Theoretical Approaches to Dialogue Analysis: Selected Papers from the IADA Chicago 2004 Conference 9783110928198, 9783484750333

Citation preview

Beiträge zur Dialogforschung

Band 33

Herausgegeben von Franz Hundsnurscher und Edda Weigand

Theoretical Approaches to Dialogue Analysis Selected Papers from the IADA Chicago 2004 Conference

Edited by Lawrence N. Berlin

Max Niemeyer Verlag Tübingen 2007

Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http://dnb.ddb.de abrufbar. ISBN 978-3-484-75033-3

ISSN 0940-5992

© Max Niemeyer Verlag, Tübingen 2007 Ein Imprint der Walter de Gruyter GmbH & Co. KG http://www.niemeyer.de Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Printed in Germany. Gedruckt auf alterungsbeständigem Papier. Druck und Einband: AZ Druck und Datentechnik GmbH, Kempten

Table of Contents INTRODUCTION Lawrence N. Berlin Bridging the Atlantic

1

PART I: DIALOGUE AS TEXT Francesco Carota Gathering Common Ground on the Negotiated Topic: The Role of the Contrastive Markers of Italian

13

Karin Aijmer Dialogue Analysis in a Cross-Linguistic Perspective

25

Joanna Szerszunowicz A Comparative Analysis of Small Talk in English and Polish

39

Johanna Miecznikowski Modality and Conversational Structure in French

49

PART II: DIALOGUE AS INTERACTION Franz Hundsnurscher The Principles of Dialogue Grammar

61

Anita Fetzer Validity Claims in Context: Monologue Meets Dialogue

77

Ernest W. B. Hess-Lüttich Dialogue Theory - A Sociosemiotic Approach: On the Pragmatics of Literary Communication

89

Maria Langleben On the Explicitness of Themes in Dialogue

107

VI

Lenke Nemeth A Subjectivity Formation Model for Dialogue Analysis in Drama

123

PART III: DIALOGUE AS DISCOURSE Lawrence N. Berlin Grounded Theory and its Benefits for Dialogue Analysis: The Beginning of a Beautiful Friendship

131

RichardHallett & Judith Kaplan-Weinger And Now for Your Moment of Zen: A Multi-modal Discourse Analysis of Humor in "The Daily Show with Jon Stewart"

151

Herta Maurer-Lausegger Audiovisual Dialectology: Methodology and Theoretical Considerations

163

Herve Varenne The Production of Difference in Interaction: On Culturing Conversation through Play

177

Douglas Macbeth Sequential Analysis in an Ethnomethodological Key: Order without Theory

199

Christian Hudelot The Use of a Functional Dialogic Model of Verbal Interaction to Compare how Daycare Workers and Teachers Scaffold 3-year-old Children

215

Mireille Froment Teacher and Preschool-Level Pupils' Dialogue around a Cartesian Table: Categorizations, Accentuations and Emergence of Routines

229

Lawrence N. Berlin Bridging the Atlantic

1. Introduction: Defining and Framing Dialogue

In April 2004, a group of international scholars convened in Chicago under the aegis of the International Association for Dialogue Analysis to discuss issues related to theoretical approaches to dialogue analysis. Representing various theoretical and methodological perspectives, the selected papers presented in this volume offer a view of the breadth of dialogue analysis from its more traditional origins to recent developments in the areas of discourse studies. While the term "dialogue" has gained much popularity in the public domain (even undergoing a functional shift in English to be used as a verb: to dialog), more traditional scholarship in dialogue analysis is not as familiar to scholars in the Americas. The conference and this volume, then, aim to provide an overview of dialogue analysis and its developments along with more familiar trends in discourse analysis and discourse studies. Dialogue is defined by most English dictionaries as a conversation between two or more (i.e., two-way communication). Hundsnurscher (1980, this volume; see also Weigand 1989) characterizes dialogue in terms of a verbal interaction, the intended outcome of which is to produce a common goal among speakers through a series of moves. Consider, then, Schegloff s (2001) definition of discourse as emanating from "multi-unit talk production" or "the product of conversation" (230). Taken in this way, dialogue and discourse do not appear vastly different and may be conceived of in terms of process and product. Yet, despite the ostensible divide between dialogue and discourse, there has been a great deal of cross-fertilization among the scholars from both the European and American academe, rendering the lines less starkly drawn between the two. Compare, for example, the classification of discourse presented by Jaworski and Coupland (1999), and further refined by Schiffrin, Tannen, and Hamilton (2001). They categorize discourse according to three components: 1) features (i.e., elements - linguistic and nonlinguistic - that form part of the social practice associated with language 2) structure (i.e., beyond the level of sentence); and 3) function (i.e., how one uses the language). This framework parallels my own estimation of three essential considerations in dialogue analysis: 1) identifying the "unit of analysis"; 2) defining the nature of "interaction"; and

2

Lawrence N. Berlin

3) determining the role of "context". These considerations are taken up in the next section. Moreover, they are taken generally to define the structure of the chapters which will be organized along the following lines: I. Dialogue as Text II. Dialogue as Interaction III. Dialogue as Discourse The reader is encouraged to examine the various chapters for their treatment of the three considerations as well as their interpretation of dialogue while proceeding through the text.

1.1. Identifying the Unit of Analysis In early linguistic studies, dialogue was idealized as a verbal interaction between two interlocutors (a speaker and a hearer) and the basic dialogic unit was conceived of as two moves: initiation and response (Halliday 1984). This "exchange" was viewed as "language as behavior," the natural counterpart to "language as code" (i.e., the traditional grammatical analysis). Thus, the dialogue embodies the semiotic acts of "signifying" or trying to create meaning through language (or from the original Greek dialogos meaning 'through words') and an understanding of its underlying systems. Furthermore, when combined with the work of Austin (1962) and Searle (1969), much has been accomplished to help us understand that dialogue is not simply comprised of an initiative and a response (Flanders 1970), but extends into an ever-expanding variety of structures - moves, turns, transactions, exchanges, sequences, etc. (see Coulthard, Montgomery, and Brazil 1981; Sinclair and Coulthard 1975). While any of these could serve as a unit of analysis, some researchers choose to start from different points, smaller or larger. Thus, the phenomenon under investigation could be as broad or as narrow as the researcher wishes it to be, but the chosen unit of analysis and subsequent design must fit the question being asked. To assist in this matter, Lincoln and Cuba (1985) assert that any unit of analysis, regardless of the size, must minimally meet two criteria: independence and relevance. Independence relates to the notion of mutual exclusivity. Relevance is directly concerned with the research question in that the units of analysis, whatever they may be, are adequate to respond to the research question. In Theoretical Approaches to Dialogue Analysis, the various scholars define their units of analysis within the context of their own research, ranging from the varied uses of lexical items within the same language (Carota this volume; Miecznikowski this volume) or in cross-linguistic comparisons (Aijmer this volume; Szerszunowicz this volume), to themes as represented in literary texts (Langleben this volume).

Bridging the Atlantic

3

1.2. Defining the Nature of Interaction It was Hymes' (1964, 1972, 1974) original work that extended the unit of analysis beyond a purely linguistic unit into the analysis of speech as a communicative event (i.e., his "speaking" model). This extension has added new dimensions to doing things with words and recognizing how and when the accomplishment of those functions is deemed appropriate in a given language (i.e., developing a "communicative competence"). In Halliday's (1978, 1984) discussions of interaction, he has maintained that it is necessary to understand the connection between the linguistic system and conversational process. The merging of what he refers to as the logical-philosophical "code" and the ethnographic-descriptive "behavior" would prove to be a powerful step in the development of a comprehensive theory of dialogue. This evolving picture also situates dialogue analysis within the study of pragmatics (Bardovi-Harlig and Mahan-Taylor 2003; Jucker 1995; Jucker, Fritz, and Lebsanft 1999). As such, scholars in the field of pragmatics have been equally interested in examining how status, intentions (Brown and Levinson 1987; Grice 1975), themes, genres (Hyland 2002), inter alia function within dialogue as forces that can motivate interlocutors to verbally interact in certain ways. Likewise, particular features have often been the focus of study (e.g., discourse markers, suprasegmentale, keyword and phrase repetition). Considering how two-way communication through words is not necessarily limited in time and space, dialogue analysts have begun to push the envelope beyond the traditional study of a pair of speakers exchanging some words. For instance, with the advent of technology as a means of communicating, visual semiotics has also begun to play a role in the analysis of dialogue as images included in e-mail correspondence, text-messaging, and advertisements where these images are intended to convey express messages to particular audiences (cf. Kress and van Leeuwen 1996). While a more traditional interpretation might argue that these components are outside the scope of dialogue analysis, their intended use to convey meaning from a speaker/writer/presenter to a hearer/reader/receiver - is not compromised by their inclusion. Moreover, their inclusion does not exclude or negate the most basic level of dialogue: through words; it merely expands it. The various chapters in Theoretical Approaches to Dialogue Analysis also range in their treatment of interaction, using both traditional and unique interpretations. Thus, while Hundsnurscher (this volume) presents a traditional, two-person interaction, Varenne (this volume); Hudelot (this volume); and Froment (this volume) examine group interactions. Taking a unique approach to interaction represented through a combined news/talk show format, Hallett and Kaplan-Weinger (this volume) explore the intended message and humor of a television show, wherein the hearer/reader/receiver's response depends on shared meanings and is gauged by the increasing popularity of the show. Maurer-Lausegger (this volume) deals with the observer 's paradox uniquely by making the videocamera a partici-

4

Lawrence N. Berlin

pant in the interaction and Berlin (this volume) offers examples where nonverbal responses can be considered as part of the conversation.

1.3. Determining the Role of Context Within dialogue analysis, researchers often debate the importance of context. While some would prefer to focus solely (or at least primarily) on the language, others contend that features such as speaker intention can't be interpreted without consideration of the relationship of the interlocutors and the broader situation in which the language is being used. Context has a duality, however, in that the language itself indexes a particular context and it establishes a framework within which language can occur (Goffman 1974; Gumperz 1992). To expand on the earlier discussion of units of analysis, Halliday (1978) elaborates his code and behavior into "context of culture" and "context of situation" respectively. In this sense, language is situated within the broader culture and behavior occurs relative to the situation; such an interpretation fulfills the positioning of dialogue in a broader pragmatic framework and, simultaneously, reconciles the various disciplines that have seen language as central and have contributed to the theoretical foundations of linguistics. Thus, rather than taking us beyond words, it leads to a deeper understanding of their meaning (cf. Duranti & Goodwin 1992). The duality of context can also manifest itself in terms of the paradigm the researcher follows. Within an ethnographic perspective, for instance, Spradley (1980) identifies the context along similar lines as Goffman (1974); he provides a framework for communication to take place, including a setting (or "field of action" after Goffinan) and an activity ("focal event"). Spradley further includes the participants when examining, suggesting that their relationship fills an equal place in the framework. Taking another tack, context can be construed in terms of the influence it exerts on the communication. Thus, Bakhtin's (1981) notions of dialogicity and historicity are revealed in the synchronic and diachronic influences that have informed the texts (cf. Moerman 1988; Urban 1991). Within Theoretical Approaches to Dialogue Analysis, many of the authors view context as critical to the analysis. Fetzer (this volume), for example, sees dialogue as contextdependent, echoing Goffman and Spradley's notion of a framing context for dialogue. Berlin (this volume) and Nemeth (this volume) explore Bakhtin's multiple "voices" as they occur within dialogue and index multiple layers beyond the immediate talk. Macbeth (this volume) and Hess-Luttich (this volume) manage to bridge the two; the former in his discussion of the "double move" and the latter in his detailed model for literary dialogue.

Bridging the Atlantic

2. Organization of the Volume

Organizationally, the chapters are divided into three sections according to their primary focus: Dialogue as Text, Dialogue as Interaction, and Dialogue as Discourse. The selected papers included in this volume feature scholars from both sides of the Atlantic whose contributions cover multiple perspectives. Thus, Theoretical Approaches to Dialogue Analysis presents a unique work that provides essential information for the scholar, researcher, and student of dialogue, as well as those interested in discourse, language, linguistics, pragmatics, and research methodology.

2.1. Dialogue as Text The authors included in this section have focused their work on textual data, identifying particular lexical constituents and their syntactic and/or semantic interface as primary in their analysis. While all chapters deal with the interactional and contextual parts of dialogue, predominance is given to the examination of more traditional linguistic units of analysis in the emergence of dialogue as text. To begin this section, Francesca Carota discusses contrastive discourse markers in Italian. Using examples of exchanges from a spoken corpus, she examines the linguistic environment of the contrastive connectors. Taking a typological approach, Carota then goes on to provide a descriptive taxonomy in which she details the functions of the various contrastive discourse markers and provides possible explanations for their usage. Karin Aijmer begins this section with Dialogue Analysis in a Cross-Linguistic Perspective. She uses parallel (translated) corpora to conduct a typological comparison of English, Swedish, and German. By focusing on a single lexical-semantic element, Aijmer chooses a cognate, the modal adverb surelyΔ to explore how a simple modal can become grammaticalized, both intralingually and/or interlingually through the process of translation; that is, translation in parallel corpora can evince syntactic or even semantic changes that indicate the importance of intersubjectivity in dialogue, or the repositioning of a phrase relative to specific interlocutors. In the next chapter, Joanna Szerszunowicz combines the work of the two former authors by taking a typological look at data from cross-linguistic corpora. Specifically, she focuses on small talk as a genre and describes the spontaneous speech of second language learners of English and Polish. The analysis identifies common topics generated by the learners and

6

Lawrence N, Berlin

provides their differential reactions (based on their first languages and cultures) to the importance of the various topics. In Modality and Conversational Structure in French, Johanna Miecznikowski follows a similar path to Aijmer's earlier chapter in that both begin with a corpus, both choose a particular lexical item, and both investigate modality. It is here, however, that the similarities end. Miecznikowski, like many other contributors in this volume, contends that the focus in dialogue analysis should be on use, not merely form. Using a data-driven interactionist approach, she establishes that the multiple surface forms obtained (of the verb vouloir) in this dialogue analysis indicate a clear, sequential organization in the discourse. Thus, the surface manifestation of vouloir indexes user preferences for particular forms relative to the type of interaction. Miecznikowski therefore proposes that an interactionist perspective could prove informative for researchers in many areas of dialogue analysis from those who favor a more linguistic approach to conversation analysts.

2.2. Dialogue as Interaction Moving beyond a primary focus on the textual, the authors in this section extend the units of analysis to include exchanges between two or more participants as interaction takes a central focus. Though representing differing theoretical perspectives and methodologies, these chapters reveal essential information about the nature of interaction and, in some cases, the critical influence that context can have on interaction. Consequently, in each chapter as the data are rendered, noteworthy aspects of the study of dialogue are made apparent. Again, the reader will find the treatment by each author of the units of analysis, the definition of interaction, and the role of context to be informative in developing a greater understanding of the breadth of the field of dialogue analysis. The chapters focusing on dialogue as interaction begin with an offering by Franz Hundsnurscher, The Principles of Dialogue Grammar. He proceeds from a linguistic perspective and presents a traditional representation of dialogic interaction based in classical argumentation. To this end, Hundsnurscher identifies the quintessential unit of analysis within dialogue, defining it minimally in terms of three moves engaged in by at least two competent speakers: an initiation (in the form of a speech act), a reaction, and some form of uptake (relative to the reaction). He gradually builds up the model by using global types of illocutionary speech acts to demonstrate (i.e., representatives, directives, commissives, and expressives). The complicating information that may arise within the units of dialogue, however, affirm his assertion that perlocution is more essential to dialogue analysis than illocution, as the reactions and uptake may take on a variety of forms depending on previous moves by interlocutors. In this manner, we gain a fundamental view of two-way com-

Bridging the Atlantic

J

munication. Hundsnurscher establishes a starting point for dialogue analysis by framing the most basic structure for conducting research. Following Hundsnurscher, Anita Fetzer elaborates a more pragmatic view of dialogue, taking on the challenge of defining it in terms of both process and product. In her chapter, Validity Claims in Context: Monologue Meets Dialogue, she provides an overview of speech act theory, relevance theory, and conversation analysis as they relate to a deeper understanding of dialogic interaction within a communicative context. Contending that dialogue is a cooperative endeavor, Fetzer offers a unique approach by incorporating Habermas' notion of validity claims to express how coparticipants' contributions within a dialogue must have reference to the objective, social, and subjective worlds. As such, and in relation to the Gricean Cooperative Principle, Fetzer suggests that dialogue emerges as a resolution of individuals' inferencing and reasoning strategies (i.e., coparticipants' individual internal intentions and planned monologues) and collective inferencing and reasoning strategies (i.e., efforts to cooperate and meet interlocutors' face needs). Consequently, a theoretical perspective surfaces that outlines an intricate definition of interaction. Ernest W. B. Hess-Lüttich extends dialogue theory beyond speech act theory and conversation analysis. Taking A Sociosemiotic Approach: On the Pragmatics of Literary Communication, he defines dialogue in terms of interaction, recognizing the limitations of terms such as 'interaction' which may be understood both in a generic, lay sense, and according to one or more disciplines wherein they take on a very specific use. Ultimately, Hess-Lüttich positions literary dialogue on multiple levels, between author and reader, between characters within a text (literature, play, or movie script), and - in the latter two cases - between actors and audience. Going beyond mere turn-taking and isolated speech acts, he presents an elaborate model that provides a pragmatic approach for the analysis of dialogue as defined by the social relationships between those communicating through a multiplicity of channels. In On the Explicitness of Themes in Dialogue, Maria Langleben's contribution focuses once again on interaction as represented through dialogue in literature. Using Pushkin's Boris Godunov to exemplify her discussion, Langleben details the seven types of themes that can appear in a comprehensive thematic analysis. As an approach to dialogue analysis of literary works, especially the scripted dialogue within plays, Langleben identifies repetition and dialogic tension as key elements in the analysis. Through anaphoric and cataphoric identification of repetitive units of analysis (i.e., from individual words to entire propositions), themes emerge within a well-constructed text to provide cohesion. The chorus of voices work together to produce various levels of interaction - reminiscent of HessLüttich - which, in turn, reveal the themes and make known the ultimate meaning of the author.

g

Lawrence N. Berlin

To round out this section, Lenke Nemeth shares A Subjectivity Formation Model for Dialogue Analysis in Drama. Focusing on extended segments of dialogue from the play Sexual Perversity in Chicago, Nemeth demonstrates how repetition in ostensibly discontinuous, elliptical dialogue can index essential components of character development, leading to a more complete understanding of the author's intentions. Subsequently, applying concepts forwarded by Bakhtin, she makes transparent meanings that the author intends to communicate to the audience by showing the interaction between the characters/interlocutors and their immediate context as a reciprocal process whereby extended stretches of dialogue construct a unique identity.

2.3. Dialogue as Discourse The final section of the volume focuses on discourse, taking the stance that context holds a pivotal role in defining archetypes or orders of discourse from which text derives. Dialogic interaction, then, emerges as the fulfillment of context-driven behavior wherein actors understand the roles they fill and utilize that knowledge in their linguistic and extralinguistic performance. In the first chapter of the section, Lawrence N. Berlin also explores a form of thematic analysis. In Grounded Theory and its Benefits for Dialogue Analysis, he claims that grounded theory enables researchers to overcome several methodological problems present in more traditional quantitative or merely descriptive analyses. Going through the procedures in step by step format, Berlin uses the script from the American movie classic Casablanca to illustrate how seemingly different approaches and designs can be integrated through the grounded theory methodology. Ultimately, the development of a matrix for constant comparison permits the triangulation of findings and the emergence of a theoretical model grounded in the data. Giving primacy to the interaction and the synchronic and diachronic contextual elements, Berlin arrives at his interpretive model weaving together critical discourse analysis, speech act theory, the cooperative principle, and politeness theory; thus, he asserts that seemingly disparate approaches - whether used independently or in tandem - will lead to the same conclusions under the rubric of grounded theory. Richard Hallett and Judith Kaplan-Weinger also focus on the centrality of context. In "And Now for Your Moment of Zen, " they use a multimodal approach to demonstrate how it can be manipulated to specific ends in the construction of hegemony. Hallett and Kaplan-Weinger reveal how the discourse of humor emerges as the result of combining and exploiting recognized contextual elements in the dialogue of The Daily Show with Jon Stewart, a news-variety show that been documented by several sources as having become the primary source of information for a growing number of people under 30. This critical

Bridging the Atlantic

9

discourse analysis advances intertextuality as the key to analysis as Hallett and KaplanWeinger explain why an examination of the combined spoken, written, and visual elements is imperative to link the dialogue synchronically and diachronically to other recognized forms of discourse. In another multimodal approach, Herta Maurer-Lausegger renders a "how-to" guide par excellence in Audiovisual Dialectology: Methodology and Theoretical Considerations. Combining technological advances with more traditional pursuits - in this case, the study of dialectology - she details a method that of conducting field research that could eventually function to preserve endangered languages. Within discourse studies, the methodology Maurer-Lausegger shares provides a way to record dialogues within their authentic contexts, demonstrating genuine interaction that can be digitized and analyzed for a variety of purposes. Continuing into the realm of conversation analysis and ethnomethodology as a means to explore discourse, Herve Varenne presents an anthropological viewpoint on dialogue in relation to culture. He distinguishes between "immortal" facts (i.e., cultural facts) as arbitrary and static versus the engagement of dialogue (i.e., "culturing through language) as temporal. Subsequently, he claims that "culturing" through language is instructed rather than learned. In other words, the existence of certain "immortal" institutions (e.g., hospitals, schools) does not determine the language use that unfolds there (i.e., the discourse); on the contrary, Varenne asserts that, in the enactment of dialogue as a form of play, humans will always attempt to push beyond the perceived limits of what is allowed. Asserting that certain acts - in this case, childbirth - are not as immutable as anthropologists (or structural linguists) would like to think, he traces the development of the dialogue as it creates the potential for altering the existing culture as it presents new facts that become part of the fabric of the discourse of giving birth. Douglas Macbeth takes an interesting approach by positing an anti-theory. In his chapter, Sequential Analysis in an Ethnomethodological Key: Order without Theory, he elaborates on the existence of "pre-theoretical worlds." In a philosophical discussion, he recognizes the reality of human experience long before it was ever theorized. With the introduction of positivism and positivist thinking, however, Macbeth declares that humans have redefined their own reality to the extent that it becomes unthinkable to imagine it without a theory. He further cautions that even qualitative research within a naturalistic, postpositivist paradigm risks reifying a positivist mindset by attempting to reduce human experience to a theory. Tracing the emergence of Discourse (with a capital "D") from the discourses that no doubt existed before they gained such prominence in research in language-in-use, Macbeth presents a dialogue from a classroom to illustrate how people engage in sequential, rule-governed language behavior not because of theory, but in spite of it.

10

Lawrence N. Berlin

Following with another Bakhtinian-inspired study, Christian Hudelot compares the results of children's dialogue output relative to the input of the adult interlocutors. In The Use of a Functional Dialogic Model of Verbal Interaction, Hudelot evaluates the quantity and quality of the children's contributions in didactic interactions with daycare workers versus preschool teachers. The subsequent categorization of the various segments suggests that the inherent features of teachers' scaffolding may promote the acquisition of child language in general, as well as the specific acquisition of the nature of social interaction. As such, the study of child dialogue (even within didactic discourse frames) becomes informative as dialogue analysts can begin to understand how units of dialogue may be constructed, how interaction is triggered, and what role context plays in a formative environment. Finally, Froment continues with an examination of the archetypal teacher-student discourse, focusing on how the teachers use their verbal contributions didactically to facilitate interaction and promote learning. Following the notion of Vygotsky's "zone of proximal development", the Cartesian table, then, becomes a locus for language development in preschool children. The variety expressed in Theoretical Approaches to Dialogue Analysis makes it apparent that dialogue analysis remains an exciting and evolving field of study. In any single discipline, the depth and breadth of research conducted advance only portions of a picture that will eventually lead to a comprehensive theory. As an interdisciplinary field, dialogue analysis often yields theories, methods, and analyses that appear on the surface to be conflicting, but in reality enrich the field and provide a more robust outcome because of the multiple foci explored. I believe that this text will serve as an instructive tool for the novice, as well as a resource for the scholar for years to come.

References Austin, John L. (1962): How to do things with words. - Oxford: Clarendon Press. Bakhtin, Mikhail (1981): The dialogic imagination. Four essays. Michael Holquist (ed.), trans. Caryl Emerson, Michael Holquist. - Austin: University of Texas Press. Bardovi-Harlig, Kathleen/Mahan-Taylor, Rebecca (2003): Teaching pragmatics. US Department of State, Bureau of Educational and Cultural Affairs, Office of English Language Programs. (http://exchanges.state.gov/education/engteaching/pragmatics/intro.htm) Brown, Penelope/Levinson, Stephen C. (1987): Politeness. Some universale in language usage. Cambridge: Cambridge University Press. Coulthard, Malcolm/Montgomery, Martin/Brazil, David (1981): Developing a description of spoken discourse. - In Malcolm Coulthard, Martin Montgomery (eds.): Studies in discourse analysis, 150. London, Boston and Henley: Routledge & Kegan Paul.

Bridging the Atlantic

\\

Duranti, Alessandro/Goodwin, Charles (eds.) (1992): Rethinking context. Language as an interactive phenomenon. - Cambridge: Cambridge University Press. Flanders, Ned A. (1970): Analysing teaching behaviour, - Reading, MA: Addison-Wesley. Goffman, Erving (1974): Frame analysis. - Boston: Northeastern University Press. Grice, H. Paul (1975): Logic and Conversation. - In Peter Cole, Jerry L. Morgan (eds.): Speech acts. Syntax and semantics, Vol. 3, 41-58. New York: Academic. Gumperz, John (1992): Contextualization and understanding. - In: Alessandro Duranti, Charles Goodwin (eds.): Rethinking context. Language as an interactive phenomenon, 229-252. Cambridge: Cambridge University Press. Halliday, Michael A. K. (1978): Language as social semiotic. The social interpretation of language and meaning. - London: Edward Arnold. Halliday, Michael A. K. (1984): Language as code and language as behaviour. A systemicfunctional interpretation of the nature and ontogenesis of dialogue. - In: Robin P. Fawcett, Michael A. K. Halliday, Sydney M. Lamb, Adam Makkai (eds.): The semiotics of culture and language. Language as social semiotic, Vol. 1, 3-35. London/Wolfeboro, NH: Frances Pinter. Hundsnurscher, Franz (1980): Konversationsanalyse versus Dialoggrammatik. - In: Akten des VI Internat. Germanisten-Kongresses Basel. Band 2, S. 89-95. Tübingen: Max Niemeyer Verlag. Hyland, Ken (2002): Genre. Language, context, and literacy. - Annual Review of Applied Linguistics 22, 113-135. Hymes, Dell (1964): Language in culture and society. A reader in linguistics and anthropology. New York: Harper and Row. Hymes, Dell (1972): Models of the interaction of language and social life. - In: John Gumperz, Dell Hymes (eds.): Directions in sociolinguistics. The ethnography of communication, 35-71. New York: Holt, Reinhart, and Winston. Hymes, Dell (1974): Foundations in sociolinguistics. An ethnographic approach. - Philadelphia: University of Pennsylvania Press. Jaworski, Adam/Coupland, Nikolas (1999): The discourse reader. - London and New York: Routledge. Jucker, Andreas (1995): Discourse analysis and relevance. - In: Franz Hundsnurscher, Edda Weigand (eds.): Future perspectives of dialogue analysis, 121-146. Tübingen: Max Neimeyer Verlag. Jucker, Andreas H./Fritz, Gerd/Lebsanft, Franz (eds.) (1999): Historical dialogue analysis. Amsterdam and Philadelphia: John Benjamins. Kress, Günther/van Leeuwen, Theo (1996): Reading images. The grammar of visual design. London: Routledge. Lincoln, Yvonna SVGuba, Egon G. (1985): Naturalistic inquiry. - Newbury Park, CA: Sage. Moerman, Michael (1988): Talking culture. Ethnography and conversation analysis. - Philadelphia: University of Pennsylvania Press. Schegloff, Emanuel A. (2001): Discourse as an interactional achievement III. The omnirelevance of action. - In" Deborah Schiffrin, Deborah Tannen, Heidi E. Hamilton (eds.): The handbook of discourse analysis, 229-249. Oxford/Maiden, MA: Blackwell Publishers. Schiffrin, Deborah/Tannen, Deborah/Hamilton, Heidi E. (eds.) (2001): The handbook of discourse analysis. - Oxford/Maiden, MA: Blackwell Publishers. Searle, John (1969): Speech acts. - Cambridge: Cambridge University Press. Sinclair, John M./Coulthard, R. Malcolm (1975): Towards an analysis of discourse. - London: Oxford University Press. Spradley, John P. (1980): Participant observation. - New York: Holt, Rinehart and Winston. Urban, Greg (1991): A discourse-centered approach to culture. Native South American myths and rituals. - Austin, TX: University of Texas Press. Weigand, Edda (1989): Sprache als dialog. Sprechakttaxonomie und kommunikative Grammatik. Tübingen: Max Niemeyer Verlag.

Francesca Carola Gathering Common Ground on the Negotiated Topic: The Role of the Contrastive Markers of Italian

1. Introduction

The contrastive connectives of Italian, such as ma ('but'), pero ('but'; 'however'), and invece ('instead'), permeate spoken dialogue. For instance, as exemplified in (1), the abovementioned connectives open the possibility for continuing statement like (a), realized here by the utterances (b), (c), and (d). (l)(a) Lo spettacolo e stato divertente stasera. The show was amazing tonight, (b) Ma a ehe ora efinito! But when did it finish? (c)Pero e stato molto lungo. But/However, it was very long, (d) Invece ieri la cena com 'e andatal Besides that, how was the dinner yesterday?

A question of interest is why these linguistic elements are used in the contexts just sketched and how they contribute to the semantics and the pragmatics of their host utterances. It is widely acknowledged in the literature that the corresponding contrastive connectives of English, namely but, however, and instead can operate at the discourse level and form a homogeneous class within the main category formed by the discourse markers (Fräser 1998). A direct parallel with the respective contrastive connectives of Italian, however, is not straightforward, since their discourse functions have not been investigated systematically and remain poorly understood. In addition, no attempt to cluster them based on their potential discourse role has yet been undertaken. Single connectives, in particular ma, have received special attention in previous work. In other words, the connective has been interpreted as a multifunctional marker, the synchronic uses of which would derive from a diachronic change (Bertinetto and Marconi 1984). Seen from a meta-textual perspective, ma signals the initial boundary of an utterance, a turn, or a wider unit in spoken dialogue, and marks the beginning of a paragraph in written text. In both cases, when opening a discourse unit, ma often correlates with a topic

14

Francesco Carola

shift (Serianni 2001). From an interactional point of view, it has been considered instead as a turn-taking device, whereby the speaker interrupts the turn of her interlocutor (Bazzanella 1995). Corpus data acquired from information-oriented dialogues bring evidence that the turn initial position does not correlate significantly with ma. In fact, the connective tends to occur at the beginning of an utterance, after positive feedback (e.g. si, d'accordd), produced within the same turn, like the example (2) turn (b) and example (3) turn (b) illustrate respectively. (2)(a) L' Eurostarparte alle 14.00. Le να bene I Orariol (b) Si. Ma mi dice se ci sono delle soluzioni meno caret (a) The train Eurostar will leave at 2 PM. Is the hour ok for you? (b) Yes. But can you tell me if there is any cheaper solution? (3) (a) Controllo sec'e un treno ehe parte prima. (b) D 'accordo. Ma non c 'e un treno diretto? (a) I will look for an earlier train. (b) Okay, but is there no direct train?

It emerges from corpus-data that also pero and invece share the positional behaviour just described with ma, occurring at the utterance opening after immediately preceding positive feedback. Moving from these empirical observations, we advance the hypothesis that the discourse behavior of ma, pero, and invece1 is sensitive to essential organizing principles of dialogue structure, namely dialogue topic management and the grounding process. The purpose of the chapter is to explain the pragmatic functions of the contrastive connectives, as well as to determine the contextual mechanisms underlying their discourse behavior motivating the interlocutors' preference towards the contextual use of a particular contrastive element. To this end, a corpus study was undertaken, focusing on the use of the contrastive connectives in the communicative contexts referred to so far. The organization of the chapter is as follows. Section 2 sets up the theoretical framework on which the work is based and clarifies the terminology employed; in section 3, some representative examples from the corpus are analyzed; in section 4, a taxonomy of the contrastive connectives and their function is proposed. Lastly, section 5 raises some questions to be addressed in future work.

Pero is an adverbial connective (Lonzi 1991), because of its free position within a clause or a sentence.

Gathering Common Ground

15

2. An Integrated Framework for Contrast!ve Markers

Collaboration and negotiation are intrinsic features of information-oriented dialogue. This typology of dialogue is shaped by the mutual cooperation between the co-participants involved into the conversation. Their active roles are fully engaged in establishing a shared path of navigation towards the accomplishment of a common conversational project (Bagerter, et al. 2004), which concerns the task of exchanging relevant information. Such a joint activity requires that the information to be discussed according to task-related needs is determined on line at any time of the conversation, assuming the form of a negotiation (Clark 1996; Linell 1999). In the present chapter, a conceptual distinction is introduced between the negotiation of the domain-dependent task (i.e., the discussion of some alternative solutions for the same task-related issue [Larsson 2002]), like the alternative "train-plane" in the example (4), and the negotiation concerning the cognitive, ideational meta-scheme of dialogue (i.e., the dialogue topic). It is proposed to configure the latter in terms of meta-negotiation (Carota 2005). Accordingly, in example (5), I'andata ('the departure') is the topic that has already been negotiated and completed. Conversely, /'/ ritorno ('the return') is a new macro-topic which will orient the next contribution by the interlocutor in which /'/ ritorno is established as being the ideational reference-point for the incoming utterance. (4)Preferisce viaggiare in treno o in aereol Do you prefer to travel by train or by plane? (5)(a) L 'andata va bene. Poi vorrei vedere gli orari del ritorno. (b) Allora, per il ritorno, vediamo...ci sarebbe un treno alle 8.30, iroppo prestol (a) The depature is okay. Afterward, I would like to know the return time. (b) So, for the return, let's see... there would be a train at 8.30 am; is that too early?

Importantly, it is assumed here that the meta-negotiation involves not only the management of the topic, but also the activity of its monitoring and its grounding. The term grounding refers to a collaborative process, by which the co-participants coordinate their private mental states (i.e., their attention, intentions, beliefs, and attitudes) at every level of communication, in order to achieve a common ground on the information under negotiation. The process of grounding is usually schematically described as involving a presentation phase, in which a co-participant contributes or presents content that the partner tries to register, and an acceptance phase, whereby the interlocutor accepts the contributed content and provides some feedback in order to evidence that the presented topic has been grounded (Clark 1996). The contribution model just sketched has been often

16

Francesco Carota

criticized for admitting the possibility of graded evidence for grounding, in which an acceptance can also represent a contribution to be accepted and so on, along a regression potentially fed ad infinitum (Traum 1999). To avoid this inconvenience, two degrees of linguistic evidence for grounding can be postulated: a positive degree (+grounded), corresponding to the acceptance of some contribution and signalled linguistically by positive feedback, or an acknowledgement, such as the abovementioned si, okay, certo and a negative degree (-grounded), signalled by negative feedback (Traum 1999). Elaborating on these observations, the notion of common ground is conceived here as referring to a shared cognitive context in which the co-participants' mental states get coordinated, tending to converge towards a mutual agreement2 with respect to the information to be negotiated (see also Traum and Dillenbourg 1998). The common ground is thus assumed to be a specific kind of cognitive context, which offers the contextual analytic line of the present chapter. The context is mirrored in the cotext, the dialogue concrete linguistic dimension, constituted by 1) the feedback; 2) co-referential and phoric links between dialogue entities; and 3) information structural phenomena which package the information. The dialogue entities are considered to be the objects, states, and events which have been linguistically introduced in the dialogue cotext. They have the referential status of being given if already mentioned in previous discourse, or being new if not previously mentioned in discourse (Prince 1981; Gundel, et al. 1993). The dialogue entities are represented as being active, semi-active, or inactive in the co-participant's mental states, using the terminology proposed in Chafe (1994).3 They are said to be co-activated if they are active with respect to the co-participants' common ground. Information structure refers to the organization of the utterance in terms of theme/rheme patterns (cf. Langleben this volume), which are encoded linguistically through the interface syntax/intonation in Italian (Cresti 2000; Lombardi Vallauri 2002). The theme is conceived of as representing the optional element of the utterance, which sets a topical coordinate at the local level of the host utterance, specifying the relationship between the utterance itself and the global dialogue topic.4 The rheme, on the other hand, is the obligatory element of the utterance, which performs the speaker's communicative goal by introducing some inactive information which has to be co-activated by the co-participants and submitted to their common ground. Being a pragmatic operator parallel to either the theme or the rheme or The notion of common ground is distinct from the philosophical notion of shared knowledge which represents semantic knowledge about the world. The notion of consciousness employed by Chafe (1994) is replaced here with that of mental state described above. Cresti (2000) points out that the theme ("topic" in her terminology) usually occupies the first position of an utterance and is marked by rising intonation.

Gathering Common Ground

\J

both, the focus highlights the salient information to be brought to the interlocutor's attention in order to be grounded.5 The interaction between the contextual and the cotextual factors depicted so far, along with the interplay of the local and the global levels, determine the semantic-pragmatic strategies which shape the dialogue structure6 during communication. Dialogue structure is assumed to represent the cotextual analytic line in the present corpus study. Consistent with these perspectives, the connectives under study are proposed to be contrastive markers. The linguistic elements used by a co-participant to express her personal attitude toward the way the information in the local host utterance globally relates to the dialogue cotext and to the cognitive context. This hypothesis was tested and supported through a preliminary corpus study of ma, pern, and invece.

3. Corpus Study

The material used here is drawn form the Adam Corpus (Soria, et al. 2000), which is a collection of information-seeking dialogues from the travel domain and consists of 58.000 words. The corpus analysis will focus on contrastive connectives7 which occur in open questions, as well as in declarative utterances performing either a statement or a request dialogue act.

3.1. Ma In the turn 3, example (6)8, the sequence si, ma starts a turn belonging to an acceptance phase, where si is an acknowledgement dialogue act, which positively ratifies the accept-

6

7

8

The prosodic focus is the peak in the intonational contour which is signalled by the fundamental frequency F0. Dialogue structure is modelled based on both common ground and topic. A detailed model of the dialogue topic structure cannot be made explicit here. For an overview, see Jaeger and Oshima (2002), Carota (2004, 2005). Due to the shortness of the present chapter, it presents only a restricted number of representative examples. However, the interpretation proposed here is based on the analysis of about 200 occurrences of the contrastive connectives. The notation used is the following: in 1 C/O, for example, the first number indicates the rum; C indicates the client, Ο the operator; text in brace brackets {...} followed by :|t| is marked as the theme, in square brackets [...] followed by F as the thematic focus.

18

Francesco Carota

-ance of the topic in the speaker's common ground. (6)(a) O: { // [riTORnof }:|t| cosa volevtf (b) C: { // ritorno } :|t| le avevo delta [doMEnicaf domenicapomeriggio [sul PREsto]F. (c)O: Si. Ma come [oRAriof? (a) The return, what did you want? (b) The return, as I said, it is on Monday Monday early in the afternoon.

(c) Yes. But what about the schedule? The host utterance of ma is a question asking for new information about the discourse entity orario ('schedule'), brought to the hearer's attention by means of focus and coreferring to the temporal discourse entity under focus in turn 2: sul presto ('early'). More precisely, the utterance is an elliptical, open question and can be resolved9 with respect to the main global discourse topic represented by the given entity ritorno ('return'), which is currently under meta-negotiation (as the contrastive theme {// \ritornoY} in turn 1, as the theme {// ritorno} in turn 2 indicates. As orario is a subcoordinate (i.e., a particular aspect of a previously established main global topic), it can be argued that ma correlates with a subtopic shift within a main global dialogue topic (i.e., flight), which is higher in the hierarchy of topicality. In conclusion, the ma-question starts a dialogue topic unit embedded in an overarching topic unit. In similar positions, ma expresses a contrast with previously given information which does not satisfy the expectations underlying the request in turn 3. In fact, before being grounded, the given information in focus, sul presto provided by the interlocutor in the previous turn, needs to be either corrected or clarified according to the entity currently highlighted by the focus orario. By using ma, the speaker temporarily blocks the current acceptance phase in order to manifest to the interlocutor that the presented topic needs to be partially changed or corrected before being grounded. For this reason, ma introduces a sort of request for repair of sul presto. It is used to reorient the grounding process toward a new perspective, by presenting a new subtopical coordinate to the common ground, which has to be accepted as being the current topic under meta-negotiation. In this sense, ma displays the speaker's attitude towards the status of the topic in the common ground and is significantly interrelated with the grounding process.

The treatment of elliptical questions follows Ginzburg (1998).

Gathering Common Ground

19

3.2. Pero In example (7), the contrastive marker pero follows positive feedback, by which the speaker accepts not only the informative content presented in the preceding turn, as in the previous example, but also the task proposed in the preceding question in turn 1. (7)(a) O: Leprenoto un [POsto]F7 (b) C: 5/. Pero vorrei prenotare anche [il HTOrno]f. (c)O: D'accordo. (a) May I reserve a place?

(b) Yes, but I would like to book the return too. (c) Okay. The host declarative utterance performs a request dialogue act, whereby the speaker adds some inactive new information through the discourse entity rilorno. Specifically, the new information is evidenced by means of both focus and the additive particle anche ('too') (König 1993), which signals that the predication of the current rheme also holds for another homologous alternative already given in the previous context (Krifka 1999) (i.e., "the departure"). From the viewpoint of the discourse structure, such an alternative represents the previously managed subtopical coordinate of a main dialogue topic "reservation" and is situated at the same dialogue topic structure level than the current subtopical coordinate ritorno. The introduction of the inactive information within the pero host utterance changes the current topical perspective: specifically, the given subtopical coordinate has to be replaced along with its related subtask. In this context, pero regulates the change of perspective, by introducing a new presentation to the common ground. It conveys an intersubjective meaning, whereby the speaker evidences that the task has not been fully accomplished. As a result, the interlocutor's attention needs to be reoriented toward a new meta-negotiation introducing a new subtopical coordinate in order to fix the instruction for performing a new task. From a structural viewpoint, pero10 cues an overlapping initial boundary for a new subtopic unit and its corresponding task, as well as for a new common ground unit.

3.3. Invece In example (8), turn 4, invece follows positive feedback and opens an elliptical open question. This ellipsis can be resolved by referring to the topic in turn 1 (i.e., voli ['flights']). It marks a deeper discourse boundary than those signalled by ma in the example (1).

20

Francesco Carola (8)(a)C: Vediamo unpo'quali sono [i VOLif. (b) O:S/. Allora, (prima di mezzogiorno}: |t|' {con Γ Alpi Eagles }: |t|2 ci sarebbe un volo ehe parte da Roma alle died mattina. (c)[..·] (d) 0:Ah, va bene. Invece con [la MeriDIAna]¥r? (a) Let's have a look at the flights. (b) Yes. So, before midday, with Alpi Eagles, there is a flight which leaves from Rome at 10. (c)[..·] (d) Ah, Okay. What about with Meridiana?

The host utterance asks for new information (i.e., / νοίϊ) related to given information (i.e., Meridiana), which has to be newly co-activated in speakers' mental states and established as the current task-related topic. Such a context profiles a prototypical negotiation in which an alternative solution to the same task-related issue is discussed: invece operates a change of perspective within the ongoing negotiation by introducing the topical coordinate to be discussed as an alternative to the previous one. Although the dialogue unit (turns 2-3) related to the alternative in the immediately preceding cotext has been grounded through the acknowledgement va bene, the grounding on the main dialogue topic / voli, presented in turn 1 is still open. Consequently, invece is implicated in the speaker's strategy of reorienting the ongoing negotiation, by revising a parameter of the previous alternative solution to the current task-related issue. Similarly, in example (9), invece follows positive feedback and reorients the information flow. However, in the present context invece does not introduce any alternative solution to the same issue in the sense considered so far. The host utterance represents a declarative utterance which opens a new orientation to the common ground. Furthermore, a left-detached contrastive theme packages the discourse entity ritorno, which co-refers to a previously mentioned given information (i.e., per rientrare), but needs to be co-activated currently as opposed to the previously meta-negotiated entity andata ('departure'), which has to be substituted in the common ground. (9)(a)O: {Venerdl}: |t| c'e un treno alle 8.35. (b) C: Benissimo. (c) O: OK. [Per [il riTORnof invece |t|} c' e' un treno alle [quindici]¥ da Roma. Puo* andar benel (d) C:Alle quindicipotrebbe andar bene. (a) On Friday, there is a train at 8.35 a.m. (b) Perfect. (c) Okay. As for the return instead there is a flight from Rome at 3 p.m. Is it ok? (d) At 3 p.m., it could be okay.

Gathering Common Ground

21

The effect of the contrast conveyed by invece is that of providing explicit instruction, aimed at replacing the alternative sub topical coordinate ('the departure')11 with the one which is indicated by the contrastive theme in the host utterance. Both the subtopical coordinates 'departure' and 'return' are subsumed under the same main topic which is still active in the speakers' mental states and still needs to be grounded.

4. Towards a preliminary taxonomy

Relying on the prior observations, a taxonomy of the contrastive markers and their functions can be sketched. It can be argued that ma conveys a corrective contrast (example 6), which arises when, within the same topic, task, and common ground unit, some accessible and active given topical coordinate has to be partially corrected because the given information is not accepted in the common ground by the speaker or it is perceived as being insufficient for the current conversational purposes. Due to such a contrast, the grounding process, which is still viable for the information just presented, is temporarily suspended. Contrastively, invece (example 8) expresses an oppositive contrast, which takes place within a negotiation, in which a polar alternative solution to a same task-related issue is placed at the same level of topicality and subordinated to the same main topic. In this circumstance in which the grounding process is still open and viable, invece indicates that a new local orientation is opened and the inactive information has to be accessed through the common ground. Pero and invece, presented in the examples (7) and (9), respectively, convey a substitutive contrast which occurs within a meta-negotiation; after that a previous topic and task have been grounded. The effect obtained by such a contrast is that the grounding process has to be (re)opened because of both a topic shift and the presentation of a new task. The interpretation of the contrastive markers used in information-oriented dialogue turns out to be sensitive to 1) the interplay of the grounding; 2) the negotiation of alternative solutions to a same task-related issue; and 3) the meta-negotiation of the topic. Contrastive markers form a homogeneous subset of discourse markers which signal the different polarities that the speaker establishes between the information introduced in their host utterance and some piece of information given in the preceding cotext. For this reason, they correlate with different boundaries of the dialogue topical structure (captured consisSuch a coordinate has been grounded through the immediately preceding acknowledgement act benissimo.

22

Francesco Carola

tently with the parameters assumed), and with the speaker's attitude toward the information state in the cognitive context of the common ground. Lastly, they function as crucial change of perspective and reorientation devices employed by the speakers during their coordination activity in order to reciprocally regulate their cognitive states of activation with respect to the information exchanged during the co-construction of conversation. Besides their significant role in anchoring the local host utterance to the global dialogue structure and contributing to discourse coherence, contrastive markers help to constantly monitor the status of the information in the common ground. For this reason, the contrast they convey can be explained as being fully collaborative.

5. Directions for Further Work

A systematic comparison between corpora representative of a wider spectrum of speech typologies is a special desideratum to verify whether and to which extent the interpretation of the contrastive markers is dependent on the particular domain of the data taken into consideration. Furthermore, in order to provide a unified account for other contrastive markers, the framework proposed in this chapter would need to be extended and generalized to other contrastive markers, as well as to the discourse markers of Italian more generally. Lastly, the interrelation of the contrastive markers with other kinds of contrastiveness, such as contrastive themes, still needs to be explored in depth. These are some of the aspects which offer a task for future work.

References Bangerter, Adrian/Clark, Herbert H./Katz, Anita R. (2004): Navigating joint projects in telephone conversations. - Discourse Processes 37, 1-23. Bazzanella, Carla (1995): I segnali discorsivi. - In: L. Renzi, G. Salvi, A. Cardinaletti (eds.): Grande grammatica italiana di consultazione, Vol. Ill, 225-257. Bologna: II Mulino. Bertinetto, P. M./Marconi, D. (1984): Ma in italiano. Parte seconda. Proiezioni diacroniche. -Lingua e Stile 19, 475-509. Carota, Francesca (2004): On some effects of lexical contrast in information seeking dialogues. - In: E. Vallduvi, J. Ginzburg (eds.): Proceedings of the Eighth Workshop on the Semantics and Pragmatics of Dialogue, SEMDIAL, Catalog '04, 148-149. Barcelona: Universidad Pompeu Fabra.

Gathering Common Ground

23

Carota, Francesca (2005): Fondo cognitivo comune e negoziazione deH'infonnazione. Funzionalita dei marcatori discorsivi in italiano [Doctoral Thesis]. Pisa: Universitä degli Studi di Pisa. Chafe, Wallace (1994): Discourse, consciousness, and time. - Chicago/London: University of Chicago Press. Clark, Herbert (1996): Using language. - Cambridge: Cambridge University Press. Cresti, Emanuela (2000): Corpus di italiano parlato. - Firenze: Accademia della Crusca. Fräser, Bruce (1998): Contrastive discourse markers in English. - In: Andreas H. Jucker, Yael Ziv (eds.): Discourse markers. Descriptions and theory, 301-326. Amsterdam: John Benjamins. Ginzburg, Jonathan (1998): Clarifying utterances. - Proceedings of the 2nd Workshop on the Formal Semantics and Pragmatics of Dialogue, 11-30. Enschede: Universiteit Twente, Faculteit Informatica. Gundel, Jeannette/Hedberg, K. Nancy/Zacharsky, Roy (1993): Cognitive status and the form of referring expressions. - Language 69, 274-307. Jaeger, Tim Florian/Oshima, David (2002): Towards a dynamic model of topic-marking. Proceedings of the Workshop Information Structure in Context, 153-167. University of Stuttgart. Krifka, Manfred (1999): Additive particles under stress. Proceedings of SALT 8, 111-128. - Cornell, NY: CLC Publications. König, Ekkehard (1993): Focus particles. - In: Joachim Jacobs, Arnim von Stechow, Wolfgang Sternefeld, Theo Vennemann (eds.): Syntax. Ein internationales Handbuch, 978-987. Berlin: de Gruyter. Larsson, Staffan (2002): Issues under negotiation. - Proceedings of the Third SIGdial Workshop on Discourse and Dialogue, Vol. II, 103-112. Morristown, NJ: Association for Computational Linguistics. Linell, Pirn (1999): Approaching dialogue. - Amsterdam: John Benjamins. Lombardi-Vallauri, Edoardo (2002): La struttura informativa dell'enunciato. - Firenze: La Nuova Italia. Lonzi, Lidia (1995): II sintagma awerbiale. - In: L. Renzi, G. Salvi, A. Cardinaletti (eds.): Grande grammatica italiana di consultazione, 341-412. - Bologna: II Mulino. Prince, Ellen (1981): Toward a taxonomy of given-new information. - In: P. Cole (ed.): Radical pragmatics, 223-255. - New York: Academic Press. Serianni, Luca (2001): Gli italiani scritti. - Firenze: La Nuova Italia. Soria, Claudia/Cattoni, Roldano/Danieli, Morena (2000): ADAM. An architecture for XML-based dialogue annotation on multiple levels. - Proceedings of the First SIGDIAL Workshop on Discourse and Dialogue, Vol. 10, 9-18. Morristown, NJ: Association for Computational Linguistics. Traum, David (1999): Computational models of grounding in collaborative systems. - Working Notes ofAAAI Fall Symposium on Psychological Models of Communication, 124-131. Traum, David/Dillenbourg, Pierre (1998): Towards a normative model of grounding in communication. - Proceedings of the ESSLLI'98 Workshop on Mutual Knowledge, Common Ground and Public Information, 52-55. Saarbruecken, Germany: Universität des Saarlandes.

Karin Aijmer

Dialogue Analysis in a Cross-Linguistic Perspective

1. Introduction

The present chapter is part of study of adverbs of epistemic certainty as a lexical-semantic field carried out with my colleague, Anne-Marie Simon-Vandenbergen (Aijmer and Simon-Vandenbergen, forthcoming). In order to find out more about what words such as surely, certainly, and obviously mean, we used translations into several target languages. In our research on modality (see also Miecznikowski this volume), we noticed that clearly and obviously were similar in meaning, but not quite the same when we compared their translations into Swedish, Dutch, German, and French. Some of these differences were very subtle and involved frequencies. For example, clearly was more often translated as a manner adverbial and was less often rendered by a word meaning naturally than the related obviously. On the other hand, when we compared the translations of certainly and obviously, they had hardly any translations in common. One would expect certainly and surely to be closely related, but in fact they share only a few translations (Aijmer 2002). From such observations, it was a natural step to also become interested in crosslinguistic cognates and their functional similarities and differences. One would, for example, expect French evidemment and English evidently to mean the same thing; or clearly and clairement to give another French and an English example. However, French clairement is a manner adverbial while clearly is mainly a sentence adverbial and French evidemment refers to evidence ('from perceptual evidence') while evidently in present-day English has been described with reference to conclusion (He is evidently right Ί conclude that he is right'); strong subjective epistemic inviting the inference of some concession or doubt on the speaker's part (Traugott 1989). The topic of this chapter is the cognate adverbs English surely, Swedish sakert, and German sicher. Intuitively, I could say that the Swedish word did not have the same meaning as surely. It therefore seemed to be an interesting topic to study the meanings of the three words and to try to explain similarities and differences. In Norwegian, Danish, and Dutch, we also find corresponding lexical cognates (cf. also French suremeni), but I have restricted myself to languages where translation data have been available to me. It is perhaps not immediately obvious that the adverbs are etymological cognates since surely is

26

Karin Aijmer

borrowed from Old French. Compare Trask's definition of cognate as "narrowly, and most usually, one of two or more words or morphemes which are directly descended from a single ancestral form in the single common ancestor of the languages in which the words or morphemes are found, with no borrowing" (Trask 2000: 62). All three words come from Latin securus which means without care, although this meaning is no longer present in the adverb. The method I will use is to compare translations for the three words. Translations give a mirror image of the meanings of the words in another language and can therefore be used to confirm hypotheses about what the adverbs mean and to give a detailed picture of meanings and functions. The method, which is itself of fairly old date (cf. Viberg 2002: 121 and the references there to Wandruszka 1969 and Bally 1950), has experienced a revival because of the availability of computerized bilingual and multilingual corpora. An important part of the study is to propose a reinterpretation of modal adverbs. The analysis which is proposed is based on an interpretation of Bakhtin's ideas ofdialogism or heteroglossia. In a heteroglossic perspective, texts address "alternative realities as expressed in previous texts" and they anticipate future realities. As a consequence, "every meaning within a text occurs in a social context there a number or alternative or contrary meanings could have been made" (White 2000: 11). It follows that the meaning of an adverb such as surely may have little or no connection with a degree of certainty or commitment. It is used instead as a modal resource by means of which speakers or writers express and negotiate interpersonal attitudes (cf. White 2000/2003).

2. Methods to Study Adverbs of Certainty

Adverbs of certainty, such as surely, are typically context-bound and their meanings are traditionally analyzed from a syntagmatic perspective; they are described in terms of the syntactic environment, in particular position in the utterance and collocation (i.e., the company they keep in the utterance). Let us see what we can learn from this perspective. To begin with, we notice that surely occurs in different positions in the utterance: (l)"Surely," Andrew observed, "there can't be many families left that control big drug companies." (AH 1) (2)"He surely would have discovered it if farming was his business, but it wasn't". (SKI)

Dialogue Analysis in a Cross-Linguistic Perspective

27

The same adverb is used, but in different positions. Does surely have the same meaning initially and medially? In Quirk, Greenbaum, Leech, and Svartvik (1985), a distinction is made between surely as disjunct (i.e., initial position with the meaning of conviction) and subjunct (i.e., position close to the verb and emphatic meaning). The type of subject in the sentence with surely can also make a difference to the meaning of the adverb (cf. Downing 2001). Surely has a different meaning in (3) and (4): (3) You can surely make a wonderful biography, Simon (RDA1) (4)It surely sounded like him from what you described (SGI)

In (3), the subject refers to the hearer. Surely expresses more uncertainty than in (4) where the subject is it. There is not a single answer to the question of how we should study the meanings of words like surely which seem to be sensitive to the context for their interpretation. For example, Greenbaum (1969) used informant tests to investigate the meaning of adverbs and their relation to each other. The results of informant testing showed that certainly, indeed, and surely were felt to be more similar in meaning hi different positions than actually and really. I am sympathetic to a multiplicity of methods to study the adverbs of certainty. Distributional properties (such as the subject and position) are, for instance, important cues to the interpretation of adverbs. Translations are another method helping us to analyze the meaning of words which are multifunctional and difficult to describe. Translators have to make explicit a meaning found in the source language in the translation into the target language. Translations are therefore important empirical evidence about meanings and complement the picture that we get from the syntagmatic analysis of factors such as position. In a number of papers, Dyvik (1998) has suggested a procedure using translation corpora which makes it possible to define semantic properties of words such as synonymy, ambiguity and vagueness. The method is perhaps easiest to understand when words are ambiguous. Genuine ambiguity such as the distinction between two meanings of bank (bankι: 'place where you put your savings,' and bank2 with reference to the river bank} is accidental and not duplicated in the translations into one or more other languages. We would, for instance, expect bank to be translated in at least two different ways into another language. Also vague or context-bound words are interesting to study in a translation perspective although the picture we get is less clear. If, for instance, surely changes its meaning in a particular context, this would be reflected hi the translation. As a consequence, we get a rich material from translations indicating not only meanings, but implicatures, presuppositions, and rhetorical effects which depend on the context. The different meanings of a word

28

Karin A ijmer

can be exhibited as a translation paradigm or translation image showing meanings and functions as translation alternatives. The method will be further explored in the following discussion. We are lucky to have corpora of translations into several languages which can be used to study the semantic properties of words contrastively. The translation paradigms which constitute my data come from the English-Swedish Parallel Corpus (Altenberg and Aijmer 2000) and the Oslo Multilingual Corpus (the German-English data).' The English-Swedish Parallel corpus consists of about 4 million words of fiction and non-fiction texts. The material from the Oslo Multilingual Corpus consists of 12 fiction texts of 10,000-15,000 words each. There are many possibilities to use the corpora for contrastive studies which include looking at both translations and sources. The study undertaken here is restricted in scope. I will look at the translations of English surely, Swedish säkert, and German sicher in order to compare the meanings and functions of the adverbs.

2.1. English Surely My starting-point will be what the translations show about the meaning of surely. Table 1 illustrates a paradigm resulting from translations. The purpose is to describe translation as a method and I will not attempt to describe all the functions of surely exhibited in table 1. The meaning of surely is complex and different meanings and functions can be discovered by looking at the translations. One meaning is exemplified by Swedish definitivt, säkert, and utan tvivel ('without a doubt'). Other meanings are indicated by naturligtvis (connection with expectation), mäste (deduction), or väl (appeal to the hearer). Let us look at the examples in more detail to see how the meanings shown in the translations can be explained. I will focus on examples where surely does not indicate a degree of certainty only (i.e., commitment to the proposition), but has other functions. (5)There is no doubt that this reform of structural and Cohesion Funds represents the most significant changes in the last ten years with a view to meeting the challenges of enlargement and EMU. Enlargement will surely offer opportunities to our regions but we will want to ensure that the proposals presented by the Commission match our objectives and priorities as socialists, namely to continue the Delors legacy of promoting jobs.... (EMCC1) Det räder inget tvivel om art denna reform av Struktur- och sammanhällningsfondema representerar de mest betydelsefulla forändringarna de senaste tio ären med sikte pä att möta de utmaningar som utvidgningen av EMU innebär. En utvidgning kommer utan tvivel att erbjuda möjligheter for vära regioner, men vi vill se till att de av kommissionen presenterade förslagen See http://www.hf.uio.no/germari/sprik/english/corpus.shtrnl.

Dialogue Analysis in a Cross-Linguistic Perspective

29

stammer overens med vära mälsättningar och prioriteringar som socialister, nämligen att fora vidare arvet frän Delors att verka for att skapa arbetstillfällen.... Type säkert visst ('surely not') Val+? ('don't you think') mäste ('must') men väl ('but surely') naturligtvis (Of course') väl ändä ('surely anyhow') nog ('probably') otvivelaktigt väl ('undoubtedly surely') sake« +? men+val+? ('but surely?') mäste naturligtvis ('must of course') mäste väl ('must surely') säkert (manner adverbial) definitivt ('definitely') menar du att ('do you mean that') men nog ('but probably') nog... väl ('probably... surely') nog väl ändä ('probably surely anyhow') nog+? ('probably'?) ju ('as you know') verkligen ('really') utan rvivel ('without any doubt') i alia fall ('anyhow') högst (fb'rvänande) 'highly surprising' dock ('anyhow') 0 Table 1. Translations of English surely into Swedish

Number 12 3 3 3 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 5

The translator has used a strong expression of certainty, utan tvivel ('without a doubt'). The translation indicates that surely is not in the first place epistemic. Surely is basically a fighting word, a description I borrow from Downing (2001) to refer to its rhetorical functions. In (5), the writer agrees with a preceding position rather than challenges it. However, in a wider context we can see that agreement is a stepping-stone for presenting a stronger argument. In (6), the translation gives a clear indication of the rhetorical potential of surely. The Swedish modal particley« in the translation conveys not only certainty, but carries the presupposition that something is generally known (i.e., should be known by the hearer). As in the previous example, surely has got little to do with certainty or commitment to the truth of the proposition.

30

Karin Aijmer (6)1 agree with Mr Liese that the seven-year authorisation period is unsatisfactorily arbitrary. If there is doubt about any genetically modified material, then surely it should not be permitted for release in the first place. There are, however, certain amendments - for example Amendment No 68 - which call for more caution and which we, British Conservatives, want to support. (EJAC1) Jag haller med Liese om att en tillständsperiod pä sju är är otillfredsställande godtyckligt. Om det fmns nägra tvivel kring ett genetiskt modifierat material sä borde detju inte tillätas användas over huvud taget. Det fmns emellertid vissa ändringsförslag - till exempel ändringsförslag 68 - som manar till mer försiktighet och som vi, de brittiska konservativa, vill ge värt stöd.

As an epistemic adverb with a following should, surely seems to have the meaning deduction (conclusion). In addition, surely has a dialogic dimension which is made clear by the translation: The writer aligns himself with a position which is taken for granted rather than being at issue. Surely is rhetorically useful in examples introduced by the issue is or the answer is which present something as new, important information. The translations show that surely is used to convey the meaning "as must, of course, be the case", "as must be the case in the order of things." In (7), the context is a but-clause and surely is also used to emphasize the contrast ("in others a consideration of the environmental impact may be lacking"). (7) We are very unhappy with this directive because we recognise that the Member States all have their own methods of town and country planning. In some cases, the consideration of the environmental impact is already very complex. In others it may be lacking, but the answer is surely to make the 1985 directive universally applicable. Mr Gahrton takes us a long way (EJAC1) Vi är mycket missnojda med detta direktiv eftersom vi vet att medlemsstaterna alia har sina egna metoder for fysisk planering. I vissa fall är beaktandet av de miljömässiga effektema redan nu mycket komplicerad. I andra fall kan det vara bristfälligt, men lösnlngen mäste ändä vara att göra 1985 ärs direktiv universellt tillämpbart. Garthon for oss en läng vag.

Example (8) is similar. (8) The issue here is that if we are to have European standards then surely we must be entitled to expect that these standards are accepted by the participating Member States. (EHUD1) Frägan här är att om vi skall ha gemensamma standarder, mäste vi naturligtvis ha ratten att förvänta oss att dessa standarder accepteras av de deltagande medlemsstaterna. (lit. 'must we naturally')

The example is taken from a debate in the European Union. The translation (mäste naturligtvis: 'must of course') shows that surely is used rhetorically to strengthen the argument that 'we must be entitled to expect that these standards are accepted by the members of the

Dialogue Analysis in a Cross-Linguistic Perspective

31

union.' However, this claim is far from self-evident or certain. The speaker uses the adverb because he has an advantage to gain if the recipient accepts the claim. Example (9) illustrates a frequent use of surely. (9) On three occasions this year I have been refused entry into Terminal D on my laissez-passer. I am told that it is not an official document. On the last occasion, last Monday, I was actually held aside and not allowed to pass until I produced my national passport. Surely the French realise that we have a Union and that we have a laissez-passer which gives us the right to travel anywhere within the 15 nations. (ECOR1) Tre ganger detta ar har jag blivit vägrad tillträde till Terminal D med min laissez-passer. Man har sagt mig art det inte är nägot officiellt dokument. Sista gangen, fb'rra mlndagen, blev jag faktiskt ford till sidan och fick inte passera innan jag visade upp mitt nationella pass. Nog vet väl fransmännen art vi har en union och att vi har en laissez-passer som ger oss ratten art resa var vi vill i de 15 länderna.

In (9), the factive verb realise carries the presupposition that we have a Union and we have a laissez-passer which gives us the right to travel. Surely signals a discrepancy between what is known (or presupposed to be true) - namely that there is a laissez-passer giving people the right to travel between countries and what is referred to in the preceding text (i.e., the speaker was held up by the French). The translation (nog vet vol. 'certainly you agree') captures the writer's reaction that there is a clash with expectations and that something is unreasonable, surprising, etc., as well as the appeal to the hearer/reader. In example (10), a similar translation (nog horde väl) has been used. The translation helps us to see that the speaker's position implies that there is another preferable alternative to what actually happened. (10) "Thank you" said the Major; "I'm glad we understand each other. If you would ask MaryJacobine to be at home tomorrow at eleven o'clock, I shall have the honour of calling upon the ladies." SURELY THE SENATOR might have argued a little more, said the Daimon Maimas. He buckled under very quickly, wouldn't you say? (RDA1) "Jag är glad att vi förstär varandra. Om ni vill ha godheten att be Mary-Jacobine vara hemma i morgon klockan elva, ska jag med glädje besöka damerna." - NOG BORDE VÄL senatorn ha protesterat lite mer, sa daimon Maimas. Han gav sig väldigt lätt, tycker du inte det?

To sum up, the translations give a good picture of how surely is used by the writer or speaker to establish a relationship with the preceding text or with the reader. We can, of course, discuss the meanings of the adverbs without taking into account their translations; however, translations help us to see when adverbs have different, sometimes incompatible meanings. In the dialogical framework proposed here, surely is a resource making it possible for the speaker/writer to actively engage in the interaction by taking up positions not

Karin Aijmer

32

only to the text, but also to alternative propositions (e.g., a preceding proposition, an assumption, what is presupposed or known to be true, or an alternative preferable proposition). Clearly we can expect the same framework to motivate the development of different meanings in other languages. An interesting question is therefore: To what extent do we find similar functional developments in Swedish or German?

2.2. Swedish Säkert Type certainly probably No doubt be sure to must I'm sure surely undoubtedly sure with certainty doubtless be sure that be certain to probably be bound to be certain of all the same definitely For sure I expect obviously necessarily be bound to usually it goes without saying there is a strong likelihood that safely 0 other Table 2. Translations of Swedish säkert into English

Number 23 7 5 4 4 4 3 3 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 4 2

A quick look at the translations from Swedish to English shows us a different picture for säkert. As might be expected from the analysis of surely, säkert also has an indeterminate meaning with several English correspondents in the corpus.

Dialogue Analysis in a Cross-Linguistic Perspective

33

One example was a manner adverbial indicating that säkert has not been completely grammaticalized: (11) Men nu lag de säkert i innerfickan. (LG l) But now he had the fuses safely in his inside pocket.

In (12), säkert is ambivalent between two interpretations. It can be both a manner adverbial ('in a carefree manner') and a sentence adverbial ('there was no doubt about it'). (12)... att han längsamt men säkert hade supit ner sig under arbetets gang. (LG l) .... that he had slowly but surely got himself drunk in the course of the work.

The most frequent translation was certainly or a synonym (cf. translations such as doubtless, no doubt, undoubtedly, be bound to, necessarily, must, definitely, there is a strong likelihood, or it goes without saying). Notice that the cognate surely occurred in the translation only three times. There must be a reason for using säkert other than to express how reliable it is. In example (13), which is taken from a debate on climate changes, säkert was translated as undoubtedly. As shown by the translation, säkert is used strategically to suggest that the knowledge is objective (i.e., shared knowledge): (13) Klimatforändringar (27/01/1999) Herr ordforande, fru kommissionär! Av alia utmaningar som vi star infor pä miljöomrädet är säkert klimatfrägan den största och mest komplicerade. Den berör ju alia jordens länder pä ett pätagligt satt. Utsläppen hotar oss inte pä kort sikt, men utgör ett mycket allvarligt hot pä läng sikt. (EVIR1) Climate change (27/01/1999) Mr. President, Commissioner, of all the environmental challenges facing us, climate change is undoubtedly the most complex. Every country in the world is tangibly affected. Emissions do not pose a threat in the short term, but in the long term they are infinitely more serious.

In (14), the translator has chosen 'there is a strong likelihood' to express objectivity (i.e., commitment to the proposition shared by the language community): (14) Vi mäste slä vakt om GJP sä länge världsmarknaden inte fungerar. Det dröjer säkert ätskilliga är innan vi har en fungerande världsmarknad att lita till. Jag vill ocksä instämma i vad Fischler sade, att vi inte skall amerikanisera det europeiska jordbruket. (EOLS1)

34

Karin Aijmer Until there is a fully functioning global market, the CAP needs to be safeguarded. And there is a strong likelihood that it will be some time before the market functions reliably.

The translation with 'bound to be' in (15) suggests that the speaker counters some potential resistance: (15) "Vad är det som har hänt?" frägade Robert Akerblom när de tog i hand. "Förmodligen ingenting," sä Wallander. "Allt har säkert sin naturliga förklaring." (HM1) "What happened, do you think?" asked Robert Akerblom as they shook hands. "Probably nothing at all," said Wallander. "There's bound to be a natural explanation."

When säkert is translated by 'I'm sure,' it is more subjective. In (16) and (17), there is an indication that something will not work out as expected. The speaker, therefore, wants to comfort or reassure the hearer. (16) "Det skall säkert gä bra," sa Martin Beck. (SW1) "I'm sure everything will work out fine," said Martin Beck. (17)"Inte for att du är man - gor det inte sä enkelt for dig - men du är säkert en mycket konventionell människa i alia fall," säger jag. (MSI) "Not because you're a man. You mustn't simplify in this way. But I'm sure you're a very conventional person nonetheless."

The translation data suggest that it is difficult to draw a sharp boundary between certainty and uncertainty. We find both elements indicating a high degree of certainty and elements conventionally associated with uncertainty. As Halliday (1994: 362) has pointed out, there is a fundamental paradox in the expression of certainty: "we only say we are certain when we are not." As an adverb, säkert also means 'probably' (i.e., weak probability rather than strong probability), as in (18). (18) "Det är nog inte mycket jag kan göra," sa Wallander. "Hon har säkert redan försvunnit. I vilket fall är det ingenting att oroa sig for." (HM1) "There's not really much I can do," said Wallander. "She's probably gone by now. And in that case, there's nothing to worry about."

Dialogue Analysis in a Cross-Linguistic Perspective

35

The translation with 'probably' suggests that the speaker is uncertain and does not want to commit himself further. To sum up, Swedish säkert is also a manner adverbial. It has a fuzzy epistemic meaning ranging between high probability (i.e., subjective certainty) and low probability. The meaning of certainty could be strengthened in certain environments where the speaker takes up a strong stance in order to sound more objective. It can also be more subjective with speech act meanings, such as comforting or reassurance.

2.3. German Sicher Type must ('ganz sicher') no doubt with tolerable certainty ('ziemlich sicher') certainly ('ganz sicher') But it is quite certain that But certainly ('aber sicher') could possibly (negative clause) Tag question it was surely what is certain is that after all he was pretty sure ofthat really securely (manner)

Number 2

0

2

Table 3. Translations of German sicher into English

German sicher is gradable (scalar) as in ganz sicher, ziemlich sicher (cf. *quite surely). Sicher is also a manner adverbial, as in (19). (19) Sie schwebt hinter meinem Rücken vorbei, tippt mir flüchtig auf die Schulter, küßt Mutter, küßt Vater und landet sicher auf einem Stuhl. (JUB l) She floats past my back, tapping me lightly on the shoulder, kisses her mother, kisses her father, and lands securely on a chair.

Unlike the Swedish cognate, it is used to express doubt and confirmation in (20), as indicated by the translation with a tag question.

36

Karin Aijmer (20) Von Frau Varela erfuhr sie, daß man eine neue Werft ihrer Bestimmung übergeben hatte, daß dem verstorbenen, liebenswerten Kanonikus ein jüngerer, schroffer, gefolgt war, daß das Kasino wegen Baufälligkeit und das sozialistische Volksheim wegen staatsgefährdender Agitation geschlossen worden waren und daß eine gewisse Frau Pelayo, sie erinnerte sich doch sicherl, plötzlich verstorben war. (ERH1) Senora Varela told her that a new shipyard had been approved; that the late, beloved canon had been succeeded by a younger, gruffer one; that the municipal clubhouse and the socialist adult-education center had been closed, the one owing to dilapidation, the other to agitation against the authority of the state; and that a certain Senora Pelayo - you remember her, don't youl - had died a sudden death.

In (21), the translation shows (aber) sicher with the meaning high certainty: The speaker concedes an earlier argument in order to strengthen an opposing viewpoint (cf. 'what is certain is that...,' 'but it is quite certain that...,' 'it was surely...' as translations of certainty). (21) Vielleicht hielt sie sich für krank, vielleicht war sie es, aber sicher war sie auf orientalische Art sehr faul, und unter dem teuflisch lebendigen Großvater hatte sie bestimmt zu leiden. (EC1) Perhaps she thought she was ill, perhaps she really was, but she was certainly lazy in an Oriental way, and she must have suffered under Grandfather, who was fiendishly lively.

To sum up, sicher can be a manner adverbial. It occurs in the meaning of certainty to take up a stance to a position that the speaker does not agree with. It is used with intersubjective meaning to appeal to the hearer. Moreover, it is a gradable adverb which often appears with degree adverbials (ganz, ziemlich).

3. Grammaticalization

When we discuss epistemic modality, both intralingually and interlingually, grammaticalization comes into the picture. Grammaticalization explains that lexical material develops into grammatical categories and the processes whereby this takes place. Hopper and Traugott (1993: 1-2) explain grammaticalization as follows: "it refers to that part of the study of language that focuses on how grammatical form and constructions arise, how they are used, and how they shape the language." They also use the term in a second sense focusing on the processes resulting in grammaticalization: "The term 'grammaticalization' also refers to the actual phenomena of language, that the framework of grammaticalization

Dialogue Analysis in a Cross-Linguistic Perspective

37

seeks to address, most especially the processes whereby items become more grammatical through time." The attention to the lexical sources for grammaticalization has also led to an interest in semantic change and in the semantic paths accompanying the grammatical processes (Traugott and Dasher 2002). It is clear that not only grammatical categories, but also lexical categories exhibit patterns of meaning which can be explained as a result of semantic change and grammaticalization. The methods of studying grammaticalization are mainly diachronic. However, translations can also give a good picture of the different meanings and functions involved in grammaticalization. Let me give an example; if clearly is translated both as a manner adverbial and a sentence adverb, we can explain its meaning as a sentence adverb as the result of grammaticalization. Studies in grammaticalization focusing on semantic change have mainly dealt with one particular language. I would like to suggest that we can get a better picture of semantic change and the regularities involved in grammaticalization by looking at etymological cognates and their translations in one or more than one language. The study suggests that intersubjectification, the fact that the speaker engages in a dialogue with other positions (i.e., other voices) is an important process in the development of new meanings which can be grammaticalized. In a heteroglossic perspective, as advocated for example by White (2003), speakers take up a diverging or converging stance not only to the text but to beliefs or assumptions which can be attributed to the hearer, an imagined reader, or to people in general. The heteroglossic approach explains that surely and its cognates in other languages can have a number of meanings and functions which are not accounted for as modal only. We have seen that the adverbs can be used for strategizing; surely can convey the implication that hearer/reader already knows something. By means of surely, the speaker or writer can also convey an attitude to the preceding context, such as surprise that something has taken place and an appeal to the hearer to confirm something. Swedish säkert has not developed dialogic uses to the same extent as English surely, and German sicher has functions with no correspondence in Swedish (e.g., appeal to the hearer). We would explain this by saying that grammaticalization has developed more slowly or that the process has been arrested because of competing elements in the language such as modal particles.

References Aijmer, Karin (2002): Modal adverbs of certainty and uncertainty. - In: Stig Johansson, Hilde Hasselgärd, Cathrine Fabricius-Hansen, Bergljot Behrens (eds.): Information structure in a crosslinguistic perspective, 97-112. Amsterdam: Rodopi.

38

Karin Aijmer

Aijmer, Karin/Simon-Vandenbergen, Anne-Marie (forthcoming): - A study of English adverbs. The semantic field of modal certainty. Altenberg, Bengt/Aijmer, Karin (2000): The English-Swedish Parallel Corpus: A resource for contrastive research and translation studies. - In: Marianne Hundt, Christian Mair (eds.): Corpus linguistics and linguistic theory. Papers from the 20th International Conference on English Language Research on Computerized Corpora (ICAME 20), Freiburg im Breisgau, 1999, 15-33. Amsterdam, Philadelphia: Rodopi. Bakhtin, Mikhail (1981): The dialogical imagination. Austin, TX: University of Texas Press. Bally, Charles (1950): - Linguistique generate et linguistique fran9aise. 3e ed. Berne: A. Francke. Chafe, Wallace (1986). Evidentiality in English conversation and academic writing. - In: Wallace Chafe, Johanna Nichols (eds.): Evidentiality. The linguistic coding of epistemology, 261-272. Norwood, NJ: Ablex. Downing, Angela (2001): 'Surely you knew!' Surely as a marker of evidentiality and stance. -Functions of Language 8/2, 253-285. Dyvik, Helge (1998): A translational basis for semantics. - In: Signe Oksefjell, Stig Johansson (eds.): Corpora and cross-linguistic research. Theory, method, and case studies, 51-86. Amsterdam: Rodopi. Dyvik, Helge (2004): Translations as semantic mirrors. From parallel corpus to Wordnet. - In: Karin Aijmer, Bengt Altenberg (eds.): Advances in corpus linguistics. Papers from the 23rd International Conference on English Language Research on Computerized Corpora (ICAME 23), Göteborg 22-26 May, 2002, 311-326. Amsterdam, Philadelphia: Rodopi. Greenbaum, Sidney (1969): Studies in adverbial usage. - London: Longman. Halliday, Michael A. K. (1994): An introduction to functional grammar (2nd ed.). - London: Arnold. Hopper, Paul J./Traugott, Elizabeth Closs (1993): Grammaticalization. - Cambridge: Cambridge University Press. Quirk, Randolph/Greenbaum, Sidney/Leech, Geoffrey/Svartvik, Jan (1985): A comprehensive grammar of the English language. - London: Longman. Trask, Robert Lawrence (2000): The dictionary of historical and comparative linguistics. - Edinburgh: Edinburgh University Press. Traugott, Elizabeth Closs (1989): On the rise of epistemic meanings in English: An example of subjectification in semantic change. -Language 65, 31-55. Traugott, Elizabeth Closs and Dasher, Richard B. (2002): Regularity in semantic change. - Cambridge: Cambridge University Press. Viberg, Äke (2002): Polysemy and disambiguation cues across languages: The case of Swedish fa and English get. - In: Bengt Altenberg, Sylviane Granger (eds.): Lexis in contrast. Corpus-based approaches, 119-150. Amsterdam, Philadelphia: John Benjamins. Wandruszka, Mario (1969): Sprachen. Vergleichbar und unvergleichlich. - München: Piper. White, Peter (2000): Dialogue and inter-subjectivity. Reinterpreting the semantics of modality and hedging. - In Malcolm Coulthard, Janet Cotterill, Francis Rock (eds.): Dialogue analysis VII. Working with dialogue, 67-80. Tübingen: Max Niemeyer Verlag. White, Peter (2003): Beyond modality and hedging. A dialogic view of the language of intersubjective stance. - Text 23/2, 259-84.

Joanna Szerszunowicz

A Comparative Analysis of Small Talk in English and Polish

1. Introduction

The intercultural perspective of research in the field of linguistics enables researchers to conduct a simultaneous analysis of numerous phenomena viewed both as cross-linguistic and cross-cultural ones. For instance, the studies on the genres of the spoken language are closely connected with the research on the culture of a given ethnic group, since the extralinguistic factors influence the genres functioning in the spoken language. While presenting the communities of language users, Kramsch emphasizes a rule important for the analysis of spoken genres,1 commenting on the differences observed in particular communities. Not only the grammatical, lexical, and phonological features of their language (for example, teenage talks, professional jargon, political rhetoric) differentiate them from others, but also the topics they choose to talk about, the way they present information, the style with which they interact, in other words, their discourse accent. (Kramsch 2000: 7)

According to Duszak (1998), the typology of speech can be presented through discoveries and descriptions of particular discourse forms performing special functions and connected with certain historical, cultural and situational conditions. The classification consists of a number of genres (i.e., types of communication events resulting from the co-ocurrence of specific characteristics of the text and the context). The dominating communication aim of a given genre inflicts restrictions on the contents and the form (Swales 1990). Therefore, genres, such as the joke or the complaint, reflect the ethnic culture and the social structure of a given community of language users as well as the ideology of the generation and political changes (Duszak 1998; Saville-Troike 1982). Therefore, significant differences on the register of the genres are observed in a contrastive analysis. First of all, some genres of one language may not exist in another one. For instance, the Hebrew speaking dugri does not function in the Polish culture (Katriel 1986). Kramsch exemplifies the thesis with a comparison of responses to a compliment in two cultures. In the French culture there is a tendency to minimize the compliment (a typical response: Oh really), while in the American culture a compliment is acknowledged (a typical response: Thank you).

40

Joanna Szerszunowicz

Second, a genre existent in one language can be found in another, but in a form modified by cultural factors. Therefore, the semantic formulas of respective genres differ significantly, as it is the case with the Polish kawal and the English joke (Wierzbicka 1999). It should be stressed that so far the term genre has not been common in the description of the spoken variety of the Polish language.2 Moreover, small talk has not been researched in Poland, either in the cultural or in the linguistic perspective. Although the genre is so frequent in everyday contacts, there is a tendency to disregard small talk as a research problem. The collection of papers on the conversation recently published in Poland (Kita 2003) does not contain any comments on small talk. Moreover, there is a tendency to disregard small talk as a genre in the classifications by Polish linguists, for instance Wilkon (2003: 51) lists the genres of speech as follows: (1) natural (colloquial) speech (chat, story, quarrel, gossip, joke, lament); (2) careful speech (debate, flirt, conversation, anecdote, maxim, interview); (3) official speech (address, conference, commentary, negotiations), not mentioning small talk. Therefore, a question arises whether the small talk genre exists in the Polish language and culture. The term itself is hardly translatable into Polish, since the equivalent of small talk found in a bilingual dictionary (i.e., rozmowa towarzyska 'informal conversation') has different connotations than the English small talk.3 Yet, it can be assumed that the communication contexts in the Polish culture require a script for a Polish small talk genre. However, it can be concluded that this genre may not be crystallized enough to be distinguished in the way it is singled out in the British or American cultures.

2. Method

In order to analyze the small talk genre in a contrastive perspective, two corpora of small talk texts (100 texts each) were collected. Two groups of students participated in the ex periment conducted at the University of Biarystok in 2003. One of the groups was cornSo far few Polish genres of speech have been described. In Polish linguistics, the term genre of speech was used first by Furdal (1982) and Wierzbicka (1983). In the recently published academic course books and dictionaries the term is present in most cases (Gajda 2001; Urbanczyk and Kucala 1999; Wolanska 2003). It is also worth mentioning a recent monograph on the complaint (Wyrwas 2002). The dictionary consulted was Wielki slownik angielsko-polski PWN-Oxford English-Polish Dictionary (2002). Warsaw: Wydawnictwo Naukowe PWN. The Polish equivalent rozmowa towarzyska may be used to refer to a much more formal kind of discourse than the English small talk.

A Comparative Anaylsis of Small Talk

41

posed of third-year Polish philology students, the other one consisted of students of English philology. The students of Polish philology do a one-year course of Practical Stylistics and they are familiar with the term genre of speech, the students of English philology were provided with handouts containing a lecture on speech genres. The term small talk was explained to all participants who were informed about the aim and method of the experiment. The participants were asked to record and then prepare a script of a small talk conversation they had, specifying the subject and marking its importance on a ten-point scale. The students were also asked to list the topics of small talk conversations, which they consider to be typical. Moreover, they answered some questions regarding the small talk genre. The key issue was to observe how the genre differs as a result of cultural filter imposed by the choice of language and linguistic convention in each culture. The group of Polish philology students conducted their small talk conversations in Polish, while the group of English philology did theirs in English.4

3. Analysis

The comparison of small talk dialogues produced by the two groups of students renders it possible to observe numerous similarities and some differences, since the English philology students, aware of the genre convention in English, modeled their dialogues on the English pattern, while the Polish philology students conducted their dialogues according to the rules of their mother tongue. Therefore, in the case of students of English, the cultural filter shaped their perception of the small talk as a genre. It should be stressed that the students of English philology developed an awareness of cultural and social conventions, which influenced their command of the genres of English. As to the structure, the small talk discourse in Polish and English does not differ significantly. On the basis of the corpora of collected scripts, the minimal pattern of the genre, which can be extended with optional elements, includes the following components:

The students are supposed to have a sufficient awareness of the small talk in English due to their stays abroad and everyday contacts with American and British native speakers who work at the University of Biafystok. Most of the experiment participants have been to an English-speaking country at least once in their life (in total: 87%; the USA - 37%, the UK - 41%, Canada - 6%, Australia - 3%).

42

Joanna Szerszunowicz (1) (2) (3) (4) (5) (6)

Greetings; Asking about the interlocutors' feelings; Answer. Asking about the other person's feelings; Answer; Farewell; Farewell.

In the case of small talk in the English culture, the answers to the questions regarding the interlocutor's health are supposed to be positive, while in the Polish culture the answers do not always come in a fossilized form, since the truth value of the answers depends greatly on the relation between the speakers. The smaller the distance between the interlocutors, the greater the likelihood of hearing a negative response (e.g., mentioning ailments, etc.), which in fact is considered to be more sincere (Jakubowska 1998). In more formal contacts, routine formulae are interpreted at a non-literal level, so their truth value is largely irrelevant (Saville-Troike 1982). Apart from the initiators, other formulaic expressions appearing in small talk are preclosing signals, which are common in both corpora (Polish corpus: 68; English corpus: 74). The invitation-like forms, functioning as terminators, are not so numerous (Polish corpus: 32; English corpus: 18) and can be described as utterly formulaic, since their function is to introduce good intentions into the discourse. The terminators used in the small talk texts analyzed are not confined to this type of verbal interaction and they appear in other genres of speech. The most significant difference observed in the corpora analyzed regards the choice of topics chosen by the interlocutors. The vast majority of Polish philology students decided on more serious topics (e.g., politics, health, economy) than the English philology students (e.g., television programs, music, weather). The topics chosen by the students of Polish philology cannot be treated as appropriate for small talk conversation complying with the English small talk convention. It should be emphasized that even though the topics introduced into the discourse by the Polish philology students can be considered atypical of small talk, it is not only the subject to be analyzed, but the speaker's and the interlocutor's perception of the subject. The topic of the conversation may sound very serious; however, the seriousness of the topic can be superficial since the subject is treated as a matter of little importance. For example, health is an issue of great importance, yet the analysis shows that the students of Polish philology tend to discuss minor ailments, such as headaches (e.g., Fine, but I had a splitting headache yesterday), tiredness (e.g., (I) didn't sleep at all. 1 really feel dizzy. Good for nothing and I have to do the test) or colds (e.g., From the very morning I have been shivering, (it) must be the rain that caught us yesterday. (I) hope it 's not flu, just a cold).

A Comparative Anaylsis of Small Talk

topic Polish current affairs, politics, etc. 39% European Union 27% health 7% films, books, etc. 7% family, friends, etc. 12% money 27% personal problems 54% pets 6% school, work 35% shopping 7% sports 9% weather 18% weekend 8% Table 1. Topics of Small Talk Conversations Conducted by Participants

43

English 3% 0% 4% 9% 3% 2% 6% 7% 24% 3% 6% 26% 11%

Therefore, classifying topics into serious and light does not serve the purpose of the analysis, since in the Polish culture certain topics, inappropriate for small talk by the British/American standard, are viewed as subjects of little importance. Therefore, if such items as the presentation of current affairs, politics and economy are viewed as uncontroversial, therefore neutral for both interlocutors, they are acceptable by the Polish standards of the small talk genre. In the experiment, both groups were asked to mark the degree of the importance of the topic of a small talk dialogue on a ten-point scale. It is noteworthy that although the registers of topic differed in both groups, the results of the participants' choices regarding the importance of the topic discussed were similar. In both groups the topics chosen were not considered to be personally important for the interlocutors, so the average degree of personal importance of the topics does not differ significantly. As to the differences in the choice of the topics, they result from a different sense of politeness and directness in Polish and English/American cultures. Moreover, the mentality of the nation is reflected in the realizations of the genre. The Polish tendency to complain and be self-deprecating can be noticed in the pessimistic answers to the how-are-you questions and in the elements of criticism towards numerous phenomena. The British/American model of small talk genre reflects the can-do attitude, since it is more optimistic and cheerful. The observations leads to the conclusion that the genre, as it is the case with small talk, are complex, since they, rooted in a given cultural context, they change in time and space. Taking the above facts into account, it should be stressed that a change was observed in the general tendency of the Polish way of answering the question referring to the situation or the mood of the interlocutor, since there is a significant increase in the number of people who try to sound optimistic. The tendency to introduce elements of positive feelings into the discourse may result from the influence of the western structure of the small talk genre observed by Poles in American movies or during their stays abroad. Therefore, the conclu-

44

Joanna Szerszunowicz

sion can be reached that the notion of cheerfulness, presented by Wierzbicka as typical of American culture (1999), is gradually becoming more and more important in the Polish culture, and consequently in the Polish small talk. Moreover, the students who filled in the questionnaire on small talk in both cultures commented on the similarities between Polish and American/English genres.5 They noticed the reflection of the keep-smiling attitude of interlocutors as well as the tendency to exploit formulaic expressions at the expense of the speakers' creativity. They referred to the new characteristics of the genre by describing it as more and more American, business-like, insincere and always polite, routine and nice. points Polish English 1 4% 6% 2 7% 18% 3 32% 34% 4 36% 31% 5 11% 5% 6 4% 3% 7 4% 2% 8 2% 1% 9 0% 0% 10 0% 0% Table 2. The Evaluation of Importance of Subjects Mentioned in Small Talk Conversations

It should be stressed that a variety of extralinguistic factors are the key factors influencing the small talk genre. The political and economic development of the country influences the mentality of Poles. At present the acknowledgement of being successful in terms of position, financial status, education, private life, etc. is viewed as a positive phenomenon in the Polish culture. Previously, the notion of being successful was mostly associated with communist party membership. Nowadays the change in Poles' attitude towards being successful is noticeable also in the genres of speech. According to Swales (1990), social groups create their own inventory of genres according to their needs, thus, in a new situation either there is a need for a new genre or existing genres are modified, so that they could meet the needs of the community. The changes, which have taken place in Poland in the last decade, account for the tendency to avoid complaining, since the act of complaining amounts to the acknowledgement of being an unsuccessful person. Therefore, the linguistic means implemented in small talk by Polish speakers serve the purpose of sounding positive. Such words as great and okay appear with a similar frequency in the Polish and English texts analyzed. The fact can be interpreted as the actualization of the cultural norms of friendliness, happiness and cheerfulness. Wierzbicka (1999) The questionnaire contained the following questions: Have you observed any changes in the Polish small talk genre? If so, how would you describe the changes? What influences the Polish small talk?

A Comparative Anaylsis of Small Talk

45

points out that the Polish language does not possess a word comparable to the American great, yet it seems that the positive values, irrespective of the means used to express them, are introduced into the Polish small talk dialogues. The conclusion is that the Polish cultural script has changed, which is proved by the analysis of the corpora.6 The modification of the script involves changing the inventory of lexical means implemented in the Polish small talk. For instance, one of the features of the Polish corpus is the accumulation of idiomatic expressions in the form of structure-indicating devices. Such phrases performing the ritual function tend to appear both in the Polish corpus and in the English one. However, in the Polish realizations of the small talk genre the rate of occurrences of context-bound idioms is slightly lower than in English. Yet, it should be stressed that the inventories of small talk idiomatic signals of beginning or ending the conversation are comparable in both languages analyzed. The explication of the small talk genre presented by Wierzbicka refers to works by Bakhtin, Wittgenstein, Austin, and Searle, as well as to her own semantics of elementary units. Wierzbicka models each of the genre described through a combination of simple sentences expressing the assumptions, intentions and other mental acts of the speaker, which defines a given genre of speech (1983: 129). Since the elementary semantic units are the elements of each genre explication, the comparison of particular genres is facilitated, which is especially important in a cross-cultural perspective. If the explication method is applied in order to analyze small talk, the majority of the statements describing both the Polish and the English structure of the genre are identical. However, the explication of the English small talk based on the culturally modified corpus compared with the Polish structure shows some significant differences. The English genre is presented through the following statements: (1) (2) (3) (4) (5)

I am telling you something many people tell each other every day; I know that you expect me to do it; I want to tell you something you often hear when people see you; I want you to feel good; I want to say something nice;

(6) I want to sound neutral. The Polish small talk genre is defined by means of a set of similar, yet not identical statements: (1) I am telling you something many people tell each other every day; (2) I know you expect me to do; It is also confirmed by observations of everyday communication. The shop assistants, clerks, etc. are much more polite and attentive than a decade ago, which results from social, political, and economic changes.

46

Joanna Szerszunowicz (3) I want to say something to you because I see you.

In fact, it can be assumed that the subjects of Polish small talks do not have to be neutral and the quality of 'being nice' is not a key characteristic of the small talk dialogues in the Polish culture. However, it should be stressed that there is a tendency to avoid unpleasant and very serious issues. Therefore, the Polish small talk is still in the state of evolution and in the nearest future it is likely to develop into a more defined genre with characteristic of the English/American small talk. Consequently, presenting the description of the Polish small talk is particularly difficult, yet it is important to analyze the genre and observe its development. Moreover, the analyses should cover more aspects of research (i.e., prosodic features) since small talk is primarily a genre of speech and body language. Such elements influence the discourse to a great extent. According to Fernando (1996: 161): The essence of small talk is that it is low in factual information [...]. However, though low in the kind of factual information crucial to the progression of a verbal interaction such as a service encounter, small talk conveys other sorts of information as a result of: (1) checking on the physical and emotional state of other; (2) expressing opinions conveying evaluations [...]: good, bad and ambivalent even though such information may sometimes be inaccurate, that is, a white lie. On the other hand, it could be accurate. Facial expression, eye-contact, and tone of voice are probably the best guides to the veracity of I'm fine or It 's really great in discourses [...].

The exclusion of the above-mentioned factors, which are not within the scope of the present paper, leaves some of the questions regarding the small talk genre unanswered. However, certain linguistic and cultural phenomena regarding the small talk genre are signaled in the analysis.

4. Conclusion

The comparison of the Polish small talk dialogues and the culturally modified English texts proves that there are more similarities than differences between the two corpora. The Polish genre is becoming closer to the English one due to the extralinguistic factors influencing the mentality of Poles. The binary opposition approach in the analysis of the small talk genre, i.e. the presentation of Polish and English small talk genres as sincere - insincere,

A Comparative Anaylsis of Small Talk

47

personal - impersonal, etc., is an oversimplified view of the phenomena. The studies on the small talk genre should be conducted with a focus on the extralinguistic factors, since the Polish small talk is becoming an increasingly distinctive genre of speech, which is proved by the fact that there are more restrictions on topics and register than before. Therefore, the structure of a small talk can be described as a universal pattern, whose modifications result from cultural differences. The analysis of corpora consisting of more texts than the ones researched in the present study is necessary to present a more complex explication of the small talk genre. The comparison of American, British and Polish small talk corpora is necessary for presenting explications of the small talk genre in a contrastive perspective, since the small talk texts provided by the students of English philology can be treated only as a culturally modified versions, based on a stereotype of the American/English small talk. Since the small talk dialogues are particularly common in everyday communications, the results of small talk research are of importance for teaching Polish/English both as a first and as a second language, especially in the fields of intercultural skills, translating/interpreting and linguistics, particularly sociolinguistics, practical stylistics and oral genre studies.

References Duszak, Anna (1998): Tekst, dyskurs, komunikacja mi?dzyj?zykowa. - Warszawa: Wydawnictwo Naukowe PWN. Fernando, Chitra (1996): Idioms and idiomacity. - Oxford: Oxford University Press. Furdal, Antoni (1982): Genologia lingwistyczna. - Biuletyn PTJXXXIX, 61-70. Gajda, Stanislaw (2001): Gatunkowe wzorce wypowiedzi. - In: Jerzy Bartminski (ed.): Wspolczesny j?zyk polski, 255-280. Lublin: Wydawnictwo Uniwersytetu Marii Curie-Skiodowskiej. Hasan, Ruqaiya (1978): Text in the systemic-functional model. - In: Wolfgang U. Dressler (ed.): Current trends in textlinguistics, 228-244. Berlin: Walter de Gruyter. Jakubowska, Ewa (1999): Cross-cultural dimensions of politeness in the case of Polish and English. - Katowice: Wydawnictwo Uniwersytetu Sla^skiego. Kramsch, Claire (2000): Language and culture. - Oxford: Oxford University Press. Katriel, Tamar (1986): Talking straight. Dugri speech in Israeli Sabra culture. - Cambridge: Cambridge University Press. Saville-Troike, Muriel (1982): The ethnography of communication. An introduction. - Oxford: Basil Blackwell Publications. Swales, John (1990): Genre analysis. English in academic and research setting. - Cambridge: Cambridge University Press. Urbanczyk, Stanislaw/Kucala, Marian (eds.) (1999): Encyklopedia j^zyka polskiego. - Wroclaw: Ossolineum. Wierzbicka, Anna (1983): Genry mowy. - In: Teresa Dobrzynska, Elzbieta Janus (eds.): Tekst i zdanie, 125-137. Wroclaw: Ossolineum.

48

Joanna Szerszunowicz

Wierzbicka, Anna (1999): Akty i gatunki mowy w roznych kulturach. - In: Jerzy Bartminski (ed.): Jfzyk - umysl - kultura, 228-270. Warszawa: Wydawnictwo Naukowe PWN. Wilkon, Aleksander (2003): Gatunki mowione. - In: Malgorzata Zofia Kita (ed.): Porozmawiajmy o rozmowie, 46-58. Katowice: Wydawnictwo Uniwersytetu Sla^skiego. Wolanska, Ewa (2003): Gatunek wypowiedzi. - In: Edyta Bankowska, Agnieszka Mikotajczuk (eds.): Praktyczna stylistyka nie tylko dla polonistow, 94-107. Warszawa: Ksiajzka i Wiedza. Wyrwas, Katarzyna (2002): Skarga jako gatunek mowy. - Katowice: Wydawnictwo Uniwersytetu Slajskiego.

Johanna Miecznikowski

Modality and Conversational Structure in French

In an interactionist perspective, linguistic forms are to be studied not only as elements of speech acts, but also as to their contribution to the structuring of verbal interaction. In this paper, an approach of this kind is applied to the conditional form in French in its attenuative reading. An exploratory study is reported in which the uses of the conditional and the present indicative form of vouloir in a corpus have been compared. It suggests that the attenuative variant of the conditional underlines the interactive negotiability of sequence openings and thematic transitions and contributes specifically to the formulation of overtly modalized propositions that are new and unexpected with regard to preceding discourse. This finding is relevant to the linguistic description of the French conditional, but also more generally to the understanding of the interrelationship between modality and sequential structure.

1. Introduction

A dialogic view of communication, which does not reduce communicational processes to the transmission of knowledge or to series of unrelated acts (cf. Fetzer this volume), has important consequences not only for the study of these processes, but also for the study of language as one of the semiotic systems involved. If communication is regarded as a type of complex social event with specific structural characteristics, to be studied in its own right, investigating the relationship of language to these characteristics becomes an important task of linguistics. Linguistic forms are then to be studied not only as elements of the speech event1 or the speech act, but also as to their contribution to the structuring of verbal interaction. In this paper, a linguistic approach of this kind will be applied to a particular linguistic marker (i.e., the conditional form in French in its so-called attenuative reading). The interactional functions of this marker have been described in pragmatic frameworks stressing the close relatedness of modality, indirectness of speech acts and politeness (Diller 1977; Speech event translates to enunciation, a key-term of the French pragmatic tradition.

50

Johanna Miecznikowski

Ludwig 1988; Roulet 1993). In what follows, I will revisit the attenuating conditional in a perspective that is influenced by conversation analysis2 (CA) and interactional linguistics (cf. Ford, Fox, and Thompson 2002; Mondada 2001; Selting and Couper-Kuhlen 2000/2001). It inherits from CA its focus on sequential structure (e.g., Macbeth this volume), but also its empiricism, which distinguishes CA from other approaches to interaction, such as dialogue grammar (e.g., Hundsnurscher this volume). On the basis of an exploratory study, I will show some advantages of a data-driven interactionist approach to grammatical forms. I will argue, in particular, that it contributes both to linguistics, by enhancing the adequacy of linguistic description, and to an analysis of dialogue that takes into account the use participants make of language to regulate their interaction.

2. The Attenuating Conditional in French

The French conditional form is highly polysemous (e.g., Grevisse 1980; Haillet 2002; Ludwig 1988; Riegel 1996). In addition to its function as a "future of the past," it can express the counterfactuality or potentiality of a state of affairs dependent on a condition, polyphony, fictionality (e.g., in the context of children's play), and finally it is used to modalize communicative acts such as requests, suggestions, advice, or offers. It is the latter reading, illustrated by the following (authentic) examples, that this paper will concentrate on: (1) Je voudrais faire quelques petits commentaires. (2) Moi je vous proposerais de rester ici.

This variant of the conditional is traditionally attributed an attenuative function, which makes the communicative act expressed more polite (Grevisse 1980). Also in more recent works in pragmatics, we find the idea of polite attenuation. Ludwig (1988, 194) states that the variant in question occurs in indirectly formulated appeals, in Bühler's sense, and terms it "Signal des abgedämpften Nachdrucks." Roulet (1993) treats it, together with other markers such as the modal verbs with which it is combined so frequently, as a means that allows the speaker to attenuate the face-threatening potential of speech acts (cf. Brown and Levinson 1987). This type of description captures quite well certain intuitions one can Cf. Gülich and Mondada (2001) for a recent introduction.

Modality and Conversational Structure in French

51

have concerning isolated utterances, such as (1) and (2) above. In particular, it accounts for the fact that the attenuating conditional hardly contributes anything to the prepositional content of these utterances, and plausibly locates the function of this marker on the interactional level. However, degrees of attenuation or face-threat are difficult to estimate and to control for empirically. Moreover, a description in terms of attenuation fails to provide a satisfactory explanation for those cases found in corpora in which, in cotexts comparable to those illustrated by (1) and (2), the conditional does not occur. Are those speakers just being less polite? Under which circumstances precisely speakers choose the attenuating conditional? In a conversationalist perspective, I have formulated the working hypothesis that the choice of the French conditional in requests and the like is not only related to the management of interpersonal relations, but also to the temporally unfolding structure of discourse and interaction on a) the thematic level, and b) the level of interactional sequences, from adjacency pairs to longer sequences. This working hypothesis is the point of departure of a larger, still ongoing study on the use of the conditional in spoken French and Italian. In what follows, it will guide an exploratory study focusing on the use of the modal verb vouloir in a corpus, which will lead to more detailed hypotheses to be tested by further analyses (cf. Aijmer this volume, for the use of corpora in dialogue analysis).

3. An Exploratory Study: Vouloir

3.1. Design of the Analysis The conditional form of vouloir is frequently of the attenuative type. This is especially true if it is used in the 1st person, a use that is regularly interpreted in the sense of a request. This exploratory analysis concentrates therefore on 1st person tokens, and more precisely on the 1st person singular. Since the 1st person of vouloir can express requests in other tenses/modes as well - the idea of a request is related to the expression of the speaker's volition - a comparison between different tenses/modes can provide useful insights as to the functions of the conditional in this cotext. For the sake of brevity, only tokens of vouloir in the conditional I vs. in the present indicative have been included for comparison here; a more comprehensive analysis will have to take into account also the imperfect (je voulais), the future indicative (je voudrai), and the conditional II (/' 'aurais voulu).

52

Johanna Miecznikowski

The use of vouloir has been studied on the basis of a corpus gathered within a research project on the interactional construction of scientific discourse.3 The corpus consists of tape-recorded and transcribed meetings of research groups. It is plurilingual, including French, German, English, and Italian parts; in a rough approximation, the French part counts about 400,000 words. This corpus is not sociolinguistically representative, but it provides interesting data for the present study because it is rich in modalizations of all kind. In accordance with CA methodological principles, no rigid pre-established analytical grid has been applied to the linguistic forms examined. In the course of the analysis, however, the following dimensions have proved to be particularly relevant: (l)The type of state of affairs denoted by the proposition modalized by vouloir; (2)The status of this state of affairs with regard to the speakers' background knowledge; (3)The construction of thematic boundaries vs. thematic continuity; (4) The co-occurrence with discourse markers and further modal markers; (5)The placement of the utterance in the turn; (6) The sequential implications of the utterance.

3.2. Occurrence of Vouloir in the Corpus In the corpus, the modal verb vouloir occurs 824 times in total. The conditional I of vouloir numbers 154 tokens, of which the majority, namely 134, are in the 1st person singular.4 Of these, 126 have been included in the analysis as instances of the attenuating conditional, whereas I have excluded 4 tokens in direct reported speech, 2 instances used in a counterfactual sense within a hypothetical construction, as well as 2 unclassifiable tokens in interrupted utterances. The present indicative of vouloir numbers 441 tokens, of which 136 are in the 1st person singular and 7 are in the 1st person plural. Vouloir is not used at all in the 2nd person of the conditional, whereas we count 82 tokens of the 2nd person present indicative.

The project, which I have participated in, has been directed by L. Mondada at Basel University, from 1997 to 2001 (cf. Miecznikowski in press). I would like to thank the project leader and the Swiss National Foundation, which has funded the project (no. 1214-051022.97), for allowing further use of the data gathered. The conditional form occurs 1972 times in total.

Modality and Conversational Structure in French

53

3.3. Use of je voudrais Je voudrais is mainly used with verbs of saying, to announce and categorize a following portion of talk (116 instances). In about half of the cases (56 instances), this kind of metacommunicative utterance is placed turn-initially, where it announces that the speaker will take the floor to produce a longer turn, and specifies what kind of turn he/she will produce. In the remaining 60 cases, the metacommunicative announcement is placed within a turn, where it is accompanied by discourse-structuring devices of all types: (l)It is often preceded by pauses or falling intonation; (2) It is regularly accompanied by expressions indicating a thematic boundary, such as alors, maintenant, la deuxieme chose, le dernier volet, pour finir, donner un exemple, etc.; (3) In some cases, it is accompanied by expressions indicating argumentative contrast, such as mais or quand-meme.

Within a turn, je voudrais+\erb of saying thus contribute to the thematic structuring of discourse. Moreover, sometimes this metacommunicative use oije voudrais has a turnstructuring function related to the sequential organization of interaction, which is comparable to the introducing function of je voudrais at the beginning of turns. This is the case when the utterance in question is placed after a sequence of discourse which reacts to a previous turn (a second pair part), and introduces a part of the turn which has the character of an initiative (a first pair part). The structuring function of je voudrais+\erb of saying is illustrated by the following extract of a videoconference among surgeons, which follows a presentation on a special kind of surgery (T.E. M.): (3)(TC12038/V) (3.1) X oui . eu:h done nous nous avons aussi une experience dans cette: eu:h Chirurgie de T.E.M yes . uh so we we have also some experience in this: uh: T.E.M. surgery (3.2) on Γ a presentee ici meme: 'il y a un an je crois il y a quelques mois je ne we 've presented it here: One year ago I think a couple of months ago I don't (3.3) sais pas si vous vous souvenez/ ((fast))'. et euh notre serie ici est assez limitee know if you remember/ ((fast))'. and uh our series is quite limited (3.4) nous avons euh trente et un cas\. alors euh je voudrais faire quelques petits we have thirty-one cases\. so uh I would like to make some small (3.5) commentaires/ eu:h concernant la. la technique chirurgicale elle-meme ((....)) comments/ uh: concerning the. the surgical technique itself ((....))

The utterance containing ye voudrais (3.4) is preceded by a falling intonation contour, a micro-pause and the discourse marker alors + euh. It announces a longer portion of talk (faire quelques petits commentaires, 3.4-3.5), specifying which topic it will be about (3.5).

54

Johanna Miecznikowski

In addition, the thematic boundary at which it occurs coincides with the transition from a preliminary, mainly reactive part of the turn to the main part of the turn (i.e., comments), which implies sequentially a reaction by the presenter to whom it is addressed. Aside from communicative actions by the speaker, je voudrais can also express the speaker's intentions as to a future state of affairs that is in some way relevant to the hearer. In sequential terms, the utterances in question occupy first pair parts - corresponding to the prototypical requests and suggestions exemplifying the attenuating conditional in the literature - or dispreferred second pair parts, implying the rejection of a request or suggestion previously formulated by an interlocutor. On the thematic level, the propositions modalized are new with respect to the universe of discourse and/or contrast with respect to previous discourse, and are foregrounded. Example (4), a further extract of a videoconference, illustrates the use of je voudrais in dispreferred seconds: (4)(TC 12028/V) (4.1) DU est-ce que tu penses est-ce que tu penses qu'il faut=euh . qu'on fasse une reunion=euh do you think do you think we should uh . organize a meeting uh (4.2) Lelacq N.N. et nous/=euh on peut venir ä Strasbourg une fois pour voir euh . ce Lelacq N.N. andwe/uh we can come to Strasbourg some day to figure out uh. what (4.3) qu'on veut faire avec ce site/ ou bien tu penses que 9a sert a rien\ we want to do with the site/ or maybe you think there 's no use\ (4.4) MA [si je pense [on the contrary I think (4.5) DU [t'as pas le temps de toute fa£on\ [you don't have time anyway\ (4.6) MA euh ce que ce que je voudrais c'est que lui ((=Lelacq)) il soit absolument la si N.N. uh what I would want is him ((=Lelacq)) to certainly be there if N.N. makes the effort (4.7) se deplace c'est evident je vais lui en parier et puis je te donnerai la reponse\... oke/ to come that 's clear I will talk to him and then I'll give you an answer\... okay/

Je voudrais (4.6) is placed in a reaction to a suggestion by the speaker DU that is dispreferred in the sense that it delays any reaction in the sense of an agreement (cf. 4.7). The intention expressed by je voudrais, justifying this delay, is foregrounded thanks to a pseudo-cleft construction and presented as constituting an argumentative contrast with regard to DU's suggestion.

Modality and Conversational Structure in French

55

3.4. Use of je veux Like je voudrais,je veux is combined, in the vast majority of cases, with verbs of saying. Of these, dire is the most frequent. We count 97 tokens of je veux dire, used either parenthetically or with a complement clause introduced by que. Both variants of je veux dire express autonymical modalization (Authier-Revuz 1995), or reformulation of previous talk. Unlike ye voudrais+verb of saying, they do not announce a longer stretch of talk; they have a strictly local scope and are rather retrospective/reformulative than prospective. The same type of metacommunicative scope can be attributed to 2 further tokens of je veux combined with other verbs of saying. In addition, je veux occurs in the expression ye veux bien (le croire) (4 tokens) as a signal of reception and agreement with regard to a statement by a precedent speaker - another metacommunicative cotext in which the conditional of vouloir is not to be found. On the other hand, ye veux+verb of saying can have the function of a metacommunicative announcement (19 tokens). Even if this case is comparable to the use of je voudrais in announcements, the circumstances under which it is to be found are different, however. On one hand, we do not find any turn-introducing announcements with ye veux, apart from the rare case where the taking of the floor by the current speaker has been prepared by another speaker, and the former is thus merely re-announcing his/her turn. On the other hand, turninternally y'e vettx+verb of saying serves to announce a portion of talk that is in strong argumentative and thematic continuity with prior talk, rather than to mark a thematic boundary. It is combined with markers, such as done, enfin, en fait, et, and tends not to occur after points of conclusion clearly marked as such by syntax, intonation, or pauses. Moreover, the portion of talk to follow is frequently presented as being brief or of minor importance, using negation (of the type je ne veux pas rentrer dans les details) or expressions such as juste, seulement, petit+NP. The use of je veux in turn-internal metacommunicative announcements, as described above, is illustrated by example (5), which is extracted from a talk given by a surgeon: (5)(TC08018/V) (5.1) SE (5.2) (5.3)

((....)) on distingue . je veux juste vous montrer un petit un petit schema d'un d'un ((....)) we distinguish . I just want to show you a little a little schema of one of the of des articles/ (h) on distingue en fait/ . '(4s) ((ajusts the transparency))' voila\. one of the articles/ (h) we distinguish actually/ . '(4s) ((ajusts the tr.)) ' here we are\ . on distingue en fait ((1 line omitted)) ces quatre formes . eu:h d'un xxx cliniwe distinguish actually ((1 line omitted)) these four forms . uh: of a clinic xxx/ ((....))

The speaker chooses je veux to introduce his presentation of a transparency (5.1). He combines the announcement with markers of downgrading (juste, petit schema, 5.1); on the

56

Johanna Miecznikowski

syntactic and intonational level, he formulates it as a parenthesis, tightly integrating it into on-going talk. Like ye voudraisje veux serves most often metacommunicative purposes, but not exclusively so. In 14 cases in the corpus, it expresses an intention related to a future state of affairs. In contrast to the attenuating conditional of vouloir, however, the utterances in question have more often the character of a reaction than of an initiative, and more precisely of a preferred reaction to a suggestion or request, indicating acceptation and consent. In many cases, moreover, the intention expressed by je veux is not directly relevant to the hearer and the ongoing interaction. On the thematic level, the propositions are often already part of the common ground because a) the speaker paraphrases and reasserts a previously mentioned intention; b) the intention in question is known and uncontroversial in the given context; or c) the proposition is presented as presupposed (e.g., in a clause introduced by puisque). In other cases, the utterances containing ye veux establish at least a strong thematic and argumentative continuity with respect to prior talk (e.g., in the case of preferred second pair parts). Consider, for instance, the following extract of a dialogue between two researchers: (6)(TC08018/V) (6.1) DU il m'a dit que si il ENvoie ce rapport a Berne/.. on arrete 'N.N. ((project name))'\ he told me that if he SENDS this rapport to Bern/.. they -will stop N.N. \ (6.2) MA d'accord\= Isee\= (6.3) DU =et moi je veux PAS/ nous on veut PAS/ et done euh il faudra qu'on attende d'avoir —and I DON'T want (that to happen)/ we DON'T want (that)/ and therefore we will have to (6.4) UN des faits/. euh: des trues ä montrer/ wait until we have ONE of the facts/. er: some things to show/ Here negated je veux (6.3) expresses the intention of the speaker, DU, not to accept and maybe to prevent the impending stop of the project both participants are engaged in. With respect to the background knowledge of the participants, this intention can be assumed to be both known to MA and shared by him. As to the thematic development of the sequence, the je veux clause creates strong topical continuity, by referring back to the proposition "o« arrete" (6.1) by means of a zero anaphora. It is integrated, as a premise, into an argument leading to a conclusion introduced by done (6.3-6.4), and has no clear sequential implications of its own on the interactional level.

Modality and Conversational Structure in French

57

4. Discussion

The observations about vouloir presented above support the idea of an attenuating function of the French conditional in some respects. In particular, they show that, in interactional contexts as those of the corpus examined here, in which we expect speakers to behave politely, the conditional is used regularly in requests and suggestions occupying first pair parts or dispreferred seconds, whereas we do not encounter ye veux in this sequential context. Insofar as these conversational acts can be considered as potentially face-threatening, the regular presence of an attenuative marker is expectable. Conversely, je voudrais does not occur in positive reactions to suggestions and requests and in signals of agreement; indeed, attenuation in these contexts arguably is neither necessary nor even beneficial with respect to face management.5 However, an explanation in terms of polite attenuation cannot account for all patterns of use observed in the corpus, especially concerning the very frequent use of je veux and ye voudrais in metacommunicative utterances. First of all, there is a type of metacommunicative operation, namely announcements, in which ye voudrais is preferred, but ye veux is used as well. Politeness strategies might explain the preference of je voudrais at the beginning of turns, since it announces a longer occupation of the conversational floor. The turninternal alternation of the two verb forms as well as those rare cases of turn-initial reannouncements usingy'e veux, however, are related primarily to the thematic and argumentative development of discourse and are difficult to explain in terms of politeness. The latter claim applies also to reformulations and autonymical modalizations of the typey'e veux dire, where we findy'e veux, but not ye voudrais. The findings reported above indicate in fact that the attenuating conditional of vouloir, unlike the present indicative, is associated with new and foregrounded information and with the introduction of thematic sequences after a thematic boundary. This fact is congruent also with the metacommunicative use of the attenuating conditional at the beginning of turns and with its use in first pair parts and dispreferred seconds: it is placed at the beginning of a sequence or, in the case of dispreferred seconds, at a sort of break in the chain of expectable conversational moves. The results of this exploratory study thus suggest that the attenuation expressed by the conditional is related to the sequential organization of discourse and interaction and to the management of shared knowledge, thus confirming the underlying working hypothesis of the study. They lead to the stronger claim that, rather 5

A substitution test replacingy'e veux withy'e voudrais (e.g., in the expressiony'e veux bien le croire) shows that the attenuative variant of the conditional is in fact blocked in these contexts; the conditional must be interpreted in a potential or even counterfactual sense here.

58

Johanna Miecznikowski

than being a generic marker of attenuation used to reduce face-threat, this variant of the conditional underlines the interactive negotiability of sequence openings and thematic transitions6 and contributes specifically to the formulation of overtly modalized propositions that are new and unexpected with regard to preceding discourse. This claim generates a series of questions to be examined in further research. On the one hand, it constitutes a hypothesis to be examined with regard to other cotexts than that of the modal verb vouloir. On the other hand, the reanalysis of the attenuating conditional proposed here will have to be integrated into a comprehensive description of the French conditional as a polysemous marker. In particular, the question will have to be addressed if the discourse structuring functions observed in the case of the attenuating conditional are related to more general semantic and syntactic properties the conditional has in other readings as well, and hypotheses will have to be formulated as to possible diachronic paths leading to the grammaticalization of these functions (cf. Aijmer this volume, for a study of parallel corpora examining different surface manifestations). A related question concerns the language specificity of the observations made here: do we expect similar relationships between modality and sequentiality in other languages, or are the discourse structuring functions of the conditional specific to the French modal system?

5. Conclusion

The interactionist approach adopted in this paper leads to new questions and to unexpected answers concerning an old subject in linguistics (i.e., modality), and in particular the French conditional form. The exploratory study reported here suggests that interactional linguistics might provide a valuable theoretical and methodological framework for the analysis of tense and mood in interaction. It shows that one of the central claims made in that framework, namely that language is shaped by interaction and, conversely, helps structuring interaction (e.g., Selting and Couper-Kuhlen 2000),7 applies also to an area that has been considered rather in a pragmatic than in a conversation analytical perspective so far. Cf. Kallmeyer (1978) for a discussion of focussing operations, in a perspective including both transitions between interactional sequences (cf. Schegloff and Sacks 1973) and the thematic organization of discourse. "Die Verbindung zwischen Sprachstrukturen und dem sequenziellen Interaktionskontext ist wechselseitig: Sprachstrukturen werden von der Interaktionsstruktur geprägt und tragen gleichzeitig dazu bei, genau diese Interaktionsstruktur herzustellen" (Selting and Couper-Kuhlen 2000: 82).

Modality and Conversational Structure in French

59

The examination of particular linguistic markers such as the conditional form is of interest not only to linguistics, but also to conversation analysis or other types of empirical dialogue analysis. The objects of research of such approaches being conversations as social events and types of sequences or operations, the functional range of linguistic units and constructions often remains outside the scope of analysis. To be able to interpret the use of linguistic means in situ - a precious source of information - conversation analysts have then to rely on semasiologically-oriented research on such units and constructions that takes into account their occurrence in various sequential environments. Analyses as the one discussed in this chapter are of that kind, thus contributing to the dialogue between linguistics and the study of interaction (cf. Aijmer this volume).

References Authier-Revuz, Jacqueline (1995): Ces mots qui ne vont pas de soi. Boucles enonciatives et noncoincidences du dire. Paris: Larousse. Brown, Penelope/Levinson, Stephen C. (1987): Politeness. Some universals in language usage (2nd ed.). - Cambridge: Cambridge University Press. Diller, Anne-Marie (1977): Le conditionnel, marqueur de derivation illocutoire. Semantikos 2/1, 117. Ford, Cecilia E./Fox, Barbara A./Thompson, Sandra A. (eds.) (2002): The language of turn and sequence. - Oxford: Oxford University Press. Grevisse, Maurice (1980): Le bon usage. - Paris: Duculot. Gülich, Elisabeth/Mondada, Lorenza (2001): Analyse conversationnelle. - In: Gunter Holtus, Michael Metzeltin, Christian Schmitt (eds.): Lexikon der Romanistischen Linguistik, 196-251. Tübingen: Max Niemeyer Verlag. Haillet, Pierre-Patrick (2002): Le conditionnel en fran9ais. Une approche polyphonique. Paris: Ophrys. Kallmeyer, Werner (1978): Fokuswechsel und Fokussierungen als Aktivitäten der Gesprächskonstitution. - In Reinhard Meyer-Hermann (ed.): Sprechen, Handeln, Interaktion, 191-241. Tübingen: Max Niemeyer Verlag. Ludwig, Ralph (1988): Modalität und Modus im gesprochenen Französisch. - Tübingen: Gunter Narr. Miecznikowski, Joanna (in press): Le traitement de problemes lexicaux lors de discussions scientifiques en situation plurilingue. - Bern: Peter Lang. Mondada, Lorenza (2001): Pourune linguistique interactionnelle. Marges linguistiques l, 1-21. Riegel, Martin (1996): Grammaire methodique du fran9ais. - Paris: Presses Universitaires de France. Roulet, Eddy (1993): Des formes et des emplois des modalisateurs de proposition dans 1'interaction verbale. - In: Norbert Dittmar, Astrid N. Reich (eds.): Modality in language acquisition, 27-39. Berlin: Mouton De Gruyter. Schegloff, Emanuel A./Sacks, Harvey (1973): Opening up closings. - Semiotica 8/3, 289-327. Selling, Margret/Couper-Kuhlen, Elizabeth (2000): Argumente für die Entwicklung einer 'interaktionalen Linguistik'. Gesprächsforschung l, 76-95. Retrieved in April 2003 from www.gespraechsforschug-ozs.de.

60

Johanna Miecznikowski

Selling, Margret/Couper-Kuhlen, Elizabeth (eds.) (2001): Studies in interactional linguistics. - Amsterdam: John Benjamins.

Appendix Transcription Conventions [] (2s) xxx /\ : par& = (il va) (h) wirklich exTRA '' ((laughs))

N.N.

begin / end of overlap. pauses (' 2 sec.) pauses in seconds (' 2 sec.) incomprehensible segment rising / falling intonation stretched vowel truncation continuation of turn across line break rapid continuation tentative transcription inbreath vowel in cap.s: high pitch dynamic accent delimitation of phenomena phenomena not transcribed indicated between (()) comments by the transcriber name

Franz Hundsnurscher The Principles of Dialogue Grammar

Dialogue grammar, as I see it, can be imagined as a sort of generative-transformational grammar, and it can be reconstructed along the lines pointed out for syntax by Chomsky (1965). Just as in syntax we are confronted with a bewildering mess of talk of all kinds, people are constantly engaging in new ever-changing conversations; they generally know the type of conversation they are in, and they generally know what they are after, whether they were able to make their point or not, and whether they are being interrupted and have to resume some line of argument or narration. These are all aspects of communicative competence that will have to be made explicit and explained in a comprehensive theory of verbal communication. As the concept of well-formedness is central to grammar, so some concept of a well-formed dialogue will have to be worked out, along with the deepstructure of dialogue and the vagaries of conversational surface structure as a result of special kinds of transformations. This is a huge program which I can only sketch here and give a few examples for illustration. Although it may be very interesting and sometimes even entertaining to investigate the various forms and ways in which people talk to each other, as linguists we cannot content ourselves with just picking out pieces of authentic verbal interaction, describing and analyzing individual aspects to a certain point, and then going on to other pieces of authentic talk in different situations and pointing out yet other coincidental phenomena. We simply cannot go on like this because authentic talk is limitless and omnipresent, and simply pointing out how variegated and different it is from speaker group to speaker group and from situation to situation will get us nowhere linguistically if we have no reliable methodological framework to work with and if we are not clear about what to look for in analyzing dialogue. I think we have to concentrate on communicative competence, not on performance, and reconstruct the basic patterns before going on to more complicated models. In works of conversation analysis, I have seen lots of transcriptions of authentic conversation with every pause and overlap, every cough and giggle painstakingly registered, but, when it comes to interpreting all these data and factual observations, they are either met with trivial comment or simply neglected, and the results are to a large extent inconclusive and void (Hundsnurscher 1980). Yet the scholars doing conversation or discourse analysis are immensely proud of their empirical and scientific approach. One reason for this kind of attitude and procedure could be that linguists imitate, more or less consciously, the methods of literary interpretation where the most minute detail

62

Franz Hundsnurscher

may contribute to aesthetic appreciation. But ordinary language conversation is not to be confounded with the poetic language of literary tests there is normally no great effort at stylistic refinement or symbolic meaning. I also doubt if it is the linguist's task to work out the psychological or sociological significance that may lie behind some individual person's verbal behavior. The linguist's foremost task with respect to communicative verbal interaction is, to my mind, to discover and explicate the patterns and structures underlying language use in general and to characterize talk as a form of rule-governed behavior. It is wrong, I think, to deviate too far from the core concept of linguistics, the rules-and-units approach, that is, from the concept of grammar (Franke 1990; Hundsnurscher and Franke 1985). I am not against empirical research in linguistics, but such research must be method-ologically sound and theory-guided, and linguists working in the field of dialogue analysis should be able to spell out their theoretical framework and their methodological guidelines when looking at and dealing with empirical data. The most plausible basis for a dialogue theory is, I think, a general action theory as developed, for instance by Aristotle (Bien 1972) or, in more recent times, by von Wright (1974) and others (cf. Franke 1990; Fritz 1982; Fritz and Hundsnurscher 1994; Kilian 2002; Weigand 1989). For my purpose of dialogue grammar, I start from very simple version of such a theory as a triple structure consisting of conditions, means, and goals (see figure 1). Action

action conditions

action nwans

action goals

Figure 1.

A person will be successful in performing action (i.e., attaining his goal) when, under appropriate conditions, he uses the adequate means. Austin's (1962) and Searle's (1969) concept of a speech act can be modeled along these lines and this will yield the general structure of speech acts or verbal actions in general (see figure 2; see also Fetzer this volume). Speech Act

CoramQjiicatiYe Conditions

Figure 2.

Utterance Forms

lliocntion & Perlocuiion

The Principles of Dialogue Grammar

63

Moving on from speech act theory to dialogue theory, one will have to consider several points of difference; I shall, for a start, mention only four: (l)The basic condition, of course, is, that there are two competent speakers, speaker 1 (Spl) and speaker 2 (Sp2), that take alternating turns in uttering sentences. Even at this elementary level, the problem of speaker instantiation and constellation will have to be considered: There may be more than two persons involved in the conversation, and this will influence the course of the conversation. Or the persons involved may have a categorically different status. In the case of soliloquy, one is said to be talking to or with oneself; in other cases, one speaker will talk to a group of persons, for instance, when preaching or lecturing. In the case of a prayer, people imagine themselves to be talking with God, or in the case of poetry, one may be said to be talking to the moon. There are people who talk to animals, houseplants, or other objects. We cannot consider such talk as being the basic or standard type of language use, but in the long run such phenomena will also have to be investigated. (2)Speech act theory (Austin 1962; Searle 1969) usually concentrates on single sentences as instances of speech acts; in dialogue, we will have utterances below sentence level (e.g., yes or no, why, all right, etc.) which we might call minimal moves and clusters of utterances comprising several sentences which we might call complex moves, or embedded monologues. (3)Another important point is the relation of utterance form to speech act type. Normally, there is not just one utterance form available to perform an illocutionary act, but there are sets from which to choose the one that the speaker thinks is best suited for the occasion. He may, for instance, phrase a general request (e.g., Do this for me) in many different ways, for example: (a) This has to be done. (b) Would you be kind enough to do this"? (c) I request you to do this. (d) Could you do this for we? (e) Haven Ί you done this yet! (f) You just do this. (4)Whereas speech act theory mainly concentrates on illocution, perlocution is all-important in dialogue because, by interacting verbally, not only do people want to be understood by a hearer, but also they want the speech act to take effect and to come to an understanding with the other speaker; this is the general goal or engaging in dialogue. Only by explicit reaction on the hearer's part can Spl be sure what his partner thinks of be sure what his partner thinks of his verbal act and whether or not Sp2 will agree.

On the basis of such an elementary action frame, we can proceed to a minimal model of dialogic interaction (see figure 3). Speaker 1

Speaker 2

Ist move

2"1 move

ISPA

ReSPA

initial speech act

reactive speech art

FigureS.

Franz Hundsnurscher

64

In his initial speech act, Spl exposes his communicative goal and expects some reaction by Sp2. Sp2, in fact, will have the choice between different reactions and will normally act according to his or her own interest; the clear case of reaction will be yes or no, to what has been said. Thus, I prefer to put it, he will either give a positive or a negative reply concerning Spl's intention (see figure 4). Spl

Spl

pus. R. ISPA

n*g·

Figure 4. A positive reply is, of course, what Spl is aspiring to attain, but, as Sp2 will normally act in his own interest, he is the one to decide and, in case he does not agree with Spl, Sp2 will giv e a negative reply. This is a model of minimal dialogue structure and it gives us the basic unit of communicative interaction (i.e., the coherent speech act pair consisting of an initial speech act [ISPA] and a response [ReSPA]). The pair with a positive reply in the ReSPA-position can be regarded as a closed sequence because Spl 's expectations have been fulfilled by Sp2's consent; I call this a well-formed dialogue sequence. As elsewhere, in grammatical coherence and well-formedness are the basic principles of dialogue structure as well as of syntax. The pair with a negative reply in the ReSPA-position is an open sequence because no accord among the speakers has been attained. Spl' s expectations with respect to Sp2's cooperative support have not been fulfilled. At this point of the interaction, Spl has to decide whether to content himself with having failed to gain Sp2's consent and thus abstaining and forsaking his original communicative goal, which means resignation, or to stick to his goal and try again to win Sp2 over, which means insisting (see figure 5). Sp 1

Sp 2

Sp 1

Resign

Figure 5.

#

The Principles of Dialogue Grammar

65

It is easy to see what the basic underlying structure for dialogic sequence formation is: Resignation on Spl's side can fact be interpreted as consenting to Sp2's negative reply, and this will also yield a well-formed sequence because there is consent at the end. Although Spl did not succeed, Sp 2 did in securing his communicative goal (i.e., in making it clear that he does not agree with Spl). If, on the other hand, Spl insists in the third move, the sequence again is open because insisting is, in fact, a negative reply to Sp2's negative reply; thus, we are left with an open sequence, and Sp 2 has to decide again how to proceed. This could go on ad inflnitum, but conversational conventions will bring about a stop somewhere, depending on Spl's stubbornness and Sp2's patience. Insisting behavior extending over more than seven moves will normally be looked upon as impolite and childish and will lead to problems of relationship (see figure 6). Spl 1st move

Spl »eve

3rd mov«

Sp2 4* mov«

Spl 5* move

Sp2 6* mov«

Spl 7* mov«

(Break off)

Figure 6.

We could call this the base component of dialogic deep structure, and as such it is still far removed from authentic dialogic interaction in many respects. The main problem for a methodology of dialogue analysis is how to enrich this very plain and abstract structure in order to be able to catch all the relevant aspects of authentic conversation and account for them in a systematic way. In this context, I can point out only a few aspects of the program of dialogue analysis and dialogue grammar. The first point is the implications of illocution specification, or the initial speech act. As a preliminary approach, the global types of illocutionary forces as worked out by Austin (1962) and Searle (1969) will do, and it can be shown that this has immediate conse-

Franz Hundsnurscher

66

quences for the specification of the corresponding reactive speech acts in the second move, and we will have ask what the respective illocutions of the positive or negative replies will turn out to be. We can get some clues by considering the vernacular vocabulary (see figure 7). ISPA

ReSPA (pos. R.) agree, assent

Representative e.g., stating (neg, R.) disagree, contradict (pos. R.) comply with, obey Directive e.g., ordering (neg. R.) refuse, object (pos. R.) accept, trust Coin missive e.g., promising (neg. R.) decline, reject (pos. R.) shew sympathy, approve Expressive e.g., lamenting rejoicing

(neg. R.) scoff, disapprove

Figure 7.

N.B. Declaratives are, because of their (institutional) set-up, not supposed to initiate dialogues; they either succeed in making something valid or they fail in this. Each of the five global illocutions comprises an extensive class of illocutionary subtypes as can be seen with representatives where we have to distinguish between stating, maintaining claiming, pretending, presuming, guessing, predicting, projecting, etc., or with directives where we have commanding, ordering, urging, instructing, admonishing, advising, asking, begging, etc. Specified illocutions are bound up with special reactions and with special utterance forms, so there is great diversification even among the first two moves in a conversation already. The second point lies in the fact that positive and negative reply as the only alternative in the second move over simplification, because there are a number of move-possibilities between yes and no. I shall give a few examples. Instead of giving a definite reply as an answer, one may have reasons not to tie oneself down exactly, but to leave things open. One can do this in at least three ways:

The Principles of Dialogue Grammar

67

(l)By postponing one's decision (a) We are going fishing next weekend; you may join us. (b)Fll have to ask my wife first, before I can decide. (2)By avoiding a strict, binding decision altogether (In recent German, this has even led to the coining of a new word, jein, which tends to be used more and more frequently) (a) I think Chancellor Schröder is doing a good job. (b)Jein, he is not bad; but at least half of what he does is downright bungling. (3)By launching a counter-initiative speech act in the second move (a) Let's go to the cinema. (b) Why don't we just go for a stroll on the beach!

At first sight, all these alternatives can be considered as negative replies (as Schiller would say, "Der Andere hört in allem nur das Nein."), but they differ from definite negative replies in that they demand special treatment and leave open the possibility of being transformed into a positive reply, especially in the first case. If the condition put forward Sp2 in the second move is fulfilled, Sp2 is bound to consent in the end. In the second case, consent might be secured by modifying the initial statement in the third move: (c) You are right; he does make some faults at times.

The third case is somewhat different; in fact, it amounts to initiating a different game from the one suggested by Spl and this is an attempt by Sp2 to take the initiative upon himself instead of just deciding on Spl's proposal (see figure 8).

"

Postponing Def| Avoiding Dec] Counter-Initiative move I

first move

alternatives of second move

Figure 8.

The interesting thing about a refined treatment of the second move is that it opens up new and diversified sequences, as in figure 9.

Franz Hundsnurscher

68 (Move 1): Dad, can I use your car1} (Move 2): No, why don 'tyou take your own7 (Move 3): It's broken down. (Move 4): Well, in that case, okay; here are the keys. Spl move 1

Sp2 move 2 (doable move)

ESPA

provisional neg. R. postponing decision by back-inquiry

Sp3 move 3

Sp4 move 4 (doable move)

posR

answer

accepting answer

Figure 9.

The choice of a special reactive speech act gives rise to enlarged sequence structures as seen in the example. Starting from illocution-specified initial speech acts, we can construct illocutionspecific dialogue sequences. In case the ISPA is a representative speech act (e.g., a statement); then, if there is disagreement between Spl and Sp2, the normal dialogical procedure is argumentation. If we take Toulmin's (1969) "Scheme of Argument" we can see that Toulmin gives a strictly monologic version of argumentation which he sketches as in figure 10. So,Q,C (qualifier) (claim)

D (datum)

(Toalmin, p. 101) Since W (warrant)

Unless R (rebuttal) (Unless) Both his parents were aliens/ He has become a naturalized American

Figure 10. This is, of course, a way of presenting the syllogistic monologue. Looking at argumentation from a dialogic angle, however, one can start either with the claim or with the datum and unfold the sequence as in figures 11 and 12:

69

The Principles of Dialogue Grammar First version:

claim

neg. R.

INSIST

neg. R.

"Harry is a British subject. "

"No, I don Ί belitve it. " "How do you know?"

"But it is a fact that he is born in Bermuda. **

"So what?"

PERSIST

Give in

"A man born in Bermuda wiU generally be a British subject."

"Oh, I didn't know that. Well, then he is of course a British subject."

Figure 11. Second version:

datum

neg. R.

INSIST (claim)

neg. R.

"Harry was born in Bermuda."

"So what?"

"Well, a man born in Bermuda will generally be a British subject."

"But there will be exceptions to this general rule. "

PERSIST

Give in

"Of course there are: if a man fs parents were aliens or if he has become a naturalized American meanwhile."

"I see; Harry like most people bom in Bermuda will presumably be a British subject."

Figure 12. A radical interpretation of these examples would be that INSIST-structure underlies all types of dialogue. There have been attempts following the Aristotelean tradition of logic and rhetoric to reduce all types of discourse to underlying argumentational structure (cf.

Franz Hundsnurscher

70

Bien 1972), but I think that all discourse, seen in a general dialogic perspective, can in fact be traced back to an underlying INSISTING-sequence structure within a broad concept of INSISTING that is, namely, not consenting in the second, third, or following move. There is some plausibility for the syllogistic argument approach to discourse as long as we have to deal with representatives. But there are other types of discourse that are connected with other illocutionary types (e.g., directives, commissives, and expressives) and they call for a different type of treatment. A directive ISPA will give rise to a type of dialogue structure that may better be characterized as an enforcing sequence. The problem here is not as in the case of a representative ISPA where the problem is how to get Sp2 to believe in what he has been told; rather, with a directive ISPA, the problem is how to get Sp2 to be prepared to do what he has been told to, and the sequence here is kept moving with a negative reply in the second move by Sp2. Besides threatening Sp2 by announcing negative sanctions or bribing him with offers, there are a number of more subtle ways to coax someone into doing what he was not willing to do at first, either by pointing out the necessity of doing things in order to avoid detriment or to gain an advantage, or by appealing to Sp2's sense of duty, solidarity or sympathy (see figure 13). Spl move 1

Sp2 move!

- »dmimiiiation - managemcM/ directorial board - etc

Figure 2.

conle«! «f reception 2, )

pniwiptes at iheatncal lent COIKIilUtHMI

τ

r

tight·. optical·. towtd· acouitK . ckOric ' haptical electro, seme* Π»»ΙΚΙΗΓ 1 1 » waves Ctlarilttis tfiuft

τ

icon »ymbol indet impute

stgm

τ

verbal-, paracommuflieaior. verbal, / a t tocM mbtcfl non*«rtal code» » „i— — . ipeaaton v^ tadrs ^ commumcatof ai hypothctkal actor

theatrical sign repertoire« bodily «prcukm. facial cxprcumn / miming, geuurn, voice, speed), intonation, ipeech tempo, para-verbal cod«, coitumes. make-up / nvnk. hairdreitinf ι hair-oo. «ige architecture, «tenery ' »et itage-piciure. decoration, properties, protection, towdtpcakcrs. li|ht, mtntc, sound. '~ The crowd is wailing . [THE BABY is silent] PEASANT WOMAN. Now, what is this? Just when You ought to cry, you don 't\ I'll show you! Here comes the bogie-man! Cry, naughty child! Throws the baby on the ground. THE BABY shrieks PEASANT WOMAN That's right, that's right!

3. Meta-Theme

In a literary dialogue, meaningful combinations of speeches are repeated sometimes at a substantial distance with no perceptible link between them. The distance between the iterated speeches (if measured by time or space, physical or textual) is great enough to preclude any possibility for the collocutors to connect the speeches and perceive the iteration. In this case, D-tension on the level of the speakers is absent. However, the reader who has access to the whole text is able to grasp the missing link. If the connection between the distanced speeches is not lost on the reader, then a meta-theme arises and, unknowable to the collocutors, the D-tension, or rather meta-D-tension, is launched on a higher level. Beyond the comprehension of the speakers, inaudible to them, the meta-theme is addressed by the author directly to the reader. The intention of the author is transmitted by the subtle use of distant iterations. The meta-theme consists of syntactically structured iterations of sentences or phrases, with minimal modifications. It is characterized by a set of features which contains all the markers of implicitness (its substance is deduced, the second speaker is the other, the iteration is at a great distance, direction is irrelevant, and D-tension is none). As a result, the meta-theme is strongly implicit. To show the workings of the meta-theme, let us consider three more examples. A deeper proposition ' The person X is eager to know the secret thoughts of the other person F is repeated in the text of "Boris Godunov" three times, in three synonymic manifestations, at a great distance. The distant occurrences of the invariant proposition are obviously disjoint:

120

Maria Langleben

they are not linked on the surface, they are pronounced by different speakers in different situations and contexts. (12)

(a) Gregory, a young monk, and future Impostor, wakes up in his cell at the Chudov Monastery, and watches Pimen, an old monk-chronicler, writing. GREGORY. I love the peaceful sight, When, his calm soul deep in the past immersed, He pens his chronicle. Oft have I longed To guess what 'tis he writes of. Is' t perchance The dark dominion of the Tartars? Is it Ivan's brutal executions? Of the tempestuous Council of Novgorod?

(b) Two scenes later, at the palace of the Czar in Moscow, two courtiers are talking: 1st COURTIER. Where is the sovereign? 2nd COURTIER. In his bed-chamber He's closeted with some magician. 1st COURTIER. Ay, that's the kind of intercourse he loves: Magicians, sorcerers, and fortune-tellers, Ever he seeks to dip into the future, Just like some pretty girl. Fain would I know, What is it he is seeking for? 2nd COURTIER. Here he comes. Would you ask him? (c) Four scenes later, during the nocturnal tryst at Sambor, Marina Mniszek addresses The Impostor Dimitry (formerly Gregory): MARINA. No need for words. I well believe Thou lovest; but listen: with thy stormy, doubtful fate I have resolved to join my own; but one thing, Dimitry, I require: I claim that thou Disclose to me the secrets of thy soul, Thy hopes, thy plans, even thy fears

In the above three excerpts, one can find the following statements, unconnected to each other, but similar: (l)Gregory says about Pimen: Oft have I longed to guess what tis he writes of. (2) A courtier says about Czar Boris: Fain would I know, what is it he is seeking for? (3) Marina says to Dimitry (Gregory): I claim that thou disclose to me the secrets of thy soul.

The three statements are leading to the same deeper invariant, ''The person X is eager to know the secret thoughts of the other person Γ . Thus, a single recurrent idea, a wish to

On the Explicitness of Themes

121

reveal the inmost thoughts of the other, becomes a meta-theme passing through the whole drama. One may think that the request is repeated only superficially, that it is random and aimless and, consequently, the suggested invariant may be discarded as spurious. But one can also conjecture that, by this formal repetition, Pushkin wished to draw the attention of the reader to something important. The author's intention is palpable in the immediate sequel of each request, similar in the three cases: the secretive person does open his mind. One after another, the fatal secrets are disclosed, and one finds out that, under different disguises, they all are aimed at the heart of the matter: the murder of Prince Dimitry: Pimen, in his chronicle, charges Boris with the murder of a child Prince, the true Dimitry; Boris charges himself with the murder of Prince Dimitry; Dimitry-Gregory is a false pretender, not a Prince, the true Dimitry.

The drama is united and invisibly ruled by the crime that took place beyond the plot, six years earlier.6 The pivotal event is introduced by various means, including the carefully crafted, latent meta-theme. The iteration of the same proposition serves as one of the capillaries by which the author's message clandestinely permeates the text.

References Brown, Gillian/Yule, George (1983): Discourse analysis. - Cambridge: Cambridge University Press. Dascal, Marcelo (1992): On the pragmatic structure of conversation. - In: John R. Searle, Herman Parret, Jef Vershueren (eds.): (On) Searle on conversation, 35-56. Amsterdam/Philadelphia: John Benjamins. Fries, Peter H. (1983): On the status of theme in English arguments from discourse. - In: Jänos S. Petöfi, Emel Sözer (eds.): Micro and macro connexity of texts, 116-152. Hamburg: Buske. Fries, Peter H. (1986): Language features, textual coherence and reading. - Word 37, 13-29. Hasan, Ruqaiya (1984): Coherence and cohesive harmony. - In: James Flood (ed.): Understanding reading comprehension, 181-219. Newark, DE: International Reading Association, lakubinskii, Lev P. (1923): dialogicheskoi rechi (On dialogue'). - In: Lev V. Shcherba (ed.): Russkaia rech'. Collection of papers, 96-194. Petrograd. Karamzin, Nikolai M. (1845/1989): Istoriia Gosudarstva Rossijskogo ('History of Russia') (5th ed.). - Moskow: Kniga. Kireevskii, Ivan V. (1832): Obozrenie russkoj literatury za 1831 ( review of Russian literature in 183 ). - In Leonid G. Frizman (ed.): Evropeets: Zhurnal Kireevskogo, 79-95. Moscow: Nauka. A contemporary critic of Pushkin, Kireevskii (1832/1989: 86), in his perceptive appraisal of "Boris Godunov", wrote: "The shadow of murdered Dimitry is reigning in the tragedy from the beginning to the end, it rules over the course of events, binds together all persons and scenes." See also Vinokur (1935: 456-457).

122

Maria Langleben

Langleben, Maria/Langleben, Daniel (1995): Iterational structures in anomalous texts. - In: Brita Wärwik, Sanna-Kaisa Tanskanen, Risto Hiltunen (eds.): Organization in discourse. Proceedings from the Turku Conference, 329-340. Turku: University of Turku. Langleben, Maria (1998): Thematic development in literary dialogue. - In Svetla Cmejrkovd, Jaja Hoffmanovä, Olga Müllerova, Jindra Svetlä (eds.): Dialoganalyse VI, 345-354. Tübingen: Max Niemeyer Verlag. Pushkin, Alexander S. (1935): Boris Godunov. - In: Complete works, vol. 7, 1-98. Academy of Sciences of SSSR. Pushkin, Alexander S. (1937): Boris Godunov [trans. Alfred Hayes]. - In: Avrahm Yarmolinsky (ed.): The works of Alexander Pushkin, 333-411. London: The Nonesuch Press; New York: Random House, 1937. Searle, John R. (1992): Conversation. - In: John R. Searle, Herman Parret, Jef Vershueren (eds.): (On) Searle on conversation, 7-29. Amsterdam/Philadelphia: John Benjamins. Searle, John R./Parret, Herman/Vershueren, Jef (eds.) (1992): (On) Searle on conversation. Amsterdam/Philadelphia: John Benjamins. Tomashevsky, Boris (1925): Theory of literature. - Leningrad: Gosizdat. Vinokur, Gregory O. (1935): Boris Godunov (a commentary). - In: Complete works, vol. 7, 385-503. Academy of Sciences of SSSR.

Lenke Nemeth

A Subjectivity Formation Model for Dialogue Analysis in Drama

This chapter offers a theoretical approach to the analysis of discontinuous, elliptical, and repetitive dialogue mostly occurring in postmodern dramatic texts. Described as symptomatic of communication failure leading up to aborted human relationships, this kind of dialogue has been studied effectively with conversation analytical methods (i.e., Gricean [Grice 1975] maxims, turn-taking system). Limited to the mere identification of the characters' communicative strategies, these methodological traditions tend to fail to substantiate reasons for the characters' miscommunication and/or non-communication. Premised on the assumption that the lack of dialogical self causes the characters' inarticulacy, the proposed subjectivity formation model, prompted by Mikhail Bakhtin's concept of identity, can be effectively applied to the study of the mode of (de)constructing the self. To illustrate the operation of the model, I will supply brief analyses of dialogue stretches from "Sexual Perversity in Chicago" (1974) and "Oleanna" (1993) written by the Chicago-bom playwright David Mamet (1947-). Postmodernism as a cultural and intellectual movement has radically transformed our notions about the limitations of the genre-defining elements of drama (character, plot, and dialogue). As a result of the disruption of the linear narrative and the effacement of the plot design, dialogue functions as the only means of character portrayal in the postmodern dramatic texts of American playwrights who have come to the literary scene in the past thirty years. The question arises, however, in what way the discontinuous and elliptical dialogue, stripped off its extra-communicational narrative aspects can reveal the inner world of the characters. The theoretical approach proposed in this chapter combines linguistic and literary techniques which account for the subtext in the selected dramatic exchanges. Discontinuous, distorted, and nonsensical as the Mametian dialogue may seem at first glance, it features as a crucial means of characterization. It's highly fragmented, elliptical, and repetitive nature has invariably been stated as indicative of the characters' lack of communication and/or miscommunication. Conversation analytic methods (cf. Macbeth this volume; Varenne this volume) can effectively be used to examine the exact mode of how patterns of conversational strategies are at work, how they are negotiated, and how certain, generally accepted conventions are flouted leading to a communication failure, miscommunication or non-communication. The ensuing brief analysis of a segment of dialogue between the two protagonists, the professor and his student in the opening scene in "Oleanna" (1993), aims to provide sound

124

Lenke Nemeth

linguistic evidence for the miscommunication apparent in their talk. However, in order to describe the particularities in the characters' talk more thoroughly, I find it necessary to introduce the categories of what I term meta- and diversion sequences. They are embedded sequences that do not forward the conversation, but hold in abeyance the answer either via meta-sequences (i.e., the repetition throughout a few utterances of what has already been said), or via diversion sequences (i.e., deliberate digressions from a topic introduced earlier). For easy reference the turns are numbered. Extract 1 (1) JOHN. I'm sorry ... (2) CAROL. (Pause) What is a term of art? (3) JOHN. (Pause) I'm sorry . ..? (4) CAROL. (Pause) What is a term of art? (5) JOHN. Is that you want to talk about? (6) CAROL to talk about. . .? (7) JOHN. Let's take the mysticism out of it, shall we? Carol? (Pause) Don't you think? I'll tell you: when you have some thing. Which must be broached. (Pause) Don't you think . . .? (Pause) (8) CAROL. .. . don't I think .. .? (9) JOHN. Mmm? (10) CAROL.... did I . . . ? (11) JOHN.... what? (12) CAROL. Did . . . did I . . . did I say something wr . . . (13) JOHN. (Pause) No. I'm sorry. No. You're right. I'm very sorry. I'm somewhat rushed. As you see. I'm sorry. You're right. (Pause) What is a term of aril It seems to mean a term, which has come, through its use, to mean something more specific than the words would, to someone not acquainted with them . .. indicate. That, I believe, is what a term of art, would mean. (Pause) (Mamet 1993: 4) The Gricean (1975) maxims (cf. Berlin this volume; Cole and Morgan 1975; Fetzer this volume) can adequately categorize the types of communicational strategies employed by the Mametian characters. In this scene, all the four Gricean maxims are flouted: the professor answers the student's question in (2) only in (13), after a series of embedded utterances including meta-sequences (4)-(5) and (8)-(12) as well as a diversion (7). Another brief segment of dialogue between the young couple, Dan and Deb in Mamet's "Sexual Perversity in Chicago" (1974: 52) also shows similar patterning. Extract 2 DAN and DEB's apartment. The morning. They are each getting ready for work. (1) DANNY. Do we have any shampoo? (2) DEB. I don't know. (3) DANNY. You wash your hair at least twice a day. Shampoo is a staple item of your existence. Of course you know. (4) DEB. All right. I do. Know. (5) DANNY. Do we have any shampoo?

A Subjectivity Formation Model (6) (7) (8) (9) (10) (11) (12) (13) (14)

125

DEB. I don't know. Is your hair dirty? DANNY. Does my hair look dirty? DEB. Does it feel dirty? (Pause) It looks dirty. DANNY. It feels greasy. I hate it when my hair feels greasy. DEB. Well, I'm not going to look. If you want to know if there's any shampoo, you go look for it. [. ..] DANNY. I'm going to wash my hair. Is there any shampoo? DEB. Yes. And no. DANNY. Now what's that supposed to mean? DEB. Everything. And nothing.

The pattern arising here is as follows: (5)-(10) are meta-sequences followed by diversion, in which a new topic is introduced (omitted from the extract), while (11)-(14) return to the original topic. Within the confines of the chapter, further examples cannot be provided, yet based on an extensive study of dialogue stretches from various Mamet's plays, it is legitimate to conclude that the dialogue exchanges evolving in these plays are typically Mametian: the characters exhibit mutually alienating strategies rather than mutually supportive ones (Nemeth 1995). Usually, the following pattern emerges: a topic is introduced, then it is repeated and confirmed, next a diversion topic is inserted, and finally, the original topic may recur and some kind of answer may be provided. Evidently, the characters exhibit mutually alienating strategies rather than mutually supportive ones. A deficiency of this kind of descriptive method, however, lies in the fact that it fails to account for the characters' unresponsiveness and inarticulacy. It seems to be fairly easy to conclude that the characters' conversational failure as well as their inarticulacy and isolation, can be attributed to the Mametian agents' uncertainties. Yet, the question arises whether it is possible to substantiate the somewhat impressionistic claim with linguistic evidence. I find that the Russian philosopher of language, literary scholar, and theorist, Bakhtin's (1995/1996) system of thought provides an adequate theoretical input and methodological point of entry into Mamet's texts. In what follows, I will discuss in what way Bakhtin's insights pertaining to language as uttered discourse and his subjectivity formation model prove to be most helpful for the approach I am proposing. Bakhtin's tripartite model of language that assigns equal significance to each element of a verbal transaction - speaker, listener, and context - bears special relevance to literary analysis (cf. Hess-Liittich this volume; Langleben this volume). By comprising formal and contextual aspects of language, Bakhtin underscores their intrinsic unity as well as points of intersection between these entities. Bakhtin's (1996: 279) concept of internal dialogism of the word bears particular relevance to understanding the forces and elements shaping the style of the characters' Ian-

126

Lenke Nemeth

guage in a literary work. Dialogue is not simply "a compositional form in the structuring of speech," but conceived as "the internal dialogism of the word [...] that penetrates its entire structure, all its semantic and expressive layers, [which] has such a power to shape style [and that it finds expression] in a series of peculiar features in semantics, syntax and stylistics." Thereby, discourse is shaped by a dual process: first, "the word, directed toward its object, enters a dialogically agitated and tension-filled environment of alien words, value judgements [. . .] (276); second, "every word is directed toward an answer and cannot escape the profound influence of the answering word that it anticipates" (280). In a living conversation, "the contradictory environment of alien words is present to the speaker not in the object, but rather in the consciousness of the listener, as his apperceptive background, pregnant with responses and objections [and] every utterance is directed toward this apperceptive background of understanding, which is not a linguistic background but rather one composed of specific objects and emotional expressions" (281). In Bakhtinian terms, the adequate operation of the apperceptive background is the key to conducting meaningful conversation. Improper or dysfunctional internal dialogism may occur when the speaker encounters passive understanding, when "nothing new can be introduced into his discourse. [Consequently,] the purely negative demands, [. . .] leave the speaker in his own personal context, within his boundaries" (Bakhtin 1996: 281); in other words, the speaker remains in his solitary confinement, which inevitably results in his/her inarticulacy, unresponsiveness, or in certain instances it leads to even violence. As the exchange between Dan and Deb demonstrates, their budding relationship is much thwarted by the malfunction of Deb's apperceptive background. Further rationalizing the exchanges, it is legitimate to claim that if dialogue as the means and the route of gaining knowledge about the world and each other ceases to operate between characters, it leads to an undialogical relationship between them, which derives from the characters' undialogical selves. Alternately, dialogue that retains its functional validity serves as a route, a means to construct a coherent self, which is dialogical (i.e., capable of a dialogical relationship). In light of this reasoning, the Mametian dialogue, discontinuous and repetitive, most often devoid of any meaningful exchange of ideas and often replete with aggressive, obscene outbursts, testifies to the absence or the malfunctioning of the characters' coherent selves. In order to understand the causes of the Mametian characters' inability to build a meaningful relationship with each other, next, I will turn to consider factors that constribute to the construction of a coherent self.

A Subjectivity Formation Model

127

I propose subjectivity formation model: Self -> Other-» Self This model is inspired by and abstracted from Bakhtin's (1995) notions of self and other as explicated in his essay "Author and Hero in Aesthetic Activity" (originally published in 1920/24); it illuminates factors determining the meaningful and functionally valid construction of the self. The simplicity of the model underlies the process-like nature of identity formation. It posits that defining and positioning the self in the external world of actuality is unthinkable without a relation of self to other, and, conversely, of other to self. Bakhtin's (1995: 4) views on identity constitution are "couched in aesthetic terms" in his essay, thus his understanding of an "aesthetically productive relationship" between the author and hero serves as a basis for the proposed subjectivity model. Bakhtin defines an aesthetic activity proper in terms of a kinetic movement between the author and hero: the author does not view or consider his/her hero from a fixed position; on the contrary, he/she moves into and away from his/her hero. This relationship can be diagrammatically illustrated as follows: Author -> Hero -> Author

Second, Bakhtin (1995) maintains that "the organizing power in all aesthetic forms is the axiological category of the other, the relationship to the other, enriched by an axiological 'excess' of seeing for the purposes of achieving a transgredient consummation" (189). In a similar fashion, constructing the self can only be conceived as a process, since the properly functioning self should constantly move into and away from the other. Hence the diagrammatic representation of the initially proposed model should be refined to the pattern illustrated below - to suggest the infiniteness of the process. Self -> Other -» Self -> Other -* Self...

By analogy, in the interpersonal nature of relationship between the self and the other it is only through the other's excess of seeing that the self can position himself/herself in the world. In order to be able to find and define his/her being in the world, the self has to position or project himself/herself in the place of the other, and virtually, has to adopt the other's perspective. The self can experience as well as place, and enrich himself/herself in the external world only in and through the category of the other. Thus, constructing the self (and its satellites) is always a result of a social comparison process, which is realized through dialogue, or more precisely, through a dialogical relationship between the self and

128

Lenke Nemeth

the other. Thus, dialogue, or the principle of dialogicity, has a crucial role in building the self. In contrast with a purely theoretical subject removed almost entirely from the multifariousness of reality, Bakhtin theorizes it as a process and takes into account theoretical and practical considerations in his treatment of the subject. In Bakhtin's understanding self must experience reality, and participate in the eventness of the present moment. Bakhtin argues for, what I would term dialogized gender, thus creating an interface between the two. The prerequisite of the formation of subjectivity is a dialogic relation between self and the various competing discourses. Conversely, the interdependence existing between subjectivity and the network of discourses proceeds from their dialogic relation. As the examples cited here demonstrate, the subjectivity/identity formation paradigm emerges in a truncated or distorted version in Mamet's plays. The dialogical relationship between the self and the other, which is essential for the meaningful and functionally valid construction of the self, ceases to operate or is objectified in a somewhat fragmented and distorted version. Mamet diagnoses a malaise of contemporary American society, the lack of the dialogical self. The self appears in a distorted or truncated version: either 1) the self is unable or unwilling to project himself/herself into and experience the other; or 2) the self adopts the perspective and the discourse of the other in order to perpetuate that state of affairs. The first version typically applies to women characters in the private realm of life (Deb), whereas the second to women protagonists appearing in the outer world (Carol). In both cases, and on the levels of self-definition and identity construction, the result is a failure of the protagonists to find and define their selves in the male-dominated outer world. Linking the purely formal description of the dialogue with its contextual aspects with the help of some of Bakhtin's concepts allows for disclosing subtextual elements in the dialogic exchanges of Mamet's characters. By presenting the crucial consequences of the lack of a dialogical self, namely, utter isolation, I believe, Mamet offers a remedy for the malaise: he affirms the necessity of a dialogical relationship with the other, which is, in the broadest sense, a promise of relief from the isolation.

References Bakhtin, Mikhail M. (1995): Author and hero in aesthetic activity. - In: Michael Holquist, Vadim Liapunov (eds.): Art and answerability. Early philosophical essays by M. M. Bakhtin, 4-256. Austin: University of Texas Press. Bakhtin, Mikhail M. (1996): Discourse and the novel. - In: Michael Holquist (ed.): The dialogic imagination, 276-281. Austin: University of Texas Press.

A Subjectivity Formation Model

129

Cole, Peter/Morgan, Jerry L. (eds.) (1975): Speech acts. Syntax and semantics, vol. 3. - New York: Academic Press. Grice, H. Paul (1975): Logic and conversation. - In: Peter Cole, Jerry L. Morgan (eds.): Speech acts. Syntax and semantics, vol. 3, 41-58. New York: Academic. Mamet, David (1974): Sexual perversity in Chicago. - New York: Weidenfeld. Mamet, David (1993): Oleanna. - New York: Vintage.

Lawrence N. Berlin Grounded Theory and its Benefits for Dialogue Analysis: The Beginning of a Beautiful Friendship

1. Introduction

Grounded Theory (GT) is a methodology that employs an inductive, interactive approach toward the building of theory grounded in the data being examined. Its benefit to the investigation of dialogue phenomena, whether naturally occurring or scripted, lies in both its procedural and conceptual dimensions. In its procedure, GT is iterative and interactive, enabling the dialogue analyst to navigate between linguistic and contextual features toward the eventual creation of theory, utilizing a combination of tools. In so doing, GT resolves a shortcoming that has often been asserted with regard to dialogue analysis: a linguistic analysis alone is insufficient for developing a complete understanding of dialogue (cf. Fairclough 1995; Kumaravadivelu 1999; Weigand 2002). With regards to the latter dimension, GT proceeds from the same concept-building stance as dialogue itself. As a speaker or writer attempts to give voice to ideas, the grounded theorist analyzes dialogue inductively to let the ideas speak for themselves, striking a balance between the spoken or written word and the context in which it occurs. In this paper, I will present examples of the various stages of GT analysis and its power to address some of the current challenges confronting scholars in dialogue analysis.

2. Background

Grounded Theory was first introduced in the 1960's by two sociologists working at the University of California at San Francisco, Glaser and Strauss (1967). Glaser completed his graduate studies at Columbia University under the tutelage of Lazarsfeld, "known as an innovator in quantitative methods" (Strauss and Corbin: 1998b, 10). Strauss received his degree at the University of Chicago and was influenced by interactionist and pragmatist writers of the time. Together, their brainchild, GT, worked to resolve philosophical contra-

132

Lawrence N. Berlin

dictions that existed within their understanding of what it meant to do research. The team was confronted with the a positivist paradigm which validated quantitative research that was essentially hypothesis-testing, never aimed at generating new theory, but merely attempting to verify former theories and account for counterexamples when they were encountered in data. Glaser and Strauss decided to pursue a path that aimed at the heart of true inquiry, a recognition that innovation comes about not through replication, but through creation. Essentially, in GT, "data collection, analysis, and eventual theory stand in close relationship to one another. A researcher does not begin a project with a preconceived theory in mind [...] Rather, the researcher begins with an area of study and allows the theory to emerge from the data" (Strauss and Corbin 1998b: 12). Thus, the interactive and iterative analytic process of GT combines a more traditional analysis of linguistic data; yet, through a thick description of the situational components (Geertz 1973; cf. Varenne this volume), GT leads to a broader understanding of the phenomena under investigation by situating the language within its broader context, and verifying its ultimate meaning through an investigation of synchronic and diachronic influences (cf. Maurer-Lausegger this volume).

3. Rationale

There are two major reasons for using GT. First, it is capable of incorporating both quantitative and qualitative paradigms, leading to the eventual integration of text and context into a fuller theory of dialogue analysis. Second, and consequent to the first, it is aimed at theory-building rather than merely hypothesis-testing. This allows for the description and eventual explanation of phenomena and overcomes the problem of trying to fit data into earlier taxonomies. In other words, it compensates for shortcomings in prescriptive approaches by accounting for all the data and enabling innovative and grounded responses to dealing with phenomena that lie outside earlier frameworks. Traditionally, a quantitative approach had long been deemed the most acceptable way to conduct systematic inquiry, with quantitative researchers often attempting to discredit qualitative research as being less than scientific. In a recent talk, however, Agar (2004) outlined extant parallels between established science, specifically nonlinear dynamic systems (NDS), and ethnography. He expressly mentioned NDS' adaptive nature, making use of feedback, changing with circumstances, and being sensitive to conditions. These probabilities and conditions (or dimensions, as they are called in GT) are part and parcel of the theory-building process.

Grounded Theory and its Benefits Qualitative Credibility

Quantitative Internal Validity

Dependability

Reliability

Confirmability

Transferability

Generalizability

133 Soundness and systematicity of collection and analysis procedures Contextualization/description enabling replicability across studies Verification process through comparison of internal consistency within a data source, across data sources (triangulation), and among auditors and participants Establishment of compatibility with other participants, places, and situations

Table 1. Comparison of Paradigmatic Trustworthiness/Robustness GT provides an approach compatible with mixed studies. Though initially conceived of within a qualitative paradigm, it does not negate the incorporation of quantitative methods (Strauss and Corbin: 1998a, 1998b), but endeavors to reconcile the two paradigms through its establishment of a framework of trustworthiness (see table 1). Thus, while the procedure insists that the initiation of the process excludes the imposition of earlier findings thereby freeing the researcher from preconceived notions and enabling the data to speak for themselves - it does not expect that systematic inquiry will proceed in a vacuum; indeed, GT - as with any legitimate approach to undertaking research - requires dialogue analysts to familiarize themselves with relevant scholarship that may feed into their ultimate theory. Aside from its potential capacity to overcome the paradigmatic wars, GT has other strengths that make it a valuable approach to dialogue analysis. It begins with a heuristic approach rather than a narrow analytic view of research. Instead of merely hypothesistesting, or attempting to impose previous taxonomies on dialogue, it allows for the recognition of variability and the influence of context in its overall analysis (Goodwin and Duranti 1992; Firth and Wagner 1997). Halliday referred to the ultimate shortcomings of the two perspectives: linguisticoriented and context-oriented when seen as disparate. Termed logical-philosophical (or language-as-code) and ethnographic-descriptive (or language-as-behavior), Halliday (1984: 3) cited the need for "a unified 'code-and-behavior' linguistics." While he did not see the two perspectives as irreconcilable, he did claim that it was not uncommon for the two to become exaggerated to the extent that they seemed so. Indeed, Lyons' (1968 as cited in Halliday 1984) comment that "linguistic theory, at the present time at least, is not, and cannot be, concerned with the production and understanding of utterances in their actual situations of use" only provided a temporary and dissatisfactory reprieve from the

134

Lawrence N. Berlin

work of developing an inclusive theory for dialogue analysis. Speech act theory (Austin 1962; Searle 1969), still in its infancy at the time, was already striving to position utterances within a larger framework of performance, expanded by Gordon and Lakoff s (1975) conversational postulates, Grice's (1975) maxims, and Brown and Levinson's (1978) theory of politeness. While these contributions led to a deeper understanding of the code essentially pursuing a bottom up approach - they were not ultimately complete, suggesting that language phenomena could be interpreted devoid of contextual considerations. Context, defined by Schegloff (1992) as comprising both external and intrainteractional types, began to broaden our comprehension of dialogue beyond the code. Claiming that these two types of context would lead to a complete theory of social interaction, they did not completely satisfy dialogue analysts as they took a more top down look at dialogue. This more ethnographic-descriptive stance saw context as primary, presenting "an interpretive commentary on a single highly-valued instance of language use" (Halliday 1984: 4). Nevertheless, this approach can lead to a viewpoint as reductionist as the former, where linguistic behavior appears to emerge as a result of setting and situation. Though the early stages of the analytic process of GT are seen as essentially linguistic (Strauss and Corbin 1998a, 1998b), it focuses on the interaction of language and meaning (i.e., parole and langue) in an effort to delimit units of analysis that correspond to the question under investigation and ultimately provide a level of abstraction that can apply across multiple levels of interest. The emergent themes resulting from the stages of coding enable a level of abstraction that not only easily renders the grounded theory compatible with other research and its findings, but also provides a framework through which different studies can be compared with one another. In turn, this interactive process not only reconciles the two perspectives, but also provides for the generalizability, or transferability, of findings to other contexts. The other advantage of GT is its focus on description rather than prescription. Thus, theory is derived from what speakers actually do, rather than what it is believed they should do. Its initial aim is not to prove or disprove other hypotheses and theories; instead, it starts from the data themselves to describe phenomena and their environments, rather than attempting to impose constraints on them, and sorts information to provide a domain or thematic analysis which can, in turn, serve as a taxonomy for explaining phenomena (see also Langleben this volume). This is not to say that it cannot ultimately prove or disprove an existing hypothesis or theory, but that it does not begin with a deductive approach, attempting to apply findings to new situations. As Dewey (1934: 50) said, "If the artist does not perfect a new vision in his process of doing, he acts mechanically and repeats some old model fixed like a blueprint in his mind." On the other hand, GT is open to the integration of other research findings toward the development of a theory; after all, the impetus for undertaking research never exists in a vacuum, but is informed on some level

Grounded Theory and its Benefits

135

by the body knowledge held by the researcher. With GT, however, the unfolding theory is one that is grounded in the data and their contexts, open to an exploration of the multiple factors that could influence them. Other approaches tend to focus on the analysis of an individual phenomenon or an individual constituent thereof in isolation, often producing results that must somehow be integrated into a larger theory, but which can, at times, be incompatible with studies investigating the same or presumably related phenomena. As a result, attempts at building theory in this manner are often arduous, requiring backtracking which may, in turn, result in the reduced efficacy of the original findings or complete failure to integrate them into a greater understanding of dialogue.

4. Procedures and Examples

Glaser & Strauss (1967) Data Collection Comparing Incidents

Integrating Categories & Properties Delimiting the Theory

Strauss &Corbin( 1998)

Berlin

Data Collection

Initial process of gathering data

Microanalysis

Open Coding

Axial Coding

Line by line review of data (e.g., transcripts, texts) through which initial impressions are drawn and jotted in the margins (N.B. essentially linguistic in nature, words, phrases, and their meanings are explored) Initial jottings are reexamined and categorized; categories are given labels.

Selective Coding

Preliminary categories are integrated and abstracted to higher order themes through the exploration of properties and dimensions; these themes are then placed in a matrix. Writing the Writing Interpretive process of describing the Theory model/theory Table 2. Comparison and Explanation of Grounded Theory Methodology

The analytic process combines different levels of coding by which a theory is eventually derived. As stated earlier, it is iterative, interactive, and inductive, leading to an overall trustworthiness through the establishment of the criteria of credibility, dependability, confirmability, and transferability. Glaser and Strauss (1967: 105) originally delegated four stages of GT analysis: "(1) comparing incidents applicable to each category, (2) integrating categories and their properties, (3) delimiting the theory, and (4) writing the theory." These

136

Lawrence N. Berlin

stages were later refined by Strauss and Corbin (1998b) to include microanalysis (open coding and axial coding), selective coding, and writing. These two descriptions merge into the following framework (see table 2). The decisive advantage of GT is its analytic process. In its initial phase of analysis, GT is like any other approach to commencing an examination of language data, whether naturally occurring or from texts. Described by Strauss and Corbin (1998b: 6) as a methodology that "enjoins taking with great seriousness the words and actions of the people studied," it recognizes how the close connection between the data collection and the analysis permits a finer-grained examination of linguistic, extralinguistic, and contextual features, putting it in line with traditions in sociolinguistics based on the work of Charles Ferguson (1959) and Joshua Fishman (1967) and the defining of domains, and in anthropology based on the work of James Spradley (1980) and his domain analysis. The employment of constant comparison, cited as an analytic tool used in the management and analysis of data to verify conclusions and enhance analytic validity (Huberman and Miles 1998), begins at the same time as data collection commences rather than being postponed until its completion. This enables the researcher to construct and verify hypotheses throughout the collection process, permitting control over the progression of the study through what Agar (2004) refers to as a "continual conversation with the data" (personal communication). Thus, as new data are gathered, and as data from different sources are compared with one another, original working hypotheses (which emerge from the microanalysis) are verified, modified, or discarded, and a theory emerges that is grounded in the data and context of the research. Merriam (1998: 159) gives a clear, concise account of the process involved in applying constant comparison. She explains that the basic strategy of the "method is compatible with the inductive, concept-building orientation of all qualitative research." In essence, a data source such as a transcript is reviewed several times by the researcher. During each reading, the reader may add notes, questions, and/or comments in the margins of the transcript. These impressionistic interpretations and evaluations lead to characterizations, or categorical labels, for each unit of analysis. Merriam (1998: 179) goes on to define a unit as a "meaningful (or potentially meaningful) segment of data." Lincoln and Guba (1985) propose that each unit can be defined by establishing the criteria of relevance and independence. In other words, each of the units should reveal some piece of information that is relevant to the particular question under investigation (Lambert and Ervin-Tripp 1993), and should be able to stand alone as the smallest piece of information that induces meaning independent of any other part of the data. In the process of naming each unit, the categorical labels that are established are again compared with one another to look for similarities. As these labels are sorted and grouped into properties and their dimensions, themes are abstracted for higher order categories.

Grounded Theory and its Benefits

137

Issues such as the frequency of co-occurrence of certain themes and their tendency to intersect place the researcher in a position to identify a conditional or consequential matrix which eventually will lead to the writing of theory.

4.1. Microanalysis Microanalysis incorporates two phases: open and axial coding. Both forms of coding work in a reciprocal relationship through which they inform each other. The reciprocity consists of doing a line-by-line reading to delimit units of analysis and recording initial impressions, or jottings, about the units. It is, in fact, inconsequential whether the units are defined first and then described, or whether the description suggests the units. In either case, though, the delimiting of units of analysis will ultimately depend upon the researcher, his or her specific level of interest (e.g., speech acts, turns, conversational exchanges), and the capacity of the units to meet the minimum criteria of relevance and independence (Lincoln and Guba 1985). This latter distinction, however, should not be construed as an indicator for distinct and mutually exclusive classifications of themes (cf. Langleben this volume). In fact, it is the very assignment of more than one theme per unit of analysis that contributes to the hypothesis building and ongoing verification process, as well as the development of a matrix that will serve in the creation of a theory. To demonstrate, I will be using the text alluded to in the title, the script for the American classic film, "Casablanca" (Epstein et al. 1942). More specifically, I will focus on the dialogue between two characters, Richard Rick Blaine and Louis Renault, their character development and their unique brand of friendship. Upon initial investigation, one word that is used consistently with reference to Rick is "cynical". Not only are there stage directions that present him in this manner (l)-(2), but also the word occurs in spoken lines where different characters address him directly (3)(4)(l)He [Louis] hurries away. Rick smiles cynically. (Wallis, et al. 1942: 27) (2)RICK (cynically): I should have known. (82) (3)UGARTE: You are a very cynical person, Rick, if you'll forgive me for saying so. (13) (4) RENAULT: Because, my dear Ricky, I suspect that under that cynical shell you're at heart a sentimentalist. (26)

The immediate linguistic context in the above examples further appears to solidify this hypothesis. In (2), Louis has not quite finished boasting how he is "master of his own fate" when he hurries away upon learning that Major Strasser of the Gestapo is waiting for him, which prompts Rick's response. In (3), Ugarte has just indicated the murdered couriers and

138

Lawrence N. Berlin

Rick responds by suggesting that their martyrdom has rendered them among the "Honored Dead", a reference which immediately precedes Ugarte's retort. In these early stages of analysis, the benefit of transcription cannot be emphasized enough. Researchers interested in corpus linguistics, for example, can download segments of corpora prior to or after having conducted a search for repetitive keywords, concepts, or feature under investigation. Once available in text form, any type of tagging can be employed to enter jottings, or impressionistic notations, into the margins. While this can certainly be done manually, entering them directly into a computer program will make sorting through the data later much easier. In the latter case, "Microsoft Word" or a research program, such as "Ethnograph", can be utilized for coding. Likewise, a concordance program can be used to identify units of analysis through a combination of means, including a word search function, a lemmatiser, and intuitive breaks indicated in the script itself (e.g., sentence boundaries, paragraphs). While the micro-level analysis suggested here is below the level of dialogue, it does, nonetheless, demonstrate how various tools can be utilized to render information in the preliminary generation of hypotheses in the development of GT. Addressing the level of dialogue, it first becomes necessary to establish what that dialogue should look like. Is it simply a couplet as in Flanders' (1970) initiative and response, or does it require some level of uptake on the part of the initiator to indicate completion, reassertion, or redirection as suggested by Hundsnurscher (this volume)? I would suggest that either definition can be employed in GT. In the following exchange, Louis, through direct and indirect reference, reveals his character to be that of a womanizer. In segment (5), Rick has just ordered Sacha, a bartender, to take Yvonne home, a woman who expected Rick the previous evening and whom he is now sending away, suggesting she has had too much to drink. (5)RENAULT: How extravagant you are, throwing away women like that. Someday they may be scarce. Rick sits down at the table. RENAULT (amused): You know, I think now I shall pay a call on Yvonne, maybe get her on the rebound, eh? RICK: When it comes to women, you're a true democrat. (Wallis, et al. 1942: 20)

The question remains about how we should interpret this piece of dialogue. Following Flanders, Rick sitting down becomes nothing more than a stage direction, and the second component by Renault is a continuation of the initiative. On the other hand, if Hundsnurcher's assertion is correct, then Rick's lack of response to Louis' initiative is, in itself, a nonverbal form of response that leads to the uptake or reassertion that ensues in the form of a question and direct reference to Yvonne. In such a case, Rick is obliged to reply,

Grounded Theory and its Benefits

139

but chooses to do so with yet another example of cynicism, further emphasizing the hypothesis that this is a motivating factor that drives his language use choices (i.e., the way he takes part in the language game). This initial exchange, the first encounter in the script between Louis and Rick, not only achieves a level of insight into both of their characters, but also succeeds in indexing multiple levels of context. Through hypothesis-testing - working with a preconceived notion we are able to analyze this data. Taking it at face value alone, Louis' initial reference to the possible scarcity of women could relate to the fact that they find themselves in the midst of wartime. However, GT's concept-building approach requires us to reanalyze and integrate these data into the emerging theory in light of subsequent exchanges Rick's reply - the choice of "democrat" - further hints at the multiple layers of the situation as well as his own multi-layered self (Bakhtin 1981; cf. Nemeth this volume). Rick, an American expatriot, finds himself in a region referred to on several occasions by Louis as "unoccupied France". The positioning, a type of situational irony cannot be deemed as unimportant, for while much of Europe was under the influence of the Third Reich, including Vichy France, the mention of "unoccupied France" is itself a misnomer since the characters find themselves in Morocco, a country occupied by France at the time. More importantly to the understanding of the textual references, it's a colony where the unlikely possibility for departure rests on the whims of the person in charge of issuing exit permits - Louis. Power coupled with chauvinism is further realized in the allusion to a visa problem in segment (6). (6)As lisa and Laszlo leave, an OFFICER comes in. RENAULT: Undoubtedly their next step will be to the black market. OFFICER: Excuse me, Captain. Another visa problem has come up. RENAULT: Show her in. OFFICER: Yes, sir. Renault looks at himself in the mirror and straightens his tie. (Wallis, et al. 1942: 66)

I would argue that, without having some insight into Louis' character through earlier text and understanding the situation and wider context in which the utterances occur, we might not grasp the co-reference (and double entendre) of visa problem and her beyond a superficial level. The iterative process of GT, checking units of analysis against each other to verify or modify hypotheses, helps in avoiding erroneous interpretations (cf. Langleben this volume). The text alone points to interpretational incongruities, but does not provide a cohesive reading of the data. For instance, in the following (7), should we construe Rick's remarks merely as the ranting of a drunken man? This cannot be the case since, I would say due to expediency of time, a scriptwriter would not include superfluous dialogue.

140

Lawrence N. Berlin

(7) Sam sits down at the piano and starts to play softly, improvising. (7.1) RICK: They grab Ugarte and she walks in. Well, that's the way it goes. One in, one out. Sam? (7.2) SAM: Yeah, boss? (7.3) RICK: Sam, if it's December 1941 in Casablanca, what time is it in New York? (7.4) SAM: Uh, my watch stopped. (7.5) RICK: I bet they're asleep in New York. I'll bet they're asleep all over America. (Wallis, et al. 1942:50-51)

The act of drinking to drunkenness does not appear to be part of the character's modus operandi, but instigated by something else. In line (7.1), Rick's initiative plainly states what has happened, yet his reference to the time in New York in lines (7.3) and (7.5) points to something more; it is almost an expression of regret, one that provides an insight into the more complex nature of his character (i.e., beyond the cynicism). In (8), his apparent wallowing in regret takes a different turn, one that indexes the basis for his cynical nature. (8) Suddenly he pounds the table and buries his head in his arms. Then he raises his head, trying to regain control.

RICK: Of all the gin joints in all the towns in all the world, she walks into mine. He holds his head in his hands. (Wallis, et al. 1942: 51) In the immortal line in (8), the "she" referred to - lisa - appears to be the cause of the cynicism. Unrequited love has hardened Rick's heart. This becomes apparent later in the firm during the famous flashback scene when Rick and lisa are together in Paris. But, in real time, we would not have this orientation enabling a fuller understanding of his character. In fact, the flashback scene had not been a part of the original play, "Everybody Goes to Rick's" by Burnett and Alison on which "Casablanca" was based. Neither was it included in the original script (which was incomplete, even at the commencement of filming) nor written by the original scriptwriters. It was added afterward to give a necessary context to the story. Thus, while the text alone identifies, or signifies (Fairclough 1995), the propositional semantic content, it remains insufficient in the determining the context of situation and the context of culture (Malinowski 1923). Halliday (1984: 9) sees culture and the social context of language behavior as semiotic constructs. He states: Exchanges of meaning succeed, if they do (and it is remarkable how often they do succeed), because each of those taking part has some idea of what others will mean and assumes that they have some idea of what he will mean. As speakers and listeners, we project the linguistic system on to the social system [...], interpreting verbal meanings as the expression of the meanings that are inherent in the culture.

Grounded Theory and its Benefits

141

Yet, what contrivances do we invent when the complex nature of culture or cultures presents us with dialogue that doesn't succeed? In (9.2), Louis' initiative turn can be construed as a simple request, yet it isn't perceived as such by Rick. His sharp response in line (9.3) indicates that he sees an attempt on the part of Louis to pry into his past, apparently a taboo where Rick is concerned. Louis, nevertheless, has not achieved his communicative goal, so he takes up the query again in line (9.4) and again in line (9.6) until he finally receives a response in line (9.8) (cf. Fetzer this volume; Hundsnurscher this volume). Despite its infelicitous nature (Brown and Levinson 1987), he decides to drop that line of questioning and takes up a new one in line (9.9). (9) (9.1) Rick and Renault look up when they hear the BUZZ of a plane taking off from the adjacent airfield. The plane flies directly over their heads. (9.2) RENAULT: The plane to Lisbon, (pause) You would like to be on it? (9.3) RICK: (curtly) Why? What's in Lisbon? (9.4) RENAULT: The clipper to America. (9.5) Rick doesn't answer. His look isn't a happy one. (9.6) RENAULT: I have often speculated on why you don't return to America. Did you abscond with the church funds? Did you run off with a senator's wife? I like to think you killed a man. It's the romantic in me. (9.7) Rick still looks in the direction of the airport. (9.8) RICK: It was a combination of all three. (9.9) RENAULT: And what in heaven's name brought you to Casablanca? (9.10) RICK: My health. I came to Casablanca for the waters. (9.11) RENAULT: Waters? What waters? We're in the desert. (9.12)RICK: I was misinformed. (Wallis, et al. 1942: 20-21)

Segment (9) suggests a lot about the relationship between Louis and Rick, and, perhaps more poignantly, Louis' deeper insight into Rick's character. The predicted cynical responses (9.5) and (9.8) do not speak of cynicism, but his potential inability to cope with Louis' humor. Rick's moves in this language game (9.10) and (9.12) may actually serve to put the two on level footing (cf. Goffrnan 1981); by engaging in this type of flippant language behavior, Rick is stepping outside of the status differential (i.e., prefect of police civilian) and re-engaging their equal footing, yet he achieves this by flouting conversational maxims (Grice 1975).

4.2. Selective coding At the next stage of GT, selective coding, it is possible to begin constructing the matrix for the emergent theory. Thus far, LOVE might be posited as a property guiding men's lives and linguistic behavior. The dimensions of this continuum range from the type of idealized

142

Lawrence N. Berlin

love that Rick experienced with lisa in Paris, one that can change human nature to cynicism when shattered, and the type of physical love that Louis seems to hold in esteem. Another possible property that can be posited relates to the nature of CHOICE. The opposing dimensions range from destiny (i.e., minus choice) to free will. The two dimensions initially seem to be personified in the characters of Louis and Rick: the former who states that he "will take what comes" (10) and the latter who claims to 'stick his neck out for nobody'

(Π). STRASSER: You repeat "Third Reich" as though you expected there to be others. RENAULT: Well, personally, major, I will take what comes. [...](Wallis, etal. 1942:31) (11) [referring to Ugarte] RENAULT: If you are thinking of warning him, don't put yourself out. He cannot possibly escape. RICK: I stick my neck out for nobody. RENAULT: A wise foreign policy. (Wallis, et al. 1942: 22) Rick also seems to live by free will, as alluded to in (12) where Louis recounts Rick's previous choices to aid different sides in earlier conflicts. (12) RENAULT: Oh, laugh if you will, but I happen to be familiar with your record. Let me point out just two items. In 1935 you ran guns in Ethiopia. In 1936, you fought in Spain on the Loyalist side. RICK: And got well paid for it on both occasions. RENAULT: The winning side would have paid you much better. (Wallis, et al. 1942: 26) Despite these references, Rick still demonstrates more complexity and underscores the tension between destiny and free will when, as in the earlier excerpt (8), he infers that fate may have had a hand in lisa's reappearance into his life.

4.3. The cooperative principle and politeness It is at this juncture that it is possible to refer to and/or include other research toward the fuller development of theory. In the previous section, an allusion was made to Grice's (1975) cooperative principle and Brown and Levinson's (1987) theory of politeness. While the interaction between Rick and Louis in excerpt (9) seems more like ritual insults (Labov 1972), the work of Grice and that of Brown and Levinson can add to the overall analysis (cf. Fetzer this volume). In the following excerpt (13), Rick and Louis make a wager, yet

Grounded Theory and its Benefits

143

as the dialogue proceeds, it becomes apparent that Louis is less than forthcoming with necessary information.

(13) (13.1) (13.2) (13.3) (13.4) (13.5) (13.6) (13.7) (13.8) (13.9) (13.10) (13.11) (13.12) (13.13) (13.14) (13.15) (13.16) (13.17) (13.18) (13.19) (13.20) (13.21) (13.22) (13.23) (13.24) (13.25) (13.26) (13.27)

RENAULT: ...There is a man who's arrived in Casablanca on his way to America. He will offer a fortune to anyone who will furnish him with an exit visa. RICK: Yeah? What's his name? RENAULT: Victor Laszlo. RICK: Victor Laszlo? Renault watches Rick 's reaction. RENAULT: Rick, that is the first time I have ever seen you so impressed. RICK: Well, he's succeeded in impressing half the world. RENAULT: It is my duty to see that he doesn't impress the other half. Rick, Laszlo must never reach America. He stays in Casablanca. RICK: It'll be interesting to see how he manages. RENAULT: Manages what? RICK: His escape. RENAULT: Oh, but I just told you. RICK: — Stop it. He escaped from a concentration camp and the Nazis have been chasing him all over Europe. RENAULT: This is the end of the chase. RICK: Twenty thousand francs says it isn't. They sit to discuss the matter in earnest. RENAULT: Is that a serious offer? RICK: I just paid out twenty. I'd like to get it back. RENAULT: Make it ten. I am only a poor corrupt official. RICK: Okay. RENAULT: Done. No matter how clever he is, he still needs an exit visa, or I should say, two. RICK: Why two? RENAULT: He is traveling with a lady. RICK: He'll take one. RENAULT: I think not. I have seen the lady. And if he did not leave her in Marseilles, or in Oran, he certainly won't leave her in Casablanca. RICK: Maybe he's not quite as romantic as you are. RENAULT: It doesn't matter. There is no exit visa for him. (Wallis, et al. 1942: 24-27)

In line (13.15), Rick offers Louis the bet which is questioned in line (13.17) and renegotiated in line (13.19) via a counteroffer. In lines (13.20) and (13.21), the bet is finalized, but, in the same turn - line (13.21) - Louis manages to flout several maxims simultaneously. By providing extra information after the wager is established, he blatantly flouts the maxim of Quantity. Next, by adding the information about needing two tickets, he clearly flouts Quantity again, and possibly flouts the maxim of Relation by taking the conversation into a different direction, one which Rick dismisses by his response in line (13.24). Throughout the excerpt, however, Louis is also flouting the maxim of Manner in that he provides pieces of information over a prolonged stretch of dialogue.

144

Lawrence N. Berlin

Brown and Levinson's theory of politeness helps us to understand these utterances not so much in terms of conversational implicature, but rather as engaging in an intrinsic facethreatening act. Specifically, they refer to "those acts that predicate some positive future act of S [the speaker] toward H [the hearer], and in so doing put some pressure on H to accept or reject them, and possibly to incur a debt" (1987: 66). This definition clearly relates to the wager made between the two men, but the way in which they carry it out provides another example of the type of vying for position - mostly playfully - that characterizes the verbal enactment of their relationship. Moreover, and in strengthening the developing theory, (13) offers yet another example to index Rick's cynicism as the result of a failed love affair. We also begin to see Louis as multi-dimensional in line (13.25) as his utterance begins to broaden our understanding of his character by straddling the property LOVE. Though he offers the physical in his mention of the lady's appearance, he also suggests the ideal by presenting the notion that there is something more to love that would make a man give up his own freedom.

4.4. Speech acts Another fruitful area of research into language phenomena that can be incorporated into GT and its fuller interpretation of dialogue is speech act theory (Austin 1962; Searle 1969). In the following excerpt (14), the performative of WARNING is explored. In the following sequence, Rick has just arranged for Louis to release Laszlo form jail on the pretext of framing him for a more solid justification for his incarceration, and, presumably, to leave the country with lisa, Laszlo's wife. Louis, hidden off camera, surprises Laszlo and lisa when he appears and announces that he is arresting Laszlo (14). (14) (14.1) (14.2) (14.3)

(14.4) (14.5) (14.6) (14.7) (14.8) (14.9) (14.10) (14.11) (14.12) (14.13)

RENAULT: Oh, you are surprised about my friend Ricky? Obviously the situation delights Renault. He smiles as he turns toward Rick. RENAULT: The explanation is quite simple. Love, it seems, has triumphed over virtue. Thank In Rick 's hand is a gun, which he levels at Renault. RICK: ~ Not so fast, Louis. Nobody's going to be arrested. Not for a while yet. RENAULT: Have you taken leave of your senses? RICK: I have. Sit down over there. RENAULT: Put that gun down. Renault then walks toward Rick. Rick puts out his arm to stop him. RICK: Louis, I wouldn't like to shoot you, but I will, if you take one more step. [WARNING; COMMISSIVE] Renault halts for a moment and studies Rick. RENAULT: Under the circumstances, I will sit down. He walks to a table and sits.

Grounded Theory and its Benefits

\ 45

(14.14) RICK: {sharply} Keep your hands on the table. [WARNING: DIRECTIVE] (14.15) He takes out a cigarette case. (14.16) RENAULT: I suppose you know what you're doing, but I wonder if you realize what this means? (14.17) RICK: I do. We've got plenty of time to discuss that later. (14.18) RENAULT: Call off your watch-dogs you said. (14.19) RICK: Just the same, you call the airport and let me hear you tell them. And remember, this gun's pointed right at your heart. [WARNING: DECLARATIVE] (14.20) As Renault picks up the phone and dials, Rick takes back the letters. (14.21) RENAULT: That is my least vulnerable spot. (Wallis, et al. 1942: 116-118)

Searle (1975: 60-61) explains that: In indirect speech acts the speaker communicates to the hearer more than he actually says by way of relying on their mutually shared background information, both linguistic and nonlinguistic, together with the general powers of rationality and inference on the part of the hearer. To be more specific, the apparatus necessary to explain the indirect path of indirect speech acts includes a theory of speech acts, certain general principles of cooperative conversation [...], and mutually shared factual background information of the speaker and the hearer, together with an ability on the part of the hearer to make inferences.

By reinterpreting Austin's broad notion of performatives, it becomes clear that Rick issues warnings in three different forms: a commissive in line (14.10), a directive in line (14.14), and a declarative in line (14.19) (Searle: 1979). Moreover, the quote from Searle (1975) indicates the benefit of context to the interpretation - Halliday's (1984: 7) notion of "narrowing the gap between language as code and language as behaviour, and using each to explain the other" (cf. Chouliaraki and Fairclough 1999; Fairclough 1995; Kumaravadivelu 1999; Weigand 2002). Indeed, a thick description is necessary to better understand the situatedness of language within its broader contexts (Geertz 1973; cf. Hymes 1996; Urban 1991). Chouliaraki and Fairclough (1999: 140-141) refer to the terms of realization and instantiation where the former is the actual potential of words toward representation of meaning in the narrower linguistic sense (i.e., semantic, lexicogrammatic, and pragmatic), and the latter relates to "language as an open dynamic system in relationship to its (social) environment." From this broader perspective, the realization of repetitive keywords (as discussed earlier in section 4.1) and repetitive speech acts (as shown above) does not only suggest confirmability in emergent themes, but also credibility as they function indexically to define, or delimit, the semantic space of the themes, and to instantiate the themes within the developing theory.

146

Lawrence N. Berlin

4.5. Matrix and theory building

Figure 1. Developing Matrix for Grounded Theory

It is at this final stage of GT where the procedural and the conceptual come together, where the hypotheses are refined, where the themes merge into the grounded theory. The preliminary matrix took on the form in figure 1 where the two properties intersect and can be said to set up quadrants for linguistic behavior. Though the several examples support this description, it is unsatisfactory in its analysis, for Casablanca has not just lasted for more than 50 years as a classic film, but has survived in the American psyche. To develop an adequate explanation, something else needs to be added to the matrix. Again, letting the text speak for itself, it is Louis who provides the answer. At the end of the film, he asserts (15): (15) RENAULT: ... Love, it seems, has triumphed over virtue. (Wallis, et al. 1942: 117)

Yet his assertion is both erroneous in terms of the film's outcome and untenable in the emergent theory, for virtue is not in competition with love, but part of a separate theme. Still, the revised matrix is incomplete (see figure 2); VIRTUE is not itself a property that exists as such in a plus-minus relationship, but represents one dimension of a property, balanced by another dimension forwarded in the text by Louis from the wager he makes with Rick (16). (16) RENAULT: [...] I am only a poor corrupt official. (Wallis, et al. 1942: 25)

Consequently, a more accurate property label emerges from the tension between these two dimensions (see figure 3). MORALITY, as a property, is not only descriptive, but also explanatory as it surpasses a mere binary distinction between good and evil, right and wrong - too facile an explanation of the human condition and impetus for engagement in dialogue.

Grounded Theory and its Benefits

147

Figure 2. Developing Matrix for Grounded Theory (revised)

Figure 3. Matrix for Grounded Theory

This new property provides an aspect heretofore missing in the analysis, one that underlies the relationship between Rick and Louis. The harmony that exists between the two men is not a realization that they exist as polar opposites; it is recognition of the potential for movement in either direction on any of the properties. As the analysis has revealed that Rick is not simply a cynical, one-dimensional character, so, too, Louis in his final decision to cover Rick's shooting of Major Strasser by instructing the officers to "round up the usual suspects" as Laszlo and lisa escape on the plane to Lisbon leads to his own redemption and emergence as a multi-dimensional character. Ultimately, though, the goal of GT is to derive theory. The matrix offers a description of the core themes found in the script of "Casablanca" and begins to address an explanation of its broad and enduring appeal. The properties present dimensions of humanity that exist

148

Lawrence N. Berlin

for all of us and provide possibilities through their combination to find individual balance. However, multiple attempts to reproduce the film since the original have proved unsuccessful, despite the seeming universal appeal. It is, therefore, necessary to consider why this has been the case. A linguistic analysis alone does not fill the gap in the explanation. It is at this juncture that context completes the theory by advancing the explanation that the synchronic situatedness of the film, a locale that depicted desperation at a time when the world was truly at war, combines with the text to make the film meaningful beyond its mere words. Applied to dialogue analysis, then, GT can integrate previous research and multiple approaches by providing a level of abstraction for comparison and indexing interrelationships while demonstrating the relevance of contextual features to the overall analysis.

5. Conclusion

I have attempted to introduce Grounded Theory as a productive way to analyze dialogue. Though originally introduced as a qualitative approach to inquiry from the field of sociology, I submit that GT solves several issues that currently remain in dialogue analysis. Namely, through its interactive, iterative procedure, it takes a holistic approach, enabling researchers to incorporate considerations of context; and through its conceptualization, it proceeds inductively from a descriptive stance, generating hypotheses and eventual theory from the data, producing a level of abstraction that could resolve the issue of seemingly incompatible designs, methods, and findings.

References Agar, Michael (2004): We gave met the other and we're all nonlinear: Ethnography and Wolfram's 'new kind of science'. - Keynote delivered February 27, 2004 at the Ethnography in Education Research Forum, University of Pennsylvania, Philadelphia. Austin, John L. (1962): How to do things with words. - Oxford: Clarendon Press. Bakhtin, Mikhail (1981): The dialogic imagination. Four essays. Michael Holquist (ed.), trans. Caryl Emerson, Michael Holquist. - Austin: University of Texas Press. Brown, Penelope/Levinson, Stephen C. (1987): Politeness. Some universals in language usage. Cambridge: Cambridge University Press.

Grounded Theory and its Benefits

149

Chouliaraki, Lilie/Fairclough, Norman (1999): Discourse in late modernity. Rethinking critical discourse analysis. - Edinburgh: Edinburgh University Press. Dewey, John (1934): Art as experience. - New York: Minton, Blach. Fairclough, Norman (1995): Critical discourse analysis. The critical study of language. - London, New York: Longman. Ferguson, Charles (1959): Diglossia. - Word 15, 325-340. Firth, Alan/Wagner, Johannes (1997): On Discourse, communication, and (some) fundamental concepts in SLA research. Modern Language Journal 81, 285-300. Fishman, Joshua (1967): Bilingualism with and without diglossia. Diglossia with and without bilingualism. The Journal of Social Issues 23, 29-38. Flanders, Ned A. (1970): Analysing teaching behaviour. - Reading, MA: Addison-Wesley. Gardner, Richard C./Lambert, Wallace E. (1972): Attitudes and motivation in second language learning. - Rowley, MA: Newbury House. Geertz, Clifford (1973): The interpretation of cultures. New York: Basic Books. Glaser, Barney G./Strauss, Anselm L. (1967): The discovery of grounded theory. Strategies for qualitative research. - New York: Aldine de Gruyter. Goffman, Erving (1981): Forms of talk. - Phildelphia: University of Pennsylvania Press. Goodwin, Charles/Duranti, Alessandro (1992): Rethinking context. An introduction. - In: Alessandro Duranti, Charles Goodwin (eds.): Rethinking context. Language as an interactive phenomenon, 1-42. Cambridge: Cambridge University Press. Gordon, David/Lakoff, George (1975): Conversational postulates. - In: Peter Cole, Jerry L. Morgan (eds.): Speech acts. Syntax and semantics, Vol. 3, 83-106. New York: Academic. Grice, H. Paul (1975): Logic and Conversation. - In Peter Cole, Jerry L. Morgan (eds.): Speech acts. Syntax and semantics, Vol. 3,41-58. New York: Academic. Halliday, Michael A. K. (1984): Language as code and language as behaviour. A systemicfunctional interpretation of the nature and ontogenesis of dialogue. - In: Robin P. Fawcett, Michael A. K. Halliday, Sydney M. Lamb, Adam Makkai (eds.): The semiotics of culture and language. Language as social semiotic, Vol. 1, 3-35. London/Wolfeboro, NH: Frances Pinter. Huberman, A. Michael/Miles, Matthew B. (1998): Data management and analysis methods. - In: Norman K. Denzin, Yvonna S. Lincoln (eds.): Collecting and interpreting qualitative materials, 179-210. Thousand Oaks, CA: Sage Publications. Hymes, Dell (1996): Ethnography, linguistics, narrative inequality. Toward an understanding of voice. Bristol, UK: Taylor and Francis. Kumaravadivelu, B. (1999): Critical classroom discourse analysis. - TESOL Quarterly 33, 453-484. Labov, William (1972): Language in the inner city: Studies in the Black English Vernacular. Philadelphia: University of Pennsylvania Press. Lampert, Martin D./Ervin-Tripp, Susan M. (1993): Structured coding for the study of language and social interaction. - In: Jane A. Edwards, Martin D. Lampert (eds.): Talking data. Transcription and coding in discourse research, 169-206. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Lincoln, Yvonna S./Guba, Egon G. (1985): Naturalistic inquiry. - Newbury Park, CA: Sage. Merriam, Sharan B. (1988): Case study research in education. A qualitative approach. - San Francisco: Jossey-Bass Publishers. Merriam, Sharan B. (1998): Qualitative research and case study applications in education. San Francisco: Jossey-Bass Publishers. Schegloff, Emanuel A. (1992): In another context. - In: Alessandro Duranti, Charles Goodwin (eds.): Rethinking context. Language as an interactive phenomenon, 191-227. Cambridge: Cambridge University Press. Searle, John (1969): Speech acts. - Cambridge: Cambridge University Press. Searle, John (1975): Indirect speech acts. - In: Peter Cole, Jerry L. Morgan (eds.): Speech acts. Syntax and semantics, Vol. 3, 59-82. New York: Academic. Searle, John (1979): Expression and meaning. - Cambridge: Cambridge University Press.

150

Lawrence N. Berlin

Spradley, John P. (1980): Participant observation. - New York: Holt, Rinehart and Winston. Strauss, Anselm/Corbin, Juliet (1998a): Grounded theory methodology. An overview. - In: Norman K. Denzin, Yvonna S. Lincoln (eds.): Strategies of qualitative inquiry, 158-183. Thousand Oaks, CA: Sage Publications. Strauss, Anselm/Corbin, Juliet (1998b): Basics of qualitative research. Techniques and procedures for developing grounded theory (2nd ed.). Thousand Oaks, CA: Sage Publications. Urban, Greg (1991): A discourse-centered approach to culture. Native South American myths and rituals. - Austin, TX: University of Texas Press. Wallis, Hal B./Wamer, Jack L. (Producer), Curtiz, Michael (Director). (1942). Casablanca [Motion picture]. Burbank, CA: Warner Brothers. Weigand, Edda (2002): The language myth and linguistics humanised. - In: Roy Harris (ed.): The language myth in Western culture, 55-83. Richmond, Surrey, UK: Curzon Press.

Richard W. Hallett and Judith Kaplan-Weinger And Now for Your Moment of Zen: A Multi-Modal Discourse Analysis of Humor in "The Daily Show with Jon Stewart"

This analysis examines the discourse of humor of "The Daily Show with Jon Stewart". The program incorporates spoken, written, and visual modalities in the telling of the day's news. In contrast to the interdependent working of these modalities in serious news programming where they produce primarily an informing message, these modalities work interdependently and intertextually in "The Daily Show" to produce messages that inform as well as satirically comment on political and social issues. In the years this program has been broadcast, it has become a leading source of entertainment as well as news for a significant part of the population. In this way, "The Daily Show" has become the very kind of hegemonic structure it lampoons. This research adds to the theory of the discourse of humor and extends dialogic analysis to include the multi-modal analysis and hegemonic functioning of spoken, written and visual texts.

1. Humor and Hegemony

In the context of hegemony, defined by Gramsci (1971: 12) as "'spontaneous' consent given by the great masses of the population to the general direction imposed on social life by the dominant fundamental group", power usually rests with those in the social order who control the political and/or economic structure of a community. In the United States, such hegemonic power has traditionally rested with those who hold political and/or economic status (e.g., elected government officials, wealthy corporate executives, and entertainment personalities). This hegemony, in turn, authorizes those in power to extol and, perhaps surreptitiously, promote their ideologies (see van Dijk n.d.; cf. Berlin this volume). Traditionally, the media have assumed the responsibility (with the unstated, perhaps, acceptance of those in power) of informing "the rest of us" of the acts and decisions of these individuals and their effect on us. In this role, the media have also come to be part of the very hegemonic structure they, in theory, are meant to inform us about. Their power is earned as they, themselves, become experts and personalities. However, the media's in-

152

Λ. Hallett & J. Kaplan- Weinger

formation-bearing role is not its exclusive role. Previous research in media discourse has recognized "that the professional demands of ''impartiality' impose a series of constraints on the televisual newsworker, [such that] attention has been directed to the specific mechanisms by which this over-accessing of powerful voices as news sources works to underwrite the ideological rules of 'objectivity' and 'balance'" (Allan 1998: 112). The news media have not only created their own hegemonic structure, but this hegemony has also become part of the larger hegemonic structure of political power (cf. Berlin 2005). Alongside their role as impartial broadcasters of news, the media sometimes present satirical versions of the news. While satire of the news in the United States is nothing new,1 the fact that many in the populace are eschewing traditional news programming for fake news sources, appears to be a recent development. With "The Daily Show with Jon Stewart" (hereafter "Daily Show"), a recent entrant into the genre of fake news programming, the medium of television has taken another step in the hegemonic relationship of media and power in both reacting to the traditional structure while constructing a new hegemonic system of its own. Conventional media writers appear to be aware of this fact; for example, Johnson of "The Chicago Tribune" writes "The Daily Show" has "done it both by having the medium's sharpest nose for bovine waste product2 and by, like its spiritual kinsman 'The Onion',3 satirizing the media even more than it does politicians" (2004: para. 3). "Daily Show" airs Monday through Thursday on "Comedy Central", a US network on basic cable. Its presence on a network named "Comedy Central" is one's first clue that "Daily Show" is a non-traditional news program. The program recently won an Emmy award in the category of best variety show, beating out major network programs such as "The Late Show with David Letterman"4 and "The Tonight Show with Jay Leno".5 "Daily Show ratings are up 15% over 2002 to 900,000 viewers [...] and the show beats CNN, MSNBC, and Fox News among young adults at 11 p.m. (Daily also airs a weekly edition on CNN International, which reaches some 200 countries)" (Fretts 2003: 33). In January 2003, the Pew Research Center for the People and the Press (2003: 11) reported, "Today,

For example, the Weekend Update segment, a parody of news events, has been a feature of the program "Saturday Night Live" (SNL) for the duration of its 29 years on the air. Before SNL, there was a news parody segment on "Rowan and Martin's Laugh-In", a program that aired for six years on American television. The expression is a play on words for bullshit - used in colloquial American English to refer to a falsity. 'The Onion" is a satire of newspapers, described by multiple Internet sources as a "farcical newspaper featuring world, national and community news" (see http://www.theonion.com). "The Late Show with David Letterman" is a late night talk show that airs Monday through Friday nights on the CBS network in the United States. "The Tonight Show with Jay Leno" is a late night talk show that airs Monday through Friday nights on the NBC network in the United States.

And Now for Your Moment of Zen

153

21% of people under age 30 say they regularly learn about the campaign and the candi dates from comedy shows like 'Saturday Night Live' and the 'Daily Show', twice as many as said this four years ago (9%) [...] And this is particularly true for younger men, 27% of whom regularly learn about the campaign from comedy shows, compared with 14% of young women". Moreover, the Pew Research Center survey shows that these fake news programs are, through their satire, informing their audiences, noting that "27% of all respondents under age 30 say they learn things about the candidates and campaigns from late night and comedy programming that they did not know previously" (2003: 11). The fact that "Daily Show" is the primary source of news for a number of Americans is a source of hand-wringing for some political pundits (Carson 2004). However, according to Bill O'Reilly of the Annenberg Public Policy Center (2004: 1), "Daily Show viewers have higher campaign knowledge than national news viewers and newspaper readers [...] When looking at young people who watch The Daily Show, we find they score higher on campaign knowledge than young people who do not watch the show". In the 'Oaily Show", we have an opportunity to examine the influence of humor on the construction of hegemony. In the case of the 'Oaily Show", this is a humor that not only exposes and satirizes political and media hegemony, but in the process creates a hegemonic identity for itself. As Rosenthal (2004: 42), TV critic of the "Chicago Tribune" proclaims, "Comedy Central's 'The Daily Show with Jon Stewart' is the gold standard of topical humor these days, lampooning not only the news but the media that process it and package it like so much low-grade lunch meat."6 In analyzing the role of satiric humor in the construction and maintenance of hegemony, we adopt a critical discourse analytic perspective that focuses on the use of strategies of paradigmatic intertextuality (Bakhtin 1999; Kristeva 1986, as cited in Fairclough 1999) and visual semiotic design. Harris (2004) explains that: the essence of satire is aggression or criticism, and criticism (previous to the era of existentialistic nihilism) has always implied a systematic measure of good and bad. An object is criticized because it falls short of some standard which the critic desires that it should reach. Inseparable from any definition of satire is its corrective purpose, expressed through a critical mode which ridicules or otherwise attacks those conditions needing reformation in the opinion of the satirist, (para. 1)

Thrall, Hibbard, and Holman (1960, as cited in Harris 2004: para. 2) define satire as "a literary manner which blends a critical attitude with humor and wit to the end that human institutions or humanity may be improved. The true satirist is conscious of the frailty of institutions of man's devising and attempts through laughter not so much to tear them down as to inspire a remodeling". The expression is a play on words for baloney - used in colloquial American English to refer to nonsense.

154

Λ Hallett & J. Kaplan- Weinger

Intertextual analysis is a particularly appropriate model for the analysis of the construction of hegemony through satire as it "selectively draw[s] upon orders of discourse [...] available to text producers and interpreters in particular social circumstances. [Fairclough explains that texts are dependent] upon society and history in the form of the resources made available within the order of the discourse" (1999: 184). He also notes that discourses, narratives, registers, etc., may combine so that, as Kristeva explains, intertextuality is "the insertion of history (society) into a text and of this text into history" (1986, as cited in Fairclough 1999: 185). Allen (2000: 1) adds that Texts, whether they be literary or non-literary, are viewed by modern theorists as lacking in any kind of independent meaning. They are what theorists now call intertextual [...] Meaning becomes something which exists between a text and all other texts to which is refers and relates, moving out from the independent text into a network of textual relations.

Multimodal in its textual form (i.e., in its use of spoken, written, and visual texts), "Daily Show" is an intertext in that it has been constructed from a variety of sources. Bauman (2004: 2) extends the concept of intertextuality to encompass all of social life as being "discursively constituted, produced and reproduced in situated acts of speaking and other signifying practices, that are simultaneously anchored in their situational contexts of use and transcendent of them, linked by interdiscursive ties to other situations, other acts, other utterances." Intertextuality is crucial for our understanding of how texts are built upon and by other texts. Nonetheless, as Allen (2000: 59) notes, "Each theorist comes to intertextuality hoping it will provide an informing tool or model for interpretation, but each theorist soon realizes that, as a concept, intertextuality plunges one into a series of oppositions and questions." To date, no one has proposed a model for intertextual analysis; therefore, what we offer here is not a definitive methodology, but rather our own procedure. When watching "Daily Show", we realized that our understanding of the satire was based on our knowledge of events and people that were not mentioned explicitly in the show. Humor resided in phrases, words, images, and movements that connected to something outside the program itself. We asked ourselves what prior knowledge we had to have to interpret the meaning of what was said and shown. It was our familiarity with certain genres like television news and variety programming, as well as our recognition of the satiric frame of the show, that led to our appreciation of the text as an intertext. Some may conceive of a text as having its own meaning; however, for researchers such as Barthes (1977: 146), a text is "a multidimensional space in which a variety of writings, none of them original, blend and crash." Therefore, this analysis explores the variety of sources that are brought together to create the intertext that is the "Daily Show". We argue that one

And Now for Your Moment of Zen

155

can enjoy "Daily Show" as a self-contained text, but when one has the appropriate background knowledge, one can appreciate it as an intertext as well.

2. "The Daily Show with Jon Stewart'

With a few exceptions, the format of "Daily Show" is generally the following. First, the anchor, Jon Stewart (JS) welcomes the audience and makes a few comments. He then begins a segment called "Headlines", during which current, real news stories are reported and commented upon. Sometimes the stories involve a set-up in which one of the "Daily Show" regulars acts as a correspondent submitting a report. In these cases, the actor is always introduced to the audience as the show's senior correspondent (e.g., the Senior Middle East Correspondent, the Senior National Security Analyst Correspondent). Viewers soon get in on the joke as everyone is considered a "Senior X Correspondent" and the same player can be many different types of senior correspondent. Following the first commercial break, another news story might be discussed and/or a segment featuring one of the cast members may air. The latter generally involves one of the regulars serving as a reporter covering a story. Often these reports contain interviews with real people. Some of the regulars have recurring segments that sometimes air during this section (e.g., "Back in Black" with Lewis Black or "This Week in God" with Stephen Colbert). An interview with a celebrity or political figure usually follows the next commercial break. After that segment comes yet another commercial break. At the end of the program, JS thanks his guests and the audience and provides a preview (teaser) of a segment on an upcoming show. The broadcast ends with "A Moment of Zen" in which videotape of something shown earlier in the program is replayed with no voiceover. This footage is generally a video vignette that was lampooned in some way during the course of the program. The function of this moment is to reinforce the outrageousness of the footage in an ironic way by calling something Zen-like that is clearly not.

156

R- Hallett & J. Kaplan- Weinger

3. Framing and Positioning in the Construction of Hegemony

To understand the functioning of language in the creation of hegemonic structures, it is useful to consider the concept of interactive frames defined by Tannen (1993: 59) as "a definition of what is going on in interaction, without which no utterance (or movement or gesture) could be interpreted." According to Goffman (1974), a frame is the context in which to interpret the meaning of a discourse. Recognition of the frame triggers expectations of behavior and of meaning within that discourse.

3.1. News Program Framing The "Daily Show" appears to be operating within two frames simultaneously. In one frame, the program is presented as a typical news program. This frame is established from the very beginning as the date is superimposed on a waving American flag as a voice-over announces, "From Comedy Central's World News Headquarters in New York". The set for the fake news program closely resembles a regular news set; there is a spinning world globe next to the anchor's desk and there is a world map behind the anchorman who is dressed in a suit with a white shirt and dark tie and is seemingly scribbling notes during the introduction. Still operating in the news frame, JS, the news anchor, conducts interviews with one of the program's aforementioned senior correspondents, Rob Corddry (RC). JS and RC occupy the roles, respectively, of anchor and field reporter through a process Davies and Harre (1990) call positioning, defined as an act "of social interaction, not as a relatively fixed end product but as one who is constituted and reconstituted through the various discursive practices in which they participate. Accordingly, who one is always an open question with a shifting answer depending upon the positions made available within one's own and others' discursive practices" (para. 9). One's position shifts with one's decision - conscious or not - about membership and affiliation in a community. Also establishing the news frame of the program are the choices in phrasing and lexical items. In the episode7 analyzed in this paper, the main (lead) story was the capture of Saddam Hussein that had taken place over the weekend.8 JS begins the discussion of this capture by referring to it as, "Our top story". When JS introduces RC, "our senior Mideast 7 8

The episode being analyzed originally aired on Monday, December 15, 2003. The analysis presented here focuses on the first section of the program (i.e., from the beginning of the program to the first commercial break).

And Now for Your Moment of Zen

157

Correspondent" he continues, "He's been uh obviously in Iraq uh following the story from Saddam's hometown of Tikrit" and then thanks RC "for joining us". When interviewing RC, JS asks, in typical report style, "What's the mood like?" Still operating in the news frame, JS seeks to establish RC's credibility as a correspondent by stating, "You were obviously there when Saddam was first brought in. You've been with the 4th ID since this thing started." By using these phrases from traditional news programming, JS and RC capitalize on intertextuality to establish news show credibility in support of the developing hegemony of the 'Oaily Show". As in traditional news programs, "Daily Show" incorporates actual news footage into the context of the program. Accordingly, there are video vignettes of Major General Ray Odierno9 from a briefing on Saddam's capture, the military's medical exam of Saddam Hussein, and Australian Prime Minister John Howard's reaction to Saddam's capture. Typical of features found in a regular news format are the "Daily Show's" use of still photos of newsmakers and graphics to illustrate the story content. Likewise, the program uses captions such as headings (e.g., "LIVE TIKRIT" and "Voice of Translator") as well as footers (e.g., "SADDAM CAPTURED", as well as quotes and identification of speakers). Again, intertextuality plays an important role as the source of the humor lies in the direct borrowing of the conventions of one genre into the other. A viewer unfamiliar with the source text may not understand the humorous devices. Each of the above features, borrowed from network and local televised news broadcasts, serve intertextually to position "Daily Show" as a traditional news program. Each of these, then, serves to construct for the "Daily Show" - on the order of the construction that occurs with standard news programs - a hegemonic force as knower, purveyor, and even source of news (see Bell 1998). However, at the same time as "Daily Show" positions itself as a news program through its intertextual strategy, it also borrows from standard variety show programming to position itself within the comedy and interview program genre.

3.2. Variety Program Framing Evidence of intertextuality from variety programs is apparent from the beginning of each "Daily Show" episode. After the voiceover, the background music switches to the genre of rock and roll music. The camera then quickly scans part of the audience. As a result of this shot, the viewer becomes aware not only of the fact that there is a studio audience present (something generally not found in traditional news programs), but also that the audience is Commander of the 4th Infantry Division of U.S. forces in Iraq; the soldiers under his command captured Saddam Hussein.

158

Ä Hol Jett & J. Kaplan- Weinger

entertained, as evidenced by their applause. The first words uttered by JS are informal: "Hi, everybody", again something that a viewer might expect from a host of a variety program, not an anchorman of a serious news program. Still operating in this frame, RC gives a nod to former "Tonight Show" host, Johnny Carson,10 by using his signature trademark expression Hi-oo and his accompanying golf swing gesture to punctuate a joke. In the context of how "Daily Show" uses intertextuality to frame and position itself contrastively as both a news program and a variety program, there exists as well humor based on satire. Various lexical and visual signs signify that the discourses are meant to be understood as having, in the words of Harris (2004: para. 1-3), a corrective purpose, "expressed through a critical mode which ridicules or otherwise attacks those conditions needing reformation in the opinion of the satirist. [...] Far from being simply destructive, satire is implicitly constructive, and the satirists themselves [...] often depict themselves as such constructive critics." It is common for variety programs to satirize current events and for news programming to describe them. Just as other variety shows satirize current events, so does "Daily Show" as seen in examples (1) through (9): (l)In Baghdad we've seen parades, tears of joy, celebrating. Even those guerilla attacks have a jovial air almost as if they're not exploding things AT us, but exploding things WITH us. (2) Well his breath wasn't exactly winter fresh,11 but I'd hardly call it weapons-grade. (3) Well in Baghdad many took to the streets firing rifles into the air in joyous celebration, but in Saddam's hometown of Tikrit others took to the street denouncing the coalition and firing rifles into the air in bitter hatred. The rifle thing sort of the aloha of Iraq. Uh could be yea could be nay. Uh really like rifles. (4) Well they checked him for disease and lice and then, of course, the footage everyone 's seen of the US searching Saddam Hussein for weapons of mass destruction. (5)Here you see they 're now they 're combing the area for more uh bioweapons. (6) Oh now they're looking in this cave for possible chemical weapons - maybe a nuke? In fact, inspectors did find this suspicious capsule they think may contain seren gas. (7)JS: I think that is his uvula. RC: His uvula of mass destruction. (%)And then in footage that's already losing its considerable shock value, doctors checked the bedraggled Saddam for lice and pronounced him a member of the nnn-need a bath party. (9)The big news out of Iraq tonight is this. After months and months of searching we finally got the guy who had nothing to do with September 11th.

As the above discussion and examples show, what makes the "Daily Show" unique is its blending of news and variety roles to position itself as a humorous, yet well-informed and critical source of news thus establishing its influence on the viewing public.

11

Johnny Carson was an earlier and well-known host of "The Tonight Show" on the NBC network from 1962 to 1992. Winter fresh is an advertising slogan used for Wrigley's® chewing gum, a popular brand in the United States.

And Now for Your Moment of Zen

159

3.3. Other Media Framing While numerous individual visual and lexical devices from both news and variety programming construct for the 'Oaily Show" a place within the genre of satire, this genre is also constructed through other instances of intertextuality. These include the home decorating genre (10), literature (11), and sports (12). (10) Mw if he's planning on spending some time there he had to make it livable. Couple of mirrors to create the illusion of spa::ce,12 a light airy color like a butterscotch or a seafoam REALly brightens up a coffin-sized hole. Some recessed lighting and VOILA you 've turned your filthy hole into a cozy ditch. (11) Translator: God bless us everyone. JS: Their own version of Tiny Tim. (12) JS: But surely no discussion of this or any momentous world event would be complete without insight and analysis from Pittsburgh Steelers'^ all-time leader in passing yards. Jim Brown15: Was a big day internationally of course with the US troops capturing Saddam Hussein over there in Iraq so a big day indeed. Terry Bradshaw16: Osama, you 're next, my friend. JS: Terry Bradshaw issuing a fatwa.

4. Conclusion

Through this analysis of visual and lexical texts in Ά fake news program, we have explored ways in which intertextuality functions not to only identify and critique hegemony, but also to construct and promote it. The analysis illustrates that through the manipulation of texts and through the borrowing of texts across domains - intertextuality - "Daily Show" wrests power from the hegemonic political and media structures and allows it to settle with the populace. Through satire, the existing hegemony and its accompanying ideologies are made laughable and are weakened; at the same time, the populace is educated, informed, made knowledgeable and, thereby, strengthened. McGregor (n.d.: para. 4) explains, "One of the central attributes of dominant discourse is its power to interpret conditions, issues, 1

13

14 15 16

Following transcription conventions, a colon (:) indicates a lengthened segment and capitalization denotes emphasis. Reference to a character in Charles Dickens' "A Christmas Carol".

The Pittsburgh Steelers is a professional American football team. A former professional American football player who is now a sports commentator/announcer. Another former professional American football player who is now a sports commentator/announcer.

160

R. Hallett & J. Kaplan- Weinger

and events in favour of the elite." By studying such dominant discourse, our attention is drawn "to power imbalances, social inequities, non-democratic practices, and other injustices in hopes of spurring people to corrective actions" (Fairclough 1993, as cited in McGregor n.d.: para. 6) - a relatively new role for variety programming, yet a role that is increasingly more accepted and respected.

References Allan, Stuart (1998): News from NowHere. Televisual news discourse. - In: Allan Bell, Peter Garrett (eds.): Approaches to media discourse, 105-141. Oxford: Blackwell. Allen, Graham (2000): Intertextuality. - London: Routledge. Annenberg Public Policy Center (2004, September 21): National Annenberg Election Survey (NAES 04). (http://www.annenbergpublicpolicycenter.org/naes/2004_03_late-night-knowledge-2_921_pr.pdf) Bahktin, Mikhail M. (1999): The problem of speech genres. - In: Adam Jaworksi, Nikolas Coupland (eds.): The discourse reader, 121-132. London: Routledge. Barthes, Roland (1977): Image - music - text (Stephen Heath, trans.). - London: Fontana. Bauman, Richard (2004): A world of others' words. Cross-cultural perspectives on intertextuality. London: Blackwell. Bell, Allan/Garrett, Peter (eds.) (1998): Approaches to media discourse. - Oxford: Blackwell. Berlin, Lawrence N. (2005). Media manipulation. - In: Anne Betten, Monika Dannerer (eds.): Dialogue analysis IX. Dialogue in literature and the media, Part 2, 173-182. Tübingen: Max Niemeyer Verlag. Carson, Tom (2004, October 3): Last comic standing. Jon Stewart is the perfect comedian for these imperfect times. [Review of the book America: The book] The New York Times Book Review, 2021. Davies, Bronwyn/Harre, Rom (1990): Positioning. The discursive production of selves. (http://www.massey.ac.nz/~alock/position/position.htm) Fairclough, Norman (1999): Linguistic and intertextual analysis within discourse analysis. - In: Adam Jaworski, Nikolas Coup land (eds.): The discourse reader, 183-211. London: Routledge. Fretts, Bruce (2003, October 31): In Jon we trust. - Entertainment Weekly, 735. Goffman, Erving (1974): Frame analysis: An essay on the organization of experience. - New York: Harper and Row. Gramsci, Antonio (1971): Selections from the prison notebooks. -New York: International. Harris, Robert (2004, October 24): The purpose and method of satire. (http://www.virtualsalt.com/satire.htm) Johnson, Stephen (2004, February 22): Stewart's astute campaign special. Chicago Tribune. (http://wliia.net/main/modules.pnp?op=modload&name=News&file=index&catid=&topic=4&alls tories=l) [posted March 24, 2004] McGregor, Sue L. T. (n.d.): Critical discourse analysis. A primer. (http://www.kon.Org/archives/forum/l 5- l/mcgregor_print.doc) Pew Research Center for the People and the Press. (2004, January 11): Cable and Internet loom large in fragmented political news universe perceptions of partisan basis seen as growing, especially by Democrats. (http://people-press.org/reports/display.php37ReportnTN200)

And Now for Your Moment of Zen

161

Rosenthal, Phil (2004, January 26): We like our news with a grin. Chicago Sun Times. (http://wliia.net/main/index.php?module=subjects&ftinc=printpage&pageid=372&scope=all) [posted January 26, 2004] Tannen, Deborah/Wallat, Cynthia (1993): Interactive frames and knowledge Schemas in interaction. Examples from a medical examination/interview. - In: D. Tannen (ed.): Framing in discourse, 5776. Oxford: Oxford University Press, van Dijk, Teun (n.d.): Ideology and discourse. A multidisciplinary introduction. Internet course for the Oberta de Catalunya (UOC). (http://www.discourse-in-society.org/ideo-dis2.htm)

Herta Maurer-Lausegger

Audiovisual Dialectology: Methodology and Theoretical Considerations

This chapter first describes the dialectological-ethnographic film project "Documentary Records of Past Folk Culture" based at the University of Klagenfurt. This pilot project documents bilingual Carinthian dialect speakers of various ages, talking about selected topics in different situations on video in cooperation with a professional camera team. This form of audiovisual dialectology represents an innovative approach to research which is not only attracting great attention among dialectologists and ethnologists, but also researchers in other areas of cultural studies. Over the years, practical experience in the field and in the film studio has led to the development of a wide variety of different approaches. Some working methods are presented which were used during the research project (19942004), concentrating particularly on dialogue situations. Some theoretical considerations on audiovisual dialectology are followed by a detailed description of the methodology.

1. Introduction

Relentless changes in social structures and technological progress over the last few decades have fundamentally changed interpersonal communication, linguistic behavior, and linguistic culture as well as people's general attitudes towards language. Social restructuring is just one factor associated with constantly changing patterns of communicative behavior, against which dialects are slowly but surely losing out. Public broadcasting and other audiovisual and electronic media play an important role in this process because they have a decisive influence on the population's general awareness of language and culture. Nowadays, all social strata in the western world, even in the most geographically isolated communities, have easy access to communication and information media. Similarly, over the last few decades, modern communications technology has also made its way into a wide range of research fields, including dialectology. In the age of multimedia, traditional research methods from individual disciplines are starting to synthesize into trans- and interdisciplinary approaches, leading to cross-media thinking and facilitating the

164

Herta Maurer-Lausegger

"mutual illumination of various theoretical and methodological approaches" (Rajewsky 2002: 11). As Rajewsky points out, "film illustrates the relativity of time, space and visual perspective, the fragmentary nature of perception and the simultaneity of experiences; the impulse to move is also most convincingly accommodated by this medium" (2002: 29). Thanks to visualization, moving pictures, and the possibility of multidimensional representations of information, conventional ways of thinking and observing are given a fundamentally different perspective.

1.1. Intermedial Relations and Intermedia Studies The concepts of intermedial relations and intermedia studies have become more common since the 1980s. Intermediality first emerged in connection with film research "at the beginning of the 20th century as film and cultural theoreticians and authors began to take a critical look at what was then the new medium of film" (Rajewsky 2002: 8 fh.). Since 1994-95: this concept has been employed in almost all publications on the relationship between literature and the technical or electronic media as well as on the relationships between these media. In the traditional field of interart studies, in contrast, intermediality has only been used comparatively recently and only occasionally; in the English-speaking world, the term intermedia (studies) is far more common while intermediality is rather rare. (Rajewsky 2002: 9)

Researchers in this field will have to define the concept of intermedial relations and its relationship to intertextuality more precisely in the future (cf. Hallett and Kaplan-Weinger this volume): It would appear that creating a link between intermediality (in the narrower sense of intermedial relations) and the concept of intertextuality, as researchers have been striving to do since the early 1990s, really would be wise and helpful - but only if it is possible to clarify the fundamental concept of intertextuality, from which the concept of intermediality is derived. (Rajewsky 2002: 57 fh.)

1.2. Multimedia Technology in Dialect Research At about the same time, multimedia communications technology began to appear in dialect research. Indeed, audiovisual dialectology can be defined as the use of multimedia technology in dialect research, involving both auditory and visual processes. Conventional audiovisual media have been strongly influenced by developments in modern information

Audiovisual Dialectology

165

and communications technology which allow new forms of presentation to be used. These innovations can also be used to great effect in dialect research.

2. The Research Project "Documentary Records of Past Folk Culture in Dialect"

In 1994, a dialectological and ethnographic film project was launched by the Institute of Slavic Studies at the University of Klagenfurt in Austria. Entitled "Documentary Records of Past Folk Culture", it involves recording the language of bilingual dialect speakers from southern Carinthia and producing documentary video films. It is also part of a project set up by Professor Gerhard Neweklowsky in 1984 entitled "Research on Slovene Dialects in Carinthia" for which I took over responsibility in 1994. When I became project leader, I decided to make use of multimedia technology in dialectological field research. Our audiovisual dialectology mainly concentrates on the language spoken by the older generation. Bilingual speakers are asked to talk about specific agricultural implements and familiar routines in their local Slovene dialect and sometimes also in German. The video documentaries thus provide a comprehensive picture of the dialectal and ethnographic characteristics of southern Carinthia for generations to come. From 1994 to 1998, the audiovisual documentaries were produced in cooperation with the Institute of Media and Communication Studies at the University of Klagenfurt. Since 1998, the video documentaries have been made in close cooperation with experts at Artis, a film company in Klagenfurt. Between 1994 and 2004, seven video projects were completed in bilingual Carinthia. Since the project was launched, different methods of audiovisual dialectology have been developed and refined. Since November 1998, the research project has concentrated on documenting both agricultural or rural terminology and objects in audiovisual form, the main aim being to record spontaneous spoken dialect in simulated yet natural situations. Observations and experiences from everyday rural life, which is still characterized by the use of traditional farming skills and implements, are recorded on video in the form of dialogues, group discussions, and monologues (cf. Maurer-Lausegger 2004b). At present, the research project is concentrating on the audiovisual documentation of monologues and dialogues.

166

Herta Maurer-Lausegger

2.1. The Video Films and Accompanying Pamphlets The videos which have been produced so far concentrate on the following topics: water mills and sawmills, sheep breeding, agricultural implements for arable farming, sledges, baking bread in the black kitchen, and reminiscences of rural life in the past, among others. All in all, fifteen documentaries have been made on these topics, seven in Slovene, six in German, and two with English subtitles. They were published with an accompanying pamphlet by Hermagoras Verlag in Klagenfurt as part of a special series entitled "Documentaries in Dialect". The accompanying pamphlets, which are illustrated with old photographs, have a phonological transcription of the film's text with a German, English, or standard Slovene translation. They also include detailed information on how the individual films were made. In September 2000, the "Cultural Studies and Business" initiative was called into being at http://www.kwfilm.com, a trilingual internet platform (in German, Slovene, English) (cf. Maurer-Lausegger 2004a).

2.2. Objectives of Audiovisual Dialect Documentation As audiovisual technology facilitates the use of a wide variety of methods when recording linguistic data in the field, the resulting materials can be used for many different purposes. The driving force behind the film and internet project is the collection of data relevant to linguistics and cultural studies, but it also touches on sociocultural aspects. The main objective of the project is to document the local Slovene vernaculars spoken in different parts of southern Carinthia by recording the language used by bilingual speakers from different age groups. The video material is highly interesting from different academic and linguistic points of view. In addition to spontaneous spoken language, audiovisual recordings can be used for research into syntactic phenomena in spontaneous spoken dialect in one-to-one and group conversations, or as a basis for analyzing dialogue and the way in which it is organized in particular situational contexts. In this case, the typical syntactic patterns of spoken language are of particular interest, especially all forms of emphasis like topicalization, asides, and extraposition, and so forth, as well as ellipsis, selfcorrection, and other linguistic characteristics (cf. Maurer-Lausegger 1996: 267 fn.; Scheutz and Haudum 1982: 308). The documentaries are also screened in public, for example, as part of university courses (dialectology and ethnology), seminars, excursions, school events, and cultural evenings (cf. Maurer-Lausegger 2004a).

Audiovisual Dialectology

167

3. Audiovisual Dialectology

The last decade in particular has seen radical changes come about in media studies. As modern technology continues to develop, new forms of cultural and academic communication are available for use. In our multimedia age, a great many new methods of presentation and observation have been introduced in linguistic research. By recording the context, audiovisual dialectology presents dialect and interaction in their situational context. Language and culture and the way in which they are processed are attracting more and more attention in linguistic, media, communication, and cultural research. Audiovisual and electronic media make it possible to record acoustic information (text, sound) and pictures (the context of the situation, simulation, etc.) synchronically and diachronically. The dialectological sound track plus images and the ensuing documentary film demand a very high degree of processing from the recipients. Audiovisual data represent a complicated interweaving of interrelated semantic and semiotic elements which complement each other, endowing the meaning of the text in a classic sense with an additional, deeper dimension (cf. Hallett and Kaplan-Weinger this volume). The multimedia approach to audiovisual dialectology makes it possible to observe the dialect from a new and more profound perspective which was not previously accessible in dialect research (cf. Maurer-Lausegger 2004c).

3.1. Audiovisual Technology in Dialect Research Until the early 1990s, no attention was paid to audiovisual documentation in dialect research. The methodological foundations for this innovative discipline were developed between 1994 and 2004 in the multimedia dialectological-ethnographical research project entitled "Documentary Records of Past Folk Culture in Dialect". This process is founded on many years of experience with dialectological field research in the bilingual area of southern Carinthia, where German and Slovene are spoken, as well as filming experience with three different camera teams and theoretical knowledge from related disciplines. The many and varied methods of audiovisual dialectology were developed step by step in our film project and at present attempts are underway to substantiate them on a theoretical basis (cf. Maurer-Lausegger 2004c).

168

Herta Maurer-Lausegger

3.2. The Topics under Investigation in a Film Project The topics are chosen according to the methods of classic dialectology. As Putschke (1982: 241) defines it, they are "restricted to the language spoken by the local 'uneducated1 population."1 At the same time, they are also typical of classic ethnology. Traditional crafts and farming skills were chosen as topics as they are part of the dialect speakers' immediate field of experience. The documentaries record spontaneous language in simulated yet natural situations.2

3.3. Theoretical Reflections on Audiovisual Dialectology The media have attracted the attention of many different disciplines over the years. Preoccupied with predominantly political questions in the 1970s (in Germany), today's media theories cover a very wide range of perspectives in almost all faculties of the arts and social sciences. The specialist literature describes a great number of subject-specific approaches dealing with a wide range of questions and methods; "it is impossible to speak of One single' media theory" (Kloock and Spahr 1997: 8). In specialist literature, interdisciplinary culture theories exist side by side, usually generalized in relation to symbols, discourse, text, or narrative (cf. Ort 2003: 24 m.). 3.3.1. The Current Status of Research Not much theoretical work has been published in the literature of cultural studies, ethnography, and related subjects.3 The same is true for audiovisual dialectology, where attempts to set it up as its own subdiscipline have only been made within the last few years. The use of film in ethnographic field research does not actually differ radically from the more usual methods employed in field research. Where film is specifically mentioned in the literature, it is usually to point out that documentary films have largely been neglected by most recent theories on film. As dialectological literature does not provide any theoretical foundations for the topic of "film and field research", I have had to rely on theoretical knowledge from related disciAs the project leader is currently concentrating her research on drawing up a dialect monography of the vernacular of Windisch Bleiberg/Slovenji Plajberk, the documentaries will primarily focus on such topics and situations. All quotations from German and Slovene were translated by the author of this paper. Simulated yet natural situations "have the same basic structures as natural communication. Those intrusions into natural communication events which are unavoidable when collecting linguistic data are reduced to a minimum and carefully controlled" (Wodak 1982: 541). Cf. Ballhaus (1995: 13); Beck (1997: 161); Schillemanns (1995).

Audiovisual Dialectology

169

plines, but especially on the insights gained from my own practical experiences in this field. Audiovisual dialectology can be understood as the use of multimedia technology in dialect research. Its realization in many different forms and audiovisual processing combined with the ensuing methodological and theoretical insights firmly place dialectology in an intermedial and interdisciplinary context of cultural studies. 3.3.2. A First Theoretical Definition A first theoretical definition of audiovisual dialectology was formulated in 2000, but it will be necessary to work on a more detailed classification relating to the many and varied research interests and different technical options open to researchers in the field. The use of video technology in dialectological field work is a method of documenting linguistic and cultural matters synchronically which allows the viewer to hear the spoken dialect as well as other sounds and noises and to observe the production of spontaneous speech from a realistic perspective which also includes nonverbal communication and the social context. The chronological sequences of pictures on the video can be reproduced so that this process of hearing and observing can be experienced and analysed at any time and in every location where the appropriate technical equipment is available. The more often and the more intensively this process takes place, the more profoundly this realistic context can be tapped into. (Maurer-Lausegger 2000: 191 fh.; Maurer-Lausegger 2004a: 27)

Audiovisual recordings, or the way speech is presented in film, are an extract from reality. They reflect what is being talked about. Audiovisual techniques preserve events from the past in a realistic context. When filming a situation, spoken text is accompanied by contextual information about nonverbal communication with its very heterogeneous para- and extralinguistic phenomena. These often provide invaluable help in interpreting a text.4 In other words, audiovisual data not only provide text, but also information on the location, time, and plot, always including extralinguistic factors. General atmospheric and background noises (Vielmuth 1998) can be given an important role when changing locations, for example. Noises can also include linguistically relevant information which can only be conveyed and deciphered audiovisually. Audiovisual material can be reproduced and allows more precise interpretation. When available, slow motion close-up shots of articulation can be very helpful when working on a phonetic transcription, for example, as the details are visible, provided that the speaker is in the picture. It is also possible to look at individual frames in succession. In other words, the observation process can be repeated as often as necessary. In an even newer development, the Internet makes it possible for everybody to access these sources as often and for as long as desirable. According to Hufschmidt (1982: 567), paralinguistic phenomena are frequently dealt with separately, not in association with nonverbal phenomena.

170

Herta Maurer-Lausegger

Video recordings allow the viewer to take an active part in what is going on in a realistic context. As this reality can be reproduced chronologically on video, both the dialect and the sequence of events can be experienced anew time and again in all their variety. The process of perceiving and observing things from the past allows us to study dialect as a cultural object in more depth and breadth (cf. Maurer-Lausegger 2000: 203). 3.3.3. Filming in the First and Second Person Theoretically, there are numerous applications for audiovisual methodology in dialectology, both in connection with field research and the finished product, the film. The methodology really only depends on the limiting and facilitating conditions of the actual context, as is the case in other documentaries on cultural matters as well (Beck 1997).5 Thus, it will be essential to clarify the many different components which are available to researchers while filming in the field and processing the filmed material. Important insights can also be gained from a detailed study of related disciplines. At this stage, I would just like to single out one example to stand for many others: In audiovisual documentation of terminology and objects from the perspective of contextual research, the subjective camera plays the part of fas first person. In other words, the camera speaks from the perspective of the dialect speaker. Here I would like to refer to a few ideas mentioned by Vielmuth (1998: 124 in.): The first one is that cameras do not only reproduce things objectively: they can also portray a scene subjectively, from the viewpoint of the observer. When the cameraman moves while filming, the camera is released from its fixed position on a tripod and becomes a subjective camera. The result is more dynamic, with a continually changing perspective. The effect of such a personal angle of vision on the viewers is to give them the feeling that they are actually part of the scene themselves.

The subjective camera takes on the perspective of an observer, involving the viewers in what is going on in the film. When filming a flower, for example, the camera becomes subjective when it slips into the role of the flower and observes the surroundings from this perspective. In contrast, the camera in the second person has a linguistic and ethnographic task to do: it records the speech of the dialect speaker and the way in which dialect is produced in an actual context while attempting to preserve the ethnographic details on film. When the camera focuses on details in close up, it provides even more important information, such as the time of day (cf. Maurer-Lausegger 2000). "In German 'Sachkulturforschung' (researching objects in context), the concept of context is principally employed as a hermeneutic instrument with the help of which a valid interpretation of an object should be reconstructed for a specific situation against the background of a potentially wide variety of possible social and cultural interpretations. Here context is a category which limits meaning" (Beck 1997: 341).

A udiovisual Dialectology

171

3.4. The Methodology of Audiovisual Dialectological Field Research The following description of the working methods of audiovisual dialectology refers exclusively to recording bilingual speakers in bilingual southern Carinthia where the field researcher is confronted with an extremely delicate sociolinguistic situation. It cannot be assumed that all of our experiences are valid in different sociolinguistic contexts. In addition, as our films are not only intended for academic research, but also for the general public, very high demands are placed on technical requirements. When I first used video technology in dialectological field research, no special methodological or theoretical principles were mentioned in the literature to indicate how to do it. Consequently, I relied on many years of experience working with a cassette recorder for classic dialectology. I also studied references in the literature for social sciences, media studies, and cultural studies, and drew on my personal experience with film production in the early 1980s. Practice in the use of video cameras in dialectological field research also allowed me to gather valuable experience which proved to be very fruitful when working out methodological principles and when starting to think about theoretical approaches to audiovisual dialectology. Audiovisual documentation of dialect can be presented in numerous different ways. Depending on the cameraperson's skills and understanding of what is required, the technical equipment, the financial means, and the actual purpose of the study, decisions have to be made about which method should be used in a particular case. Every situation, or "the way in which linguistic data are obtained" (König 1982: 475), requires a specific method for filming, taking a whole series of parameters into account. I have used several techniques to obtain our audiovisual linguistic data and they require detailed theoretical backup.6 In order to make use of audiovisual techniques in dialectology, experience is required in many different fields. Alongside classic dialectological working methods, field researchers have to take account of the relevant sociolinguistic, situational, and communicative factors in the specific situation. Coordinated teamwork is one of the most indispensable prerequisites for a simulated yet natural situation to take place in front of the camera. It is also vital that the requirements of dialectology are optimally supported by the technology used. 3.4.1. Our Dialect Speakers As classic dialectological interests are the prime motivation for our recordings (i.e. the description of dialectal systems), the dialect speakers tend to be chosen from among the oldest generation of farmers in the village who speak the dialect in its most traditional form. Here, the ethnic identity of the speaker is hardly relevant, if at all. Every now and It is not possible to go into the individual aspects in this paper for technical reasons.

172

Herta Maurer-Lausegger

then, members of the younger generations are included in dialogues and group discussions. The recordings are of spontaneous conversations in the local Slovene dialect (and sometimes m German) with all their "slurs, short cuts, and ambiguities" which "often do not correspond to what would be used for writing the same thing down," as pointed out by König (1982: 478). 3.4.2. Locations The dialect speakers are filmed in familiar surroundings, such as their living room, kitchen, or yard. Since 1998, the contextual perspective (i.e., the immediate context in which the objects being researched are used) has also been taken into account7 as part of their actual cultural environment. Beck (1997: 156) points out that: the 'contextual perspective of the objects being researched forces the researcher to go out into the field where the implements are employed and the pieces of furniture are used' (Bringeus: 1968, 163) in order to observe how people actually use these things and why they are useful according to ethnographical methods of field research. In contrast to historical methods of analysing objects, actions can be directly observed and do not have to be reconstructed with the help of written sources or the actual object. 3.4.3. Recording Dialect in Front of the Camera The following considerations are essential in undertaking audiovisual dialectology: (l)When recording spontaneous spoken dialect, not all the methods of conventional filming can be used. As dialect is generally only spoken in its most natural form in spontaneous speech, in order to obtain representative material, it is not possible to make use of a screenplay in the classic sense of the word. The researcher does have to prepare the recording very thoroughly in advance, however, to take away the dialect speakers' fear of the camera and to ensure that everything connected with the filming runs smoothly. (2) The researcher works very closely with the cameraperson to draw up a specific concept for directing which takes maximum account of the cultural, dialectological and technical aspects. When filming, instructions to and from the camera team should be kept to a minimum, thus guaranteeing what should be an optimal portrayal of the scene.8 (3)Classic dialectological objectives (recording very traditional forms of the local dialect) involving several dialect speakers often require close-ups at several points. Several takes and different versions of the same scene are almost unavoidable. It is the researcher's task to maintain the naturalness of the conversation by reacting in the right way. This is also the case when it comes to dealing with unexpected disruptive factors during filming. (4) When filming groups of dialect speakers (whether they are sitting down or standing), experience has shown that there should not be more than two or three people in the group. In this case, it is advisable for one of the dialect speakers to take over responsibility for guiding the

8

Researchers interested in objects in context criticize the fact that "integration of the analysed objects into a genuine social context of use" (Beck 1997: 156) does not yet appear to have gained acceptance in ethnological research. Visual sociology proceeds in a similar methodological manner (see Schändlinger 1998: 78).

Audiovisual Dialectology

173

conversation. If at all possible, the researcher should take on the role of an external coordinator and only contribute to the conversation in exceptional circumstances. In certain situations, it is even necessary for the researcher not to be present at all during filming (or at least for part of the time). An example for this would be to prevent children from making eye contact with the researcher when they are meant to be talking to another adult. (5)When recording spontaneous spoken dialect, integrating the researcher as an active partner within the group of dialect speakers is only advisable when the dialectologist speaks the same dialect as the people being filmed. (6) When the special focus is on phonetics, the camera should principally focus on the person who is speaking at any one time. In such a case, other things which might be going on within the context of the conversation are not covered by the camera. The exception to this is if a second camera is used to record the overall scene. This has only happened in one of our projects so far when a camera team from the Austrian Broadcasting Company was filming at the same time.9 (7) It is important for the researcher to make notes on any changes in the situational context so that events which are not recorded on film can also be documented.I0 (8) Special attention has to be paid to the appropriateness of the images and the fact that the speaker must always be seen to be speaking. Voiceovers, for example, could be easily manipulated. At the same time, the sound quality has to be excellent for later dialectological analysis. For these reasons, only one dialect speaker is used to present the information even if other people are, by necessity, occasionally involved in what is being filmed. (9)During the technically demanding audiovisual documentation of terminology and objects from the perspective of contextual research, text is not touched at all in the editing process. The sequences also remain largely unchanged so that the methodological procedure is comprehensible in the finished film. Whether it is a question of recording terminology or dialect, documentation with a video camera closely follows the ethnographic principle of aiming to reconstruct various aspects in the life of dialect speakers as accurately as possible (cf. Engelbrecht 1995).

4. Audiovisual Presentation of Monologues and Dialogues

Three different constellations from our video documentary, "Cej so tiste stezice ... Spomini na nekdanje kmecko zivljenje" ('Following the tracks of yesteryear ... reminiscences of rural life in the past'), illustrate just some of the many ways in which audiovisual documentation of dialect can be realized. In all three cases, the speakers were sitting down and recorded with one camera in a fixed position while the sound was recorded by an assistant. Zooming and tracking with the camera focus the viewers' attention on the dynamics of the conversation and the expressive gestures and facial expressions with which the speaker tries to emphasize certain points.

10

The material recorded filmed with the second camera cannot be used for copyright reasons. This is particularly relevant when recording nonverbal communication (see also Hufschmidt 1982).

174

Heria Maurer-Lausegger

Situation 1: The researcher as interview partner. In this case the researcher takes part in the conversation and is filmed along with the dialect speaker. The researcher steers the dialogue in a particular direction, creating a simulated yet natural situation which should be maintained during filming as well. Once absorbed in the conversation, the speakers usually forget that they are being filmed. In contrast with free speech, conversations guided by the researcher include smooth transitions when changing the subject and provide impulses to talk about different topics, thus preventing the dialogue from being broken off or interrupted. Situation 2: Free conversation between two people. Here the camera records an unguided free conversation in a simulated yet natural situation. Both speakers have constantly changing roles in the interaction and speak for different amounts of time, depending on who takes on the role of guiding the conversation. The second person only occasionally manages to dominate the interaction. Both speakers use their local dialect without deviating from their usual linguistic system. This method of filming is the most natural when recording the most traditional form of the vernacular used in rural areas. Situation 3: Monologue in front of the camera. The third method of documenting freely spoken dialect is in the form of a monologue with only one person in front of the camera while the researcher is a passive listener beyond the camera's angle of vision. Standing right next to the camera, the researcher takes on the role of an interested listener, maintaining the naturalness of the language. The dialect speaker tends to look at the audience standing next to the camera. In such situations it is important that the cameraperson/assistant speaks the local dialect as otherwise the speaker's dialectal system can be influenced by their presence. Such phenomena are not to be observed on our sequences, which is quite an achievement.

5. Conclusion

As part of the project, "Documentary Records of Past Folk Culture" at the University of Klagenfurt, various methods of audiovisual dialectology have been developed with the help of a professional film team. Based on practical and academic findings, the first theoretical considerations were formulated by the project leader in 2000. Some aspects of this project have been presented in this paper. Using different methods of audiovisual presentation of dialect undoubtedly opens up a whole new series of perspectives to dialectological research. Conventional research methods and theoretical approaches in other subsections of dialectology can be extended and complemented by a broad range of new visions from communication, media and cultural studies when representing the visual elements of dialect in context.

A udiovisual Dialectology

175

References Ballhaus, Edmund (1995): Film und Feldforschung. Überlegungen zur Frage der Authentizität im kulturwissenschaftlichen Film. - In: Beate Engelbrecht, Edmund Ballhaus (eds.): Der ethnographische Film. Eine Einführung in Methoden und Praxis, 13-46. Berlin: Reimer. Beck, Stefan (1997): Umgang mit Technik. Kulturelle Praxen und kulturwissenschaftliche Forschungskonzepte. - Berlin: Akad. Verlag. Bringeus, Nils-Arvid (1968): Das Studium der Innovationen. - Zeitschrift för Volkskunde 64, 161185. Engelbrecht, Beate (1995): Film als Methode in der Ethnologie. - In: Beate Engelbrecht, Edmund Ballhaus (eds.): Der ethnographische Film. Eine Einführung in Methoden und Praxis, 143-186. Berlin: Reimer. Kloock, Daniela/Spahr, Angela (1997): Medientheorien. Eine Einführung (= UTB für Wissenschaft: Uni-Taschenbücher, 1986). - München: Fink. König, Werner (1982): Probleme der Repräsentativität in der Dialektologie. - In: Werner Besch, Ulrich Knoop, Wolfgang Putschke, Herbert Ernst Wiegand (eds.): Dialektologie. Ein Handbuch zur deutschen und allgemeinen Dialektforschung. 1. Halbband (Handbücher zur Sprach- und Kommunikationswissenschaft 1,1), 463-485. Berlin/New York: Mouton de Gruyter. Maurer-Lausegger, Herta (1996): Die Mundart von Diex/Djek§e in Kämten. Eine zweisprachige Videodokumentation. - In: Ingeborg Ohnheiser (ed.): Wechselbeziehungen zwischen slawischen Sprachen, Literaturen und Kulturen in Vergangenheit und Gegenwart, 257-268. Innsbruck: Verlag des Instituts für Sprachwissenschaft. Maurer-Lausegger, Herta (1997): Filmska dokumentacija nareöja. Metodologija spraäevanja v dvojeziöni situaciji na juznem Koroäkem. - In: Nasko Kriznar (ed.): Etnoloäki film med tradicijo in vizijo, 141-150. Ljubljana: Zalozba ZRC SAZU. Maurer-Lausegger, Herta (1998): Dialektforschung mit der Videokamera im südlichen Kärnten. Ein Bericht über Erfahrungen, Methoden und Probleme. - In: Robert Pittner, Karin Pittner (eds.): Beiträge zu Sprache und Sprachen 2. Vorträge der 5. Münchner Linguistik-Tage, München 1995, 5-19. München: Lincom Europa. Maurer-Lausegger, Herta (2000): Die audiovisuelle Dialektforschung - eine Kulturwissenschaft. In: Wolfgang Stadier, Eva Binder, Helmut Kalb (eds.): Junge Slawistik in Österreich. Beiträge zum 1. Arbeitstreffen, Innsbruck, 24. - 26.02.1999, 187-207. Innsbruck: Verlag des Instituts für Sprachwissenschaft. Maurer-Lausegger, Herta (2004a): Audiovisual dialectology. The film research project "Documentary records of past folk culture in dialect" and its role in the linguistic culture of Carinthia. Tidsskriflfor Sprogforskning 2/1, 25-40. Maurer-Lausegger, Herta (2004b): Audiovisuelle Dialekt- und Terminologieforschung. Fragen zur Übersetzung und Synchronisation dialektologischer Filme. - In: Irmeli Heiin (ed.): Dialektübersetzung und Dialekte in Multimedia, 21-44. Frankfurt am Main/Berlin/Bern/Bruxelles/New York/Oxford/Wien: Peter Lang Verlag. Maurer-Lausegger, Herta (2004c): Wiedergabe des Dialekts in audiovisuellen Medien (inklusive Internet). - In: Livia Adamcova (ed.): Beiträge zu Sprache & Sprache. Vortrage der 11. Jahrestagung der Gesus in Bratislava, 267-274. München: Lincom Europa. Ort, Claus-Michael (2003): Kulturbegriffe und Kulturtheorien. - In: Ansgar Nünning, Vera Nünning (eds.): Konzepte der Kulturwissenschaften. Theoretische Grundlagen - Ansätze - Perspektiven, 19-38. Stuttgart/Weimar: Verlag J. B. Metzler. Rajewsky, Irina O. (2002): Intermedialität. (= UTB für Wissenschaft: Uni-Taschenbücher, 2261). Tübingen/Basel: Francke.

176

Herta Maurer-Lausegger

Schändlinger, Robert (1998): Erfahrungsbilder. Visuelle Soziologie und dokumentarischer Film. Konstanz: UVK Medien. Scheutz, Hans Peter/Haudum, Robert (1982): Theorieansätze einer kommunikativen Dialektologie. In: Werner Besch, Ulrich Knoop, Wolfgang Putschke, Herbert Ernst Wiegand (eds.): Dialektologie. Ein Handbuch zur deutschen und allgemeinen Dialektforschung. 1. Halbband (Handbücher zur Sprach- und Kommunikationswissenschaft 1,1), 295-315. Berlin/New York: Mouton de Gruyter. Schillemanns, Sandra (1995): Die Vernachlässigung des Dokumentarfilms in der neueren Filmtheorie. - In: Manfred Hattendorf (ed.): Perspektiven des Dokumentarfilms, 11-28. München: Diskurs-Film-Verl. Schaudig u. Ledig. Vielmuth, Ulrich (1998): Ratgeber für Videofilmer: Tips und Tricks vom Profi. - Köln: DuMont. Wodak, Ruth (1982): Erhebung von Sprachdaten in natürlicher oder simuliert-natürlicher Sprechsituation. - In: Werner Besch, Ulrich Knoop, Wolfgang Putschke, Herbert Ernst Wiegand (eds.): Dialektologie. Ein Handbuch zur deutschen und allgemeinen Dialektforschung. 1. Halbband (Handbücher zur Sprach- und Kommunikationswissenschaft 1,1), 539-544. Berlin/New York: Mouton de Gruyter. Documentary records of past folk culture in dialect available at (http://www.kwfilm.com).

Herve Varenne (in collaboration with Mary Cotter)

The Production of Difference in Interaction: On Culturing Conversation through Play

1. Introduction

Many classical problems must make it difficult to think of conversation, dialogue, discourse, speech in context, etc., in terms of culture. And yet it must be done - carefully. This chapter explores why this must be done and how it might be done. To do this I bring together some of Garfinkel's latest language (2002), classical considerations from cultural anthropology (Geertz 1973; Sapir 1921), and some more recent ones (Boon 1999; de Certeau 1984/1980). To illustrate the argument, I present a summary analysis of one moment of joking during a hospital labor. Garfmkel, provocatively, now writes about "immortal social facts" in a determinedly Durkheimian mode. His prototype for such facts is the "service line" - the kind of lines one encounters when one enters a post office to purchase stamps. Service lines are immortal in the sense that they exist before any particular person join the line; they perdure after the person leaves; and they are controlled by all other people in the line as they instruct each other about what to do next. In his career, Garfmkel has extensively explored how social facts are maintained, and what their consequences can make on individual experiences. He has focused on order while paying close attention to extraordinary cases, but he has not quite faced the fact that one can line up for service in many different ways, as people who cross any number of boundaries soon experience. He has not given clear guidance on how to deal with the production of such differences in the actual history of human interaction. My main goal is to explore ways to account for the production of difference in terms that take seriously both ethnomethodological and anthropological traditions. To concretize these considerations, I build on work by Cotter (1996) who reported in great detail the case of a woman (and her husband, doctors, a nurse, and a researcher) laboring towards child delivery in a contemporary hospital. Hospitals, like post offices, reveal themselves in what people produce there together. These products, however temporary, are immortal in that they are already there when women (and any other with them) enter delivery rooms, when they organize their body in the terms prescribed for them, and when they do any number of

178

Herve Varenne

things that might make their particular delivery extraordinary. But hospital delivery rooms do not determine what can be done there. One can tell lies there, and put oneself at such physical danger as to redirect the exact development of a labor. And one can joke about all this - together with the co-participants in the event. Sacks (1975) once argued that "everyone has to lie". Everyone also has to play - and through such play, new orders may emerge. Immortal facts, I argue, are not simply immortal. They are also arbitrary in the classic senses we inherit from Saussure (1981/1915) and, more recently, Bourdieu (1977/1972). "Arbitrariness", in this tradition, refers to one major consideration with two aspects, one logical and the other political. Anthropologists and others have extensively demonstrated through comparative analysis that any common human activity can be effectively conducted in a range of different ways (a logical argument), and that any particular way becomes overwhelmingly consequential for the population that must live by it (a political argument). They have also extensively demonstrated that people, everywhere, resist those who tell them what to do. They play with what is given them and often they make new objects with it. The arbitrariness of patterns particular to times and populations can be introduced in terms of three relations of the observed pattern to aspects of human experience: biology, social organization, and social process. First, immortal social facts are arbitrary in relation to the organization of the delivery of the services (e.g., stamps, a healthy baby, "education for all"). Neither rationality, biology, nor even ideology can determine the exact shape of the evolving pattern. The biology of human childbirth may be a universal problem, but all solutions in human history have been particular, and they keep changing across and within national boundaries (Davis-Floyd 1992; Davis-Floyd and Sargent 1997; Fried 1980; Jordan 1978; Rapp 1999). Second, immortal social facts are also arbitrary in terms of the political patterns that keep them alive. Childbirth, historians tell us, was for a long time controlled by women. More recently, it has been controlled by particular kinds of men with special "medical" authority (Apple 1990; Apple and Golden 1997; Leavitt 1986). Third, immortal social facts are arbitrary in terms of the new services that will have to be delivered precisely because of the arbitrariness induced by the organization of biology. The medicalization of childbirth has multiplied the kinds of people who must be involved for any particular birth to be treated as normal. The "doctor" is now four different specialists: gynecologist, obstetrician, anesthesiologist, and pediatrician. Each has a specific range of authority on the laboring body, and this authority may exercise itself only at certain times.' This has been well delineated in Stratton's (2003) work on the medicalization of hearing loss in Sweden.

The Production of Difference in Interaction

\ 79

In brief, I am convinced we must accept that all immortal facts are cultural facts. They are the product of historical evolution, they constrain but do not determine, and yet they perdure - mostly. Mostly indeed, for the very arbitrariness of any fact leaves open the door for further evolution. Ethnomethodology must face culture. But to talk about "culture" comes with its own danger. For many years, noting difference and arbitrariness has led to a series of often unexamined hypotheses: routine performance in the terms of a particular cultural arbitrariness would be the product of the knowledge participants gained through early enculturation or socialization; individuals without such knowledge might have trouble getting the service; trouble among the people waiting for the service might be explained through a reference to differing socialization about how to get or deliver service; this might or might not be articulated into the possible "national", "class", "gender", or "race" differentiations that may also be involved. Only confusion is produced by moving from the observations of institutionalized difference to hypotheses about the learning (i.e., internalization) of rules or habits. Cultural orders are not the product of personal orders - even when it can be shown that participation does have an effect on personal orders (e.g., character, self, identity). Cultural orders are the evolved product of the active organization of people who, in the process, constitute this order and thus the conditions they will face from then on. The great merit of ethnomethodology has been its providing much stronger accounts of this constitution. What ethnomethodology has not quite done is provide accounts of conversational processes that allow for unpredictable and unrepeatable drift in the organization of constituted facts (cf. Macbeth this volume). Current accounts are strongest when dealing with the processes that maintain order even as they recognize the continual reality of disordering moves (cf. Hundsnurscher this volume). Such moves must be seen as opening the way to transformation, what I refer to as "culturings". Garfinkel is well aware that disordering founded in the very form of the order is always possible: people do "screw around" (2002: 257). He is interested in these problems as sources for instructions that establish what is consequential at particular times both for people to each other and also for the observer. He does not systematically explore the conditions that may reconstitute the original fact so that it will have different "apparent phenomenal properties" (2002: 256) for future participants. For example, one can imagine a setting where some people (say school administrators) might meet because "there was some screwing around in the parking lot while the students were getting into line to board buses". Such a meeting would have its own properties, of course, but it may also produce an alteration in the properties of future gettings into line to board buses at that school (e.g., by adding security guards). Single events, historical happenstance, can become factual and produce new culture.

180

Herve Varenne

This chapter uses one moment of screwing around to concretize these considerations. The emphasis in this account is on potentialities activated rather than on the reconstitution of the order. This is also accomplished so that, in this case, nothing much appeared to change. But, in the telling of a story about screwing around, some difference was introduced, factualized, and made locally consequential: all participants dealt with the story as story-and a funny story at that. But it may also have had other consequences though not clearly in the time frame of the observation. In the end, a new subculture was produced one which participatory structures clearly allowed for such joking. The remainder of the chapter has two main parts. First, I briefly trace how much cultural anthropology - and I do stress "cultural" - can gain by confronting conversational analysis and ethnomethodology as precisely the strongest theoretical framework to continue the broad development of culture as a fundamental concept in the social sciences. Second, I focus on one aspect of this development, the aspect that has to do with the very production of differences that may then be institutionalized into a "fact", what is usually talked about as "a" culture or subculture, way of life, etc. This exploration itself has three aspects that will be differentially stressed at various points: (1)1 focus on play that is necessary because immortal cultural facts are also sets of instructions that cannot determine what kind of response they will produce (Bateson 1955; Boon 1999; Geertz 1973; McDermott and Tylbor 1983); (2)1 focus on "enunciation", the word I use here to index the very act of speaking as it is happening, what Merleau-Ponty (1973) called "le langage parlanf (sometimes translated as 'speech'). Every collective performance that can be seen by all to constitute the immortal cultural fact must also be unique. This uniqueness is potentially noticeable locally and may only trigger various forms of active ignore-ance (Garfinkel 2002: 222-223), instruction, sanction, etc. Whether it "makes a difference" is another matter. (3)1 focus on what can make collective enunciation "make a difference", that is, on the factualization of expression through processes we might refer to as "inscription" or the constitution of new consequentialities - processes that reveal the power of the present over the future.

2. An Argument for Cultural Anthropology to Journey into Conversation (and for Ethnomethodology to Journey into Culturings)

It is an altogether fair generalization to say that ethnomethodologists have paid little attention to what anthropology has had to say, and that anthropologists have similarly ignored ethnomethodology. Whatever the reason for this, there is enough convergence in the core interests of at least some in each discipline that it might be worth tracing what the two traditions can contribute to each other. The key step in anthropology came when it was rec-

The Production of Difference in Interaction

181

ognized that the study of language is the best model for the study of culture. But, at the time when Garfinkel started working with sociolinguists and conversational analysts, LeviStrauss (1963/1958), who made the argument most cogently in anthropology, proposed that anthropologists consider the work of Saussure and Jakobson above all. Most anthropologists resonated to the general argument but they found the particular structuralist model too bloodless. They shifted to a literary model that appeared to allow best for the symbolic power of symbols and other expressive forms (see also Hess-Lüttich this volume). In their writing, culture was to be text. Much followed from this - particularly a disinterest in language in productive action. Still, to insist that culture is text does focus attention on descriptive adequacy and the practical politics of the text (both as it is being written, and as it is taken by any reader). When addressed together, these are precisely some of the matters that moved ethnomethodology. When addressed separately, as they mostly were in anthropology, the general interest in language became a concern with "discourse" as Foucault meant the word, and a concern with "practice" as Bourdieu "theorized" it. But neither Foucault nor Bourdieu, in their own intellectual practice, encouraged facing the practical details of constituting human worlds. For various reasons, the alternate route that would have anthropologists face these details remains one few take. It may be that most anthropologists take seriously one critique of ethnomethodology that Bourdieu repeatedly made: that it only accounts for such arguably minor matters as the opening of telephone conversations, and cannot deal with, say, class reproduction through schooling. This does not follow. Anthropologists such as McDermott (1979, 1993) analyzing classrooms, or Lave (1984, 1988) observing everyday cognition in structured fields, demonstrated that close attention to the apparent details of the constitution of moments through speech and movement can help in the more carefully restating of classical problems for those concerned with the organization of large populations and the movement of people through this organization (see Hundsnurscher this volume, for an opposing opinion). But such a focus leaves open the question of the historical particularization of this organization - the matter that most concerned cultural anthropologists. In the context of medicalized childbirth, for example, it makes sense to talk of an "American way" that is somehow different from a "Swedish" or "Dutch way" (DavisFloyd 1992; Jordan 1977/1993; Leavitt 1986; Sargent 1997). America, in the ongoing constitution of its medicine, schooling, business, politics, is an immortal social fact of overwhelming significance inside and outside the national boundaries of the United States. It is also historically unique and it must be possible to account for this uniqueness. What remains difficult is writing more careful accounts of the properties of any set of practices that make it recognizable as American. It is even more difficult to account for the reconstitution of these particular properties across time and social space. And it has been all but impossible to suggest how, exactly, it can change.

182

Herve Varenne

The danger is to make the maintenance of America dependent on psychological processes of inculcation as if America resided in the personal constitution of Americans. The solution is to return to the close observation of people dealing with each other and revealing what is most significant to them at the time. In my own work in small towns, school and family settings (Varenne 1977, 1984, 1992), I came to argue that America resides partially in the paradigm of ways of telling experience in various settings, each with its specified rhetorical forms, and its own authority to transform the career of those being talked about. I could observe how the retelling of experience in certain settings could erase many features of everyday experiences, and how it could highlight others as consequential. There are settings where telling what something really means can have major consequences. In this way, broad settings are constituted in which all, in the United States and increasingly outside of it, must live their lives. As I started exploring how this erasure is accomplished I found ethnomethodology more and more useful. With McDermott, I looked at various institutional moments when voices of intellectual authority produce massive reorganizations with longlasting effects on those involved. In our "Successful Failure" (Varenne and McDermott 1998), we accounted for the particular kinds of (dis)abilities found in the United States in terms of the local reconstitution of a particular cultural arbitrariness: the one that makes "passing/failing" an examination, for instance, key to movement within what I would call "polities of practice".2 "America", in this perspective, refers to the pattern of linkages between various settings that are maintained by the movement of people as agents o/the institutions (treated as settings) that delegate them into particular other settings and hold them accountable for what they do there. Concretely, it means that "doctors" are not only "doctors" because they are so constructed by "patients" in the real time of a medical examination; they are also "doctors" because they are so constructed for such examination time in the real, though perhaps indirect, time of "medical school".3 In these ways, classical concerns with "social structure" can be made alive in a manner that can be accounted for in full ethnomethodological terms. But this is not quite enough if the account does not also allow for the production of "America" as distinct from say, "Sweden" in the detail of the organization of its polities for example its organization of childbirth. To focus only on the method of the production and identification of order conspires to make historical particularity in the principles of any This phrase is intended to echo the phrase "community of practice" that was first proposed by Lave (Lave and Wenger 1991) to handle the complexity of the movement of people through the fields of identified knowledge and then transformed by Wenger (1998) in such a way as to hide to political work that continually reconstitute the differentiation of people working together. As Cotter (1996) argued, this can make it necessary actively to ignore the status of "doctor" at interactional moments when to do so would dis-order the local setting.

The Production of Difference in Interaction

\ 83

ordering a secondary topic. My concern is to pay attention to the production of new orders linked, but distinct from the old orders that used to prevail. All immortal orders are also temporary. They appear in history, and then they disappear. The reality of drift, evolution, and transformation is precisely what made anthropology necessary as a counterpoint to sociology. "Culture" is too powerful a concept to abandon. But, to preserve culture, one must precisely move away from what has become the most common usages of the word. Culture, when it is defined in passing by someone not necessarily concerned with the concept, is usually presented as somehow having to do with learning. Culture, for me, has to do with the institutionalization of difference in a full factual way. Culture belongs to the world of social facts as understood by Durkheim (1982/1895), to the world of langue as understood by Saussure (1981/1915), and to the world of interaction as understood by Garfinkel (2002).4 It has to do with what individuals do not control in their lives. It has to do with what people receive and with what, together, they do with what they receive. Given that much of what people do is enforce differentiations on each other, there is justification from such critical thinkers as Bourdieu to present culture as inherently violent over many if not all the individuals who cannot escape living by the distinctions it makes. Thus, "in America", it is all but impossible not to be measured as a distinct individual, not to be identified as a person with these measured characteristics, not to be treated differentially in terms of these measurements. And it is not possible to escape this process through which the more exact the measurement, the less one will be able directly to challenge the status that one will eventually be assigned (e.g., "successful athlete", "gifted student", "depressed man"). Nothing is more violent, in a meritocracy, than a perfectly fair test. The "freedom" to be one individual self has a very dark side, as well as the very bright one for which America likes to take credit. It is in this sense that I maintain that America is an immortal cultural fact. Once, I summarized the main aspects of this argument in the following terms: "Culture has less to do with the habits we have internalized than with the houses that we inhabit" (Varenne and McDermott 1998: 14). I would now add that culture also has to do with the redecoration of these houses (Miller 1986), and, though more difficult, the building of new partitions and rooms. Culture has to do with the temporal processes that make the immortal facts people work at subverting. In this chapter, I have only space for showing how people do, in the detail of their participation, "redecorate their rooms", to expand the house metaphor. I do this partially by confronting scholars from Garfinkel to Merleau-Ponty. Culture, I have argued, is not about learning. In the details of its unfolding, it is about "instructing" and not quite doing what In other words, I am divorcing this presentation of culture from the tradition that starts with Weber (1949), and then moves on to Parsons and Shils (1951) to culminate with Geertz (1973) and his students.

184

Herve Varenne

one has just been told to do. Garfinkel (2002) argues powerfully that cities, streets, laws, regulations, customs, etc., all exist in the instructions all those who must live by them cannot help but give each other. And, in the complex societies we all inhabit, cities, streets, etc., all exist through the instruction those who instruct us in our daily lives themselves receive - in a long chain of instructors where one finds oneself alternatively instructor and instructed - perhaps contradictorily so. Those instructions are required because of: (l)the arbitrariness of the overall conditions (similar conditions do not always produce similar effects); (2) the constitution of this arbitrariness through distinctions and substractions that must always leave many aspects of experience outside what can easily be handled (fundamental to all structural linguistics); (3)the incompleteness of the constituted world (thus the "etc." principle in ethnomethodology); (4) the unpredictability of what will have to be dealt with next, and the arbitrariness in that next move; and (5) the incompleteness and unpredictability that make all the "learning" that may occur not only of limited usefulness, but even possibly dangerous.5

Both Bourdieu and Garfinkel emphasize that social order cannot be understood in terms of rules people would follow. But Bourdieu moves from there to the habitus (cf. Fetzer this volume), thereby returning to inculcation and learning. Garfinkel moves to instructed achievements by "congregations" (to use the word Garfinkel use for what I call "polities of practice"). Moving to instruction makes the most sense. It makes sense because it allows for more inclusive accounts of the kind of interactional sequences we wish to understand. Most importantly it makes sense because it suggests the mechanism for the production of extraordinary "understandings" (i.e., the kind of textual representations that constitute poetry, philosophy, the sciences) as well as for the production of new arbitraries, both locally and at the level of congregations of congregations, or overall polities. I am developing a line of inquiry that I ground with Merleau-Ponty's discussion of speaking through ordered and ordering structures in his "Prose of the World" (1973/1969). This applies most directly to the kind of minimal inflection of a medical interaction that I am using as an example. In conclusion, I will suggest further how this might be generalized to fulfill the program I am setting. Better than most, Merleau-Ponty understood that personal expression, what he called "enunciation", precisely required a structure to take shape. He was building on Saussure and thus his sense of structure does not appear to have the dynamic quality that ethnomethodology acknowledges when it talks about "ordering" as an ongoing process. But his insistence that every speech act must be somewhat different from any other cannot be 5

Think about driving in unfamiliar territories: best to unlearn much that one took for granted though what exactly that might be is not even itself obvious.

The Production of Difference in Interaction

185

ignored. It makes sense to face the facticity of constituted order because one can observe people always taking them into account as they struggle. But they do struggle and thus what may look to casual observers (or to analysts attempting to model the major features of the order) like a process of reproduction cannot actually be so, precisely because the speaker (and of course the audience) has never been in the total situation that required the speech act. At this point, those interested in order would look for the instruction that will replace the possibly disordering act within the parameters of what can be actively ignored. But those interested in the production of new orders would focus on the work everyone else must now do to deal with the difference that the disordering act produced. They would investigate the temporal echoes of this difference: how long will it be remembered in the very acts that all will then perform? We have many examples of such investigations: When Jakobson (1960) wrote about poetry as play on the sound systems of particular languages, he was also writing about the extraordinary matters that can be produced with an order that had not, apparently, been designed to produce it. Levi-Strauss (1966/1962) generalized this when he wrote about bricolage. Boon (1999) has been going much further as he writes about the "extra-vagance" of all cultural productions. Recently Miller (1986, 2001) has concretely investigated all this in the realm of the most commodified objects. The evidence that people strain the limits of constraints they cannot escape is everywhere (cf. Hundsnurscher this volume). The remainder of the paper is another case of a struggle with and through the re-ordering moves of all participants.

3. Making a Difference

I now try to make all this somewhat more concrete through an abbreviated look at one moment within a longer conversation controlled by interlocking polities, some local and some not local at all. This moment includes four persons laughing in unison followed by another person stating, with laughter in his voice "She is dangerous, you know!" followed by another person repeating "She is dangerous!" (see also Langleben this volume; Nemeth this volume, for treatment of repetition). This moment took place in the early 1990s, in a hospital located within a prosperous suburb of New York City. It took place during the labor and delivery of a professional woman, a medical doctor herself. This woman, "Lonnie" as she is called here, is surrounded by her husband, a nurse, a researcher, a gynecologist, and an anesthesiologist who together constituted the local polity. Together this crowd worked in the terms of an immortal social fact. I now explore this fact from the point of view of its arbitrariness, that is:

186

Herve Varenne

(l)the arbitrariness of the American organization for childbirth; (2)the further arbitrariness that had been produced at the local level; (3)the arbitrariness that all participants can introduce (even if they may not or should not).6

3.1. Differences that Made Initial Conditions There is a long tradition in cultural anthropology stressing the multiplicity of human ways for childbirth and the ever more complex practices modernity has made. It has been demonstrated again and again (Davis-Floyd 1992; Davis-Floyd and Sargent 1997; Frieds 1980; Jordan 1978) that there cannot be any such thing as a "natural" childbirth that would only be a private affair between a biological female and the young of the species who is being bom. Everywhere that anthropologists have looked, childbirth, like motherhood in general, triggers massive cultural responses (Drummond 1978; Scheper-Hugues 1992). Giving birth is necessarily a social and political event. The major institutions of any societies are involved. Nowhere is this truer than in the industrial societies of the past two centuries. The radical transformations that have occurred during this period only serve to make Drummond's point ever more poignant: every new technological development, from forceps to epidurals may have "solved" some earlier problems and undoubtedly saved lives, as well as arguably made the process of giving birth more comfortable. They have also produced new interactional problems as Rapp (1999), for example, has demonstrated. It may still be the hope of those in the medical sciences that the process of childbirth can be taken out of the vagaries of ideology, prejudice or politics, and given to "experts" whose legitimacy cannot be challenged both because they work within the boundaries of applied rationality and because of the evident good that their work produces. But precisely on the matter of childbirth, even a cursory look at its recent history reveals that this hope is not realistic: This analysis is based on extensive work done on this case by Cotter (1996). A more detailed analysis of this case can be found in our "Dr. Mom?" (Varenne and Cotter 2006). For this work she followed Lonnie throughout her pregnancy, attending and videotaping all prenatal visits to the gynecologist. She videotaped the last two hours of the hospital labor. She focused on the interactional production of contractions as a particularly telling instance when what appears commonsensically as something that happens within the body of the laboring woman can be shown to involve all participants as, together with the woman, they account for the strength of the contraction, and the amount of pain Lonnie is really feeling. This is brought to sharp focus after the epidural when the reading of a trace on the monitor becomes key in the contested construction of how well the labor is progressing. Cotter discusses in detail the many kinds of talk performed during the labor and their participatory structures. She gives particular attention to the local significance of the woman's professional status and its apparent erasure. This argument is developed in a draft paper available at the following URL: http://varenne.tc.columbia.edu/hv/doc/ this site also contains most of the transcripts used in Cotter's work.

The Production of Difference in Interaction

\ 87

every new procedure triggers new forms of contestation. I mention only the matter of painkillers. The national health systems in Europe and the United States differ on such a fundamental matter as the "best" (i.e., most "rational") way of administering painkillers. The regulators of some nation-states "trust" people in pain to self-administer while the regulators of other nation-states do not. Such decisions have massive practical consequences in the detail of medical interactions. The decisions place the person of focus, and those others most significant at the particular moment, in situations that constrain all of them in very different ways. People in pain in America must work to convince medical personnel that they need painkillers (and medical personnel must take the position of someone who is waiting to be convinced - even if they already are). In Sweden, the work will be different, and so will be what is to be normal medical interaction, what will be trouble, and what kinds of instructions the participants must give each other whatever they are doing. The hospital where the case summarily presented here took place is one of those that will not let the delivering woman, Lonnie, self-administer painkillers. Lonnie is thus in the position of having to make the case that she needs painkillers not only in general (e.g., before the labor starts), but also continually throughout the labor. As Jordan (1993/1978) documented, this produces specific kinds of conversations. These conversations involve an examination by an authoritative person of a human body who, for the duration, is subjected to what Foucault would call a "panoptic" gaze. These conversational practices constitute the hegemony of medical modernity in the most local of polities when pain is brought out by any of the co-participants as potentially requiring further action. These conversations cannot be escaped. But, and this is central to my argument, being caught, on a hospital bed, in a continuing discussion about the pain one is experiencing, having to make a rhetorically powerful argument that this pain is sufficient to warrant more painkillers, all this required activity cannot be presented as the product of misunderstanding or apathy. Rather, the overall situation must trigger a heightened "culturing" response that is productive of something new that all must face - even if only to ignore it. It is somewhat different from the most expectable line of interactional development. Thereby it produces a new, possibly very local and temporarily short, arbitrary re-organization of the original conditions - productive, in other words of a subculture.

3.2. Differences that are Further Elaborated A little bit more information about the moment introduced earlier will help illustrate what "culturing" can produce. The moment comes half-way through the labor, after an epidural

188

Herve Varenne

has been administered.7 The anesthesiologist comes in and asks How are we doing! Everything, including the royal -we indexes the moment as an expert examination that produces an initial sequence quite similar to those Jordan documents. It produces a challenge from Lonnie: Do you feel guilty that I am still feeling some painl (a paraphrase of sec. 43 [see Appendix for transcript]) and a defense by the anesthesiologist / think I did a good job. These exchanges are on a bantering mode leading to a counter-challenge by the anesthesiologist: Did you really feel no pain in your earlier labors when they had to use forceps and you were in the Ο.Κ.Ί (paraphrase of sec 57-65). This is followed by general laughter and it allows, or perhaps invites, Lonnie to tell a story about her preceding labor. I refer to that earlier labor as the "storied labor" since there are no independent statements about it besides Lonnie's story. She starts with the classic beginning for the telling of a (possibly) tall tale: Listen, do you want to hear about the C-sectioril A frame that is also a challenge has been set and all join; for close to a minute a joke is told through them (in Sacks 1974 sense). The joke is told in the shape of a story in the telling of which all participate. Here is a paraphrase: Listen! I had pushed for three hours and the epidural had worn off. They said I had to have a C section. This little Chinese anesthesiologist came and he could not get the tubing undone. He was running around screaming "give me a cramp." I didn't trust him. He was trying to do my sensory level with the alcohol wipe. I kept saying that I could feel the alcohol. So he gave me too much. I was in the recovery room for five hours. I thought I was never going to walk around.

This may not appear particularly funny to some readers, but it does start in laughter, the co-participants amplify (sec. 97-98, 102, 104, 109-11) and chuckle (sec. 91). And the whole thing culminates with the laughter mentioned earlier (sec. 118). This laughter is itself amplified with the two She is dangerous utterances, both of which are spoken with laughter in the voice (sec. 119-121). In the process of the story, many matters are indexed: race, foreignness, trust, medical knowledge, etc., I focus on only one of these indexes: the index to pain and particularly to asking for pain medication. The story would not work if it did not play with a central feature of this situation: a woman in labor must ask for painkillers, anesthesiologists must check using authoritative means: if you know something about these means, you might be able to lie about your pain and thus receive extra painkillers - you might be able to pass effectively as being in pain. We are not here in the world of habitus. We are in the world of Garfinkel on passing, and also in the world of de Certeau on laperruque, and the world of Levi-Strauss on bricolage. 1

See Chapter 12 in Cotter's (1996) "Labor Negotiations" for further details, other stories, and several other examples of joking (e.g., labor can also be fun).

The Production of Difference in Interaction

\ 89

We are also in the world of what Geertz (1972) called "deep play" and possibly even at the edge of the world of what Garfinkel (1956) called "degradation ceremonies". Playing with painkillers is rarely funny, and it may always be dangerous. The punchline of the story we are reading is They gave me so much, I was in the recovery room for five hours, numb [...] / thought I was like never going to walk again. This is the utterance that is followed by general laughter, and by the comments about Lonnie being dangerous. There is no ignorance or confusion here, and there may not have been any at the time of the storied labor - except perhaps the induced ignorance of the anesthesiologist about the state of Lonnie's body that led him to deliver a second dose of anesthetics. One of the great merits of Geertz's (1972) famous paper on the Balinese cockfight is his introduction of the notion of "deep play" as capturing something powerful that made sense of something he had observed in Bali when powerful Balinese men apparently put their full status on the line. His particular analysis concludes that, when they do this through their cocks, nothing much happens: the status rarely changes. It may be so in such cases, but it need not be so. In the case we are examining, Lonnie's radical departure from the expected is a different matter: her many statuses as wife, mother, even medical doctor could have been transformed if something had gone really wrong. Her act did change the course of the storied labor as it made the C-section more dangerous than it would otherwise have been. Thus she did make history for herself and her family. We can imagine what might have happened if her baby had died: some further radical things might have changed not only for her or her husband but even perhaps for the anesthesiologist, the hospital, and others, in an unpredictable reverberation up and down various networks of institutions. Sanctioned delivery practices do change and the change may be triggered by the collective appropriation of a single act. Following Cotter (1996), I am not going to explore why Lonnie lied during her storied labor. I do not speculate either as to why the participants reacted the way they did. I simply note first that such a sequence is allowed by the very organization of hospital childbirth. To the very extent that one has to ask for painkillers, that being given the painkillers is a conversational process directly implicating "trust" at various points, then the openings for all kinds of other troubles are built in - including the possibility of lying about pain. All these openings would be closed if one did not have to ask. These openings, the ensuing troubles, instructions, and the repairs after sanctions, thus necessarily make each actual instance unique in the detail of its unfolding. Hospital labor, like medicine in general may be as hegemonic as many have claimed, but it does not determine what is to happen in any medical interaction. Rather it produces particular forms of trouble and opportunities that we should expect participants to exploit in a precisely unpredictable fashion. The interesting question now becomes: how can a unique instance of collective enunciation make any difference? Structural analyses are famous for their power to reveal the ca-

190

Herve Varenne

nonical in the detail of the production of the particular as particular. They are also infamous for hiding the very production of the particular. But this erasure is not a necessary feature of a structural analysis, on the contrary perhaps. By focusing our attention on the work everyone performs to produce what will then be accepted as orderly, structural analysis can help us notice the very acts of erasure, or, on the contrary, the acts of inscription that can then become fact for future participants. In this case, it does appear that Lonnie's adventures in the storied labor made some significant difference at the time (by necessitating a reconstitution of the delivery from "normal" to "difficult"). Interestingly, it appears to have become part of the history of the family - the husband clearly already knew the story Lonnie told. In other words, the "difference" that Lonnie introduced at the instant when, for whatever reason, she lied to the Chinese anesthesiologist, made no difference to him, made some difference to her gynecologist then as the delivery shifted to emergency Caesarean (a "normal" reconstitutive process in any event), and then made some difference in the family. I can now focus on the production of difference telling the story about lying made.

3.3. Differences that May Not Make a Difference Having a child in a hospital is not a routine matter. It is not produced by the acquired "common sense" of the participants, but rather by the instructions they give each other as to what to pay attention to and what to ignore. It will be, in every case that we examine closely, an extraordinary event, even if the extraordinariness is "forgotten" later (i.e., even if it does not lead to transformation in either the life of the participants or the organization of the institution). But forgetting that which made a difference in a performance must not be confused with the fact that a difference was produced. Indeed, in many cases, this forgetting must be an active process of, to mention recent Garfinkel (2002) again, ignoreance; it may well be that hegemonic powers exercise themselves more through the active erasure of noticeable difference than through the psychological processes that would be necessary for the difference not to have occurred.8 Be that as it may, I propose that Lonnie's joking narrative be understood in the context of what Merleau-Ponty (1973/1969) discussed as le langage parlant. At the moment of enunciation, when Mead (1913) would say that / speak, the very structuring of language Foucault, as well as most of those who follow him, presents the success of the panopticon as produced by what happens within those who are being watched by an invisible warden. Less attention is paid to the activity of the warden and what he must do to remain invisible. I suggest the warden's work may involve a lot of actively ignore-ing what the inmates are doing, an ignoreance that the inmates would then take into their own accounts.

The Production of Difference in Interaction

191

(and interactional patterns) must produce a unique and unpredictable statement that can only be "textualized" in a secondary step. In that secondary step, an utterance may be transformed into a myth,9 or it may be actively forgotten by being reintegrated within the regular discourse that is supposed to account for such utterances. In many ways, the story Lonnie tells about the earlier labor is such a family myth told in a setting well-organized for the telling of such myths. The telling itself is potentially productive of something in its actual setting during the current labor. What it is productive of is precisely what I am in danger of erasing if I write it, "interpret it", as a warning by Lonnie to the anesthesiologist not to take her seriously when she asks for more painkillers. I could have used the language of semi-conscious strategies and intention and claimed that, given the overall sequencing of the tale, it must have been intended to be such a warning. Or, to take a more social deterministic position, even if this was not Lonnie pre-story intention, the warning may be what the story produced in the anesthesiologist. My issue here is not to decide which (speech) act the story produced, but to stress the unmotivated, arbitrary, reality of the utterance as spoken, and then as disappeared. It is not that it does not make "common" sense: all participants work with it as a humorous story. It is rather that the manner of the sense-making is not pre-ordained or simply rational (orderly) within the overall factual sequence (i.e., "hospital labor"). The relationship of any utterance (or sequence of utterances) to any model we might have of canonical sequences is to be dealt with as the same kind of relationship Jakobson (1960) posited between the phonetics and rhythms of any language and its poetry. Lonnie's tale is play with a property of hospital labor and the discourse of the medical examination in general. It is also play that, on the day of its performance, did nothing we could observe - except make people laugh. It was not mentioned again in the following hour of labor.10 None of the participants refer back to it, not even jokingly. It is as if the story had never been told.

10

I am indexing here Levi-Strauss' (1971: 560) definition of a myth as a past utterance from a single author as collectively known and worked with. Arguably, Levi-Strauss is directly developing here Merleau-Ponty (1973/1964) and restating Mead's (1913) analysis of the postulatable / as subject of the objective me that becomes known in the rhetorical forms of a particular polity of practice at particular times within its operation. It is possible that the anesthesiologist was more on his guard after the story than he may have been. He may have rationed painkillers later in the labor, but we have no evidence of conversations he may have had with the nurse or other personnel.

192

Herve Varenne

4. Making Cultural Facts for the Future

Through a continuing discursive process, a "modern" "American" way for hospital labor has evolved altogether extravagantly (Boon 1999), over several million years of hominid evolution, and over at least 100,000 years of human transformations, all the while continuing to preserve all biological and sociobiological structures. In the 1990s, this American way for labor may have become so powerful to appear "normal" if not "natural" to many in the United States. This may be mostly the case for those who have never actually had to give birth to a child, or who end up talking about giving birth outside the moments of the actual experience. At that moment, one woman, doctors, nurses, and now any number of "others", form a temporary version of a more general polity of practice and they produce a unique version of the sequence through various kinds of play. This act may become fact in the history of the participants as it is available for further elaboration in the detail of the conversation that constitutes future labors. Then again, this act may become "nothing" and disappear in the blur of the "normal". Or else it may make history in radically different contexts, as the case I have presented did for Cotter, myself, and now the audiences of the papers we have written. The call here is for conversation analysts to pay more specific attention to the processes that make a difference in history. This will involve analyses of the processes that actively erase evolved differences, hide them, and make it appear "nothing happened" that would require the meting of consequences by those who sanction wayward performances. Much of this work will address classical issues in what has been called "reproduction", what is sometimes also written about as reconstruction or reconstitution. Mullooly and Varenne (2006) have explored such erasure in the context of classroom play where teachers may be involved in the active ignore-ance of what may have been required they do if they had performed any kind of acknowledgment (cf. Hudelot this volume). But we do need much more work, starting at the most local of levels where discourse analysis is most effective. Then, we might be able to see, perhaps through various forms of quasi-pronominal anaphora, when something that is marked as having occurred in an earlier sequence is used to produce difference in further sequences - and also, and possibly more common, to erase the possibility of difference by reconstituting the extraordinary past as ordinary for all future purposes. But we must also move on to longer sequences bringing together ever more speakers. Some cases of deep play produce change in the status of the protagonists. Violent revolutions do take place and, even if it can be argued that much that is reconstituted after such revolutions echoes older patterns, some revolutions do make history.

The Production of Difference in Interaction

\ 93

References Apple, Rima D. (ed.) (1990): Women, health, and medicine. A historical handbook. - New York: Garland Publishing. Apple, Rima D./Golden, Janet (eds.) (1997): Mothers and motherhood. Readings in American history. - Columbus, OH: Ohio State University Press. Bateson, Gregory (1955): The message 'This is play'. - In: Bertram Schaffher (ed.): Group processes. Transactions of the second conference, 145-242. New York: Josiah Macy, Jr. Foundation. Boon, James A. (1999): Verging on extra-vagance. Anthropology, history, religion, literature, arts... showbiz. - Princeton, NJ: Princeton University Press. Bourdieu, P. (1977/1972): Outline of a theory of practice. Trans. Richard Nice. - Cambridge: Cambridge University Press. Cotter, Mary (1996): Labor negotiations. A study of interactions during hospital childbirth. - New York: Doctoral Dissertation, Teachers College, Columbia University. Davis-Floyd, Robbie (1992): Birth as an American rite of passage. - Berkeley: University of California Press. Davis-Floyd, Robbie/Sargent, Carolyn Fishel (eds.) (1997): Childbirth and authoritative knowledge. - Berkeley: University of California Press, de Certeau, Michel (1984/1980): The practice of everyday life. - Berkeley: University of California Press. Drummond, Lee (1978): Transatlantic nanny. Notes on a comparative semiotics of the family in English-speaking societies. -American Ethnologist 5, 30-43. Durkheim, Emile (1982/1895): The rules of the sociological method. Trans. W.D. Halls. - New York: The Free Press. Eakins, Pamela (ed.) (1986): The American way of birth. - Philadelphia: Temple University Press. Fried, Martha/Fried, Morton (1980): Transitions. Four rituals in eight cultures. - New York: W.W. Norton. Garfinkel, Harold (1956): Conditions of successful degradation ceremonies. - American Journal of Sociology 61,420-424. Garfinkel, Harold (2002): Ethnomethodology's program. Working out Durkheim's aphorism. Lanham, MD: Rowman and Littlefield. Geertz, Clifford (1973): The interpretation of cultures. -New York: Basic Books. Goodwin, Charles (1981): Conversational organization. Interaction between speakers and hearers. New York: Academic Press. Jakobson, Roman (1960): Concluding statement. Linguistics and poetics. - In: Thomas A. Sebeok (ed.): Style in language, 350-377. New York: Wiley. Jordan, B. (1993/1978): Birth in four cultures. A crosscultural investigation of childbirth in Yucatan, Holland, Sweden and the United States (2nd ed.). - Prospect Heights, IL: Waveland Press. Lave, Jean (ed.) (1984): Everyday cognition. Its development in social context. - Cambridge, MA: Harvard University Press. Lave, Jean (1988): Cognition in practice. Mind, mathematics, and culture in everyday life. - New York: Cambridge University Press. Lave, Jean/Wenger, Etienne (1991): Situated learning. Legitimate peripheral participation. - New York: Cambridge University Press. Leavitt, Judith (1986): Brought to bed. Child bearing in America, 1750 to 1950. - New York: Oxford University Press. Levi-Strauss, Claude (1963/1958): Structural anthropology. Trans. Claire Jacobson and Brooke Schoepf. - New York: Basic Books.

194

Herve Varenne

McDermott, Ray P. (1993): The acquisition of a child by a learning disability. - In: Seth Chaiklin, Jean Lave (eds.): Understanding practice. Perspectives on activity and context (Learning in doing. Social, cognitive and computational perspectives), 269-305. New York: Cambridge University Press. McDermott, Ray P./Gospodinoff, Kenneth (1979): Social context for ethnic borders and school failure. In - Aaron Wolfgang (ed.): Nonverbal behavior. Perspectives, applications, intercultural insights 175-195. New York: Academic Press. McDermott, Ray P./Tylbor, Henry (1983). On the necessity of collusion in conversation. - Text 3/3, 277-297. Mead, George H. (1913): The social self. - The Journal of Philosophy, Psychology and Scientific Methods 10, 374-380. Merleau-Ponty, Maurice (1973/1969): The prose of the world. Trans. John O'Neil. Evanston, IL: Northwestern University Press. Miller, Daniel (1986): Appropriating the state on the council estate. - Man 23, 353-372. Miller, Daniel (ed.) (2001): Car cultures. -New York: Berg. Mullooly, James/Varenne, Herve (2006): Playing with pedagogical authority. - In: Judy Pace and Anne Hemmings (eds.): Classroom authority. Theory, research, and practice, 63-86. Mahwah, NJ: Lawrence Erlbaum Associates. Ochs, Elinor (1979): Transcription as theory. - In: Elinor Ochs and Bambi Schieffelin (eds.): Developmental pragmatics, 43-72. New York: Academic Press. Parsons, Talcott/Shils, Edward (eds.) (1951): Toward a general theory of action. -New York: Harper and Row. Rapp, Rayna (1999): Testing women, testing the fetus. The social impact of amniocentesis in America. - New York: Routledge. Sacks, Harvey (1974): An analysis of the course of a joke's telling in conversation. - In: Richard Bauman, Joel Sherzer (eds.): Explorations in the ethnography of speaking, 337-353. New York: Cambridge University Press. Sacks, Harvey (1975): Everyone has to lie. - In: Mary Sanches and Ben G. Blount (eds.): Sociocultural dimensions of language use, 57-79. New York: Academic Press. Sapir, Edward (1921): Language. - New York: Harcourt, Brace and World. Sapir, Edward (1966/1949): Culture, language and personality. Selected essays. - Berkeley: University of California Press. Sargent, Carolyn Fishel/Bascope, Grace (1997): Ways of knowing about birth in three cultures. - In: Robbie Davis-Floyd, Carolyn Fishel Sargent (eds.): Childbirth and authoritative knowledge, 183208. Berkeley: University of California Press. Saussure, Ferdinand de (1981/1915): Cours de linguistique generate. Edition critique proparee par T. de Mauro. - Paris: Payot. Scheper-Hughes, Nancy (1992): Death without weeping. The violence of everyday life in Brazil. Berkeley: University of California Press. Stratton, Alison (2003): The cultural work of hard of hearing in Sweden. - New York: Doctoral dissertation, Columbia University. Varenne, Herve (1977): Americans together. Structured diversity in a Midwestern town. - New York: Teachers College Press. Varenne, Herve (1983): American school language. Culturally patterned conflicts in a suburban high school. - New York: Irvington Publishers. Varenne, Herve (1992): Ambiguous harmony. Family talk in America. - Norwood, NJ: Ablex Publishing Corp. Varenne, Herve/Cotter, Mary (2006): Dr. Mom? Conversational play and the submergence of professional status in childbirth. - Human Studies, 1 -29. Varenne, Herve/McDermott, Ray P. (1998): Successful failure. - Boulder, CO: Westview. Weber, Max (1949): The methodology of the social sciences. Trans. Edward A. Shils/Henry A. Finch. - New York: The Free Press.

The Production of Difference in Interaction

195

Wenger, Etienne (1998): Communities of practice. Learning, meaning, and identity. - New York: Cambridge University Press.

Appendix A Transcript - 9:17:50 to 9:19:51

frame #

Wife

Husband

no, in my other labors

ι

I didn't feel any pain

!

Anesthesiologist

57 58 59

918:50 ....yeah

1:00

really? ....

61 62

.... the

1

j

one with the forceps

L 1

and the .... in the O.R. and

1

what was the other one? [GENERAL ********* LAUGHTER]

!

yeah ....

I

Cotter sec.

':

63 64

65 66 67 68 69

listen you want to hear iiabout .... 919:00

the C section?

I

il.10

if.... I pushed for

71

•jthree hours, no but the ifepidural

72

I

?

had worn off. ....

'I pushed for three hours I I... then a

jdifferent anesthesiologist jcame, athis little Chinese fellow, it 'was in Hightown hospital, 9J9:iO so they said they have to do a C section ilxxx because you're not Unot pushing well

73 74 75 76 77 78 79 1.20

81 82 83

196

Herve Varenne

Wife

frame #

Husband

|

Anesthesiologist

Cotter sec. 84

("" ' "

jHe couldn't take the tubing pondone

85

|

86 87 88 89

....couldn't get the clamp apart

I

I.... he was running around IScreaming 'cramp' 'cramp'

1.30

fgive me a cramp' .... I

iwe couldn't he couldn't

|****|

j

Iget the thing apart to give =me |the bolus, I didn't trust him ilmean .... j if he couldn't find a clamp then what good is he?

Iget the thing off so

919:30 Sthe he was trying to do

•f ·'····" "···"·

,my sensory level .... with the

.... ah ah

alcohol wipe oh, that's xxx I kept saying even though I couldn't couldn't ifeel the alcohol ι tkept saying I could Ifeel it ....

yeah, I can

919:40

still feel it yeah give me !.... xxx gave me so much U'm telling you .... I was ' in the recovery room for five hours .... numb .... they brought me back I thought I was like never going to walk again ι ;Ihadso much ....

919:50 1

;

91 92 93 94

j.... he couldn't

.

!***!

more medication ....

[GENERAL ***** LAUGHTER] she is dangerous you know

'she is dangerous because

95 96 97 98 99 1.40 101 102 103 104 105 106 107 108 109 1.50 111 112 113 114 115 116 117 118 119 2:00

l_.IL_J.j

.... !******!

she'll keep asking for it ....

121

The Production of Difference in Interaction

197

On the Research, the Taping, and the Transcription Conventions Used Here The original research (Cotter 1996) was conducted in the early 1990s and involved videotaping all prenatal visits by Lonnie (a pseudonym) to her obstetrician, as well as the labor and the first get together of the parents, the new baby and his sibling. All observations and taping were done in full view of the participants who were aware of the research goal and had given their permission according to the standards of the time. The videotapes were transcribed through paraphrases of the combined visual and audio stream at about oneminute intervals. Selections were selected for more detailed transcriptions of the recorded interaction. Our first priority in these transcriptions was the representation of three central properties of the event: co-participation of all presents, postural shifts, and semantic content. We decided not to use the usual conventions from conversational analysis as they tend to obscure the active co-participation of participants who may be silent at any particular time (Goodwin 1981), and do not allow for easy incorporation of visual information that can now be made available through thumbnails of frame grabs. Details about the transcription conventions, and the justification for audio transcription can be found elsewhere (Ochs 1979; Varenne 1992). In brief, each line of the transcript represents one second. [....] indicates a half second of silence. One frame grab is included for each 10 seconds. We did not attempt to catch most of the variations in the phonetic aspects of the verbal stream since we were not using these for analysis.

Douglas Macbeth

Sequential Analysis in an Ethnomethodological Key: Order without Theory

This chapter discusses the possibility of pursing studies of language-in-use that have no particular need for theory. It does so in the particulars of ethnomethodology (EM) and conversation analysis (CA). In disciplining their inquiries to the methodic character of member methods, EM and CA propose an analytic program that has no more use for formal theory than the natives have. They set out to demonstrate that the promise of naturalistic inquiry possesses a basis-for-inquiry premised not in sociological theory, but in the durable order of everyday life that precedes theory. The chapter concludes with an analysis of a class coming to order as a practical exhibit of order without theory.

1. Introduction

I want to begin with the central phrase of my title: "sequential analysis in an ethnomethodological key". By sequential analysis, I of course mean the program in studies of natural conversation initiated by Sacks, Schegloff, and their students and colleagues. And in linking it to ethnomethodology, I am only referring to the well known history of how their initiatives and those of Garfinkel found each other, and found each in the other an elaboration of a common interest in what can be called, in Garfinkel's phrase, "the routine grounds of everyday life" (1967). Their respective and common enterprises have been thoughtfully reviewed in a collection of subsequent works (cf., Button 1991; Heritage 1984; Livingston 1987; Lynch 1993, 2000; Maynard and dayman 1991). Within that shared history, three inflections are central to my argument. The first is their common disciplinary homes: Garfinkel, Sacks, and Schegloff are sociologists. The classic problems of social order are central to their disciplinary matrix. A second piece of context is how it was that their inquiries were part of a more general dissatisfaction in American social science, and sociology in particular, with its familiar forms of analysis. Summed by Garfinkel as a program of formal analysis, the conventional wisdom was - and is - that the meaning, order and structure of ordinary worlds were to be explained via the discernment

200

Douglas Macbeth

and arrangement of generic, theorized structures yielding a synoptic point of view, or in more contemporary parlance, a "god's eye view". The criticisms of this wisdom were on behalf of something like naturalistic inquiry, or the premise that there were fields of social life - extraordinary in their diversity and extensiveness - that could only be taken up by direct study. Both the dissatisfactions and the proposed remedies were leveraged upon a view of (again borrowing from Garfinkel) "ordinary immortal society" as a pre-theoretic life world. The dissatisfactions with normative social science were themselves diverse. But in my view, Garfinkel's ethnomethodology wrote the most radical break with social science as we know it. By "radical" I don't mean politically radical, but analytically radical: a decisive break or departure from our analytic habits. Garfinkel repeatedly discusses EM's program as the study of worlds already in possession of disciplined discourses on things like order, logic, structure, rules and roles. By this account, ordinary social worlds already own social sciences' identifying topics, owning them as practical work in the world. Assembling "the objective realities of social facts", for example, is an ongoing task and achievement of everyday life. The very term ethnomethodology points to this domain. Rather than standing for a professional methodology, it only and exactly formulates a topical domain: the endless ways in which cultural members assemble cogent worlds in common, worlds that can be inquired of, disputed, reasoned with, agreed to, resisted, assessed, reckoned, described, etc. EM thus points to an orderliness that has a life in advance of professional analysis. This life is what others doing professional analysis take interest in. This is, on the one hand, an oddly simple-minded thing to say. On the other, the premise of a pre-theoretical world has, to say the least, its critics. The sum of my remarks is to address those criticisms. A third understanding has to do with the notion of sequential analysis in an ethnomethodological key. Of those member methods and what they achieve - things like common understanding, and thus order and recurrence - the sequential analysis of natural conversation was pointing to a primordial site of their enactment. As Garfinkel and Sacks (1970) see it, the mastery of natural language is the central identifying mark of cultural membership. It is also, and this is near the seat of their radical break with formal, professional analysis, through and through an analytic exercise, as in how we analyze the projectable completion of a turn underway and thus locate where our speaking might begin. From this it follows that ordinary cultural members are the first analysts on the scene. It was to this field of ongoing analyses that EM was pointing, and it is an analytic community that is not consulting theory as we know it. Their bases for "proceeding with their analyses" are entirely different, and Garfinkel, Sacks, and their colleagues and students took interest in what those bases are. In member methods, they sighted a field of methodic practices showing grammars of ordinary action to which professional sociological reason-

Sequential Analysis in an Ethnomethodological Key

201

ing, including theoretical reasoning, had no privilege or prior standing. They were positing something like pre-theoretical or pre-scientislic worlds as actual venues for naturalistic study. They weren't pursuing causal accounts, but production accounts of how sensible worlds in common are produced, in situ. These are descriptive programs whose aim is "not to override, undermine or discount the endogenous analysis; rather, it is to formulate how it is achieved in and as a methodic procedure" (Lynch 2000: 525). As it turns out, however, this first possibility of a life world available for explication as of its members' concerted methods for rendering their worlds sensible has fallen into dispute. What's in dispute is not so much whatever EM has to say about it, but it itself: the possibility of a life world already in possession of order, structure and coherence, whose possession it comes to endogenously, as culture, forms of life, and native pedagogies for teaching and showing how things like reason, civility, identity, conversation and the rest are done. The dispute is quite central to the possibilities of "order without theory", but before turning to it directly, I want to point out how much of EM's and CA's program has found its way into our common analytic culture. I mean here roughly the qualitative turn in social science. Though they own no single source, there are now prevailing orientations in the literature, and especially in natural language studies, of the following kinds that EM and CA have pursued over the last 40 years: that order is situated, meaning contextual, that action and context stand in a reflexive relation of mutual constitution and elaboration, that order, structure and recurrence are meaning's achievements, that ordinary worlds are already meaningful for their inhabitants, that discourse, and things like turn taking, sequence organization and talk-in-interaction more generally are produced, managed and oriented to by the parties to the occasion, and allied understandings. At least in the particulars of EM and CA, we could say that these understandings developed in the gaps of classical theory, in the places where classical theory took little interest, had no use, and could have no use. These were among the phenomena lost to the very methods of formal analysis. But we can also say that these gaps have been filled in an entirely different fashion by the contemporary turn in social theory, especially in Discourse theories. What I especially have in mind is the recent and maturing development of critical discourse analysis (CDA), which now has a substantial presence in educational studies and language study more generally1 (cf. Hallett and Kaplan-Weinger this volume). In CDA's view of things, these propositions about pre-theoretical worlds whose coherence can be found without benefit of a theoretical authorization are specifically mistaken.

A more considered argument along these same lines will be found in Macbeth (2003).

202

Douglas Macbeth

2. The Critique of Pre-Theoretical Worlds

The rejoinder to the posit of everyday, pre-theoretical worlds is a complex one. On the one hand, observations about pre-theoretical worlds are too palpable to contest directly. We see and live them as the routine grounds of our everyday lives, in freeway traffic, grocery shopping and ordinary conversation. Instead of contesting them directly, the critique of pre-theoretical worlds turns on their analytic status and presses certain implications. We can identify two versions of the objection to talk about pre-theoretical worlds, what we can call a weak version and a strong version.

2.1. The Weak Version The weak version articulates the familiar critique of "mere description". The characterization is familiar enough. It sets aside what could be deeply serious and analytically rigorous about analytic description, and leverages its critique on what everybody knows is the probity of "explanation", and with explanation the promise of causal links as we find them in natural science and presume them available to the human sciences. This is a very large issue in the philosophy of social science, through Weber, Dewey, Schutz, Rorty, Winch, Wittgenstein, and others. But perhaps more discrediting still, even as of this weak version, is the charge that in positing pre-theoretic worlds and a descriptive program for taking interest in them, we are simply returning to a classic empiricist genre, an objectivist premise that posits a world that could be so described. In the particulars of sequential analysis, the critique is familiar enough in the exchanges between Schegloff and Werthall (1998) and Schegloff and Biling (1999), and allied discussions of transcript as an emblem of objectivism (cf. Hundsnurscher this volume, for a critique). This indeed is a very serious charge - of a kind of residual positivism - for which some clarification may be helpful. Given the posit of pre-theoretical worlds, already ordered by the practiced forms of life that have taken root there, a proposal to study those grounds is something very much empirical. There are clearly empirical attachments here, as there are for any and every proposal for naturalistic inquiry. The error of the charge, however, is to mis-identify "empirical" with "empiricist". As Coulter (1983) clarifies, the "empirical" is not to be confused with empiricist programs for recovering "central tendencies", probabilities in the aggregate, and then fashioning from them claims of general structure and recurrence. What we find in EM and CA is something quite different.

Sequential Analysis in an Ethnomethodological Key

203

Working from the ground up, as it were, the practical intelligibility of actual occasions is the first organizational problematic on the scene. The aim is to study, in Lynch and Bogen's helpful phrase, "intelligible actions performed on singular occasions" (1996: 265), as in occasions of work in a science lab, or a story told, or, in the particulars of our materials, a class coming to order. These are not "objectivist" questions, but inquiries into demonstrable, accountable sense and meaning and the grammars of action that organize them.

2.2. The Strong Version The strong version is both stronger, and more interesting. It is, in sum, that any imaginable pre-theoretic world - including every imaginable ground for inquiry within or about it - is in truth, if the term can still be used, premised by prior organizations that can only be accessed theoretically. In this way, the strong version rules the positing out of relevance: Such places - pre-theoretical worlds - don't exist as analytically adequate grounds for inquiry. The characterization is implicit to the weak version. In the strong version, it is more brightly drawn. By this reading, order, meaning and most certainly structure are each an expression of prior states of affairs, even deeply historical states of affairs, that only theory can disclose. Every intelligible statement is endowed, authorized, issues from, or otherwise is an expression of a theoretical genealogy, whether of social structure, social facts, or power and historical epistemes. Culture itself, meaning not our accounts of it but itself, is an orderliness that can only be excavated theoretically; its organization, and thus explanation, lives in affairs that are only available that way. By this account, sense and meaning themselves own a theoretical foundation, whose analysis will not be accountably complete until those foundations have been established. The arguments of both versions show their continuities with traditions of general theory. Theory construction is inquiry's aim and authorization (cf. Berlin this volume). Yet there is, to my eyes, something novel in the contemporary discourses on discourse theory. I want to call it a third move - a double move - with respect to the critique of pre-theoretic worlds. It is a move that overtakes central programmatics of EM, CA, and other programs having to do with the primacy of local or ethnographic worlds and the order of social life in situ. What was for prior generations of social theory a colloquial landscape of no use to formal analysis in its welter of detail, has become a new and rich field of evidences of theory's range, probities, and synthetic powers. This move can be seen in contemporary discourse studies and critical discourse analysis especially.

204

Douglas Macbeth

2.3. The Double Move The "double move" shows a practice of theorizing that has been joined to naturalistic inquiry in a way that takes up ethnographic settings in their circumstantial detail as theoretical articulations. What it achieves is a vision of theory with its explanatory powers fully intact, but one that has use for vernacular worlds. Among other things, this theoreticalnaturalism proposes a new address to the perennial task of articulating the macro and micro social orders, and a new program for achieving the suture (cf. Berlin this volume, for a treatment of Grounded Theory). This naturalism is organized by a first move that theorizes the common topic of our inquiries, that is, "discourse", or conversation, or talk-in-interaction (e.g., Hallett and Kaplan-Weinger this volume). Discourse, which had been in the common parlance of our analytic community a natural object, an activity in the world, vernacular, unremarkably prolific in its production and diversity, and a topic in the study of what this talking species does, now becomes, and from the outset, a. formal object, abstract rather than concrete, generic rather than vernacular, conceptual and even historical, rather than praxiological. What I'm referring to here is of course an understanding of discourse with a capital "D", as Gee (1996) puts it, or "Discourse". Leveraged from the hugely influential continental literatures of the last 30 years - the work of Foucault and the critical theorists central among them - and quickly taken up elsewhere, Discourse is now an utterly familiar object in the literature. Rather than closely account for it, I only mean to point to it as a development that no one working in an academic department in social science or the humanities could possibly miss. The second move has to do with how discourse as vernacular practice has not, therefore, been erased. It is rather repositioned as a kind of parole to the formal structures of Discourse, expressing those prior, prevailing, historic-and-theoretic Discourses of power, knowledge, epistemes, and the like (cf. Chouliaraki and Fairclough 1999; Fairclough 1995; Freebody, et al. 1991; Gee 1996; Gutierrez and Larson 1994; Lemke 1995; Luke 1997; van Dijk 1993). In this way, these moves show the drilling down of theory's foundations not only conceptually (or theoretically), but also ethnographically and naturalistically. The double move insinuates theory's province into everyday life, identifying underlying causatives for its every vernacular expression. From this theoretical enfolding of everyday worlds the double move achieves a new form of theory that is both synoptic and local. Gee provides an emblematic formulation in a brief but forceful passage that displays the relation of Discourses to the ordinary, everyday expressions they order. Both synoptic and microscopic, they render the world - in every case - an exhibit on their own behalf: "Discourses are ways of being in the world, or

Sequential Analysis in an Ethnomethodological Key

205

forms of life which integrate words, acts, values, beliefs, attitudes, and social identities, as well as gestures, glances, body positions, and clothes" (1996: 127). It deserves a careful reading. The list stands on behalf of every imaginable expression of sense, meaning, structure or recurrence - including the formulation "forms of life" which was offered by Wittgenstein as therapeutic treatment for what he described as philosophy's "craving for generality" (1960: 18). The entirety of "saying and doing" verbs are to be understood as expressions of this first estate; Discourses (capital D) animate every way of speaking, every interest, task, gesture or occasion and, of course, what anyone happens to say. It is difficult to imagine a more aggressive, or perhaps even fundamentalist analytic program. Enormously generalizing, it seems unavoidably reductionist. The double move is one that both and at once naturalizes and renders abstract. Its practical effect has been to revise our ways of speaking of vernacular discourse and how we take interest in it.

4. Some Consequences

The notion of pre-theoretical worlds is entirely mistaken by these accounts, though we can observe that if indeed this is so, and theory's province is co-extensive with sensible worlds, then, among other things, this would be an extraordinary aggrandization for theorists: Theory's register is what theorists professionally own. In this light, it has the look of a privileging exercise, one that can only be democratized by insisting that, after all, and by these self-same arguments, we're all theorists to begin with. Thus, there's no privileging here, and the argument comes full circle. With these generic structures already in hand, we proceed much as normative social theory proceeded 50 years ago, with a program of what Garfinkel (1988) wryly described as one of "detailing generalities", wherein we stipulate to formal structures and then search for and claim their evidences. Though not at all a novel exercise, when theorized power is insinuated into the local order, the narrative becomes hyper-masterly, and nothing, as it were, escapes its gaze.2 A further consequence, in the particulars of EM and CA, is to rewrite some central insights and formulations. One follows from Sacks' "Notes on Methodology" (1984a), where he takes up the prevailing certainty in social science that social worlds are largely desultory landscapes, punctuated here and there by the formal structures and/or functions that No less wryly, de Certeau describes the search for evidence of theory's play: "Procedures without discourse are collected and located in an area organized by the past and giving them the role, a determining one for theory, of being constituted as wild >reserves< for enlightened knowledge" (1984:65).

206

Douglas Macbeth

give them order. To it he develops an account of social worlds possessing "order at all points". This conceptualization is quite central to natural language study. It gives grounds to figure that we could productively take interest in ordinary occasions of language use insofar as the very order of discourse is an orderliness that will be encountered at all points. To this stunning innovation, CDA writes a transforming amendment: Yes, we can take interest in ordinary worlds at all points, on the understanding that what we will find there is something like "theory at all points". Thus, the localism, the finitism is preserved, while inserting an entirely different, formal and professional genealogy of order. So too for the formulation of EM's program as inquiries into "intelligible actions performed on singular occasions" (Lynch and Bogen 1996). The occasionally of sense and meaning is preserved in CDA, but refers now not to the work of ordinary members, but rather to the expressions of prevailing Discourses, "on singular occasions". Continuous with these consequences is yet another. It has to do with how social science is positioned with respect to the worlds it surveys. The classical positioning, hugely taken for granted to this day, is that social science offers expert accounts. Thus, as a knowledge production engine, social science stands to common sense as a better form of knowledge, more considered, formal, synthetic, generalizable, etc. But expertise tends to be a zero sum game; we are authorized in the measure that others are not, and thus claims to see and detect the play of political-material-historic Discourses in the mereness of a native's gesture unavoidably entails a de-authorization of the person whose gesture it is to understand its sense and meaning. Fairclough's formulation of the "partially unsighted discourse subject" (1995: 45) is one of its more decisive formulations, and it is these very deficits of knowing that render resistance and the like heroic. Given an ambivalent authorization to know, those who resist are heroic in that classically ambivalent sense. Collecting the arguments so far, we may have come to a "relativist's impasse": Can there or can there not be a life world - endlessly many life worlds - in advance of theory's provisions, each possessing and displaying a methodic order of practice indifferent to and in advance of whatever formal accounts would make of them? Though I hope the lines are drawn well enough to be recognizable, I doubt that either side can pin the other. Arguments don't work that way, and nowhere is the "under determination thesis" more relentlessly in evidence than in arguments about the tasks, purposes and warrants of a social science. Rather than more argument, I want to consider some materials recorded in a 5th grade classroom in a public school in San Francisco.3 It is one of my favorite sequences and I regularly teach with it. I want to take it up to the question of whether we can find evidence The school is in a working-class district and is exemplary of the diversity we find in many urban American schools. The students are culturally diverse and include immigrant, nonnative speakers of English.

Sequential Analysis in an Ethnomethodological Key

207

for the order of everyday worlds as the order of competent practice. My interest is in how classrooms "come to order" and in so doing bring into alignment structures, purposes, identities, relations and even power as pre-theoretic, vernacular organizations in the room (cf. Froment this volume; Hudelot this volume, for other treatments of classroom interaction).

5. An Exhibit

The sequence begins as students are entering the room and taking their seats. There are multiple conversations going on as the teacher stands at the board surveying the scene. She writes and announces the page of their lesson for the day, speaking over the din of voices, and then begins with a first question that, as it turns out, is reformulated several times. If and when she receives an answer to her question, she and the class will have in hand the first sequence of their instruction, and arguably first evidence of a lesson underway. With an eye to this task, my interest in the sequence is in the pre-theoretic organization of a pause. A pause is an organizational thing that emerges exactly and only insofar as every party who witnesses it is engaged in its joint production; every pause is available to anyone's next remark as that which brings it to an end. They are exemplary of the proposition that organizational things are reflexive to the practices for achieving them. So what, then, in the particulars of our sequence, is the praxiological - and pre-theoretical - organization of the pause that we see in line 23 and its place in their joint construction of the order of the room? (see Appendix for transcript notations) (Beginning of a 4th grade math class.) 1. T: Kay:::. Page two hundred, an' twenty two.

2.

(8.0) ((chatter))

3. 4. 5. 6.

Kay, up till now::, (1.0) ((chatter)) you've been tawking about (.) jus- reg:ular fractions that equal each other, right?

7.

(1.0) ((chatter))

8. Can enyone give me two fractions that equal each other? 9. S: Urn, two- two

10.T:

// Two fractions that mean same thing.

11. 12. 13.S: 14. 15.S:

(2.5) ((chatter)) Any two fractions that mean the same thing. one an' two fifths. (2.5) ((chatter)) Equal fractions?

208

Douglas Macbeth 16 17.T: 18.S: 19. 20.T: 21. 22. 23. 24.S: 25.T: 26.T: 27. 28.

(1.5) ((chatter)) An:y two fractions. = = Two, two, two. (5.0) ((chatter, singing)) Let's start with one hhalf, maybe. (2.0) ((writes on board)) Who remembers a fraction that- etjuals one half? (2.0) ((chatter drops)) * One fifth. * (0.5) Who remembers a fraction that equals one half. R'member we can make it do anything. (3.0) What do we do to make an equal fraction, Mario...

I first want to consider the sequence of questions across lines 8, 10, 12 and 17 as an analytic organization. Each is a reformulation of its prior, all are built from the first one in line 8, and each is coupled to a trailing duration, most visible in lines 11, 14-16 and 19. In those durations, the teacher is assessing and revising her questions in light of what she finds. In this way each next question displays her analysis of the play of the last one, and each, and thus the sequence, is reflexively tied to the discourse in the room and her turn-by-tum analysis of it. Generally, she finds no responses that offer grounds for continuing. There are of course some responses, in lines 9, 13, 15 and 18, but they are apparently not promising, and the students' banter amongst themselves continues. We can also note a kind of slope to her questions. Though each is a reformulation of its prior, and all are iterations of the first one, we can see a progressive simplification of what, by the terms of the question, would constitute an answer. Thus at the outset, an answer would show two fractions that equal each other. By line 17, ''''any two fractions'1'' will do, and this shape points to how an "answer" is not entirely what the teacher is trying to produce. A reply is the better description. An answer may address, correctly or not, the substance of her question. But a reply will address the interactional order of her questioning, and reflexively produce it. So it is not only an answer that she is working to produce, but a discursive organization, and if it seems elusive it is because each of her questions calls for and cannot proceed without a student who self-selects or volunteers her candidacy. These first questions by the teacher (lines 8, 10, 12, and 17) do not themselves nominate next speakers, they solicit them. Said differently, they are prefacing questions oriented to the work of next speaker selection, and to the prefacing question Can anyone give me two fractions'? the students, in their continued orientations elsewhere, are virtually replying 'not me' (see Schegloff 1980 on 'pre' sequences). Though the students' conversations cannot be heard nearly as well as the teacher's remarks, we can nonetheless say this much about their analytic work. As they are conducting

Sequential Analysis in an Ethnomethodological Key

209

their several conversations with one another, they are also listening to the teacher's remarks to everyone.4 They are listening to what is said both within and across the conversations in the room for how places for speaking and listening are organized. These are analytic hearings, and their conversations are mutually contexting. And that the students are listening this way can be heard in and around line 23 where the "chatter" appreciably drops. We can hear it just as they do in its production, and notice - as they do - that line 22 is a different kind of question. 22.Who remembers a fraction that- eguals one half? 23.(2.0) ((chatter drops))

As a way of speaking of the difference between this question (line 22) and the prior ones, we can say that this one is not only a question of knowledge, it is a question that mobilizes a cohort. That is, the question Who remembers? identifies in its very intelligibility a membership category (i.e., the class). Its sense and sequential implicativeness is leveraged by addressing those who were there last time, and are therefore already candidates for next turn nomination (see Payne and Hustler 1980, on the work of "cohorting"). Reflexively, the question organizes a cohort of candidate respondents made of their now-relevant identities as "members of the class". In this light, the question Who remembers? can be seen as an analytic move by the teacher, shaped by her ongoing analysis of the talk in the room. It displays her analysis of the interactional orientations evidenced by the several student conversations that are competing with her own (i.e., an orientation to not becoming engaged in the teacher's gambit of prefacing questions), and sets them to work in a different fashion. It gives their demonstrable orientations and aversions a different object to avert. For the students, the question sets in place their occasioned identity as "the class" by shaping a context of choice for those who are competent to hear its terms, namely: For those of you who were here last time, you may continue your other engagements at risk of being nominated as one who might remember, or you may suspend your competing conversations and find anonymity from within the cohort organization of "the class". The choice urges them to become ordinary students, unremarkably there and without nomination-relevant features, and to even hear the choice is to become implicated in the production of the normal order of the room. Borrowing on Sacks' (1984b) phrase, the teacher's question encourages the students to become engaged in "doing being ordinary". Having produced the nomination relevance of the class, any next remark or engagement by any See Sacks (1992) on the warrant to listen in natural conversation. We need only posit an orientation to the possibility of speaking next to provide for how non-speakers would be listening to what is being said.

210

Douglas Macbeth

one of them can becomes grounds for an actual nomination, and this is the organizational thing that the students - each of them - see (and as we see in line 28, Mario is selected for the first direct question of the lesson). In this way, and as their individual and co-occurring analyses of these sequential-interactional horizons, they produce the pause and, in so doing, produce themselves as the cohort. Thus, by the teacher's production and their recognition of a discursive organization laden with social identities and moral organizations of rights and obligations, the order of the room and the identity of the cohort is reflexively assembled by students and teacher alike. This is the interactional work and achievement we hear across lines 22 and 23, and we could usefully say that they are the first "reflexive analysts" on the scene.

6. Conclusion

Hopefully, the sense of my subtitle - "order without theory" - has been clarified, if not agreed to. For this teacher and these students the relevant fields of order and meaning are local, praxiological, and endogenous to the intelligibility of things like a pause. Order, structure, identities and relations are thereby organized, and show organizations that are not only technical, but social-moral, too. The notion of pre-theoretical worlds does not, of course, mean that such worlds cannot then be theorized. Of course they can. To theorize is a marvelously unrestrained verb form; it will hook up to anything, and we can pose for any shred of sense or artifact unseen genealogies. That we can theorize whatever we choose cannot be in dispute. The argument is rather that the choice cannot be enforced, and what we find in EM and CA is, in Lynch's discussion, a disciplined "silence" about theory.5 This silence does not mean that those who do this work do not read theory, or do not otherwise have analytic commitments grounded in best understandings of how ordinary worlds work. As Lynch also observes, throughout Garfinkel's corpus we find a "comprehensive vision of how ordinary society organizes itself (1999: 219). It is rather that those best understandings return us to a world too diverse in its structures, sensibilities, practices, methodologies, cohorts, contexts, forms of life, language games, and the achievements of each to be accounted for by "a profesTheorizing is a curious topic that enforces its way of speaking on seemingly every effort to take it up critically. As Kavanaugh notes: "to espouse a position against theory is to make an integral move within the discourse of theory. [Citing Mitchell, he continues:] The antitheoretical polemic is one of the characteristic genres of theoretical discourse" (1989: 2). Nonetheless, I hope I have offered more than a polemic and other than theory.

Sequential Analysis in an Ethnomethodological Key

211

sionally fashioned nexus of definitions, propositions, and a priori expectancies" (Lynch and Bogen 1996: 273). Serres (1989) gives us the fetching image of theory's disputes as a dispute over who will be standing last in line in an epistemic queue, thus having last words.6 By the line of argument I am making, a certain leveling of the implicit epistemic queue is entailed. Sequential analysis in an ethnomethodological key argues that social science has no privileged standing to the plenum of everyday life, but only, if it cares, a disciplined interest in how this first province of meaningful action is assembled from within its situated occasions. The double move, however, amends the image in a very different way: Theory now stands both at the end and at the beginning of knowledge production. It is what comes last in understanding the production of normative order, and first in every accounting of any actual expression of it. Perhaps so, perhaps not. Arguments will not settle the matter. And studies of members' methodic practices need only that much of an opening to recommend their project.

References Billig, Michael (1999): Whose terms? Whose ordinariness? Rhetoric and ideology in conversation analysis. - Discourse and Society 10/4, 543-558. Button, Graham (ed.) (1991): Ethnomethodology and the human sciences. - Cambridge: Cambridge University Press. Chouliaraki, Lilie/Fairclough, Norman (1999): Discourse in late modernity. - Edinburgh: Edinburgh University Press. Coulter, Jeff (1983): Contingent and a priori structures in sequential analysis. - Human Studies 6/4, 361-376. de Certeau, Michel (1984): The practice of everyday life. - Berkeley, CA: University of California Press. Fairclough, Norman (1995): Critical discourse analysis. - London: Longman. Freebody, Peter/Luke, Allan/Gilbert, Pam (1991): Reading positions and practices in the classroom. - Curriculum Inquiry 21,435-457. Garfinkel, Harold (1967): Studies in ethnomethodology. - Englewood Cliffs, NJ: Prentice Hall. In his treatment of "Panoptic Theory", Serres speaks of a "hypocritical method" and how it consists of: always placing yourself behind the other, thus creating a line. [...] philosophies that base their theories on the human sciences try to occupy sites that will themselves escape all criticism; they try to find the outer limit of the chain or the end of the line. They are addicted to reasoning by extrema. (1999: 33-34)

212

Douglas Macbeth

Garfinkel, Harold (1988): Evidence for locally produced, naturally accountable phenomena of order*, logic, meaning, method, etc., in and as of the essentially unavoidable and irremediable haeceeity of immortal ordinary society. - Sociological Theory 6/1, 103-109. Garfinkel, Harold (2002): Ethnomethodology's program. Working out Durkheim's aphorism. Lanham, MD: Roman and Littlefield. Garfinkel, Harold/Sacks, Harvey (1970): On formal structures of practical actions. - In: John C. McKinney, Edward A. Tiryakian (eds.): Theoretical sociology, 337-365. New York: AppletonCrofts. Gee, James P. (1996): Social linguistics and literacies. Ideology in discourses (2nd ed.). Philadelphia: Palmer Press. Gutierrez, Kris/Larson, Joanne (1994): Language borders. Recitation as hegemonic discourse. Internationaljournal of Educational Reform 3/1, 22-36. Heritage, John (1984): Garfinkel and ethnomethodology. - Cambridge: Polity Press. Kavanaugh, Thomas M. (1989): The limits of theory. - Palo Alto, CA: Stanford University Press. Lemke, Jay L. (1995): Textual politics. Discourse and social dynamics. - London: Taylor and Francis. Livingston, Eric (1987): Making sense of ethnomethodology. - London: Routledge and Kegan Paul. Luke, Allan (1997): Text and discourse in education. An introduction to critical discourse analysis. Review of Research in Education 2/1, 3-48. Lynch, Michael (1993): Scientific practice and ordinary action. Ethnomethodology and social studies of science. - Cambridge: Cambridge University Press. Lynch, Michael (1999): Silence in context. Ethnomethodology and social theory. - Human Studies 22,211-233. Lynch, Michael (2000): The ethnomethodological foundations of conversation analysis. - Text 20/4, 517-532. Lynch, Michael/Bogen, David (1996): Spectacle of history. Speech, text and memory at the IranContra hearings. Durham, NC: Duke University Press. Macbeth, Douglas (2003): Learning lessons reconsidered. On the differences between the naturalistic and critical analysis of classroom discourse. American Educational Research Journal 40/1, 23 9280. Maynard, Douglas/dayman, Steven (1991): The diversity of ethnomethodology. - Annual Review of Sociology 17, 385-418. Payne, George/Hustler, David (1980): Teaching the class. The practical management of a cohort. British Journal of Sociology of Education 1/1, 49-66. Sacks, Harvey (1984a): Notes on methodology. - In: J. Maxwell Atkinson, John Heritage (eds.): Structures of social action, 21-27. Cambridge: Cambridge University Press. Sacks, Harvey (1984b): Doing being ordinary. — In: J. Maxwell Atkinson, John Heritage (eds.): Structures of social action, 413-429. Cambridge: Cambridge University Press. Sacks, Harvey (1992): Lectures on conversation, Vols. I-II. Gail Jefferson (Ed.). Oxford: Blackwell. Sacks, Harvey/Schegloff, Emanuel A./Jefferson, Gail (1974): A simplest systematics for the organisation of turn-taking for conversation. - Language 50, 696-735 Schegloff, Emanuel A. (1980): Preliminaries to preliminaries. 'Can I ask you a question?' Sociological Inquiry 50, 104-152. Schegloff, Emanuel A. (1998): Reply to Wetherell. - Discourse and Society 9, 413-416. Schegloff, Emanuel A. (1999): 'Schegloff s texts' as Billig's data. A critical reply. - Discourse and Society 10/4, 558-572. Serres, Michel (1989). Panoptic theory. — In: Thomas M. Kavanaugh (ed.): The limits of theory, 2547. Palo Alto, CA: Stanford University Press. van Dijk, Teun (1993): Principles of critical discourse analysis. Discourse and Society 4, 249-284. Wetherell, Margaret (1998): Positioning and interpretative repertoires. Conversation analyses and post-structuralism in dialogue. - Discourse and Society 9, 387-412.

Sequential Analysis in an Ethnomethodological Key

213

Wittgenstein, Ludwig (1960): The blue and brown books (2nd ed.). - New York: Harper Torchbooks.

Appendix

Notations Transcript notations are derived from the conventions developed by Sacks, et al. (1974). They aim to render the sequential production of interaction, as it sounds. Punctuations note pace and intonation rather than grammar. Pauses are noted in seconds, e.g., (2.0). Micro pauses are noted by (.). [underlining] shows emphasis or loudness. **

[asterisk] notes soft speaking. [hyphen] indicates the point where a word is cut off in its production.

:

[colon] indicates a sound stretch on a word or word portion, e.g., no::

II

[double slash] notes the point at which one speaker begins, overlapping another.

=

[equal sign] notes speaker transition without gap or overlap.

(()) Indicates scenic descriptions not available in the transcribed talk.

Christian Hudelot

The Use of a Functional Dialogic Model of Verbal Interaction to Compare how Daycare Workers and Teachers Scaffold 3-year-old Children

This model is called dialogical because it is largely inspired by work of Bakhtin (1981), and especially by the fact that wording is a responsive activity in which the speaker always anticipates the potential reply of the listener. It is called functional because, unlike those who emphasize the cohesive or structural organization of the dialogue following Francois (1995), we consider that dialogue is constituted by both factors which make it comprehensible, that is, features which add to (grammatical) cohesion, or intralinguistic, and factors which contribute to the interest of the dialogue. These relationships between continuity and change may differ from one interaction to another and manifest the specific organization of particular verbal interactions. As we consider dialogue analysis an empirical discipline, this model will be tested by contrasting how daycare workers and teachers scaffold a group of two- to three-year-olds when they are asked to describe a picture.

1. A Dialogical Approach to Dialogue

Aims of the Pie sent interaction

Psychosocial space Socio-culturai conditions

Figure 1. Dialogical Process

216

Christian Hudelot

From a dialogical point of view, a verbal interaction can not be reduced to a sequence of utterances in which the second utterance will simply complete the previous one according to pre-defined rules (cf. Hundsnurscher this volume; Varenne this volume). In fact, the first utterance is not itself independent of possible replies to it. As Volosinov (1973/1930) said, followed by authors such as Francois (1987) and Grize (1990), the utterance is produced for an interlocutor, or (to be more precise) for its comprehension or its possible replies: for what Bakhtin would call its evaluative reception. In addition, it is a fact that if we communicate something to someone in a certain form, we also communicate it with a more or less specific aim, in a specific social situation (i.e., with respect to sociocultural conditions of achievement [Kerbrat-Orecchioni 1987]). Given the above, we then assume that if the way one sees the interlocutor's beliefs and skills or capacities partly determines verbal behavior. Since school teachers and daycare center workers are trained differently, we should find some differences in the way they behave with groups of same-age children who are asked to describe a poster-sized picture. The aim of this study is to compare how groups of 2- to 3-year-old children are scaffolded by the adults who regularly care for them. The settings are daycare centers and preschools in France and the children are in a situation where they are asked to describe a picture. The data analyzed are taken from bi-monthly recordings of five groups composed of (four to five) children with their daycare worker, preschool teacher, or pediatric nurse all interacting verbally together. The children in the daycare center were boys and girls between the ages of 2:8 and 3:0 years (the average age was 2:10) and those of the preschool were between the ages of 3:2 and 3:5 years (the average age was 3:3). The same preschool teacher was in charge of the groups (labeled EMI, EM2 and EM3); two different daycare center workers worked with their groups (labeled CR1 and CR2). The groups were isolated in a room occasionally frequented by the children; and during each session, a picture that they had never seen previously was presented to them. The children sat in front of the 60 χ 40 cm picture and the adult was expected to make them talk about it. The data was recorded with a camcorder by Josse and Hudelot (see also Josse and Hudelot 1996).

2. Dialogue, Continuity, and Change

Unlike those who emphasize the cohesive or structural organization of the dialogue, following Francois (1995), we consider that dialogue is constituted on the subsequent basis: first, there are factors which make it comprehensible, for example, features of (grammati-

The Use of a Functional Dialogic Model

217

cal) cohesion, or intra-linguistic meaning; second, there are factors which contribute to the interest of the dialogue. This relationship of continuity and change reveals itself in two ways: it may be found within the discourse of a particular speaker, between the discourses of two different speakers, or, most usually, in both cases. In this chapter, I will deal with three broad areas of continuity and change: referential cohesion, discourse position, and degrees of dialogical move relevance. Referential cohesion constitutes the first area in which continuity and change may be observed. We can isolate some kinds of sequences (saynette to quote Fra^ois' expression) during which the speakers' utterances are linked to the picture. We can then distinguish different sequences depending on whether the speaker refers to an element of the picture, speaks about something related but not directly linked to the picture, or speaks about something without any relation to it. There are also sequences during which the speakers are themselves involved in managing the behavior or the discipline of the group. By interactional position we are not referring to the socially prescribed roles such as teacher and pupil, but rather the way in which these roles are enacted in particular circumstances (Franfois 1995; cf. Varenne this volume). For instance, in these dialogues, we observe an enormous difference between the contributions of the adults who address requests and produce assessments, and the children who mainly reply. And finally, we can consider the type of language games, or discourse genres, played by the scaffolding adult and the replying children. Regarding the topic relation or the most frequent adjacency pairs, the groups show a common profile. They all have spoken mainly about the picture and facts, have mentioned things not directly connected with it, or have been engaged in regulative behavior (see figure 2). But there are no significant differences between daycare center and preschool apart from the fact that no item is connected to the picture in the second daycare center group and that the adults of these groups seem to be more often involved in the regulation of the children's behavior (see figure 3). This seems to invalidate the hypothesis of a difference between the types of scaffolding of teachers and of daycare workers. In fact, these results lead to the conclusion that the task

EMI

EM2

EM3

CR1

CR2

Ξ about Picture · relating to P B managing S not relating n indeterminate

Figure 2. Relationship between the Discourse and the Picture

218

Christian Hudelot

AEM1

AEM2

ACR1

ACR2

C EM1

C EM2

CCR1

CCR2

Ξ Q Ouvertes · Q fermees Q Assertions Q App. Des. mi Ordres S Autres

Figure 3. Adults Act like Teachers and Children like Pupils (A = Adult; E = Children)

was within the discursive competence of the groups of children. Two groups (recorded in another daycare center by Mifiana 1995) were not able to complete the task successfully. Figure 3 also reveals the fact that the differences are not due to the status of the adult, but to the way in which the group behaved.

3. Scaffolding

In this situation, the picture is newly discovered by both the adults and the groups of children. Neither group had previously done the task with this specific picture. This is important when we consider the scaffolding process. Even though we admit that there is a wide range of ways to scaffold, we have to recognize at best a rough dichotomy: we can help a child to learn a strictly coded activity (e.g., to tie one's shoelaces) versus situations where the adult is able to help the child to do something he or she will not be able to do alone (e.g., relating a nightmare). Nevertheless, according to Callanan (1991) we assume that the scaffolding metaphor may be applied to the study of picture wording if broader strokes are used to define the level of scaffolding. What are the links between the children's interventions and those of their interlocutor? As it is common in situations of this type, the majority of the items are mainly induced by the adults. This means that the children's items not only follow up the adult discourse, but also are, so to speak, generated by the adults' interventions. On the other hand, the differences between induced and non-induced moves that are linked to the adults' discourse are significantly greater within the daycare center groups than in the preschool groups. Curiously, the statements of the children are more often auto-continuous in the daycare center groups than in the preschool groups. This is due to the fact, in particular, that these children tend to repeat themselves more often.

The Use of a Functional Dialogic Model

219

20 --

EM1

EM2

EM3

CR1

CR2

p about Picture · relating to P B managing B not relating · indeterminate

Figure 4. Relationship between Children's and Adult's Discourse

If we consider the discursive genre of the sequences (see figure 4) - that is to say the way content is manifested in lexical and grammatical coding - we may distinguish four broad genres: (l)The group can denominate or lexically categorize facts, characters and/or objects; (2) They may interpret the actions through their consequences or causes, or through the intentions or feelings of the characters (3) They may discuss and argue; and finally (4)They may carry out imaginary tests (e. g. guessing what would happen if someone were pulling the rope which is around the boy's wrists).

Figure 5 shows that the wording of the picture is mainly achieved by naming facts and objects or by producing a commentary on them. It can be seen in figure 6 that, although the

EMI

EM2

EM3

CR1

CR2

D about Picture · relating to P Qmanaging m not relating B indeterminate

Figure 5. Language Games during the "Wording" of the Picture

ShoIMP ShoSPO DenIMPL Den SPO DBS IIIPL Des SPO W IMP infSPO (Sho » sbowrotss; Den = deBOTamaäoa; Des - description: Icf = inference: IMPL »implicate, SPO = ipontanecus

Figure 6. Percentage of Adult Types of Utterances Compared to Children Types

220

Christian Hudelot

profile of each group is quite specific, the preschool and day daycare center groups differ from each other in two important ways. First, discussion and imaginary tests are found solely in the preschool groups. Second, the daycare center groups spend more than half of their time labeling the elements of the picture.

4. Adult Positioning

We certainly understand scaffolding differently from the way Bruner (1983) does. This question has been presented elsewhere (Hudelot 1997; Vasseur and Hudelot 1997) and scaffolding might be considered as the trace of the process of co-elaborating a linguistic formulation profitable to the speaker and benefiting to the ongoing interaction, whether or not this process is the result of a request, a prevention, or a repair. We propose considering three different aspects of the adults' interventions to isolate four positions in the adult discourse. First, we notice that, during the interactive task, the interventions of the adult may be linked mainly either to the management of the interaction, to the content of the task or to the children's linguistic formulation. Second, his/her intervention is connected with his/her own discourse or with that of the children. Finally, his/her intervention either implies a reply or an answer, or on the contrary, is induced or simply linked to the preceding dialogical move. Position 0. Although the adult intervention is linked to the task, it is not directly connected with the children's formulations. This is where the task is presented or a new topic is introduced. Position 1. The adult position is an anticipative or proactive one. This is evidently the place for interrogative utterances to occur, but it is not the only one. However, it should be noticed that the interventions identified as position 1 are not necessary in an absolute first or initiative position, since they can be linked to a previous intervention. PI (1)P1(1) Absolute PI position, can be addressee unspecified or be oriented to a particular child

AD: Bon alors j'ai resolu mon probleme moi! Tiens! Qu'est-ce que c'est que ca? AD: Okay, I've solved my problem. Hey, what's that? PI (2) self continuity (2) Insisting So: 'inaudible' rouge. AD: Quoi qui est rouge? (PI (3)) So: La. 'pointage'

The Use of a Functional Dialogic Model AD: C'est quoi? (PI (3)) Jo: 'pointage' Ca aussi c'est rouge. AD: Ca aussi c'est rouge. (P3 (1)) Mais c'est quoi qui est rouge comme ?a la? (PI (2)) So: AD: So: AD: J: AD:

(...)red. What's red? (PI (3)) Here, 'pointing' What is it? (PI (3)) 'pointing' That is red, too. That is red, too (P3 (1)). But what is red like this, here? (PI (2))

(3) Repeat the question to get the point of view of each child AD: Ju: AD: Va: AD: Am: AD:

Bein, je sais pas. Et toi tu dis quoi? ('Ju) C'est un loup. C'est un loup ... ('Va) Et toi, tu penses que c'est quoi ? Un loup. Un loup! ('Am) Amelia c'est un loup ou c'est un chien? Un chien. Un chien ... Alors deux chiens et deux loups. Salwa ... C'est un chien ou un loup ?

AD: Ju: AD: Va: AD: Am: AD:

Well, I don't know. And you, what would you say? ('Ju) It's a wolf. It's a Wolf... ('Va) And you, what do you think it is? A wolf. A wolf. ('Am) Amelia, is it a wolf or is it a dog? A dog. A dog ... So two dogs and two wolves. Salwa ... Is it a dog or a wolf?

PI (3) other continuity (4) Repeating the children's denomination while soliciting a new categorization To: AD: Or: AD: Ju: AD:

Ca c'est des maisons fa. C'est les maisons de qui? C'est un canard. Qu'est-ce qu'il fait? I va Fattraper kx! Qu'est -ce qu'i va lui faire? II va le manger? (P2 (1))

To: AD: Or: AD: Ju: AD:

They are houses. Whose house is it? Oh, that's a duck. What is he doing? It will catch it kx! He will catch it. What did you expect him to do to him? Will he eat it? (P2 (1))

(5) Eliciting formulation after a pointing movement or a non verbal reply. Gr:

C'est c'est, i veut attraper 93. 'pointage'

221

222

Christian Hudelot AD: C'est quoi 93? Gr: 'pointage' Et euh mai un ma- un maAD: C'est quoi 93? Gr: AD: Gr: AD:

It's it's, he wants to grab that, 'pointing' What is that? 'pointing' And uh ma- a ma- a maWhat is it?

(6) Eliciting justification Ma: AD: Ma: AD: Ma: AD:

L'est pas contente äPourquoi il n'est pas content? Mais la a pas d'sab'. Y'a pas de sable? (P3(2)) Non. C'est quoi alors?

Ma: AD: Ma: AD: Ma: AD:

He is not pleased atWhy is he not pleased? But there's no sand. There is no sand? (P3(2)) No. So, what is it?

Position 2. The adult speaks for the children, he/she pronounces the beginning of an answer, suggests his/her own wording and so forth.

P2(l) (7) answering in the place of the child Ju: I va Fattraper kx. AD: I va 1'attraper- (P3(l)) Qu'est -ce qu'i va lui faire? (Pl(3)) I va le manger? (P2 (1)) Ju: It will catch it kx! AD: It will catch- (P3(l)) What did you expect him to do? (Pl(3)) Will he eat it? (P2 (1)) (8) Maintaining the thematic cohesion but suggesting her owns formulations Gr: AD: Jo: AD:

C'est des feuilles! Ce sont des feuilles qui sont tombees de l'arbre? Oui. Peut-etre que la musique etait trop forte, hein?

Gr: AD: Jo: AD:

They are leaves. Have these leaves fallen from the tree? Yes. Maybe the music was too loud, wasn't it?

The Use of a Functional Dialogic Model

223

P2(2) (9) Semantic extension and implicit lexical correction or inferences AD: La: AD: La: AD: An: AD:

Alors eux est-ce que tu crois qu'ils ont peur eux? (P2 (1)) Non. Eh ben non (P3 (1)), parce que qu'est-ce qu'ils sont en train de fie eux? (PI (2)) Lire un livre. Ca les inquiete pas du tout le loup hein! II a tire la langue. II have, il have.

AD: La: AD: La: AD: An: AD:

So, do you think that those ones are frightened? (P2 (1)) No. Oh no (P3 (1)), because what are they doing, all of them? (PI (2)) Reading a book. That doesn't worry them at all, does it? He stuck out his tongue. He is slobbering, he is slobbering.

(10) Summing up the group discourse AD: ... Alors on en etait ou dans tout 53? ('groupe) La 'pointe sur 1'image' on avail dit que c'etait les oreilles du loup, lä c'etait les oreilles du p'tit garcon, 'point e vers 1'image' lä c'etait les griffes 'pointe' du p'tit gar9on ou du loup? Alors je sais plus moi! AD: ... So what was the point? ('group) Here 'pointing at the picture' we said that it was the wolfs ears, here were the little boy's ears, 'pointing at the picture' here were the nails 'pointing' of the little boy or the claws of the wolf? So I really don't know!

P2(3) (11) The adult is the mediator, the spokesperson of the children's discourse Ma: 'pointe' § mais, mais, mais... To: § II est oü bebe? AD: II est ou l'bebe? Demande Theo, il est oü 1'bebe? Ma: 'pointing' but, but, but... To: Where is the baby? AD: Where is the baby? Theo is asking where the baby is. Position 3. This retroactive position is the place where acknowledgement, negative or positive evaluations (Sinclair and Coulthard, 1992) take place, along with reformulations, formal corrections, repetitions marked with doubt, request for specification and so on. P3 retroaction (12) P3 (1) evaluating positively or negatively; acknowledgement (of receipt) AD: Ry: AD: Cy:

'pointe' Oui, comment il s'appelle? Le hibou. Le hibou, oui c'est le hibou Ryme. 'tapotant C'est le matin, on dort le matin.

la jambe de Ry'

224

Christian Hudelot AD: Peut-etre le matin. AD: Ry: AD: Cy: AD:

'pointing' Yes, what's this called? The owl. The owl, yes Ryme it's an owl. ' tapping Ryme's leg' It's in the morning, we sleep in the morning. Perhaps, in the morning.

(13) P3 (Γ) Correcting while proposing another lexical choice. Cy:

Les grandes oreilles du, il (g)riffe pas le canard ... moi j'ai fait des, j'ai vu des r, des girafes. AD: Des girafes? Cy: J'ai vu des girafes et des Dumbo. An: 'inaudible' AD: Ca s'appelle un elephant Dumbo. Cy: Moi, j'ai vu un Dumbo la tele". Cy: AD: Cy: An: AD: Cy:

The big ears of the, he isn't clawing the duck... me I did, I saw giraffes. Giraffes? I saw giraffes and Dumbos. (...) Dumbo is called an elephant. Me, I saw a Dumbo on TV.

(14) P3 (Γ) The adult is justifying her evaluations Cy: AD: Cy: AD:

La il (g)riffe le canard le loup. 'pointe' II a griffe le canard? Oui. Non regarde il Γ a pas encore attrape,

Cy: AD: Cy: AD:

Here the wolf is clawing the duck, 'pointing' Did he actually claw the duck? Yes. No, look, he hasn't caught it yet.

(15) P3 (2) Doubtful reply AD: On dirait quoi? Jo: Un oiseau. AD: Un oiseau? AD: What would you say it is? Jo: A bird. AD: A bird? Figures 7, 8, and 9, respectively, present adult positioning during the labeling sequences (Position 1), during the interpretative sequences (Position 2), and during the discussion

The Use of a Functional Dialogic Model

225

sequences (Position 3). While these results are not immediately apparent, when analyzed more qualitatively, the adult dialogic moves show several differences. First of all, in situations of denomination of the image elements, the preschool teacher more often carried out the answering movements instead of the children, or proposed more lexical coding than the daycare center workers did. They also tended to propose extensions which semantically enriched what the child had just said. On the other hand, the daycare workers more often had recourse, in P3 position, to repetition of the children's answers in a doubtful intonation. Within the interpretative episodes, the teacher showed the same tendency of proposing semantic extension in the P2 position. But there is also a difference with respect to the retroactive movements: the teacher is the only one to justify her own lexical coding. Within

EM1 •A induced Dsetf connect

EM2 a A linked Qmonologal

EM3

CR1

HC induced Qlnittation

CR2

EC linked Ξ unimerpf stable

Figure 7. Adult Positions during the Labeling Sequences (A = Adult; C = Child)

EMI

EM2

EM3

CR1

CR2

EDPO ·Ρ1 ΠΡ2 SP3 BMIS BIND

Figure 8. Position during the Interpretive Sequences 60 50 40 30 20 10

I EM1

EM2

EM3

CR1

QPO ·Ρ1 8Ρ2 ΠΡ3 EIMIS BiNO

Figure 9. Adult Positions during the Discussion Sequences

CR2

226

Christian Hudelot

the discussion episodes (which exist only in the preschool groups), one can note a large number of miscellaneous items and (especially) movements of focusing the children's attention on the elements of the image, or requests made to the children to ask questions to their classmates.

4. Conclusion

Finally, these results support the assumption that when they deal with a similar task in a group of preschool children, teachers and daycare center workers act differently. In the daycare center, the adults expected that the children would express themselves individually. On the contrary, at preschool, the adult tried to help the children, acting as a group, to produce a discourse and to discover some relationships that they would not have been able to produce or to discover if they had not been speaking to each other. A frequently heard saying sums this up in a clear-cut way: the daycare center workers say that they are interested in the singularity of each child, and the preschool teachers are dealing more with pupils as a group to whom they must teach something.

References Bakhtin, Mikhail M. (1981): The dialogic imagination. Four essays. - Austin: University of Texas Press. Bruner, Jerome S. (1983): Le developpement de l'enfant. Savoir faire, savoir dire. - Paris: PUF. Callanan, Maureen A. (1991): Parent-child collaboration in young children's understanding of category hierarchies. - In: Susan A. Gelman, James P. Byrnes (eds.): Perspectives on language and thought. Interrelations in development, 440-484. Cambridge: Cambridge University Press. Fran9ois, Frederic (1987): Morphologie, syntaxe et discours. - In: Jean-Α. Rondal and Jean-Pierre Thibaut (eds.): Problemes de psycholinguistique, 175-247. Bruxelles: Pierre Mardaga. Fran9ois, Frediric (1995): Organising principles, discourse movement and atmosphere in dialogue. A clinical interview between a psychiatrist and a patient. - International Journal of Psycholinguistics 11/2, 205-225. Fran9ois, Frederic/Hudelot, Christian/Sabeau-Jouannet, Emilie (1984): Conduites linguistiques chez le jeune enfant. — Paris: PUF. Grize, Jean-Blaise (1990): Logique et langage. - Paris: Ophrys. Grossen, Michele/Salazar Orvig, Anne (eds.) (1998): Clinical interviews as verbal interactions. A multidisciplinary outlook. - Pragmatics 8/2,

The Use of a Functional Dialogic Model

221

Hudelot, Christian (1993): Quand 1'adulte aide l'enfant a parier. Quelques aspects de l'etayage langagier. Dialoganalyse IV. referate der 4. Arbeistagung Basel 1992 Teil 2, 215-222. Hudelot, Christian (1997): Modalites d'interventions de 1'adulte dans la gestion d'un petit groupe d'enfants de moyenne section de maternelle en situation de description d'image. - Cahiers ^'Acquisition et de Pathologie du Langage 14, 123-155. Hudelot, Christian (1999): Etayage langagier de l'enseignant dans le dialogue maitre-eleve. - In: Michel Gilly, Jean-Paul Roux, Alain Trognon (eds.): Apprendre dans l'interaction, 219-240. Nancy: PUN. Hudelot, Christian/Josse, Denise (1996): Gestes indicatifs dans une situation de lecture d'image. Petits groupes d'enfants de 2-3 ans avec un adulte familier. - Actes du Colloque Modes d'interaction et de communication chez le jeune enfant, Florence. Kerbrat-Orecchioni, Catherine (1987): La mise en place. - In: Jacques Cosnier, Catherine KerbratOrecchioni (eds.): Decrire la conversation, 359-352. Paris: Editions du CNRS. Minana, C. (1998): Groupes d'enfants de 2-3 ans en situation de lecture d'image. Comparaison entre institutrice et auxiliaire de puericulture. - Memoire pour le Certificat de Capacite d'Orthophoniste sous la direction de Christian Hudelot. Universite Paris 6. Moeschler, Jacques/Reboul, Anne (1994): Dictionnaire encyclopedique de pragmatique. - Paris: Seuil. Salazar Orvig, Anne/Hudelot, Christian (1989): Enchainements, continuites et deplacements dialogiques chez le jeune enfant. - Verbum 12/1, 99-115. Sinclair, John/Coultard, Malcolm (1992): Toward analysis of discourse. - In: Malcolm Coulthard (ed.): Advances in spoken discourse analysis, 1-34. London: Routledge. Vasseur, Marie-Therese/Hudelot, Christian (1997): Peut-on se passer de la notion d'etayage pour rendre compte de I'elaboration langagiere en LI et en L2. Cahiers d'Acquisition et de Pathologie du Langage 15, 109-135. Veneziano, Edy (2000): Interaction, conversation et acquisition du langage dans les trois premieres annees. - In: Michele Kail, Michel Fayol (eds.): L'acquisition du langage. Le langage en emergence. De la naissance ä trois ans, 231-265. Paris: PUF. Volo§inov, Valentin N. (1973/1930): Marxism and the philosophy of language. Trans, by Ladislav Matejka, I. R. Titunik. - Cambridge, MA/London: Harvard University Press. Wood, David J./Bruner, Jerome S./Ross, Gail (1976): The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry 17, 89-100.

Mireille Froment

Teacher and Preschool-Level Pupils' Dialogue around a Cartesian Table: Categorizations, Accentuations and Emergence of Routines.

The teaching-learning situation analyzed in this chapter is an initiation to the classification of elements in a Cartesian table among a group of preschool level children, aged 3-4. The didactic format creates a context and prepares a practice space. The pupils are encultured to a routine which represents how to do, speak, and think, and which teaches intellectual scaffolding and knowledge building. The analysis shows how the teacher scaffolds learning for the children. She focuses their attention on the various objects she brings, and demonstrates how categorizations, thematic movements, and routines work together to build common referents and elaborate new knowledge. Moreover, the analysis highlights the gap between the teachers's plans and the concrete dialogue, and between the teacher's aims and the pupils' instantaneous perceptions.

1. Introduction

The analysis of a scholastic dialogue leads to specific problems related both to its existence in an institutional environment and its developmental purpose. School is a socially preexisting context. The roles (i.e., teacher/pupil) are set before any intervention of the actors. The teacher/pupil relationship is asymetric. As Vygotsky (1998/1933) points out, this asymmetry is a condition for development, but this condition is not sufficient. Indeed its effects can be either positive or negative. Within the guidance of the didactic dialogue, it can be observed particularly in the gap between teacher's long term aims and pupils' instantaneous perceptions. The management of coherence (i.e., theme and dialogue movements) by the teacher contrasts with pupils's open associations. The didactic dialogue is structured to build knowledge. School transmits patterns, or ways of thinking, through genres called scholastic genres. A part of its objective is to enculturate pupils into patterns and ways of thinking that can be considered, in a Vygotskian perspective, as mental tools. School also has the feature of heterogeneous learning practices. Bruner (1983) classifies them into 1) learning through imitation (the relationship of doing as a teacher or another

230

Mireille Froment

pupil is important in the school context); or 2) learning through verbal transmission or personal activity. These different learning practices are not exclusive of each other, but can complete each other or be dominant. The teaching-learning situation analyzed here is an initiation to the classification of elements in a Cartesian table (Duval 2003) between a group of preschool-level children, aged 3-4. The teacher first uses elements from a doll's tea party (i.e., glasses, cups, plates, flatware), then colored shapes. In each case, the elements have to be classified according to the two criteria of shape and color. The purpose of the analysis is to point out how the situation makes sense for every participant in the dialogue, how the pupils' answers make sense for the teacher, how the different points of view are structured into the didactic dialogue, and how the points of view are modified according to the pedagogical aim. The interpretation is a constant aspect of the dialogue (Froment and Carcassonne 2002) and takes a particular importance in the didactic dialogue.

2. The Teacher Organizes a Favorable Situation for Pupils to Build and Acquire Knowledge

The teacher interacts with pupils through the use of language and various manipulatives. The didactic format which is used to build a teaching/learning situation is a way of regulating and organizing interactions, but also a practical interpretation of their content. The patterns lead to a way of making sense of the situation. The didactic format is a medium between the participants and the common referent. It has effects on the discourse and on the subject. The formatted teaching/learning setting entails four pedagogic sessions. The first one has not been recorded. Those four sessions form a sequence (i.e., an organized series of sessions); each of them has the same purpose: to construct a concept and to classify objects in a Cartesian table, but manipulatives (e.g., elements of doll's tea party, ropes and squared paper) change from one session to another. In the first and second sessions, elements of a doll's tea party (i.e., cups ands plates) are used in a context which modifies their meaning. They work as elements of a class to be identified, based on their shape and color. In the third and fourth sessions, cups and plates are represented by colored paper labels. Modifications also concern the representation of the Cartesian table. In the first session, ropes represented the boxes of the table. In the second and the third sessions, a square piece of paper is used, and in the fourth, the teacher draws the table on the blackboard.

Dialogue around a Cartesian Table

231

During the first three sessions, pupils sit on the carpet on which various objects have been displayed; for the fourth session, they have to look at the blackboard. During the whole sequence, the teacher guides the interaction. She says who can try to classify the objects in the table and she guides the exchanges (i.e., she attributes roles to pupils). The pupil who classifies one object in the table has to act (by choosing a square) and to encode the action through words. The other pupils have to discuss and evaluate the action. They can also suggest another way of classification. The didactic format runs its course; it relies upon a developmental structure which can be labeled "contextualization, decontextualization and recontextualization". Each session of the sequence is in a repeat-and-restructure relationship with the former one. Concretely, the time of restructuring can be observed in the very first part of a session (first called saynete), which the teacher begins and regulates. It is an important time for intersubjectivity to be built. The referential objects (i.e., elements of a doll's tea party) are only encoded as colors and shapes. Those familiar objects are pointed out from an unusual point of view, which has to be constructed in order to be shared by the participants. In the second session, after the teacher has reviewed the first one, she points out the modifications she made to the equipment: 29-teacher-12: 30-Karim-7: 31-teacher-13: 32-Lucie-2: 33-teacher-14: 34-Sabine-2: 35-teacher-15:

36-pupils together: 37-teacher-16:

38-pupils: 39-pupils: 40-pupils: 41-pupils: 42-pupils:

hum [she sets down a squared paper on the carpet] like squaring! like squaring, you are right. Look! and objects too [she lays a basket with elements of doll's tea party on the carpet] elements of a doll's tea party! which come from a doll's tea party, exactly. Then, have I forgotten something? and the rope! oh no, we don't need the rope today, I set down the table here on the floor [noise] look I brought that [the basket with elements of a doll's tea party], some plates, some some glasses we will say cups because we could mix up color and object [the French word for the color green and for the object "glass" is the same], well, then, we will say "cups" since there are little handles, here [she points each cup handle] a cup a cup a cup a plate a plate

The pupils enumerate the first session's objects, present or not, (34-Sabine-2); the teacher rapidly indicates (35-teacher-15) that the table takes the place of the ropes. She recategorizes squaring as table. The elements of a doll's tea party offer less diversity than during

232

Mireille Froment

the first session. She keeps two classes of objects: "cups and plates". She makes the pupils enumerate the elements of the doll's tea party like in the first session. The turns the pupils take correspond to the enrollment order, and serves to focus their attention on the objects they will speak of. In the third session, the presentation of the equipment takes a longer time. (1 to 66) 1-teacher-1: 2-pupils together: 3-teacher-2: 4-pupils together: 5-Sabine-l: 6-teacher-3: 7-Karim 1: 8-teacher-4: 9-Lea-l: 10-teacher-5: ll-Karim-2: 12-teacher-6:

13-Brice-l: 14-teacher-7: 15-Assia-l: 16-Lucie-l: 17-teacher-8: 18-pupils together: 19-teacher-9: 20-pupils together: 21-teacher-10: 22-Sabine-2: 23-teacher-ll: 24-Sabine-3: 25-teacher-12: 26-Sabine-4: 27-teacher-13: 28-Sabine-5: 29-Lucie-2: 30-teacher-14:

Well, what have I put on the floor? a-H-carpet a carpet!-f~H- a carpet with flowers? no :::: with little squares with little squares, well it's a ::: A SQUARED paper! Well, you are right, it's called with blue and green lines a squared paper. You see, with many little squares. Hush hush:::Mira ; Mira be quiet we have to fill it yes :::we will fill it, exactly. We will not fill it anyhow. This time I am bringing neither cups nor plates, nor flatware. I am bringing [she shows the objects] little papers little papers! A CIRcle it's yellow [she puts the papers on the carpet] a circle, and it's yellow, that's right [noise] hush :::I'll put it anywhere, just like that, OK, so here, it's a circle blue :::a blue circle oh ::::Tomy stop it please ! blue :::blue :::blue::: [she puts a paper on the carpet] hush ::: a blue circle [she shows another circle] green::: it's a green circle let's put it here ah :::! put it anywhere-H-H-for the time being oh no! [several pupils cough] wait, we only look at everything in the basket there's yellow, there's blue, there's green there's red bravo, you win, you win, there's a red circle!

The square piece of paper is categorized as carpet, then as squared paper by Karim (as in the second session). Karim (ll-Karim-2) says what the task is; he anticipates the teacher's discourse and speaks as if he were the teacher. The reiteration of the didactic format allows him to anticipate what will happen and what will be said. A background is built, made of knowledge and habits. The presentation of the colored paper shapes takes time. The

Dialogue around a Cartesian Table

233

teacher makes the task into a game by asking: What have I got in my basketf It's an opportunity for pupils to name and describe the colored paper shapes, one by one. The task supposes that pupils can recognize shapes and colors. In the fourth session, the presentation is shorter: 1-teacher-:

2-Brice-l: 3-Karim-l: 4-Alexandre-1: 5-pupils together: 6-teacher-2: 7-pupils together: 8-teacher-3: 9-Assafoune-l: ΙΟ-pupils together: ll-teacher-4: 12-Assafoune-2: 13-pupils together: 14-teacher-5: 15-Alexandre-2: 16-teacher-6:

hush :::so, pay attention a little bit. Oh ::Tomy it's not OK, sit down correctly, if you have no room enough on the bench, sit on the carpet. BUT I don't want you to turn your back on me. Oh yes, and besides there is a place here, let us move a little, move a little Nasim, move a little Aminedine, move over Diola, move over Alexandre, move near me, to make more room for him, Tomy, that's good. Look at what I::: am drawing here some some some squares colors, up colors squares, colors so, there are colors, that's right, what colors? blue :::red :::green::: that's a shape, which one? a square a square and that? a circle a circle and how is called my drawing here? a squaring it can be called a squaring BUT we will use it to do something. In fact, it's called a table + a Cartesian table, since we can enter by two places. We can enter here "colors" like this, we can enter there, we can enter this way, "shapes", this way and that way. What do you think of, look [she shows a circle]

The teacher proceeds differently. She calls upon the children to speak so that they describe shapes and colors. She uses the lexemes shapes, colors, and Cartesian table to categorize the different objects. Each session presents a new stage which is re-interpreted through the previous one. Generalization and meta-use of the sessions are part of every didactic sequence (Nonnon 1990), of its functionality level and of its inscription in a longer temporality. Bruner's conclusions (1983) about formatted sessions can be recalled: an organized and repeated structure offers the possibility to be creative by introducing variations. During the sessions, the specific interaction type, presented previously in the didactic formatted setting, builds a context supposed to be favorable for the acquisition of knowledge by the pupils. But, between the teacher's plans - what she calls upon the children to do and the concrete dialogue - there is a gap as obvious as it is inevitable. The concrete exchanges of the dialogue constitute the only observable data on 1) what is asked; 2) the diversity of the answers; 3)

234

Mireille Froment

the variability in the interpretations; and 4) the singularity of associations. We analyze this movement from a dynamic point of view: sometimes what is called for by the teacher is a context in which children's answers make sense (i.e., sometimes children's answers lead to a change in the context).

3. Teacher and Pupils' Dialogue. Categorizations, Accentuations and Emergence of Routines.

In each analyzed session, the teacher asks the pupils to do something, makes them do it, and makes them recall the previous session. The recollection is, in fact, a recontextualization of the last session's actions and utterances according to what the teacher deems useful for building knowledge (i.e., how to act and how to speak). The recall is intended to make present something no longer present through words; it makes semiotics enter into dialogue (i.e., doing something through verbalization only becomes verbalizing). From a cognitive point of view, pupils move from doing to knowing how to talk about what has been done. At this stage, the teacher's scaffolding concerns language and cognition. It relies on lexical recategorizations, thematisation movements, and discursive movements around words: 1-teacher-1: 2-Karim-1: 3-Lucie-l: 4Brice-l: 5-teacher-2: 6-Karim-2: 7-teacher-3: 8-pupiIs together: 9-teacher-4: ΙΟ-pupils together: ll-teacher-5: 12-Sabine-l: 13-Brice-2: 14-Karim-2: 15-teacher-6: 16-pupils together: 17-Mira-l: 18-Karim-4: 19-teacher-7: 20-Mira-2:

so, Alexander wasn't at school yesterday, we have to explain to him what we did. Do you remember what we did yesterday? we put we put papers with colors :: then we put plates, cups we put forks and spoons ropes too ! yes ::: did we put them anywhere on the carpet? first we put plates, we put glasses [inaudible] and ropes and did we put plates, cups, flatware, ropes anywhere on the carpet? no ::: we put everything in the middle? no ::: how did we put them? we first put plates and here we put the ropes! the ropes and then we put them straight [horizontally] the ropes we put them straight and we mixed plates, cups and flatware? no ::: we put them § we put them § hush, Karim let Mira speak! we put together glasses with glasses, flatware with flatware and plates with plates and cups with cups

Dialogue around a Cartesian Table 21 -teacher-8: 22-pupils together: 23-teacher-9: 24-Karim-5: 25-teacher-10: 26-pupils together: 27-teacher-11: 28-Karim-6:

29-teacher-12:

235

yes and did we mix all cups together? no ::: and what else did we do? BUT there were some others in other places, and cups in another place we put cups on one side, plates together, flatware together. BUT we mixed all plates together? no::: how did we separate flatware, plates and cups? first we put all the plates, then at the back we put glasses and cups-H-and glasses++and then we++\ve put them straight and then+++we arranged-H-we put them so [he shows with his finger] hum [she lays on the carpet a large squared paper]

The second part of the session is particularly interesting because of the teacher's discursive movements around predicates: put, put together, separate, either used alone or as a repetition of a pupil's response. Thus, the part of the session shown here is entirely devoted to the verbal recall of the previous session. It ends when the teacher takes note that the children do not identify the second classification criterion: color. At that moment, she makes a discursive move: from her asking pupils to say, to her asking them to do and say (i.e., she asks the children to classify the objects). The teacher's question (1-teacher-l) is an open question. Karim, Lucie, and Brice answer with a factual statement; the predicate is put. Children discuss their actions. The predicate is specified by the object (put glasses; put ropes). They enumerate their actions, according to a temporal succession. The predicate put plays the role of an attractive word in the dialogue. A word which is repeated and which leads to discursive movements is said to be an attractive word. A word becomes attractive in a certain situation and gains a specific density. The teacher makes a specific discursive move around the predicate put; she asks (5-teacher-2) Did we put them anywhere on the carpet"? She repeats the predicate introduced by the children, but adds the category in a way that her point of view can be interpreted. The meaning of a particular act is modified when integrated in another operational level, either more general (e.g., put changed into a class), or integrated into another meaning plane; that is the case here. From 7 to 28, the relationship between the teacher's questions and the unchanging point of view in the children's answers is nearly a misunderstanding. The teacher modifies her strategy she introduces a new categorization by choosing new predicates, mix (15-teacher-6), then separate (27-teacher-11). Thus, we observe that lexical recategorizations, thematic movements, and discursive movements around words are the teacher's main strategies: (l)Her question first introduces a move around the children's predicate; (2) She introduces a new predicate in her question (15-teacher-6); (3)She repeats the predicate (15-teacher-6), but only for cups (21-teacher-8). For the first time, from cups, plates, glasses, flatware viewed as a whole, she selects cups and builds a subclass (Grize 1999);

236

Mireille Froment

(4)She introduces a new predicate (27-teacher-ll): she repeats cups, plates, glasses, flatware viewed as a whole and chooses the predicate separate.

Children's answers do not follow the teacher's guidance; they do not produce the second classification criterion (i.e., the intraclass criterion: color). We can take the measure of the difference between teacher's proposal and the concrete dialogue. We can say that she is out of the children's proximal development zone (Vygotsky 1998/1933). She modifies the way she offers scaffolding for the children. She asks the children to act and to speak when acting (i.e., to put words on their acts). The move between speakingthinking of something absent (the recall of the last session) and speaking-thinking of the present (the here and now) is important from a didactic point of view. How to say something and how to do something are elaborated through one session's didactic dialogue and repeated in the next one. This is how routines are built.

3.1. Emergence of routines Routines (i.e., how to do things, how to say things, ways of thinking) are repeated from one session to another; altogether, they become stable and run their course. In our corpus, the teacher works at implanting routines to help pupils perform the task, or to turn them into mental tools (Vygotsky 1998/1933). As soon as the second session begins, under the teacher's intervention, a routine emerges. It combines action and words (i.e., a pupil puts an object in a square and comments on this action). 43-teacher-17: 44-pupils together: 45-teacher-18: 46-Yohann-1: 47-teacher-19: 48-Sabine-3: 49-teacher-20: 50-pupils together:

Who knows how to organize them properly? I, I, I::: Yohann, do you know? Hullo! try to put the cups, go on0! Well one cup has to be put on one square, OK? OK [Yohann puts a cup on a line] go on, I let you do0 and the others will pay attention [Yohann puts every cups on the squared paper] there are not many cups! that's good, he puts all the cups, so, are they all on the same line? yes:::

The teacher gives pupils roles and discursive places; she indicates with precision what discursive genre is called for, for example, an appraising comment (49-teacher-20). The teacher's purpose is to build a class composed of those sub-classes: plates and cups. A criterion has to be elaborated. The teacher takes the initiative:

Dialogue around a Cartesian Table 61-teacher-26: 62-pupils together: 63-Karim-8: 64-teacher-27: 65-Karim-9: 66-teacher-28: 67-Sabine-4: 68-teacher-29: 69-Yohann-2: 70-teacher-30: 71-Omar-l: 72-teacher-31: 73-Lucie-3: 74-teacher-32: 75-pupils together: 76-teacher-33: 79-Omar: 80-teacher-35:

237

so, yesterday we mixed all the cups like that? no ::yes :: no :: no ::: no, hey, how did we do? we organized them we organized them we put them here [she shows the cups] what is the difference between these cups? because we put them what is different between this cup, hush, between this one and that one, what is different? they are large they all have the same size! what is the difference? they do not have the same color! you are right Lucie, they have not the same color! this one is blue, and that one is yellow:: and yesterday, we mixed all colored cups together? we did not stick the glasses! we separate the colored cups, so, who could put them by color?

The teacher introduces difference as a theme to build the subclass color inside the class cups. But pupils work in the dark; there is a misunderstanding between the teacher and the pupils. Pupils repeat what they said before or describe what they see. The teacher poses a new question (76-teacher-33). Here there is a decomposition move (i.e., she separates the stages of the comparison) and a recomposition move (i.e., she comes back to the class cups). With 80-teacher-35, the meaning of properly becomes explicit: we separate the colored cups. In the third session, the second session's emergent routine is repeated; it becomes stable and complex when introducing the second criterion explicitly (color): 87-teacher-37: 88-Sabine-ll: 89-teacher-38: 90-Sabine-12: 91-teacher-39: 92-Karim-6: 93-pupils together: 94-teacher-40: 95-Karim-7: 96-teacher-41: 97-Karim-8: 98-teacher-42: 99-Karin-9:

so, what has she done here? I, I agree you agree Sabine, why do you agree? she puts squares with squares and circles with circles yes, that's good, circles and squares are well separated here I, I don't agree! me neither:: who is able to explain to me, before moving paper shapes, why he or she disagrees?-H-sit down-Η-1 ask, who is able to explain to me, you Karim? eh++ eh++ why? go on because++ because a red has to be put on the red square, a circle has to be put on the yellow square, and the red ones on the red square ah, you believe that the circles have to be put on the square of the same color? yes, upon it!

238

Mireille Froment

100-pupils together: no :: :no ::: 101-teacher-43: so we are not allowed to put them together, we are not allowed. Well Karim, we take note of your proposal, that's good because circles are together and squares are together. BUT are the blue things together? 102-pupils together: no :::

Pupils take the appraising discursive place and play with the routine: (l)They begin to play repeating I agree, I disagree (93) (2) They begin to know how to speak and to justify : Sabine (90) justifies what she says by describing; Karim (79) justifies what he says, with the color argument but from his point of view, (3)The chidren begin to take initiatives

The teacher (101-teacher-43) presents a short synthesis - it might be what she is expecting from the pupils. In that context, almost at the end of the session, she starts a new routine: verification. She tells the pupils how to name what they are doing when reading a Cartesian table; she introduces two lexemes: line and column: 144-teacher-63:

I'll mix and dispose otherwise, so, a square here, a triangle there, a circle there, the blue one here, the red one and the yellow one. AND now I'll give to each of you a little form that has to be put in the right square. I put the first one like that, you'll understand, it's 145-pupils together: a green circle::: 146-teacher-64: so I put it with the circles, so either here or there, where circles can be seen. BUT what is that circle's color? 147-pupils together: green::: 148-teacher-65: so, in the green line, so pay attention, the circle column and the green line 149-pupilstogether: green:::

During the fourth session, the teacher manages for the children to repeat and link the two routines. She teaches the pupils how to use their minds (Bruner 1996). But if the children repeat the color and are able to say where the shape has to be placed, they do not repeat the routine "line and column".

4. Conclusion

In the course of a didactic dialogue, pupils are called on to take a role and a discursive place for a given content. The didactic format creates a context and prepares a place for

Dialogue around a Cartesian Table

239

learning. The pupil is encultured to a routine which represents how to do and how to speak and think, which teaches intellectual procedure and a way to build knowledge. The analysis shows 1) how the teacher scaffolds the children's learning; 2) how she draws their attention to the various objects she brings; and 3) how categorizations, thematic movements and routines work together to build the common referents and elaborate new knowledge. Yet, the analysis also identifies the gap between the teacher's intent and the concrete dialogue, between the teacher's aims and the way the didactic situation makes sense for the children. On the teacher's side, there are intentions; on the pupils' side, there are vast possibilities of interpretation, so that a didactic session contains a large part of improvising. The analysis is led a posteriori from the researcher's point of view at a macro-level (the saynete progression) and at a micro level (the development of the exchanges). Thus, we can observe the teacher's strategies and how she reacts to the children's discourse. Further, we can distinguish her recategorizations, her thematic movements, and her discursive movements around attractive words. On the children's side, they can only have initiatives at the micro level. We observe how they resist taking note of the second criterion and we observe how they start to play with routines and how they work in the dark. All this constitutes examples of the gap between intent and concrete dialogue. But we observe how children start to justify what they say - to discuss their actions and to put words on actions, too.

References Bruner, Jerome S. (1983): Le developpement de l'enfant. Savoir faire, savoir dire. - Paris: PUF. Bruner, Jerome S. (1996): L'education, entree dans la culture. - Paris: Retz. Duval, Raymond (2003): Comment analyser le fonctionnement representationnel des tableaux et leur. diversite ? Spirales 32, 7-31. Francois, Frederic (1994): Pratiques de oral. Dialogue, jeux et variations des figures du sens. Paris: Nathan. Froment, Mireille/Carcassonne, Marie (2002): Genres et mouvements discursifs. Un eclairage sur la reflexivite des oraux ä l'ecole primaire. - In: Jean-Charles Chabanne, Dominique Bucheton (ed.): Parier et ecrire pour penser, apprendre et se construire, 151-168. Paris: PUF. Grize, Jean-Blaise (1990): Logique et langage. - Paris: Ophrys. Nonnon, Elisabeth (1990): Est-ce qu'on apprend en discutant? Un exemple d'interaction maitreeleves en section d'education specialisee. - In: Fra^ois Frederic (ed.): La communication inegale, 147-212. Neuchätel-Paris: Delachaux etNiestle. Vygotsky, Lev S. (1998/1933): Theorie des emotions. Etude historico-psychologique. - Paris: L'Harmattan.