interactants were instructed to have a natural conversation, what ensued was more like a conversational interview, i.e. one partner was eliciting responses from the other, although about topics that are typical of everyday conversation. Certainly, JB was able to employ each of the turn allocation techniques: for example, turn 15 - "And Dulcie's ..." - me more; turn 17 - "Eh?" - you next; turn 21 - "They can't..." me next. A variety of turn construction units were used by all speakers, although the distribution of backchannels reflected the speaker/listener roles of the interactants. The Unfamiliar Layperson made more use of backchannels (14) in comparison with the Familiar Clinician (9) and the Unfamiliar Clinician (8). As the conversation progressed between UL and JB, the Unfamiliar Layperson was observed to 'ride with' a considerable degree of ambiguity, for example, listened using only backchannels for a lengthy turn by JB (60 clauses). This listening strategy contrasted with the scaffolding offered by both Clinicians, examples of which can be seen in tl6, t20, t26. This finding suggests that the scaffolding associated with the role of clinician in the context of therapy may continue in such conversational interviews (Silvast 1990a, b; Sobiecka-Koszel 1991) The structure of turn-taking provides the structure for repair of trouble in conversation (see Schegloff 1979, 1987, 1992 and Schegloff, Jefferson and Sacks 1977 for detailed discussion of repair mechanisms). In all conversations, JB undertook self initiated same turn self repair for the greater proportion of repairs, but this did not vary across conversations with different partners. He demonstrated the ability to utilise the turn taking structure to undertake all types of repair over the course of the three conversations. Some examples can be found at tl3, with a same turn self correction of a neologism (albeit unsuccessful); at tl7, with an other initiated self repair; at t25, with a self initiated self repair. The present research adds to previous research that challenges the extent to which it can be said that speakers with Wernicke's aphasia fail to self-monitor (Marshall, Robson, Pring and Chiat 1998). Ferguson (1994) found that the nature of the activity affected the type and distribution of repairs, since structured testing limited opportunities to repair, but some semi-structured situations promoted opportunities to repair. The issue with regard to unstructured conversation comes back to Goffman's (1981) insights regarding the amount of repair work any conversation can bear without so disrupting the conversational flow that the interaction ceases to be conversational. Perhaps, the considerable repair work observed in these conversations is at the outer limit, and hence there is, in a way, a

Understanding Paragrammatism: Contributions from CA and SFL

275

'ceiling effect' that makes for similarity of repair regardless of partner, as found in this data for JB. Using a Conversation Analysis (CA) approach widens the context of both assessment and treatment. Necessarily, the exchanges we need to observe should involve people in interaction, not stimulus/response pairs. We need to observe and analyse what clients do with people other than clinicians and in the settings of their daily life on more than one occasion. We need to pay as much attention to the partner as to the aphasic speaker, so that the aphasia ceases to be the 'disability' of the person with aphasia, and can be recognised as a joint 'handicap' shared with other members of society. For JB, we can see turn-taking and repair in interaction as a relative strength. This is particularly important for him, given his reduced auditory comprehension, since as Sacks, Schegloff and Jefferson (1974) point out, it is through the turn-taking that we can both display and negotiate our understanding, for example, t l 7 JB "Eh?". From the starting point of using CA as a sociolinguistic approach to investigate paragrammatism, we can ask whether the findings of syntactic error in whether paragrammatic language are revealing the impairment, or paragrammatism is created in the attempt to repair the underlying difficulty. This question is possibly hidden from a psycholinguistic view, because of fundamental and theoretically-motivated methodological differences in what is considered to be the data for analysis. For purposes of psycholinguistic research into speech error data, it would be usual procedure to 'clean up' the data, through the exclusion of restarts, repetitions and revisions, as well as to separate out all ungrammatical or nonsensical sentences. For example, following the guidelines for narrative analysis by Saffran, Berndt and Schwartz (1989) would result in the deletion of approximately half the data. If we exclude this part of the data, then we lose the opportunity to view the revision and adaptation of the aphasic speaker. Conversation Analysis has shown us how speakers use conversational structure to negotiate meanings, but it does not seek to explore the nature of the meanings being exchanged. For that, we turn now to Systemic Functional Linguistics.

4. Systemic Functional Linguistics The particular area of interest that emerges when we hear JB talking, is the question, how is it that his use of language can be 'confused and erroneous' (Butterworth et al 1990), yet somehow communicative, at one and the same time. It is evident that he has considerable problems retrieving lexical items, yet what resources remain available to him to express the relationships between what he can retrieve? To look at this, I have analysed the logical dependency relationships (taxis) and the logico-semantic relationships between clauses forming clause

276

Alison

Ferguson

complexes. The analysis is based on procedures set out by Martin, Matthiessen and Painter (1997) and Halliday (1994).

5. Logical Dependency and Logico-Semantic Relations The asymmetrical speaker/listener roles highlighted in the previous CA findings are reflected in the greater number of all clauses produced by JB compared with the non-aphasic speakers, as can also be seen in the representative example presented here. In line with the non-aphasic speakers' listener role is their higher proportional use of minor clauses (UL 22%, UC 33%, FC 50%) compared with JB (9% with UL, 11% with UC, 9% with FC). See for example, t26 FC "Down the south coast?", with the use of a minor clause allowing minimal interruption while executing the repair. JB tended to use more clause complexes than his partners, which may reflect again the speaker/listener roles of the interactants (JB 62% with UL, 43% with UC, 54% with FC, compared with UL 39%, UC 19%, FC 18%). The average number of clauses per clause complex was similar, ranging between 2.5 to 3 clauses per clause complex. For example, t l 6 FC "And how long is it since you have seen them?", t27 JB "I haven't been there lately, and we went there a lot, you know." These observations reflect the previously reported view from the syndrome approach regarding the relative preservation of syntactic complexity. Since the numbers of clause complexes for the non-aphasic speakers were low, comparisons with JB need to be made cautiously regarding the logical dependency relationships (taxis) found between clauses. However, comparison with other research on aphasic and normal speakers is of interest (Armstrong 1992, 1995; Nesbitt and Plum 1988 - Table 1). Armstrong's (1992) aphasic speaker FP tended to make far more use of paratactic relations than hypotactic, and to make frequent use of extension. Armstrong suggested that FP was using his relative ease of access to these relations strategically, to allow him to clarify his meaning. JB, on the other hand, made more use of hypotactic relations, and made frequent use of elaboration, and this appeared to be in his attempts to clarify his meaning. From a psycholinguistic perspective, such 'over-use' has been seen as a reduction in syntactic complexity (Gleason, Goodglass, Obler, Green, Hyde and Weintraub 1980). Instead, from an SFL perspective, it could be suggested that the logical dependency and logico-semantic relationships provide an important and available resource for these aphasic speakers to make their meanings clearer. In this way, we can see the aphasic speaker attempting to use his relatively spared grammatical resources to tackle his lexical difficulties.

Understanding Paragrammatism:

Contributions from CA and SFL

JB 1 194

FP 2 249

ED* 219

NP 4 2,733

%

%

%

%

parataxis hypotaxis

55 45

89 11

68 32

70 30

projection expansion

82 18

92 8

85 15

84 16

projection'. locution idea

5 95

71 29

19 81

59 41

expansion: elaboration extension enhancement

50 30 20

33 54 13

20 48 32

23 51 27

(# clauses)

277

Table 1: Choices in the network of clause complex relations Could it be that I am simply seeing a glass half full, as it were, rather than the half empty glass depicted through the psycholinguistic perspective? I am suggesting that the findings support the view of paragrammatism as an adaptive response of the impaired semantic system. To begin to test this view, we need to consider what a lexico-grammatical impairment would look like, from a Systemic Functional Linguistic (SFL) perspective. I suggest that such an impairment would be evidenced by lack of access to the full range of semantic options. Evidence for adaptation would be action taken by the speaker subsequent to the incongruent realisation, for example, repetition or revision. In either case, in view of the previous psycholinguistic research showing problems with thematic roles, I suggest that we should see incongruence between Processes and Functional Participant Roles. JB made use of all options from the Transitivity system with regard to Process type. He made most use of material, then in order of frequency, relational, mental, behavioural, existential and verbal processes. The pattern of distribution was similar across conversations with different partners. JB variably achieved congruence between Processes and Functional Participant Roles across all options, without any pattern of error associated with Process type. These 1 2 3 4

JB combined data selected for this Table. FP at 6 mpo data selected for this Table - Armstrong (1992). Armstrong (1992). Nesbitt and Plum (1988).

278

Alison Ferguson

findings support the view that JB's difficulties relate to variable access to the range of semantic options, rather than signalling a particular deficit in the structures or processing for the Transitivity system. These findings are at odds with the psycholinguistic research suggesting specific difficulties with the processing of thematic roles (Shapiro, Gordon, Hack and Killackey 1993). There did appear to be a difference in the distribution of the incongruence of roles with Process options across the conversations with different partners. With the two clinicians, JB's incongruent formulations occurred 21-24% of the time, whereas with the layperson, JB's incongruent formulations occurred 43% of the time. The reasons for this may lie in what was noted in the previous CA as 'scaffolding' by clinicians, where the clinicians appear to take action to steer the discourse toward greater clarity of meaning. Through introducing some CA methodology into the SFL analysis, it was possible to look at actions taken by either speaker subsequent to incongruence (for example, repetition, revision, reformulation by either speaker). JB was observed to take substantial action to clarify incongruence (taking action to clarify incongruence on 69% of opportunities with UL, 64% with UC, 58% with FC). The Unfamiliar Layperson took this opportunity less often than other speakers (UL 6%, UC 24%, FC 21%). For example, see t l 3 where JB follows the incomplete Actor + material Process "He had", with the incomplete Carrier + relational Process "He is", with a further successful reformulation as Carrier + relational Process + Attribute, "He is only 41". At tl4, we can see FC attempting action subsequent to JB's neologism, "Right, so", although she is unable to complete this clarifying action. This part of the analysis identifies just what is going on linguistically at the points where it was previously observed interactionally that UL was 'riding with' considerable non-understanding during the conversation. The findings from the analysis also add to the previous CA findings regarding the considerable repairwork undertaken by speakers with Wernicke's aphasia. The SFL analysis has been able to describe the more subtle linguistic reformulation undertaken in the process of repair. It is this extensive reformulation that I would suggest gives rise to the apparent 'confusion' in the discourse and disrupts the grammatical organisation. That is, the conversational features of Wernicke's aphasia represent the speakers' adaptation to the underlying linguistic impairment involving problems with lexical processing (see Miller and Ellis 1987 for a cognitive neuropsychological account from a connectionist framework which is compatible with SFL notions of selection from a system network of options). I have argued that the findings from the SFL analysis support a view of paragrammatism as a result of a fundamental lexical-semantic impairment. I would argue that error analyses and measures used in previous research from a psycholinguistic perspective (e.g. Edwards and Bastiaanse 1998 - Type Token Ratio), are inadequate to the task of establishing or discounting the presence or

Understanding Paragrammatism:

Contributions from CA and SFL

279

significance of lexical vs syntactic deficit, since the meaning relationships within the clause are not accounted for. That is, I am not suggesting that the syntactic difficulties are insubstantial, however I suggest that such research has not convincingly excluded the lexical deficit hypothesis for paragrammatism. We could also use other aspects of SFL to explore JB's discourse further. However, the analyses presented here allowed us to extend our understanding of Wernicke's aphasia through the contrasts between analyses. Therapy directions based on SFL have been discussed with regard to aphasia by Armstrong (1991, 1993) and with regard to cognitive-communication disability associated with traumatic brain injury by Togher (1997). There is a strong feeling amongst those working in the area to resist the temptation to target treatment goals toward differences considered 'abnormal' as identified through analysis, e.g. for JB, to work on decreasing his use of Extension. Instead, the direction for therapy provided by SFL is far broader than such an approach. From an SFL approach to therapy, the question as to what is or is not 'normal' ceases to be the concern it is when working within a psycholinguistic framework. The question of concern is how the individual is using his available linguistic resources within the context of situation, and can we see ways to increase the availability of resources, or to compensate for their inaccessibility? While this might conceivably involve working on increased use of options observed to be accessible but rarely used, it might just as conceivably involve harnessing resources from other system networks, or for example in JB's case, to see if ways can be found to make the reformulation strategy more efficient and hence less confusing to the listener.

6. Conclusion In this paper I have tried to demonstrate the different ways these various perspectives can shed complementary light on paragrammatism. The SFL analysis presented here provides an interesting theoretical counterpoint to each of the other perspectives discussed. Firstly, this particular SFL analysis shows how it is that a syndrome approach might validly consider the grammatical form as being relatively spared, since we have seen evidence of access to the full range of structural and semantic options. Secondly, the SFL analysis has supported proponents within cognitive-neuropsychology for a lexico-semantic explanation of the grammatical difficulties observed in Wernicke's aphasia. From this author's perspective, these grammatical difficulties arise from the adaptation of the speaker in the attempt to surmount lexical processing difficulties. Thirdly, CA laid bare the bones of how meaning is negotiated between interactants despite the aphasic difficulties, and this pushed the SFL analysis further to explore the revisions and reformulations subsequent to problematic usage, and provides important evidence

280

Alison

Ferguson

to support the view that paragrammatism is the result of adaptation, rather than representing the impairment of Wernicke's aphasia.

References Armstrong, E. M. (1991): The potential of cohesion analysis in the analysis and treatment of aphasic discourse. Clinical Linguistics and Phonetics, 5 (1), 39-51. Armstrong, E. M. (1992): Clause complex relations in aphasic discourse: A longitudinal case study. Journal of Neurolinguistics, 7 (4), 261-275. Armstrong, E. M. (1993): Aphasia rehabilitation: A sociolinguistic perspective. In: A. L. Holland, and M. M. Forbes (eds.) Aphasia treatment: World perspectives. San Diego, CA: Singular Press. Armstrong, E. M. (1995): Clause complexing in aphasic discourse. In: A. Ferguson (Ed.), Proceedings of the Second National Aphasiology Symposium of Australia (Melbourne, Sept. 29-30, 1994). Newcastle, NSW: Department of Linguistics, University ofNewcastle. Blumstein, S. E„ Milberg, W. P., Dworetzky, B., Rosen, A., and Gershberg, F. (1991): Syntactic priming effects in aphasia: An investigation of local syntactic dependencies. Brain and Language, 40, 393^121. Butterworth, B., Panzeri, M., Semenza, C., and Ferreri, T. (1990): Paragrammatism: A longitudinal study of an Italian patient. Language and Cognitive Processes, 5 (2), 115-140. Davis, G. A. (1993): A survey of adult aphasia and related language disorders. Englewood Cliffs, NJ: Prentice-Hall. Edwards, S., and Bastiaanse, R. (1998): Diversity in the lexical and syntactic abilities of fluent aphasic speakers. Aphasiology, 12 (2), 99-117. Ferguson, A. (1994): The influence of aphasia, familiarity and activity on conversational repair. Aphasiology, 8 (2), 143-157. Ferguson, A. (1998): Conversational turn-taking and repair in fluent aphasia. Aphasiology, 12 (11), 1007-1031. Gleason, J. B., Goodglass, H., Obler, L., Green, E., Hyde, M. R., and Weintraub, S. (1980): Narrative strategies of aphasic and normal-speaking subjects. Journal of Speech and Hearing Research, 23 (2), 370-382. Goffman, E. (1981): Forms of talk. Oxford: Blackwell. Goodglass, H., and Kaplan, E. (1983): The assessment of aphasia and related disorders. Philadelphia: Lea and Febiger. Halliday, M. A. K. (1994): An Introduction to Functional Grammar. London: Arnold. Heeschen, C. (1985): Agrammatism versus paragrammatism: A fictitious opposition. In: M. Kean (Ed.), Agrammatism (Ch. 8). Orlando: Academic Press. Heeschen, C., and Kolk, H. (1988): Agrammatism and paragrammatism. Aphasiology, 2 (3/4), 2 9 9 302. Kolk, H., and Heeschen, C. (1992): Agrammatism, paragrammatism and the management of language. Language and Cognitive Processes, 7 (2), 89-129. Marshall, J., Robson, J., Pring, T., and Chiat, S. (1998): Why does monitoring fail in jargon aphasia? Comprehension, judgment, and therapy evidence. Brain and Language, 63 (1), 79-107. Martin, J. R., Matthiessen, C. M. I. M., and Painter, C. (1997): Working with Functional Grammar. London: Arnold. Miller, D., and Ellis, A. W. (1987): Speech and writing errors in "neologistic jargonaphasia": A lexical activation hypothesis. In: M. Coltheart, G. Sartori, and R. Job (eds.), The cognitive neuropsychology of language (Ch. 12, 253-271). London: Lawrence Erlbaum.

Understanding Paragrammatism:

Contributions from CA andSFL

281

Nesbitt, C., and Plum, G. (1988): Probabilities in a systemic-functional grammar: the clause complex in English. In: R. Fawcett, and D. Young (eds.), New developments in Systemic Linguistics (Vol. 2: Theory and applications, 6-38). London: Pinter. Sacks, H., Schegloff, E. A., and Jefferson, G. (1974): A simplest systematic for the organization of turn-taking for conversation. Language, 50, 696-735. Saffian, E. M., Berndt, R. S., and Schwartz, M. F. (1989): The quantitative analysis of agrammatic production: Procedure and data. Brain and Language, 440-479. Saffran, E. M., Schwartz, M. F., and Linebarger, M. C. (1998) Semantic influences on thematic role assignment: Evidence from normals and aphasics. Brain and Language, 62 (2), 255-297. Schegloff, E. A. (1979): Identification and recognition in telephone conversation openings. In: G. Psathas (ed.), Everyday language: Studies in ethnomethodology New York: Irvington. (23-78). Schegloff, E. A. (1987): Recycled turn beginnings: A precise repair mechanism in conversation's turn-taking organization. In: G. Button, and J. R. E. Lee (eds.), Talk and social organisation Clevedon: Multilingual Matters. (70-85). Schegloff, E. A. (1992): Repair after next turn: The last structurally provided defense of intersubjectivity in conversation. The American Journal of Sociology. 97 (5), 1295-1345. Schegloff, E. A., Jefferson, G., and Sacks, H. (1977): The preference for self correction in the organization of repair in conversation. Language, 53, 361-382. Schienberg, S., and Holland, A. (1980): Conversational turn-taking in Wernicke Aphasia. In: R. H. Brookshire (Ed.), Clinical aphasiology (106-110). Minneapolis, MN: BRK Publisher. Schwartz, M. F. (1987): Patterns of speech production deficit within and across aphasia syndromes: Application of a psycholinguistic model. In: M. Coltheart, G. Sartori, and R. Job (eds.), The cognitive neuropsychology of language London: Lawrence Erlbaum. (Ch.8, 163-199). Shapiro, L. P., Gordon, B., Hack, N., and Killacky, J. (1993): Verb-argument structure processing in complex sentences in Broca's and Wernicke's aphasia. Brain and Language, 45, 423^147. Silvast, M. (1990a): Aphasia therapy dialogues. Paper presented at International Aphasia Rehabilitation Congress, Edinburgh, September. Silvast, M. (1990b): Conversation analysis and the study of interaction in aphasia. In: M. Leiwo, and R. Aulanko (eds.) Studies in logopedics and phonetics 1, Helsinki: Department of Phonetics, University of Helsinki. (Series B: Phonetics, Logopedics and Speech Communication 2, 199— 210). Sobiecka-Koszel, G. (1991): Communication strategies: The therapist as the dominant partner. Aphasiology, 5 (2), 197-199. Togher, L. (1997): Operationalizing discourse therapy. Aphasiology, 11 (6), 621-625. Zurif, E., Swinney, D., Prather, P., Solomon, J., and Bushell, C. (1993): An on-line analysis of syntactic processing in Broca's and Wernicke's aphasia. Brain and Language, 45, 448^164.

Acknowledgements The author would like to thank Lynne Mortensen for the original inspiration for the direction of this research, as well as for her assistance with the analysis.

Branca Telles Ribeiro Listening to Narratives in Psychiatric Interviews1 1. Introduction A psychiatric interview is a difficult speech event. It is troublesome for the psychiatrist, who must follow a strict institutional agenda so as to probe in a principled way into the patient's problems. It is troublesome for the patient, who expects a listener to her many stories and is frequently interrupted by the doctor's requests. Things get even more complicated when the patient's talk gets progressively "off the track." This study analyzes four psychiatric interviews between Dr. Mauro Cardozo, a clinical psychiatrist and the same patient, Rejane Rodrigues, a 31 year old woman with manic depression).2 For the most part, the talk is quite difficult to follow as the patient is in a manic episode. She often speaks too fast while bringing up a series of unrelated references; she also introduces idiosyncratic topics, making the listener's task very demanding, as we can see in the following interaction:3 Doctor: Patient: Doctor: Patient: Doctor: Patient:

1

2

3

and why did you come here? I came here 'cause God brought me here to my house. the house my mother Anna Rodrigues built on the rocks, from David's throne. your mother built you a house on the rocks? that's right mmm mmm. but whatwas alive

Many thanks to the narrative seminar group at Harvard University for insightful discussions, specifically Elliot Mishler, Jane Liebschutz, Trudy Duffy and Carmela Perez. Our sincere thanks to Susan Hoyle for her keen editorial comments. Extended appreciation to Sandra Jones, Lauren Storck, and Jacqueline Scarbrough for encouraging discussions on earlier drafts. We are also grateful to Drs. Jo5o Ferreira da Silva Filho and Romildo Bueno for supporting the research project on Context and Coherence in Psychotic Discourse at the Institute of Psiquiatry of the Federal University of Rio de Janeiro. Grants from CNPq — the Brazilian National Research Council — and FAPERJ — the Rio de Janeiro State Research Foundation — sponsor the project. These interviews were videotaped within a two month period, during the patient's stay at the hospital. We are indebted to the doctor and the patient for their time and availability. All names have been changed to protect the confidentiality of the participants. Both participants are Brazilian; thus, the interviews were conducted in Portuguese. Disturbances in the use of language are displayed in several psychiatric conditions, such as mania, severe depression, organic disorders, and schizophrenia.

Branca Telles Ribeiro

284 Doctor: Patient:

you were alive? yeah LIVED (shouting), and speaking to everyone, here I am Lord. come into the presence, the Holy Ghost will concede it all (inside now).

In spite of such disjunctive talk, there are moments of coordinated conversation between the doctor and the patient. In these moments, bits of a story can be pieced together, thus providing the doctor with crucial clues to the patient's past and present illness, family and social history, and chief complaint. Listening to and understanding such stories is, therefore, essential to the psychiatrist, who must make an assessment of thought, language, and communication disorders. This paper focuses on a specific narrative - the Bozo story - which seems to accomplish at least three functions in these interviews: 1) it creates textual cohesion in a discourse often filled with unrelated topics; 2) it provides a referent - Bozo, the clown - which seems to clarify contradictory utterances and interactional moves, providing the listener with important clues as to the patient's performance and identity; 3) it integrates different interactive contexts which would otherwise conflict, thus creating some conversational coherence. We will see how narrative analysis works in describing the patient's complex layering of reference. In these four interviews, narrative analysis turned out to be the most useful tool for displaying consistency of information in an otherwise incoherent discourse. 4

2. The Unfolding of Stories in the Interview Can stories emerge within a psychiatric interview? We often hear that "informants' stories do not mirror a world "out there." They are constructed, creatively authored, rhetorical, replete with assumptions, and interpretive" (Riessman 1993). A doctor must then be willing to be a competent listener: someone who hears, listens and understands. He/she must also be willing to listen to stories as metaphors for embedded meaning. Listening to stories implies a double footing on the doctor's part: he/she is not only the "inquirer" portrayed by Goffman (1981: 143), the active professional pursuing information with a prior institutional topic agenda; he or she is also a 4

Stories have been the object of investigation in literary studies, semiotics, conversational and discourse analysis. In sociolinguistics, narratives have been widely researched following Labov's definition of a story (Labov 1972, 1997), Tannen's involvement strategies and frame analysis (1984, 1986, 1989, 1993), Linde's life stories (1993), among others. Mishler's work (1986, 1992, 1999) indicates how prominent narratives are in research interviews as well as in clinical discourse. Specifically medical and psychiatric interviews are of interest to this work. Most recently narrative analysis has been proposed as a new paradigm for discourse investigation (Brockmeier and Harré 1997).

Listening to Narratives in Psychiatric

Interviews

285

receptive and available person, placed in a secondary position in the speech exchange situation, that is, the listener to a story.5 Several stories come up in these interviews. They range from stories about the patient's previous traumas, prior institutionalizations, moments of depression, suicide attempts, and mostly unhappy love stories. The Bozo narrative is the most prominent love story. It is also highly successful in getting the doctor's attention. A sociolinguistic narrative analysis of the four interviews, following Labov (1972, 1997), reveals textual cohesion in an otherwise topically unrelated discourse. The patient produces some pieces of the Bozo story in the first interview, but she restates and expands on these pieces in the subsequent interviews. If all the interviews are analyzed as contributing to a single narrative, consistent information emerges out of apparent incoherence. 2.1. The Bozo Narrative: A Cohesive Narrative Structure As with any narrative, let's start from the beginning. Rejane Rodrigues, the patient, first introduces and summarizes in a nutshell the story, preparing the audience for more to follow. This information is presented in the abstract sections (1st, 3rd, and 4th interviews): abstract:

"I was married to Bozo" "Bozo the clown."

This summary gets expanded in the 3rd and 4th interviews: abstract: abstract:

"my problem was being in love." "but I love one man in my life, he's Beto, he's the Bozo."

We also find more of the same in the orientation parts, where the patient affirms and restates the following: Beto is the Bozo, he has been named "Bozo the clown" by people at his work place; he works as a physical therapist in a hospital where Rejane had been institutionalized before; at this hospital, in Valen9a, Jacarepagua, he took care of her and healed her injured foot, and also at this hospital they "got married" at some June festivity. Facts and fiction get blurred by this wedding reference: "we married on that day," "on the day of the Caipira." 6 This ambiguity is left unresolved. Except for the 2nd interview, each encounter has at least one orientation section where the above information is reiterated. 5

6

Briefly stated, a footing is how participants position or align themselves towards each other and towards what they say or do in interaction. Thus "the projected self is somehow at issue" and so is "the production or reception of an utterance" (Gofiman 1981: 128). Footings capture the ever changing nature of social roles in discourse and interaction. The Caipira costume party takes place at Saint John's festivities held in June. Besides children's games, adolescents frequently participate in a make-believe wedding ceremony, when couples "get married" and witnesses clap and sing.

Branca Telles Ribeiro

286

Particularly valuable information is conveyed in the evaluation sections. They indicate the point of the story and address the "so what" question (Labov 1972, 1997). Taking into consideration not only the semantic meaning but also the pragmatic value of propositions in the evaluation sections, we see Rejane displaying ambivalence toward her lover. She shifts from a series of complaints to a series of positive evaluations. Beto/Bozo is thus someone who is a "pest" or an "ass," who "did not come and followed [her]," but he is also someone who is "quite sexy and attractive," someone to whom she says "I love you," and who is also "a Christian in God." The referring form "the clown" may well capture this ambivalence since it evokes both negative attributes (a rude, awkward, illmannered person; an incompetent or silly person) and positive traits (a playful, funny, bright person). We will see later how this form may be quite appropriate to the development and performance of this narrative. In the complicating action - the core of a narrative - Rejane informs us that she "has been possessed by some Beto" and that they "have been fooling around." The notion of "possession" merits further discussion, since it triggers the issue of irrationality, present both in mental illness as well as in a love affair. A person "in love" is frequently described as irrational and may sometimes even display a disruption of social behavior. The notion of "possession" also implies a "lack of control" which may indicate that someone else has control over her, someone like Bozo. It may also mean that she has relinquished control over to Bozo. A Bozo story is embedded in the 2nd interview, when Rejane refers to a different boyfriend who had sex with her at a very young age: Complicating Action "[Bozo] had sex with me when I was sweet thirteen, when I was still a virgin." "he had sex with me under the couch, behind the couch, behind doors (my mother running after us) while we were making the most of it." "the one who really got me was Eric."

This embedded story stands by itself. The juxtaposition of Eric as the Bozo indicates that the Bozo identity may stand for other love or sexual relationships, not only for the Beto love narrative. In the 4th interview, Rejane recapitulates and expands on the Beto/Bozo story: Complicating action:

an interesting (detail) he would run away from me, I would run away from him, everything was video-taped, and later I and he stood by it (to run away), he tried everything to run away with me

Listening to Narratives in Psychiatric

Interviews

287

but 11 couldn't see well I used glasses. Complicating action:

to run away, the garage door was open. he was going to take me away on that day (laughs) but I couldn't see.

Prior information is restated once more: at the party, Beto and Rejane played tricks on one another, then they made a commitment to run away together. Again there is consistency in the repetition of descriptive actions which shift from "play" to "commitment". New information is introduced when the narrator tells us that Beto really tried hard and the door was open, but unfortunately she couldn't find her way out, she couldn't see. While external circumstances were favorable, internal difficulties did not forward the action. Finally the narrative wraps up in the last two interviews with a coda section: 3rd Interview coda:

4th Interview: coda:

"I met him, dated him, married him. it did not work."

"then I was discharged." "he told me he would look for me, he did not say anylonger 'I love you' to me. he did not look for me." "it has been a year."

From the above discussion, we see how structurally cohesive the text is.7 2.2. The Name Bozo: Clarifying Paradoxical Meaning The name Bozo, whenever it occurs, is a relevant contextual ization cue (Gumperz 1982) within the story schema - for Beto, the Bozo - as well as a cue for the ambiguous layering of personal footings proposed by the patient at a given moment. Let's first discuss Rejane as the Bozo. At different moments, she refers to herself as a clown ("I am a clown"). At other times she performs as a clown and she is aware of it: "I behave like a clown." She teases the doctor or the video audience as in this indirect sexual reference: Doctor: Patient:

7

but this is a hospital here, right? for those that like to get shots, yeah.

There are also problems which a narrative analysis indicates, such as the lack of details in the action sequence, contradictory information in the orientation sections, and inadequate use of deixis. These points merit further investigation; however they are beyond the scope of this paper.

Branca Telles Ribeiro

288 Doctor: Patient: Doctor: Patient: Doctor: Patient: Doctor:

do you like to get shots? only from Bozo. only from whom?

(

)

from whom you like to get shots? from Bozo. from Bozo? (laughs) really?

A "shot" or "injection" has the explicit meaning of a medical liquid injected in the body. It also has an implicit sexual connotation. The interpretation and effect on the audience (laughter) derives from its second meaning, not the first. Here Rejane alternates between "sexy" (erotic) and "clownish" (playful) footings with the doctor and with her audience (potential viewers of the videotape). These play interactional contexts allow for ambiguity and denial of intentionality. Second, the name Bozo is used in the Beto/Rejane love story. While she states that Beto is her husband, she tells us that he is also a clown. The trickster identity reinvokes ambiguity which is not resolved as she says "we married in mystery." There is added uncertainty to this wedding as it took place in the hospital during a playful ceremony. Bozo has also metaphorical connotations which fit any attractive man with whom the patient was previously involved (such as Eric) or may be involved with (projected fantasies with the doctor). There is, therefore, the doctor as the Bozo because Dr. Cardozo fits most of Bozo's attributes: Doctor: Patient: Doctor: Patient: Doctor: Patient: Doctor: Patient:

so you really liked being here? oh- there's nothing better. really? I see you. I see handsome men, sexy, looking at me, filming me. oh really? (smiling) then I feel- (laughs) so you like this, (laughs) oh I like it.

Finally, Bozo is also an indirect reference to the emotional state of euphoria (laughter and fun). This state is particularly interesting since the patient has a manic depressive disorder and one of the symptoms of mania is euphoria. The clown brings about the perlocutionary effect of making the audience laugh (Austin 1962). As we have seen in several excepts, Rejane often makes the doctor laugh. She also makes us, the audience, laugh a great deal. However, on the reverse side, the Bozo metaphor encapsulates the tragic clown. Depression, sadness and suicide attempts are issues brought up by both participants in these interviews.

Listening to Narratives in Psychiatric

2.3.

Interviews

289

Multiple Contextual Embeddings and Story Bits: What is This All About?

Let's now discuss briefly the interactional contexts which frame the enactment of the story: the inquiry footings and the personal play footings. First, we should examine the presuppositions which underlie the psychiatric interview. In the inquiry, we see the doctor requesting further information so as to double-check for consistency of content (topic development) and reliability of information. Often the patient's talk displays loose associations (a communicative disorder known as derailment) and an eagerness to go on with a new idea (pressure of speech): Doctor: Patient: Doctor:

do you believe you're thinking too fast, Rejane? that's what God gave me, an empty mind. an empty mind, but I want to know if [you're thinking] too fast, too quick.

We see the patient somewhat providing for topic ratification and development (as a response for "thinking too fast," she says she has "an empty mind"); but there may be abrupt shifts of reference without an explicit topic association: Doctor: Patient: Doctor: Patient:

he is the Bozo? we have been fooling around, but we're on a honeymoon, after 90 days things go away ( ) mmm mmm.= international.

In the inquiry, the doctor also investigates which aspects of the story can be credible: Who's Beto? Who's the Bozo? Dr. Cardozo asks information and clarification questions which test for content as well as for communication skills (topic maintenance and topic continuity). The point is that truthfulness matters in the inquiry frame of the psychiatric interview.8 The question "what is real?" matters given the institutional contexts under which this narrative is told: a psychiatric hospital with a patient in a manic episode. The doctor's clarification and confirmation requests indicate uncertain and doubtful footings: "Bozo [is here]?", "ah really? but Bozo is your husband?", "and he is also here?". A comparison of the Bozo narrative as it unfolds in the four interviews indicates the doctor's effort to get reliable information. On the other hand, in each new encounter, there are shifts towards the use of more personal footings (which entail different presuppositions about the doctorpatient encounter). The doctor progressively shifts footings from a distant, 8

We understand "truth" and "credibility" necessarily related to the "plausibility of a chain of objective events" (Ochs and Capps 1997).

Branca Telles Ribeiro

290

unrelated alignment (1st interview) to a display of attention and curiosity (2nd interview) and finally to repetitively pursuing information (3rd interview) on who this Bozo character/persona could be and what he means/meant to the patient ("you talked a lot about Bozo," "but finally, who is Bozo?"). This change in footing and listernership enabled the unfolding of a detailed narrative in the last interview with beginning, middle and end. The development and conclusion of this narrative also indicates a way of balancing and wrapping things up. Though the doctor (and the audience) is left uncertain as to what exactly happened between Rejane and Beto - the two main protagonists - secrecy seems to be a major ingredient in keeping the doctor (and the audience) involved with the narrative. By playing on ambiguity, the patient successfully keeps her interlocutor "hooked", accomplishing what she initially set out in the 1st interview when she told him "turn yourself on" and Dr. Cardozo responded "I'm turned on." Rather than truth, involvement matters for the successful development of this narrative. It is how participants (the doctor and the patient) co-constructed and negotiated a story which favored a complete narrative in the 4th interview. Rather than the pursuit of information - typical of the inquiry - in the story and play footings involvement is of essence. In these contexts, what seems to be at stake for the patient is getting the attention and listernership of the doctor. To this end, this story is highly successful.9

3. Conclusion While topic analysis may point to different psycho-pathological problems (such as incoherence, tangentiality, loss of goal) in a patient's discourse (Ribeiro 1994, Chaika 1990), a narrative analysis investigates ways in which a story is told in bits and pieces and through multiple contextual embeddings. Throughout these four interviews, the telling and retellings of a story worked to accomplish some textual cohesion and conversational coherence in an otherwise complex and incoherent discourse. We saw that a narrative may unfold at different levels of contextual embedding. In the case of the Bozo story, we may have two or three different stories taking place at the same time, each one embedded within the other without necessarily having a subordinate relationship to each other: there is the Beto love story, also the lover Eric as the Bozo, and even the patient's projected fantasies with a playful doctor. Piecing these bits of information together may indicate to the psychiatrist relevant information about his patient's social and family history, present illness and chief complaint.

9

For an extended discussion on discourse and involvement strategies see Tannen (1989).

Listening to Narratives in Psychiatric Interviews

291

The doctor's double footings - as an inquirer and as a listener to a story - are crucial for the successful development of the psychiatric interview. While credibility and truthfulness are relevant for the inquiry, involvement allows narratives, whether life or health stories, to unfold. Listening to narratives may therefore be a tool for psychiatrists to gather relevant information on their patients. It may also have therapeutic effects: at the end of these interviews, Rejane thanks the doctor (and the audience) for the opportunity to speak and be heard. Counteracting her sense of isolation ("no one to talk to, except to God"), being summoned to an interview can be viewed as an invitation to talk, to chat, and to interact. At the very end of the 4th interview, Rejane says "this [talk] relieved me." There is a feeling of gratitude for being listened to.

References Austin, J. (1962): How to Do Things with Words. Cambridge, MA: Harvard University Press. Brockmeier, J. and R. Harre (1997): Narrative: Problems and Promises of an Alternative Paradigm. Research on Language and social Interaction, 30(4), 263-283. Chaika, E. (1990): Understanding Psychotic Speech: beyond Freud and Chomsky. Springfield, 111.: Charles C. Thomas. Goffman, E. (1981): Footing. In: Forms of Talk. Philadelphia: University of Pennsylvania Press, 124-59. Gumperz, J. J. (1982): Discourse Strategies. Cambridge: Cambridge University Press. Labov, W. (1997): Some Further Steps in Narrative Analysis. In: Journal of Narrative and Life History 7 (1-4). Lawrence Erlbaum Ass., 395-415. - (1972): The Transformation of Experience in Narrative Syntax. In: Language in the Inner City, 354-96. Philadelphia: University of Pennsylvania Press. Linde, C. (1993): Life Stories. Oxford and NY: Oxford University Press. Mishler, E. G. (1992): Work, Identity, and Narrative: An Artist-Craftsman's Story. In G. C. Rosenwald and R. L. Ochberg (eds.), Storied Lives. New Haven: Yale Univ. Press. Mishler, E. G. (1999): Storylines. Craftartists' Narratives of Identity. Cambridge: Harvard University Press. Mishler, E. G. [1986] (1990): Research Interviewing: Context and Narrative. Cambridge, MA: Harvard University Press. Ochs, E. and L. Capps. (1997): Narrative Authenticity. In: Journal of Narrative and Life History, 7(1-4), 83-89. Lawrence Erlbaum. Paz, 0 . (1996): The Double Flame: Essays on Love and Eroticism. London: Harvill. Ribeiro, B. T. (1994): Coherence in Psychotic Discourse. NY: Oxford University Press. Ribeiro, B. T. (1996): Conflict Talk in a Psychiatric Discharge Interview: Struggling between personal and official footings. In: C. R. Caldas-Coulthard and M. Coulthard (eds.) Texts and Practices: Readings in Critical Discourse Analysis. London: Routledge, 179-93. Riessman, C. K. (1993): Narrative Analysis London: Sage. Schiffrin, D. (1996): Narrative as self-portrait: Sociolinguistic Constructions of Identity. In: Language in Society. 25:2,167-203. Tannen, D. (1984): Conversational Style. Norwood, NJ: Ablex. Tannen, D. (1986): That's not what I meant!. NY: Ballantines. Tannen, D. (1989): Talking Voices. Cambridge: Cambridge University Press. Tannen, D. (1993): Framing in Discourse. NY: Oxford U. Press.

292

Branca Telles Ribeiro

Appendix: Transcription Conventions sentence-final falling intonation phrase-final intonation, indicating more talk to come sentence-final rising intonation abrupt cutting off of sound (words) uncertain transcription ( ) transcription impossible [words] translation (adjustments made from the original Portuguese text) = latching (utterance linked by = indicate no break in flow of talk) overlapping speech: two people talking at the same time underline emphatic stress CAPS very emphatic stress, loudness or shouting (pause) noticeable break in rhythm with no speech; length of time was not assigned , ?

Klaus P. Schneider

Diminutives in Discourse: Sequential Aspects of Diminutive Use in Spoken Interaction 1. Introduction Although diminutives have been studied extensively during the past 150 years or so, pragmatic analyses are rare. These are all speech-act based, focusing almost exclusively on micro-pragmatic aspects. By contrast, the present paper adopts a discourse perspective, the communicative functions of diminutive forms beyond the speech act, considering adjacency pairs, interactive exchanges, and larger discourse units. Special emphasis will be placed on re-initiations, re-runs, and follow-ups, and on specific negotiation patterns. The approach adopted in this paper has implications for the general study of grammar in spoken interaction.

2. What are Diminutives? Prototypical diminutives are derived from nouns by suffixation (e.g., Span, casa > casita; Russ. dorn > domik; G. Haus > Häuschen). But diminutives are not a morphological category. Their distinctive feature is their meaning. Diminutives denote smallness AND express attitude which may be positive or negative, depending on the specific interplay of linguistic AND situational factors in a given context.1 This meaning is expressed by various formal means in different languages depending on their respective typological make-up. I distinguish between three types of diminutive forms: 1) synthetic diminutives formed by morphological processes, 2) analytical diminutives formed by syntactic processes, and 3) inherent diminutives which are semantically, but not formally related to other items in the lexical system of a given language. Contrary to popular belief, all three types exist in English. Analytical diminutives formed by little + noun are generally considered the most important type and the language-specific functional equivalent to prototypical diminutive formation in languages such as Russian, Polish, Spanish or Italian.2 But English also has the synthetic type. In the literature, more than eighty formatives of English are treated as diminutive suffixes.3 Among these, {IE} and {S} are 1 2 3

For details see Schneider (forthcoming: ch. 2.1. and 8.5.). Cf., e.g., Strang (1968: 138); Dressler/Merlini Barbaresi (1994: 114); Schneider (forthcoming: ch. 6.3.). See Schneider (forthcoming: ch. 6.1.2.).

Klaus P. Schneider

294

particularly productive; cf. doggie, girlie, footie; fatty/fats, Debbie/Debs etc.4 However, the synthetic type will not be discussed in the present paper.5 Examples of inherent diminutives, which play an important role in the following analysis, are sip, drop, and nip, which contrast with draught (while gulp or swig may be considered inherent augmentatives).

3. Diminutives in Macro-Pragmatic Perspective The following analysis focuses on diminutives in English offer sequences. It comes in three parts. First, offer sequences are characterised as negotiating discourse and involving relatively complex facework. Second, diminutive use in offer sequences is discussed on the basis of examples from my corpus. And third, diminutive use in other sequences is examined to find out whether generalisations across sequence types can be made about macro-pragmatic aspects of diminutive use. 3.1. OFFER Sequences Offer sequences minimally consist of two moves, offer and offer response. These two moves form a simple exchange involving two interactants, the benefactor and the beneficiary. Offers are essentially commissive speech acts. Speakers commit themselves to a future action, namely to do something at their own cost and for the benefit of their co-conversationalists. By doing this, they put their positive face at risk. Therefore offer acceptance is preferred, while refusals are dispreferred. At the same time, accepting an offer could be regarded as violating the maxim 'Minimize benefit to self (Leech's GENEROSITY maxim; Leech 1983: 132). Beneficiaries may appear greedy if they accept at once. Therefore many cultures have developed polite norms according to which an offer must never be accepted in the first reaction to it. Beneficiaries are expected to decline the offer initially, while benefactors are expected to renew their offer thereafter. It is only then that beneficiaries may accept, if they wish. In Ireland, for example, offer sequences minimally consist of four moves, not just two: (I)

4 5

Move Move Move Move

1: 2: 3: 4:

offer ritual refusal offer renewal (re-initiation) offer acceptance or refusal

e.g. A: Want a cup of tea? B: No, thanks. A: Are you sure? B: Okay, I'll have one. (fabricated)

Cf., e.g., Muhlhausler (1983) and Poynton (1989). For a discussion of {IE}, {S}, and twelve other English diminutive suffixes see Schneider (forthcoming: ch. 6.2).

Diminutives in Discourse: Sequential Aspects of Diminutive Use in Spoken Interaction

295

The second move is called 'ritual refusal', because the polite norm requires a refusal in this position even if the beneficiary intends to accept the offer. This may cause serious problems in cross-cultural communication. The German norm, e.g., differs from the Irish norm. In Germany, beneficiaries are supposed to accept an offer immediately, they will not be given a second chance. Many Irish students after their year abroad in Germany report on this phenomenon (and on how they suffered).6 Offers of food and drink of the type discussed under (1) (Goffman 1971: 65) calls them 'ritual offerings') typically occur in party contexts or at social gatherings of a similar kind. In such contexts, offers characteristically occur more than once. Hosts normally follow the maxim 'Maximize benefit to other' (Leech's TACT maxim), at the same time they observe the complementary maxim 'Maximize cost to self (Leech's GENEROSITY maxim; Leech 1983: 132); they do not wish to appear stingy. By contrast, guests, as mentioned above, do not wish to appear greedy. They should be, and normally are, aware of the cost to the benefactor and act in accordance with the maxims 'Minimize cost to other' and 'Minimize benefit to self (which are alternative formulations of the TACT maxim and the GENEROSITY maxim respectively). However, in accepting an invitation to a party guests also accept to be catered for and entertained by their hosts. Therefore guests are expected to acknowledge the trouble their hosts have gone to in preparing food and providing drink and show appreciation. This makes it difficult to refuse an offer. These conflicting norms and expectations have to be balanced out with each new offer in the course of the respective event. But with each new offer the danger of appearing greedy increases. Terminologically, I differentiate between initial offers and subsequent offers. Initial offers are not preceded by any other offer, they open the first offer sequence in the course of a given event. By contrast, subsequent offers open offer sequences that occur later in that same event, they are external to the first offer sequence. All offer sequences in the same discourse are disjunct, which means that stretches of discourse devoted to other topics and serving different purposes occur between them. (I do not consider here offer sequences in subsequent speech events involving the same participants, although this may also be an interesting area for investigation, bearing in mind that invitations and social favours create obligations. To my knowledge, this whole area has not been researched in pragmatics.) Inside offer sequences, I distinguish between initiative offers and offer renewals. An initiative offer is the first move in each offer sequence. Offer renewals appear as third or later moves. Offer renewals in turn 5 and thereafter can be characterized as pressing. Pressing is a discourse strategy which is 6

For details see Barron (in prep.).

296

Klaus P. Schneider

expected in some cultures, while it is frowned upon or regarded as impolite in other cultures. Offer responses can be positive or negative. The former are called offer acceptance, the latter offer refusals. Offer refusals in second turns of a sequence can be ritual refusals, as mentioned above, whereas in later turns they can be considered genuine (at least in English speaking communities). 3.2. Diminutives in OFFER Sequences Diminutives (DIMs) are used in all types of offers, i.e. in initial offers and subsequent offers, also in initiative offers and offer renewals. In offer responses, however, DIMs seem to occur in offer acceptance only; there is no need for them in refusals. The discourse functions of DIMs can be derived from the face work involved in offer sequences as described above. Let us now consider some examples from my corpus. My material consists of natural and of fictional discourse. The latter is taken from drama, more particularly from conversation pieces, but examples from dialogue in prose are also included. In offer sequences, DIMs seem to be used most frequently in offer acceptance. The following examples illustrate this use: (2)

(3)

N: Honey? What would you like? H: I don't know, dear... A little brandy maybe. A: Have a drink. Scotch or Bourbon? B: A little drop of Scotch would be welcome.

(ALB 21)

(COB, s.v. little 5.)

In using a diminutive, the beneficiaries observe the maxims 'Minimize benefit to self and 'Minimize cost to other'. They play down the amount of the offered good. DIMs are also used in offers, as can be seen from the following examples: (4) (5)

Would you like a little cigarette? (Offering cakes) Have a little one.

(LEI 21) (Köhler 1994: 29)

On the surface, these offers violate the relevant politeness maxims 'Maximize benefit to other' and 'Maximize cost to self, since the diminutive forms refer to the goods that are offered. The use of these forms, however, can be explained in much the same way as DIM use in offer acceptance. By diminishing cost and benefit, the benefactors reduce the social risk involved for the beneficiaries and make it easier for them to accept. In other words, the benefactors adopt their addressees' perspective, anticipating their polite refusals.

Diminutives in Discourse: Sequential Aspects of Diminutive Use in Spoken Interaction

297

DIMs are more frequently used in offers opening subsequent sequences in which refusals are more likely to occur. While examples (4) and (5) are initial offers, (6) and (7) are subsequent offers, as can be seen from the use of the expressions another and top-up: (6) (7)

Would you like another drop? Give us your glass. I'll give you a little top-up.

(Kohler 1994: 29) (LEI 57)

If a DIM is used in the offer, no DIM can be used in acceptance. And vice versa, a DIM will only be used in acceptance, if no DIM is used in the offer. Consider the following examples: (8)

(9)

N: Do you want a ...? H: Yes dear... a little more brandy, maybe. Just a drop. A: More cheese? B: Just a smidgin, please.

(ALB 34)

(DCE2, s.v.

smidgin/smidgen)

Dressier and Merlini Barbaresi (1994: ch. 3.5.) claim that each speech act may contain one DIM only. This is a micro-pragmatic statement. In the light of my data, I suggest that each adjacency pair consisting of offer and offer response may contain one DIM only, either in the offer or in offer acceptance, but not in both. The examples under (4), (5), (6), and (7) are all initiative offers (or sequence openers). Let us now consider offer renewals: (10) B:... Now, who's for another drink? ... How about you Sue? S: No, thank you. B: Yeah, come on, Sue, give you a little top-up, just to settle your nerves. That's it. (LEI 32)

The following dialogue involves three speakers. L and B are host and hostess, T is a guest: (11)

L: Cigar? T: No, thank you. I've just given up. L: Are you sure? B: Yeah, go on, Tone, take a little cigar, enjoy yourself go on, take one. L: Yes, go on, take one. T: Thank you. [He takes one.]

(LEI 38)

In both cases, the initial offer, which does not contain a DIM, is refused, and the renewed offer, which does contain a DIM, is accepted. While L in (11) reacts to the refusal by using the conventional question routine Are you sure?, B, his wife, is more pushy. Apart from the DIM, she uses expressions characteristic of the strategy I have termed pressing (go on; cf. also come on in ex. 10) and a

Klaus P. Schneider

298

supportive move (enjoy yourself, cf. also just to settle your nerves in ex. 10) as component parts in her persuasive strategy realising the offer renewal. A more exaggerated example of this strategic use of a DIM in offer renewal is found in the following dialogue from "Father Ted", a TV comedy programme set in Ireland. In this excerpt, Mrs Doyle, the housekeeper, offers Father Ted, the priest, a piece of cake7: (12)

MD: FT: MD: FT: MD: FT:

And speaking of cake, I have cake. I'm fine for cake, Mrs Doyle. Are you sure, Father? (Pause) ... ... (Pause) I won't, thanks. Ah, you will. Sure it's only small, it's just a tiny little cake, you won't even feel it going in, you won't even know you 're eating it. No, ... (FAT)

Similar examples can be found in almost every episode of this series. Mrs Doyle usually uses many go ons, but she rarely wins the battle. Father Ted is in a different position than guests usually are; after all, he is Mrs Doyle's employer. It is worth noting that in examples (10), (11), and (12) no DIM is used before the offer renewal containing the DIM. This suggests that, in addition to what I said earlier, whole offer sequences, even long ones, may contain no more than one DIM. I have, however, one example in my data in which DIMs are used in moves 1 and 7, the sequence-intial offer and the third renewal. This may indicate that adjacency and length of sequence are crucial factors determining the use of DIMs in offer sequences. Further research is needed here.

3.3. Diminutives in Other Sequences The functions of DIMs in other sequences are similar to the functions of DIMs in offer sequences. However, this seems to apply only to speech acts related to offers and offer acceptance regarding their communicative purpose. In other illocutionary types, DIMs seem to have different functions.8 Since offer acceptance can be classified as reactive requests, it is not surprising that DIMs are also used as cost/benefit minimizers in initiative requests: (13) (14)

Could I have a little brandy? (ALB 76) Ijust want to ask you if I can borrow a tiny bit

ofpaper. (Brown and Levinson 1987: 177)

7 8

I am grateful to Anne Barron for this example. See Schneider (forthcoming: ch. 7).

Diminutives in Discourse: Sequential Aspects of Diminutive Use in Spoken Interaction

299

Suggestions are similar to both requests and offers in that they predicate a future action of the S and the H. The following examples illustrate the use of DIMs in initiative suggestions: (15) (16)

Fancy a little dance, Tone? Why don't we have a little game?

(LEI 58) (OSB 15)

In case an initiative suggestion is not accepted, a DIM can be used in the reinitiation of this speech act, as a persuasive strategy: (17)

'Shall we have a kiss?' she said, after a while, stroking my face lightly. I panicked. 'What?' 'Just a little kiss to start with, to see how we get along. ...'

(KUR 201)

In this particular function, DIMs are also employed in reformulations. Reformulations are turn-internal re-initiations using the same realisation format as the original formulation in a modified form. The following example is a request realised by an imperative (a realisation type termed 'mood derivable' by BlumKulka et al. 1989 in their taxonomy of realisation types for requests): (18)

You're a scientist, aren't you? C'mon ... make an experiment... make a little experiment. (ALB 98)

In the pause following the original formulation the speaker notices that the addressee hesitates. So she uses that same formulation again, this time using the DIM form. The opposite case, I propose, is not conceivable. It is unlikely that a DIM will appear in the original formulation and then be dropped in the reformulation.

4. Conclusion The discussion can be summarised as follows. DIMs occur in offers, suggestions, and requests; they can be used in initiations, re-initiations, and reformulations of these speech acts. In these cases, the speaker employs DIMs in a persuasive strategy to negotiate a positive outcome, i.e. acceptance in the case of offers and suggestions, and compliance in the case of requests. The use of DIMs in response to the initiative speech act seems to be particular to offer acceptance. All sequence types seem to contain one diminutive form only. Further research is needed to shed more light on the use of diminutives in other speech act sequences, adopting the approach advocated in this paper.9 Other phenomena related to diminutives should also be studied from a macro-pragmatic perspective. These include forms traditionally classified as augmentative, 9

For preliminary results see Schneider (forthcoming: ch. 7).

Klaus P. Schneider

300

excessives, and elatives, and the area of speech act modifiers known as upgraders and downgraders (cf., e.g., Bazzanella et al.1991; Blum-Kulka et al. 1989). Finally, I suggest that the macro-pragmatic approach can be used for contrastive investigations and to serve as a frame of reference in studies of crosscultural communication, and of cross-cultural misunderstandings in particular.

References Barron, A. (in prep): A longitudinal study of the effect of the year abroad on the pragmatic competence of second language learners. Bazzanella, C., Caffi, C. and Sbisa, M. (1991): 'Scalar Dimensions of Illocutionary Force'. In: Zagar, I. (ed.) Speech Acts: Fiction or Reality? Ljubljana: IPrA Distribution Centre for Yugoslavia at the Institute for Social Sciences, 63-76. Blum-Kulka, S., House, J. and Kasper, G. (1989): 'Investigating Cross-Cultural Pragmatics: An Introductory Overview'. In: Blum-Kulka et al. (eds.) Cross-Cultural Pragmatics: Requests and Apologies. Norwood, N.J.: Ablex, 1-34. Brown, P. and Levinson, S. (1987): Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Dressler, W. and Merlini Barbaresi, L. (1994): Morphopragmatics: Diminutives and Intensifies in Italian, German, and Other Languages. Berlin/New York: Mouton de Gruyter. Goffman, E. (1971): Relations in Public. Harmondsworth: Penguin. Köhler, D. (1994): Zur pragmatischen Funktion der synthetischen Diminutive im Deutschen und Englischen: Eine vergleichende Studie. (Unpublished M.A. thesis.) Dublin: University College Dublin. Leech, G. (1983): Principles of Pragmatics. London: Longman. Schneider, K. (forthcoming): 'Size and Attitude': Expressive Wortbildung und diminutivische Ausdrücke in der Englischen Alltagskommunikation.

Corpus (only data used in this paper) ALB COB DCE2 FAT KUR LEI OSB

Albee, E. (1965): Who's afraid of Virginia Woolf? (1962). Harmondsworth: Penguin. Sinclair, J. (ed.) (1987): Collins COBUILD English language dictionary. London/Glasgow/Stuttgart: Collins/Klett. Summers, D. (ed.) (1987): Longman dictionary of contemporary English. 2 nd edition. Berlin/München: Langenscheidt/Longman. Father Ted, The Second Sermon, Chapter One: Hell, 1996. Kureishi, H. (1990): The buddha of suburbia. London: Faber and Faber. Leigh, M. (1983): Abigail's party ./Goose-pimples. Harmondsworth: Penguin. Osborne, J. (1960): Look back in anger. (1956). London: Faber.

Carol Marley Trading Talk? Interaction in Written Dating Advertisements 1. Trade or Talk? Camomile seeks Earl Grey. I am NW, 30's, prof, woman. UR caring, intell. prof. RU my cup of tea? ML20625 1

This advertisement embodies a tension between speech and writing, process and product, interaction and transaction, that is central to its genre. Dating ads have frequently been likened to small ads columns, but small ads feature language in an ancillary role, as a means of achieving a terminal goal, the physical trade of money for goods or services. In the case of dating ads, the goods or services themselves and thus the goal itself - are much more firmly rooted in language exchange. The advertiser is in fact seeking not so much a product as to open a sustainable spoken dialogue with a suitable partner for longer term interaction. The ad itself can therefore be seen as a written medium initiation of a spoken exchange - the opening of a conversation.

2. A Written Medium Genre Coupland's view of dating ads as 'discourses of the commodified self (1996), reflects their element of small ad style trading in goods and services; the highly conventionalised rhetorical structure she describes also focuses on the acquisition of a product. Her 'textual framework' for written ads consists most frequently of the following sequential elements (ibid: 193), illustrated here with the text of my first example: 1. ADVERTISER 2. seeks 3. TARGET 4. GOALS 5. (COMMENT) 6. REFERENCE

1

Camomile I am NW, 30's, prof, woman. seeks Earl Grey UR caring, intell. prof. RU my cup of tea? ML20625

All examples are retyped as they appeared in the original (except for line width and omission of the Message Line reference), retaining bold face, abbreviations and apparent typos. Underlining for ease of reference is my only editorial intervention.

302

Carol

Marley

This advert sits rather uneasily in Coupland's suggested textual framework, in two respects. Firstly, the Advertiser and Target realisations are not continuous in the ad itself, but arranged in alternation between the two, an arrangement that in part draws on a familiar textual device, the general-particular pattern (see, for example, Hoey 1983 and Winter 1992). Another part of its effect, however, is that it also provides a monologic echo of the alternation between speakers that characterises dialogue. Secondly, the writer certainly takes the 'fleeting opportunity...to deconventionalize the advert' (Coupland 1996: 194), offered by the optional fifth slot, but the label 'comment' fails to capture the interactional thrust of 'RU my cup of tea?' In fact, this part of the ad would be better accommodated by one of the categories Coupland suggests for the spoken part of her corpus. She describes the discourse structure of spoken self-advertisements in terms of sequenced 'moves' (ibid: 196), with conversation-based categories such as 'greeting', 'metacommentary', 'promises' and 'thanking/farewell/pre-closing'. Also included towards the end of the sequence is a category of 'challenge', which seems much better to capture the interactional function of my advertiser's closing question. 2 The interactivity between advertiser and reader is considered in more detail by Shalom, who points out that the 'imprecision and vagueness' of the frequent, stereotypical lexis of the genre serves to enhance negotiation by 'engaging the intended reader in a kind of "do I fit?" dance with the text' (1997: 187). She also comments that the 'informal, spoken style further reinforces the interactivity of the text' and notes the use of imperatives, which she attributes to the genre's connection with advertising (ibid: 191). A distinction can be drawn, though, between interactivity and interaction. Thompson and Thetela (1995) suggest that written text is often seen primarily in terms of interactivity, an information-oriented approach concerned with the ways a writer shapes their text to the (ideational) needs of the reader. Interactional analysis, on the other hand, is function-oriented. It has its basis in the exchange of interpersonal roles effected by the language functions of questioning, commanding etc and has thus been associated with studies of speech. But Thompson and Thetela show that written advertisements not only enact and project such interactional roles for both writer and reader, they also manage this interaction to particular ends. Oral and literate strategies, moreover, are not confined to spoken and written media respectively (McCarthy 1993); written text may approximate speech by importing features of the spoken mode or vice versa.

2

Nair comments briefly on a similar phenomenon found in her data, which 'seems designed to initiate a conversation, voiced rather than merely scripted, between the advertiser and an unknown but potentially exciting partner' (1992: 238).

Trading Talk? Interaction in Dating

Advertisements

303

In the analysis that follows I will attempt to show how my advertisers create interactional roles by their use of questions and commands and how they exploit the spoken mode in their written medium ads, in order to open an exchange of the talk that may lead to and sustain the desired relationship.

3. Analysis My data are drawn from a weekly lonely hearts column, SoulMates, covering the first three months of its appearance, from June to August 1995, in The Guardian, a British daily newspaper. The corpus consists of three complete columns over the three months, amounting to 584 ads in all.

3.1. Interactional Roles The ads in my corpus each constitute the written opening move, or Initiation, in a potential exchange; different types of language function are associated with different types of Initiation; and different types of Initiation set up different expectations as to the appropriate Response. Halliday (1994: 69) summarises in a table I have adapted slightly below: Function Give goods-&-services Demand goods-&-services Give information Demand information

Initiation offer command statement question

Expected Response acceptance undertaking acknowledgement answer

Alternative Response rejection refusal contradiction disclaimer

Table 1: Language functions, Initiations and Responses

Coupland's generic framework suggests that although the underlying function of the ads is to give and demand goods and services, the norm for dating ad writers is rather to give information about what they seek, and this is borne out by the majority of my data. However, the exchange will fail in its interactional intent if it fails to elicit an overt response - and herein lies the danger of an Initiation that takes the form of a statement. As Thompson and Thetela point out: In many cases, . . . the expected response to a statement is best expressed in negative terms: allowing the speaker/writer to continue without contradiction, challenge or loss of attention is in itself sufficient acknowledgement by the reader/listener for the interaction to progress. The other functions, on the other hand, do normally expect an overt response (1995: 112).

If an ad writer, then, constructs their ad solely in declarative mood, merely making a statement, they cast themselves as giver of information while their reader, in the reciprocal role of receiver, need be nothing more than attentive by way of a response. Of course, since this is not face-to-face interaction, the writer has no

Carol Marley

304

means of monitoring even this degree of success, and cannot actually elicit a 'real time' response of any other sort. About 10% of my advertisers, though, do indeed opt to enact one of the other roles as a means of exerting greater pressure on their readers to respond.

3.2. Questions Thompson and Thetela identify a category of questions that occur at the end of the text, leaving the reader to complete the interaction (op cit: 115). From the examples they give, it seems that the required response in commercial ads is a preshaped, minimal one - of agreement. The text-final questions that occur in my data, however, do not appear to be quite so straightforward. Examples 2 and 3 illustrate the type of question that most commonly occurs at the end of my ads: Example 2: Romantic liasonette sought by semi-attached, slim, yng 50 male. Likes VU, Foucault, Europe. Hols Sept? Example 3: Jewish masseur, 37, with 1 and a half ears and plenty contradictions WLTM sexy, intell. woman. Walks, talks . . . ?

These questions certainly hand over to the reader to continue the interaction, casting them in the role of answerer, but the possible answers are not so constrained as in Thompson and Thetela's data. In fact, the function of these questions is effectively one of modalisation: they occur in the Goals slot as an invitation to the reader to speculate on the possible future outcome of the interaction, a point beyond the present of the text and indeed beyond their immediate response to what the advertiser offers. A somewhat different type of text-final question does focus on the present of the ad and of an immediate response. My first example, ending with RU my cup of tea?, provided one instance of such a question; others can be seen in examples 4 and 5: Example 4: Attract., exuberant, vivacious, adventurous, females, seeking handsome, hedonous males, for evening of intrepid fun, stimulating convers., friendship and laughter. Have you got what it takes? Example 5: Sunderland male, WLTM a caring intellig. woman, 28—35, to share life. Enjoys music, reading, walking in the Lakes. Fit the bill?

Trading Talk? Interaction in Dating

Advertisements

305

Again these questions require a response from the reader, but they are less speculative - they are more of a challenge for the reader to rise to, firstly by agreeing verbally, and then implicitly by actually replying to the ad. Part of this effect derives in examples 1 and 4 from the less elliptical syntax and explicitly personal naming (cf Thompson and Thetela, op cit, on projected roles) of the participants, as 'I' in the earlier text, and/or 'you' in the question. The ellipsis of examples 2, 3 and 5, on the other hand, varies in degree, but has a consistent effect of introducing a feature of the spoken mode: Halliday describes such exophoric ellipsis as 'taking advantage of the rhetorical structure of the situation, specifically the roles of speaker and hearer' (1994: 63). It thus presents a more conversational style in support of the interactional roles enacted by the use of a question. In the case of example 5, it is worth noting that the ellipsis of the Mood element means the advertiser's question is actually ambiguous, and so opens up wider ground for the interaction. The writer demands information of the reader, but the information the reader gives in response could relate either to you - whether they consider themselves to fit the advertiser's bill, or to I - whether the advertiser fits their own bill. In the case of the earlier examples, more radical ellipsis projects the readerin-the-text even more effectively as a hearer-in-the-speech-situation: the syntax is not explicitly interrogative, but the use of a question mark signals that these are que-clarative utterances. In other words, the writer relies upon the reader to construct an auditory impression of their text, hearing it with the rising tone (Brazil 1985) a speaker would use to present these declarative forms as questions. In example 3, the effect is further reinforced by the use of dots to indicate a pause signalling the end of the 'speaker's' turn and creating a silence for the hearer/reader to fill (Sacks et al 1978). The text-final questions, then, conclude the Initiation by handing over to the reader for them to respond and take responsibility for the next move in the interaction. Another group of questions that regularly occur in my data are textinitial ones, opening the interaction. Unlike the written adverts Thompson and Thetela analyse (op cit), my ad writers do not provide the answers to their own questions, preferring the other pattern they identify, that of posing a yes/no question for the reader to provide a minimal response. And usually the following text in the ad shows that the reader is assumed to have done so: Example 6: American? Intell.? Witty? Celtic woman, 25, PhD, likes travel, mountains, pubs, good conver. London. Example 7: Anguished urges? Youthful, sensual scholar offers v. discreet affair to lovely lusty lady, 25— 40. South.

306

Carol

Marley

These text-initial questions are similar to the text-final ones in that the reader's response seems to be less constrained to agreement than in Thompson and Thetela's data. These questions pro-actively engage the reader in measuring themselves for how well they fit against a kind of checklist of pre-requisites for further exchange. But those whose reply is negative need read no further; only those who are able to reply in the affirmative with a 'preferred second pair part' (Levinson 1983), are now bona fide addressees of the advertiser. By contrast with the text-final questions, the advertiser then retains responsibility for the following move. There is an unspoken/unwritten 'if so, then' mediating the demand for information in the advertiser's question and their giving of information in the next textual move, resulting in a more fully developed exchange structure that might tentatively be represented as follows: Advertiser: Reader:

Are you American? [Yes]

Advertiser:

[OK] Are you intelligent? [Yes]

Reader: Advertiser: Reader: Advertiser: Reader:

[OK] Are you witty? [Yes] [OK, then] I am a Celtic woman etc ?

I - elicit R - reply F - accept I - elicit R - reply F - accept I - elicit R - reply F - accept I - inform R?

This use of questions actually extends the interaction of the ad over at least two, and possibly more, exchanges. The conditional logical relation between the moves actually realised in the text (reflected in the need for a Follow Up move in my expansion above) is very occasionally made explicit in the ad: Example 8: Are all men around 60 longing for women 20 yrs younger? If not call my ML and get a surprise.

While text-final questions hand over the turn and responsibility to the reader, then, text-initial questions generally exert a greater degree of management by extending over several exchanges and implying that at least one reader has been able to supply a match in respect of the information the advertiser demands. Both these uses exploit the relative constraints of yes/no question forms, but nevertheless seem quite readily to allow a 'wrong' answer, in order to weed out unsuitable respondents. The technically more open wh-question forms are not widely used by my advertisers but when they are, their function is similar to yes/no questions,

Trading Talk? Interaction in Dating

Advertisements

307

because they are limited as to addressee and pragmatically loaded with a preferred response. In fact, they seem to be restricted to demanding information as to the location of the advertiser's Target, in order to elicit a response of 'I'm here': Example 9: Attract, grad, 37, wonders if there are any articulate, caring, funny men left. Where are vou all? Lond. Example 10: Choicest blend /Earl Grey, where is Camomile's special cup of tea? I'm prof. 38, NW. UR caring, intellig., kind. L'pool.

Example 10 appeared approximately 3 weeks after my first example, apparently placed by the same advertiser. Interestingly, as well as having broadened the label referring to her Target - now offering Choicest blend as an alternative to Earl Grey - the writer appears to have opted for a less challenging reformulation and repositioning of her question. It is tempting (but probably not wise!) to speculate that she reaped the rewards of this, and of the increased interaction that results from using a text-initial question so as to involve the reader in more than a single exchange: no more ads from Camomile appeared during the 3 month span of my data.

3.3. Commands As Table 1 showed, questions function to elicit a verbal response from a reader; commands, however, are ultimately intended to prompt an action response, with the verbal undertaking shown in Halliday's outline being an interim response if the action is not to be carried out immediately (or additional, if it is). In view of this, the most natural place for commands to occur in my ads, and indeed the most frequent placement actually found in the data, is at the end of the text. Example 8, above, provided such an instance, with a command in which the 'first person' advertiser addresses the 'second person' reader as a separate entity, directing them to call my ML etc. The example below provides an alternative type, a let's imperative, used by my advertisers to include both writer and reader together in a joint course of action:

Example 11: Judge Dredd seeks Tank Girl. 35, into computer art, cycling and films. You wild, warm, friendly and sexy. Let's surf into the summer! Ldn.

As with text-final questions, text-final imperatives serve to hand responsibility for sustaining the exchange over to the reader. Similarly, too, a distinction can be

308

Carol Marley

made as to whether or not they focus on the immediate present of the ad. Those beginning with let's again invoke speculation on a delayed future outcome of the reader's more immediate, physical response of replying to the ad. The other imperatives further divide; firstly into those that refer directly to the present exchange, lexicalising the action that constitutes the next move, and directing the reader to call, ring or phone, as example 8 does. Some indeed refer even more urgently and explicitly to immediacy of response; the next example combines commands of both types and culminates in an 'aggravated' instruction to act at once: Example 12: Soul mated lover needed. "Send me your picture diary and loves pullover, let's see if it fits". Phone now. Harrogate.

Other commands, on the other hand, refer only indirectly to the present exchange: Example 13: 29, profess, man into outdoors/indoors, doing/thinking. WLTM F, 21-30. Go on! I'd like to meet you.

The urging of this command has a function similar, but not identical, to text-initial questions. The indirectness of Go on! or Do it! has to be decoded as meaning call me or its equivalent, and using this inexplicit formulation casts the reader in the role of someone already considering replying to the initiation and just needing a little encouragement to pick up the phone. In other words, this strategy seems designed to extend the exchange, as questions may, and projects a reader-in-thetext who is caught semi-frozen in the act of deciding whether to respond. These commands differ in structural effect though: such exhortation from the writer must be seen as intervening in the original exchange, as a bound Initiation attempting to keep it on track, rather than effecting real progress into a subsequent exchange as was seen with the use of questions. Speculation (with a let's imperative) on the putative relationship apart, the number of commands and assumed responses which could perform a function strictly parallel to questions and assumed answers must be seen as severely limited, since the assumed action performed by the reader must be restricted to the here-and-now of the text. However, as the next example shows, it is indeed possible - though rare - to chain commands via an assumed response across two exchanges: Example 14: Untangle the following anagram: lal nme rea satabds, then contact one who isn't. Ldn.

Trading Talk? Interaction in Dating Advertisements

309

3.4. Serial Exchanges Example 15: Do you like Jazz, cinema, going out to breakfast? I do, I'm 29, black, female graduate. Call me. London.

This example illustrates the most obvious strategy for sustaining an interaction across several exchanges within a single ad, that of combining eliciting and directive exchanges. After opening with a question which assumes a three-part response (not necessarily positive) from the reader, this writer exploits expectations that conversational openings are reciprocal (Levinson, 1983). That is, by providing an apparent response of her own, she seems to assume not only that the reader provides information in reply, but also that they politely take their own turn to initiate by returning the question: Advertiser: Reader: Advertiser:

Do you like Jazz, cinema, going out to breakfast? [I do (not) What about you?] I do

I R I R

There are thus arguably four 'turns' in this ad, spread across two exchanges, before the writer makes a third initiation with a statement, I'm 29 etc, that the reader need only attend to by way of a response. Immediately on the heels of this, though, she then enacts another set of interactional roles for herself and her reader, from whom she now requires another overt response with the final command. She thus steers a varied interactional path through several exchanges before handing over to a reader who by now is effectively a co-conversationalist. Notably, too, this interaction is conducted on a much more even footing than that in example 8, where the implicit IRF structure was strongly redolent of the unequal power relations of teacher-pupil and doctor-patient interactions. Although I have concentrated so far on the interactional technique commonly exploited, that of selecting initiating roles that require an overt response from the reader, the last example also introduced an alternative strategy of assuming a responding role. This is used by a small number of advertisers to effect the highest degree of interactional intimacy found in my data. 3.5. Responses So far the enacted roles I have discussed have paralleled those Thompson and Thetela (op cit) found in commercial adverts fairly closely. My advertisers' use of responding moves, however, provides a more significant departure from their findings that the ad writer usually explicitly formulates a question attributable to

310

Carol Marley

the reader-in-the-text before going on to provide the answer to it. The closest writers in my corpus come to this strategy is shown in the following example: Example 16: Gay guy, 32, WLTM someone 30-45 for interesting life. Me? I'm an undiscovered little gem. London.

As in earlier examples, the highly elliptical form of this question contributes to its conversational flavour. But there is also another way in which this usage belongs to a spoken, rather than written mode, since it is the kind of turn-initial echo question familiar in face-to-face interaction. So although the question is indeed textually realised, it actually presupposes a missing turn. This turn can be reconstructed as consisting of a Response from the reader (either verbal or silent acknowledgement of the advertiser's informing opening move), plus their own Initiation, requesting the advertiser to give more information about himself. The next textual move then forms the advertiser's Response to the reader's Initiation, in a second exchange. However, an echo question such as this actually implies a further responding role for the reader, whose confirmation - even if silent rather than verbalised - is required for the interaction to proceed. Effectively, then, this advertiser increases the interactional intimacy of responding to a reader's question by returning it for agreement before extending further into a third exchange with the answer itself. Other advertisers, by contrast, omit any textualisation of the reader's question, but nevertheless supply information that presupposes someone has asked them for it: Example 17: Yes they do exist! Tall, fit, prof, male, low mileage, good cond., no baggage, WLTM sim. female, 28-35. NW. Example 18: Arr. the same old thing. 30 y.o. man needs woman of same age with brains and hips! CardiftfBristol.

The yes and arr in these examples provide linguistic signals of response: the reader-questions presupposed must be construed as something along the lines of 'Do males who are desirable and unattached exist?' and 'Why are you advertising?' These two ads differ from examples 15 and 16 in that the answers to the implicit questions notably occur in text-initial position thus plunging the reader into mid-interaction with the writer, who is moreover assumed to have initiated it. Each of them also makes use of an aspect of the spoken mode that

Trading Talk? Interaction in Dating Advertisements

311

occurred more frequently in my data than any of the other interactional devices I have so far considered.

3.6. Exclamations I suggested in 3.2 that my advertisers exploited intonational aspects of the spoken mode, using question marks as an alternative to interrogative syntax, in order to encourage the reader into an overt response. Although I found no use of exclamative syntax in my data, I found plentiful use of orthographic exclamations, which again require the reader to reconstruct the auditory experience of face-toface interaction. These exclamations fall into three main categories in terms of what they appear to signal to the reader. Often they signal a joke or a humorous incongruity: Example 19: Male, 35, separated, hates pseuds, likes TV, pubs, films, music, sex - reverse order! Slim, not too ugly. N. Yorks.

More frequently they occur at the start of the ad as attention grabbers, sometimes explicitly so with phrases such as Attention please! or SOS!, and sometimes implicitly so, by foregrounding the advertiser's self-description, for example, Hunk-U-Like! and Not a boring accountant! Most frequently of all, however, they initially appear to do nothing more than provide a textual indication of emphatic intonation, giving 'auditory' remarkability to utterances that may otherwise be entirely unremarkable: Example 20: Gay male, 32. Art lover, music lover. Seeks similar for fun, friendship, maybe more! Derbys. Example 21: Dry witted, cynical, 26, needs softening with affection and romance! Into jazz and house conversions!

However, the reader's reconstruction of the intonation of the exclamation is likely to involve high termination (Brazil 1985), and as Coulthard points out such a choice at the end of an initiating move 'constrains the other speaker to make a contribution' (1985: 141). In other words, using an exclamation mark appears to be another strategy employed by my advertisers to achieve an overt response from their reader-interlocutors. Coulthard goes so far as to suggest that informing initiations of this sort, with high termination constraining the next speaker to verbalise their reaction, are actually borderline between 'the most extreme type of inform [and] the mildest of elicit' (ibid: 142).

312

Carol Marley

4. Conclusion My analysis has shown that, in spite of the written mode and accompanying generic conventions, a proportion of dating advertisers exploit or approximate aspects of the spoken mode and enact overt interactional roles for themselves and project implicit ones for their readers. I have argued that the use of initiating commands, questions and exclamations exerts a pressure on the reader to make an overt response to the advertisers' opening moves and become a participant in this incipient conversation. Alternatively, or in combination with enacting initiating roles for themselves, advertisers may enact responding roles, with the effect of rendering the reader a co-conversationalist already by pitching them straight into the middle of an exchange. These are arguably very similar to the strategies Thompson and Thetela have found commercial advertisers to use. But two considerations point to a significant difference. Firstly my advertisers pose questions that leave the reader freer to respond as they choose, rather than as the writer constrains them to. Secondly, my advertisers do not explicitly enact initiations or responses for a reader-in-the-text; unlike their commercial counterparts, they do not provide textualisations of both an initiation and its response, in the kind of mimicked interaction Thompson and Thetela liken to 'that between a ventriloquist and his dummy' (op cit: 115). My advertisers may encourage the reader's participation by exerting pressure to respond or implying they have initiated, but they do not take the liberty of speaking their words for them as Thompson and Thetela's writers do. Commercial writers, then, display towards their readers a higher degree of presumption, a degree that could indeed be offensive or patronising to the reader of a dating ad and so counterproductive for the writer. However, in other respects, such as personal attributes of their desired other, these writers presume a great deal. In fact, I would suggest the better explanation is to be found in the tension I referred to at the outset of this paper between the process of interacting and the product-oriented trade in commodified selves. Both the differences I have suggested point to a greater tendency for my advertisers to narrow down, rather than widen, the type of reader they wish to encourage as respondents, by allowing them to self-select only if they meet the writer's requirements. Dating advertisers may indeed wish to interact and open a conversation with their readers, but not at all costs; they are also participants in a transaction whose goal is sustainable interaction. From this angle, they are traders offering themselves as 'goods' in return not for money but for other goods. And while it may not matter who buys your old sofa for hard cash so long as the deal is done, in this case the 'goods' you receive in return are precisely your reader and partner in talk.

Trading Talk? Interaction in Dating Advertisements

313

References Brazil, D. (1985): The Communicative Value of Intonation in English, Discourse Analysis Monograph no. 8, English Language Research, Birmingham. Coulthard, R. M. (1985): An Introduction to Discourse Analysis (2nd edn), London and New York: Longman. Coupland, J. (1996): Dating advertisements: discourses of the commodified self, Discourse and Society 7 (2): 187-207. Halliday, M. A. K. (1994): An Introduction to Functional Grammar (2nd edn), London: Edward Arnold. Hoey, M. P. (1983): On the Surface of Discourse, London: George Allen and Unwin. Levinson, S. C. (1983): Pragmatics, Cambridge: Cambridge University Press. McCarthy, M. (1993): Spoken discourse markers in written text. In: J. M. Sinclair, M. P. Hoey and G. Fox, (eds.), Techniques of Description, 170-182, London: Routledge. Nair, R. B. (1992): Gender, genre and generative grammar: Deconstructing the matrimonial column. In: M. Toolan, (ed.), Language, Text and Context: Essays in Stylistics, London and New York: Routledge, 227-254. Sacks, H., Schegloff, E. A. and Jefferson, G. (1974): A simplest systematics for the organisation of turn-taking for conversation, Language 50 (4): 696-735. Shalom, C. (1997): That great supermarket of desire: Attributes of the desired other in personal advertisements. In: K. Harvey and C. Shalom, (eds.), Language and Desire, London: Routledge, 186-203. Thompson, G. and Thetela, P. (1995): The sound of one hand clapping: The management of interaction in written discourse, Text 15 (1): 103-127. Winter, E. O. (1992): The notion of unspecific versus specific as one way of analysing the information of a fund-raising letter. In: W. C. Mann and S. A. Thompson, (eds.), Discourse Description: Diverse Linguistic Analyses of a Fund-raising Text, Amsterdam: John Benjamins, 131-170.

Acknowledgement I would like to thank Sarah Waite-Gleave for her encouraging and telling comments on an earlier draft of this work.

Fathi Dali Code-Switching and Dialogism in Lolita1 Many critics have written on the importance of language in Lolita, and the language they meant is English. They have praised Nabokov for his handling of the English language and have given many examples. In this paper however, we hope to study Humbert Humbert's - HH from now on - and, to a lesser extent Nabokov's use or manipulation of French, and to examine how it is integrated into the English text. HH was born in Paris. His mother was English but she died when he was only three and his mother's sister took care of him. HH's father was "a Swiss citizen of mixed French and Austrian descent" (9). HH was raised partly in the South of France, on the Riviera. His father read to him Don Quixote and Les Misérables, and HH attended "an English day school". At the age of thirteen HH was sent to a lycée in Lyon. He later became a "college student in London and Paris" (15). He developed an interest in English Literature. He lived in Paris for a long period of time, composed pastiches in English (16), "launched upon an Histoire abrégée de la poésie Anglaisé" (16), taught English first to adults and then in a school for boys, and "started to compile that manual of French literature for English speaking students which was to occupy me throughout the forties - and the last volume of which was almost ready for press by the time of my arrest" (16). HH is bilingual, fluent in both English and French and his English is idiomatic, "casually littered with linguistic small change from both sides of the Atlantic." (Wood 1994: 110). HH likes to consider himself and to be considered a cultivated European in general and a cultivated Frenchman in particular: "as we French scholars may ..." (202). In another instance he says: "soyons logiques, crowed the cocky gallic part of my brain" (238). According to Wood (1994: 112), Lolita's first publisher asked Nabokov "to degallicise the text a bit". Which he did. But for Wood "it can have been only a bit, since there is scarcely a page without a slither into French (...)" (112). It is the thesis of this paper that since Nabokov was reluctant to "degallicise" his text these instances and instants of code-switching are significant. While it is true that HH can say anything he wants in English, still can we simplify and say that the French expressions and words are just a "signalling of strangeness" as Wood claims (113) or that HH resolutely refuses "to be confused with the natives" (idem)?

1

All quotations are from A. Appel Jr.'s 1990 annotated edition of Lolita.

316

Fathi Dali

Code-switching or code mixing takes place when a speaker, in the course of his speech in one language uses a word or group of words in another language. Speakers who switch codes are usually bilingual or multilingual. In general it is difficult to "predict when a switch will occur without knowing what a speaker intend[s] to say next" (Holmes: 48). But, as Holmes also says, "it is often possible to account for the switches after they have occurred." (idem). The analyses of the reasons behind code-switching are all the more difficult when we deal with a highly cultivated and skilled bilingual like HH, who, paradoxically enough, unlike other code-switchers, switches back to his own language, i.e. French. The first and least important type of code-switching in Lolita is the one in which HH uses technical or pseudo-technical words or expressions: "For many bilinguals certain kinds of referential context are more appropriately or more easily expressed in one language than the other." (Holmes: 44). Thus HH uses chocolat glacé (13), a typically French drink de l'Époque, pot au feu (25), préfecture, etc. Sometimes HH, or rather Nabokov, uses, in italics, French words which have become part of the English vocabulary such as for example trousseau (278). It does not necessarily mean that HH/Nabokov did not know that 'trousseau' is also an English word. It could simply mean that HH used it in French, (including the French pronunciation) when addressing Lolita, as part of his last attempt to seduce her. The second type of code-switching is "when a speaker switches code to quote a person." (Holmes: 44). In Lolita for example, in part one chapter 6, HH quotes in French and with obvious delight a Parisian prostitute. He likes among other things the way her Parisian childish lips exploded on bas, when "she said with great gusto: 'je vais m'acheter des bas'". HH also quotes extensively his friend Gaston Godin who speaks only French because his "English was a burlesque" (181). HH also indulges himself by quoting French poets like "dear old Ronsard" (214) and Belleau (47). But although these types of code-switching are acceptable for the sake of realism, exoticism etc, it is important to notice that HH rarely translates these quotations or gives a hint as to what they mean. HH expects his American reader to be able to understand French, to come closer to him, to share with him the pleasure of understanding French, to be his accomplice. HH wants his reader to share the joy of parodying "Gros Gaston", to share the sad feeling of homesickness towards Paris. He needs the readers' solidarity, or at least of those American readers who are sophisticated enough to understand and appreciate French. Speech fillers, called also "emblematic switching or tag switching (Holmes 1992: 42), are another type of code switching used lavishly by HH. This consists of using "an interjection, a tag, or a sentence filler in another language (...)" (Holmes, idem). Fillers like "I mean", "You see", "You know" are the most common among speakers of English. In general every speaker tends to use one or

Code-Switching

and Dialogism in Lolita

317

two particular fillers. In Lolita HH uses several, depending on what he is saying. Take the following example (66): "(...) I could do nothing about it, but nothing, mais rien.'''' In this example it is as if the English words do not convey the meaning so efficiently as the French ones. We notice here that "but nothing", which is not a proper exclamatory filler in English, anticipates and parodies 'mais rien\ a proper filler in French. HH uses many French speech fillers among which could be mentioned: "comme on dit" (125 and 157), "entre nous soit dit" (128), "soit-disant" (74 and 147), "que dis-je" (168), "done" (217), "que sais-je?" (241), and so on. These fillers have no effect whatsoever on the meaning of the sentence where they occur, nor do they show that HH has the mania of using one filler all the time. Appel, whose annotation of Lolita is of invaluable help to the student of the novel, took great care to translate all the French words and phrases in it. Looking at his translations of these fillers one can see that the French fillers, like any filler in any language, are not translatable, that they are typically French. When translated these expressions lose their flavour, and perhaps one of their major functions of "ethnic identity markers" (Holmes 1992: 42). Certainly HH writes his narrative in English, because it is addressed to an American jury, but apparently he cannot help availing himself of French words when they provide "le mot juste" (as G. Flaubert would say). Sometimes it is more than a simple filler that is at stake. HH at times prefers a French syntactic structure to an English one. Thus, as Mrs Chatfield exclaims, "For shame, Mr. Humbert! The poor boy has just been killed in Korea," HH wonders "didn't she think 'vient de' with the infinitive expressing recent events so much more neatly than the English 'just' with the past?" (290). All languages being equal, HH's judgement is subjective and unscientific. That is HH would have said: "The poor boy vient d'être killed in Korea." But if he considers that part of his brain is "gallic" (238) i.e. that he is bilingual, then it is perhaps understandable that he feels that, for example, "arrière pensée" (78), or "enfin seuls" (119), or "le grand moment" (123), convey "more neatly" what he means to say, or suit the circumstances better than those English equivalents which Appel, for instance, tried to provide. We are reminded here of N. Sarraute's comment on Joyce's Finnegans Wake (1939), however different from Lolita that novel is. She says that that novel is an illustration of "the extraordinary complexity of a poetic universe for which one language is not enough, which needs several languages." (Aubert, Jolas 1979: 67). It seems that English was not enough to satisfy HH when he sought to compose his creative universe. In other cases we discover that instead of expressing himself in French and then giving the (approximate) meaning in English HH proceeds the other way round by appending the French word as in this example, "a burst of furious sarcasm - un ricanement - " (224). In an unusual way the French word comes as a

318

Fathi Dali

reinforcement of the idea expressed in English. It is enhanced by its position at the end of the sentence. In another context and to stress his love for Lolita, HH says (248): "I Loved you. I was a pentapode monster, but I loved you. I was despicable and brutal, and turpid, and everything, mais je t'aimais, je t 'aimais T To say "I loved you" in English more than once, does not seem to HH to be as expressive or as sincere as when it is said in French. HH here hopes to justify and excuse his past brutality and bestiality by stressing his love in his own language and by putting the French phrase in a prominent position. It is as if English also were unable to convey his message. Distressed HH simply takes refuge in the language he loves and trusts the most. There exists in Lolita yet another type of code-switching, which is at the same time stylistic and humorous. We are promised from its very beginning that this novel is going to be written in a "fancy prose style" (9), and later HH confesses that he has only words to play with. Thus, the reader's attention is attracted by the many examples of alliterations that HH likes to produce from time to time: "The implied sun pulsated in the supplied poplars" (60) (where the voiceless bilabial stop /p/ is combined with the alveolar liquid /l/, and the alveolar fricative /s/); or, "A florid-faced fellow with sapphire-blue eyes (where the labiodental fricative /f/ is repeated four times). The reader who cares to look for more examples of alliteration will find that there are many of them, from the long ones which repeat one or more sounds in the same phrases to those where the sounds are repeated only twice in two juxtaposed words, paralleling the narrator's pseudonym Humbert Humbert. For example the sounds /£/ and /s/ in "fuss and faces" (166), or /f/ in "fattish and faded" (303), or the cluster of sounds /bl/ in "black and bloody" (305) etc. HH indulges in the same exercise by alliterating words or groups of words in French or by combining French and English words. For example the sounds /v/ and /r/ in (174): a vieillard encore vert - or was it green rot?"; or again the sounds /k/ and /s/ in: "Monique wanted to know if she should remove the layer of red from her lips avant qu 'on se couche in case I planned to kiss her" (22). In both examples we have code-switching and alliteration combined, a game HH is fond of. Mixing French and English HH manages also to produce some puns as when he describes the Parisian prostitute's hotel room as "bed and bidet" (22), which parodies 'bread and breakfast'. He calls himself in another context "a manqué talenC (15) where the French adjective sounds like the English noun 'monkey', and brings to mind HH reference to himself many times as an ape. Later, he tries, sheepishly, the same game of play on words when the nurse refuses to let him touch an envelope she is taking to Lolita. He said (243): "Je croyais que c 'était un bill - not a billet doux". HH is aware, and is made to understand by both the nurse and Lolita, that his jeu de mot is a flop. We must admit that he himself sometimes wonders at the ineptitude of his puns as, for example, when he rings the bell of

Code-Switching

and Dialogism in Lolita

319

Lolita's house and nobody answers (269): "Personne. Je resonne. Repersonne. From what depth this re-nonsense?" It is perhaps possible to interpret this "renonsense" by saying that HH finds himself too weak to resist a pun. These games of alliteration, these plays on words, these puns, introduce a linguistic playfulness which is central to Nabokov and which exhibits his deep delight in the sheer pleasure of words, all words from all languages: English, French, German, Latin. But we may also try to explain the reasons behind at least some cases of codeswitching through Nabokov's attitude towards translation. Nabokov, who stood for a dogged literalism in translation, said (Appel, Jr. 1972: 51): "The only object and justification of translation is the conveying of the most exact information possible and this can only be achieved by a literal translation with notes." It is easy to test this theory on the type of code-switching which involves a referential content, but with HH, we have the impression that, even a "literal translation with notes" would fall short of conveying accurately the meaning, or the feeling, or the sensations contained in the French words or expressions he uses, whether with a referential context or not. For example, in (105): "One might suppose that (...) I would have sunk back, heaving a sigh of delicious relief. Eh bien, pas du tout!" The English "Well not at all" (Notes: 374) - a literal translation of the French exclamation - not only lacks that defmiteness and the regret of the French source, but, most of all, it is not a proper filler and comes to the French speaker's mind as it came to HH's, automatically. In some cases the textual environment seems to trigger the filler in HH's train of thought, though the filler cannot be predicted. Take for example (128): "(a rather sordid affair, entre nous soit dit. The English words "sordid affair" are literally the same in French: "une affaire sordide", hence the French filler which solicits the addressee's complicity. And later, HH tells us about their tour, which was "far from being an indolent partie de plaisir". He says that its "sole raison d'être (...) was to keep my companion in passable humor from kiss to kiss." (154). Not only does "sole" - an English pun: sole, soul - sound like "seule", but even "passable humor" looks like a literal translation of the French cliché "humeur passable". It is interesting at this stage to consider the American characters' attitude to HH's code-switching. We start with Charlotte's positive attitude. Sociolinguists would say she converges. In love with HH, she therefore admires his "proficiency in intra-sentential code-switching" (Holmes 1992: 52). She knows some French and tries to emulate HH when she addresses him or her daughter Lolita - in his presence - or when she writes him a love letter. Charlotte, whom HH despises and whose "awful French" (68) he hates and parodies, is the only American character in the novel who accommodates her speech to his (Giles et al 1987), who has had an admiring attitude to his juggling with both English and French. She finds it glamorous and high-brow to switch codes.

320

Fathi

Dali

The other American characters either refrain from expressing any opinion about HH's code-switching or have a negative attitude to it, i.e. they diverge, like, for example, Lolita's nurse who says irritably after the biWbillet doux flop (243): "And my pappy can parlay-voo as well as yours". Lolita's attitude is also definitely negative, although she too switches to French when she is after something from HH. Thus, when the nurse complains about HH's French, Lolita tells him (243): "do you mind very much cutting out the French? It annoys everybody." But when she wants, for example, to leave school and go "for a long trip again" (297), or when she wants HH to do her a favour, she accommodates her speech to HH's and uses French: '"I choose? C'est entendu?' she asked wobbling a little beside me. Used French only when she was a very good little girl." But in general, the American characters' reactions to HH's code-switching are politely negative. Still, HH never thinks of apologising for it. It may be that HH is stubborn in his desire to seduce Lolita, or to impress people, or to enhance his status "by his ability to manipulate two or more codes proficiently" (Holmes: 52). It may make him feel different and superior, still, the result is different from what he has expected. His code switching irritates more than anything else, not only because the US in the fifties was staunchly monolingual, but also because the Americans resented his arrogance and discourtesy. Some critics try to explain HH's mania. For example, Frosch (1987: 89) speaks about "displacement or incongruity" in Lolita: incongruity "appears in HH's old world, European manner - aristocratic, starchy, and genteel - set in a brassy America of motels and movie magazines, and in his formal, elegant style of speaking posed against Lolita's slang". Frosch (idem) reminds us that "the geographical, linguistic, and temporal aspects of HH's dislocation are often related to Nabokov's own exile". Nabokov was multilingual - Russian, English, German, French - and Lolita was his second novel in English, his first being The Real Life of Sebastian Knight (1941). In an interview conducted in September 1966 by A. Appel, Jr. and to a remark on how "dramatic" it was "to think of you bidding farewell to one language and embarking on a new life in another..." by writing Sebastian Knight in English, Nabokov replied that he knew he "would eventually land in America." Therefore he "switched to English" after he was convinced that he "could use English as a wistful standby for Russian." Nabokov added that he still felt "the pangs of that substitution..." (Appel, Jr. 1972: 57). It is legitimate to parallel HH's problems with his author's "private tragedy". Indeed each of these two expatriates had to "abandon his natural idiom" at a crucial moment of his life, although not in similar circumstances. However, and although the interpretations mentioned above (whether linguistic or psychological) are not to be discarded totally, it is my argument that the major explanation for HH's code-switching lies within the dialogic design that permeates the whole novel, for dialogism, as described by Bakhtin (1981), is intrinsic to the novel. It is a quality that includes

Code-Switching and Dialogism in Lolita

321

many voices, and many levels of language and dialects. Code-switching frustrates any desire to see the world in a monologic fashion. It insists that the world be composed of a multiplicity of codes and languages. And, as I shall shortly propose, code-switching in Lolita opens the novel up ever more widely to the code of everyday language - the cliché - and to the code of art, i.e. the history and tradition of the novel. We start here by examining HH's own attitude towards his own code switching. He calls "raison d'être" and "partie de plaisir" "clichés" (154): "these French clichés are symptomatic". Indeed, when we examine the list of the French words or phrases used by HH we find that most of them, if not all of them, are clichés (a French word!): mon oncle d'Amérique (27), soit disant, casé, qui prenait son temps, à titre documentaire, vin triste, comme on dit (127 and 157), etc. The French expression 'comme on dit' inevitably introduces a cliché, and in this case it is an American cliché. Furthermore, and significantly, at the very beginning of the second part of the novel, HH starts toying with a French phrase "wows connûmes" which has, for him, "a Flaubertian intonation" (145), and with which he opens three consecutive paragraphs, "(this is royal fun)", he exclaimed gleefully. To my mind the clue provided by this example is the French novelist Flaubert. Besides, HH did say later (211): "In my youth I once read a French detective tale where the clues were actually in italics". Appel, Jr. thought it was "one of the works of Maurice Leblanc", (Notes: 410) but failed to give a title. This explanation is quite possible, but Gustave Flaubert's Madame Bovary (1856) not "a detective tale" of course is teeming with italics, though their significance is different from those in Lolita. On the other hand, Maddox (1983: 83-84), pointing at the similarities - and differences - between Lolita and Madame Bovary, notices that Lolita took piano lessons "with a Miss Emperor (as we French scholars may conveniently call her)" (202), - an allusion to Mile Lempereur from whom Emma Bovary pretended to be "taking her piano lessons" Maddox; idem). A (deliberate?) slip of the tongue made by HH ("mopping up Gustave's - I mean Gaston's - king side" followed by a telephone call by Miss Emperor to report to HH that Lolita is missing her piano lessons is one more clue as to the parodie similarity between Emma Bovary and Lolita Haze. As for the cliché, we know that it is one of the important stylistic features used by French realists, and in particular by Flaubert and Balzac. Besides the many critics who wrote about this (C. Duchet 1975, Herschberg-Pierrot 1979, 1980, Amossy and Rosen 1982), Nabokov himself, in his lecture on Madame Bovary (1980), said that Flaubert used to collect all types of clichés. All the French words and phrases italicised in Lolita have been said before either by laymen or by poets and writers. They are the "déjà dit", the "comme-on-dit", a reproduction, and therefore they have no originality, no singularity. Starting from

322

Fathi Dali

this postulate we can also say that many of the quotations in Nabokov's novel (i.e. Mérimée, Ronsard, Eliot etc) were overused and turned into clichés, that HH's story itself, to a large extent, is a cliché, a replica of Poe's and of Dante's, and even of young HH's amorous adventure on the French Riviera. One can even say that Lolita and HH are themselves clichés, stereotypes of characters who have already existed in world literature. The novel Lolita, as all the critics (Appel, Jr. 1967, Barbedette 1986, etc) agree, is indisputably a parodie novel, i.e. a reproduction of some pre-existing stories. One aspect of parody is that it "may imply a flattering tribute to the original writer" (Holman 1972: 380). If we study Lolita in this context, i.e. that of the parodie novel, we come to the conclusion that, with its other functions developed at the beginning of this paper, code-switching becomes one of the more important components of the dialogicity of the novel in general, and especially as it is set by Nabokov for his novel. Poe, Dante, Rimbeaud, Ronsard, Joyce, Flaubert, and many other writers are quoted, invoked, mentioned, parodied in Lolita. Languages and levels of languages are used, toyed with. All this contributes to the dialogicity of this novel. We witness the coexistence of two - or more - discourses within many of HH's phrases and sentences who likes to cross boundaries, not in an aggressive way but by paying tribute to the original owner. HH is on a constant move, accommodating, converging, migrating from one discourse to another - like a honey-gathering butterfly - resenting a monovoiced, monocultural discourse.

References Amossy, R. and Elisheva Rosen (1982): Les Discours du cliché. Paris: SEDES. Appel, Jr. A. (1972): Vladimir Nabokov, an interview in The Contemporary Writer, 35-60. Edited by L. S. Dembo and C. N. Pondrom, The University of Wisconsin Press. Appel, Jr. A. (1987): The Springboard of Parody. In: Vladimir Nabokov's Lolita, 35-51. Edited by Harold Bloom, New York: Chelsea House Publishers. Aubert, J. and Jolas, M. (1979): Joyce and Paris, Paris, CNRS and Publications de l'Université de Lille 3. Bakhtin, M. M. (1981): The Dialogic Imagination. Austin: University ofTexas Press. Barbedette, G. (1986): Entre l'exil et la Parodie. In: Dossier sur Vladimir Nabokov, in Magazine Littéraire (23) Sept. 1986, 18-23, Paris. Duchet, C. (1975): Signifiance et in-signifiance: le discours italique dans Madame Bovary. In: La Production du sens chez Flaubert, 358-378. Frosh, T. R. (1987): Parody and Authenticity in Lolita. In: Vladimir Nobakov's Lolita, 83-97. Edited by Harold Bloom. New York: Chelsea House publishers. Giles, H., Mulac, A., Bradac, J. J., and Johnson, P. (1987): Speech Accommodation Theory: The First Decade and Beyond. In: M. L. McLaughlin (ed.) Communication Yearbook 10, Beverley Hills: Sage, 13-48. Hershberg-Pierrot, A. (1979): Clichés, Stéréotypie et Stratégie discursive dans le discours de Lieuvain: Madame Bovary, II, 8. In: Littérature (36), Décembre 1979, 88-103, Paris: Larousse.

Code-Switching and Dialogism in Lolita

323

Hershberg-Pierrot, A. (1980): Problématiques du Cliché: sur Flaubert. In: Poétique, (43), Sept 1980, Paris: Seuil. Holman, C. H. (1972): A Handbook to Literature. New York: Odyssey. Holmes, J. (1992): An Introduction to Sociolinguistics. London and New York, Longman. Maddox, L. (1983): Nabokov's Novels in English. University of Georgia Press. Nabokov, V. (1970): Lolita, the annotated edition. Preface, introduction, and notes by A. Appel Jr., New York: Vintage books. Nabokov, V. (1980): Lectures on Literature. Harcourt: New York. Wood, M. (1994): The Magician's Doubts, Nabokov and the risks of Fiction. London: Chatto and Windus.

Murray Knowles

'Suffering is Good for the Soul': Dialogue and Ideology in Traditional Juvenile Fiction 1. Introduction In any discourse situation there will exist relations of expectation between participants. In speech interlocutors may command, request, persuade, inform and so on. In written texts, however, interaction or the rhetoric of discourse has a different perspective. In spoken discourse, the sending and receiving of a message is almost instantaneous and assumptions can be, and are, made by participants of each other as messages are encoded and decoded. In literary discourse, however, we have ' a kind of discourse where the writer can assume relatively little about the receiver of his message or the context in which it will be received' (Leech and Short 1981: 258). Of course, as Leech and Short (op cit. 259) go on to point out, writers, while not having direct knowledge of their readers, can make assumptions about common experience, cultural background and shared knowledge. In literature written for children, there is an added dimension in that: The realisation of a text, and especially of a text for children, is closely involved with questions o f control, and of the techniques through which power is exercised over, or shared with the reader. (Hunt 1988: 163)

All literary texts will have levels of discourse that involve author and reader in addition to the level involving interaction between characters in the story. All levels are important in assessing the question of authorial control. In this paper, however, I am concerned with the discourse that pertains between writer and reader. I will focus, in particular, on a lexical description which permits a critical and stylistic analysis notably through the identification of significant lexical items and their collocates. I wish to consider patterns of vocabulary and their usefulness in investigating point of view in texts. Point of view refers to that aspect of storytelling which is associated with authorial control or the narrative framework employed by the author. Simpson (1993: 11) sees the psychological as the most important of the categories of point of view that he discusses because: It will encompass the means by which a fictional world is slanted in a particular way or the means by which narrators construct, in linguistic terms, their own view of the story they tell.

Murray

326

Knowles

2. Children's Literature: A Discourse Partnership I made reference earlier to 'relations of expectation' that exist between partners in dialogue and the implications these have for the encoding and decoding of texts. Participants do not have to be 'face to face': Dialogue can be understood in the broader sense, meaning not only direct, face-to-face vocalized verbalized communication between persons, but also verbal communication of any type whatsoever. (Voloshinov 1973:4)

I am concerned with the power relations that exist in the 'verbal communication' between the authors of one particular genre of children's literature and their child readers and the way in which these relations are manifested in language. It is, after all, indisputable that the effects, whatever they might be, which literature might work on children must be mediated largely through the language which constitutes the texts in question. In fiction, the reality-creating potential of language comes to the fore particularly clearly, and writers have a heightened degree of creative licence. It is the aim of this paper to attempt to highlight one of the ways in which adult writers used this degree of linguistic licence when writing for children and the degree to which such writers are themselves constrained by the modes of expression current at the time. In both instances, the writer/reader relationship and current modes of expression, questions of ideology must be central. Modern children's literature has it roots in the mid-nineteenth century as authors turned to writing books for children which provided for entertainment as well as a moral purpose. From the 1840s on, there developed a mass output of popular children's fiction. As the century progressed, stereotypical representations of male and female were increasingly the norm. Books were now produced specifically for girls, using domestic and religious themes, and for boys, where school, games, and Empire were linked; both types are ideologically connected, both were stabilising forces. (Hunt 1994: 52)

My discussion is based upon a corpus of children's literature constructed over the years at the University of Birmingham. There are two sub-corpora and the one relevant to this paper is the Victorian sub-corpus (around 760,000 words) and the texts selected for its composition come from the 'classic' authors of boys' school and adventure stories such as R. M. Ballantyne, G. A. Henty, Thomas Hughes, W. H. G. Kingston, F. Marryat and Talbot Baines Reed. The works produced by these writers represent what I call Traditional Juvenile Fiction. (The other sub-corpus is based on my own survey of children's reading habits in the 1980s and 1990s). In recent research, (see Knowles and Malmkjaer 1996), I was concerned with

Dialogue and Ideology in Traditional Juvenile Fiction

327

ideology and its linguistic representation in children's literature over a period of one hundred and fifty years. Part of this concern had to do with items of vocabulary over a period of time wherein profound social changes have occurred. These changes are reflected in the language in terms of vanished or highly restricted words, and differences in patterns of collocation. As a point of entry, I considered social institutions such as Family, Friendship, Gender, Race etc. and these in turn provide a list of key lexical items or search-words (see below) as a basis for preliminary investigations, notably of collocation.

3. Collocation The notion of collocation is derived from Firth (1957a; 1957b) and since then has been further refined by, among others Halliday (1966) and Sinclair (1966; 1987a; 1987b; 1991). The notion, at its simplest, relates to the tendency of words to cooccur with other words. In carrying out a lexical analysis, sets of concordance lines, in which the node or search word is displayed in centre position, provide the main analytic tool. In obtaining concordance lines, I made use of MicroConcord (Scott and Johns 1993), a computer software program designed to search through large amounts of text to find all the occurrences of a word or words. From this information, concordance lines can be obtained which provide information about the collocational properties of a word or item. MicroConcord provides the researcher with a concordance known as KWIC or Key Word in Context (the search word or node referred to above). Thus, in considering a particular social institution as represented in children's literature such as Family a researcher might decide that his/her search word be mother, or father, brother or sister or, indeed, family itself. Concordance lines will then be obtained for the item under investigation. In early investigations of the corpus, I noticed the high frequency of occurrence of the items British and English in the Victorian sub-corpus. There were 256 occurrences, a ratio of 1: 2,956. In Tom Brown's Schooldays, a narrative of 101,403 words there are 28 occurrences of the item, a ratio of 1: 3,621.1 offer for consideration the concordance lines taken from the Tom Brown's Schooldays file for the particular associations or 'semantic prosodies' (see Louw 1993) of the item English: 1. reat Goddess Thirst, like the 2. out the lesson in Greek and 3. worthy of the name; you 4. is very odd how almost all 5. natural and English way for 6. most likely will be, as you are 7. boys, but to make them good 8. bad foreign music than good

English English. English English English English English English

and French by the stream And now, Arthur, finish const boys for whom this book i boys love danger; you can boys to settle their quarrels. T boys, good future citizens; an boys, good future citizens; and cheeses. Good, perhaps, if it be

328

Murray 9. subduing the earth in most 10. one of you, in any common 11. und and the expression of 12. ch to answer for all through 13. mothers , also; aye, and of 14. ciety, instead of in the old 15. ch to say on the subject of 16. one times. I know only two 17. ty word, but that's the plain 18. to write down eight lines in 19. e question of how to adapt 20. hours of sunshiny life by 21. enham station. If you love 22. foolish or meaningless, as 23. that public schools in the 24. ing, which means in plain 25. write it out in Greek and 26. with fists is the natural and 27. giving them good spirited 28. That great event in the

English English English English English English English English English English, English English English English English English English. English English English

Knowles

counties, and leaving their mar country neighbourhood. Will y country holiday-making; that, i country-sides, but it never play fathers, and sisters, and brothers home duties; because farmers' s mothers, also; and of English fat neighbourhoods thoroughly, an of it. If the chickens were dead of the most matter-of-fact kind public school education to nerv rivers. Every little pitiful coa scenery and have a few hours to schoolboys, at least as the sch sense, are peculiar to England to go to hell, and the other twe Very good, h e ' s caught me, and way for English boys to settle t words, instead of the wretched year, the Derby, was celebrated

These concordance lines demonstrate some interesting language patterns for which more context can be obtained. For example, there are, not surprisingly, six occurrences of the item boy/s or schoolboys. It is a boys' book, written for boys with a large part of the narrative set in a public school. There is the inference of didacticism in the manner in which the you collocates with English boys in line 3. The author is directly addressing his implied reader and it appears an utterance of import is about to be delivered. There is a high percentage of items which connote rural life and the author appears to be making positive evaluations with the collocational partnerships set up between English and cheeses (8); counties (9); country neighbourhood (10); country holiday-making (11); country-sides (12); neighbourhoods (16); sunshiny (20); rivers (20); love (21); scenery (21). Hughes, in fact, is mourning what he sees as the passing of an 'English way of life' and he pronounces on this accordingly. Note also the clustering of the item good in lines 7 and 8. Good English boys in 7 carries the evaluation that to be any other kind of boy is not so good and the occurrences of natural in lines 5 and 26 reinforce this. To be natural is to be English is to be good. The emphasis on the 'positiveness' of English is further substantiated by the antonymous relationship between badforeign music and good English cheeses in 8. There is even linguistic evaluation in 17, 18, 24 and 27 where the inference is that 'good' English' is plain, matter-of-fact, good spirited. There are also the 'home' referents with mothers in 13 and 15; fathers, sisters, and brothers in 13 and home duties in 14. Line 28 adds to the image of the simple, pastoral idyll with which this writer seems to be concerned by adding the notion

Dialogue and Ideology in Traditional Juvenile

329

Fiction

that to be English is to be a sportsman, with the collocates that great event and the Derby. The modern corpus of 862,951 words had only 63 occurrences of English, a ratio of 1: 13,697. There are no public school stories in the modern corpus, so I selected one of Robert Leeson's Grange Hill books which are set in an English comprehensive or state school. Admittedly, this narrative is much shorter than Tom Brown's Schooldays being only 28,029 words in length. There are only two occurrences of English which means a ratio of 1: 14,014. The concordance lines speak for themselves: 1. next language. So, they speak 2. Claire. 'You know Luigi's

English English

here. I've news. It's another language isn't all that good. You're taking

4. Collocates of Collocates The rather simplistic nationalism of the 1840s became, by the 1880s (and persisted well into the twentieth century) the ideology of imperialism. The concepts of bravery, truthfulness, faithfulness to one's friends and companions were very much a part of the traditional writer's ideological baggage. The lexical set friend, companion, comrade, proved to be important nodal items through their frequency of occurrence and the collocates obtained from considering these items, provide significant lexical information about Victorian boys' narratives. It is these collocational patterns that assist in identifying semantic change. Amongst the most frequent of the collocates were: Brave, Gallant,

Stout

In this set, while all items will be recognised by the modern reader, brave can be considered the core word. Stout, in the context of courage, is highly marked and gallant is also marked as being somewhat old fashioned. Brave is required to interpret the other two items and bravery as well as nationalism was a concept much promoted by the Victorian writers for boys as a positive virtue. Halliday's (1966) paper 'Lexis as a Linguistic Level' is seminal in the study of vocabulary with its focus on how collocation can be used to generate lexical sets. Thirty years on from that paper, we have available, as I have demonstrated, the insights of corpus-based, computer-aided techniques which permit the speedy and precise investigation of language patterns in text. I decided, therefore, to access my data using the items in the set above, the collocates of friend, companion, comrade, to see if the 'collocates of collocates' could provide further information on linguistic representation in Traditional Juvenile Fiction. From the concordance lines obtained for the lexical set - brave, gallant, stout - it was immediately noticeable that there was a high frequency of occurrence of the item

Murray

330

Knowles

fellow. There turned out to be 588 occurrences of fellow in the Victorian corpus. The item was then checked against the Oxford English Dictionary so that an impression of its usage in Victorian context might be gained. OED had 6.5 columns for fellow. Amongst the contextualised nineteenth century entries were the synonyms companions, associates, comrades. Furthermore, I was informed that it is often used with 'good and jolly' as in 'an agreeable or pleasant companion' (OED 1971: 980). Additional information given was that fellow was a familiar synonym for man and that it was commonly used with the following qualifying adjectives: good, bad, brave, clever, foolish old, young. Phrases like 'what a poor fellow' were often used exclamatorily as an expression of pity and ' M y dear fellow' was common in familiar address. Mention was also made of the item as a customary title of address to a servant 'or other person of humble station' (ibid: 980) and that it was applied by schoolboys to themselves and to each other. All of these combinations occur frequently in the Victorian corpus with the following items being the most frequent collocations generated: poor old little good young dear big black

71 41 28 28 25 16 13 13

occurrences occurrences occurrences occurrences occurrences occurrences occurrences occurrences

It is clear that here is an item which had some considerable cultural significance in the nineteenth century and I thought it interesting to compare with late twentieth century usage. There are only 24 occurrences (i.e. roughly 4%) of fellow in the Modern fiction corpus. The two most frequent collocates in -1 position (one word to the left offellow) are given below: little young

4 occurrences 4 occurrences

There was one occurrence of poor in -2 position (two words to the left of fellow) as in: poor, little fellow. Furthermore, in one of the occurrences of young fellow, the combination was in spoken English where the reference was to a possible boyfriend. It is in informal spoken contexts that the item is probably most frequent in contemporary usage where it means boyfriend or lover. Fellow as a nodal item in the modern corpus did not generate anything like the frequency or range of collocates that it did in nineteenth century boys' books. In addition, when I

Dialogue and ideology in Traditional Juvenile Fiction

331

referred to the COBUILD dictionary fellow was given as 'rather old-fashioned English' when used in expressions such as My dear fellow (1987: 526). Other COBUILD information was also interesting. The extra column provides additional semantic information on entries. It gave comrades as a synonym offellows in the sense of 'people who share work or an activity with you or who are like you in some way'. Associate was also given as a superordinate. The entry for comrade listed friend and companion, especially in terms of sharing difficulties or dangers. The entry also noted that, like fellow , it is 'a rather old fashioned word' (ibid: 287). So, while there are links with past usage fellow is an item which does not translate easily into contemporary usage being now very restricted in range and context. In the Victorian context, however, I can, via my consideration of collocates of collocates, add this item to my original set of nodal items (see above). So, along with friends, companions and comrades, fellows is a significant lexical item within the semantic field of 'maleness'. It also generated another wider set of significant items, as I noted in the list of collocating adjectives above. It is one of these that I now consider. In the list just referred to, the item poor is the most frequent collocate of fellow in the Victorian corpus. There are 71 occurrences both in direct and indirect speech. The fact that poor was collocating very frequently with the male item fellow seemed to me to be particularly significant. Preliminary investigation indicated that this pairing signalled a strong indication of impending calamity in the narratives. I decided, therefore, to examine the collocates generated by poor fellow as node. I treat the combination as a fixed item. There is a total of 117 lexical items generated by the collocates of poor fellow. These are presented in sets in figure 1 on the next page. Items specifically referring to the processes of dying or being killed are listed in column 1. They include verbs and other words to be found in the semantic field of death. Items which refer to other physically unpleasant processes are listed in column 2. Items which carry additional connotations of suffering are listed in column 3. Of the items left after the categorisation of the information in figure 1, the following are significant: saved, lifted (to safety), hold, tend, lives x2, survivors, God, prayers x2, water x3, cooling, beverage. All these items could well be evaluated as positive. Indeed, they appear to fall into three naturally occurring lexical sets: a. saved, lifted, hold, tend, lives, survivors b. God, prayers c. water, cooling, beverage

332

Murray

1. Die x2 Shot x2 Hit x3 Never spoke again Is fast enough (dead) Lost x2 Sunk x2 Torn from (killed) Spring up (having been shot) Burying Covered up (a body) Nothing to be Done Sink under Lay (in death) Gone Never to rise again Fallen Killed outright x3 Killed Slaughtering Stark Cold Dead x3 Body

2. lose heart groans x2 crying x2 bleeding blown off ill-treating wounding washed off maimed wounded racked pain wounds groans cries (of agony) bound up agony fever

Knowles

3. Steal out Leave alone Hopelessness Dreadfully Escaped (from one danger to another) Low spirited Homesick Left (abandoned) Leave alone Clinging Compelled Fear Distress Fate (of poor fellows being slaughtered)

Figure 1 : Collocates of Poor Fellow These three sets are antonymous to the items categorised in figure 1 above. As positive evaluation they are far outnumbered by lexical signals of negative attribution. It would seem, therefore, that in Victorian usage poor as a modifier of fellow was certainly an expression of pity. Within the context of literature written for children, as manifested in my corpus, this was most often associated with death. The lexical sets categorised above all arose from consideration of the institution of Friendship where fellow asserted itself to be a culturally significant item. I have already noted its infrequent usage in modem children's literature. A final useful point of comparison can be made with twentieth century usage as contained in the COBUILD Bank of English corpus. There are 9,520 concordance lines containing/e//ow in the COBUILD corpus. Poor occurs 22,293 times in 167 million words (the size of the corpus at the time of this study) and collocates with fellow on 57 occasions in -1 position. Of the occurrences of poor fellow, 19 are from the British books section of the COBUILD corpus which accounts for 23 million words. Overall, poor does not appear in the list containing the 22 most frequent collocates o f f e l l o w .

Dialogue and Ideology in Traditional Juvenile

Fiction

333

5. Conclusion Collocation can be a very useful tool in interpreting the dialogue of narrative text. Naturally, concordance lines can only be used as part of an interpretative effort and it is necessary, if one wants to make any strong claims concerning ideology in narrative, to consider substantially more text than they provide. Sets of concordance lines are, however, useful starting points for further textual searches, or as checks against claims one might wish to make. With computational techniques valuable data can be obtained relatively quickly which is especially useful when you are attempting to account for patterns of language in large corpora. The lexical analysis with which this paper has been concerned focused upon collocational patterns as shown by my Victorian corpus. Firstly, there was the rural nationalism of mid-nineteenth century England as exemplified in the collocates of English. This is hardly a typical association for the item in the late twentieth century. Secondly, the emergence of fellow as a key item is a good example of how, in a relatively short space of time, a word can cease to be culturally or ideologically significant. In the traditional juvenile story, fellow can be seen as a synonym for friend, companion, comrade. Perhaps it is better considered as a general word or superordinate within which specific degrees of friendship are included. These reflect social intimacy, class and ethnic diversity, and the attitudes expressed towards them. Such a lexical analysis also reveals an underlying ideology from the point of view of the young male hero. The perils of nature (storms at sea, for example) and the possibility of death in battle are part of the 'real world' of Victorian masculine activity. The assumption is that the hero's eventual return to his family can be seen as a just reward for doing his 'just duty' in upholding the social order as enshrined in the pax Britannica. Exposure to such dangers is very much a part of the young male's moral training which commenced at public school at home in England. In terms of reader/writer interaction then: The reader perceives the world of the text and the events which occur in it from the hero's point of view, or the point of view of a narrator who admires him and places him in the foreground, so that the story imposes his perspective and his evaluations. Therefore one of the overriding meanings which readers construct from these stories is that it is the hero who is of primary importance and the activities of such men that matter in the world: (Hourihan 1997: 38)

Finally, in any text, as a writer must use what is given to construct something that may be new, a reader can only assimilate the new within what is already known. Language is, after all, a social phenomenon and the relationship between language and literature is a social practice engaged in by a writer with a reader in mind.

334

Murray Knowles

What the child reader takes on board from any one socialising agent will be greatly influenced by what s/he has already received from other agents. Literature is a very important socialising agent; it is, however, only one amongst many.

References COBUILD (1987): Collins COBUILD English Language Dictionary. Glasgow: HarperCollins. Firth, J. R. (1957a): Papers in Linguistics 1934-1951. Oxford: Oxford University Press. Firth, J. R. (1957b): A Synopsis of Linguistic Theory 1930-55. (Special Volume of the Philological Society), Oxford: Basil Blackwell. Halliday, M. A. K. (1966): Lexis as a linguistic level. In: C. Bazell, J. Catford, M. Halliday and H. Robins (eds.), In Memory of J.R. Firth, London: Longman, 148-62. Hourihan, M. (1997): Deconstructing the Hero: Literary Theory and Children's Literature. London: Routledge. Hughes, T. (1857): Tom Brown's Schooldays. (1880) London: Macmillan. Hunt, P. (1988): Degrees of Control: Stylistics and the Discourse of Children's Literature. In: N. Coupland (ed.) Styles of Discourse, London: Croom Helm, 163-82. Hunt, P. (1994): An Introduction to Children's Literature. Oxford: Oxford University Press. Knowles, G. M. and K. M. Malmkjaer (1996): Language and Control in Children's Literature. London: Routledge. Leeson, R. (1985): Reading and Righting. Glasgow: HarperCollins. Louw, B. (1993): Irony in the text or insincerity in the writer? - The diagnostic potential of semantic prosodies, in M. Baker, G. Francis and E. Tognini-Bonelli (eds.) Text and Technology: In: Honour of John Sinclair, Amsterdam: John Benjamins, 157-76. Oxford English Dictionary (compact edition) (1971): Oxford: Oxford University Press. Reynolds, K. (1994): Children's Literature in the 1890s and the 1990s. Plymouth: Northcote House. Scott, M. and T. Johns (1993): MicroConcord. Oxford: Oxford University Press. Sinclair, J.McH. (1966): Beginning the study of lexis. In: C. Bazell, J. Catford, M. Halliday and R. Robins (eds.), In Memory of J.R. Firth. London: Longman, 410-30. Sinclair, J.McH. (1987a): The nature of the evidence. In: J. McH. Sinclair (ed.), Looking Up: An Account of the COBUILD Project in Lexical Computing. Glasgow: HarperCollins, 150-9. Sinclair, J.McH. (1987b): Collocation: A progress report. In: R. Steele and T. Threadgold (eds.), Language Topics: Essays in Honour of Michael Halliday, vol. 3. Amsterdam: John Benjamins, 319-31. Sinclair, J.McH. (1991): Corpus, Concordance, Collocation. Oxford: Oxford University Press. Simpson, P. (1993): Language, Ideology and Point of View. London: Routledge. Voloshinov, V. I. (1973): Marxism and the Philosophy of Language. New York: Seminar Press.

Dorota Pacek

Wordplay and Nonsense in Polish Translations of Dialogues in Alice's Adventures in Wonderland 1. Introduction This paper is concerned with the strategies used by several translators of Alice books in translating dialogues into Polish. For several reasons, dialogues are one of the most striking features of the books. Firstly, there is a wealth of puns, neologisms, nonsense and portmanteaux words, all of which create specific problems for translators. Carroll plays with the language, exploiting the fact that some words have several different meanings. This leads in many instances to the creation of nonsense dialogue and apparent lack of logic - as Alice says, some characters sound 'perfectly idiotic'. In this article I shall analyse the ways in which four post-war Polish translators deal with some of the stylistic difficulties of Carroll's text from the point of view of compensation and translation norms. 1.1.

Background Information

The story of Alice's adventures was told by Carroll to the Liddell sisters for the first time in July 1862 during a boat trip near Oxford and Alice's Adventures in Wonderland was first published in its present form in 1866. The edition used for the purpose of this paper is the 1970 The Annotated Alice edited by Martin Gardner. Alice has been translated into many languages and in many cases several translations exist. The reason for so many attempts is the fact that Alice is a true challenge for translators. Not only is it full of various types of plays on words, verses, parody and nonsense, it has also now become a classic, a book which, although meant for children, (and more precisely for one specific child), is nowadays more often read and appreciated by adults. As Phillips points out, although the Alice books were written "originally for children, [they] have transcended that audience. Like all true classics, they are multileveled and continue to entertain one with subsequent readings" (Phillips 1972: xxii). Since Alice nowadays functions on two levels - as a book for children and as a book for adults, and since it is often more appreciated by the adults (i.e. not by the 'intended audience') there are specific difficulties in rendering it into a foreign language. Ideally the translation should retain its 'childlike' quality, that is amuse,

336

Dorota Pacek

be exciting and easy for children to read, but it should also preserve those characteristics that are most appreciated by adults, such as clever paradoxes, nonsense, unusual twists of meaning, parodies and puns. This 'dual personality' of Alice books may influence the translator's decision about his translating strategy; he might choose to render it as a book for children, or as a book for adults, or as a book that both children and adults would enjoy. The choice of approach will depend on many factors, one of them being translation norms and traditions in a given target culture, an issue to which I shall return later.

1.2. Polish Translations To date there have been six translations of Alice's Adventures in Wonderland into Polish. Here I will only discuss the latest four translations, as the first two, from 1910 and the 1930s are now unavailable. The four post-war translations were done in 1955 by Marianowicz, in 1972 by Slomczynski, in 1986 by Stiller, and in 1997 by Kozak.

2. Alice Books and Compensation Texts, depending on their purpose, target audience, and genre, fall on a cline from those where content is the most important element, and style of little or no importance, to those where the stylistic means used to convey a message constitute a prominent, and in some instances the most important, feature of the text. Alice books are one of the best examples of the second category, especially if we take into account dialogues which are particularly full of puns and other stylistic devices. Any translation, and literary translation in particular, is a process where losses are unavoidable. Losses can be of many different kinds, but can be subsumed under the two general categories of meaning/content and form/style. It is often said that meaning is more important than form, and since it is often impossible to retain both, meaning should have priority over form. This might be true for technical and other run-of-the-mill texts, but such a view has little validity in the case of imaginative, creative literature. Often, in such texts, the form is the meaning, or is a means for creating special, unusual meanings. In this paper, I will only concentrate on the imaginative and creative language of dialogues in Alice, and the ways Polish translators have coped with the difficulties of preserving the style, as the more stylistically loaded the original text, the more it resists translation. Therefore, I would now like to turn to the notion of compensation as one of the devices which allow translators to minimise translation losses in terms of style.

Wordplay and Nonsense in Polish Translations of Dialogues in 'Alice '

337

3. Compensation Compensation is not a well-defined and well-understood concept, and there are many conflicting definitions. Moreover, the literature shows that the term 'compensation' is often used in the every-day, common sense meaning of the word. There are several problems with most existing definitions of compensation. They could be summarised as follows: 1.

2. 3.

There is no agreement as to what features of translation this concept should comprise. If it is to include all the possible losses of lexical, syntactic, stylistic, cultural, and pragmatic types, then the concept is too general and all-embracing to be of much use to translation theory or practice. None of the definitions give any indication of what specific procedures can be considered as compensatory. There is either a lack of concern with the problem of where compensation can occur, or it is suggested that compensation can only appear in TT at some distance from the point where a specific effect or text feature is lost in ST.

The most recent and exhaustive analysis of the concept of compensation is presented by Harvey (1995). Harvey outlines the problems involved in defining compensation by saying that an analysis of examples given by various scholars "reveals that practically anything that did not involve straightforward correspondence was subsumed under this label" (Harvey 1995: 67). Since compensation embraces too many possible aspects of mismatch between two languages, he limits this notion to "stylistic, text-specific features and effects" (ibid.: 71), and excludes systemic, lexical or cultural transfers. He holds that stylistic effects are not system-bound, but text-bound, i.e. although they exist in a language system in general, they are employed for "stylistic purposes within a specific text" (ibid.: 75). The investigation of compensation in the case of all the translations analysed here will be based on Harvey's framework of compensation, which is summarised in table form on the next page. From the point of view of this paper, the most interesting and important category in Harvey's framework is that of parallel compensation, as it has been explicitly excluded by most scholars - they either ignore the issue of location, or only allow the possibility of contiguous or displaced/generalised compensation. However, in Polish translations of serious literature, including Alice, the most common type of compensation is the parallel one.

Dorota Pacek

338 Type

Sub-type Stylistic

Typological correspondence Stylisticsystemic

Definition Effects in ST and TT are text-specific and contribute uniquely to the colour, tone and register of that particular text. The effects have a stylistic value where they occur in text, but these draw upon part of conventional systemic resources of the language.

Example Neologisms Phonoaesthetic devices

Idioms Exploration of grammar for rhetorical effects

Linguistic device in ST compensated for by the use of the same type of linguistic device in the TT

Pun by pun, phonoaesthetic device by another

Analogical

Linguistic devices in ST and TT different, but both metalinguistic in character.

Noncorrespondence

ST and TT effects share no linguistic features at all.

Dialect by puns, reflexive verbs by possessive pronouns and adjectives Neologism by conventional idiom, tense system vs lexical borrowing

Parallel

Compensation occurs at exactly the same place in ST and TT.

Contiguous

Compensation occurs in TT a short distance from the lost effect of the ST Long distance. Whole text as Unit of Translation. Compensation not tied to any specific instances of ST loss, but number and quality equal; to naturalise TT.

Direct Linguistic correspondence

Topographical correspondence

Displaced Generalised

Figure 1: Harvey's framework (1995)

4. Translation Analysis - Dialogue One I have chosen three different dialogue extracts, to exemplify some of the types of difficulties that the Polish translators faced.

Wordplay and Nonsense in Polish Translations of Dialogues in 'Alice '

339

Firstly, I will look at the well-known conversation between Alice, the Mock Turtle and the Gryphon, where the Gryphon and the Mock Turtle describe to Alice the school subjects they were taught in their youth. Here is the sample text: "Reeling and Writhing, of course, to begin with," the Mock Turtle replied; "and then the different branches of Arithmetic - Ambition, Distraction, Uglification and Derision". "I never heard of Uglification," Alice ventured to say. "What is it?" The Gryphon lifted up both paws in surprise. "Never heard of uglifying!" it exclaimed. "You know what to beautify is, I suppose?" "Yes," said Alice doubtfully: "it means to-make-anything-prettier." "Well then," the Gryphon went on, "if you don't know what to uglify is, you are a simpleton." [...] "What else had you to learn?" "Well, there was Mystery," the Mock Turtle replied, [...], - "Mystery, ancient and modern, with Seography; then Drawling — the Drawling-master was an old conger-eel, that used to come once a week: he taught us Drawling, Stretching, and Fainting in Coils."

This dialogue is exceptionally dense in terms of puns, which are built around a list of real school subjects which, because they are taught under the sea, are replaced by other nouns, based on paronymy. The density of puns suggests that this extract might be particularly difficult for translators. However, because all these puns are of the vertical type (paradigmatic in linguistic terms), i.e. they are realised in a single word, they created little problems for the translators - it was just a question of finding Polish words which are similar in sound and spelling to Polish school subjects. As these puns are of less interest in terms of translation difficulties, I have put examples in the Appendix. Here, I want to concentrate on the final pun of the dialogue, as it seems to have created most problems for the translators. The last part of the conversation (see below) is more challenging, as it is based on homophony, and represents a horizontal type of pun (paradigmatic in linguistic terms), i.e. the words in question (lesson-lessen) are "in a relation of contiguity by occurring one after another in the text" (Delabastita 1996: 128). As mentioned above, this pun is based on homophony, where two English words have to be substituted by different words in Polish. "And how many hours a day did you do lessons?" said Alice [...]. "Ten hours the first day," said the mock Turtle: "nine the next, and so on." "What a curious plan!" exclaimed Alice. "That's the reason they're called lessons," the Gryphon remarked: "because they lessen from day to day."

Stiller: "A ile mieliscie jednego dtua lekcji?" spytala Alicja, zeby czym prgdzej zmienic temat. "And how many lessons did you have each day?" asked Alice to change the subject quickly.

Dorota Pacek

340

"Pierwszego dnia dziesiec - odpowiedzial Faiszywy Zotw - a nastepnego dnia dziewigc i tak dalej." "Ten the first day - answered the Mock Turtle - nine the next and so on." "Jaki to dziwny plan zajec!" wykrzykn^la Alicja. "What a strange timetable!" exclaimed Alice. "Dlatego to nazywa sie lekcje - wyjasnii Gryfon - bo codziennie jest lekcej i lekcej." "That's why they are called lessons - explained the Gryphon - because it's easier/lighter and easier/lighter from day to day."

Stiller followed Carroll's pun very closely, translating almost word by word, thus not employing compensation but direct translation. He achieved this by utilising the word 'lekcej', the comparative form of the adjective 'lekki' (light, easy), which is a rather archaic form, only occasionally used nowadays in some regional dialects. However, it works well in the context, and does not strike the reader as an 'incorrect' form, but rather brings an additional humorous effect into the pun. Here, the translator was to some extent 'lucky' that there exist in Polish two words that are similar in meaning to the English words used by Carroll, and equally similar in terms of sound and spelling. Such similarities between two languages are usually purely accidental. Kozak follows the same line as Stiller in her translation, so the comments above apply equally to her version.

Slomczynski: Slomczynski employed a different pun, hinging on the homonymy of the first word 'pi^tki', which is plural dative from the word 'piec' i.e. five, and the second word pi^tki' which is plural nominative for 'Fridays'. Although he replaced Carroll's pun by his own, thus employing parallel compensation, it follows the logic of Carroll's writing and stretches over several utterances, giving it cohesion and coherence in a manner akin to Carroll's play on words. - A miewaliscie piqfld? - zapytala pospiesznie Alicja, chcac szybko zmienic temat. - And did you have/get fives ('five' is the equivalent of 'A' grade in British schools)? asked Alice hurriedly, to change the subject. - Tak, ponad pigcdziesiqt rocznie, cztery miesiecznie i raz w tygodniu. - Yes, more than fifty per year, four per month and once per week. - Zawsze raz w tygodniu ? - zawolala zdumiona Alicja. - Always once per week? exclaimed surprised Alice. - Oczywiscie, ze tak - powiedzial Gryf - przeciez jest tylko jedenpiqtek w tygodniu. - Obviously yes - said the Gryphon - there is just one Friday per week.

In Carroll's dialogue the pun is not only amusing in itself, but also a pivotal element of the whole conversation, stretching over several utterances and thus retaining cohesion of the rather nonsensical dialogue. Slomczynski's translation works well, as it not only compensates for the lost pun by a different one, but also, as in the case of the original, leads to a quite nonsensical exchange of utterances.

Wordplay and Nonsense in Polish Translations of Dialogues in 'Alice '

341

In this respect his translation is more successful than that of Marianowicz, as it follows more closely the 'spirit' and style of Carroll's writing. Marianowicz replaced the whole dialogue by his own, where the pun under consideration is replaced by two different ones, one of which is based on paronymy and one on homophony. So although he does compensate for loss, much of the cohesion of the extract, as well as its contrived logic, are lost in the process:

Marianowicz: - A czy mieliscie czesto wypracowania? zapytata Alicja [...] - And did you often have essays (to write)? asked Alice.

-1 owszem, mielismy wyprasowania domowe mniej wiecej raz na tydzien - odrzekl Smok. - Certainly, we had home ironing more or less twice a week. [...]

- A jak u was bylo z cwiczeniami? - And what about exercises?

- O, swietnie, znakomicie. Zapewniam cie, ze cwiczen nam nie braklo. Bylismy cwiczeni przy kazdej okazji - odparl dumnie Niby Zobv. - Oh, super, excellent. I assure you that we did not lack in exercises. disciplined/drilled on every occasion - proudly answered the Mock Turtle.

We were

Marianowicz employs this type of translation quite often. Although his translation has been criticised, especially by Stiller, for not being very faithful, especially in terms of meaning/sense, his solutions are always very witty and sometimes even are more humorous than the original.

5. Translation Analysis - Dialogue Two Another dialogue I would like to examine here follows the one discussed above, and is also between Alice and the Gryphon and the Mock Turtle. It is perhaps even more challenging for translators than the previous one, as it consists of a series of different puns, on different words, all of them based on homophony and paronymy and leading to some nonsensical exchanges: [...] "I never knew so much about a whiting before." "I can tell you more than that, if you like," said the Gryphon. "Do you know why it's called a whiting?" "I never thought about it," said Alice. "Why?" "It does the boots and shoes" the Gryphon replied very solemnly. Alice was thoroughly puzzled. "Does boots and shoes!" she repeated in a wondering tone. "Why, what are your shoes done with?" said the Gryphon. "I mean, what makes them so shiny?" [...] "They are done with blacking, I believe."

Dorota Pacek

342

"Boots and shoes under the sea," the Gryphon went on in a deep voice, "are done with whiting. Now you know." "And what are they made of7" Alice asked [...] "Soles and eels, of course," the Gryphon replied [...]. "If I'd be the whiting," said Alice, [...] "I'd have said to the porpoise, 'Keep back, please! We don't want you with us!'" "They were obliged to have him with them," the Mock Turtle said. "No wise fish would go anywhere without a porpoise." "Wouldn't it really?' asked Alice [...] "Of course not," said the Mock Turtle. "Why, if a fish came to me, and told me he was going a journey, I should say 'With what porpoise?'" "Don't you mean 'purpose'?' said Alice. "I mean what I say," the Mock Turtle replied [...]

Marianowicz: [...]- Dotychczas nigdy jeszcze nie dowiedziaiam sie tak wielu rzeczy o rybach. - Until now I've never learned so many things about fish. - [...] Na przyklad, jak ci sie zdaje, dlaczego ryby tak swietnie tatkza? [...] For example, why do you think fish dance so well? - Doprawdy nie wiem - odpowiedziala Alicja. - Dlaczego? -1 really don't know - answered Alice. - Why? - Dlatego, ¿e SQ niezwykle wysportowane. Widac to wyraznie, kiedy sie sledzi ich ruchy. - Because they are very agile. It is quite clear, when you watch/observe their movements. - Czyje ruchy? - zapytala Alicja, gubiqc wqtek rozmowy. - Whose movements? - asked Alice, loosing the plot. - Sledzi, rzecz jasna. Wez pod uwage, ¿e ryby cwicza sie bez przerwy w biegach przez ptotki i w cwiczeniach na linach. - Herrings', obviously. Consider, that fish constantly practice hurdle racing and the tightrope.

[...]

- Tak, tak - wtrocii Niby Zotw - wiele ryb dokazuje istnych cudow w tej dziedzinie. Na przyklad minogi... - It is so - added the Mock Turtle - many fishes work wonders in this area. For example lamprey... - Wlasnie - rzekl Smok. - Taniec wyrobil mi nogi zupelnie nadzwyczajnie. [...] - Exactly - said the Dragon. - Dancing developed/trained my legs extraordinarily. [...] -... czy okonie... - ciognal [• • •] Niby Zolw. - ...or perch... - continued the Mock Turtle. - Oko nie - rzekl smutnie Smok. - Niestety, taniec nie wyrobil mi oka. [...] - Eye no - said the Dragon sadly. - Unfortunately, dancing did not train my eye. [...]

All original puns, including the one about porpoise, which is one of the best remembered and liked of Carroll's puns, are replaced by a different sequence of punning. It works very well, as all the puns are based on Polish names of different fish, which helps to maintain cohesion of this sequence. Also, the punning section refers back to the conversation about dancing, between Alice, the Mock turtle and

Wordplay and Nonsense in Polish Translations of Dialogues in 'Alice '

343

the Gryphon, which directly precedes this extract. This also adds to the overall cohesion of the text. Moreover, the whole extract, but especially the exchange with the word sledzi (herring/observe), is nonsensical, witty, and sounds even more absurd that the equivalent section of Carroll's text. It might be argued that it results in 'overtranslation'. However, it would seem that it is a conscious compensatory strategy on the part of the translator, as in some other parts of Carroll's text he 'undertranslates' either by omitting some fragments of text, or by translating only sense-by-sense, without retaining the play on words. Stomczynski: [...]- Czy wiesz skqdpochodzi nazwa bialoryba? [...] - Do you know here the name 'whiting' conies from? [...] - Bo sluzy ona do czyszczenia but&w i pantofelkow - powiedziai Gryf bardzo powaznie. - Because you use it for cleaning shoes and court shoes/paramecium. - said the Gryphon. [...] Do czyszczenia butow? -powiedzialazdumionym glosem. [...] For cleaning shoes? - she said in a wandering tone. - No, czym sq wyczyszczone twoje buciki? Chc% powiedziec, co powoduje, ze tak Isniq? - Why, what are your shoes cleaned with? I mean, what makes them so shiny? [...]- Smaruje siqje czernidlem, prawda? [...] - They are done with blacking, right? - Buciki i pantofelki w glqbinach morskich - ciqgnqi Gryf niskim glosem - bieli siq bialorybq. [...] - Shoes and court shoes/paramecium in the sea - went on the Gryphon in a deep voice - are whited with whiting. [...] - A kto czysci te buciki? [...] - And who cleans those shoes? [...] - Tylko flqdry ich nie czyszczq [...] - Only plaice do not clean them. [...] - Gdybym byla bialorybq - powiedziala Alicja, [...]- powiedzialabym do Morswina: "Nie nadeptuj mi na ogon, nie chcq twojego towarzystwa w drodze do wody ". - If I were the whiting, I'd say to porpoise: "Don't tread on my tail, I don't want your company while travelling to water". - Musialaby jednak przedstawic te dowody - powiedziai 2obviciel. - She would have to present those proofs/evidence. - Jakie? -powiedziala Alicja nie zrozumiawszy. - What (proofs)? said Alice, not understanding. - Oczywiscie -powiedziai ¿ohviciel. - Gdyby ryba przyszla do mnie i powiedziala: "nie lubi% towarzystwa morswina w drodze... ", spytalbym dokladnie tak samo jak ty: "dowody? " - Of course - said the Mock Turtle. - If a fish came to me and said: "I don't like porpoise's company while travelling...", I'd ask exactly the same question as you: "(What are your) Proofs?' - Alez ja powiedzialam: "do wody "! - But I said: "to water". - Przeciezwlasnie to mowiq! - powiedziai Zobviciel urazonym tonem. - But that's exactly what I said! said the Mock Turtle in an offended tone.

344

Dorota Pacek

Throughout the book, Slomczynski's strategy is to keep as close to the original as possible. This is also what he has done in the first part of this extract. He is able to follow very closely Carroll's puns due to the fact that the Polish equivalent of the word 'whiting' is very similar in Polish - literally 'whitefish'. This allows Slomczynski to translate the first series of puns almost word-forword, and he resorts to compensation only once. Although this section of the dialogue is rather 'heavy' in translation, and less funny than the original, Slomczynski used one ingenious idea to compensate for the lost effect of the 'soles and eels' pun. Instead, in translation, the Gryphon replies to Alice's question that 'only plaice do not clean' their shoes. In Polish the word 'flgdra' (plaice) is often used in colloquial language to refer to a sloven or a slob. The compensatory strategy used here works well, as it creates a truly humorous effect in this extract. The second series of punning (porpoise-purpose) is replaced by a different one, also based on homophony. Here, the pun is based on the phonological identity of the words 'dowody' (proofs) and 'do wody' (to water). This creates a similar chain of puns as in the original and compensates well for the loss of the original pun.

6. Concluding Remarks From the analysis of these Polish translations of Alice it is clear that different translators use different translation strategies. They either translate directly, almost word by word, but only if such a translation allows for the preservation of original stylistic features, or resort to compensation if direct translation is not possible. The treatment of wordplay depends on translation norms in a given society, as according to Toury (1995), norms are one of the main factors determining translators' strategies in rendering the text in the target language. Translation norms can impose the rule of adequacy, where the maximum preservation of the features of the source text is required, or the rule of acceptability, where the source text can be modified to some extent to fit the social, literary and linguistic conventions and expectations of the target culture. In Poland, the situation seems to be that texts considered to be high literature in general follow the norm of adequacy, while popular literature follows the norm of acceptability. As far as Alice's translations are concerned, all the translations seem to have followed the rule of adequacy in preserving the original features exactly as they appear in the original, or by compensating for any losses there and then. Slomczynski and Stiller both strive to keep as close to the original as possible, with Slomczynski going to extremes in his efforts to preserve all the textual

Wordplay and Nonsense in Polish Translations of Dialogues in 'Alice '

345

features of the original, thus exoticising the text to an extent greater than any other of the translators. Both his and Stiller's translations seem to be more suitable for adult readers, as by attempting to account for all textual features as closely as possible, they very often create Polish that sounds contrived, is barely acceptable stylistically, and where puns are sometimes difficult to follow on first reading. Marianowicz and Kozak, on the other hand, seem to be less concerned with exact following of the original and translate more freely. Although in many instances they do follow Carroll's version closely, in other cases they replace Carroll's puns and other stylistic devices with their own, more in agreement with the rules of the Polish language. Thus they naturalise the text for the Polish reader, and also create much lighter, more 'readable' versions, which are equally enjoyable for children and adults. In any case, all translators attempt to account for all stylistic features by using parallel compensation where direct translation is not possible. Marianowicz is the only translator who employs contiguous and displaced compensation as well, as he sometimes does not account for Carroll's stylistic devices where they occur in the original. It would seem then, that although parallel compensation has been explicitly excluded by most translation scholars, it is in fact the most common type of compensation occurring in Polish translations of high literature, which is the result of translation norms and expectations prevalent in Polish society.

References Carroll, L. (1955): Alicja w Krainie Czarow. Warszawa: Nasza Ksiegarnia. [Translated by Antoni Marianowicz]. Carroll, L. (1972): Przygody Alicji w Krainie Czarow. Warszawa: Czytelnik [Translated by Maciej Slomczynski]. Carroll, L. (1990): Przygody Alicji w Krainie Czarow. Warszawa: Lettrex. [Translated by Robert Stiller], Carroll, L. (1997): Alicja w Krainie Czarow. Warszawa: ABC Future. [Translated by Jolanta Kozak]. Delabastita, D. (1996): Introduction. The Translator 2/2. Special edition D. Delabastita (ed.) Wordplay and Translation, 127-139. Gardner, M. (1970): The Annotated Alice. Penguin Books. Harvey, K. (1995): 'A Descriptive Framework of Compensation'. The Translator 1/1, 65-86. Phillips, R. (ed.) (1972): Aspects of Alice. London: Victor Gollancz Ltd. Toury, G. (1995): Descriptive Translation Studies and Beyond. Amsterdam/Philadelphia: John Benjamins Publishing Company.

Dorota Pacek

346

Appendix Solutions to some vertical puns employed by four Polish translators. Carroll Reeling

Marianowicz Zgrzytanie (grating/rasping)

Siomczyriski Chlapecadlo (chlapac=splash; abecadlo=alphabet)

Stiller Czyhanie (lying in wait)

Kozak Czmychanie (fleeing)

Writhing

Zwisanie (hanging)

Portografia (port + ortography)

Zwisanie (hanging)

Prychanie (snorting)

Ambition

Podawanie (giving/passing)

Wodowanie (launching)

Wodowanie (launching)

Dogadzanie (pleasing)

Distraction

Obejmowanie (embracing)

Obejmowanie (embracing)

Obejmowanie (embracing)

Obgadywanie (gossiping)

Uglification

Mrozenie (freezing)

Dnozenie (multiplication + bottom)

Mrozenie (freezing)

Ci?zkowazenie (heavyweighting)

Derision

Gdzielenie (g+division)

Brzydzielenie (analogical)

Obrzydzielenie (analogical)

Gl?dzenie (prattling)

Mystery

Zimnastyka (cold + gymnastics) etc.

Histeria (hysteria)

Histeria

Histeria

etc.

etc.

etc.

Etc.

In Polish: Reading = czytanie Writing = pisanie Addition = dodawanie Subtraction = odejmowanie Multiplication = mnozenie Division = dzielenie

Josiane Boutonnet

Verbal Humour as the Ultimate Level in Language Learning: Accessing the Contextual Properties of Humour in Everyday Conversation 1. Introduction Verbal humour is a form of play which can break the rules of the linguistic system, the pragmatic conventions for language use, or both. According to Chiaro (1992) verbal humour includes every conceivable way in which language is used with the intent to amuse and Ross (1998) defines humour as something that makes a person smile or laugh. Nash (1985) describes humour as a serious business which takes many forms, whereas Palmer (1994: 3) includes 'everything that is actually or potentially funny, and includes the processes by which this funniness occurs'. In this paper I explore verbal humour as it occurs in everyday conversations and the complex processes required for the participants to engage in this activity. The ability to produce and understand verbal humour does seem to form part of our communicative competence (Hymes 1972) as a native speaker. However, it cannot be assumed a priori as Alexander (1997) posits that verbal humour is an intuitively grasped aspect of human social communication, even though in many children and adult native speakers of a language, its mastery comes to form part of their general language competence. Humour competence as defined by Attardo (1994) is nonetheless learned in social contexts, irrespective of the fluent speaker's awareness of the processes by which this has occurred. Intuition requiring immediate perception or knowledge would ignore the complex work required of the individual to reconstruct the 'intent' of a proposition. A number of studies provide ample evidence for a developmental path for humour acquisition in preadolescent children (McGhee 1976, 1983, McGhee and Chapman 1980, Chapman 1973) attesting, for example, to particular cognitive capacities which act as prerequisites for understanding specific forms of humour. Although a nomothetic approach - developmental changes in humour that are generally characteristic of individuals as they move from infancy through old age - as presented by McGhee (1983: 118) cannot be used to assess the progression of an adult learning a foreign or second language, it is plausible that there exists a developmental route for the acquisition of humorous devices in a foreign language. Verbal humour is a skill which has rarely been valued in the formal teaching context and is often peripheral to the business of rule learning, despite

348

Josiane

Boutonnet

the psychological advantages gained from using humour and the benefits it offers learners in the classroom (cf. Vizmuller 1979/80, Sopher 1981, Goodman 1983, Alexander 1997). Yet, in the naturalistic environment, Second Language Learners (SLL) will need to make sense of humour whilst working out the structural properties of that language, and in the process of learning a semantic system, construct their own identity within a socio-semiotic system. As Halliday argues (1973: 11) language is the medium through which one becomes a personality in consequence of his/her membership of society. For the Second Language Learner, becoming a member will partly depend on his/her ability to take on a humorous role in some situations.

2. Contextual Properties Linguistic expressions cannot be accurately understood except with reference to the context in which they occur. The term context encompasses much more than the immediate situation and the participants in the interaction, as the whole of the culture is meaningful. By identifying aspects of the context which are relied upon by the hearer to access the humorous intent in someone's speech, it can be shown that it is not possible to separate linguistic from socio-cultural knowledge. Both Hymes' (1962) notion of the speech event, his SPEAKING mnemonic, which determines the way individuals talk in a particular situation, and Levinson's (1979) activity type which views the language itself as determining the event, are combined in this approach. This author agrees that informal situations are more likely to generate informal chats, including what Cook (1996) labels 'space-filling discourse'. Yet, individuals' uses of language shape everyday interactions. Casual conversations host a number of discourse types, as participants switch from one register to another, so there is room for manoeuvre even within that type of setting. As Maybin argues (1996: 5), language takes meanings from the context in which it is used, and I would add that participants form one essential component as they engage in the use of language to achieve social goals.

3. Playing with Language Verbal humour depends on a study of the linguistic data. Much has been written on word play, especially punning, and the richness of resources found in English for the purpose of amusing. For detailed analyses of 'poetic jokes' (Hockett 1977), or 'accidents in the design of English' (Pepicello and Weisberg 1983) see Nash 1985, Chiaro 1992, Alexander 1998; I myself will focus here on the interplay between language and contextual aspects affecting the production and reception of verbal humour. An inductive methodology typical of Ethnography and Conversation Analysis is adopted, and explanatory accounts are based on data collected by the author.

Verbal Humour as the Ultimate Level in Language

Learning

349

3.1. First Example In the local context of the communicative event verbal humour is a device used by speakers to amuse their audience. Reaction to the stimulus is often overtly expressed non-verbally (smile, laughter...) or verbally. In some instances the audience withholds its response so as to censure the act: Jl) A young couple in their early twenties (G - male, J - female) living in Britain, were sharing their house with a German friend (D - female) of the same age. They were on holiday and G was spending his time in the garden shed restoring some item of furniture. They would all take it in turn to make each other tea. During one of J's turn to bring tea to G in the shed - G was inside the shed, turned away from the shed door - J tapped G's backside to alert him to her arrival and this was his response: Oh stop it D! Both G and J laughed. J then reported the joke to D who did not share the joke, but at the time successfully concealed it, and explained years later that she had felt anxious about it, and had resolved to find accommodation somewhere else, for fear of causing disharmony between the couple.

The joke has its basis in nominal substitution, where J's name has been replaced by D's name, thereby implying a degree of intimacy between G and D. The tone in which it was delivered conveyed the meaning 'stop teasing me!' with a deliberate intention to tease J. D was a fluent English speaker, and had been living in Britain for twelve months. She nevertheless failed to join in the humorous act reported to her, despite her knowledge that a joke was intended, she did not accept it as funny. Her anxiety can be attributed to her not having established a joking relationship (Radcliffe-Brown 1952) with the participants (she had been living with them a short while). The term 'joking relationship' here refers to a tacit understanding between two or more individuals that joking is permitted, and perhaps expected in some circumstances, with the same combination of friendliness and antagonism suggested by Radcliffe-Brown. According to D's personal world-view - being familiar with someone else's partner did not constitute a matter for joking, this act was threatening because it was positioning her in a role in a constructed story which went against her perception of herself. It is unlikely that the reception of the joke can be attributed to the fact that English is D's second language, as her proficiency was near-native. Rather, it highlights the importance of rapport when poking fun at others. It also draws attention to the various kinds of personal 'baggage' individuals carry around with them, a product of their up-bringing and life experiences, which comes to form the means by which they view and act upon the world.

Josiane Boutonnet

350

All three participants have now known each other for eighteen years and have since discussed the event. Due to shared experiences and increased understanding of each other, all three have added to their jointly produced 'conversational joking'; the original joke has been kept alive as an expression of intimacy, and has been altered, adapted and extended to fit a number of other humorous situations. D has devised a number of responses i.e. "shsh.. she might hear us.." and even acts as the initiator on other occasions. This strategy often shapes the interactions which follow by creating an atmosphere of fun and makes for high involvement in the Norrickean sense (1994) (cf also Gumperz 1982), i.e. 'the coherent give and take of talk in interaction'. Cameron (1989) argues that it is not true that participants necessarily share the same view of what is going on in interaction and in unequal encounters, where participants do not enjoy the same degree of power, communication can be fraught. This is certainly true of cross-cultural encounters, where knowledge of the linguistic system, and the pragmatic rules for its use, constitute a form of power. The outsider who is the butt of the humour will suffer embarrassment whilst gaining an insight into the culture's values. It can be argued that the joke played on D in J1 can function on two levels: on the one hand it might have been used by G to show intimacy towards J, and J to show intimacy towards D - it is not clear whether G intended it to be reported back to D. On the other hand it can also signal aggression by making the recipient feel uneasy. J could have been testing D's reaction too. Tannen (1992) refers to the "paradox of power and solidarity" when a single turn at talk can lead to dual interpretation, as illustrated above. The metamessage 'this constitutes play' - was not accepted by the recipient and when a refusal to engage in the humorous act (i.e. by taking offence) is the outcome, it backfires on the initial speaker. The strategy in this case redefines the personal boundaries and can also shape future interactions, making for low involvement in interaction and reduced contact. 3.2. Second Example J2) During a seminar discussion where a group of five students were discussing the relationship between language and thought, one participant mentioned the stage during which some children do not distinguish linguistically between their father and other men, and use the word 'dada' for all. A: ... so children might well use... that word...to refer to other men they know... you know... like the neighbour, the milkman ... and it doesn't mean... B: [ well, it might be... (All participants giggle... before resuming the discussion)

Verbal Humour as the Ultimate Level in Language

Learning

351

The student's contribution here cannot be analysed in terms of word play, but its content, closely following the word 'milkman' in the preceding turn is the trigger which generates giggling. It relies on the participants believing B's contribution is relevant to the ongoing discourse and thus interpretation of an investigative kind is required. It also relies on participants sharing knowledge of the connotations associated with the word 'milkman', namely that of someone who is used by the culture to diffuse tensions relating to infidelity. The student's interruption is interpreted as humorous by the audience and points to the importance of language as the means to bind people together and to invoke shared knowledge between the participants. As well as communicating ideational meaning, talk is the vehicle for communicating interpersonal meaning (Halliday 1978). Language can be used to act on a situation, i.e. tense situations can be diffused by a speaker's humorous contribution, topics can be renegotiated etc. B's move is not a bid for a turn, but acts as a device to detract from the serious business, possibly because the student's own contribution had been negligible up to that point. It is possible that B's contribution is the result of the previous discussion which the students have had as part of the same activity, discussing issues of sexism in English. Thus it could be both anaphoric and at the same time act as a device to shift the discourse back to the topic of sexism in language. It is also possible that the individual is someone with a snappy answering conversational style. He might have established himself as a 'joker' in the group, someone who has a talent (or at least a taste) for conversational impromptu humour. B places himself politically, as someone who is willing to challenge notions of Political Correctness, albeit in jest. Without sound knowledge of the person, it is probably unlikely that one could identify the speaker's position, although his intentions can be identified: amusing - and possibly testing - his audience, as well as possibly distracting the participants away from the topic. Humour provides an excellent means of negotiating understandings about the world, and if rebuked the perpetrator can retreat to a safe position: "it was only a joke". The inserted utterance is not novel, and B has in fact performed a wellrehearsed idea which has been expressed in a variety of forms in many conversations, and with the same implied meaning, drawing on sexist Members' Resources, to borrow Fairclough's terminology (1989). Whether one chooses to resist the assumptions made by the speaker's remark, and sanctions by disapproval, the speaker can be assured of having drawn on the same meanings. Only in the case of young children, or outsiders to the culture would meaning and understanding not necessarily be matched. French speakers might infer what is meant from the format, which would not be too dissimilar in French, but French people do not have milkmen, and for the French bystander, making sense of the

Josiane

352

Boutonnet

implicature would require establishing parallels between the two cultures. The milkman has a counterpart in the postman in French society, who is used in a similar way in conversations. Someone is who is not suitably acculturated might be totally bemused or might take the referential proposition literally. Non-literal meanings which require interpretive inferencing are particularly demanding on the SLL. In example J2, subsequent discourse should indicate whether the humorous act is used by the group strategically to help establish and reaffirm dominance in a status-hierarchy (Chapman 1983) or whether to subvert power relations. The fact that in our example J2 it does not disrupt the discussion and merely constitutes an aside, rather than a diversion, does not diminish the fact that it might trigger a fullblown debate on sexist attitudes at a later stage in the conversation. Thus, although humour is located in short sequences of two or three part turns, one has to consider the overall discourse. Speakers build on ideas and jointly produce world views, as well as reproduce old ones. How humorous contributions fit together is crucial, turning separate experiences into shared experiences, revealing group allegiances and communicating attitudes.

4. The Language Learner Producing Humour For the language learner in a naturalistic environment, learning the target language comes from the need to learn how to mean in a variety of situations. Verbal play and verbal humour generally are forms of a language game in which speakers take part on a daily basis in most verbal encounters. Learning how to play the game is another way of understanding the culture and of demonstrating assimilation. However, it is possible for the SLL to be the initiator of humour - unintentionally - due to errors in pronunciation, syntax or a failed attempt to make use of idiomatic expressions. J3) B is a Spanish native speaker who has been living with a family in England for 12 months. She wishes to ask G (who is male) to wake her up early the next day, as she has an important examination. B: Can you get up me tomorrow? G: Well, if you like. (Turning to his wife) Is that OK with you?

The humour evidently derives from the misplacing of the preposition. The learner might be aware of the double-entendre in this creative use of language and self-correct; on the other hand some form of explanation might be expected if the learner realises the potentially comic effect, and, although this would draw attention to the gaffe committed, it would have the double effect of signalling inclusion and providing an opportunity for learning about the language as well as the culture.

Verbal Humour as the Ultimate Level in Language

Learning

353

One of the problems facing the adult learner however is coping with being the object of laughter. If according to Hagege (1996: 176) 'I'angoisse du ridicule n'habitepas 1'enfant' it is a major inhibiting factor in second language production in adults. Cross-linguistic differences and influences (Odlin 1989) and a learner profile can yield information about the second language learner's performance and ultimately inform pedagogical methods and teaching practice. According to Alexander (1997: 178) the negotiation of verbal humour needs to form part of a learning objective in its own right, not least because of the pervasive nature of humour in much verbal social interaction. Native speakers embrace the SLL as one of the speech community, when s/he has finally graduated to the stage of appreciating humour, regardless of her/his overall linguistic competence in that language - the speaker might not be proficient in some domains. La Fave et al (1976) describe this achievement as 'virtuous', one which is greatly valued and admired in any speaker, but it is one which is in essence ethnocentric, relying on the participants sharing cultural values and world views.

5. Contextual Clues For an utterance to be taken as funny, or for the humorous intent to be recognised, participants in interaction make use of a complex set of clues which are constantly checked against the individual's frame of reference. The following example is one which is used often by G on a number of selected visitors to the house: J4) A friend X arrives at G and J's house, one of the children opens the door and X says "hello". On hearing X enter the house, G who, is in another room of the house with his partner J, comments: G: I hate X, don't you? J laughs. X joins both G and J, laughing G: ...oopps! and the conversation begins.

In this instance the humour derives from a number of linguistic and prosodic clues, as well as a set of conditions in the Searlean sense (Searle 1971): -

voicing a negative opinion about a friend knowing the person concerned is overhearing (flouting the Approbation Maxim, Thomas, 1995); being caught in the act of saying something which is not normally voiced openly; the knowledge that all participants must not take the utterance literally, that it is meant as a 'joke': because the utterance was said in a loud voice so that X would overhear, X knows that G and J know that it would be overheard by X, X knows that G and J are

Josiane

354

-

Boutonnet

aware that X is entering the house, the participants have a joking relationship, G is known for making jokes; the fact that both hearer and speaker know the utterance to have been said by G to his partner with X being part of the audience.

Here we have a combination of factors: in terms of prosody, loudness of voice and intonation, function of tag question - using Holmes' (1995: 79-86) classification of tag questions, don't you would function as a facilitative tag, if asking for confirmation; however, here it functions as a rhetorical question. Also, G's facial expression, which as X enters the room is one of delight, with "...oopps" being expressed as feigned surprise. Those sound and visual clues interact with the participants' knowledge outlined above, that a Maxim has been flouted with the purpose of causing amusement and rather than showing hostility, the comment acts as a unifying device.

6. Conclusion Aspects of context are multi-faceted and cannot be reducible to the linguistic event itself. As mentioned above, the whole of the culture is meaningful, and this is what proves to be especially difficult for the SLL whose ability to fit the puzzle bits together is hindered by knowledge gaps. Verbal humour takes full advantage of knowledge acquired by a speech community. Acquiring the knowledge to understand and perform a variety of humorous acts might in fact be a life-long endeavour. One significant aspect of an individual's sociolinguistic competence is knowledge of the norms associated with her own culture, and for the second language learner, of those of other cultures she comes in contact with. To achieve successful communication and because verbal humour relies on participants drawing on mutual understanding, she needs to acquire a substantial amount of background knowledge besides the language system itself. Rarely are individuals explicitly briefed about what constitutes the 'ridiculous'. Rules for appropriate language use, and the means available in English for making a creative use of humour, are often learned at some personal cost, due to a mismatch between the native speaker's poetic potential and that of the learner. Verbal humour in informal conversations remains misunderstood until the cues signifying humorous intent are spotted and the necessary reconstruction work accomplished. And even then, for humour to be appreciated, some degree of acculturation (Schumann, 1986) might be required. As an outsider to a language it is perhaps easy to assume one might be able to access a language as one enters the country, because the language appears to be simply there, to exist as an identifiable entity. But language is embedded in social

Verbal Humour as the Ultimate Level in Language Learning

355

activities, and inseparable from its speakers. The SLL needs to tap into sociocultural and pragmatic knowledge, as well as to acquire the structural properties of English. As Alexander (1997: 9) clearly states there is a dimension to language learning 'which goes beyond the language system proper', and becoming fluent in a second language is about reaching those parts which the system cannot reach.

References Attardo, S. (1994): Linguistic theories of humor, New York and Berlin: Mouton de Gruyter. Cameron, D. (1989): Rules, power and communication, In: Forrester, M. A. 1996: 116-125. Chapman, A. J. (1973b): Social facilitation of laughter in children, Journal of Experimental Social Psychology, 9, 528-541. Chapman, A. J. and Foot, H. C. (eds.) (1976): Humour and laughter: Theory, Research and Applications, London, New York: John Wiley and Sons. Chapman, A. J. and Foot, H. C. (eds.) (1977): It's a funny thing, humour, Oxford: Pergamon. Chapman, A. J. (1983): Humour and Laugher in Social Interaction and Some Implications for Humor Research, in McGhee and Goldstein, 1983. Chiaro, D. (1992): The Language of Jokes - Analysing Verbal Play, London and New York: Routledge. Cook, G. (1996): Language Play in English, in Maybin and Mercer (eds.), 198-234. Fairclough, N. (1989): Language and power, London: Longman. Forrester, M. A. (1996): Psychology of language - A critical Introduction, London: Sage Publications Ltd. Goldstein, J. H. and McGhee, P. E. (1972): The psychology of humor, New York and London: Academic Press. Goodman, J. (1983): How to Get more Smileage Out of Your Life: Making Sense of Humor, Then Serving It, in McGhee and Goldstein, 1983, 1-21. Gumperz, J. J. (1982): Discourse strategies, Cambridge: Cambridge University Press. Hagege, C. (1996): L'enfant aux deux langues, Paris: Editions Odile Jacob. Halliday, M. A. K. (1973): Explorations in the Functions of Language, London: Edward Arnold. Halliday, M. A. K. (1978): Language as social semiotic: the social interpretation of language and meaning, London: Edward Arnold. Hockett, C. F. (1977): Jokes. In The View from Language: Selected Essays, 1948-1974, Athens, Georgia: University of Georgia Press, 257-289. Holmes, J. (1995): Women, men and politeness, London: Longman. Hymes, D. (1972): On communicative competence, In: Pride, J. B. and Holmes, J. (eds.) Sociolinguistics, Harmondsworth: Penguin, 269-285. La Fave, L. Haddad, J. and Maesen, W. A. (1976): Superiority, Enhanced Self Esteem, and Perceived Incongruity Humour Theory. In: Chapman and Foot (eds.), 1976, 63-91. Maybin, J. and Mercer, N. (eds.) (1996): Using English - From Conversation to Canon, London and New York: Routledge. McGhee, P. E. (1976): Children's Appreciation of Humour: a Test of the Cognitive Congruency Principle, Child Development, 47, 420-426. McGhee, P. E. and Chapman, A. J. (eds.) (1980): Children's Humour, Chichester, England, Wiley. McGhee, P. E. and Goldstein, J. H. (eds.) (1983): Handbook of Humor Research, New York: Springer-Verlag, Vols. 1 and 2. Nash, W. (1985): The Language of Humour, London and New York: Longman Ltd. Norrick, N. R. (1994): Involvement and Joking in Conversation, Journal of Pragmatics, 22, 4 0 9 430.

356

Josiane

Boutonnet

Palmer, J. (1994): Taking Humour Seriously, London and New York: Routledge. Radcliffe-Brown, J. (1952): On Joking Relationships, In: Palmer, J. 1994, 13-16. Ross, A. (1998): The Language of Humour, London and New York: Routledge. Schumann, J. H. (1986): Research in the Acculturation Model for Second Language Acquisition, Journal of Multilingual and Multicultural Development, 7, 379-392. Searle, J. R. (ed.) (1971): The Philosophy of Language, Oxford: Oxford University Press. Sopher, H. (1981): Laugh and Learn, English Language Teaching Journal, 35, 431-6. Tannen, D. (1992): That's not What I Meant! London: Virago Press Ltd. Thomas, J. (1995): Meaning in Interaction, London and New York: Longman. Vizmuller, J. (1979/80): Psychological Reasons for Using Humour in a Pedagogical Setting, Canadian Modern Language Review, 36, 266-271. Zhao, Y. (1988): The Information-Conveying Aspect of Jokes, Humor, Vols. 1-3, Berlin: Mouton.

Susan Mandala

Talk in the Mind: Scripted Dialogues and Mental Scripts 1. Introduction In this paper, I argue that scripted pedagogical dialogues are biased towards one participant's point of view. While the existence of such a bias clearly has implications for dialogue development, another question to ask is why such a bias might exist at all. It is the latter question that is the subject of this paper. The original purpose of the research on which this work is based was to compare the exchange structure of advice-giving events in naturally occurring conversation with similar exchanges represented in English language teaching textbook dialogues, i.e. written representations of conversation designed for language teaching purposes. The aim was to see if the represented exchange structure of the advice-giving events deviated from the natural exchange structure, and if so, to investigate what effect such deviations had on the naturalness of the dialogues. Carter and McCarthy (1995) have claimed that teaching dialogues frequently fail to represent interactions in a naturalistic way because they tend to follow written norms that are inappropriate in representations of conversational speech. However, a dependence on written norms did not emerge as a central concern in the teaching dialogues studied here. Rather, these were found to be pragmatically unsound because they tended to represent conversation from only one point of view, typically that of the advice-giver rather than the recipient. It is significant that it is the advice-giver's point of view that is favoured in the scripted dialogues. Drawing on Malone's (1995: 149) work, we can say that the speech event of advice-giving casts two discourse roles, that of the advicegiver and that of the advice-recipient (see also Decapua and Huber 1995). The advice-giver's role can be considered the higher status role as this participant is either perceived as more knowledgeable, or presumes so to be (Decapua and Huber 1995). There is, however, evidence to suggest that the higher and lower status roles in advice-giving can be recast (Malone 1995). In a study of physicists meeting to discuss preparations for a conference, Jacoby and Gonzalez (1991: 158) note that participants who receive advice are cast as novices. They can recast themselves as experts by rejecting their interlocutor's directive moves, as in example (1) below:

Susan Mandala

358

(1) Segment 5 - RO Lab 10-17-90. Data from Jacoby and Gonzalez 1991: 171, lines 29-42. Indications of overlapping speech and non-linguistic behaviour removed for clarity of example: 29 Ron:

But (.) Marsha has just made a suggestion that by shifting your time scale you might be able to make it go away. (0.6)

40

(But) I cannot shift the time scale (.) li-linear (ly). (.)

Miguel:

According to Jacoby and Gonzalez (1991: 173), Miguel, a doctoral candidate, is the lower-status participant in terms of knowledge and institutional rank with respect to Ron, a tenured professor and the lead investigator. By rejecting Ron's suggestion that Miguel take Marsha's advice, Miguel manages to deny Ron's expert status and recast himself as a knowledgeable participant (Jacoby and Gonzalez 1991: 172). Evidence given by Heritage and Sefi (1992) in a study of advice given to new mothers by health visitors can be similarly interpreted. Heritage and Sefi (1992: 367, 391) suggest that new mothers are concerned to reinforce their ability as able carers when they request or receive advice. Evidence of such concern is indicated in example (2): (2) [1C1:44],

Data from Heritage and Sefi 1992: 406, lines 17-20.

17

HV:

hh I think if you've got three cats it might be a good idea to have a cat-net. you've got three cats if might be a good idea to have a cat-net.

M:

Yeh I've got one I've definitely got one

18 19

As Heritage and Sefi (1992: 407) explain, the mother's response in line 19 indicates that she has already dealt with the problem and thus does not actually need the advice. Like the rejected directives identified by Jacoby and Gonzalez (1991), such moves can be interpreted as attempts by the advice recipients to recast themselves as knowledgeable higher-status participants. Such recastings may result in what is essentially a power struggle as participants compete for the higher status role. Evidence for this view comes from work by Jefferson and Lee (1981), who set out to study troubles-tellings but found advice and advice rejections instead. In their data, tellers of troubles who receive advice instead of commiseration react negatively to that advice by rejecting it. Jefferson and Lee (1981: 410) account for these rejections by suggesting that such interactions suddenly contain two speakers, the troublesteller, the first speaker, and the advice-giver, who had previously been the troubles-recipient. Before the interaction can continue smoothly, the participants

Talk in the Mind: Scripted Dialogues and Mental Scripts

359

have to realign themselves. Either the troubles-teller must become the advicerecipient, or the advice-giver must revert to the role of troubles-recipient (Jefferson and Lee 1981: 412). As Jefferson and Lee (1981) point out, participants are competing to be givers or tellers, rather than recipients. Does this suggest that the giver and teller roles have attached to them greater status, at least in terms of participant perception? In examples (1) and (2) above, it was suggested that advice-recipients attempted to raise their status by rejecting advice through the display of knowledge. In these cases, status is relatively easy to characterise in terms of institutional rank and expertise. In Jefferson and Lee's data, however, participants are similarly competing for status in casual interactions. While we cannot appeal to a notion of social status to explain competition for the giver/teller role, we might suppose, as Decapua and Huber (1995) do, that advice-givers in casual talk adopt higher status roles through the presumption of greater knowledge. However, this does not account for the desire to remain the troubles-teller, as knowledge is not at issue in troubles-tellings. It may be possible to interpret these findings in terms of speech event expectations. In addition to their argument about the sudden existence of two speakers, Jefferson and Lee (1981) suggest that the advice-giver as speaker inappropriately introduces a service encounter element into an intimate troublestelling. Arguing along similar lines, Traverso (1999) points out that participants in service encounters are expected to perform different kinds of verbal acts in relation to their differing roles, and these expectations stem from situational constraints placed on service encounter interactions. We can go one step further here and suggest that speech events not only cast different roles, but may in certain cases cast one of them as relatively dominant, and the other as relatively reactive. Applying this to Jefferson and Lee's data, we can suggest that in telling a trouble, the troubles-teller has cast him or herself as the dominant teller, and the other participant as the reactive troubles recipient. By giving advice instead of receiving the trouble, the recipient rejects the reactive role, becoming instead the dominant advice-giver. By rejecting the advice, the troubles-teller resists being cast as the reactive advice-recipient. As Jefferson and Lee (1981) demonstrate, such interactions are unstable and tend to falter until one of the participants accepts a reactive role.

2. Scripted Pedagogical Dialogues as Imagined Discourse The data investigated here suggest that the textbook advice-giving encounters are biased towards the dominant discourse role of the advice-giver. It might be supposed that dialogue writers are consciously taking the decision to produce dialogues that exhibit this tendency. However, given that the purpose of these

Susan Mandala

360

dialogues is most often to present new language for the learner, such a decision seems unlikely. Rather, it is more probable that dialogue development is largely an intuitive process, and this results in the production of dialogues which exhibit a dominant role bias. Preliminary evidence for the role of intuition in scripted dialogue development comes from a questionnaire distributed to English language teaching textbook authors1. The questionnaire was designed to elicit information on how authors produced the dialogues that appear in their course books. Of primary interest was whether the authors used natural speech data in the development of scripted dialogues, whether they followed their intuition as native speakers, or whether they used some combination of the two strategies. Of the twenty-two authors to whom the survey was distributed, fifteen responded. Of those fifteen, three responded by saying the questionnaire was too difficult to answer. Thus, the results are based on the responses of twelve informants. This is a relatively small informant pool and it is recognised that further research is needed. However, the preliminary results are instructive for the purposes of the present discussion. The respondents were all native speakers of British English, the majority had over five years of experience as materials developers and held at least a masters degree in linguistics, applied linguistics or a related discipline in combination with a TEFL/TESL qualification. The two questions most relevant to our purposes are listed below2. 11. What is typically your starting point in dialogue development? (e.g. intuition, dramatic/literary models, television/movie dialogues, conversations you have had or overheard, a corpus of speech data) 13.

In 10 steps or less, please describe how you prepare dialogues for inclusion in your course books. If you prepare different types of dialogues in different ways, please specify.

The responses to these questions suggested that the typical scripted dialogue may begin in the mind, not in examples of natural speech. That is, the dialogues go primarily from mind to page, not from natural speech transcript to page. Eleven respondents answered question eleven, and intuition was cited as the starting point in nine of the responses. Four of these nine gave as their starting point intuition combined with conversations the authors had participated in or overheard. While natural conversation figures in the responses, it must be pointed out that the empirical data cited is remembered rather than edited from taped material. The use of corpora was mentioned only once as useful, but only as a way of checking intuition. The responses to question thirteen confirmed those of question eleven. In most cases, the first step given was intuitive, such as thinking of a structure or function that needed to be presented, and then placing this in a plausible context. 1

The questionnaire can be obtained from the author.

Talk in the Mind: Scripted Dialogues and Mental Scripts

361

Such responses indicate that scripted pedagogical dialogues may represent imagined discourse, and as such they are a key into our mental discourse world.

3. Scripted Pedagogical Dialogues as Mental Scripts If it is accepted that scripted dialogues represent imagined discourse, they can be further interpreted as representations of mental scripts. According to Schank and Abelson (1977: 41), a script "is a predetermined, stereotyped sequence of actions that defines a well-known situation." In other words, scripts are the mental representations of everyday situations and can be viewed as speech event prototypes. Because scripted dialogues are in most cases intended as models of natural talk and are not typically produced for artistic or dramatic purposes, it can also be suggested that scripted pedagogical dialogues represent the mental script of a given conversation in a given situation. In the present case, the script is that of an advice-giving encounter. The components of the script, the imagined discourse acts, are likely to approximate not what occurs in natural discourse, but the idea of what is likely to occur in the mind of the speaker, the speech act prototype (see Blum-Kulka et al. 1989 and Hill et al. 1986, cited therein, on the identification of prototypes which stand for a range of speech act variations in performance). Schank and Abelson (1977: 41) also point out that "every script has associated with it a number of roles. When a script is called for use, i.e. 'instantiated' the actors in the story assume the roles within the instantiated script." Presumably, this role part would include the speech act prototypes for the actors. According to Schank and Abelson (1977: 42), a script is "written from one particular role's point of view," but that "scripts from many perspectives are combined to form what might be considered the 'whole view' . . . . Such a 'whole view' is rarely, if ever, needed or called up in actual understanding." As will be shown in the next section, a comparison between advice-giving events in scripted pedagogical dialogues and advice-giving events in natural speech2 indicates that the 'whole view' may be rarely used because perspectives are not typically combined in imagined discourse. Rather, the perspective of one participant, the dominant role participant, tends to govern the interaction.

2

A more detailed discussion of the textbook and natural talk data can be found in Mandala (forthcoming).

Susan Mandala

362

4. Dominant Role-Centricity in Scripted Pedagogical Dialogues In order to compare the utterances in the actual and scripted advice-giving encounters, the following four categories are used: reactive role participant advice deflection, dominant role participant sustained rejections, dominant role participant overruled rejections, and topic changes3.

4.1. Advice Deflects Advice in natural talk is frequently met with resistance as well as rejection (Hutchby 1995; Bresnahan 1992; Heritage and Sefi 1992; Jefferson and Lee 1981; Barber 1979; Kress and Fowler 1979). Advice that is not accepted, rejected or resisted is left unresolved and can be described as deflected advice. An example of deflected advice is given in (3): (3)

CANCODE:# 90095002

S1 ¡Initiation: S2:Negative Response:

I think it might be worth getting in touch just to find out. Well Well I I they're down on the list. I can't get in touch with them at the moment. There's nobody in in the day.

As can be seen in the example, deflecting advice is something the reactive role participant in natural talk actively does, in this case by delaying compliance. This in itself is not an interesting finding, as it is well accepted that talk is interactional and that goals are negotiable. What is interesting here is that when textbook advice is left unresolved, it is rarely because the reactive role participant delays compliance. Rather, the advice is left unresolved primarily because the dialogue simply ends after the advice-giver issues the advice, as happens in examples (4) and (5) below. (4)

Textbook Data: Olearski et al. Formula 3. 1991: 151.

C2:Challenge: C1 :Positive Response: C2: Challenge CI Challenge:

3

I enjoy smoking. Why should I give it up? Well, doctors say it's harmful to your health. Oh, I never take any notice of doctors. They're always making a fuss over nothing. As far as I'm concerned you're taking an unnecessary risk. When you hear the latest statistics, you'll soon come round to my way of thinking.

The analysis given in the data throughout this paper is a modified form of Tsui's (1994) linguistic conversation analysis.

363

Talk in the Mind: Scripted Dialogues and Mental Scripts (5) Textbook Data: O'Neill and McLean. Kernel 3: 1982: 64. But you're a good guitarist. accepted the job yet? No, not yet.

Jenny:Challenge: Nick:Negative Response: Nick:Initiation: Jenny.Challenge:

You'll be all right.

Have you

You see, I told them . . . Well, if I were you , I'd accept as soon as possible. This is your big chance, Nick. Take it.

In neither case can we know how the last piece of advice was received by the reactive role participant, as the dialogues end without reaching that point. In both (4) and (5), it is the dominant role participant who has the last word. Notice as well that the advice-giver's last words are fairly aggressive. In terms of politeness theory (Brown and Levinson 1987), the advice in both cases is indicated explicitly (As far as I'm concerned in 4; Well, if I were you in 5) and delivered without redress. In neither case, however, does the reactive role participant have a represented rejoinder. In example (5) particularly this effect is heightened by the advice-giver's interruption of the recipient's account in line 4. Interestingly, what is perhaps the beginning of a genuine advice deflect or a display of competency (Heritage and Sefi 1992) in response to earlier advice is written to fail, as Nick's You see, I told them... is cut off by Jenny's advice in lines 5-6. As further evidence of the dominant role participant bias in the scripted dialogues, consider a natural interaction in which the advice-giver goes fully on record (Brown and Levinson 1987) with their advice. Example (6) is similar to (5) above in that one participant is advising the other to take a job. (6) Cobuild: # S0000000005 SI initiation:

S2:Positive Response: S3 initiation:

Well if you want my opinion FX if you don't do it soon you're gonna lose the opportunity and you'll be sorry Cos opportunities don't wait for people. That's true isn't it I think I think I lose a lot of confidence because I'm not in a salon long enough. And I think I need that confidence.

As we can see, Jenny's represented advice-giving move in lines 5-6 of (5) is lexically and pragmatically faithful to S i ' s natural advice-giving move in lines 1-3 of (6). What is missing from the scripted dialogue is the recipient's voice, which ends with Jenny's point of view predominating. In the natural sample, the recipient is truly reactive, as there is a response to the advice. S3 issues a selfdenigrating move (see Tsui 1994: 149) which is nevertheless a display of competence (Heritage and Sefi 1992). While expressing her lack of confidence, S3 also reinforces her ability to make her own decisions.

Susan Mandala

364

4.2. Sustained Rejections If an advice recipient rejects advice, the advice-giver may sustain adopting the recipient's point of view. In a sustained rejection, does not object to the recipient's rejection of their advice, as Here, SI has been advising S2, and S2 rejects a piece of advice in

that rejection by the advice-giver in example (7). lines 1-3:

(7) CANCODE:# 90112001 S2:Challenge:

. . It'll be what you gain minus what you lose. And it the the S3 :Back Channel: Mm. end results wouldn't be quite as worthwhile as it 53 :Back Channel: Mhm. S4: be quite 54 /SI :Positive Response: expensive to do. SI: Oh no probably not no. It would probably be pretty expensive

As we can see in line 6, S4 and SI support S2's rejection of S i ' s advice by indicating that the proposed changes (changes SI has suggested) would indeed be expensive. Here, what is in effect happening is that the advice-giver ratifies the recipient's bid for the dominant discourse role. Sustained rejections are primarily oriented towards the advice recipient in that they support the recipient's point of view. While they were found to occur in both the natural data and the scripted data, fewer sustained rejections in the scripted data could be considered oriented towards the recipient. Rather, sustained rejections in the scripted dialogue data tended to support the advice-giver. In examples (8) and (9) below, we see rejections that are sustained, but no evidence of movement on the giver's part towards the recipient's point of view: (8) Textbook Data: Olearski et al. Formula 3. 1991: 73. C2 has a cold, and CI suggests staying home. CI initiation: C2:Challenge: CI:Positive Response:

Can't you stay at home? Not really. Whereas before I could stay at home, now I can't afford to. Well, that's life, isn't it?

(9) Textbook Data: Maley and Duff. Variations on a Theme. 1978: 142. CI has advised C2 to go to the police, and C2 queries this advice before rejecting it. We begin the extract with CI's response to C2's query. C1 :Positive Response: C2:Challenge:

They're sure to notice it. So? They still don't know who's found it.

Talk in the Mind: Scripted Dialogues and Mental

Scripts

C1 :Positive Response:

As you like.

CI :Initiation:

I'd still go if I were you.

365

C l ' s move in line 4 of (8) is a positive response to C2's report (lines 2-3) in that it can be regarded as an acknowledgement of the reported state of affairs (see Tsui 1994: 182). However, it is clearly the advice-giver's position that is supported here, not the advice-recipient's. Note also that line 4 is the end of the dialogue. CI gives advice, sees that advice rejected, and then simply ducks out of any further interaction. While the rejection is sustained, the prevailing point of view here is that of the dominant role participant. Example (9) follows a similar pattern to (8), even though it is from a different book published thirteen years previously. In (9) C2 rejects the advice (line 2). While the rejection is sustained in line 3, note that the sustained rejection is oriented towards the advice-giver, and immediately followed by a reiteration of the advice. C l ' s reaction to the advice rejection basically amounts to 'Fine. Have it your way, but don't say I didn't warn you.' Sustaining a rejection in this way reinforces the advice-giver's position rather than supporting the advice recipient, further indication that the scripted dialogues favour the point of view of the dominant role participant. 4.3. Overruled Rejections Overruled rejections occur when the advice-giver objects to the recipient's rejection of the advice. In contrast to sustained rejections, overruled rejections are oriented towards the advice-giver, the dominant role participant. Examples are given in (10) and (11) below: (10) Cobuild: # S0000000492. SI is suggesting that S2 join the civil service. SI Initiation:

You ought to join it I think you'd do absolutely brilliantly.

S2:Challenge:

I'm too old.

S1 :Challenge:

To join the civil service?

S2:Positive Response:

Yeah.

SI :Challenge:

You're joking.

(11) Cobuild: # S0000000448 SI initiation:

Erm you're you're just gonna have to sit down with FX and and r really talk it out.

SI Challenge:

Well there's there's nothing for her to say it's for me to decide whether I'm gonna pay it or not really isn't it.

S2:Challenge:

No, no because you see the thing is . . . (move continues)

Susan Mandala

366

In (10) SI overrules S2's rejection by indicating that S2 has no valid grounds for making the rejection (lines 3 and 5). S2 does essentially the same thing with S i ' s rejection in (11) (line 5). As we can see in the examples, the reactive role participants are failing in their bids for dominant role status. If, as we are suggesting, scripted dialogues tend to be biased towards the dominant role participant, we would expect overruled rejections, moves which are by their nature oriented towards the dominant role participant, to be well-modelled in the scripted dialogues. And this is indeed what we find. Consider examples (12) and (13) from the scripted data. (12) Textbook Data: O'Neill and McLean. Kernel 3. 1982: 64. Jenny has advised Nick to take a job playing in a band. The extract begins with Nick's rejection of that advice, in which he notes that taking a secure job might be more beneficial in the long run than playing the guitar. Nick:Initiation: Jenny:Challenge:

It's a very safe job, you see. Playing in a band is a bit risky. But you're a good guitarist. You'll be all right.

(13) Textbook Data: Hargreaves and Fletcher. Making Polite Noises. 1985: 26. CI initiation:

Why don't you get a decent job for a change?

C2:Challenge:

But I like my job.

CI Challenge:

Look, digging gardens is not a job for a University graduate.

In both (12) and (13), the advice-givers object to the rejection of their advice. In (12) Jenny asserts that Nick is a sufficiently talented guitarist who will not need to worry about job security, and in (13) CI asserts that gardening is not an appropriate job for C2 even though C2 enjoys it. In both cases, the represented overruled rejections are well-modelled with respect to the natural samples. In both (10) and (12), the dominant role participants assert that the reactive role participants are in error. In both (11) and (13), the overruled rejections are aggravated, signified by S2's emphasised disagreement in (11), line 5 (No, no), and C l ' s impatient focuser Look in (13) line 3. Overruled rejections are modelled well in the scripted dialogues, but it must be remembered that they are components of the advice-giving encounter which naturally favour the dominant role participant. 4.4. Topic Changes On the whole, changes of topic were not frequent in the scripted dialogue data. However, those that did occur did not typically serve to end the advice-giving encounter. Rather, most of the observed topic changes were written to fail, and this resulted in the advice-giver's view again predominating. An example is given in (14) below. In this extract, Jane has been advising Richard to quit smoking, and

Talk in the Mind: Scripted Dialogues and Mental Scripts

367

Richard has consistently rejected her advice. The extract begins after Jane has suggested that Richard eat sweets rather than smoke. (14) Textbook Data: Jones. Functions of English. 1981: 59. Jane:Challenge: Richard:Challenge: Jane:Negative Response: Richard:Follow-up:turn-pass: Jane:Initiation: Richard:Positive Response: Jane:Initiation:

Well do you think that matters? Don't you think it's better to be fat than unhealthy? No, I don't. I'd rather be f a t . . I'd rather be thin than fat certainly. Oh dear, oh dear . . . w e l l . . . Anyway... Hey! I've got a great idea! What's that? Why don't you go to a hypnotist? My sister did!

Richard attempts to shut down the current topic in line 6, and at first it appears that Jane responds by introducing a new topic in line 7. In line 9, however, we find that she has actually introduced more advice. Because the topic change is written so that it does not succeed, the dominant role participant is permitted to keep that position. We have thus far shown that textbook advice-recipients do not typically issue genuine advice deflects. Consequently, it is the advice-giver, the dominant role participant, who typically 'has the last word' in the scripted dialogues. In addition, we have shown that sustained rejections, recipient-oriented features of advice-giving encounters, are frequently oriented towards the dominant role participant in the scripted textbook dialogues. By contrast, overruled rejections, naturally oriented towards the dominant role participant, remain so in the scripted dialogues. There is also evidence to suggest that topic changes, places in discourse where roles may potentially be recast, are written to fail in scripted dialogues. Taken together, these factors add up to a considerable bias towards the dominant role participant in scripted pedagogical dialogues.

5. Further Evidence It can be argued that the evidence presented thus far is weak, based as it is on only one speech event, advice-giving. However, studies of scripted pedagogical dialogues concerned with other speech events point to the same bias. In a study of direction-giving encounters and representations of such interactions in textbook dialogues, Scotton and Bernstein (1988) point out that while the utterances of the direction-giver, the dominant role participant, are richly represented, those of the direction receiver, the reactive role participant, are only minimally represented. Moves that direction receivers actively make, such a requests for clarification, are

368

Susan Mandala

consistently under-represented in textbook dialogues (Scotton and Bernstein 1988). It can further be argued that the pedagogical dialogue, designed for teaching purposes rather than research, is flawed evidence. However, other types of imagined discourse exhibit a parallel bias towards the dominant role participant. Scrutiny of Searle's (1976: 23) speech act taxonomy, for example, reveals that it is framed primarily according to the point of view of the dominant role participant. Representatives are used to 'tell people how things are,' but there is no category for reacting to information that we receive from others. Directives are for 'getting others to do things,' but there is no corresponding category for responding to requests or commands. Even the commissives are conceptualised as oneparticipant actions. Speakers 'commit themselves to doing things,' but apparently do not respond to the proposals of other speakers. While there is a category for 'expressing feelings and attitudes' (the expressives), there is no corresponding category for acknowledging or reacting to these expressions as voiced by other speakers. With declarations, 'we bring about changes through our utterances,' but we apparently do not agree or disagree or otherwise react to these declarations when they are made by others. The persistent lack of act categories pertaining to the role of the reactive participant in Searle's taxonomy may be due at least in part to his intuitive method, which calls for imagining what speakers and hearers will say (see Searle 1969: 22).

6. Conclusion In this paper, it has been shown that scripted pedagogical dialogues are heavily biased towards the dominant role participant. As this paper is exploratory, the work undertaken points to several avenues of further research. It is necessary to a) develop an operationalised definition of dominant and reactive roles that can be maintained across a range of speech event types; b) further substantiate scripted dialogues as imaginary discourse; and c) study other forms of imaginary discourse in greater depth to see if there is evidence of a similar dominant role bias. If scripted dialogues are taken as representations of speech event prototypes, this suggests that our mental representations of events may be similarly biased. This suggestion may seem counter intuitive, but studies of lexical representations also point to a discrepancy between our mental models and our actual performances. As Sinclair (1991) has shown, prototype definitions of lexical items do not coincide with how these same lexical items are used in natural discourse. It may be the case that we cast ourselves as the dominant role participant in our mental representations of interactive speech events. This does not mean that it is impossible to imagine the contributions of the reactive role participant. Clearly it is not, or fictional dialogue and drama would not exist.

Talk in the Mind: Scripted Dialogues and Mental Scripts

369

Rather, it is suggested that dominant role biased representations of interactive speech events may be the unmarked norm. In a guide to writing fiction, Knight (1981) points out that one of the most pervasive problems with novice writers is their failure to imagine any point of view other than their own. This suggests that normally intelligent adults may have to go to extraordinary lengths in order to hypothetically imagine another point of view. While we are required to imagine our interlocutor's perspective while engaged in conversation, our prototype versions of interactive events may be biased towards our own point of view as the dominant role participant.

References Barber, B. (1979): Communication between doctor and patient: What compliance research shows. In: J. E. Alatis and G. R. Tucker (eds.) Georgetown University Roundtable on Language and Linguistics: Language in Public Life. 119-125. Washington, DC: Georgetown University Press. Blum-Kulka, S., House, J. and Kasper, G. (1989): Investigating cross-cultural pragmatics: An introductory overview. In: S. Blum-Kulka, J. House, and G. Kasper (eds.) Cross-Cultural Pragmatics: Requests and Apologies. 1-36. Norwood, NJ: Ablex Publishing Corporation. Bresnahan, M. (1992): The effects of advisor style on overcoming client resistance in the advising interview. Discourse Processes 15 (2): 229-247. Brown, P. and Levinson, S. (1987): Politeness: Some Universals in Language Usage. Cambridge: Cambridge University Press. Carter, R. and McCarthy, M. (1995): Grammar and the spoken language. Applied Linguistics 16 (2): 141-158. Decapua, A. and Huber, L. (1995): 'If I were you . . .': Advice in American English. Multilingual 14(2): 117-132. Grimshaw, A. (ed.) (1990): Conflict Talk. Cambridge: Cambridge University Press. Hargreaves, R. and Fletcher, M. (1985): Making Polite Noises. Bell and Hyman. Heritage, J. and Sefi, S. (1992): Dilemmas of advice: Aspects of the delivery and reception of advice in interactions between health visitors and first-time mothers. In: P. Drew and J. Heritage (eds.) (1992): Talk at Work: Interaction in Institutional Settings. 359-417. Cambridge: Cambridge University Press. Hill, B., Ide, S., Ikuta, S., Kawasaki, A. and Ogino, T. (1986): Universals of linguistic politeness. Journal of Pragmatics. 10: 347-371. Hutchby, I. (1995): Aspects of recipient design in expert advice-giving on call-in radio. Discourse Processes 19 (2): 219-38. Jacoby, S. and Gonzalez, P. (1991): The constitution of expert-novice in scientific discourse. Issues in Applied Linguistics 2(2): 149-181. Jefferson, G. and Lee, J. R. E. (1981): The rejection of advice: Managing the problematic convergence of a 'troubles-telling' and a 'service encounter.' Journal of Pragmatics 5 (4): 399-422. Also appears. In: P. Drew and J. Heritage (eds.) (1992): Talk at Work: Interaction in Institutional Settings. Cambridge: Cambridge University Press.

370

Susan Mandala

Jones, L. (1981): Functions of English. 2 nd edition. Cambridge: Cambridge University Press. Knight, D. (1981): Creating Short Fiction. Cincinnati, Ohio: Writer's Digest Books. Kress, G. and Fowler, R. (1979): Interviews. In: R. Fowler (ed.) Language and Control. 63-80. London: Routledge and Kegan Paul. Maley, A. and Duff, A. (1978): Variations on a Theme. Cambridge: Cambridge University Press. Malone, M. (1995): How to do things with friends: Altercasting and recipient design. Research on Language and Social Interaction 28 (2): 147-70). Mandala, S. (forthcoming): Exiting advice. Pragmatics and Language Learning. Olearski, J., White, R., and Williams, E. (1991): Formula 3. Macmillan Publishers. O'Neill, R. and McLean, A. (1982): Kernel Three. London: Longman. Schank, R. C. and Abelson, R. P. (1977): Scripts, Plans, Goals, and Understanding. Hillsdale, NJ: Erlbaum. Schiffrin, D. (1985): Everyday argument: The organisation of diversity in talk. In: Van Dijk, T.A. (ed.) Handbook of Discourse Analysis: Vol. 3 Discourse and Dialogue. 35-46. London: Academic Press. Scotton, C. M. and Bernstein, J. (1988): Natural conversation as a model for textbook dialogues. Applied Linguistics 9 (4): 372-384. Searle, J. R. (1969): Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Searle, J. R. (1976): A Classification of illocutionary acts. Language in Society. 5 (1): 1-23. Sinclair, J. (1991): Corpus, Concordance, Collocation. Oxford: Oxford University Press. Traverso, V. (1999): Complementarity in shop encounters. Paper presented at IADA 99 Working with Dialogue, University Birmingham, April 8-10, 1999. Tsui, A. B. M. (1994): English Conversation. Oxford: Oxford University Press.

Acknowledgements The author wishes to thank all of the textbook authors who gave of their time and expertise in completing the questionnaire. The natural talk data were taken with kind permission from the Cobuild Bank of English, and from CANCODE, the Cambridge and Nottingham Corpus of Discourse in English. Grateful thanks go to Rosamund Moon, Ramesh Krishnamurthy, Mike McCarthy and Jean Hudson for their assistance.

Rosamund Moon

Lexical Norms in Spoken English 1. Introduction This paper explores similarities and differences between the lexis of conversational British English and that of fictional dialogue, using corpus data as a basis for the comparison. It focuses on the word: on forms, frequencies, and phraseology. It then suggests how word list frequencies can provide a simple way of assessing how 'conversational' a text is and where it is situated in relation to other texts or text types, through calculation of the density of key lexical items. A number of interesting things are shown up in corpus evidence, and these will be discussed below; however, there are also limitations to this study. First, corpus data is organized at the level of the word, as is the case with the corpus used here, the Bank of English: the subcorpus of spoken interaction gives no phonological information and only limited information on the context of interaction. As Weigand points out (this volume), this makes the corpus format inappropriate for the study of interaction in relation to context, contextual meaning, pragmatics, and dialogic structure. Furthermore, words have been normalized in transcription, according to regular spelling forms, and punctuation has been imposed. Second, the cross-genre or intertextual comparisons presented here are simple and in no way substitute for the detailed statistical measures which characterize the work of Biber (1988) see also, for example, Biber et al (1998: 263ff), Stubbs (1996: 125ff). Third, there are a number of reasons why spoken interaction and fictional dialogue are not comparable: see discussions by Hughes (1996: 37ff), Leech and Short (1981: 288ff), Short (1996: passim), Simpson (1989: 43ff), and Toolan (1998: 105ff). Spoken interaction and its social functions and phatic qualities have to be set against fictional dialogue and its literary functions and narrative qualities: cf Halliday's distinction between spoken language as dynamic, written language as synoptic (1989: 81). At the pragmatic level, spoken interaction and its implicatures are negotiated between the participants, whereas meaning in fictional dialogue is created for and by the third party reader, not the fictional participants in the dialogue: see Hughes (1996: 49, 54-5) Whereas false starts, vague language (Channell 1994), and hesitations are natural features of spoken interaction, these features in fictional dialogue would be interpreted as indices of character - stupidity, evasiveness, and so on - with hesitation and uncertainty conveyed instead through narrative comment such as

Rosamund Moon

372

'he stumbled over his words' or 'she added slowly' or 'falteringly': (Hughes 1996: 51-4, Toolan 1988: 120, and others).

2. The Corpus The Bank of English (BofE) contains 329 million words of current (i.e., 1990s English). This includes 20.2m words of unscripted British spoken interaction of various kinds - conversation, radio phone-ins, lectures, and so on; 18.5m words of scripted, semi-scripted, and unscripted broadcasts from the BBC World Service; and a 42.1m word subcorpus of British fiction and non-fiction. As modes, they represent respectively spoken for spoken, written for spoken, and written for pseudo-spoken for written. The following will largely be based on a comparison between the subcorpus of spoken interaction (SI), and the dialogue extracted from the subcorpus of British books (FD), with occasional reference to the BBC subcorpus (BBC).1

3. Comparisons: Type-Token Counts Table 1 sets out type-token counts for the different subcorpora under discussion. Figures are also given for the numbers of forms occurring more than once (1+), and for the percentages of the subcorpora which are accounted for by the 100 and the 1000 commonest forms. It can be seen that the SI subcorpus is characterized by having far fewer distinct word forms in it than, in fact, any other part of the BofE: SI has a smaller and less varied vocabulary, while its 100 commonest forms account for a greater percentage of text; however, proportionately more of its words occur more than once. The BBC subcorpus, in contrast, has a much larger vocabulary, in keeping with the diversity of the subject matter that it broadcasts. The FD subcorpus is situated somewhere between the two: its lexicon is neither as sparse as SI nor as wide-ranging as BBC. SUBCORPUS SI BBC FD

TOKENS 20.2m 18.5m 8.4m

TYPES 69,234 114,351 107,597

1+ 46,909 69,780 66,043

TOP 100 56% 47% 52%

TOP 1000 76% 73% 76%

Table 1: Type-token counts 1 a. Corpus data is drawn from the Bank of English corpus created by COBUILD at the University of Birmingham, and is reproduced with their permission. b. I am indebted to Jeremy Clear who created this daughter subcorpus, which allowed comparisons to be made more easily. c. I should emphasize that the data for FD presented here is provisional and approximate: we believe that the subcorpus of dialogue contains only the text in fiction which is signalled as dialogue or direct speech, but at the time of writing, the subcorpus has not been finally tidied up.

373

Lexical Norms in Spoken English

4. Comparisons: Word Lists Table 2 sets out the 50 commonest word forms in the same (sub)corpora, arranged in descending order of frequency. Figures are standardized as permillionages (that is, average frequencies per million corpus words) for ease of comparison. Items such as s and t represent clitics ('s and 7). Arguably, frequencies in SI are distorted because of the high incidence of repetition, mainly for hesitation or emphasis: this point is discussed by Nelson (1997: 119). For example, I I occurs 276 times per million words of SI, yes yes 399/million, very very 138/million, and and and 74/million; the respective permillionages for FD are 35, 20, 13, and 6. Such repetition accounts in part for the relatively high frequencies of some words, but it does not explain it entirely. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

SI the (30843) i(29977) and (24130) you (23682) it (20152) to (19785) that (18097) a(17354) s (16776) of(15123) in (11636) er (10820) t (10460) yeah (10291) they (8891) erm (8607) was (8158) we (8088) mm (7954) is (7202) know (7200) but (6543) so (5960) what (5787) there (5765) on (5594) have(5463) he (5338) do (5261) yes (5089) for (5025) right (4965) think (4747)

BBC the (73511) of (31758) to (27058) in (24153) and (20419) a (19863) that (11809) s (10263) is (8940) for (8588) it (7236) on(7183) has (6909) by (6275) be (6117) have(6003) was (5862) as (5578) from (5501) with (5445) he (5215) mr (5072) are (4876) been (4752) at (4698) but (4297) said (4159) they (4053) this (3840) which (3705) an (3535) will (3318) his (3220)

FD the (53382) to (27404) of (25949) and (25190) a (21753) ¡(18449) in (16757) you(13999) it (12876) that (12023) was (11248) he(9705) s(9683) for (7854) is (7594) with (6732) on(6590) t (6424) as (6254) she(6158) be(6002) had (5999) but (5616) her (5525) his (5506) have(5315) not (5038) we (4969) at (4890) they (4751) this (4438) me (3960) all (3909)

BofE the (55248) of (25764) to (24875) and (23474) a (21862) in (18125) that (10288) s (9657) is (9196) it (9157) for (8651) ¡(7731) was (7504) on (6899) he(6439) with (6376) as (5684) you (5348) be(5272) at (5127) by (4909) but (4678) have (4538) are (4358) his (4278) from (4265) they (4242) this (3869) not (3761) had (3444) has (3330) an (3207) we (3152)

374 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

Rosamund Moon well (4742) like (4392) be (4262) ve (4198) this (4119) with (4033) no (4004) at (3946) just (3819) all (3802) re (3791) if (3767) or (3692) mean (3687) as (3669) not (3664) about (3634)

government(3201) there (3117) not (3101) had (2935) were (2892) their (2886) one(2778) president (2697) would (2543) two (2510) who (2453) its (2448) soviet (2286) more (2164) people (2150) now (2036) new (2035)

what (3802) there (3719) my (3661) from (3606) are (3568) by (3563) were (3299) if (3265) one (3120) or (3118) him (3097) so (3089) an (2965) no (2851) can (2832) would (2821) do (2795)

were (2938) or(2863) said (2789) one (2742) there (2717) will (2689) their (2676) which (2637) she (2566) were (2565) all (2522) been (2480) who (2470) her (2353) would (2282) up (2217) if (2182)

Table 2: 50 commonest word forms: permillionages Even taking repetitions into account, there are some obvious discontinuities here, in particular in the relative positions of grammatical words, in the few lexical words which feature here, and in the presence of phatics such as erm in SI. While the is the commonest item in all the subcorpora, it is only slightly commoner in SI than I, the second commonest word; in contrast, in most other text types the is generally at least twice as common as the item in second place. This suggests a different type of deixis in SI from that in other text types. There are striking contrasts between the lexical or open set words in these lists. SI has know, right, think, well, and mean, all semi-lexical words at best, and featuring as or in discourse markers. BBC has government, people, president, Soviet, and new, reflecting the topics reported, and said, representing the metalanguage of reporting: other lexical words in the top 100 forms include country, foreign, political, states, reports, and correspondent. FD has no lexical words at all in its top 50, and the highest placed are in fact know (rank 63) and time (rank 67). While SI has 17 'lexical' words in its top 100, FD has only 9: Table 3 sets them out in rank order.2 Only see and time are higher placed in FD than SI, but see is still proportionately commoner in SI: time is the only word which is commoner in FD, with a permillionage of 1939 as opposed to 1575. Time is typically the commonest noun in written corpus word frequency lists, reflecting its use in functional phrases establishing narrative sequence and temporal deixis.

2

Rank orders rather than absolute or relative frequencies are used in much of the following: cf Sinclair (forthcoming), who makes use of rank orders in comparing word frequencies in two corpora.

Lexical Norms in Spoken English SI 63 69 77 54 21 47 66 32 98 83 85 81 99 33 94 100 34

get go going got know mean people right said say see sort things think time want well

375

FD 77 90 112 110 63 240 95 114 99 127 78 391 163 91 67 116 92

Table 3: (Semi-)lexical words: rank orders At first sight, the grammatical words of SI and FD might seem less at odds. Table 4 sets out the rank orders of articles and (subject/common-form) pronouns. Unlike the and a, an is substantially less common in SI. I, you, and it are slightly higherplaced in SI, but the real discrepancy is the substantially higher frequencies for we and they in SI, and he and she in FD. This may arise from 'we' as a marker of engagement and 'they' as a generic or indefinite in SI, with 'he' and 'she' as the topics of the unfolding narrative conveyed in FD. We and they are twice as common in SI as FD; he and she are twice as common in FD as SI. the a an

SI 1 8 91

FD 1 5 46

I you he she it we they

2 4 28 57 5 18 15

6 8 12 20 9 28 30

Table 4: Articles and pronouns: rank orders Table 5 compares conjunctions and conjuncts. There is little difference between SI and FD in rankings or frequencies for the coordinators and, but, and or, nor with

Rosamund Moon

376

if however, because/cos is substantially more frequent in SI. This may reflect the different ways in which cause and effect are established in SI and FD. Then is 50% commoner in SI, and now is slightly commoner in SI, though ranking lower than in FD: this reflects their uses as discourse markers and sequencers, as well as indicators of temporal deixis. and but or

SI 3 22 46

FD 4 23 43

if because cos

45 71 95

41 119 4534

then now

58 78

70 68

Table 5: Conjunctions and conjuncts: rank orders

Most of the grammatical items so far have been commoner and ranked higher in SI; adverbial and prepositional particles reverse the trend, and feature more strongly in FD than SI. These are set out in Table 6.3 Only about is commoner in SI: this perhaps relates to its functions of signalling approximation and of introducing noun phrases which indicate topic. Of course, many of these items are parts of larger units such as phrasal verbs, but the underlying trend seems to be that FD is either more interested in detailing circumstances, or tends to mention circumstances in adverbial/prepositional phrases: this reflects the informationgiving priority of FD. Note in particular that of is not only more highly-ranked in FD, but is nearly twice as common as in SI. I will return to of later. about at by down for from in into of on 3

SI 50 41 108 120 31 80 11 129 10 26

FD 54 29 39 100 14 37 7 73 3 17

To and like have not been included here and below. The frequency lists only give forms, undistinguished according to word class, and so to represents both preposition and infinitive particle, while like represents verb, preposition, and conjunct.

377

Lexical Norms in Spoken English out over up with

79 145 67 39

53 94 51 16

Table 6: Adverbial and prepositional particles: rank orders

5. Comparisons: Phraseology Discrepancies between SI and FD occur not only at the level of word but also at the level of phraseology. This applies throughout the whole phraseological spectrum, from idioms and fixed expressions to lexicogrammatical patterning. Cf Sinclair (forthcoming), who points out the extent to which high frequency words such as a are not necessarily independent items, but recur as part of phraseological units. Table 7 sets out seven fixed expressions which are known to be high frequency, and gives their permillionages in SI, FD, and BBC.

I mean I think you know

SI (3442) (2967) (4264)

FD (175) (478) (185)

BBC (28) (325) (13)

at (the very) least in (actual) fact of course take place

(122) (232) (422) (27)

(266) (169) (463) (62)

(371) (151) (159) (258)

Table 7: Fixed expressions: permillionages

The grossest distinctions are with the expressions I mean, I think, and you know, which are dramatically commoner in authentic SI than in FD - between 6 and 23 times as common in SI. This is not surprising, given that FD does not replicate the vagueness or hesitation of natural speech. I think demonstrates the least discrepancy - and the highest frequency in BBC: a reflection of its epistemic function in signalling speakers' commitment to their opinions. The other four fixed expressions demonstrate no general patterns, and frequencies appear to vary on an individual basis. At (the very) least, as signal of a minimum or of concession, is twice as common in FD as SI, three times as common in BBC. In (actual) fact, as signal of truth, is considerably commoner in SI than FD and BBC. Of course, as signal of truth or concession, is more than twice as common in SI and FD than in the BBC. Take place, used in recounting events, is nearly 10 times commoner in the reportage of BBC than in SI, and 4 times commoner than in FD. Some contrasts between BBC and SI may correspond to differences in goals of the two modes or text types; FD suggests a mixture of functions.

378

Rosamund Moon

Pure idioms - opaque or semi-transparent metaphorical fixed expressions are generally rare in English (Moon, 1998: 60ff), and so it is hard to make robust quantitative comparisons. Again, they seem to vary in an individual basis. Three idioms that relate to the situational context are hold your horses, talk of the devil, and walls have ears: there were no examples of any of them in SI or BBC, but 7 in FD. However, they are fairly old-fashioned expressions and tied to specific extralinguistic contexts which are unlikely to occur in BBC and are less likely to occur in the kinds of discourses recorded for the SI subcorpus. A further nine idioms were checked, and this time they all had, loosely, some metalinguistic function in that they comment on informational content, truth values, or other utterances, so that they might be expected to be less tied to situation: beat about the bush, have a bone to pick with someone, let the cat out of the bag, lie. through one's teeth, pull someone's leg, put one's foot in it, rub it in, sit on the fence, spill the beans. There was a total of 62 tokens in SI, 62 in FD (a much smaller corpus), and 21 in BBC. (Only 4 of the idioms occurred in BBC, and significantly 16 of the occurrences were for sit on the fence.) This suggests a tendency for the BBC to avoid idioms in general, perhaps as concession to its international audience, and for FD to include more idioms than are found in SI, perhaps fossilizing them. The phraseologies discussed so far have all been multiword lexical items. However, discrepancies between SI and FD can be found in terms of collocation, grammatical form, clausal position, and so on. While of course is roughly equally common in SI and FD, they occur in different patterns. In SI of course is often preceded by and or but, and prefaces some additional opinion or information: Yeah. And of course in those old houses as you say... But of course they the other side did produce... In FD these patterns are rarer: instead of course is often clause or sentence final: Nothing unreasonable, of course: all small-scale stuff. She didn't visit them every week, of course.

Similarly, well in SI is sentence initial 58% of the time, often collocating with hesitation markers, whereas in FD it is sentence initial only 31% of the time. It is not obvious why this should be so. The verb say often occurs in both SI and FD in the infinitive, in strings such as 'I was going to say...'. In SI it tends to occur after a different set of items from FD: more modal verbs or words with modal meaning such as want, try, and inclined, and adjectives and verbs such as hate, pleased, delighted, embarrassed, while regret to say is typical of FD. The verb remember, when positive, is often followed by an -ing form in SI (50% more likely than in FD): it is 3.5 times as likely to be negative in SI as in FD. Furthermore, when negative, remember is typically followed by a full noun phrase in SI, but by indefinite or relative pronouns in FD. These are isolated and subtle cases, but they give an idea of the

Lexical Norms in Spoken English

379

kinds of distinctions which may be observed. The phraseologies of SI and FD are simply not the same.

6. Implications It is easy to provide data to map out lexical distinctions between SI and FD, but less easy to show how it is relevant. It would be possible to build it into an electronic lexicon for writers of fiction, some kind of verisimilitude dialogue checker - but fiction is not necessarily trying to replicate the patterns of SI: FD has a different agenda and set of needs. Besides, it can be argued that FD is its own genre with its own norms and patterns, not just a deviant or debased form of SI or of other written-mode genres. What I am suggesting, however, is that an assessment of large-scale patterns and distributions in SI and FD makes it possible to examine further instances of FD, to attempt to assess its "realism", and see how far the lexis of individual writers or of genres, of dialogues or of the participants in those dialogues, is aberrant with respect to some established norm. If its features are more like those of SI (in general) than FD (in general), then that may imply things about the kind of communicative act which is being created in the text under examination, and the way in which it should be interpreted. This can be stylistically significant.

7. Applications: the Case of Ulysses Applications can be demonstrated by comparing both SI and FD with sections from Joyce's Ulysses. It is of course unfair to compare a highly literary Irish English text, finally published in 1922, with subcorpora of 1990s British English in which the most literary of texts are mass-market fiction. However, the principles can be shown. The eighteenth and final section, the Penelope episode, is Molly Bloom's monologue, approximately 25,000 words long.4 Table 8 sets out the rank orders of grammatical words alongside those in the SI and FD subcorpora of the BofE, to show up places where the lexis of Penelope deviates from SI and/or FD. The article ranks are very similar (with a low rank for an approximating SI rather than FD), but the pronoun ranks reflect the monologic nature, with a high rank for / and comparatively low rank for you and we in particular, whereas he is prominent (much of this monologue concerns the two men in Molly's life). With respect to conjunctions, but is substantially lower in rank than either SI or FD, and or is substantially higher. Cf McKenna and Antonia (1996) who compare the interior monologues of Bloom, Stephen, and Molly in Ulysses-, they discuss Molly's use of 4

To enable these comparisons, contracted forms such as don't have been reconstructed as don and t, in keeping with the practice of BofE.

Rosamund Moon

380

or, suggesting that it reflects her vagueness and imprecision (1996: 61-2). Finally, the 100 commonest word forms include 16 adverbs/prepositions (excluding like): cf 10 in SI, 14 in FD, and 16 in BBC. In particular, after, down, o f f , out, and up are high in rank. One reason for this lies in Molly's use of idiomatic and informal expressions (cf McKenna and Antonia (1996: 64)), another that she is mentioning circumstances in adverbial/prepositional phrases to a greater extent than either SI and FD. the a an I you it he she we they and but or if because then now about after at by down for from in into of off on out over up with

SI 1 8 91 2 4 5 28 57 18 15 3 22 46 45 71 58 78 50 176 41 108 120 31 80 11 129 10 140 26 79 145 67 39

FD 1 5 46 6 8 9 12 20 28 30 4 23 43 41 119 70 68 54 105 29 39 100 14 37 7 73 3 126 17 53 94 51 16

Penelope 1 6 99 2 27 8 4 31 49 23 3 67 25 34 82 43 90 41 50 28 89 58 18 77 9 86 7 63 16 26 80 29 13

Table 8: Grammatical words in SI, FD, and Ulysses: rank orders

This can then be compared with other sections of Ulysses: for example the very formal dialogue in the form of a catechism that takes up the seventeenth section,

381

Lexical Norms in Spoken English

the Ithaca episode, approximately 23,000 words long. Table 9 sets out the 50 commonest forms in Penelope and Ithaca. Whereas Penelope has only one nongrammatical word, old, in the top 50, Ithaca has 6, including 3 names, Bloom, Dublin, Stephen, and 3 others {place, street, years). This, together with the presence of 15 adverbs/prepositions in the top 100 forms of Ithaca, reinforces the importance of the statement of circumstances. Of is the commonest word of all, an and or are much higher than in other data examined, / is infrequent (rank 144), and we barely figures at all.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Penelope the 1199 ¡924 and 652 he 612 to 580 a 537 of 480 it 419 in 414 that 413 was 333 him 304 with 287 me 267 t 227 on 221 all 219 for 219 his 210 like 209 s 208 my 194 they 181 her 175 or 169

Ithaca of 1473 the 1466 and 749 in 526 a 498 to 370 by 266 with 232 s 177 his 174 what 172 he 167 on 142 or 132 at 127 from 127 bloom 123 for 123 had 120 an 104 as 95 did 94 its 85 was 79 which 78

26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

Penelope out 164 you 156 at 146 up 138 when 129 she 125 as 122 what 120 if 118 there 116 have 113 be 108 d 108 so 106 not 100 about 99 them 97 then 95 yes 91 too 89 had 86 old 83 is 80 we 79 after 77

Ithaca Stephen 65 not 62 their 60 her 58 street 53 that 53 be 48 him 48 were 48 she 46 been 39 it 39 when 38 one 37 all 36 between 36 would 35 more 34 two 33 years 33 dublin 30 if 30 other 30 place 30 then 28

Table 9: Comparative word frequencies in Ulysses

It is also possible to chart frequencies and distributions of the utterances of individual characters. The fifteenth section, the Circe episode, is in play format. The three characters who speak most are Bloom (6186 words), Stephen (1744 words), and Zoe (856 words). This is too little data to comment on more than the very highest frequency words, but the 20 commonest forms are shown in Table 10, and they show a little of the differing priorities of the three speakers - the you focus of Zoe, the stronger / focus of Bloom, the heavier use of the and lighter use of a by Stephen as against Bloom, perhaps reflecting different proportions of

382

Rosamund

Moon

given and new information. These suggest tendencies which can then be explored through conventional examination of the text itself.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Bloom the 257 ¡201 a 163 of 158 you 122 and 107 to 93 in 87 s 77 it 74 that 69 is 57 for 54 was 50 me 48 my 47 have 41 on 39 with 36 all 34

Stephen the 89 of 53 and 37 i 37 to 31 is 28 a 25 you 21 in 20 my 19 s 18 it 17 for 14 not 14 that 14 or 13 ali 12 me 12 his 11 this 11

Zoe you 37 the 34 a 31 s 24 i 23 and 15 for 14 that 12 it 11 of 11 to 11 on 9 19 your 9 118 me 8 what 8 him 7 no 7 who 7 with 7

Table 10: Three speakers in Circe

8. A Comparative Measure Frequency counts of individual texts show great degrees of variation, particularly in terms of which lexical items occur in texts, and, as has been seen, even grammatical words vary in frequency and rank order from text to text. Many of the distinctions between SI and FD result from the interactive communicative needs and nature of SI, and the narrative needs and nature of FD. The interactive and the narrative become two markers or parameters by means of which the nature of a text may be assessed. Although the exact order and relative frequency of items vary, there is some gross consistency between genres or modes and the frequency lists of their lexis. All 16 subcorpora of written text or of broadcast journalism in BofE show the same six words as the six commonest, and this can be confirmed from other written corpora: there is typically a sharp falling-off in frequency between them and the word in seventh place. These six words are a, and, in, of, the, and to. I have found less agreement between the SI subcorpus and other collections of spoken English, but provisionally, let us take the six commonest items found in the BofE: they are also the six commonest items in the CANCODE corpus of

Lexical Norms in Spoken

English

383

spoken English (McCarthy and Carter 1997: 23). They are and, I, it, the, to, and you. This gives us three words common to both lists (and, the, to); three words from written or scripted texts: (a, in, of)\ and three words from spoken texts (I, it, you). Of course, all of these words would normally feature amongst the most frequent in most texts, but the relative proportions can be taken as key indicators of, on the one hand, the 'writtenness' or degree of narrative or informational content of a text or part of a text, and on the other hand, the 'spokenness' or degree of interactivity, by calculating the percentage of text taken up by each of these groups of three words. This can be termed the Core Function Word (CFW) density of text, and it can then be used to compare different texts with a norm, established from a research corpus such as the BofE, to compare texts within a corpus of a particular writer's work, or to compare local parts of a text with overall figures, and so on. Cf the exploration of lexical density in texts and genres by Ure (1971) and Stubbs (1996: 7Iff). 5 The three groups of words correspond, simplistically or crudely, with Halliday's stratification of meaning into three levels or functions: ideational, interpersonal, and textual (1978: 116ff). The writtenness (W) indicators can be associated with informativeness and the ideational: typically, a introduces new information, in introduces adverbial groups in which circumstances are stated, of features in complex noun groups. The spokenness (S) indicators can be associated with interactivity and the interpersonal: I and you represent participants in the dialogue, it is typically associated with evaluative structures and endophoric or exophoric reference. The common (C) indicators can be associated with cohesion and the textual: and has a linking function, the signals given or shared knowledge, and to as infinitive marker also links, although it has other functions as well. We can now apply the CFW measure to the corpora and texts under scrutiny. The evidence of written texts in BofE suggests norms for written texts of around 10.5% C indicators, 6.5% W indicators, and 2% S indicators (actual average figures for written British English are 10.3% C, written 6.7%, and spoken 1.8%). It is less easy to feel confident about calculating the corresponding norms for spoken English from the BofE alone, but these can be set provisionally as 7.5% C indicators, 4.5% W indicators, and 7.5% S indicators. We can hypothesize that the higher the figure for the density of W indicators, the more information-centred or narrative-centred a text is; the higher the figure for the density of S indicators, the more interaction-centred a text is. BofE contains a large tranche of journalism, and the CFW densities reflect neat stratification of sub-types: the broadsheet newspapers (Times, Independent, Guardian) occupying the middle-ground; the magazines and tabloid newspapers (Sun, News of the World, Today) having lower 5

Some remarks by Sinclair (forthcoming) led me to develop this measure; any fallaciousness is entirely my own.

Rosamund Moon

384

ranges of densities of W and C indicators, but higher of S; and the semi-specialist journals New Scientist and the Economist having higher densities of W and C indicators, but lower of S. Table 11 sets out the CFW densities for the SI, BBC, and FD subcorpora in BofE, and the Ulysses episodes.

? WRITTEN NORM

COMMON INDICATORS and, the, to 10.5%

WRITTENNESS INDICATORS a, in, of 6.5%

SPOKENNESS INDICATORS I, it, you 2%

? SPOKEN NORM

7.5%

4.5%

7.5%

SI BBC FD (BofE WRITTEN

7.5% 12.1% 10.4% 10.3%

4.4% 7.6% 6.3% 6.7%

7.4% 1% 4.5% 1.8%)

Ithaca Penelope (Ulysses overall

11.4% 9.8% 10.1%

11% 5.7% 7.3%

0.3% 6% 2.8%)

Table 11: CFW density

SI establishes its own norms. The figures for BBC approximate the written norms, rather than spoken, with a high figure for C indicators: they reflect, perhaps, the cohesiveness of the broadcasts, and the amounts of both given and new information, and of detail. The figures for FD are between those for written norms and spoken norms: slightly lower in W indicators than the written norms, midway in S indicators between written and spoken norms. Ulysses overall corresponds broadly to the written norms; however, the monologue Penelope moves much closer to the spoken norms, and the catechism Ithaca moves much further away, with an extreme score for W indicators of 11%. Some measure of the differences between the three most frequent speakers in the play Circe can be detected: the score for C indicators for Bloom, Stephen, and Zoe are, respectively, 7.4%, 9%, and 7%; for W indicators, 6.6%, 5.6%, and 5.5%; and for S indicators, 6.4%, 4.3%, and 8.3%. By extracting further amounts of dialogue from Bloom and Stephen separately, it may be possible to see if the scores are repeated and if so, what the literary implications are: for example, if Stephen is generally more textual and less interpersonally-oriented than Bloom.

Lexical Norms in Spoken English

385

9. Conclusion Clearly, more investigation is needed to see if the hypothesized CFW norms can be corroborated by different corpora: in particular, to examine further corpora of spoken interaction of different kinds. One step would be to work through the mathematics, to see how far statistics supports a CFW measure; however, the point of a CFW calculation is that it is simple and fast, and does not need anything more technical than easily derived wordlist counts and basic arithmetic. Another obvious step would be to apply CFW measures to further dialogues, whether fictional or spontaneous, to see whether a coherent picture continues to emerge in the way that it appears to with the data examined so far. If it does, then assessing CFW density will provide a means of assessing the degree to which a dialogue is oriented towards the informative or the interactive, and for fiction, the degree to which a narrative unfolds through dialogue which is informative rather than phatic, expressive, or quasi-interactive. This has its own utility in the analysis of dialogue of different kinds.

References Biber, D. (1988): Variation across Speech and Writing. Cambridge: Cambridge University Press. Biber, D., S. Conrad, and R. Reppen (1998): Corpus Linguistics: Investigating Language Structure and Use. Cambridge: Cambridge University Press. Channell, J. (1994): Vague Language. Oxford: Oxford University Press. Halliday, M. A. K. (1978): Language as Social Semiotic. London: Edward Arnold. Halliday, M. A. K. (1989): Spoken and Written Language. Oxford: Oxford University Press. Hughes, R. (1996): English in Speech and Writing. London: Routledge. Joyce, J. (1922): Ulysses. (1968, London: Penguin.) Leech, G., and M. Short (1981): Style in Fiction. London: Longman. McCarthy, M., and R. Carter (1997): Written and spoken vocabulary. In: N. Schmitt and M. McCarthy (eds.) Vocabulary: Description, Acquisition and Pedagogy. Cambridge: Cambridge University Press, 20-39. McKenna, W. and A. Antonia (1996): "A few simple words" of interior monologue in Ulysses: reconfiguring the evidence, Literary and Linguistic Computing, 11/2, 55-66. Moon, R. (1998): Fixed Expressions and Idioms in English: a Corpus-based Approach. Oxford: Oxford University Press. Nelson, G. (1997): A study of the top 100 wordforms In: ICE-GB text categories, International Journal of Lexicography, 10/2, 112-134. Short, M. (1996): Exploring the Language of Poems, Plays and Prose. London: Longman. Simpson, P. (1989): Phatic communion and fictional dialogue in R. Carter and P. Simpson (eds.) Language, Discourse, and Literature. London: Routledge, 43-56. Sinclair, J. (forthcoming): A way with common words. Stubbs, M. (1996): Text and Corpus Analysis. Oxford: Blackwell. Toolan, M. (1988): Narrative. London: Routledge. Toolan, M. (1998): Language in Literature. London: Arnold. Ure, J. (1971): Lexical density and register differentiation. In G. Perren and J. L. M. Trim (eds.) Applications of Linguistics. Cambridge: Cambridge University Press, 443-52.

Charles Owen

Naturally-Occurring and Artefactual Dialogue 1. Introduction This paper examines the distinction between 'naturally occurring' dialogue and dialogue which is not 'naturally occurring'. There are two related reasons for doing this: a practical problem in forensic linguistics, and the theoretical issue which it somehow inevitably raises, both to be outlined and discussed below. Before proceeding there are two preliminaries. One terminological matter needs to be clarified. The term 'naturally occurring' is very common in the literature. I presume that it is not just a semantically empty slot-filler, but that it makes a real contrast with a different category of dialogue not covered by this label. What should we call this other category? I propose here to call it 'artefactual'. The word is inelegant, but recommends itself on the grounds that it is briefer than 'non-naturally occurring', and free of what, arguably, are unwanted connotations inherent in words such as 'artificial', 'fabricated' or even 'scripted'. Next, as a means of illustrating some of the points to be made later, the reader is invited to study the eight dialogues reproduced below, and to decide which are a transcript of naturally occurring conversation, and which are artefacts (in whole or in part), e.g. scripted material. Try to classify each dialogue as: I = U = Pos. = Pr.

=

impossible (this cannot possibly be a record of an actual conversation) unlikely (this is unlikely to be a record of an actual conversation) possible (this is possibly a record of an actual conversation, but there are still features which make it suspicious) this is probably a true record of an actual conversation (one can never be certain).

Dialogue 1 T C T C

I've just seen a police car running at top speed. What's it doing? Perhaps it's running after some gangsters who are still at large. Some gangsters? There are gangsters in all directions. Can the police ferret them out? It's true that bad characters can't be easily stamped out. But the police should still go all out to hunt for them here and there.

Dialogue 2 M W

Oh, William, are you speaking truthfully? You may believe me, sweetheart. How is the child?

Charles Owen

388 M W

Oh, William, it is very ill. William, it is not breathing! It is dead! Our little one is dead! Hush, my beloved, be comforted. Sad though it is, without the child we shall travel faster and safer.

Dialogue 3 [L and P are talking. Graham enters the room.] L P L P L P L P L P L P

You know Graham, Pat, don't worry. I am worried. I'm terrified of the others. They have threatened me with personal injury. Who are the others, Pat? Do you mean the persons who were on the job with you? Yes sir. Well who are they? We need to know the complete story. Vic Hodgson and his relation Martin, and Johnny. Do you mean Johnny Richardson? Yes sir. Tell me about the car Pat - the one you took to the farm. There was a blue Cortina Estate and a van that Johnny borrowed from someone at the Fox and Hounds. It had a white top. How did you travel, Pat? Who did you go with? I went with Vic Hodgson. He drove. Martin, his relation, came with us. I sat in the back.

Dialogue 4 A H A H A H A H A H

Are you still in philosophy? No, unfortunately. Ah. What are you doing now? I am a what do you say - a cleaner. A cleaner? What is that? Cleaning. Washing. With a brush and a bucket. I am cleaner at the bus station. You wash buses? No, not buses - the lavatories, the floors where people walk and so on. Oh. I see. You're a cleaner. Yes.

Dialogue 5 BB CO BB CO BB CO BB

Well then Charles, er...do you live in a house? a flat? or a bungalow? I live in a house, in er, Edgbaston. Right. Is that semi-detached? or detached? It's a semi-detached house. It's got a flat at the top of it. It's quite a large house. Okay, so er, tell me where you live. Where do you live? in a house? or er... I live in a - in a semi-detached, er, house, in er, West Heath. An old house or a modem house? Erm...nineteen thirties.

Dialogue 6 DC

I've got two girocheques which are also in that name. The first one is 940671032 for three hundred and eleven pounds thirty-five. It's made out to Mrs E. and the address is - although it says 16 Lidiham Road, I believe that to be 16 Lytham Road.

Naturally-Occurring MM DC MM DC MM DC MM DC MM

and Arte/actual

Dialogue

389

Yeah. Did you receive this girocheque? Yeah. You did. And a second girocheque - 940671007 for three three nine pounds thirty pence - again made to Mrs E. Did you receive that girocheque? Yeah. Did you - did you in fact cash this girocheque? Yeah. You did. Mm.

Dialogue 7 SI MA SI MA SI MA SI MA SI MA SI MA SI

Where did that money go to? Uh? That money - oh lah, I was helping with - like I was - send it - oh... Send it. Send it to like - like - well - well - sometime - because I was - send when was like Rwan - Rwanda - send it some money there. Raw - warer - Rowanda. Rwanda. But you're from Zaire. Yeah. So why would you send money to Rowanda? Because I was - because I was see in the television the people was dead. So how much - how much did you send to Rowanda? I can't - 1 can't remember again. Right. Do you know anything about the address at 9 James Close?

Dialogue 8 MS MJ MS MJ MS

MJ MS MJ MS MJ MS MJ MS MJ

Hello, Mrs Johnson? Hello. Yes, hello, it's Martin Skinner here - you may remember me from the Euro Ads from last year? Good gracious, I thought we'd heard the last of you. Oh no, no, no you're not getting rid of me that easily - I'm not stopping until every business has sent off for their Euro preparation fact sheet, so have you managed to do that? Well, er, I did actually Mr Skinner.... Excellent ...last year. Oh, very good. And I contacted my local business link. Oh well, that's excellent work - erm... So if you just um allow me to get on... Oh yes of course, I wouldn't erm... take up any more of your valuable time. Thanks very much Mrs Johnson, nice to talk to you. Good bye.

390

Charles Owen

2. The Forensic Linguist's Problem In many court cases, records of conversations constitute crucial evidence. Since the introduction of the taping of interviews in police stations under the Police And Criminal Evidence Act of 1984, courts can of course listen to the conversation as well as read a transcript (although this seldom happens in practice). However taping of interviews does not entirely remove the importance of transcripts. The following are examples of conversations for which only a written record will be available: Pre-PACE cases, dating from before the time when taping was mandatory. •

Conversations with suspects outside the formal interview, especially between arrest and arrival at the police-station.

•

Other non-interview material, taken down by the police as evidence. Statements by witnesses, or possible witnesses, come into this category. Interviews where no tape exists, in English speaking jurisdictions where taping is not mandatory (e.g. Republic of Ireland) or in non-English speaking countries where an English speaker is interviewed in English.

When an accused person disputes the written record, perhaps alleging fabrication or doctoring, a forensic linguist may be engaged to give an opinion on the likelihood of the record being accurate. Even partial mistranscription, whether deliberate or accidental, e.g. insertions, omissions, alterations etc., could in theory invalidate the whole of the written record. Inevitably, there is a certain degree of 'tidying up' in any transcription, and that is accepted. The difference between tidying up and falsification may not always be easy to decide in practice, but it is certainly a real difference in principle. It is important to be clear about the precise role of the linguist. The linguist is giving an opinion, based on the written text in front of her/him, as to whether this conversation actually took place in the way the record suggests. Notice that this is not an opinion about whether the participants were telling the truth, or whether they intended to express the meanings which appear to be there on the page. Those are matters for the court as a whole to decide. Of course, if the linguist can show that the conversation is unlikely to have occurred in the way it appears on the page, then that will influence the court's view of these other matters, but the two areas are necessarily distinct. If the linguist thinks it is not an accurate record, he or she is effectively saying that the transcript looks implausible as a record of 'naturally occurring' discourse: the conversation could not have happened like this on linguistic grounds. To a greater or lesser extent, the record is 'artefactual'. The court then has to decide whether to attach any weight to this opinion. If it does not, it is effectively saying to the linguist: 'You may think this looks artefactual, but you

Naturally-Occurring

and Artefactual

Dialogue

391

haven't convinced us. We think the accused could have said, did in fact say, these words, in that context.' More significantly, the court may even be implying that there are no solid criteria available which would enable anyone to say that a transcript is artefactual rather than a true record of what occurred naturally. This is a challenge to the forensic linguist.

3. The Theoretical Question Now the theoretical question. Simply phrased, this is: what precisely are the solid criteria which will enable us to judge whether a manuscript is a transcript of a 'naturally occurring' dialogue, or whether it is partially or wholly artefactual? In the literature on spoken discourse, there seems to be a very clear double presumption that: a. b.

the object of inquiry should be 'naturally occurring' talk, and the meaning of 'naturally occurring' is self-evident

The impulse for the first presumption surely stems very largely from the work of Malinowski and Firth, whose (1935) claim that in conversation 'we shall find the key to a better understanding of what language really is and how it works' is cited widely. Despite Firth's remarks, it is clear that not much work was possible until the widespread availability of recording equipment, a fact made clear in Abercrombie's 1965 short but classic article 'Conversation and Spoken Prose'. He has a 3-way distinction between 'reading aloud' (which includes all material learnt by heart from a written script) monologue and conversation. About the last, he says that it contains 'the most natural, the most frequent, and the most widespread occurrences of spoken language.' (1965: 3). With regard to the differences between conversation and spoken prose, his article focuses largely on intonation and performance features which are hard to capture on paper. I shall return to the significance of performance features later, but need to make two practical points here. Whatever developments there may have been in our ability to transcribe spoken language in such a way that a reader can reconstruct more accurately what happened, the methods and systems employed are normally unsuitable for forensic purposes, because a court would have difficulty in understanding them. Explanations by the forensic linguist need to be more or less instantly comprehensible to a judge and jury. Moreover, it is an unfortunate fact that there is no such thing as a 'correct' transcription. Leaving aside variations in tidying up, and the impossibility of capturing performance features of speech, it has been shown that even trained stenographers vary

392

Charles Owen

considerably in the records they produce. This applies to a range of linguistic parameters, e.g. vocabulary, syntax and who said what in each turn; it happens even in cases where a tape can be listened to over and over again. Neither of these practical problems represent insuperable obstacles to the forensic linguist in my view, but they need to be remembered. Goodwin (1981: 33ff.) offers an extended discussion of what it means for unscripted conversation to be natural. The main distinction drawn is between 'naturally occurring' and 'experimentally elicited' data, and he discusses in particular the methodological and ethical issues surrounding the collection of the former. He quotes approvingly Labov's (1972: xiii) description of natural conversation as: ...language as it is used in everyday life by members of the social order, that vehicle of communication in which they argue with their wives, joke with their friends, and deceive their enemies.

There is a possible implication here that when a doctor argues with his wife his conversation 'occurs naturally', but when he sees his patients his conversation has something unnatural about it. The same would apply to teachers joking in the staffroom one minute and teaching their pupils the next. This seems problematic. Whilst we should have no difficulty in distinguishing the two types of conversation in terms of register or predictedness, doing so in terms of 'naturalness' raises more questions than it answers. Stenstrom's more recent introductory text (1994: xii) says: 'The description is based on naturally occurring spoken interaction, as manifested in the LondonLund Corpus of English conversation.' Examples of 'naturally occurring spoken interaction' include the subcategories of 'interview', 'discussion' and 'conversation'. While these are all said to 'occur naturally' one important difference between them is certainly the degree of predictedness. A further possible implication here is that naturalness should not be seen as a property of the individual interaction, but of an accumulation (preferably large) of different types of spoken interaction. From a descriptive point of view this is attractive because it allows a great range of spoken data into the description. Whether this helps forensically is less clear, since the forensic focus is on the individual interaction. Comparing a disputed transcript with some essence of naturalness distilled from a large corpus might have superficial appeal, but the logic of a corpus-based description is that in capturing the typical it necessarily disguises or smoothes over the untypical. We would then be faced with the task of persuading a court that the disputed transcript was in fact implausible or impossible, rather than merely untypical - not an easy task.

Naturally-Occurring

and Artefactual

Dialogue

393

In the field of language-teaching discourse, there has been a steady stream of work by Sinclair and others (e.g. Sinclair 1982, Willis 1987, Trickey 1988) which has distinguished between planes or levels of discourse, sometimes labelled 'outer' and 'inner', in which 'outer' is 'natural' and 'inner' is not. 'Inner' is often scripted as well, but not necessarily so; at least, not in the sense that words are written on paper and then read. If we adopt a slightly modified definition of scripted to mean something more like 'orchestrated', then 'inner' is certainly scripted in that sense. Although the pedagogic-evaluative stance towards the inner/outer distinction varies, there does seem to be quite a good consensus that the natural can be distinguished from the artefactual not only by linguists and teachers, but also by learners, despite the occasional misunderstanding. However, this work relates to language-teaching, and it is not immediately clear that the mutually dependent categories of 'outerness' and 'innerness' apply to other forms of talk.

4. The Contribution of Stylistics There is one area of work on discourse which differs in regard to the two presumptions stated earlier. Burton's (1980) Dialogue and Discourse was the first (and possibly still the best) attempt to explore the relationship between 'naturally occurring' discourse - she specifically uses the term - and something which is manifestly 'artefactual', i.e. the dialogue of modern drama. Burton also takes it as axiomatic that 'naturally occurring' discourse is quite unlike any dialogue to be found in a drama script, and vice versa, and she is disdainful of literary critics who praise authors for capturing the true essence of real dialogue. But she does not just leave the matter there. What interests her is how consumers of drama come to perceive these fictions as 'real' or at least 'naturalistic'. This, of course, is our problem as well. In other words, despite the fact that a drama script may lack many of the features of 'naturally occurring' conversation, why is it that we are nonetheless disposed to say that it is 'naturalistic'? Correspondingly, in the case of a drama script which is absurd, what is it that makes it absurd? And most significantly for us, why is it that a court might choose to regard an artefactual dialogue as a real one? The argument which Burton proposes, and which is restated by most of the writers who have followed her, is that by studying drama scripts we can learn a lot about 'naturally occurring' conversation which we would not otherwise learn if we restricted ourselves to more clearly defined social contexts such as the classroom or the doctor's surgery, i.e. the more formal or predicted interactions. This at first seems to express a very different sentiment from that of Labov cited earlier. How, we may ask, is it possible to argue on the one hand that it is 'naturally occurring' discourse that will reveal to us the true nature of language,

394

Charles Owen

while on the other hand claiming that a study of artefacts such as dramas will enhance our knowledge of the natural? Hess-LUttich (1985: 204) states the paradox explicitly: ...there seems to be a contradiction between claiming a principal structural identity between factual and fictional dialogue, which allows the latter to be regarded as a proper basis for hypotheses on the former, and stating a modal difference, which does not allow us to quote evidence based on fiction as evidence based on reality.

The resolution is that dramatists model their work from their practical experience of 'naturally occurring' dialogue, and are thus offering us distillations or condensations of the real thing. More recently, Herman (1995) discusses socially contextualised 'authenticating conventions' as the handles by which dramatists manipulate audiences into believing that what they hear is real, while Simpson (1998) seems to echo Burton in his discussion of incongruous or absurd dialogue: ...while everyday speech and drama discourse are not homologous modes of communication, they are none the less parallel, and expectations about well-formedness in everyday speech form the benchmark against which aberrant and incongruous discourse can be measured.

Underlying the case for studying drama scripts is also the theme of 'Verfremdung' or alienation; if incongruity informs us about normality, it does so by the process of foregrounding and consequent estrangement of the reader/viewer from the familiar. Burton's idea is to extend what was originally a literaiy device to linguistics and to characterise discourse analysis as alienation. Whether Brecht would have been happy to be harnessed to this argument can only be speculated upon, but she quotes him on alienation: What is obvious is in a certain sense made incomprehensible, but this is only in order that it may be made all the easier to comprehend. Before familiarity can turn into awareness, the familiar must be stripped of its inconspicuousness.

To summarise, stylisticians see the relationship between drama and 'naturally occurring' discourse as reciprocal, i.e. both as evidence contributing to a general theory of how talk happens, despite the fact that it is not 'naturally occurring', and as a means of sharpening our perception of the dramatist's style.

5. The Forensic Linguist's Problem - Reprise Where does the forensic linguist's problem fit in to this picture? The forensic task is, in a sense, a mirror-image of the stylisticians task. The linguist does not have the luxury of knowing in advance that a manuscript is, for a fact, natural or artefactual. Still less is the primary aim of studying the text to enhance theoretical

Naturally-Occurring and Arte/actual Dialogue

395

understanding of discourse. What the forensic linguist wants to do is to focus precisely on some of those differences which have been taken as read by linguists. If linguists from Abercrombie to Simpson and so many in between are clear that 'naturally occurring' spoken discourse is quite different from spoken prose, or scripted dialogue, it should, on the face of it, be a relatively straightforward matter to persuade an intelligent judge that the difference is easy to spot. But it isn't. Is this because transcribed conversation is a pale reflection of the original? As Abercrombie says, 'one can't really get down a conversation in ordinary writing' (1965: 6). However, several writers (Burton 1980: 115, Hess-Luttich 1985: 203, Short 1989) have also stressed that the absence of typical performance dysfluencies in artefacts can be considered an asset to the linguist, since it leaves the global patterns uncluttered and more readily discernible. This argument must surely also apply to that other substantial component of spoken discourse, i.e. intonation. Losing it in transcription should not, according to this reasoning, damage the analyst's ability to make strange the remaining material - even to a judge. Thus, if the differences between natural and artefactual are so glaringly obvious at levels above that of phonic substance, it ought to be possible to show that a written conversation is or is not a record of an actual conversation. In an ideal world, it ought further to be possible to show which parts of a written conversation are 'spoken prose' or 'artefactual' and which occurred naturally. The fact that any such statement will inevitably be probabilistic is true but should not cause us too much anxiety. In most cases, we should be able not only to distinguish the implausible from the merely untypical, but to make the difference obvious to others. Among the possible reasons for this difficulty are that the distinction between 'naturally occurring' and its alternative has been exaggerated, that the differences between different types of talk are more complex than has been realised, or a combination of these reasons. The illustrative material accompanying this paper is intended to provoke thought on these issues.

6. Marginal Cases Let us consider what might be termed 'marginal cases' - marginal in terms of whether they are 'naturally occurring' or not. Radio advertisements Normally, if we switch on our television and hear people discussing the merits of a certain brand of detergent we detect at once that this is scripted. This happens even if we ignore give-away features such as jingles; we can tell just from the way the characters talk that the material is acted. By 'the way the characters talk' is

396

Charles Owen

meant not just things like exaggerated intonation, but the content, the pragmatic sequencing, the vocabulary, syntax and so on. Geis (1982) criticises these minidramas from an ideological standpoint, calling them unnatural, and also deceptive, subversive even. In an extensive rebuttal, Cook (1992: 205-210) argues that minidramas are perfectly natural in their context, i.e. advertising. Regrettably, there are two uses of the term 'natural' in circulation here. Geis is saying that these conversations would not happen in a normal social situation. Cook is saying that these foregrounded, scripted interactions are natural as communicative events given that they are intended to be heard by third parties. They are both right, and it is odd that there should be any argument about it. We should do well to remember though that any falsified record of a police interview is also intended for consumption by third parties, so, ironically, it may be natural in Cook's sense, but is not necessarily so in Geis's sense. Improvisations If we stumble into a classroom being used by drama students, we generally know whether it is a rehearsal we have interrupted, or a seminar. Yet this simple dichotomy is not foolproof. My office is just along the corridor from our drama department, and sometimes, as you are passing their rooms, people spill out into the corridor, shouting, waving their arms etc. The inference I draw is that the interactions are scripted; the participants are acting. But consider this: a student rushes up, seizes my arm and yells in my face. When I fail to respond, she flings my arm away disgustedly and pounces on some other passer-by. I'm not sure whether I should have said something or not. Am I in a 'naturally occurring' interaction or not? This is recognisable as an improvisation in which the passer-by is an involuntary stage prop. We may say that the crucial point is that improvisers are mouthing words as part of a dramatic exercise, not because they really want to communicate with each other. To return to our earlier discussion of planes of discourse, this could be a case of 'inner' discourse, similar to a scripted drama. Breaking character, as a deliberate alienation device, or through incompetence, would represent a shift to the 'outer' plane. Moreover, in my role as passer-by, I don't answer when the student shouts at me because I instinctively realise that it will make no difference. But this can be a feature of quite normal conversations as well. So the question remains: what linguistic features of this event can safely be relied on as clear markers of artefactual dialogue?

Naturally-Occurring

and Artefactual

Dialogue

397

Foreign language teaching dialogues There are two aspects: pedagogic materials and the dialogues attempted or produced by learners in class. Most dialogues in pedagogic materials are not really marginal at all, but overtly scripted, and consequently not only artefactual but highly artificial. This has been held in some quarters to be detrimental to learning. On the other hand, the problems of using 'naturally-occurring' dialogue as teaching material - the sheer difficulty confronting the learner - are also widely known. In practice, most teaching seems to co-exist happily with the concept of 'degrees of naturalness' in regard to pedagogic material. Learner-produced dialogues also clearly lie on a continuum from repetition of scripted material, through tightly-controlled practice, to schematically determined simulations, to complete autonomy. Purist proponents of the inner/outer distinction presumably hold that it is only where learners are discussing something other than what their teacher asked them to discuss that their dialogue is occurring 'naturally'. Can the flexibility with which we can discuss FLT dialogues also apply to non-pedagogic dialogues, or is this a peculiar feature of language whose main purpose is the improvement of language?

Schematically formulaic interactions The most obvious example of formulaic interactions are those in which the turns of one of the participants have been partly or wholly decided in advance 'interviews' in Stenstrom's terms. Even where only a few of these turns are predictable, they may be some of the most important organising turns. In doctorpatient interviews, for example, whether or not the turns have actually been written down seems less relevant if they have been practised hundreds of times. Even where two people have never met before, and are in a completely unstructured relationship, phatic communion can extend into long conversations of a partly predictable nature. My chosen example here is anchor-lifts at ski resorts the kind where you find yourself trapped next to someone you are unlikely ever to have a conversation with again. The close physical proximity means that complete silence is almost impossible to maintain, yet the time available means that you are unlikely to be able to develop a serious topic. The conversation may safely categorised as 'desultory' but it may also be partly or substantially 'formulaic'. These somewhat impressionistic observations seem to suggest that even socalled casual conversation may be more routine, more predictable, than sometimes suggested.

Charles Owen

398

1. The Dialogues Here I identify the dialogues and comment briefly on each. The categories which were offered at the beginning of this paper were: I = U = Pos. = Pr.

=

impossible (this cannot possibly be a record of an actual conversation) unlikely (this is unlikely to be a record of an actual conversation) possible (this is possibly a record of an actual conversation, but there are still features which make it suspicious) this is probably a true record of an actual conversation (one can never be certain).

Dialogue 1 Source: Lim (1979): Oral Drills in English Idioms. Singapore: Educational Publications Bureau. This is a classroom textbook. The introduction does not specifically say that the dialogues are scripted rather than records of actual conversations, but since the book presents 428 idioms, it seems probable that the dialogues were specially written to illustrate them. Most people who have been asked to judge this dialogue regard it as non-native in tone, but even making allowances for that, they also regard its declamatory and moralistic content as untypical of naturally occurring conversation. Usual score: I. Dialogue 2 Source: Cox, C. (1969): Maria Marten or Murder in the Red Barn - A Melodrama. London: Samuel French. The title is self-explanatory. But it might be worth noting that even by the standards of melodramatic dialogue, William's callous mitigation of the baby's death seems remarkably 'written' in style, which is doubtless deliberate. Usual score: I. Dialogue 3 Source: Dialogue extract derived from a transcript of a disputed conversation in an actual murder case. The identities have been concealed. The original document is a statement by a police officer (L in the extract) in which he reports on a conversation he claims to have had with an accused man (P). It is in the form of a narrative, but the alleged utterances by himself and P are in direct

Naturally-Occurring

and Artefactual

Dialogue

399

speech shown by speech marks. This extract simply removes all the narrative part of the report - He said... I said...etc. - and leaves the direct speech. The report is based on notes made at the time, not on a tape recording, so L's version of the conversation is necessarily a reconstruction. The question at issue was whether it was reasonable to question the authenticity of the conversation on the basis of what L reported was said. Most respondents regard this as a possible conversation, but not many are willing to label it Pr. Three frequently mentioned areas of suspicion are: • •

•

the expression 'threatened me with personal injury' - seen as untypical of spoken language, but perhaps possible from a policeman. P's use of the word 'relation' - seen as too underspecified, and thus indicative of lack of knowledge by L (either because he had forgotten, or because the conversation simply hadn't happened). the use of colour adjectives to identify the Cortina and the van - seen as informationally 'heavy' - and by inference designed for third party consumption.

Dialogue 4 Source: Stoppard, T. (1978): Professional Foul. London: Faber. A television play by a dramatist known for his skill at producing naturalistic dialogue. While some people are seduced by his skill, and give this Pr, others find it too clever to be true - and they are right. Obviously the extract is very short and carefully selected. More data might resolve the matter more firmly. Dialogue 5 Source: Willis, J. and D. Willis (1988): Collins COBUILD English Course 1: Teacher's Book. London: Collins. This is a transcript of a conversation between me and somebody with initials BB, whom I no longer remember. The point to note is that it was produced under controlled conditions. We were seated in a laboratory, handed a card with a topic, and asked to 'have a conversation' about it. This conversation was then recorded and transcribed as a model for use in the coursebook. Indications of this are present in my rather ponderous switching of the turn to BB: Okay, so er, tell me where you live. While most respondents are happy to give this Pr, there is also general recognition that the term 'naturally occurring' has to be stretched considerably to include conversations of this kind.

400

Charles Owen

Dialogues 6 and 7 Source: My transcription of a tape of a police interview in a major fraud case. The two dialogues are in fact continuous, but the initials have been changed to suggest two different dialogues. MM in 6 is MA in 7. DC and SI are different police officers. Most respondents correctly identify these as actual conversations but are less good at spotting that 7 is a continuation of 6. One obvious feature here is the amount of performance dysfluencies which I have allowed into the transcription of 7. This is partly necessitated by the fact that MM is struggling in English and is barely understandable at certain points of the tape. Where the meaning is much clearer, transcribers have a greater tendency to ignore dysfluencies. Dialogue 8 Source: My transcription of an advertisement on Classic FM Radio, March 1999. 'It's Martin Skinner here' clearly indicates a supposed telephone conversation. Respondents are usually fairly well disposed to the authenticity of this conversation until they hear the tape. Pos. and Pr. are common. In fact it is certainly scripted - quite carefully so, perhaps by someone who has studied discourse analysis.

8. Conclusion The suggestion that there is a recognisable category of marginal cases suggests confidence in the solidity of the majority of cases - that they have or have not 'occurred naturally'. This may be misleading. I feel rather that the term 'naturally occurring' is unfortunate since it oversimplifies what is a complex range of dialogue modes. Except in matters such as intonation, pausing and so on, the distinction between natural and artefactual may be less sharp than usually assumed. This all looks like bad news for the forensic linguist, whose job is to try and make a precise distinction based on a written record: any doubt in court could be damaging if not fatal to linguistic evidence. It would be a rather depressing note to conclude on if a natural corollary of Abercrombie's claim (that you can't get down ordinary conversation in writing) is that linguists have no role to play in helping courts in this area. Is the answer to the theoretical question raised in this paper really that there are no solid linguistic criteria by which we may tell a conversation which happened from one that did not?

Naturally-Occurring

and Artefactual

Dialogue

401

I do not think this is justified, but there are two problems which need to be overcome. The first is that the state of our knowledge about the police interview as a discourse type is limited. Through more systematic study of transcripts of police interviews, it ought to be possible at least to reduce, if not eliminate, the area of doubt - to increase confidence in our powers of distinction. The second is that in the eyes of the law, the default presupposition is, and probably has to be, that transcriptions are accurate, or at least accurate to a point where no real doubt exists as to what was said, as opposed to how it was said. The forensic linguist would normally be called by the defence to undermine that view in a particular case (although there is no reason in principle why the prosecution should not have recourse to linguistic opinion to sustain the default presupposition). The police, judges and lawyers are so accustomed to looking at transcripts as true records (something for which they can hardly be blamed) that it probably takes some shift of perspective to read them differently. Odd though it may sound, we still need better ways of 'alienating' courts so as to make them understand transcripts more clearly. Both these endeavours would require co-operation between the forces of the law and linguists, and perhaps some addition to training programmes in the various fields of work involved.

References Abercrombie, D. (1965): Conversation and Spoken Prose. In: Studies in Phonetics and Linguistics. London: Oxford University Press. Burton, D. (1980): Dialogue and Discourse. London: Routledge. Cook, G. (1992): The Discourse of Advertising. London: Routledge. Geis, W. (1982): The Language of Television Advertising. Academic Press. Goodwin, C. (1981): Conversational organization. Interaction between speakers and hearers. London: Academic Press. Firth, J. R. (1935): The techniques of Semantics. In: J. R. Firth 1957. Papers in Linguistics 19341951. London: Oxford University Press, 7-33. Herman, V. (1995): Dramatic Discourse: Dialogue as Interaction in Plays. London: Routledge. Hess-Luttich, E. W. B. (1985): Dramatic Discourse. In: van Dijk, T. (ed.) Discourse and Literature. Amsterdam: Benjamins. Labov, W. (1972): Sociolinguistic Patterns. Philadelphia: University of Pennsylvania Press. Short, M. (1989): Discourse Analysis and the Analysis of Drama. In: Carter, R. and P. Simpson (eds.) Language Discourse and Literature. London: Unwin Hyman, 139-168. Simpson, P. (1997): Language through Literature. An Introduction. London: Routledge. Sinclair, J. McH. and D. Brazil. (1982): Teacher Talk. London: OUP. Sinclair, J. McH. and R. M. Coulthard. (1975): Towards and Analysis of Discourse: the English used by teachers and pupils. London: OUP. Stenstrom, A.-B. (1994): An Introduction to Spoken Interaction. Harlow: Longman. Trickey, D. (1988): Establishing criteria for naturalness in student-student interaction in EFL. ELR Journal, Vol. 2, 107-140. Willis, J. (1987): Inner and Outer: Spoken Discourse in the Language Classroom. In: M. Coulthard (ed.) Discussing Discourse, University of Birmingham, ELR Monograph, 1-20.

Janet Cotterill Multiple Voices: Monologue and Dialogue in the OJ Simpson Criminal Courtroom 1. Introduction In October 1994, African-American sporting icon OJ Simpson stood trial for the brutal double homicide of his ex-wife, Nicole Brown Simpson and her friend, Ron Goldman, both of whom were Caucasian. A cornerstone of the defence case was that the murder scene had been compromised by one of the Los Angeles Police Department's scene-of-crime investigating officers, Detective Mark Fuhrman. The defence allegation was that Fuhrman had been motivated by his racist tendencies to plant evidence at the crime scene which incriminated Simpson. The criminal trial took place in the wake of the Rodney King beating, also involving the LAPD, and race became and remained a key issue in the case. The trial resulted in Simpson's acquittal despite overwhelming evidence tying him to the crimes. One of the witnesses called by the defence to support their claim that Fuhrman was a racist was Kathleen Bell, who maintained she had overheard a conversation ten years previously in which Fuhrman had made racist comments about African-Americans. During Court TV coverage of the trial, she had seen Fuhrman giving evidence and had recognised him. Bell had immediately faxed a letter to Simpson's attorney recounting the incident she claimed to have witnessed. This paper will focus on the evidence of Kathleen Bell in three different forms; first, through an analysis of her initial monologically-expressed letter to Simpson's team of attorneys which provides an initial statement of her stance towards Fuhrman and his comments. This is followed by an analysis of her elicited dialogic testimony as a sequestered witness on the stand, and the paper concludes by studying the reiteration and re-presentation of Bell's position by both sets of lawyers in the monologic closing argument phase of the trial. Before analysing the evidence of Kathleen Bell, however, it is first necessary to consider the criminal courtroom as the setting for monologue and dialogue.

2. Monologue and Dialogue in the Adversarial Courtroom The verbal activities of the participants in the adversarial trial-by-jury courtroom may are clearly delineated into the essentially monologic and the predominantly dialogic. This distinction is illustrated in the diagram below:

404

Janet

Jury Selection

Opening Arguments

Witness (Cross-) Exam

D

M

D

Closing Arguments

M

Jury Instructions & Summing Up M

Jury Deliberation

D

Verdict

D

Cotterill

Sentencing

M

Figure 1: Distribution of Monologic (M) and Dialogic (D) Phases in the Adversarial Trial There are a number of trial phases which are formally monologic in orientation; the opening and closing arguments, judicial jury instructions and the final sentencing. It is interesting to note that in these phases it is the institutionally more powerful participants - the lawyers and the judge respectively - who are sanctioned to make monologic contributions. This is of course a tendency which may be observed in a wide range of similarly power-based discourse settings; the right to produce monologue and thereby hold the floor for relatively extended periods is accorded to other 'powerful' participants in asymmetric contexts, such as doctors, teachers and managers. In the courtroom, the judge and both sets of lawyers have considerable communicative dominance over the non-participatory jury. When face-to-face interaction does occur in court, this is also controlled by the lawyers and the judge. During jury selection (known as voir dire), as well as in the verdict stage of the trial, the judge and lawyers once more have command of turn-taking priority and topic control. Even when the judicial power switches to the jury, during the deliberation stage, the jury discusses in private and amongst 'peers' and once back in the courtroom, the verdict is delivered in response to specific yes/no and option questions from the judge, and with an appointed individual as spokesperson for the group. In addition to these monologic phases, the principle of dialogic question-andanswer sequences, both implicit and explicit, applies equally at various stages of the adversarial criminal trial. One of the main aims of the opening statements by prosecution and defence, for example, is to prime the jury with the questions to which answers will be provided in the course of the trial. In this way, both sides try to create a schematic account of the crime which will seem plausible and credible to a majority of jurors. For the prosecution, this consists of the elicitation of responses which adequately fulfil the 'burden of proof. In contrast, the defence must suggest either that these responses fail in this regard, or else they must pose a series of alternative questions, proposing a conflicting version of events, and hence creating 'reasonable doubt' about the prosecution version in the minds of the jury.

Multiple

Voices: Monologue

and Dialogue

in the OJ Simpson Criminal

Trial

405

The formulation of these questions and their subsequent responses is of crucial importance, since the ultimate recipient is not the lawyer/questioner, but rather the jury, who represent what Goffman refers to as 'encircling hearers': talk ... can be played out as a display for the encircling hearers ... the arrangement employed in TV talk shows, or a lawyers examination of a witness before a jury (Goffman 1981: 137)

Despite the interactional analogy with a TV talk show audience, the jury has a more important and decisive role to play in the overall outcome of the purposive speech event, beyond the straightforward role of spectator. Indeed, they are frequently referred to as the 'factfinders' in a criminal case. Although the jury are not directly involved in the dialogic lawyer-witness exchange, the principal aim of examination is to permit the prosecution and defence to display their respective pieces of evidence for the scrutiny of the jury. In order for this to take place effectively, the information held by the witness must be elicited and transmitted explicitly and unambiguously to the jury. This elicitation involves a high degree of linguistic manipulation and reformulation, so as to ensure that the jury will prefer one version of events rather than another. The rules governing courtroom interaction are such that testimony must be elicited from the witness in the form of question-and-answer adjacency pairs, rather than through protracted monologic contributions from either witness or lawyer. This clearly entails a 'suspension of disbelief on the part of all participants, particularly when attorneys examine witnesses from their own side; both lawyer and witness are aware of the relevant evidence and must 'play out' the dialogic elicitation of this evidence in order to fulfil the requirements of their courtroom roles. The lawyer, who is familiar with the witness's story, must still 'question' the witness and respond pseudo-spontaneously to their responses, all the time hoping that the witness will not falter or stray too far from the script. In addition, the lawyer must hope that the witness will be sufficiently skilled verbally to construct a coherent and rhetorically persuasive account for the jury. Thus, through this type of storytelling-by-proxy, the lawyer must resist the temptation to intervene and take over the role of storyteller until the time comes for the powerful and coercive closing argument. Only at the very end of the evidential stage of proceedings, is the lawyer sanctioned to re-tell the story in monologic form and whilst addressing the jury directly. This is invariably accompanied by a warning from the judge who directs the jury that 'this is opinion, not evidence' so as to explicitly highlight the partisan and persuasive nature of the closing argument. It is interesting to note that despite the lawyer's position of dominance over the witness in many respects, both legal and interactional, during the examination phase, the lawyer remains fundamentally dependent upon the witness to perform effectively and produce a convincing story for the jury.

Janet

406

Cotterill

In the OJ Simpson criminal trial, the monologue-dialogue cycle began with Kathleen Bell's letter to the defence team, and this will now be the focus of a detailed analysis.

3. The Letter: An Initial, Monologic Expression of Stance 1 2 3 4 5 6 7 8

I am writing to you in regards to a story I saw on the news last night. I thought it was ridiculous that the Simpson defense team would even suggest that their (sic) might be a racial motivation involved in the trial against Mr Simpson. I then glanced up at the television and was quite shocked to see that Officer Ferman (sic) was a man that I had the misfortune of meeting, You may have received a message from your answering service that I called to say that Mr Ferman may be more of a racist than you could ever imagine.

9 10 11 12 13 14 15 16 17 18 19 20 21 22

Between 1985 and 1986,1 worked as a real estate agent in Redondo Beach for Century 21 Bob Maher Realty (now out of business). At the time, my office was located above a Marine recruiting center off Pacific Coast Highway. On occasion, I would stop in to say hello to two Marines working there. I saw Mr Ferman there a couple of times. I remember him distinctly because of his height and build. While speaking to the men, I learnt that Mr Ferman was a police officer in Westwood, and I don't know if he was telling the truth, but he said that he had been in a special division of the Marines. I don't know how the subject was raised, but Officer Ferman said that when he sees a "nigger" (as he called it) driving with a white woman, he would pull them over. I asked what if he didn't have a reason, and he said he would find one. I looked at the Marines to see if they knew he was joking, but it became obvious that he was very serious.

23 24 25 26 27 28

Officer Ferman went on to say that he would like nothing more than to see all "niggers" gathered together and killed. He said something about burning them or bombing them. I was too shaken to remember the exact words he used, however I do remember that what he said was probably the most horrible thing I had ever heard anyone say. What frightened me even more was that he was a police officer.

29 30 31 32 33 34

I am almost certain that I called the L A P D to complain about Officer Ferman, yet I did not know his last name at the time. I would think that the LAPD has some record of this. N o w I know that M r Ferman was the investigating officer, I must suggest that you check into his background further. I am certainly not a fan of Mr Simpson, but I would hate to see anyone harmed by Officer Ferman's extreme hatred.

Figure 2: Letter from Kathleen Bell to Simpson's Defence Attorney, July l S / l ^ 1995

Multiple Voices: Monologue and Dialogue in the OJ Simpson Criminal Trial

407

In her letter, Bell sets out her evaluative and judgemental stance towards a number of issues surrounding the trial of OJ Simpson. Early in the first paragraph, she dismisses the defence strategy of creating a case based upon race as absurd and inappropriate (lines 2-4): I thought it was ridiculous that the Simpson defence would even suggest that their (sic) might be a racial motivation involved in the trial against Mr Simpson.

In addition to this attempt to distance herself from the defence team and their strategy, presumably lest her point of view be rejected as partisan, Bell is equally careful to deny any sympathies with Simpson, perhaps in an attempt to pre-empt a possible prosecution charge that her allegations of racism were simply the work of a (female) fan bent on obtaining Simpson's acquittal. Bell explicitly denies any allegiance to the defendant, declaring emphatically: I am certainly not a fan of Mr Simpson, but I would hate to see anyone harmed by Officer Ferman's extreme hatred, (lines 33-4) (my italics)

This denial of commitment to the Simpson's cause was later to prove of some significance to the defence in their closing argument (see page 10 below). In her evaluation of Detective Ferman's comments about African-Americans, this distancing strategy is again in evidence. It is clearly important that she is not seen to empathise in any way with Fuhrman's racist attitudes. Bell rejects Fuhrman's use of what became known in the trial as the 'N' word, representing the term "nigger", achieving denial through a sophisticated multi-level approach. The first involves the use of an marker attributing the use of the term firmly to Furhman (lines 18/9): Officer Ferman said that when he sees a "nigger" (as he called it) driving with a white woman, he would pull them over, (my italics)

The addition of the italicised phrase gives the impression of the "extravocalisation" of the 'N' word, to use White's (1998) term. In this way Bell is evoking the heteroglossic nature of the situation, effectively saying to her audience "this is not a word that I would use, but he did. I am simply reporting his use of the word to you". This is coupled with the inclusion of parentheses, reminiscent of a journalistic-style editorial comment on an existing, but crucially, separately-authored, text. The final stage of the process adds a further layer of dissociation through her use of quotation marks. In explicitly signalling the term as emanating from Fuhrman (note the use of double rather than single quotation marks), Bell distances herself still further from the expressed sentiment.

Janet

408

Cotterill

The final significant area of evaluation in the letter concerns Bell's affective reactions to Fuhrman's behaviour and comments. She expresses these reactions explicitly by means of a range of emotive lexical choices. Bell thus describes herself as 'shocked' (line 5), 'shaken' (25), and 'frightened' (27) by Fuhrman's remarks. These lexical choices are intensified through the use of'quite', 'too' and 'even more'' respectively, adding further affective force to her reactions. Bell also uses a number of other references which evaluate the social appropriacy of Fuhrman's comments. His words are judged to be tantamount to racism, with Bell referring to him as a 'racist' (line 8), and describing his behaviour as demonstrating 'extreme hatred' (34). Finally, Fuhrman's comments are evaluated as 'probably the most horrible thing I have ever heard anyone say' (26/7). From the defence standpoint, the letter written by Kathleen Bell was a gift beyond their wildest dreams. They had an independent witness offering evidence which appeared to corroborate their claim that Fuhrman was a racist police officer. One US academic is quoted as stating that "in many ways, I don't know what the defense would have done without Fuhrman. He gave their [the defence] story a villain" (Smith 1995: 47). The defence simply needed to call Bell as a witness and elicit similar testimony from her on the stand, and they could inflict potentially serious damage upon the prosecution's case. Bell's testimony (although she claimed she had never expected to be called to testify and was extremely reluctant to appear), provided the defence with a guilt-edged opportunity to reinforce their attack on Detective Mark Fuhrman. For the prosecution, in contrast, Bell represented a considerable problem, since they had to try to discredit the seemingly credible witness and deconstruct her claims of Fuhrman's racism. Both sets of lawyers had their opportunity to examine Bell when she appeared in court late in the case. An analysis of her testimony will now be discussed.

4. The Testimony: From Monologue to Dialogue1 Much of Bell's direct examination testimony was relatively unproblematic, consisting of fairly straightforward reiteration questions aiming to encourage Bell to restate the information in the letter, producing a reasonably full and appropriate response by Bell, as in the following example: Q: Bell: Q: Bell: 1

All right, now tell me the circumstances of your meeting at Hennessey's tavern. Who were you with? I was with Andrea. What was the purpose or occasion for you being with her? Anything special? No, we just stopped ...

All data here are from Kathleen Bell's testimony, The People vs OJ Simpson, 5th Sept., 1995.

Multiple Voices: Monologue and Dialogue in the OJ Simpson Criminal Trial

409

However, there were also a number of occasions where potentially damaging holes in Bell's letter needed to be repaired. The major area of concern was Bell's (for the defence, irritating) habit of hedging some of her claims. This was clearly a problem for the defence, since any acknowledgement of heteroglossic potential in Bell's letter could have provoked a challenge from the prosecution. By indicating the possibility of another possible scenario or an alternative interpretation of events, Bell was clearly laying herself, and the defence's case, open to attack on cross-examination, and these discrepancies needed to be addressed by the defence direct examiner.

5. The Defence Case: Fixing the Hedge One instance of hedge-fixing occurs when Bell is being questioned about whether or not she called the Los Angeles police (LAPD) to complain about Officer Fuhrman. Defence attorney Bailey challenges her potentially damaging heteroglossic statement that she is "almost certain that [she] called the LAPD to complain about Officer Mark Ferman" (line 29) and that she "would think that the LAPD has some record of this" (lines 30/1). The italicised sections indicate points of weak commitment, where Bell has left a point of some doubt which could lead the jury to question her powers of recall and the prosecution to question her credibility. Bailey therefore attempts to repair the damage by asking her to search her memory once more: Q:

You said in your letter that you are almost sure you called the LAPD. What is your present recollection as to whether you did?

Bell responds effectively at this point by changing the previous hedged responses to an unequivocal: Bell:

I am positive I did

In addition to Bailey's efforts at repairs of this type, there are also a number of occasions on which Bell attempts to self-repair her commitment level, perhaps recognising an inconsistency, as in the following example: Q: Bell:

A moment ago you wanted to explain something to us that you had just recalled. Would you tell us what that was? I wrote this is a very big hurry and I didn't think there would be such a need to recall exactly what had happened in detail ... he did not say the 'N' word when he said 'If I saw a black man with a white woman driving in a car, he would pull them over'. He definitely said 'a black man'

410

Janet

Cotterill

Not all of these attempts to repair are successful, however. Here, Bailey builds up the intensity on her behalf, only to see Bell reduce it once more in her response, through the addition of even more hedging than in the original (and already problematic enough) segment of her letter: Q:

Did you respond in any way, other than becoming upset, when Mr Fuhrman told you what he would like to do with reference to burning an entire race ? I just, I looked at the Marines and I kind of, they weren't saying anything. They just shrugged their shoulders, and so I kind of got teary-eyed and left.

Bell: and

And what did you say? I was taken aback a little bit and so I kind of paused and looked at the Marines and I just said "Well what if they didn't do anything wrong?"

Q: Bell:

Bell effectively denies Bailey's attempts to prime her affective response with his highly engaged assertion that Fuhrman wished to "burn an entire race", but rather than support Bailey's attempts, Bell minimises and hedges her response, with a series of 'just's and 'kind of s. These areas of concern for the defence translate into areas of opportunity for the prosecution, who are quick to exploit the heteroglossic potential acknowledged in Bell's letter.

6. The Case for the Prosecution: Challenging Evaluative Force One of prosecutor Darden's key strategies in cross-examination was to challenge Bell on the intensity of her negative reaction to Fuhrman, given that she was alleged to have subsequently introduced him to her best friend as a potential boyfriend: 1 2 3 4 5 6 7 8 9 10 11 12 13

Q:

I take it that you understand it that this word, this epithet, is the most vile word in the English language Yes You would agree with that, wouldn't you Yes And when you heard Detective Fuhrman use this word, you were offended ? Yes You were hurt by it? Yes And you were frightened of him because of his use of this word Yes You were horrified I think you testified earlier

14

Q:

When you heard Detective Fuhrman use these words you cried,

Q: Bell: Q: Bell: Q: Bell: Q: Bell:

Q:

Bell:

Multiple Voices: Monologue and Dialogue in the OJ Simpson Criminal Trial

15 16 17 18 19 20 21 22

Bell:

Q:

Bell:

Q: Bell:

or did you cry? Somewhat Somewhat? Yes, I was teary-eyed. He could see that I was upset Despite the horror and trauma Detective Fuhrman's use of this word caused, isn't it true that you still introduced Andrea Terry to Mark Fuhrman I did not

The cumulative force of Darden's questions to Bell results in a powerful challenge to her evaluative stance with respect to Fuhrman's comments, despite Bell's final ideationally-based denial. This culminates in Darden's 'despite...' question (line 19), increasing his counter-expectational argument further with the addition of 'still' (20). Bells counters with an emphatic 'I did not' (22). There is also an interesting insert sequence in the above example in lines 16 and 17, where Darden breaks off from his accumulation of expectational adjectives to exploit a weak point in Bell's response. When she tells him that she cried 'somewhat', surely a sign of embarrassment than a genuine attempt to minimise the force of her utterance, Darden immediately seizes upon the incongruence of this minimisation in the context of the previous list of powerfully emotive adjectives. Bell does herself no favours by reformulating her emotional state with the diluted term 'teary-eyed' (line 18). Overall, however, the prosecution challenge was both half-hearted and ineffectual, both in terms of the quantity and quality of cross-examination. Realising the force of Bell's letter and its claims, and acknowledging the cumulative weight of the other taped evidence against Fuhrman, the prosecution were resigned to losing the battle over Bell. As lead prosecutor Clark later acknowledged: Now that the defense had Fuhrman on the ropes, they went for the final knockout, demanding his reappearance in court. But not before calling Kathleen Bell and a string of other witnesses to attest that Fuhrman had spouted racist remarks in their presence. (Clark 1997: 451)

Bell's testimony certainly seems to have had a profound impact on the jury. The LA Times (1995: 47) reported that "jurors looked shaken - frowning, recoiling or staring intently" during her evidence. It is clear that the issue of Fuhrman's racist comments became, for the prosecution, an exercise in damage limitation. The prosecution were faced with a credible, independent witness providing damning testimony, and this was subsequently consolidated by corroborating evidence, including a tape of Fuhrman uttering racist epithets on an entirely separate occasion. Indeed, when Fuhrman was called back to the stand by the prosecution, such was the 'mountain of

412

Janet

Cotterill

evidence' attesting to his racist tendencies, that he refused to testify further and took the Fifth Amendment. As the case drew to a close, after nine months of evidence, the prosecution were no longer contesting Fuhrman's racist tendencies. By the time the closing arguments, the monologic climax of the case, were delivered by both sides, the gap between defence and prosecution had widened considerably.

7. Closing the Circle: From Dialogue to Monologue The closing argument is a highly significant speech event, not only in terms of its rhetorical potential, but also because it affords attorneys the opportunity to reconceptualise and reframe the evidence elicited from witnesses in the course of the trial whilst addressing the jury in direct monologue. Although it is formally monologic, the true orientation of the closing may be more accurately described as in dialogic terms. In order to understand the dialogic nature of the closing argument, it is useful to consider it in the context of Hoey's (1994) description of monologue as: a dialogue in which the listener's questions or comments have not been explicitly included, but which retains indications of the assumed replies of the listener (Hoey 1994: 29)

The closing argument does indeed consist of responses to potential projected questions as Hoey suggests. However, it differs from other, more co-operativelyoriented, discourse settings such as casual conversation, in that it is fundamentally based upon a directive rather than a predominantly social or referential function. The effect of this coercive dimension is that the 'questions' which Hoey refers to are not questions to which the speaker feels the listener would like to receive answers. Rather they consist of those questions which the prosecution and defence would like to place in the minds of the jury, and so are much more prescriptive in orientation. An interesting feature of both the prosecution and defence closing arguments is the explicit presentation of these questions: It is your duty and it would be your challenge to stay focused on the question you were brought here to answer, and the only question that you were brought here to answer, did the defendant commit these murders? ... but those are not, they are not important issues (Prosecution closing argument)2

2

All of the data in this section are taken from the criminal trial closing arguments by prosecution and defence, as appropriate.

Multiple Voices: Monologue and Dialogue in the OJ Simpson Criminal Trial

413

The prosecution went to great lengths to claim that the question of Mark Fuhrman's racism expressed in Bell's letter was actually nothing more than a peripheral question, distracting the jury from their central task. Chief prosecutor Clark maintained that the real question centred on the guilt or innocence of OJ Simpson. From the very outset of her summation, Clark was unequivocal about the LAPD detective: Let me come back to Mark Fuhrman for a minute. Just so it is clear. Did he lie when he testified here in this courtroom saying that he did not use racial epithets in the last ten years? Yes. Is he a racist? Yes. ... Do we wish there were no such person on the planet? Yes. But the fact that Mark Fuhrman is a racist and lied about it on the witness stand does not mean that we haven't proven the defendant guilty beyond a reasonable doubt. And it would be a tragedy if with such overwhelming evidence, ladies and gentlemen, as we have presented to you, you found the defendant not guilty in spite of all that, because of the racist attitudes of one police officer (my italics)

Through a series of rhetorical yes/no questions, designed to respond to some of these projected (or in this discourse context, implanted) questions, Clark attempts to guide the jury away from Fuhrman, returning the spotlight to Simpson and the issue of the burden of proof. In stark contrast, a triumphalist defence team were able to devote less attention to the provision of answers, since in criminal trials, the burden of proof (and hence the requirement to answer the questions satisfactorily) rests firmly with the prosecution. In this very real sense, the legal process may be seen to favour the defence. Instead the defence concentrated on posing a range of questions which they implored the jury to ask of the prosecution case. The specific issues presented by the defence were in the form of fifteen specific questions. Amongst these questions, which included some dealing with the forensic DNA evidence, and others which concerned the debate over Simpson's means, motive and opportunity, there was a particular (perhaps disproportionate) focus on Fuhrman and Bell. Together they accounted for no fewer than three of the total fifteen questions, including the following: If Mark Fuhrman, who speaks so openly about his intense genocidal racism to a relative stranger such as Kathleen Bell, how many of his co-workers, the other detectives in this case, were also aware that he lied when he denied using the "N" word yet failed to come forward?

For lead defence attorney Cochran, the key strategy in his section of the closing argument was clearly the propagation of the seeds of doubt sown in the minds of the jury by Bell's letter questioning Fuhrman's racial integrity.

Janet Cotterill

414

8. The Final Word A total of 126 witnesses gave evidence in the Simpson criminal trial, including Kathleen Bell. However, at the end of the trial, the sole task assigned to the jury was the choice between the two versions of events presented to them. It was a relatively stark choice, as UCLA law professor Peter Arenella points out: Adversarial trials only examine two versions of the truth - OJ is guilty and the police did nothing wrong, or incompetent and corrupt police framed an innocent man. Left unexamined is a third possibility - OJ is guilty, but serious misconduct occurred (Arenella 1995: 49)

Even the form of the final verdict announcement was strictly controlled by the court in the Simpson case. In British courtrooms, the foreperson at least has the opportunity to deliver the verdict orally, if only in response to a controlling yes/no question, such as 'do you find the defendant, guilty or not guilty?'. According to US judicial conventions, however, the verdict in this case was read by a court official from a pre-printed charge sheet, with the jury foreperson simply required to fill in the appropriate space with the words 'guilty' or 'not guilty', as applicable. An extract of the Simpson verdict follows: In the matter of people of the State of California versus Orenthal James Simpson, case number BA097211. We, the jury, in the above-entitled action, find the defendant, Orenthal James Simpson, not guilty of the crime of murder in violation of penal code section 187(a), a felony, upon Nicole Brown Simpson, a human being, as charged in count 1 of the information ...

After nine months of silent contemplation, and following an improbably short deliberation period of only two hours, the only incidence of direct dialogic interaction between the court (in this instance embodied by the judge) and the jury came in the aftermath of the verdict. The jury were asked if their verdict had been unanimous and then each individual juror was asked to confirm their individual verdict: The Court: The Jury: The Clerk: Juror No. 1: The Clerk: Juror No. 2:

Ladies and gentlemen of the jury, is this your verdict, so say you one, so say you all? Yes. Juror No. 1, as to count 1, is this your verdict? Yes. Juror No. 2, as to count 1, is this your verdict? Yes...

The jury's sole verbal contribution was thus a fleeting and inconspicuous formality amidst the furore in the courtroom following the not guilty verdict. It

Multiple Voices: Monologue and Dialogue in the OJ Simpson Criminal Trial

415

was not even covered by the TV cameras, who focussed instead on the acquitted Simpson and his 'dream team' and their jubilant celebrations.

References Arenella, P. (1995): The Fuhrman Files in In Pursuit of Justice - The People vs Orenthal James Simpson. Los Angeles: LA Times, 47-53. Clark, M. (1997): Without A Doubt. New York: Viking. Goffman, E. (1981): Forms ofTalk. Philadephia: Philadephia University Press. Hoey, M. (1994): Signalling In Discourse: a functional analysis of a common discourse pattern in written and spoken English. In: M. Coulthard (ed.) Advances In Written Text Analysis, London: Routledge, 26-46 (published originally as Signalling in Discourse, 1979, ELR Monograph, University of Birmingham). Los Angeles Times (1995): In Pursuit of Justice - The People vs Orenthal James Simpson Los Angeles: LA Times. Smith, K. (1995): The Fuhrman Files. In: In Pursuit of Justice - The People vs Orenthal James Simpson. Los Angeles: LA Times, 47-53. White, P. (1998): Telling media tales: the news story as rhetoric, unpublished PhD thesis, University of Sydney, Australia.

Malcolm Coulthard

Suppressed Dialogue in a Confession Statement Preamble One November evening in 1952 two teenagers, Derek Bentley aged 19 and Chris Craig aged 16, set out with the intention of stealing money from a butcher's shop. However, the butcher was working late, so they had to abandon the idea. Instead they decided to break into a nearby confectionery warehouse. The boys 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. The trial, which lasted only two days, took place five weeks later. 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 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.

1. Introduction In the original trial the problem for the Prosecution, in making the case against Bentley, was to demonstrate that he could indeed be guilty of murder despite being under arrest when the murder was committed. At this point it would be useful to read the statement which it was claimed Bentley dictated shortly after his arrest. It is presented in full on page 418; the only change I have introduced is to number the sentences for ease of reference. Bentley's barrister spelled out for the jury the two necessary pre-conditions for them to convict him: they must be "satisfied and sure": i) ii)

that [Bentley] knew Craig had a gun and that he instigated or incited Craig to use it. (Trow 1992: 179)

418

Malcolm

Coulthard

(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 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: 'I 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)1 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: 'I 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.

Figure 1: Derek Bentley's Statement

Suppressed Dialogue in a Confession

Statement

419

The evidence adduced by the Prosecution to satisfy the jury on both points was essentially linguistic: For point i) 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 who, because of the importance of the case was the Lord Chief Justice, Lord Goddard, had made great play with this sentence arguing both that its positioning in the narrative of events, before the time when was a single policeman on the roof and also the choice of "the gun" (as opposed to "a gun") must imply that Bentley knew that Craig has a gun well before it was used - in other words "the gun" in its 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". The evidence to support point ii), 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. Thus, 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 sworn to you positively that those words were said (quoted in Trow 1992: 109).

So, as the judge emphasised, 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.

2. The Linguistic Evidence At the time of Bentley's arrest the police were allowed to collect verbal evidence from those accused of a crime in two ways: either by interview, when they were supposed to record verbatim, contemporaneously and in longhand, both their own questions and the replies elicited, or by statement, when the accused was invited to

Malcolm

420

Coulthard

write down, or, if they preferred, to dictate to a police officer, their version of events. During statement-taking the police officers were supposed not to ask substantive questions. 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 - police questions and his replies 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. (Trial transcript pp. 100—1)

There is no doubt that this was one procedure 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" (Trial transcript p. 156)

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 for most of the text the narrative of events is fairly coherent. However, the narrative seems to end with utterance (38) "The policeman then pushed me down the stairs and I did not see any more". What follows in sentences 39-42: (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)1 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.

is 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. 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

Suppressed Dialogue in a Confession

421

Statement

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 sequence of sentences ( 3 9 ^ 0 ) Bentley replied "That was an answer, sir, to a question", to which the barrister replied "I daresay it was in reply to a question" but 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 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 typical examples of 'normal' usage of negative reports in the sentences below which are taken from a crucial confession statement in another famous case, that of the Bridgewater Four: a) b) c)

Micky dumped the property but I didn't know where. Micky Hickey drove the van away, I don't know where he went to We didn't all go together, me and Vinny walked down first. (Molloy's Statement)

In the first two examples, a) and b) the second negative clause functions as a denial of an inference which the listener could have reasonably derived from the first clause. Example c) is similar but this time it is a denial of an inference which the narrator guesses the listener might have made, as there is no textual basis for the inference1. In other words such negatives are an integral part of the ongoing narrative. We find examples of negatives being used in a similar way in Bentley's statement:

1 See Pagano, 1994 for a detailed discussion of the function of negative clauses in texts.

422

Malcolm (6) (7)

Coulthard

A little later Norman Parsley and Frank Fasey called. 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. Similarly: (38) The policeman then pushed me down the stairs and I did not see any more

Here the negative explains the end of the narrative of events, in other words not seeing anything further has a clear narrative relevance. 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 corner 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 a) it must be the policeman that Craig was firing at, nor b) that it must be Craig who was doing the firing, nor c) 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 a negative statement. The fact that, like (27) 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 too would only make narrative sense if it were linked backwards or forwards to 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:

Suppressed Dialogue in a Confession

423

Statement

(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, its logic and the sequencing of the information were not under his direct control. The placing of the reporting of some of the events then depends 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 was able to assume this was Bentley's unaided narrative then the positioning of "I didn't know he was going to use the gun" would be crucial 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 about 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 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 the 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". By a remarkable co-incidence a parallel event occurred during the trial itself. Bentley's barrister, whilst eliciting a narrative of the evening's events from him produced the following sequence of questions: Q A Q A Q A

Well, after some difficulty did you then get on the roof and find Craig Yes sir, I went on Craig's drainpipe and got up Up to that time did you know that Craig had a loaded revolver No Sir When you got on to the roof what happened then Some lights in the garden; someone shone a light in the garden (Trial Transcript p. 97)

If this set of utterances were turned into a monologue narrative the 'knowledge' about the loaded gun would similarly be reported well in advance of its use with a similarly misleading effect.

424

Malcolm

Coulthard

3. Judgement Demonstrating 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 utterance "Let him have it Chris". In August 1998, 46 years after the event, the Lord Chief Justice, sitting with two senior colleagues, criticised his predecessor's summing-up and allowed the Appeal against conviction.

References Coulthard, R. M. (1993): On beginning the study of forensic texts: corpus, concordance, collocation. In: M. Hoey (ed.) Data, Description, Discourse, London: HarperCollins, 86-114. Hannam - extract from the examination of Chief Inspector Hannam in the trial of Alfred Charles Whiteway in October 1953. Pagano, A. (1994): Negatives in written text, In: R. M. Coulthard (ed.) Advances in Written Text Analysis, London: Routledge, 250-75. Trow, M. J. (1992): "Let him have it Chris", London: HarperCollins.

Ron Scollon Hidden Dialogicality: When Infelicity Becomes Fear of Infringement 1 1. From Texts to Conversation and Dialogue Jesse 'The Body' Ventura™, Governor of Minnesota newly elected late in 1998, registered his professional wrestling stage name in 1992 according to Dennis Cass (1999). As Cass points out this makes him, 'the only candidate in American history who can defend himself against unlicensed merchandise' (p. 70). The same issue of Harper's in which the Cass article appears indicates in the popular Harper's Index that the sequence 01-01-00™, the first day of the new millenium by some counts has also been registered as a trademark which prohibits unlicensed use of that numerical sequence in product names. Somewhat closer to the home of this paper is the word 'semiotics' which Coombe (1998) tells us has been registered by a consultation and training company for protected use by that company. The laws governing the use of registered trademarks and other protected 'commodity/signs' (Coombe 1998) vary from jurisdiction to jurisdiction. Perhaps it is unlikely that a journalist would use 'Jesse "The Body" Ventura, '01-01-00', or 'semiotics' in a way that would directly infringe the rights of the owners of these properties. Nevertheless, such intellectual property rights, as Coome says, legitimize 'new sources of cultural authority by giving the owners of intellectual property priority in struggles to fix social meaning' (p. 26). For about three decades now linguists and philosophers of language have been working out the various conditions of felicity which might be said to stand as the background to our ability to interpret the utterances of others. Austin (1962) and Searle (1969) taught us about speech acts. Grice (1975) taught us about the conditions which, all else being equal, would provide a felicitous interpretation of a speech act. Brown and Levinson (1978, revised as 1987) taught us how we manage to interpret our standing vis-a-vis the others with whom we are engaged in conversational and other exchanges. And Gumperz (1977, 1992) taught us how all of this is cued simultaneously as we speak through contextualization cues and processes of conversational inference.

1

In developing the ideas presented here I have enjoyed discussions with Suzanne Wong Scollon and Vicki Kit Yee Yung. Neither of them should be thought responsible for the infelicities which remain in my thinking about hidden dialogicality.

426

Ron Scollon

During this period of study face-to-face social interaction, primarily casual conversation among friends, has dominated our linguistic thinking as the representative anecdote (Burke 1945) of linguistic communication. This was in itself a departure, of course, from the word- and text-dominated analyses of the preceding linguistic past. Until the introduction of the tape recorder, linguistic analysis, in America at least, was centrally focused on words, sentences, and texts which were elicited through careful questioning from what we hoped were representative exemplars of the language or language group under study and written down in phonetic transcription for careful analysis later in the quiet of our studies. Of course, those 'Americanist' practices were predicated on even earlier Indo-Europeanist scholarship which worked through the analysis of historical developments as reflected in texts. Recently, however, going against the current of tape-recorded conversational data there has been developing a critical discourse which focuses centrally on texts of public discourse from newspaper and television news broadcasts to advertising (Hall 1980; Fairclough 1992, 1995; Caldas-Coulthard 1993, 1994, 1997). In another place (Scollon 1998) I have tried to develop the argument that these texts of news discourse are equally amenable to social interactional analysis along with face-to-face conversational data. I have tried to argue that in verbs of saying and reporting, in bylining and other attribution practices, in journalist identifications, and in layout and design whether of the newspaper or the television and radio broadcast are the same positionings of participants which we have come to see in the conversational data studied elsewhere. Pan (to appear) has taken up this theme and argued that the interpersonal face positionings analyzed by Brown and Levinson can also be read in newspaper texts in the relative positionings of journalists and the newsmakers about whom they write. Working within the project of the development of a theory of the novel, Bakhtin many years ago began an analysis which treated texts such as novels as dialogic which now in our time has, belatedly, become a refracting prism through which we view not just novels nor just texts, but all utterances including somewhat paradoxically the actual dialogues of conversation. If there is a representative anecdote of our time, it is the dialogue; and if there is a theme which penetrates our discourses about discourse it is dialogicality. In this essay I will concern myself with a particular form of dialogicality, what Bakhtin called 'hidden dialogicality'. He characterizes hidden dialogicality this way: Imagine a dialogue of two persons in which the statements of the second speaker are omitted, but in such a way that the general sense is not at all violated. The second speaker is present invisibly, his words are not there, but deep traces left by these words have a determining influence on all the present and visible words of the first speaker. We sense that this is a conversation, although only one person is speaking, and it is a conversation of the most intense

Hidden Dialogicality:

When Infelicity Becomes Fear of Infringement

427

kind, for each present, uttered word responds and reacts with its every fiber to the invisible speaker, points to something outside itself, beyond its own limits, to the unspoken words of another person (Bakhtin 1984: 197).

I will argue that with the unprecedented rise in the presence of intellectual property in contemporary life we are seeing the development of a new form of hidden dialogicality - a dialogue between the owners of intellectual properties and any who would carry on a discourse in public which might want to or need to make use of those commodity/signs. The journalist writes, 'Jesse 'The Body' Ventura™, Governor of Minnesota'; the lawyer for Ventura says silently, 'Use the term positively or expect litigation.' I will argue that in this hidden dialogue between the owners of intellectual properties and the producers of public discourse we can see an unprecedented intervention of the power of the commodity mediated through the power of the nation state's copyright and patent laws into the common discourses of daily life. That is to say, not only are we now seeing the nearly ubiquitous presence of the commodity/sign as the discursive coinage of public discourse, through the hidden dialogicality of intellectual property law we hear the voice of the commodity even when its name is not being spoken.

2. Reading Texts An American country and western music song has captured the nature of hidden dialogicality with customary wit and style: If the Phone Don't Ring, Baby, You'll Know It's Me? It could be argued that with the rise of hermeneutic schools of interpretation in the past several decades from deconstructionist literary criticism to semiotics to critical discourse analysis the reading of voices in texts has moved from the deciphering of nuanced allusions of the literary salon to the central tool of a critical project of decentering our most pervasive hegemonic powers. In most cases these voices are loud, clear, and distinct and can be seen working out in the open on the stages of public discourse. But as Widdowson (1998) and others have complained, it becomes another matter when the intertextualities and interdiscursivities we read in texts are subtle, indirect, or ambiguous, as they so often are. Here the case I want to take up is the case of the country song: When there's no voice there at all, how do we know what's missing? How do we come to identify the texts, voices, and power at work behind the texts we analyze? So as not to raise false hopes that we will finally come to some foolproof mechanical procedures, perhaps it is best to begin with a rather simple example of

2

Quoted by Nury Vittachi, Sunday Morning Post, Agenda, p. 2, February 7, 1999.

428

Ron Scollon

the sort hidden dialogicality that concerns us here. For this I will use the name, logo, and trademark design of a Hong Kong fast food restaurant. Everybody's Happy (Daigalok) In the summer of 1997 we took this following photo of a fast food restaurant in Guiyang, Guizhou, China:

There are two things to notice in this sign. The first of them is the name, yijiale ('the family's happy' or perhaps, 'happy family') or 'Family Cafe' as they have given it in English. Along with the name, there is a color scheme of red characters on a yellow background within a red ellipsis. This is the overall logo and design of the restaurant which is repeated on the smaller sign to the left which stands a few feet away across the sidewalk/pedestrian area. The second thing to notice is the character for 'happy', le to which I will return shortly. To provide some useful dialogicality, we need to look at another sign for a different fast food restaurant a few miles away in the same city, Guiyang. This one is called Jiajiale ('everybody's happy') and is given no English name.

Hidden Dialogicality:

When Infelicity Becomes Fear of Infringement

429

Here we have some dialogicality with 'Family Cafe' in the elliptical logo as well as a red and yellow color scheme, though in this case it is the background which is red and the characters which are yellow. Further, there is some play in the name with jiajiale rather than yijiale, literally 'family family happy' rather than 'one family happy'. With just these two restaurants in consideration it seems clear enough that one or the other is doing a 'take off on the other. But there is further dialogicality here to look at. This restaurant in the photo on the next page is in Hong Kong:

430

Ron Scollon

Here we have the same name, jiajiale though in its Cantonese pronunciation of 'KA KA LOK', we have a red and yellow color scheme, and we have an ellipsis. These elements seem clearly enough to be dialogic with the source of all this which is the very popular and widely franchised fast food originator of this dialogicality, 'Café de Coral' or diagalok as shown on the next page:

Hidden Dialogicality:

When Infelicity Becomes Fear of Infringement

C A F E DE

431

CORAL

////#« Perhaps most significant here is the encircled 'R' in the bottom right corner registering this name, logo, and design for this Hong Kong company. Of the three take-offs on the original, it is the Hong Kong 'KA KA LOK FAST FOOD' which is most divergent, perhaps, because it is in Hong Kong where the supporting but hidden dialogic voice of patent and copyright law is strongest. It is here in the Hong Kong example where we hear this voice speaking most strongly saying, 'You must not copy design elements or trade names so that the user might be deceived into believing that you are trading as the same company.' This is the hidden dialogicality which is introduced by intellectual property law. To be sure, as we have seen in this case, we are not to expect an absolute divergence between explicit and marked dialogic partners, on the one hand, and covert, hidden, invisible dialogic partners on the other. The dialogic partner of legal protection speaks more loudly and clearly in Hong Kong than in China and this voice has the consequence of squelching the partly imitative voices of allusion and semiotic similarity. In 'KA KA LOK FAST FOOD' we hear 'Café de Coral' saying, 'Do not appear to be using our name, logo, and design.' In the Jiajiale and Yijiale restaurants of Guiyang we hear this voice in a more attenuated way as if it were saying, 'You may be a long way from Hong Kong and there is no law to protect me, but you shouldn't be using my patented designs.'

Ron Scollon

432

3. Other Hidden Dialogicalities Those who can read Chinese will have noticed that the character used for 'happiness', le, or lok, is not the same in all cases. In Hong Kong it is invariably given in the traditional form: m. In the restaurants in Guiyang it is given in the simplified form: & In the first photograph the sign to the left had four sides and a photo taken on the opposite side shows this:

f.'

SO'CAR

That is, the character for le is given here in the traditional form, not the simplified form. If we look back at the original photo, it could be seen that the character has been painted over changing the traditional form to the simplified form. In this photo we could see - at least we could see on the original when we took the photo - that the character had had something pasted over it which had fallen off or been taken off. One further photo from the shop window shows that the first character in the word for 'market' guangchang in 'evening beer market' has been altered from the traditional form of writing to the simplified form. That is, guangchang

Hidden Dialogicality:

When Infelicity Becomes Fear of Infringement

433

rm had formerly been written as:

That is, the original sign has had the internal portion of the character altered to bring it in line with the simplified way of writing it.

This altering of characters represents a rather different but equally interesting form of hidden dialogicality in this sign in Guiyang. In China the reform of the writing system, which was formally promulgated in several waves, the most recent in 19863, introduced a number of simplified forms for writing traditional Chinese characters. To give this literacy reform the bite of law, the reforms were promulgated along with a system of fines for using the traditional forms. Thus the 3

There have been several waves of reform of the writing system in 1955, 1956, 1964 and most recently in 1986. These have been discussed by Norman (1988), DeFrancis (1984), Chen (1993), Kraus (1995), and Peterson (1995). An official Chinese source is the Yuyan Wenzi Guifan Shouce (Zengdingben) (1993).

Ron Scollon

434

use of the traditional le here in Guiyang, while not only infringing the patent rights of the Hong Kong company, 'Cafe de Coral', is also speaking against the voice of Chinese literacy reform policy.

4. Dialogicality and Unfinalizability I do not believe Bakhtin discussed hidden dialogicality except in the 'Problems of Dostoevsky's Poetics' (1984) where the treatment was relatively straightforward. There he was engaged in the preparation of a relatively exhaustive analysis of the many ways in which multiple voices can be recognized in the novel. He was clearly trying to get at the idea of the effect we have in a discourse when we are aware that we are hearing only one side of the conversation and yet can grasp what must be being said on the other side. Although Bakhtin did not mention it, one would think a good example would be the discourse one hears when a friend is talking on the telephone in our presence. For the idea to be useful, I would think we need to go beyond the relatively simplistic version given in 'Dostoevsky's Poetics' to enrich the concept with an understanding of Bakhtin's idea of dialogicality. This is, of course, not an easy matter as Bakhtin characteristically does not let the idea stabilize in clear definitions. Holquist (1981), in giving us a glossary of Bakhtin's terms in The Dialogic Imagination, has this to say about dialogization, A word, discourse, language or culture undergoes 'dialogization' when it becomes relativized, de-privileged, aware of competing definitions for the same things. Undialogized language is authoritative or absolute (p. 427)

In this interpretation, the crucial aspect of dialogization is relativization. Perhaps Bakhtin's own statement will clarify that he means to focus on what he elsewhere calls the unfinalizability of all forms of life. This unfinalizability resides in the fact that as long as life continues, it continues to have the power to redefine itself, to go against all external descriptions and definitions. A man never coincides with himself. One cannot apply to him the formula of identity A = A. In Dostoevsky's artistic thinking, the genuine life of the personality takes place at the point of non-coincidence between a man and himself, at his point of departure beyond the limits of all that he is as a material being, a being that can be spied on, defined, predicted apart from its own will, 'at second hand.' The genuine life of the personality is made available only through a dialogic penetration of that personality, during which it freely and reciprocally reveals itself (1984: 59).

He continues with the pungent statement: The truth about a man in the mouths of others, not directed to him dialogically and therefore a secondhand truth, becomes a lie degrading and deadening him (1984: 59).

Hidden Dialogicälity:

When Infelicity Becomes Fear of Infringement

435

At the same time, however, we have to see that Bakhtin rejects the view that dialogism is relativism. In Dostoevsky's Poetics (1984) he writes that dialogism has nothing in common with relativism (or with dogmatism). But it should be noted that both relativism and dogmatism equally exclude all argumentation, all authentic dialogue, by making it either unnecessary (relativism) or impossible (dogmatism) (p. 60)

My own reading of this is that the most useful concept here is to see in all texts the ways in which they relate to, which is not to say relativize, other texts. Taken from this point of view, hidden dialogicality might be understood to mean any of the multiple ways in which the intertextualities or dialogization of a text are obscured and which thereby produce an allegiance with the hegemonies of the plain, natural, or unremarkable discourses of the world as taken for granted. I do this realizing that, unless I misinterpret him, I am going against Bakhtin's own argument that hidden dialogicality is not the same as what he calls hidden polemic. In a hidden polemic the author's discourse is directed toward its own referential object, as is any other discourse, but at the same time every statement about the object is constructed in such a way that, apart from its referential meaning, a polemical blow is struck at the other's discourse on the same theme, at the other's statement about the same object (1984: 195).

The essential meaning I read here is the influence of another's argument on the polemical or rhetorical shape of one's own argument. It contrasts with direct polemic which simply would say, X says Y and I argue against that thus. These 'sideward glances', as Bakhtin calls them, seem the stuff of hidden dialogicality and I find only the polemical nature to distinguish between them.

5. Hidden Dialogicality and Discourse Representation Now if we bring together this idea of hidden dialogicality from Bakhtin and Fairclough's analysis of discourse representation it becomes clear that we will never find a sharp distinction between overt forms of intertextuality such as direct quotation and the various covert forms of intertextuality we have in metaphor, hedging, presupposition, or negation (Fairclough 1992). This may be looked at from one of two points of view. On the one hand it might be apparent that the use of direct quotation with a clear citation of the source as we find in much academic discourse but also in casual conversation is an overt representation in one text of another text which was originally produced in another discourse. This would seem a fairly overt sense of dialogicality as Bakhtin has discussed it. In this view, the less direct forms of discourse representation such as presupposition and negation which Fairclough discusses might be thought of as

436

Ron Scollon

being, or at least bordering on, hidden dialogicality. As he comments, the use of the phrase 'the Soviet threat' presupposes another voice making the claim that there is such a threat but this other voice is only heard in this discourse as a 'silent' partner. On the other hand, as I have argued elsewhere (Scollon 1998a, b, c, d) following on Bell's earlier discussion of news discourse (1991), one function of the direct quotation in journalistic discourse is to separate the journalist from responsibility for the utterance quoted. That is, if we use Goffman's (1974, 1981) notion of the distinction between the author and the principal, between the one who produces the wordings of the utterance and the one who is positioned as having to take responsibility for the consequences of the utterance, direct quotation is frequently used in journalistic discourse to shield the journalist from having to take responsibility for his own utterances. When a journalist uses direct quotation, at least in some cases he or she is responding to clearly defined legal strictures having to do with libel, defamation of character, and the like and so in this use of direct quotation is hidden another voice, the voice of newspaper policy which itself is ventriloquating legal conditions. In such a case it could be argued that direct discourse representation, while appearing to be quite an overt form of dialogicality is, in fact, responding to and hiding dialogical relationships among the author, that is, the journalist, the newsmaker, the news organization for which the journalist works, and the libel laws of the state within which the text is produced.

6. Audience Design, Referee Design Before rushing ahead to focus on still other cases of hidden dialogicality, it might be useful to ask whether or not we are talking about the same thing that Bell had in mind when he wrote of 'referee design' (Bell 1991). Audience design is relatively well understood as those characteristics of a message which take the audience into consideration. While this can range from the pronouns used to the style or register of the utterance, the central concept is that the language of the message is tuned to the receiver or addressee. Referee design was introduced to handle those cases of news and advertising discourse, mass media cases, where the language or style or register chosen reflected not characteristics of the sender or the receiver but of a third party. When an advertiser in Hong Kong uses an Italian name or Italian words in an advertisement, it is taken for granted that neither the producer of the goods nor the viewer/reader of the advertisement speak or read Italian. Italian takes on its force in this case through its association with, perhaps, fashion. The referee - the Italian fashion industry - is not present in the discourse except as what Goffman might call a bystander, but the characteristics of that bystander provide a kind of

Hidden Dialogicality:

When Infelicity Becomes Fear of

Infringement

437

dialogicality with the exchange between Chinese advertiser and Chinese consumer in Hong Kong. It is a voice which might be thought of as saying, 'You engage, at least in part, in this discourse because of our Italian-design discourse. Welcome and enjoy it.' As I have just suggested, we might want to gather into this concept some of Goffman's ideas about various less-than-currently-ratified participants in a discourse. And I would want to bear in mind that there is little to limit our imagination of the number and kinds of positions which may be taken in any particular discourse. I am particularly mindful of Goffman's idea of containment (1974) in which a frame is placed around an exchange so that it appears to be other than what it is. Goffman's interest in conmen and tricksters of all kinds should be remembered here as we open up our analysis of the voices in a text, in a discourse, to voices outside of it, behind it, under it, sneaking around through it, or hidden in the wings orchestrating it. Put another way, I would say that what I have come to here is a position that is not so interested in a formal definition of hidden dialogicality as in making sure that, when we look at texts from photographs of fast food restaurants to presidential addresses, we continue to pull back the curtain that hides the wizard behind, who is ventriloquating what appears to be the main discourse.

7. Taboos as Hidden Dialogicality I started with the idea that the ownership of commodity/signs, to use Coombe's (1998) term might be producing a hidden dialogicality in news and other forms of public discourse. I suggested, but of course have not at all demonstrated, that if there is a perceived concern with sanctions against the use of particularly semiotic properties, an utterance which reflected that concern would be engaged in a kind of hidden dialogicality with the owners of those properties and in that would be seen the intervention of the intellectual property laws of the state into common public discourse. Now we might ask if it has not always been thus. Have there not been taboos of all kinds producing hidden dialogicalities with utterance in virtually any social sphere? Is there anything really different about a journalist avoiding the registered name 'the body' in making reference to the Governor of Minnesota and the same journalist avoiding the use of a pejorative word making reference to members of a particular social group because of the concern with being labeled 'X-ist'? In other words, is hidden dialogicality basically the socio/discursive production of expectations, hopes, or fears which in other contexts might be called simply selfcensorship?

438

Ron Scollon

8. How Do You Know, Baby, When the Phone Don't Ring, It's Me? To come to a conclusion here, it seems that we are really stuck where we started with hidden dialogicality. It seems that we can only argue that hidden dialogical partners are there as dialogical contributors to a discourse to the extent we can establish that the producers of the utterances expect, fear, hope, or wish them to be there. Intellectual property, our erstwhile subject of this essay, has been transformed from a hidden dialogical partner of fairly concrete and legal dimensions into a set of fears or expectations of sanction. Can we go so far as to say that hidden dialogicality derives its dialogical interest and power exactly from the fact that it is hidden? To the extent this is true, I would argue that it should become part of the agenda of a critical study of discourse and of dialogue to seek out and discover the hidden dialogicalities in utterances. I would argue that this view moves the study of intellectual property and the commodity/sign along with the study of discursive taboos and self-censorship to central stage in our study of dialogue.

References Austin, John (1962): How to do Things with Words. Oxford: Clarendon Press. Bakhtin, Mikhail (1984): Problems of Dostoevsky's Poetics. Caiyl Emerson (ed. and tr.) Minneapolis: University of Minnesota Press. Bell, Allan (1991): The Language of the News Media. Oxford: Basil Blackwell. Brown, Penelope and Stephen Levinson (1978): Universals in Language Usage: Politeness Phenomena. In: Burke, Kenneth (1945): A Grammar of Motives. Englewood Cliffs, N. J.: Prentice-Hall. Caldas-Coulthard, Carmen Rosa (1993): From Discourse Analysis to Critical Discourse Analysis: The Differential Re-presentation of Women and Men Speaking in Written News. In: John M. Sinclair, Michael Hoey, and Gwyneth Fox (eds.) Techniques of Description: Spoken and Written Discourse. London: Routledge, 196-208. Caldas-Coulthard, Carmen Rosa (1994): On Reporting Reporting: The Representation of Speech in Factual and Factional Narratives. In: Malcolm Coulthard (ed.) Advances in Written Text Analysis. London: Routledge. Caldas-Coulthard, Carmen Rosa (1997): News as Social Practice. Florianopolis: P6s-Gradua?ao em Ingles. Cass, Dennis (1999): An Action Figure for All Seasons: Minnesota Elects a Toy Governor. Harper's, February 1999, 298: 65-71. Chen, Ping (1993): Modern Written Chinese in Development. Language in Society 22: 505-537. Coombe, Rosemary J. (1998): The Cultural Life of Intellectual Properties: Authorship, Appropriation, and the Law. Durham and London: Duke University Press. DeFrancis, John (1984): The Chinese Language: Fact and Fantasy. Honolulu: The University Press of Hawaii. Fairclough, Norman (1992): Discourse and Social Change. Cambridge: Polity Press. Fairclough, Norman (1995): Media Discourse. London: Edward Arnold. Goffman, Erving (1974): Frame Analysis. New York: Harper and Row. Goffinan, Erving (1981): Forms of Talk. Philadelphia: University of Pennsylvania Press.

Hidden Dialogicality:

When Infelicity Becomes Fear of Infringement

439

Goody, Ester (ed.): Questions and Politeness: Strategies in Social Interaction. New York: Cambridge University Press. Republished as (1987) Politeness. Cambridge: Cambridge University Press. Grice, H. P. (1975): Logic and Conversation. In: P. Cole and J. Morgan (eds.). Syntax and Semantics, Volume 9: Pragmatics. New York: Academic Press. Gumperz, John (1977): Sociocultural Knowledge in Conversational Inference. In: M. Saville-Troike (ed.), 28th Annual Round Table Monograph Series on Language and Linguistics. Washington, DC: Georgetown University Press. 191-212. Gumperz, John (1992): Contextualization and Understanding. In: A. Duranti and C. Goodwin, (eds.) Rethinking Context. Cambridge: Cambridge University Press. 229-252. Hall, Stuart (1980): Encoding/Decoding. In: S. Hall, D. Hobson, A. Lowe, and P. Willis (eds.), Culture, Media, Language: Working Papers in Cultural Studies 1972-1979. London: Hutchinson, 128-138. Harper's Index Harper's, February 1999, 298 (1785): 13. Holquist, Michael (1981): Glossary. In: Bakhtin, M. M. (1981) (Originally published in 1934-5). The Dialogic Imagination. Austin: University of Texas Press. Kraus, Richard Curt (1991): Brushes with Power: Modern Politics and the Chinese Art of Calligraphy. Berkeley: University of Calfornia Press. Michael Macovski (ed.): Dialogue and Critical Discourse. Oxford: Oxford University Press. 2 1 5 236. Norman, Jerry (1988): Chinese. New York: Cambridge University Press. Pan, Yuling (to appear): Beyond Face Strategies: Linguistic Politeness in Chinese Newspapers. In Troi Carleton and Robin Queen (eds.). Publisher to be announced. Peterson, Glen (1995): Literacy Campaigns in China under Mao and Deng. Lecture to the Contemporary China Research Centre, Tuesday, 5 December 1995, City University of Hong Kong. Scollon, Ron (1998a): Mediated Discourse as Social Interaction: A Study of News Discourse. New York: Longman. Scollon, Ron (1998b): Intertextuality Across Communities of Practice: Academics, Journalism, and Advertising. Paper presented at the Symposium on Discourse Across Languages, University of Wisconsin-Milwaukee, 17-19 September 1998. Scollon, Ron (1998c): Plagiarism and Discourse Representation as Hegemonic Social Practice: Ownership and Positioning. Paper presented in the session 'Globalism in Hong Kong Public Discourse: Post-modern ideology and local resistance' at the Annual Meetings of the American Anthropological Association, Philadelphia Dec 2-6, 1998. Scollon, Ron (1998d): Globalism and the New World Dynasty: Person, Property and Power in Intertextuality. Chairman's comments for the session 'Globalism in Hong Kong Public Discourse: Post-modern ideology and local resistance' at the Annual Meetings of the American Anthropological Association, Philadelphia Dec 2-6, 1998. Searle, John R. (1969): Speech Acts. New York: Cambridge University Press. Widdowson, H. G. (1998): Review Article: The Theory and Practice of Critical Discourse Analysis. Applied Linguistics 19(1) 136-151. Yuyan Wenzi Guifan Shouce (Zengdingben) (1993): Beijing: Yuwen Chubanshe Chuban. [Handbook of Standardized Characters]

List of Contributors Julia Bamford, Dip. Studi Aziendali e Sociali, Facoltà di Economia, Università di Siena, Piazza S. Francesco 17, Siena 53100, Italy. Christine Berger, Département de Linguistique Générale et Appliquée, Laboratoire d'Etudes Sur L'Acquisition et la Pathologie du Language, Université René Descartes Chez L'Enfant (UMR 8606, Paris V - CNRS), 12 rue Cujas, 75005 Paris, France. Josiane Boutonnet, School of Humanities, University of Wolverhampton, Castle View, Dudley, DY1 3HR, United Kingdom. Joanna ChannelI, Channell & Associates, PO Box 5494, Derby, DE22 1ZQ, United Kingdom. Janet Cotterill, Centre for Language and Communication, Cardiff University, P.O. Box 94, Cardiff CF10 3XB, United Kingdom. Malcolm Coulthard, Department of English, The University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom. Fathi Dali, Faculté des Sciences Humaines et Sociales de Tunis, Université des Lettres des Arts et des Sciences Humaines, 94 Bd du 9 Avril 1938, Tunis 1007, Tunisia. Alison Ferguson, Department of Linguistics, University of Newcastle, University Drive, Callaghan Newcastle NSW 2308, Australia. Julia Gillen, School of Education & Centre for Human Communication, Manchester Metropolitan University, 799 Wilmslow Road, Manchester M20 2RR, United Kingdom. Michèle Grosjean, Université Lumière Lyon 2 - GRIC, 5 Avenue Pierre Mendes France CP, 69676 BronCedex, France. Michèle Grossen, University of Lausanne, Department of Psychology, BFSH 2, 1018 Lausanne, Switzerland. Christian Hudelot, Département de Linguistique Générale et Appliquée, Laboratoire d'Etudes Sur L'Acquisition et la Pathologie du Language, Université René Descartes Chez L'Enfant (UMR 8606, Paris V - CNRS), 12 rue Cujas, 75005 Paris, France. Damir Kalogera, Kneza Borne 18, 10000 Zagreb, Croatia.

442

List of Contributors

Murray Knowles, Department of English, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom. Almut Koester, Rablstrasse 46, 81669 München, Germany. Gerda Lauerbach, Institute of English & American Studies, Goethe University, PO Box 11 19 32, 60054 Frankfurt/M., Germany. Susan Marídala, Department of English, Royal Holloway College, University of London, Egham Hill, Egham, Surrey, TW20 OEX, United Kingdom. Carol Marley, Department of English Language & Literature, University of Liverpool, PO Box 147, Liverpool, L69 3BX, United Kingdom. Jim Martin, Department of Linguistics, University of Sydney, Sydney, NSW, Australia 2006. Birgit Meerholz-Haerle, Plinganserstr. 114b, 81369 München, Germany. Martin Montgomery, Department of English Studies, University of Strathclyde, Glasgow Gl 1XH, Scotland. Rosamund Moon, Department of English, The University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom. Barry Natusch, College of Economics, Nihon University, 1-3-2 Misaki-machi, Chiyoda-ku, Tokyo 101, Japan. Charles Owen, Department of English, The University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom. Dorota Pacek, Department of English, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom. Luis Pérez-González, Departamento de Traducción y Interpretación, Universidad Europea de Madrid, Campus Universitario, C-17, 28670 Villaviciosa de Odón, Madrid, Spain. Arja Piirainen-Marsh, Dept of English, University of Jyvaskyla, POB 35, 40351 Jyvaskyla, Finland. Boris Pritchard, Rijeka College of Maritime Studies, University of Rijeka, Studentska 2, 51000 Rijeka, Croatia. Frances Rock, Department of English, The University of Birmingham, Edgbaston, Birmingham, B15 2TT, United Kingdom. Klaus P. Schneider, Englisches Seminar, Universität Bonn, Regina-Pacis-Weg 5, 53113 Bonn, Germany.

List of Contributors

443

Ron Scollon, Department of Linguistics, Georgetown University, Washington, DC 20007, USA. Branca Telles Ribeiro, 191 Mystic Street, Arlington MA 02474, USA. Michael Toolan, Department of English, The University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom Polly Walsh, Dip. Lingue, Facolte di Economia, Universite di Firenze, Via Curtatone 1, 50100 Florence, Italy. Edda Weigand, Universität Münster, Fachbereich 9: Philologie, wissenschaft, Bispinghof 2B, 48143 Münster, Germany.

Sprach-

Peter White, Department of English, The University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom. Agnès Witko, Université Lumière Lyon 2 - GRIC, 5 Avenue Pierre Mendes France CP, 69676 BronCedex, France.