Analysing the Language of Discourse Communities
Related Elsevier books CLARK War Words: Language, History and the Disciplining of English FIRTH (ed.) The Discourse of Negotiation: Studies of Language in the Workplace MEY (ed.) Concise Encyclopedia of Pragmatics
SPOLSKY (ed.) Concise Encyclopedia of Educational Linguistics
Related Elsevier journals Journal of Pragmatics Editor: Jacob Mey Language and Communication Editors: Roy Harris and Talbot J. Taylor Language Sciences Editor: Nigel Love Lingua Editors: Johan Rooryck, Diane Blakemore and Neil Smith Linguistics and Education Editors: David Bloome and Jay Lemke
Free specimen copies available on request
Analysing the Language of Discourse Communities Joan Cutting University of Sunderland
2000 ELSEVIER Amsterdam London New York Oxford Paris Shannon Tokyo
ELSEVIER SCIENCE Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB, UK
© 2000 Elsevier Science Ltd. All rights reserved. This work is protected under copyright by Elsevier Science, and the following terms and conditions apply to its use: Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of docu ment delivery. Special rates are available for educational institutions that wish to make photocopies for nonprofit educational classroom use. Permissions may be sought directly from Elsevier Science Global Rights Department, PO Box 800, Oxford OX5 1DX, UK; phone: (+44) 1865 843830, fax: (+44) 1865 853333, email:
[email protected]. You may also contact Global Rights directly through Elsevicr's home page (http://www.elsevier.nl), by selecting 'Obtaining Permissions'. In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (978) 7508400, fax: (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London Wl P OLP, UK; phone: (+44) 207 631 5555; fax: (+44) 207 631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works Tables of contents may be reproduced for internal circulation, but permission of Elsevier Science is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mecha nical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier Global Rights Department, at the mail, fax and email addresses noted above. Notice No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made.
First edition 2000
Library of Congress Cataloging in Publication Data A catalog record from the Library of Congress has been applied for. British Library Cataloguing in Publication Data A catalogue record from the British Library has been applied for.
ISBN: 0080438938 ©The paper used in this publication meets the requirements of ANSI/NISO Z39.481992 (Permanence of Paper). Printed in The Netherlands.
CONTENTS
Acknowledgements
ix
1 1.1 1.2
Theory And Methodology Introduction General Approach To Discourse Analysis Meaning Of Context Linguistic Analysis Functional Analysis Conversational Organisation Analysis Studies Of InGroup Markers The Study Of The MSc InGroup Language Model And Method Of Analysis Map Of The Book Conclusion
1 1 3 3 5 6 8 9 11 11 14 15
The In-Group Psychology Literature On Relationships The MSc Student Group InGroup Markers Informality Humour Topic Shifts The Recordees Conclusion
17 18 19 22 22 24 27 30 31
Knowledge Areas Introduction The Literature On Knowledge And Context The Categorisation Of Knowledge Changes In Knowledge The Common K Areas Shared Interpersonal Knowledge Further Dimensions Conclusion
33 33 33 36 44 44 45 47 49
Grammar Of The In-Group Code Introduction Overview Grammatical Reference Categories Analysis Of Grammatical Reference
51 51 51 53 60
1.3 1.4
1.5 2 2.1 2.2 2.3
2.4 2.5 3 3.1
3.2
3.3 3.4 4 4.1
vi
4.2
Changes In Grammatical Reference NonAnaphoric Definite Reference Degrees Of Explicitness Shared Interpersonal Knowledge Referents Context Specificity Further Dimensions Conclusion
62 62 63 67 68 68 71 74 75
Lexis Of The In-Group Code Introduction Overview Lexical Reference Categories Analysis Of Lexical Reference Changes In Lexical Reference Nouns Verbs
77 77 77 79 83 84 84 91
5.3 5.4
Further Dimensions Conclusion
94 95
6 6.1 6.2
Implicitness Over Utterances Introduction Clausal Ellipsis The Categories Changes In Clausal Ellipsis Conversational Implicature The Categories Changes In Humour Topic Shifts And Hitches Changes In Topic Shifts Changes In Hitches Further Dimensions Conclusion
97 97 98 98 99 102 102 103 107 107 110 115 117
Function Introduction MacroFunctions The Categories Changes In MacroFunction Speech Acts The Categories Changes In Speech Acts Function Of Implicit Language MacroFunction Speech Acts Further Dimensions
119 119 120 120 122 124 124 128 134 134 135 137
4.3
4.4 4.5 5 5.1
5.2
6.3
6.4
6.5 6.6 7 7.1 7.2
7.3
7.4
7.5
vii
7.6
Conclusion
139
8 8.1
Further Study Summary Implicit Language And InGroup Membership Changes In Form And Function Function Of Implicit Language Recommendations For Further Research Expansion Of The Model Generalisability Of The Model Social Dimensions Of The Model Applications Teaching English As A Foreign Language Clinical Pragmatics Forensic Linguistics Conclusion
141 141 141 143 145 146 146 147 150 152 152 155 156 157
8.2
8.3
8.4
Appendix I Appendix II Appendix III
159 163 165
References
167
Author Index
175
Subject Index
179
This page intentionally left blank
ACKNOWLEDGEMENTS
I should like to thank Mr. Jo Reed of the Department of French of the University of Birmingham for introducing me to phatic communion and research in linguistics, in 1975. I am grateful to Professor Alan Davies for his guidance at the outset of my research training and I am especially grateful to Dr. Hugh TrappesLomax, for his meticulous interest, valuable criticism, and encouragement throughout every stage of my research project. I should also like to thank the members of the University of Edinburgh's Department of Applied Linguistics and Institute for Applied Language Studies who offered me advice, agreed to test the intercoder reliability of the coding system, and sat the impenetrability test. I must also thank the University of Sunderland for giving me research leave to work on this book. I am, of course, very grateful to the six recordees themselves, without whose conversations there would be no book. Finally, I want to thank my family and friends for their understanding.
This page intentionally left blank
1 THEORY AND METHODOLOGY
1.1 INTRODUCTION "A certain group of prisoners got so bored with hearing each others' jokes again and again that they numbered them from one to a hundred and would amuse each other simply by the way they shouted out the numbers." (Anon). This tale illustrates the fact that groups of individuals who have interacted frequently over time develop a pool of common knowledge about the group, its activities and the individuals within the group, and seem to develop an economical way of referring to things, persons, events and situations in their shared environment. It was an interest in the elliptical nature of the language of individuals who know each other well and form a social group that motivated the present study. The question in the mind of the researcher was: what elements of the language of the casual conversations in social gatherings such as parties and coffeebreaks contribute to the vague, impenetrable nature of the language? what is it about the grammar and vocabulary used interactively that makes it so difficult for eavesdroppers to follow if they do not know what the speakers know? The interest was focused on the language of one particular type of social group: the academic discourse community, defined by Swales (1990) as a group whose members have common goals, intercommunication mechanisms, particular genres and specific lexis. The description is based
2 Analysing the language of discourse communities on a study of the casual conversations of six nativespeakerofEnglish students who took the 199192 MSc course in Applied Linguistics in the Department of Applied Linguistics (DAL) of the University of Edinburgh (EU) (N.B. Since this book was written the Department of Applied Linguistics has merged with the Department of Linguistics to form the Department of Theoretical and Applied Linguistics). These six subjects are considered to be representative of all the 1991 92 MSc students; they are also members of the academic discourse community of all Applied Linguistics MSc students of EU. This is an intense twelve month Masters course, consisting of three terms of lectures from October to May, and independent work on a dissertation in the summer months. Typically, in the autumn term, the students, who are mostly experienced language teachers working abroad, follow core courses and prepare short tutorial tasks; in the spring term, they follow selected option courses, sit a written examination and write three projects; and in the summer term, they follow option courses, write two more projects, take home a portfolio exam and prepare for their dissertation. The students are strangers to each other when they arrive, form a closeknit group throughout the year, and then return home at the end of the course, although a few stay on to do a PhD. The pressure of the course allows them little time for socialising together outside the department, apart from an occasional night out. Most social interaction takes place in lunch hours and coffee breaks; most casual conversations occur either in the department common room or in a university cafeteria. This book aims to show how the language evolves from the time the speakers first meet and becomes the ingroup code of the formed academic discourse community. Most studies reported in the literature of the language of groups and intimates are global, imprecise or unsystematic, and they describe the language as a product at a given time. Although general reference is made to the differences between the language of strangers and that of intimates, no systematic study appears to have been carried out to follow through the interactions of individuals as they form a group, to discover exactly how and when language changes over time as assumed knowledge grows. The main aim of the present study was to discover the precise changes in language that occur with increasing knowledge over time, and to use a longitudinal analysis to describe the language as a process. The study examines the development of the grammar, the lexis, and the interactive meaning across utterances and the implied meaning under utterances, and the way that these features combine over time, to contribute to the implicitness and impenetrability of the ingroup's language. It is a study of the relationship between context, interaction and implicit language, between common knowledge and the form of reference. It examines the social function of the students' conversations and how the purposes of talk change as the group forms. It analyses the function of the ingroup implicit language itself, in order to arrive at an explanation of why it is used. Familiarity affects grammar and grammatical categories are not only contextdependent but also change over time and are sensitive to members' assessments of and coconstruction of a context of background knowledge. This book aspires to add to the study of grammar in conversation, and problematises notions of grammar in context, in order to unpack the complexities of what has been called "background knowledge". Above all, it is hoped that by describing in detail the methodology of analysis and providing a systematic model for predicting the process of language change over time as individuals who start as strangers with common goals, interact together and are united by common ingroup knowledge until they form any discourse community, the book will stimulate both students and
Theory and methodology 3 experts in the field to pursue further research into the language of discourse communities. It suggests ways of expanding, generalising and applying the model to fields outside linguistics.
1.2 GENERAL APPROACH To DISCOURSE ANALYSIS The approach to discourse analysis and the interpretation of verbal social interaction used in the study is eclectic: it draws from both philosophy and sociology. It draws from philosophy, in that it is based on pragmatics (Grice, 1975; Leech, 1983; Levinson, 1983). It draws from sociology and anthropology, in that it is based on the functional principles of interactional sociolinguistics (Goffman, 1963; Gumperz, 1982; Tannen, 1989), the ethnomethodological principles of conversation analysis (Garfinkel, 1967; Jefferson, 1978; Sacks, 1972; Schegloff, 1968) and the structural principles of variation analysis (Labov, 1972). As Sadock (1985: 142) says, the best approach to discourse is to "examine structure in the light of functional requirement and function in the light of structural requirement". The approach does not add a dimension of psychology, as a psychological analysis of relationships is not appropriate in the context of the MSc course (see Chapter 2).
Meaning Of Context The view of context taken in this study is a primarily a pragmatic one. For pragmaticists, context is background knowledge of the world, shared experience, beliefs and assumptions, knowledge that guides a speaker's use of language. Communication depends on the interlocutors' cognitive ability to use the context of mutual knowledge, to draw from the presuppositional pool (Venneman, 1975). This approach particularly suits the study because its main purpose is to explore the relation between the MSc students' context of knowledge and the language used to refer to it. The study examines the way that each category of common and shared knowledge demands a specific form of reference, and the way that knowledge changes with interaction over time and affects the language. It diverges from the pragmatic methodology, whose data consists of "constructed utterances in hypothetical contexts" (Schiffrin, 1994: 12), since the study's source of data is the everyday interactions of a real group of speakers within a real context. In this sense, it follows the methodology of conversation analysis. As Ford (1993: 2) says, conversation analysis is "unique in being based strictly on naturally occurring data". The approach to context used here is also that of interactional sociolinguistics, which sees context as knowledge of social circumstances and expectations, and the situation of the here and now. The sociologist Goffman (1963) sees language as both reflecting and adding meaning and structure to social circumstances (Schiffrin, 1994: 97). He examines the way that social and interpersonal contexts "provide presuppositions for the decoding of meaning" (ibid: 105). The present study analyses the extent to which the physical circumstances of the common room and participants, and the psychological circumstances of the events on the course timetable dictate the topics chosen and the way that language is used. It deviates from Goffman's framework in that it takes a more linguistic view of interaction and also adds a dimension of psychological constructs such as goals and emotions. Interactional sociolinguistics is an appropriate approach for this study because it is devoted to the language characteristic of social groups, identities and relationships, and because it sees
4 Analysing the language of discourse communities identity as dynamic and maintained through a period of time. The linguistic anthropologist, Gumperz (1982) looks at speech communities and how the interpretation of language reflects race and ethnicity, with an aim to explaining communication difficulties between people of different sociocultural backgrounds. Tannen (1984: 31) studies the conversational style of social groups, defined by gender, race, ethnicity and class, with an aim to providing overall characterisations of features of talk. As she says "it is the creation of /.../ shared meaning communication that makes a collection of individuals into a community, unites individuals in relationships" (1989: 29). The study of MSc group is similar to Tannen's work in that it analyses language in casual conversations as an expression of group identity, but differs from both Gumperz's and Tannen's, in that it does not focus on the issues of gender, race, ethnicity and class, or suggest that ingroup members use their code to show superiority over another group, dominate or actively exclude others within the group. The present study focuses on individuals of similar backgrounds coming together with similar goals and forming a discourse community united by its common knowledge and needs. It differs from previous studies in its aim to follow through the changes in language and the social function of talk as the group forms. The approach to context used differs from that of the ethnography of communication (Hymes, 1972), which focuses on the general cultural background of beliefs and values, and a knowledge of the appropriate use of grammatical structures. The present study does not examine the effects of the general knowledge of a speech community, defined by Hymes as "a community sharing rules for the conduct and interpretation of at least one linguistic variety" (p.54) on their language. It examines the effects of growing specific knowledge of one particular discourse community, defined by their common goals, activities and experiences. The present study differs too in that it describes situation and text as dynamic and changing, whereas Hymes describes them as a static entity. Conversation analysis can also be seen as relevant to the present study, in that it takes context to be not only knowledge and situation but also text. Conversation analysts believe that language is both contextshaped and contextrenewing: it is created by context as speakers' "contributions to an ongoing sequence of actions cannot be adequately understood except by reference to the context in which they participate" (Drew and Heritage, 1992: 18), and it creates context: "Since every utterance will itself form the immediate context for some next action in a sequence, it will inevitably contribute to the contextual framework in terms of which the next action will be understood." (ibid). As Boden (1994: 99) says, "The conditions created in and through interaction thus have consequences far beyond the immediate interactional setting, embedding actors and activities in a very real world of their own making". Ochs (1996: 3) summarises the view of text as social context: "conversation analysts have become increasingly concerned with ways in which talk and interaction both organize and are organized by institutions, relationships and culturally specific environments." The present study gives importance to intertextuality (de Beaugrande, 1981), treating previous texts as part of the context of background knowledge, and taking into account the fact that in group members referring to entities not in the text, can assume knowledge of what was said in previous conversations. Exophoric reference is at the centre of the present study of implicit language. This study's approach to context does differ from that of conversation analysis, however, in that it places more emphasis on social relations and social identity within the group,
Theory and methodology 5 and above all on function: the purpose of choosing particular language forms, the illocutionary force of utterances, and the overall reason for talking.
Linguistic Analysis The approach to data analysis and description used in the study is one that starts with the structure and moves on to make statements about the function, one that looks at the language and then suggests how and why it is used. The method of analysis is therefore primarily linguistic, even though the view of context is rooted in interactional sociolinguistics which is fundamentally functional. The building of the linguistic model of analysis is based on units of reference devised from Halliday and Hasan's (1976) analysis of cohesion and deixis. The study examines referring expressions and their role in cohesion, deixis and definiteness, and lexical items of general and specific reference within the text of the data. Unlike systemicist analysis, however, the analysis is not in terms of information structure or the complex issue of theme and rheme, since the hypotheses about changes in language over time do not touch this vast field. The issue of old and new information is dealt with on a grammatical referential level, in terms of anaphora and non anaphora. It is the interest in reference and indeed inference, that places this study firmly in the field of pragmatics. Pragmatics gives importance to referring as a discourse process, and explores definite and indefinite reference, and explicit and inexplicit reference. Grice (1975) sees referring items as pragmatic processes through which the hearer infers the speaker's intentions, guided by the cooperative principle: the maxims of quantity, quality, manner and relevance which constrain the speaker's choice of referring terms. He observes that definite referring expressions depend critically on mutual beliefs and suppositions. This study is based on the pragmatic inferential model of communication; it examines how interlocutors refer vaguely to entities that are new to the text but known in the context. The model of analysis contains some of Levinson's (1983) presupposition triggers: definite descriptions, comparisons and contrasts and nonrestrictive relative clauses. It also analyses the flouting of Grice's maxims, in the humorous conversational implicature of the implicit language. It limits the analysis of conversational implicature to that which features in humour because that is where it is most visible and it functions as a strong claim for ingroupness. The approach used deviates from the pragmatic in that it does not consider the logical analysis of each utterance in terms of presupposition, ("anything that it is reasonable to believe that the speaker is assuming or taking for granted", Palmer, 1981: 171), entailment or inference (what the hearer concludes that he is "reasonably entitled to draw from a sentence or utterance", Hurford and Heasley, 1983: 279). Such analysis of data would depend heavily on the analysts' own assumptions and intuitions. The study does not contain definitions of the speakers' intentions in individual utterances, or judgements as to whether they are satisfied with their hearers' interpretation of their intentions, because the hypotheses about changes in language over time are not about the logical relation between each utterance but about the relation between the form of reference and context.
6 Analysing the language of discourse communities The interest in reference and situated inference also places this study in the field of interactional sociolinguistics. Gumperz's (1982) "contextualisation cues" are central to the model of analysis. They are "speakers' and listeners' use of verbal and nonverbal signs to relate what is said at any one time and in any one place to knowledge acquired through past experience, in order to retrieve the presuppositions that must rely on to maintain conversational involvement and assess what is intended." (Gumperz, 1992: 230). They indicate those aspects of the context that are relevant to what the speaker means and enable interactants to make inferences about one another's intentions and goals (Drew and Heritage, 1992). Gumperz's model includes prosodic features, paralinguistic signs, sequencing strategies, code choice, choice of lexical forms, formulaic expressions and syntax. He said that contextualisation cues are part of any group members' code: "Members have their own socially defined notions of code or grammatical system" (1982: 99). The present study adapts the theory of contextualisation cues to suit its purposes. It builds on Gumperz's model to make a more systematic and detailed linguistic analysis of the code choice, the choice of lexical and syntactical forms. It focuses on a code of nonanaphoric definite implicit reference and vague ingroup lexis. The study does not include Gumperz's prosodic features or paralinguistic signs, even though an exploration of phonological pointers to common knowledge, such as stress, intonation and key, would have contributed usefully to the examination of contextbound language. Speakers stress the part of the utterance that contains the new information or part of the assumed knowledge that has special focus; they choose fall or rise intonation depending on their assessment of the mutual state of knowledge between themselves and hearer (McCarthy, 1991: 106); and they use low key for reiteration or assumed to be known knowledge (Brazil and Coulthard, 1980). This aspect of phonology is unfortunately beyond the scope of the study; a systematic exploration would have required detailed analysis of every utterance. The study's method of linguistic analysis is also influenced by variation analysis (Labov, 1972), in as much as it takes a linguistic approach and centres on the shared code. Labov's approach uses quantitative and linguistic methods to code and count components, in order to look at the patterns of linguistic variation and change through the systematic investigation of the discourse community. Labov's approach suits the needs of the present project which aims to measure quantitatively the changes in language, even though his tendency towards dialectology and the study of the groups united by race or class makes his objectives quite different. It is this need for measurable units and systematic analysis that separates the present study from the techniques used by conversation analysis, which has an antiquantitative analysis approach. As Eggins and Slade (1997: 31) say, "Lack of systematicity is a problem for all aspects of the CA account of conversation."; dealing with "close up focus on small excerpts" limits their ability to deal comprehensively with complete sustained interactions.
Functional Analysis Pragmatics and interactional sociolinguistics are also at the basis of the approach used to analyse of the macrofunctions of talk and the implicit language. The politeness principles (Brown and
Theory and methodology 7 Levinson, 1978; Leech, 1983) of pragmatics state that the speaker should claim common ground and convey that he and the hearer are cooperators, and that the speaker should not presume that the hearer will cooperate or coerce him to do so. Interactional sociolinguistics analyses how involvement is created in interaction (Gumperz, 1982), how social occasions create expectations for a display of involvement (Goffman, 1971), and how strategies of rapport and involvement are expressed in conversational style (Tannen, 1984). This study explores the MSc students' adherence to the politeness principles, in its examination of group cohesion. Students' interactions are examined to find what balance of the transactional function and the interactional function (Brown and Yule, 1983) is present; they are examined to discover whether the students claim common ground to show involvement and sympathy or to test the normality of their experiences and attitudes, and whether they assert ingroupness or use ingroupness. The model for examining function at a microlevel is derived from, but quite different from, that of Speech Act Theory (Austin, 1962; Searle, 1969) and that proposed by Systemic Linguistics (Berry, 1989; Fawcett, 1973; Halliday and Hasan, 1976; Hoey, 1983). Speech Act theorists see language as being used to perform acts and they list the communicative acts: directives, commissives, expressives, etc. They begin with function and then move on to establishing linguistic realisations of the functions, which is the opposite of the bottomup approach used in this study. They deal less with actual utterances than with utterance types; the closed set of options is not linguistic. Systemic models of microfunctions contain a set of some 50 speech acts, with one linguistic form for each of these. They are a priori models that propose algorithms of act sequences that can be predicted in all verbal interactions. However, as Drew and Heritage (1992: 12) point out, it is difficult to interpret how a particular utterance was understood by participants and how it squared with speaker's intentions, and in naturallyoccurring data, "sentences and utterances are shaped to occur in particular sequential and social contexts". In addition, as any discourse analyst will know, one form can have several microfunctions, and one microfunction several forms. Neither of these models suits the purposes of a description of casual conversations because "chat" is not primarily necessitated by a practical task (Cook, 1989). The MSc students are not "doing" anything in the sense of physically carrying out a joint activity together with a visible end result; they are not "requesting", "persuading" and "promising" etc. in order to get something done. They are simply sitting in the common room together, "asserting", "informing" and maybe negotiating agreement (Firth, 1994), in order to say something as they pass the time together. Even the theory of indirect speech acts is not entirely relevant, because it is mainly concerned with indirect ways of "requesting", "persuading" and "promising" etc. in order to get something done. The microfunction model used was born of ethnomethodological observation of the data and designed to suit the characteristics of the casual conversations of a budding academic discourse community. It starts from Goffman's (1963) interpersonal rules of social interaction and adds psychology. It relates the micro to the macrofunction, the illocutionary to the perlocutionary force, building on politeness principles for claiming ingroup membership. The 50plus speech acts, along with the features of reallife spontaneous conversations such as phatic fillers and backchannels, are grouped together under nine general headings according to the object of discussion and the attitude being displayed. Speakers discuss either themselves, the interlocutor or a third party; they display either a neutral, a positive or a negative attitude. It can be seen in
8 Analysing the language of discourse communities the light of Eggins and Slade's (1997) description of "interpersonal semantics", in which talk with attitudinal colouring is known as "appraisal", and involves evaluations of reality and expressions of emotional states and judgements, and is closely related to the semantics of involvement, which involves informal ingroup language.
Conversational Organisation Analysis The model used does not include a vision of conversational organisation or overall conversation structure, principally because one is not needed in the analysis of implicit language and group formation, but also because the models that have been devised do not fit casual conversations. The Birmingham School (Sinclair and Coulthard, 1975) model of moves, based on findings in classroombased research, and consisting of the structure InitiateRespondFollowup does not seem appropriate for the analysis of casual conversations, because followups or evaluations of the interlocutor's words do not occur frequently, and because thanks to casual conversation's generally loose structure and drifting topics, it is often not clear or even helpful to distinguish an initiating move and a responding move. Casual conversations are also different from classroom exchanges in that the unequal power between participants tends to be suspended, the number of participants is small and the turns are short. The models of genres in casual conversations cannot cover all the types of informal chat in the study either: Labov's (1972) model of narrative structure, (abstractcomplicationevaluation resolution), Plum's (1988) models of "anecdotes" (remarkable event and reaction), "exemplums" (incident and interpretation), and "recounts" (orientation and record of events), and Slade's (1995) model of gossip (substantiating behaviour and pejorative evaluation) are descriptions of microstructures that can occur within casual conversations. Chat "cannot be characterised in generic terms" (Eggins and Slade, 1997: 268). The texts in the data studied do not even have a beginning, middle and end. The approach to conversational organisation analysis used in the study is closer to that of conversation analysis, which centres on the patterns and mechanics of conversation: turntaking, turn constructional units (spates of talk, environment for sentences in turns), adjacency pairs, sequential relevance, overlaps, interruptions and pauses (Garfinkel, 1967; Jefferson, 1978; Sacks, 1972; Schegloff, 1968). This is not to say that the study attempts to analyse the data in terms of the overall organisation: the focus on changes in language as the group forms are grammatical and lexical and not related to the patterns of conversation. The study is not concerned with establishing a theory of a stable set of rules of sequences and structures across utterances. However, the model used does go further than studying language at the noun phrase or utterance level: it is grammar in interaction. The study is interactive and deals with structures across clauses and styles of reference stretching over utterances. As Ford (1993: 11) says, "It is crucial to the turntaking system that interactants share an understanding of the grammatical and intonational resources of their language." Ochs et al. (1996) suggests that turn constructional units are organised by grammar. Firstly, the study examines how implicit contextualisation cues relate to each other, form lexical chains and syntactic webs through a text or texts and how the features are used interactively by interlocutors. Secondly, taking into account turntaking, it
Theory and methodology 9 analyses the grammar involved in topic shifts, in order to discover whether implicit language can occur on topic shifts and make the whole topic that follows implicit and potentially impenetrable, and to discover whether such topic shifts vary according to knowledge area assumed, and whether they increase with interaction over time. The study also takes into account sequential relevance and overlaps, in as far as it analyses clausal ellipsis and breakdowns in communication. As Goodwin (1979: 98) says, "sentences in natural conversation emerge as the products of a process of interaction between speaker and hearer and /.../ they mutually construct the turn at talk." The study examines how speakers leave sentences incomplete, either left unfinished or filled with a vague, semantically empty expression, and how interlocutors either complete the sentence for their colleagues or simply continue talking. It looks at the way that this interactive sharing of utterances varies grammatically and lexically according to the topic under discussion. It also analyses the way that the language is mutually constructed between interlocutors in the case of breakdowns and repairs of breakdowns. The study investigates whether one of the causes of hitch in the flow of speech can be the implicit language itself, occurring when the speaker references inefficiently and violates the maxim of quantity. Finally, the study suggests tendencies that can be seen as adjacency pairs, in its analysis of the microfunctions, "appraisal" and the semantics of involvement. It explores the ingroup members' positive and negative evaluations and their expressions of negative emotional states, and also their demonstrations of sympathy, empathy, advice, and whether each microfunction tends to demand a particular response. It examines each type of humorous conversational implicature and the subsequent turns to see whether the repartee of jokes, irony and banter can have a predictable reaction.
1.3 STUDIES OF INGROUP MARKERS Studies of the particular characteristics of the language of groups or familiars have shown, albeit in a superficial way, that the reference of familiars is elliptical and carries contextdependent meaning implicitly expressed. The earliest descriptions of ingroup language are those of linguists Firth (1957) and Bernstein (1971). Bernstein lists the features of his restricted code: short, contextdependent grammatically simple sentences, repetitive lexis, few subordinate clauses, more verbs, adverbs and pronouns than nouns, and a reduced range of lexical and syntactic alternatives. He says that these features "interact cumulatively and developmentally reinforce each other and so the effect of any one depends on the presence of the others" (1971: 43) but he gives no suggestion as to how this might happen. Others have made global observations, of an imprecise nature. Tyler (1978: 391) notes that spoken discourse is "notoriously elliptical and shorthand, especially between parties who know one another well and share common interests and expertise in the topic being discussed". Brown and Levinson (1978) mention contraction and ellipsis amongst their ingroup identity markers. Gumperz (1982: 131) suggests that "exclusive interaction with individuals of similar background leads to reliance on unverbalised and contextbound presuppositions in communication". Tannen (1984: 31) considers ellipsis, indirectness, implicature and unstated meanings to be interpersonal involvement signals of "high involvement style". She claims, "the more work /.../ hearers do to supply meaning, the deeper their understanding and the greater their sense of involvement with
10 Analysing the language of discourse communities both text and author" (1989: 23). None of these descriptions are longitudinal or developmental as the present study is. The language of groups of university students has been studied, and the findings with regard to the implicit form of ingroup code have relevance to the present study, but once again, they are descriptions of product rather than process. Kreckel (1981) says that the university students' language in her study consisted of "a multitude of ingroup codes, discipline specific and social group specific /.../ taking discipline or groupspecific knowledge for granted" (1981: 36), and Levy (1979), found that the students in his study talked about their selected course subjects in such a way that even the staff found it difficult to understand, but neither makes any attempt to explore how impenetrability might vary at different stages in the course. Other linguistic ingroup markers have been proposed, but the studies simply consist of lists of features. Kelly et al. (1983) list the use of "we" statements reflecting tendencies (e.g.: "we always...") and global assessment (e.g.: "we're better than..."). Berger and Bradac's (1982) markers of intimacy or "immediacy" include greater specificity (e.g.: "Jane" rather than "that person"), more spatial and temporal indicators of proximity (e.g.: "here's John" rather than "there's John"), and more forms of inclusion and mutuality (e.g.: "we" rather than "you" and "I"). Of these markers, the present study focuses only on specificity, because this fits in with the analysis of reference. The others fall outside the remit of the study of implicit language. Some studies compare the interactions of strangers with those of speakers who know each other well, in terms of function, style and strategy; their descriptions are again lists of features. Altman and Taylor (1973) say the language of "intimate" speakers is more emotional and contains more positive and negative evaluation remarks than that of "newly related" speakers. Goffman (1971: 326) says the speaker "will often present initially a front of diffidence and composure, suppressing any show of feeling until he has found out what kind of line the others will be ready to support for him." Jefferson, Sacks and Schegloff (1987: 160) have suggested that informality in terms of an informal register of idiomatic expressions, popular words, slang and obscenities is an "invitation to intimacy". Tannen (1984: 31) lists amongst her interpersonal involvement signals of "high involvement style": playful routines, irony, allusion, reference to familiar jokes and assumptions, tropes, and imagery. Planalp and Benson (1992) show that friends were more likely than acquaintances to offer opinions, ask for advice, express negative judgements, and use "jokes, sarcasm, teasing and repartee", "profanity and slang". These descriptions provide no suggestion of the route from the newly related acquaintance or low involvement stage to the intimate friend or high involvement stage. The present study traces the development of informality, emotional language and humour, in order to demonstrate that the ingroup has formed. Other studies have discovered on organisational and fluency markers, and phonological and acoustic markers of intimacy. Goffman (1971: 342) comes to the picturesque conclusion: "when a set of persons are on familiar terms /.../ then inattentiveness and interruptions are likely to become rife and talk may degenerate into a happy babble of disorganised sound". Stainton (1987) finds interruptions to be more acceptable in an environment where the participants hold a close relationship. Edmondson (1981) finds false starts and other hesitation phenomena to be indicators of speakers who know each other well. Giles and St. Clair (1979: 46), describing his sociopsychological Accommodation Theory of "the processes whereby individuals shift their
Theory and methodology 11 speech styles to become more like that of those with whom they are interacting", finds that convergence can be found in pronunciation, pause and utterance length, vocal intensities, and speech rate. Tannen (1984) finds marked pitch and amplitude shifts, a faster rate of speech and an avoidance of pauses, abrupt topic shifts and persistence of topic to be markers of high involvement style. The present study does not add all these dimensions to the linguistic one, partly because they not contribute to the study of implicit language and partly because the analysis of features such as speech rate and amplitude shifts etc. require special equipment for a systematic analysis. Moreover, nonfluent communication can have other functions, such as hedging, and it can have other causes, such as nervousness. The hesitation phenomena of omissible hedging verbs (Brown and Miller, 1980) such as "I think" and "I suppose" are more politeness ploys or softeners, planners or delaying devices allowing the speaker time to formulate his message, than markers of intimacy. The markers that the study does take from these are the abrupt topic shifts and persistence of topic in order to demonstrate that the ingroup has formed and to expand the analysis of implicit language on topic shifts.
1.4 THE STUDY OF THE Msc INGROUP LANGUAGE Model And Method Of Analysis Based on an appreciation of these approaches and the findings of studies of ingroup markers, several assumptions were made visavis the possible outcomes of the analysis of the implicit language of the MSc ingroup. It was assumed that the more time that the students spend together, the more common knowledge they share. It was hypothesised that the more knowledge the students share, the more they can refer to referents elliptically and implicitly, knowing that the specific referent is in the minds of the hearer and does not need naming or explaining. It was believed that any emotional appraisal using ingroup language can be a claim for ingroup membership in that it is a statement of inthesameboatness. It was surmised that the way of referring can be so elliptical and vague that only the group members themselves can be expected to understand it, and that even though an outsider to the group may not be able to grasp the full implication of individual utterances. The model of the ingroup's implicit language consists of the 'ingroup code' (lexical and grammatical items at the noun phrase and verb level: implicit reference, explicit nonanaphoric definite reference, special course nouns and general words) and 'other implicit features' (implicitness at the clause, utterance and exchange level: clausal vagueness, implicit language topic shifts, and conversational implicature). The method of data collection was both random and systematic. Spontaneous conversations were recorded overtly in the commonroom once a week in the first half of each of the three terms, from 4 October to 12 May. The half terms were different lengths: there were six recording days in the autumn term, five in the spring term and four in the summer term. Each day's recording lasted from 10 to 30 minutes, bringing the total number of hours of recording to about seven. A total of 29 dialogues (26,000 words) from 15 days' recording were used in the analysis. All the students on the course were told in the course welcoming address that the researcher would be recording their conversations, and their permission to record was sought before every
12 Analysing the language of discourse communities recording. They were neither asked to sit together, nor asked to talk if they did happen to sit together, nor guided in any way if they did happen to talk. Milroy favours spontaneity: "If we are to obtain any kind of insight into the structure of everyday spoken language we need to look at speech where the speaker has selected his own topic which does not emerge as a result of direct questioning" (1987: 59). The researcher stood anonymously on the other side of the room, in order to be near enough to turn off the recorder if necessary, yet far enough away to avoid being included in the conversation. It was decided to limit the study to a small group of native speakers on English and not to analyse conversations containing nonnative speakers on English (see Chapter 2 for the reasons), although it was impossible to cut out all nonnative speaker contributions because they were 18 out of the year's 34 MSc students. The recordees were students who often happened to sit in the common room; they were never altogether all at once: one day they came and talked in pairs, another in a combination of three of them, another day in a different combination of four. The main reason for randomly taking a small group was that it would permit the researcher to become familiar enough with each recordee to be able to detect any tendencies and changes that were being caused by the personalities and idiosyncrasies of the speakers, rather than by time, course events or knowledge areas. She also needed to learn about their lives inside and outside the course and to know something of their opinions and attitudes in order to share their interpersonal knowledge and thus be in a privileged position to interpret the findings in an informed manner. She was herself an MSc Applied Linguistics student in 199091. As a member of the discourse community of DAL MSc students, she could make decisions as to the reference, context and implications of the language in the study. As an outsider to the 199192 MSc in group, she could keep the exclusion of the vague language in perspective; her interpretations of meaning must at times be guesswork. There were four stages to the analysis of the data. The first stage consisted of transcribing the texts in discourse units (see Figure 1 for transcription conventions). The term 'discourse unit' denotes any stretch of talk uttered by one person that ends with a falling intonation (unless it is a closed question or a tag question with rising intonation) and a pause of more than 0.5 seconds. Text sentences were not divided at main or subordinate clause boundaries since analysis based on parts of utterances could have led to wrong assumptions. Thus a discourse unit can be several main clauses long or a single word such as "Yeah" or "Right", or sound such as "mhm" or "heh heh". The text was not divided into turns because one turn can contain several knowledge areas and topics, or moves because of the difficulty of defining where each one begins and ends. The next stage of the data analysis was to tag the text linguistically and labelling the fields functionally. See Figure 2 for the coding system and Figure 3 for a sample of fully coded transcription. In field 7, each noun, article, adjective, pronoun, and verb is tagged for grammatical features and lexical features, as is each instance of initial ellipsis, incomplete sentences and breakdowns in communication. Noun phrase tags indicate whether it is anaphoric or not, definite or indefinite, explicit or implicit, and whether it is a proper noun, a technical term, a general noun etc. The functional labels are in fields 1 to 6. The intercoder reliability of the code was tested, in order to check the functioning of this part of the code and to see whether other language teachers and linguists who used the categories to code the data would code it the same as the researcher did. Three excerpts, each about 30
Theory and methodology 13 discourse units long, were taken: 25%, 50% and 75% of the way through the data, each excerpt starting at the beginning of a new topic. Linguists and EFL teachers were asked to read an explanation of each part of the code and then to tag each referring expression or label each discourse unit in the three short excerpts which had been printed for them to read but not selected for them to listen to. A total of 30 testers took part in testing the various parts of the code. The Kappa test (the number of right answers minus the estimated score divided by total number of questions minus the estimated score) was used to calculate how closely their understanding and use of the code corresponded to the researcher's intentions.
14 Analysing the language of discourse communities Figure 1 : Transcription code and conventions At the beginning of each line: AM, BM, CM, DM the four male recordees, AF, BF the two female recordees, MM any male NS who is not one of the six recordees, FF any female NS who is not one of the six recordees, NM any male NNS NF any female NNS Within the text: XX, XY, and XZ any member of staff, X, Y and Z any student or student's family member, not present in the dialogue. = Interruption // Overlap /.../ Lines from the original omitted to make example quoted simpler (0.5) Pause (man) Possible interpretation of unclear words (single brackets) ((unintelligible)) No interpretation possible (double brackets) (heh heh heh) Laughter (as many "heh'"s as exhalations) ((sniffs)) Other nonverbal sounds (described between double brackets) ((BM enters)) Other audible activities/actions (described between double brackets)
Figure 2; Field labels of the coding system Functional Fields: 1) Dialogue and discourse unit numbers 2) Knowledge area 3) Macrofunction 4) Topic shift 5) Microfunction 6) Speakers Text Field with linguistic tags : 7) Grammatical and lexical
Figure 3 ; Sample of fully coded transcription Text field Functional fields 1 7 2 3 4 5 6 1015 2 S sc 31 BM I actually_#A did (2) have_10 a job_XAl_20_&HF set up in Argentina_21. 1016 2 S sc 31 BM same thing_NCJ2_22 happened_N103. 1017 2 S D 31 BM Fell_XLI_X10 through. 1018 2 S D 31 BM It_NEP2 was_l 1 inflation_NA2_20. 1019 2 S D 31 BM And then_#> it_NEP2 came_X10 back on again. 1020 2 S D 21 AM Oh. 1021 2 S D 31 BM By that stage_#>_NDJ2_20 I thought_X10 no way. (heh)
Theory and methodology 15 The final stage of analysis was the calculation of the densities of the elements of interest, in isolation or in combination with other elements. The percentage of the elements out of all words or units was calculated for the autumn, spring and summer terms, and for each of the knowledge areas. Measurement by density out of all words was used for the textual calculations for grammatical and lexical elements, and other implicit features, because several of these could occur within one discourse unit. Measurement by density of discourse unit was used for knowledge areas, topic shifts, macrofunctions and speech acts. Triangulation tests and tests of impenetrability were carried out. The significance of results was tested using statistical tests.
Map Of The Book Chapters 2 and 3 give the background to the main study. Chapter 2, The InGroup', describes the MSc student group and outlines the psychology literature on relationships. The chapter proceeds with an examination of ingroup markers to show that it is indeed a social group and brief outline of the triangulation test performed to check the conclusions of the study. Chapter 3, 'Knowledge Areas', describes the types of background knowledge that MSc students assume that their interlocutors have. It outlines the characteristics of the four common knowledge areas discovered: general knowledge of the world, knowledge of language teaching, knowledge of EU, The DAL and The Institute for Applied Language Studies (IALS), and knowledge of 199192 MSc course (the largest knowledge area). The chapter also shows how shared interpersonal knowledge of the interlocutor can feature in any of these areas, and that sections of dialogue depending on it increase over time. Chapters 4 to 7 give the main linguistic and functional findings. Chapter 4, 'Grammar In Of The InGroup Code', analyses nonanaphoric definite reference, and describes the categories "super explicit", "explicit" and "implicit" reference. The grammatical implicit contextualisation cues are nonanaphoric definite noun phrases, demonstrative pronouns and adverbs, and third person pronouns. It shows that there is a slight increase in implicit reference and that course topics contain the most implicit the nonanaphoric definite reference, with vague referring items pointing to specific referents in the background knowledge. Chapter 5, 'Lexis Of The InGroup Code', deals with the pragmatic meanings of nouns and verbs, and discusses the implicitness of the lexical items that have a special meaning that only ingroup members would understand. It discusses the gradability of contentfulness in lexis, analysing proper nouns, technical terms, special course nouns and general nouns and verbs, all of which are lexical implicit contexualisation cues. Chapter 6, 'Implicitness Over Utterances', contains an exploration of other features of implicit language that depend on ingroup knowledge at the clause, utterance and exchange level. The chapter describes the analysis of unfinished sentences and vague fillers. It describes humorous conversational implicature (hyperbole, irony, banter and ambiguous allusions) and how it contributes to the implicitness of the ingroup language. It also examines "hitches" or breakdowns in communication, and suggests that the ingroup code can break down. An analysis is made of how all these features combine with implicit contextualisation cues. Chapter 7, 'Function', deals with macrofunctions and microfunctions, to provide a background to the analysis of the ingroup code. It shows how interactional language expressing a positive attitude to the interlocutor and using the ingroup code becomes a way of claiming ingroupness and using the group to test normality. It concludes with a suggestion about the interactional function of the implicit language of the ingroup.
16 Analysing the language of discourse communities
Chapter 8, 'Further Study', contains a summary of the influences on the change in the language. It shows that increasing common knowledge over time is associated with an increase in implicit reference and changes in the function of communication. It shows that course events in the spring cause a peak in course knowledge dialogues and nonanaphoric definite reference, and a trough in humorous utterances, and that topic has a great influence on the implicitness of both grammar and lexis in reference. The concluding chapter also contains recommendations for research into the expansion and the generalisability of the model, and suggestions of applications of the findings to the fields of Teaching English as a Foreign Language, clinical pragmatics and forensic linguistics.
1.5 CONCLUSION Firth (1994), analysing the discourse of negotiation in the workplace, says that some linguists, working on detailed transcripts of actual, reallife instances of negotiation, attempt to uncover, in the minutiae of interactive talk, the meaningful, sociallyaccomplished features and interactionallyachieved structures of negotiation discourse. He laments that as yet, such studies are relatively few in number. This book constitutes such a study. It aims to describe the development of the implicit language in the casual conversations of an academic discourse community, that of the Applied Linguistics MSc students of EU. The aim is to discover to what extent both increasing knowledge with interaction over time and the topic area affect the forms and functions used by ingroup members of the community. This chapter has shown that previous studies of ingroup markers unsystematic descriptions of product, whereas the present study aims to describe systematically the process of the development. The chapter has outlined the approach used: the view of context and the theory behind the linguistic and functional analysis are principally ones of pragmatics and interactional sociolinguistics, but the study is enhanced by aspects of variation analysis and conversation analysis. The study examines the relation between the students' assumption of knowledge of the world and beliefs, and the form of reference used, analyses the effect of the course events on the language used. It explores the ways that interlocutors refer and infer, using the grammatical and lexical "contextualisation cues", to entities in the context of knowledge. The study takes from variation analysis the quantitative methodology to systematically code and measure the units of analysis. It is influenced by conversation analysis, in that it gives importance to intertextuality, and examines the way that implicit contextualisation cues on topic shifts, clausal ellipsis and breakdowns, the microfunctions and humour are occur interactively. Politeness principles and theories of involvement form the framework of analysis of the macrofunction of the conversations. Microfunctions are analysed according to the attitude being displayed, while speakers claim ingroup membership. The book describes the group dynamics of a discourse community in formation.
2
THE IN-GROUP
The chapter explores the psychology literature on relationships and describes the MSc student group. The review of the literature on relationships has the sole objective of defining the particular social group of the study. The objective in this chapter is not to carry out a psychological investigation into the development of relationships in the MSc student group but to discover the linguistic indicators of group dynamics. A full description of the course (its design, its content, its assessment methods, etc.) and the student group (their nationality, their occupation, etc.) is provided here so that readers may have detailed background knowledge of the context that is essential to an understanding of the conversations and indeed the implicit language. The chapter contains an examination of ingroup markers to show that the MSc student group is indeed a social group and it provides a brief outline of the triangulation test used to confirm impressions about relationships and language. If the literature is right in saying that ingroupness can be shown by the presence of markers of intimacy, then the discovery of these markers in the data analysed can be used to demonstrate that a group did indeed form or at least that its members wanted to claim ingroup membership. Once it has been demonstrated that a group formed, it can be confirmed that the language belongs to a formed discourse community, and the discussion of the ingroup language can begin. Markers of intimacy were reviewed in Chapter 1 to explain why certain features have been selected for the model of analysis in the study. In this chapter, markers are mentioned to show which ones were taken as proven ingroup markers of
18 Analysing the language of discourse communities group formation; they are informality, humour and topic length. Triangulation was needed check whether the six recordees actually formed a group of special friends (defined as people who spoke to each other at least ten to fifteen minutes three times a week) forming a minigroup within the course, or whether they were simply associates (defined as people who only spoke to each other about two to five minutes once a week) and therefore representative of all ingroup course members. It was felt necessary for the generalisability of the findings to the rest of the MSc group to know whether the characteristics of their language could be said to be representative of the language used with all members of the MSc course.
2.1 PSYCHOLOGY LITERATURE ON RELATIONSHIPS Definitions of close relationships in the literature contain psychological and social parameters to be fulfilled. Argyle and Henderson (1985: 64) define friends as "people who are liked, whose company is enjoyed, who share interests and activities, who are helpful and understanding, who can be trusted, with whom one feels comfortable and who will be emotionally supporting". This would be a tall order for the MSc students, considering the limited time that they have for socialising. Hinde (1987: 36) lists factors present in relationships between friends, such as trust, intimacy ("the extent to which the participants in a relationship reveal all aspects experiential, emotional and physical"), interpersonal perception (whether interlocutors see each other as they really are), etc., but these are difficult to measure and are not the object of this study. MSc students rarely reveal aspects of themselves and do not have enough time to interact with each other to be able to have perfect interpersonal perception. Definitions of close relationships in terms of social activity are not applicable either. Duck (1991: 64) says categorically, "Friendships do not start until people do friendly things in friendly places: they are not created merely by friendly talk.". Friendly talk is all some MSc students have time or energy for. Argyle et al. (1981: 383) state, "People will like each other if they meet frequently and under rewarding circumstances". The students' anxious conversations during project time can hardly be called "rewarding circumstances" unless "rewarding" can be understood as satisfying in terms of stress relieving through expressions of solidarity. Many writers describe the stages of development of a relationship. Duck (1991) outlines four stages that could apply to all relationships: that of selecting friends, that of developing techniques for encouraging the relationship, that of making the relationship grow, and that of maintaining the relationship. The stage of selecting friends amongst those of similar attitudes, emotional stages and intelligence does not generally apply to MSc students. The students are thrown together by the course and held together for the duration of the course by the common goal and course activities. They do not meet all their colleagues by choice. They simply find themselves sitting next to others in the lecture hall, in tutorials, in the common room. As the course progresses, they meet the people who take the same options. The students whose conversations are analysed in this study are ones who frequently sit together in the common room; the others disperse and go in groups or alone to a cafeteria, their homes, etc. Many students eventually lose contact with each other as they choose different options and concentrate on their private study in term three. There are no "stages" as such, in their interactions.
Theingroup 19 Hinde's (1987) definition of a relationship suits the present study: "a series of interactions over time between ... individuals known to each other." The MSc students can be called an "aggregation", which is Hinde's term for a group formed because of a common attachment. They can also be seen as an "ingroup", Levinson's term for any speakers who interact over time and form a social group. Most of the "ingroup members" have a "colleaguetype relationship", a relationship of "associates", although a few develop relationships that extend beyond the course. Orestrom (1983: 41) makes an astute observation about the colleaguetype relationship: "The fact that two people work at the same place does not, even if they talk with each other every day, indicate anything about their real personal relations". The MSc students could be said to be similar to "workmates" but differ from them in that the question of rolerelationships, power and dominance is not so pronounced. However fierce the competition and however impressive the group leaders may be, the students are all "in the same boat". Swales (1990) describes the academic class as forming a discourse community. The MSc student body in this study fulfil all his criteria for members of a discourse community. They have the broadly agreed common public goal of passing the course; their mechanisms for communication are mainly facetoface interaction whether in tutorials or in the common room; these mechanisms they use to provide feedback, solidarity and relief from anxieties; they acquire a special lexis.
2.2
THE Msc STUDENT GROUP
The MSc course in Applied Linguistics in the University of Edinburgh consists of three terms from October to May. It is pyramidal in design. In term one, all students must follow a broad base of ten compulsory core courses, such as Grammar, Syntax, Semantics, Phonetics and Phonology, and Language and Linguistics. In term two, they select six option courses out of the 31 on offer, examples being Discourse Studies, Classroom Based Research, Pedagogical Descriptions, Psycholinguistics and the Psychology of Language Learning, Second Language Acquisition, and Language Programme Design. In term three, they select three option courses. In the summer months, the students write their dissertation. Throughout the year, the department runs an optional monthly programme of invited lecturers. As the amount of input in the form of lectures and tutorials decreases from term to term, the amount of independent assessed work increases, and the atmosphere and group dynamics changes. In the autumn term, students are very much guided in their study and are asked to complete short tutorial tasks which they prepare together in pretutorial workshops. In this term, conversations seem to be low key presentations of the individual and exploration of the other individuals. The students start as strangers and the group does not really gel in the autumn term. In term two, the week after they return from Christmas, they have a written examination on the core courses (they answer three questions of their choice in three hours). They have to complete three pieces of independent work: one core project and option projects A and B, each of 35,000 words, by the Easter holidays. Projects can include original research, the use of questionnaires, etc. Thus in this term, they separate out into their option groups and are stressed because of the course pressures. The group seems to have gelled by the middle of the term. In term three, which is only a half term of classes, they have more independent work: option projects C and D, and a portfolio takehome exam (with 72 hours for completion of three questions, each question on a different one of their option courses) at the end of the half term. They also have to prepare for
20 Analysing the language of discourse communities the summer dissertation of 15,000 words, whose deadline is September 31. At the end of June, an external examiner moderates assessments, interviews students for quality control and helps to judge borderline cases. In the last term, they meet less therefore, and there is a slight disintegration of the group. The stress level seems lower, as if students had "got into the swing of things". Over the summer, students work on their dissertation entirely on their own; they rarely come into the department. The majority of lectures and tutorials take place in the main lecture hall, seminar room and tutors' offices in DAL. DAL has a small departmental library but students tend to go to the nearby university main library to study and make use of the computer facilities in the David Hume Tower (which houses the Arts Faculty) across the road. Some go to the library of IALS. Lectures and tutorials directly related to the Teaching of English as a Foreign Language tend to take place in IALS. Regular and permanent teaching staff in IALS teach English as a Foreign Language in ESP, EAP and teacher training courses as well giving their services to DAL. IALS adds its welcoming 'bun lunch' for DAL MSc students to the DAL welcoming 'tea and biscuits'. It allows MSc students to use its library. Students have various courses of action in DAL. Some stop their study in May, and do not do the dissertation; they are awarded a Diploma. Others do the MSc course on a parttime basis, spreading their classes and projects over a period of two years. Some students continue in the department to "convert" their MSc to an MLitt in Applied Linguistics (either in DAL or in IALS) and some then convert their MLitt to a PhD That is to say, their MSc counts as the first year towards their MLitt or PhD. One or two students a year win an IALS scholarship to do an MLitt in Applied Linguistics (a research degree equivalent to the English DPhil). The IALS scholarship funds the students to spread the MLitt over two years and in return, the Institute requires the student to undertake 480 hours of teaching, mostly in the summer courses. At the end of the year, IALS invites all MSc students to teach on its summer courses, in the Institute itself, in the David Hume Tower or two miles away in the science faculty at King's Buildings. The MSc students have similar characteristics from year to year. The course attracts both native speakers (NS) from Canada, Ireland, Scotland, England and Wales, and nonnative speakers of English (NNS) from Europe, Asia and Africa. They are mostly language teachers with teaching experience abroad, although there may occasionally be a translator or speech therapist in their midst. In the 199192 course, there were 34 students, 16 of whom were native speakers of English, and they had an average age of 31. Of the six NS whose conversations were analysed, three were English, one was Irish, one was Scottish and one Canadian. They were four men and two women between the ages of 25 and 40. NNS conversations were omitted from the analysis, because of cultural, linguistic and functional differences, that became apparent in impressionistic observation. 21 % of all hitches (breakdowns in the flow of communication) were caused by NNS lack of cultural background knowledge of Britain, which is significant considering that they occupied only 8% of all discourse units. As Cook (1989: 74) says, "Misjudgements and mismatches of schemata are particularly likely when people try to communicate across cultures and across languages". In a study that placed such importance on the influence of background knowledge, it was not convenient to have NNS knowledge problems clouding the picture, even though their presence in the commonroom is a natural part of MSc life.
The ingroup 21
Gumperz says that NNS have linguistic problems related to reference and inference: "their assumptions about what information is to be conveyed, how it is to be ordered and put into words and their ability to fill in the unverbalized information they need to make sense of what transpires may also vary." (1982: 172). The most important reason for omitting NNS conversations is that there was no way of ascertaining whether NNS use the ingroup implicit language features, which are at the centre of the study. As Trask (1995) points out, different languages demand different degrees of explicitness: for example, native Australians avoid vagueness whereas speakers of Malagasy are reluctant to be explicit, preferring noncommittal vagueness. Moreover, NNScentred conversations contain metalanguage that is not present in predominantly NS conversations. NS talking to NNS often had to explain words, repeat and ask for repetition more than they would if speaking to NS. Some NNS lacked knowledge of the meanings of idiomatic language. Example (1) shows a breakdown in communication that occurs because a NNS has to ask for the meaning of a colloquial general noun which the NS modified playfully: (1) 07036 BM But I haven't got the thingymajog in my em= /.../ 07039 BF Thingymajog! » 07040 NF What's that? > 07041 NF A jog? 07042 BF Itit should be thingymajig. (0.5) 07043 BF It means thingy. (0.5) 07044 BF When you can't find the right word. 07045 NF // Ah. The humour of the NS also seems to exclude NNS from relating closely with them. In the data, NNS occupy only 3% of discourse units containing humour; again this is significant considering that they do occupy 8% of all discourse units. Since this study examines types of humour, those who have a different cultural background could not be central to the study. On occasions, a NS gave serious words a coating of humour, and the NNS laughed, thinking that the speaker intended to be funny, and showed themselves to be out of touch. In example (2), the NS are discussing the disadvantages of going on to a higher degree. NF interprets a comment expressing a serious concern as something that can be laughed about; the laughter is not interactive: (2) 27053 DM You have to live like a student as well // for three years. > 27054 NF // (heh heh heh) 27055 CM Yeah. (1) 27056 CM Which is not as fun as living like a worker for two years. 27057 DM No. The final reason why NNS conversations were omitted was that they contained functional features that were not present in NS conversations. Levinson (1978: 253), talking of misunderstanding that can happen if the speaker is using a foreign language, says that his politeness principles model "puts into perspective the ways in which societies are not the same interactionally, and the innumerable possibilities for crosscultural misunderstandings that arise". NNS presumably bring to their conversations different politeness and cooperative principles, and speech act realisations. As Hatch (1983: xii) says, "While speech acts may be universal, the range of ways of expressing each differs from language to language". In (3), a NNS intends to ask for an explanation of the meaning of a word but the form that NF uses is interpreted as a check that she has heard correctlv. Thev are talking about the portfolio exam:
22 Analysing the language of discourse communities (3)
13163 AM 13164 AM 13165 AM » 13166 NF 13167 CM > 13168 NF 13169 FF 13170 FF
So they expect quite long answers. I mean not like like the exam. Three hours. Long answers? Long answers. Howhow long? It depends. They they haven't decided how long you've got to do it yet.
2.3 INGROUP MARKERS Informality Informality, in terms of initial ellipsis, slang and expletives, was taken as an ingroup marker in this study, since the use of informal language can be seen as a strategy to gain acceptance and in group membership. It was assumed that if all the students use an informal style, they are accommodating (Giles and St.Clair, 1979) and aligning their expression to their interlocutor's utterances in order to be accepted by the rest of the group and cohere with them. The term 'initial ellipsis' is used to mean declarative and interrogative sentences without a beginning, i.e. with the subject or operator missing, such as ones beginning "Been there" or "Done that". Quirk et al. (1985: 895) call this situational sentential ellipsis, saying that "the interpretation may depend on knowledge of a precise extralinguistic context". The element missing could be the impersonal or inanimate subject "it", "that" or "there", and/or the verb to be: "is", "was", etc. For example: (4) 05068 BM Then they start picking up March April. » 05069 BM Not surprising really. ((2)) + 05070 AF Sounds plausible. The element missing could also be the personal subject and the operator "have", "has", "would", "did", "does", etc. as in: (5) >21001 DM Have a good weekend? > 21002 DM Do anything? = » 21003 AF Do anything? (1.5) 21004 AF I don't think so actually. (5) Slang and expletives can be in the form of a noun (e.g. "crap" and "dosser"), a verb (e.g. "slag" and "knacker"), an adjective (e.g. "bloody" and "crummy") or expressions such as "hell of a", "for goodness" sake" and "what on earth". In the category 'slang' is included informal words such as "slob", "kid" and "quid", and popular phrasal verbs such as "go on about", "keep mum about" and "stick with". The category 'expletives' includes obscenities such as "bloody hell" and "screw them", and blasphemy such as "Oh God" and "Christ". Eggins and Slade (1997) say that interpersonal worlds are shared using vocatives, technicality, slang and swearing and that these are the markers of involvement, which is part of interpersonal semantics. Expletives were analysed with caution because they can be as much a product of tension and anxiety as an invitation to intimacy. Argyle et al. (1981), in their examination of environmental setting, found that interpersonal threat and the threat of a new situation are anxiety producers, as is
The ingroup 23 "overloading", in terms of excessive stimulus of information for extended periods. The MSc course can make demands that make the students feel stressed and overloaded with information, especially in the spring term. The present study did not take any expressions of anxiety as markers of intimacy, because anxious speakers often show their anxiety to total strangers. It did not, therefore, follow up the markers of extreme anxiety as suggested by psychologists: simplification, repetitions, less diversified lexical content (Osgood and Walker, 1959), "ah's", sentence changes, repetitions, omissions, tongue slips, sentence incompletion (Scherer and Giles, 1979) or tone, breathy voice, rapid and irregular tempo and speech errors (Kasl and Mahl, 1965). Analysis shows that over time there is an increase in informality, calculated as the percentage of discourse units containing initial ellipsis, slang and expletives out of all discourse units. Figure 1 shows the difference between the proportions of discourse units with all informal features and those without informal features in each of the three terms. The difference is significant: the value of x2 was 12.206, significant at the 0.005 level. Initial ellipsis increases throughout the course: the autumn term 2%, the spring term 4%, and the summer term 5%). The level of slang and expletives rises slightly: the autumn term 2%, the spring term 2%, and the summer term 3%). Figure 1: Frequencies of informal features Autumn DUswith informal features:
yes no
27 713
Term Spring 99 1686
Summer 81 1001
Although the level of slang and expletives rises slowly, these features are striking as ingroup markers. Only people who have known each other and interacted for a certain length of time are likely to assess the main male protagonist of a televised novel in the following terms: (6) 11052 AM Urn he's a bit of a:: > 11053 BF Well (1) sort of like a willy with on two legs!. 11054 AM Um. > 11055 AM He's a bit of a wimp in the film. Example (6) demonstrates the interactiveness of informality: one speaker uses it and the interactant echoes it. Only people who have interacted over a certain length of time would feel free to dramatise their narratives with expletives, as in (7): (7)> 19086 BM When I saw the Thames Estuary, I said where's the bloody water? The use of expletives can be seen to be a result of tension and anxiety, produced by the demands of the course and the competition with colleagues, as well as by the increased intimacy. When the pressure increases, the students share their nervousness about project deadlines: (8) 20024 CM // Just a week // to go. » 20025 BM // Oh God right yeah. and this tension spills over into all aspects of course life. In (9), BF remembers her class: (9) » 14072 BF Bloody hell! 14073 BF Is it eleven already?
24 Analysing the language of discourse communities Humour Humorous utterances are markers of ingroupness because humour usually implies an expression of knowledge of values and attitudes, and therefore emotion, and expressions of emotion are markers of intimacy. Fillmore says that it is clear "how snugly the spontaneous humor of natural conversation is embedded in the lives and experiences of the people among whom it is exchanged", and that "humor is an expression of the collective experience of the participants, and receives response only from those who share common concerns" (1994: 271272). The categories for analysing humour differ from those proposed by Grimshaw (1994) who says that there are two sorts of manoeuvres: those that do not exploit linguistic form, such as telling a funny story, clowning, exaggerating, downputting, and those that do exploit linguistic form, such as pretended misunderstanding, inappropriate register, blatantly inappropriate implicatures (eg: "Do have anything you want to ask before we send you away?"), playing with grammatical rules, playing variations in speech routines (eg: "thank you very much", "You're welcome very much.") and allusion to jargon (eg: "in quotes middle class group"). In the data, humour is analysed in two categories (see Figure 2), depending on whether it lies in the meaning of the words themselves or in an implied meaning. The first category in this study is called 'nonflouting' humour: speakers talk about funny situations or use funny language. It is public, in the sense that it is accessible to most hearers because it does not usually depend on knowledge of information that is not mentioned. The second category is called 'flouting' humour: speakers humorously flout the maxims of cooperation (Grice, 1975), using hyperbole, irony, banter or ambiguous allusions, (see Chapter 6 for a detailed discussion of flouting humour). 'Public flouting' humour depends on knowledge of information that is not mentioned but it is still usually accessible to most hearers. It consists of irony and hyperbole about a third party or a situation, and ambiguous implicit humorous allusions to a third party or a situation. 'Private flouting1 humour also depends on knowledge of information that is not mentioned, but it is private in the sense that it depends on shared interpersonal knowledge about the interlocutor, which makes it accessible only to in group members. It consists of banter (mock impoliteness), irony (mock politeness) about the interlocutor, and ambiguous implicit humorous allusions to the interlocutor. Figure 2: The sources of humour Nonflouting humour Public Public funny situation funny language
Flouting humour
hyperbole irony ambiguous allusion
Private banter irony ambiguous allusion
In the coding of humorous utterances, only the discourse units that contained the funny lines and any verbal response that contributed to it were analysed. This means that discourse units that contained laughter, written "heh heh heh", were not coded or counted as part of the humorous units. The presence of "heh heh heh" obviously does not indicate that something funny has been said or even that something intended as serious has been taken as funny. Students often laugh when they feel nervous or are filling in the silence between their utterances. Similarly, the absence of "heh heh heh" does not mean that there has been no humour. A student can say something amusing and his colleagues may just smile, outwardly or inwardly.
The ingroup 25 Analysis shows that the density of discourse units containing all forms of humour (calculated as a percentage of all discourse units) taken together drops dramatically in the middle of the course, and returns to the level of the beginning at the end (autumn: 9%, spring: 7%, summer: 9%). This means that as the group forms, the level of humour does not increase. The reasons for this will be discussed in Chapter 6. For the moment, it should be underlined that humour is present right from the first term. All nonflouting humour seems to be aimed at adding enjoyment to the conversation and making the speaker seem more entertaining; he "displays" and makes himself more attractive and thus may bring the hearers closer to him. Nonflouting humour consists of joking, describing fimny situations, dramatising and impersonating. When MM is telling a long story about a hypnotist show, for example, he makes his audience laugh by adding direct speech: (10) —»02100 MM And another thing the one who's wearing these glasses the the hypnotist walks in front of the guy with the glasses on and the guy with the glasses goes urghh! 02101 MM He suddenly saw through yeah. 02102 FF (hehhehheh) Nonflouting humour also consists of using funny language in the sense of mixing registers, using cliches ironically in a serious context or using elaborate and formal language for everyday events. Speakers use the language of other domains to add a sparkle to their conversations. In (11), MM uses the language of advertising to detract from himself when CM praises his new haircut; CM echoes the humorous language by using a cliche himself: (11) 27175 CM Oh sweet cut. 27176 DM Wow. 27177 NF What's that you've changed. > 27178 MM I want to look different. 27179 CM You have nice hair though. > 27180 CM Always have. > 27181 CM Always will. In (12), a joke about what initials could stand for, CM enjoys mixing vulgarity and expletives with the revered academic institution of the MSc: (12) »18009 CM // Medium Level in Shit 18010 CM What's that? > 18011 CM MSc? 18012 BM Something like that you know. 18013 BM He's so = > 18014 CM Medium Level in Shit and Crap Flouting humour can also contain an element of "display" and entertainment, and it is possible that for this reason it is present right from autumn term. In a chain of narratives about bus drivers, the hyperbole is interactive: each speaker increases the exaggeration in order to make their contribution seem more relevant: (13)—» 04049 BF We had a really neurotic bus driver this morning. /.../> 04057 DM Demented bus driver. // (hehhehheh) 04058 BF 04059 AF // (hehhehheh) 04060 AF You get the occasional one. (2.5) —> 04061 AF The wild ones are on the on the green buses actually. > 04062 AF They go at a hell of a rate.
26 Analysing the language of discourse communities > 04063 AF Don't stop for anything The main function of humorous hyperbole, irony and allusions, however, appears to be that of asserting ingroup membership. In (14), an example of humorous hyperbole, students exaggerate the consequences of the new option courses separating old friends, in order to amuse each other and to show good will: (14) 15148 BM We'll have to go out sometime. 15149 DM Yeah. > 15150 BM Before we forget each other's faces. 15151 BM // (heh heh) It's true. 15152 DM //(heh heh heh) The interactional functions of entertaining and of claiming ingroup membership, in flouting humour are very often inseparable. When, in (15), BM flouts the maxim of sincerity to exaggerate his plight writing his project: (15) 17032 BM I wrote my introduction and I thought that sounds really good. » 17033 BM I thought but (0.5) I'll be here for ever if I discuss what I've just // said I'm going to discuss. 17034 NF // (heh heh heh heh) he is entertaining his colleagues with his dramatic exaggeration, and at the same time he is stressing the shared background (they are all writing projects) and the shared values (they are all having their problems). Private flouting humour, a marker of such intimacy, is used much more than the other sources of humour right from the beginning of the year. This suggests either that the group has gelled already in autumn term, or that it is used in autumn term as a way of establishing ingroup membership. It is a marker of trust and intimacy more than ingroupness: to use interpersonal irony, banter and allusions, the speaker must be familiar enough with the interlocutor to be able to refer, uninvited, to him, his life and beliefs. Banter in the form of a mock rejection of the interlocutor is possibly one of the strongest ways of saying how much one enjoys their company. In (16), a student enters the common room and there is nowhere to sit; DM gets up: (16) 29105 DM Nono I was1 was going to move anyway. 29106 DM Shit! » 29107 DM I can't sit here and er be chatting to AF all day. 29108 AF (heh heh) 29109 DM I'm afraid. 29110 AF Oh well. Interpersonal irony and banter can sometimes have the secondary aim of "bringing the speaker down a peg", reducing the intensity or importance of what he is saying. Irony can be an excessive praising and congratulating, as in: (17) 14031 BF I beat the sandwich rush. 14032 DM Oh yeah! (heh heh) » 14033 DM That's quite clever. (5) > 14034 DM On the ball as ever. Banter occurs in nine cases out often between members of the opposite sex; this form of teasing sometimes verges on lighthearted, comradely flirting. BF enjoys teasing her male colleagues about the work that they are or are not doing. In (18), DM shows pleasure that a deadline is to be moved forward and BF reprimands him warmly:
The ingroup 27 (18)
23034 DM // O::h. 23035 BF Ah. 23036 BF Oh you dosser! 23037 BF You're an absolute dosser! 23038 DM (hehhehheh) Brilliant= Banter might also be used as a probe for potential romantic partners but the investigation of that is beyond the scope of this book.
Topic Shifts The last marker of intimacy mentioned in the literature that was taken as an ingroup marker in this study is related to topic shifts, inspired by Tannen's (1984) "high involvement style" features of abrupt topic shifts, the introduction of new topics without hesitation, and persistence of topic. In order to analyse topic shifts, the notion of topic had to be defined. There have been several approaches to the definition of topic: the formal, the surface cohesion model, the semantic, the pragmatic and the interactive (McCarthy, 1991). The definition of topic used in the present analysis is: "What speakers say about something". A topic is therefore not "what is being talked about " but "what is being said about what is being talked about". It is not, for example, "Edinburgh buses" but "What happens on Edinburgh buses"; it is not "BF's husband" but "Why BF's husband did not apply for the PhD" A topic can be given a title that is an embedded question, as in "Why X happened" or "How to do Y." The inclusion of the WHquestion word in every topic title helps to define the boundaries between topics. In the sense that the topic can be given a title, this definition is principally a semantic one. This method of marking topic boundaries may be less precise than that of using shift markers as the formal approach does, in which stretches of talk are "bounded by certain topic and/or transactional markers" (McCarthy 1991: 132) such as a fall in intonation, pauses or fillers, mitigation, selfreferential metastatement and metareference (Stubbs 1983). The method used avoids the danger of becoming too mechanical and mistakenly saying that there is a topic shift every time there is a shift marker or that there is no topic shift if none of the markers can be found. Using lexical chains to define topic boundaries can be equally dangerous because they stretch across from one topic to the next and some topics contain very little lexical cohesion. That is to say, the approach used in this study is not one of surface cohesion either. The approach to topic definition is interactive and pragmatic. As Wardhaugh (1985: 139) explains, "Something is only a topic if more than one speaker makes an utterance relevant to it". Brown and Yule (1983: 8990) say that topics are negotiated and that speakers speaking topically try to make their contributions compatible with what they think the other participants are talking about. For a topic to be a topic, more than one speaker has to make a relevant contribution and "speak topically" even though the contribution is a backchannel while the other speaker holds the floor with his "speaker's topic". The topic can thus be defined as "strings of utterances perceived as relevant to one another by participants in talk" (McCarthy, 1991: 132). Wardhaugh (1985: 139) says that the comments that participants make "cluster", that "the focus of the cluster is a topic". Sperber and Wilson (1986: 160) say that in casual conversation, a modicum of relevance should be enough but that the speaker's words must be "relevant enough to be worth the addressee's attention". When a speaker speaks relevantly, he modifies and
28 Analysing the language of discourse communities extends the mutual cognitive environment, the topic framework and the cotext whose assumptions both speaker and hearer share, thus achieving successful communication. The pragmatic approach to topic clearly requires a definition of relevance. Grice (1975) went no further than describing his cooperative Maxim of Relation as "be relevant". Leech (1983) redefined this with his Maxim of Relevance, saying that, "An utterance U is relevant to a speech situation if U can be interpreted as contributing to the conversational goal(s)" of the speaker or hearer (p.94). These goals can be both social and personal. Brown and Yule's (1983) reaction to Grice's simple instruction "Be relevant" is to ask "relevant to what?" They redefine it as speaking topically on the conversational or shared topic, making one's contribution fit closely to the topic framework (the activated features of knowledge of context, and the preceding cotext or domain of discourse). Speakers may also be speaking on a topic, in which case their contributions may be relevant to one particular entity being discussed, expressing a personal topic, but not necessarily relevant to the existing topic framework. The interest in topics in the present study centres specifically on topic shifts. A distinction was made between 'shift to a new main topic' and 'shift to a subtopic with in the existing main topic'. (19) is an example of a main topic shift, showing little or no relevance to the preceding topic. Units 06107 to 06110 come at the end of a general knowledge topic about "how to get a computer through customs", and 06111 begins a topic related to the course, about "how they got on with their tutorial task" (19) 06107 AF And yet there's no reason why you shouldn't take it out. 06108 AM Yeah 06109 AM There is because she might sell it there you see. 06110 AM There's this there's there's this tax on luxury items. » 06111 CM Yeah let's do this. 06112 CM We've only got like twenty minutes. A 'subtopic shift' occurs within the hierarchical structure of a main topic. It can have relevance to the preceding subtopic, or it can be the resumption of a previous topic after a short interruption or insert sequence. (20) demonstrates a conceptually and lexically linked drift. Before unit 18123, CM was saying who his favourite golfer was and why. This drifts easily into the subject of whether CM himself plays golf. (20) 18120 CM And he had a slump for about six years and he's just in the last year he's come back. 18121 CM It's like now there's three in a row and he's in his thirties so he's pretty young. (4) 18122 CM So. » 18123 BM Do you play golf? 18124 CM Yes. For the coding process, the first two discourse units of the shifts were labelled because it was felt that if only the first unit were labelled, this could cause the first utterance stating the new topic to be lost, since the first unit is sometimes only a hesitation, a turnholder: (21) 15082 DM So I'll just do that today. 15083 DM And try to get a few out. ((15)) » 15084 CM Oh that's... (1.5) > 15085 CM My budgie pooped on my book. ((0.5))
The ingroup 29 or a shift marker of some sort: (22) 18011 CM MSc? 18012 BM Something like that you know. 18013 BM He's so = 18014 CM Medium level in shit and crap. 18015 BM (hehhehheh)((4)) » 18016 CM It's true though. —» 18017 CM The more you study the more you realise that there are no answers. The first recorded line of dialogues was not labelled because there is no way of proving whether these units are at topic shifts, as the cassette recorder was put down in front of the recordees when the conversation was already under way. The first recorded line of a dialogue is most likely to be a continuation of an old topic, but it could also be a restatement of the old topic in order to put the researcher in the picture, or a new topic started because recordees do not want their old topic to be taped. The percentage of occurrences of discourse units with topic shifts out of all discourse units was calculated for each K area. An average for each term was found. Analysis showed that the percentage of main topic shifts remains stable while that of subtopic shifts decreases, with increased knowledge over time (see Figure 3). By spring term, therefore, once a main topic has been established, any subtopic within in it or the topic itself is likely to last longer. The persistence of topic suggests that by spring term, the students are "highly involved"; the group can be said at least to be gelling by spring term. Thus, by the summer term, group members can shift drastically from one main topic to another, as in (23) which contains an abrupt shift from a general knowledge topic about 'How people see Ferguson now that she has been filmed having her toes licked by her financial advisor', to an interpersonal but unrelated topic, beginning on unit 24028, about 'How BF and DM could get together'. (23) 24024 MM Not a royal any more. ((2)) 24025 BF Was she ever? 24026 BF Was she ever really I mean? 24027 DM She wasn't very real was she? » 24028 BF I was going to say you have to come round. 24029 DM That'd be nice. When subtopic shifts occur in the summer term, they are often resumptions of previous topics, as in (24). Before unit 29051, the students were discussing the first of two ideas that AF had for a project. This topic is interrupted by the insert sequence about whether NF likes the class that she is going to. At 29057, DM prompts AF to return to the topic of her project and to tell him the second idea: (24) 29048 AF I'm just really interested in the whole sort of business of // language and formation 29049 DM // Yes. 29050 DM It is very interesting. » 29051 NF I'd better go to my class. ((2)) 29052 AF You did this last week didn't you? // (heh) 29053 NF // Yeah. 29054 NF I really like the teacher very much. 29055 AF (heh heh) 29056 DM It's very relaxing. (1)
30 Analysing the language of discourse communities » 29057 DM 29058 AF
And the other one? Oh the other possibility I suppose is language attitudes. (3)
Figure 3: Percentage of DUs with topic shifts out of all DUs
2.4 THE RECORDEES The triangulation was carried out on 29 May 1992, at the end of the MSc course. The six recordees had to be interviewed before they left; one was to go to England, one to the Sudan, two to Japan, and one to the Caribbean. This meant that they were interviewed before all the dialogues had been transcribed, and before the coding system had been devised or the data analysed. Thus the questions that they were asked were not ones that sought confirmation of conclusions formed from analysis; they were ones that aimed to find how realistic the researcher's first impressions were and how relevant her concerns. The researcher invited the recordees to dinner. This allowed her to thank them for their cooperation, and to observe them interacting during the free overdinnertalk, as well as test them. Triangulation therefore consisted of the overdinner observations, analysis of a written questionnaire after dinner (see Appendix I) and observations of the unguided discussion that followed on naturally from the questionnaire. The questionnaire, on some of the very dialogues in which they had participated, required the recordees to say what they thought the function and form of their talk was in the common room; their answers to these specific questions will be mentioned in the appropriate chapter in this book. The free discussion revealed that they constituted an open group of four "special friends" (AF, BF, BM and DM) plus two "ingroup course members" (AM and CM) who had chosen, for whatever reason, to interact with the other four. The six recordees had maintained their interaction with each other throughout the year despite the fact that AM had taken different option courses from the other five, and AF had only done one option course, being a parttimer. They had, in fact, done "friendly things in friendly places" (Duck, 1991); they had each gone out with at least one member of the six (AM and CM included) on several occasions for a drink. As a result of their interaction, the recordees were close enough to share a great deal of interpersonal knowledge about each other: they knew about each other's spouses, problems at home, and
Theingroup 31 arrangements for future employment. The presence of the ingroup markers of informality, humour and topic length showed that they cohered as a group. They seemed to share a sense of humour, and showed solidarity over dinner by making each other laugh and responding with laughter to each other's funny comments. They did say that they felt that slang, and obscenities were reserved for special friends, yet the analysis of the data described in this chapter shows that slang, obscenities and emotional language were used by and with all six of the recordees. They said that they did not consider themselves to be an exclusive group since they had spent as much time with other colleagues as with each other. On average, they each considered about eight different students out of the whole course as "special friends", but of these only three were from the group of six recordees. Thus, their language can be taken as representative of the language used by all NS students on the course. Duck's (1991) friendship parameter of commitment, and "maintaining the relationship" did not seem to apply to any students. As far as the MSc group dynamics is concerned, the recordees said that they had entered the course knowing that they would be with their colleagues only for a year: only AM and BM said that they would like to keep in touch with each other after the course. They said that they had felt unable to develop any friendships very far because of separating off into option courses; it had been difficult to associate with many of the people whom they had met in the first term consistently all through the year, therefore. Once the data was transcribed and the analysis under way, the distribution of discourse units spoken by each of the six recordees in each of the three terms was calculated, in order to test whether the results of the linguistic and functional analysis could be said to be biased by the presence of one or two particular recordees. Figure 4 shows that the distribution each term is approximately the same. It will be noted that, throughout the course, the ratio of the total number of units of the women speaking (AF and BF) to the units of the men speaking is 2:3; this tallies with findings from gender studies about male conversation and topic dominance. This study did not enter the immense field of gender differences, however, preferring to limit itself to exploring each recordee's characteristics in terms of language features and functions of talk. A discussion the individual characteristics occurs in Chapters 3 to 7. The six recordees' individualities do not cause particular tendencies to develop in the data that could make the findings not generalisable to any group of native speakers of English.
2.5 CONCLUSION The findings in this study can be taken as typical of all six recordees, and the six are typical of the whole student group. The conversations recorded in each term contain a similar distribution of each of the six. Comments by the six recordees, during triangulation, suggest that the interaction within this group of six is representative of interaction between the NS and all other members of the MSc course. The study of intimacy markers shows that the group begins to gel in autumn term, because even then students indulge in irony and banter. It has indeed gelled by spring term: that is when informality, in the form of initial ellipsis, slang and expletives, and drastic and more persistent topic shifts increase. It confirms, therefore, that the linguistic features to be analysed in this study are indeed part of the language of a group.
32 Analysing the language of discourse communities Figure 4: Proportion of discourse units spoken by each of the six recordees Autumn term
3 KNOWLEDGE AREAS
3.1 INTRODUCTION All conversations are based on an assumption of common sociocultural background knowledge. The anthropological linguist, Malinowski (1923) saw the sociocultural context of utterances as central to meaning and action. As Clark says, "The participants in a conversation work together against a background of shared information" (1993: 4). This chapter reviews the literature on the relationship between background knowledge and levels of intimacy, and between topics and language. It then describes the part of the study that examines the way that the background knowledge assumed changes with interaction over time. This book is based on the assumption that accumulated interaction and knowledge as well as the topic itself affect the way that the language used.
The Literature On Knowledge And Context Linguists have long attempted to define the context of background knowledge; the debate continues. Clark observes that mutual knowledge is based on community membership (1993: 37). Carlson (1993: 60) says that the source of common ground is linguistic copresence and physical copresence as well as community membership. Goodwin and Duranti (1992: 10) list the dimensions of context: setting (the social and spatial framework), behavioural environment (the
34 Analysing the language of discourse communities way speakers use their bodies and behaviour to organise the space), language as context (contextualisation cues how talk provides context for other talk) and extrasituational context (how understanding of an exchange requires background knowledge beyond local talk and immediate setting). They lament "the very small amount of research that has focused explicitly on the organisation of context". This book deals with the organisation of language as context and the extrasituational context of the community, and explores the relation between the two. Chapter 2 described the setting. This chapter focuses explicitly on the organisation of the extrasituational context. Linguists have discussed the extent to which knowledge of the extrasituational context is relevant in conversation. Sperber and Wilson (1986: 64) say that if interlocutors "establish that they belong to the same community or group, they can reasonably assume mutual knowledge of all propositions normally known by group members." Few would dispute this, but they go on to say that interlocutors interpret knowledge that is intended and manifest without assuming wider knowledge and beliefs. Clark's (1993) mutual knowledge hypothesis on the other hand, posits that knowledge of context and speaker are not always sufficient for successful comprehension, and that interlocutors do use wider knowledge and beliefs that they assume they mutually share in order to make conversational inferences and interpret utterances. This chapter examines both common knowledge known by all group members and shared interpersonal knowledge of individual members, and it describes the difference between the two. The extrasituational context of conversation in institutions has been studied; the way the institutional context and conversation change over time has not. Drew and Heritage stress the role of conversation in institutions: "talkininteraction is the principal means through which lay persons pursue various practical goals and the central medium through which the daily working activities of many professionals and organizational representatives are conducted." (1992: 3) Ochs et al. (1996: 37) have investigated "ways in which talk and interaction both organize and are organized by institutions, relationships and culturally specified environments", and ways of "relating strategies for engaging in verbal interaction to the socialisation, maintenance and transformation of social realities such as family, the school, work or community political structures". Firth (1994), analysing the discourse of the workplace, says that it is the individual's competent social membership accomplished interactionally through locally ordered discourse practices, that is of interest. Boden (1994: 8499) notes that "it is through corridor chat, quick lunches, and hanging out in office doorways that the essential flux of information and mood is conveyed" and that the "conditions created in and through interaction thus have consequences far beyond the immediate interactional setting, embedding actors and activities in a very real world of their own making". Bell (in Firth 1994: 54) says, "Each workplace has in some respects a distinctive "normative order", and /.../ this will affect the nature of conflict, constrain the type of cooperation, and have an impact on the process of negotiation itself1. This chapter examines the extent to which the MSc common room dictates a norm as concerns the knowledge areas that must be referred to in discussion. Those linguists, sociologists and psychologists who have commented on the fact that knowledge changes with interaction over time and that language must be analysed in this changing context, have not gone further than a superficial analysis of this phenomenon. Heritage says that conversation analysis studies the utterance embedded in a context: "A speaker's action is context
Knowledge areas 35 shaped in that its contribution to an ongoing sequence of actions cannot adequately be understood except by reference to the context including, especially, the immediately preceding configuration of actions in which it participates." (1984: 242). It also shapes a new context for the action that will follow. Coulthard (1977) refers to knowledge as part of an ongoing process: "Common ground is not restricted to shared experiences of a particular linguistic interaction up to the moment of utterance; rather it is a product of the interpenetrating biographies of the participants, of which common involvement in a particular ongoing interaction constitutes only a part." (p. 106) but he does not study the development of "interpenetrating biographies". Sigman (1983: 1812) states that the analyst of discourse "must be able to make reference to conversations engaged in over time, that is to discourse embedded in a continuous social relationship" in order to analyse conversations, since "the significance of any one interactional engagement is regulated by the larger ongoing social process", but he does not analyse exactly what knowledge is involved in the "continuous social relationship" and how it changes. Clark says, "As the discourse proceeds, the participants accumulate shared information by adding to it with each utterance" (1993: 437). He observes that mutual knowledge based on community membership is generally preserved over long periods of time and it is constantly being renewed. Schiflrin (1994: 360) observes that "the function of an utterance (and hence its identity) must be defined in relation to (and as appropriate to) a context that is not static, but dynamic, and even more critically, a context that is still in the process of being interactively formed". The accumulation of background knowledge has been approached from the point of view of information structure. As Cook says, "communication might be defined as the conversion of new information into given information" (1989: 64). Once a speaker is sure that his hearer has the necessary background knowledge for what he is going to say, he can introduce something new, and then this new information becomes part of the background knowledge for the next utterance and all that follows. The accumulation of given, and new converted into given, builds up to form a presupposition pool (Venneman, 1975). The assumed knowledge may be "given" and "in the consciousness of the addressee at the time of utterance" (Chafe, 1976), or "known" but "unused" (Prince, 1981), in the sense that it was not in the hearer's consciousness. This book does not examine language from the point of view of information structure, or theme and rheme; it examines it in terms of endophoric reference, or textually given, and exophoric reference, or textually new (see Chapter 4). Change in common knowledge has been dealt with in terms of topics discussed at different levels of intimacy, but what abound again are descriptions of product, or knowledge at a given time, rather than ones of process, or how knowledge changes from one time to another. Sacks (1992) examined the conversations of strangers and found that in "gettingtoknowyou" conversations, speakers seek knowledge of their interlocutors' membership (eg: sex, age, race), and then class them as representative of that category (eg: "She's 48" or "He's a Negro"). Wardhaugh states: "Opening up a conversation with a complete stranger is obviously a somewhat risky endeavour there are so many unknown quantities." (1985: 118). He explains that if speakers are virtual strangers, unsure of each others' background, they must proceed cautiously and attempt to find some common ground on which to manoeuvre. Brown and Levinson say that strangers' common ground may be reduced to "an assumption of common interest in good weather or other such safe topic", whereas for close friends "it may extend to close identity of interests and desires" (1978: 64). "The more the speaker knows about the hearer, the more close to home will be the safe
36 Analysing the language of discourse communities topics he can pursue." (ibid: 112). Planalp and Benson (1992) found "mutual knowledge" to be the most common parameter that subjects used in order to distinguish between friends' (defined people known for at least a year) and acquaintances' (defined as people not talked with more than once) conversations. This book explores how the type of knowledge assumed affects the language used to refer to it. The effect of the topic area on the language used and way of talking has been examined in the literature, but the studies offer a specific description of only one particular aspect. Giles and Powesland (1975: 122) have found that "the topic of conversation in social interaction can be influential in determining speech modifications when the subject matter is high on one or more of the following dimensions; salience, emotionality, technicality, abstraction and humorousness". The dimension of salience was examined by Matarazzo (1970), who found that students increased their mean utterance duration when discussing education or their studies. Kanfer (1960) investigated the aspect of emotionality and discovered that anxiety topics affected his subjects' speech rate and accent. The dimension of technicality was researched by Moscovici (1967) who found that a car specialist discussing cars with another specialist used a greater variety of words and more technical terms than with a nonspecialist. Ratner and Rice (1963) found that speakers talking to poorly informed listeners on a technical topic used more words, repetitions and complete descriptions than they did with wellinformed listeners. Familiarity with topic was investigated by GoldmanEisler (1968) who showed that if a speaker was more familiar with a topic, his speech contained fewer pauses and a quicker articulation rate. Work on the dimension of abstraction was carried out by Lawton (1965) who found that the more abstract the topic, the more complex the grammatical structure. Exploring the dimension of humorousness, Giles (1977) found that with a humorous topic, speakers were less hesitant, and used a non standard accent, less precise enunciation and varied tempo and pitch more than with a non humorous topic. This book includes a more complete description of the relationship between topic and language: the model contains the grammatical and lexical elements mentioned here, adds interactional and functional features, and shows how they all relate to each other.
The Categorisation Of Knowledge Linguists have suggested various categories of knowledge. Knowledge has been found to have three types of status: private, common or shared interpersonal knowledge (Kreckel, 1981). This book examines only the common and the shared interpersonal knowledge. Private knowledge is that which interlocutors have about themselves and which they are most likely right in assuming that few others know. This cannot be analysed as background knowledge but only observed, if it is revealed, as new knowledge which will consequently become shared interpersonal knowledge. Common knowledge is "knowledge acquired separately" (Kreckel, 1981), knowledge of the world, of which every speaker has different amounts. As Wardhaugh (1985: 18) rightly says, "in any conversation the participants will have different kinds of knowledge about almost any topic that is likely to be mentioned". Some of the knowledge that the students bring to the course is common to them, since they are linguists and/or language teachers, and then once they have started the course, they all gain background knowledge of the course itself, common to them all. Shared interpersonal knowledge is "knowledge acquired in mutual interaction" (Kreckel, 1981), acquired through previous verbal interactions or joint activities and experiences, and including privileged knowledge about the interlocutor. According to Kreckel, it is1 the speakers' knowledge
Knowledge areas 37 of interactions in the past and the perspective of continued interactions in the future, that leads to group cohesion. Shared interpersonal knowledge about the interlocutor has been discovered to have three levels (Berger and Bradac, 1982). The lowest level is the descriptive level: knowledge about physical details and past history of the person. This is what Planalp and Benson (1992: 497) call mutual knowledge or "basic demographic information about each other" and "each other's habits and dispositions", "each other's activities, schedules and plans" and "people, events or places". The next level is the predictive: knowledge about what a person's beliefs and attitudes are that allow the interlocutor to predict how he would react in a given situation. The highest level is the explanatory level at which the interactor is able to explain why a person reacts the way he does. The analysis described in this book does not follow the Berger and Bradac model because such micro divisions of shared interpersonal knowledge were not considered necessary, given that knowledge itself is not the main focus of attention. Clark (1993: 37) differentiates between "generic" knowledge, or knowledge about kinds of things, and "particular" knowledge, or knowledge about individual or particular things. The study of knowledge described in this chapter is not based on the Graesser and Clark (1985: 301) model of "Generic Knowledge Structures" (GKS), or world knowledge "housed in semi autonomous packages", which they say, "provide the background world knowledge that is needed for constructing bridging inferences (those which fill gaps between explicit statements in order to establish conceptual connectivity) and projection inferences (those which elaborate and expand a coherent passage structure but do not fill gaps), and consist of animate beings (eg: manager, sister, dragon), inanimate concrete entities (tree, palace, new York), abstract concepts (eg: goodness, happiness, noise), causeoriented event structures (eg: seeing, fearing, rushing), and goaloriented activities (eg: giving, fighting, eating). The study of particular entities, concepts and events is described in Chapter 5, which is about the ingroup lexis. The researcher, observing the students in the common room in the spring and summer terms in the MSc course, had the impression that, as time passed, coursemembers seemed to talk more and more about the course to the exclusion of all else. She was interested in discovering if in fact over time, topics drawing on knowledge of the MSc course did increase. It may be that when the pressure of work, in the form of exams and projects, increases, as it does in the spring and summer terms, the ingroup members need to show solidarity with and seek support from each other by talking more about the situation and events that unite them. She also felt that over time, sections of conversations assuming personal knowledge of interlocutors increase, and wanted to discover whether this was also the case. Knowledge can be seen as an indivisible whole; it can also be divided up and categorised in as many different ways as there are analysts to divide it up. The construction of the common knowledge area (K area) model is necessarily an arbitrary, subjective and artificial way of dividing up the complexity of "reallife" casual conversations; the K area categories used are dictated by the researcher's vision of the data. It was, however, necessary to build an analysable model of knowledge, in order to test the hypotheses that topics based on MSc course knowledge and on interpersonal knowledge increase. Above all, a model was needed in order to analyse the relationship between K areas and language, to study the ingroup code and its function, and to observe changes in these elements within each K area over time. The noncourse K areas are
38 Analysing the language of discourse communities labelled Kl, K2, K3, and the course K area is labelled K4 (see Figure 1). Figure 2 is a list of examples of topics in the data in the K areas. Numbers identify dialogues. Figure 1: Knowledge areas Non course K areas Kl the world K2 language teaching and learning university study in general, computing as an aid, etc. K3 the University of Edinburgh (EU) the Department of Applied Linguistics (DAL); The Institute for Applied Language Studies (IALS), university buildings and what they contain, the physical here and now, scholarships and teaching, etc. The course K area K4 1991 92 EU Applied Linguistics MSc programme deadlines, specific tasks, specific study groups, particular books and articles, special ways of referring to courses, students, etc. Figure 2; Examples of topics in each K area Kl : The World World and TV 11 What happened in a TV serial. 19 What the sea pollution is like in the Mediterranean and Japan. Edinburgh and Scottish traditions 4 What happens on Edinburgh buses. 8 How BF spent her Hogmanay in Edinburgh. Speakers' homes and habits 15 What CM's budgie does in the livingroom. 21 How AF wastes time in the evening at home. Speakers' families 21 Why AF's social life is limited by her son and why he is growing so fast. 23 How DM's wife and BF's husband affect their financial situation. Speakers' trips, outings and entertainment 11 Why AM did not go for a meal after the pub; where BF went after the meal. 14 Why DM did not climb a mountain in Pitlochry. Meanings of words 7 What "thingamajig" means. 9 What the origin of AM'sname is. K2: Language Teaching and Learning Language teaching 1 Where AM, BM, and CM have taught English before coming to Edinburgh. 22 Why DM and BF had difficulty introducing innovations in language schools abroad. Studying at university 10 What the exam questions were like in CM and DM's undergraduate courses. 18 What the mnemonics for MSc and PhD are. Computers for study 9 How CM uses the tabling feature on Microsoft Word.
Knowledge areas 39 26 How BM can solve his layout problem by changing software. K3: Edinburgh University (EU) EU buildings and life 4 Whether BF's house is convenient for the department and King's Buildings. 28 Why BF's husband did not apply for the PhD IALS/DAL here and now 3 Whether AM has time to go to the DAL common room and get a coffee. 15 Why BM and DM are shutting out the sunshine, in DAL common room, and where the key to the DAL photocopier is. Doing a PhD in DAL / working for IALS 23 Why BF will not do a PhD in AL, and what happened to DM's application for IALS summer teaching. 28 Why BF is not interested in the IALS scholarship, whether she should apply, and what CM did about his proposal and interview for it. K4 : The 19911992 MSc Core and option courses 13 Why understanding syntax is easier than it seemed in the core course. 15 What AM and DM think about the Psycholinguistics courses, and what options BM and DM are doing. Lectures and tutorials 4 How much BF and BM read for the tutorial, whether they completed the task, and why BF did not copy down the examples in the lecture. 27 How many lectures CM is going to miss, whether DM could get him notes, and whether some lectures could be missed. Exam and portfolio 10 What subjects CM and DM are studying for the exam, why, what CM was doing in the library, and how to study for the Linguistics question. 13 When and how long the portfolio is. Projects and dissertation 15 How far on CM and DM are with their projects, whether CM has filled in DM's project questionnaire, and what DM's tutor said about his project. 21 What DM should write his dissertation on. Books and articles 15 What CM and DM think about certain articles. 18 What reading shows about changing fashions in linguistics and teaching approaches. In the coding of the data, the K area was established for the duration of a whole subtopic, wherever possible, and all discourse units within that subtopic were coded in the same way. That is to say, each discourse unit was not analysed individually and coded one by one. If there was only one discourse unit that drew from a K area that was different from the K area of the preceding and following discourse units, it was not indicated as different. In example (1), students are preparing to discuss a tutorial task sheet (K4). BM refers to his "free time" presumably at home, in the world outside the course (Kl) to explain the present situation in the course. This one Kl unit (03054) is not coded Kl but K4, because of the surrounding units: (1) 03052 BM I wrote some some lines here. ((!)) 03053 NM That's fairly lengthy.
40 Analysing the language of discourse communities » 03054 BM Ah but well I had I had a lot of free time! (2) 03055 BM Urn. (3) 03056 BM Oh yeah. (2) 03057 BM Where's my pen? If an exchange appeared to draw from more than one K area at a time, the predominant one was selected. Excerpt (2) illustrates this: BM is typing up his project, and the excerpt was coded as K2 and not K4 because the discussion focuses principally on the technical computer aspect of his problem: (2) 26032 CM Oh you want a table basically. (2) 26033 CM With lines. 26034 CM On one side of the table you have teacher's interaction and on the other side you have a description // of it 26035 BM // I want a table that has. 26036 CM You want a table. 26037 CM Can't get those on Works. 26038 CM You'd have to use Word for that. The four K areas relate to each other as concentric circles (see Figure 3). Bell (in Firth 1994: 54) says that "In studying the extrainstitutional context, we begin with the recognition that each workplace is situated in a wider societal context. It is surrounded, in other words, by a series of larger settings within which it is 'nested'". K4 is nested in, and is part of, K3, which is in turn included in K2, which is included in Kl. To put it another way, the course itself (K4) takes place in the context of DAL and IALS (K3); DAL and IALS are part of the language teaching and university study world (K2); language teaching and university life are but a small part of the whole world (Kl). Figure 3: K area concentric circles
This relationship in concentric circles can be demonstrated by looking at each of the K areas and their distance from K4, knowledge of the course. Starting with K3 topics, these are closest to K4
Knowledge areas 41 because K3 is knowledge about the university probably acquired since the students joined the MSc course. Even within K3 some topics are closer to K4 than others. The category of the "Here and now of IALS/DAL" topics is closer than that of EU since it requires knowledge of facilities and characteristics of the DAL MSc common room itself. The category of "Doing a PhD in AL and Working in IALS" is the closest since topics assume knowledge of IALS staff and the co ordination between DAL and IALS to organise the scholarship selection process, and they assume knowledge of MSc option course subjects that can be developed to make an MLitt study. This cannot be included within K4 because it is not about the course itself, but about an extension of the course after it finishes. Moving on now to K2 topics, these are not as close to K4 as K3 ones because K2 contains knowledge that students had before the course: their past experiences of "Language Teaching" such as introducing innovations in English schools abroad, their memories of "Studying at University" such as answering exam questions in previous university courses, and general knowledge about using "Computers for Study" such as the best software for text processing. K2 is closer to K4 than Kl, however, because it is knowledge of those parts of the outside world that are directly related to theory acquired in the course about linguistics and language teaching, and related to course components and activities such as sitting exams and writing projects. Kl is the furthest from the course, although again some categories are closer to K4 than others. Topics in the category of the "World and TV" such as current events and world pollution, and topics in the "Edinburgh and Scottish Traditions" category such as what Edinburgh buses are like and how the Scots celebrate Hogmanay have very little to do with the MSc course. Topics in the categories "Speakers' Homes and Habits", their "Families" and their "Trips, Outings and Entertainment" are closer to the course, in that they are about outofcourse situations that centre round course members, such as where they went with their spouses and children, and what happened during sports activities and outings with other course members. The problem of how to visualise the interaction between the four K areas using concentric circles raised certain theoretical questions, although it did not actually affect either the coding or the data analysis in any way. The circles could have been reversed, with K4 on the outside and Kl on the inside. It was tempting to reverse them, given that all the participants of the dialogues are the MSc students and not people outside the course. This way, the classification would have been in terms of an implication scale, starting from the perspective of the MSc student and looking inwards towards the knowledge of the world and seeing everything outside the course in terms of how it related to the course. The reversed model was rejected for a series of reasons. Firstly, it does not take into account the fact that, although most of the time students look at the world as it relates to the course, the students' view of the world, their focus, does change over the three terms of the course. In the autumn term, students focus out on K2, their past experience of language teaching and how it is relevant to the course. In the spring term the focus is most of the time on K4, the course itself, and speakers see little else than the present. In the summer term they again focus out but this time to the future and K3, whether to continue to a higher degree after the course and whether to work in the Institute. The reversed model resists this change of focus, treating all subjects as emerging through the focus on K4. The second reason for rejecting the reversed model was that the concentric circles as they stand in Figure 3 do demonstrate the fact that most conversations
42 Analysing the language of discourse communities depend to a certain extent on K4, or ingroup knowledge. Each concentric circle is one stage away from the ingroupness, showing that ingroup knowledge is gradable. The model used also accommodates the notion of inaccessibility from noncourse members. The outsiders or "overhearers", who do not share common ground, (Schober in Clark, 1993; Clark, 1997) are in the different circles outside the course. The quality of outsiderness is gradable. Each K area was examined separately in these concentric circles in order to reflect the gradability of courserelatedness, but they were also grouped in the two macrocategories, noncourse K area and course K area (see Figure 1), in order to make a broad statement about K4. If the only analysis had been of only the four K areas in isolation, the picture might not have been so clear. The question of where to put the dividing line between noncourse and course topics was not a simple one, because everything is courserelated in some way. The line could have come between Kl and K2/K3/K4; between K1/K2 and K3/K4; or between K1/K2/K3 and K4. The K1/K2/K3 K4 divide was used because it was considered that a topic is either about the course itself (projects, classes, books, etc.) or it is not about the course itself. This division is most suitable for testing the hypothesis that topics based on MSc course knowledge increases, because it isolates course knowledge more clearly than the models that have a gradability of courserelatedness within the category of course knowledge itself. Shared interpersonal knowledge was easier to define, and it was not broken own into mini categories. Whereas common knowledge is that public knowledge that one would expect most people to have about Edinburgh, language teaching or the course, shared knowledge is the privileged, interpersonal knowledge about the interlocutor, that which speakers would not expect most people on the course to know knowledge about the interlocutor's home and family setup, outofcourse activities and particular incourse activities. Shared knowledge is the descriptive, predictive and explanatory interpersonal knowledge. Figure 4 lists examples of topics. Sections of dialogue based on shared knowledge occurred throughout the data in all K areas. This category can be incorporated into the concentric circles theory and diagram, if shared knowledge is represented as lines radiating out from the centre hike spokes of a wheel, to indicate that it is all pervasive, running through the K areas. Figure 4: Examples of topics in shared knowledge Home and family (descriptive): Where they live/have lived/have visited. What their immediate family consists of and what their names are. Activities (descriptive): What their particular past, present and future activities are outside the course both social and workrelated. What their particular past, present and future activities are within the course options, tutorial groups, tutors, projects, books, and study progress. Personality and attitudes (predictive and explanatory): What their personalities are like, how they usually behave and why. What their attitude towards and feelings about each other, certain aspects and components of the course, the world outside are. What their aspirations and objectives are.
Knowledge areas 43 The one area of interpersonal knowledge missing from the recordings is that of extremely intimate personal details revealed in selfdisclosure. This may be partly explained by the fact that recordees monitor what they talk about because of it going down on permanent record, it is also partly explained by the fact that, even without the cassetterecorder, self disclosure is rarely DAL "common room talk" because the room itself dictates more public topics. Self disclosure is more likely to be "pub talk" or "coffee at someone's flat" talk. Shared interpersonal knowledge differs from K4, in that whereas K4 contains information that almost any MSc member could have about the course, shared interpersonal contains information that only a limited number of members could have about all matters, including the course. To take an example, excerpt (3) was from K4, and not coded as shared: the speakers appear to be referring to a misprint in one of their text books: (3) 13001 AM It's a real text book. It's not like Brown and Miller. 13002 AM 13003 CM Figure thirteen. 13004 CM See figure thirteen. 13005 AM (hehhehheh) 13006 CM Figure twelve. Fifigure fourteen. 13007 CM 13008 AM // (hehehheh) 13009 CM // No figure thirteen. (7) It seems probable that any student could have understood this oblique reference to the problem, since the book was set for discussion in tutorials in a compulsory core course. The distinction between K4 and shared interpersonal knowledge was not made on the grounds of impenetrability: both could make a dialogue impenetrable to an outsider. Shared interpersonal knowledge is usually exclusive to a small group within the MSc group, and possibly exclusive to only the speaker and listener, but K4 can exclude all those who are not ingroup members. Once all the data had been categorised and coded, the analysis went through four stages. A Kappa intercoder reliability test (the number of right answers minus the estimated score divided by total number of questions minus the estimated score) was carried out with two coders and produced a result of 46% for one coder and 75% for the other. The second was satisfactory, and discussion with the first coder revealed that he had made errors of interpretation of the code because he had not read the descriptions of each K area closely enough. Then, the percentage of discourse units in each K area was calculated in all the data, in order to discover which area occupied most space in the whole course. Next, the percentage of discourse units in each K area was calculated in each term, in order to find the changes over time in each K area. Finally, to investigate whether shared interpersonal knowledge increases over time, the percentage of shared interpersonal knowledge discourse units in each K area was found in each term. Shared interpersonal knowledge in noncourse K areas was compared with that of K4.
44 Analysing the language of discourse communities
3.2 CHANGES IN KNOWLEDGE The Common K Areas Throughout the whole course, most conversations centre around the course itself; that is to say K4 occupies most discourse units (see Figure 5). The three noncourse K areas may not have been the same size by definition, in the first place. Thus the proportions Kl, K2 and K3 might have been different if they had been defined differently. However, since K4 is defined in such a watertight way as knowledge of components of the course, the proportion of 42% seen here could not be different. Figure 5: Proportion of each K area in the whole course
Figure 6: Proportion of each K area per term
The main topic type of conversations seems to be determined by the term in which it occurs (see Figure 6). The autumn and summer terms are similar in that noncourse K areas predominate; the spring term is the one with the highest proportion of time spent talking on K4 topics. In the autumn term, K2 is more frequent; in the summer term all K areas occur but K3 topics predominate. The hypothesis that topics based on MSc course knowledge increases was not therefore confirmed. It would seem that the cause of change in knowledge assumed is not so
Knowledge areas 45 much the fact of interacting over time but quite simply the change of concern and focus of attention, reflecting the stage in the course and events on the timetable. In the autumn term, the students are more likely to discuss the past than the present, Kl being the knowledge of the world that they bring with them from the past and K2 being the experience of previous teaching and study that they have had. The students know that keeping to safe topics, such as amusing narratives and anecdotes and personal history about where they live and where they have taught, will observe nonfacethreatening politeness principles. The K4 topics are practical organisational ones about "who is in whose tutorial group" and "how the tutorial task was divided up." The conversations consist of presentations of self and exploration of others. In the spring term, the students come back from their Christmas recess discussing their holidays (Kl) but stressed and eager to talk about the course itself, K4, about the present, to compare notes and seek solidarity. K4 occupies a whole 57% of conversations, while students check how their colleagues' revision is going and later how they answered their exam questions, and then later how much reading they have done for projects and tutorials. Their study and growing confidence with each other allows them to exchange evaluative opinions about articles and the courses themselves. The only other K area in the spring term that features with any frequency is Kl. Once the exam has passed, students can discuss TV serials, the news and world situations, and offer details about their homes and families, such as their budgie's poop and their weekends in with friends, etc. without seeming facethreatening. In addition they discuss evenings out that they have had together. In the summer term, the noncourse K areas occupy more time than K4; K4 is less than it was in the spring term. Students feel free to check over each other's projects and comment, and they are prone to talk about the future and give each other advice about a suitable dissertation topic. K3, practically nonexistent in the autumn and spring terms, is now the biggest noncourserelated K area, occupying 33% of conversations as students wonder where they are going next, whether they will take the IALS MLitt scholarship or do a PhD. K2 is bigger than Kl: students give advice about computer software for projects, the K2 topics that are closest to K4. Kl topics, smaller here than in the autumn and spring terms, centre again on personal details about speakers' families, such as their wife's swollen ankles and their son's passion for chips, and plans for course member social activities.
Shared Interpersonal Knowledge Throughout the course, there is a marked increase in shared interpersonal knowledge assumed, taking all K areas together. In the autumn term, the percentage of discourse units (DUs) with shared interpersonal knowledge out of all DUs is 1%; in the spring term it is 7%, and in the summer term it is 8%. Figure 7 contains the raw data of shared interpersonal knowledge for each term. The value of %2 was 41.251, significant at the 0.005 level, and shows that, taking all four K areas together, the difference between the proportions of DUs containing an assumption of shared interpersonal knowledge and those not containing one in each of the three terms is significant. This confirms the hypothesis that topics based on interpersonal knowledge increase. The more experiences they share, and the more the students know of each other, the more students can and do refer to this.
46 Analysing the language of discourse communities Figure 7: Frequencies of DUs with shared interpersonal knowledge and those without in each of the three terms, taking all K areas together. Term Autumn Spring Summe DUs with yes shared interpersonal no knowledge
5 735
111 1,674
77 1,005
The density of shared interpersonal knowledge DUs depends not only on how much interaction there has been over time; it also varies according to what is being talked about (see Figure 8). Kl has the highest density of shared interpersonal knowledge throughout the course. This is not very surprising: conversations about speakers' homes and habits, families, trips, outings and entertainment are bound to contain personal references the participants. Kl shared interpersonal knowledge increases dramatically, reaching double the average of the other areas by the summer term. K2 shared interpersonal knowledge sections are smallest probably because themes about methodology and technology tend to be impersonal. Shared knowledge in K4 dialogues is very much less than that in Kl dialogues and the density does not grow so dramatically, maybe because K4 is in itself more privileged by definition and because K4 is group rather than individual knowledge. Figure 8: Percentage of shared interpersonal knowledge out of each K area
A few examples will demonstrate the different types of shared interpersonal knowledge that occur in each K area. In Kl dialogues, most shared interpersonal sections are about home and family, and social outofcourse activities. Excerpt (4) shows knowledge of what the immediate family consists of. DM shows surprise that AF does not go out as much as he does at the weekend: (4) » 21014 AF Yes but you don't have to find a baby-sitter. 21015 DM Mm. 21016 DM Mm. —> 21017 AF And you've got somebody there to go out with straight away.
Knowledge areas 47 They both know that AF is a single parent and DM is a childless married man whose wife has come to stay with him in Edinburgh. Excerpt (5) demonstrates how speakers show a knowledge of each other's attitudes and can explain each others' behaviour. In the part of the conversation immediately preceding this, a student had been talking about a weekend climbing in Pitlochry. (5) 14021 DM More than I did this weekend I'm telling you. » 14022 BF You had friends didn't you? -> 14023 BF I can imagine why you wouldn't want to. 14024 DM Resting. 14025 BF Yeah. Sure, (heh heh) 14026 BF BF apparently knows who the friends are and why DM would lack the motivation to go out climbing; she also knows that he is joking when he says that he was resting. Most shared interpersonal knowledge sections in K3 relate to personality and attitudes or to activities outside the course. Excerpt (6) illustrates knowledge of past and present activities outside the course, yet related to the Institute. Both BF and DM know that they have both talked of applying for summer teaching with the Institute: (6) » 23051 BF Did you apply for work in IALS? » 23052 BF No. » 23053 DM I did in the end yeah. 23054 BF You did. (2) 23055 BF Have you heard? 23056 DM Yeah. 23057 DM I mean he said something like well... > 23058 DM You did as well didn't you? K4 shared interpersonal knowledge is predominantly related to activities within the course and to attitudes towards and feelings about it. Excerpt (7) shows knowledge of the interlocutor's attitudes towards components of the course. DM knows BF's personal preferences: he knows that she does not like doing subheadings on the contents page of her projects. (7) 22056 DM You don't go in for all these subheadings (0.5) one one = 22057 BF In the redraft // well I 22058 DM // Oh you do yeah (0.5) » 22059 DM I know you don't like them.
3.3 FURTHER DIMENSIONS The triangulation questionnaire and informal discussion with the six recordees themselves revealed a relatively low degree of awareness of the predominance of K4 topics or the changes from term to term. As they listened to themselves on the cassette recorder, they were amused at what they saw as the spinningout of conversation on what seemed to them, looking back, irrelevant and uninteresting topics. Although some of them said that they felt the common room did dictate chat about course topics, they insisted that it did not matter much what they actually said about them. AM claimed that in his talk with ingroup course members the topic was "of NO consequence" [his underlining and capitals]. In the questionnaire, most of the recordees said that they would not risk talking about the staff and other students (K4) with all ingroup course members, reserving such topics for special friends. The male recordees said that they never spoke to any of their colleagues about their personal life and feelings, and the female recordees said that
48 Analysing the language of discourse communities this topic was reserved for special friends. Analysis showed that talk about staff and other students took place with everyone. It also showed that, by spring term, all sk of them were mentioning aspects of their personal life and could not help showing feelings, attitudes and opinions in what they said to everyone. A brief analysis was made of the K areas assumed in each recordee's talk, in order to confirm that the findings described in this chapter were not skewed by individuals and their preferred topics. The percentage of each recordee's discourse units out of each K area total was calculated. Figure 9 shows that Kl and K4 topics are divided almost equally amongst all recordees, as are sections of dialogue depending on shared interpersonal knowledge. The fact that Kl and K4 take up 72% of all the data and that they contain very similar proportions of contributions from all recordees suggests that the study's findings are typical of the six and generalisable to all members of the 199192 MSc. Figure 9: Percentage of DUs in each K area for each recordee Karea AM BM CM DM AF
Kl K2 K3 K4 Shared knowledge sections
16 13 2 23 16
12 33 4 13 12
17 32 49 23 9
19 6 38 24 27
16 8 7 10 12
BF 20 8 0 7 24
This is not to deny that individual differences exist. K2 topics are mainly BM and CM's domain (they like talking about computers, for example), and K3 topics are CM's and DM's (they like discussing higher degrees), but these K areas are the smaller ones, so this difference is not felt overall. Moreover, K2 and K3 topics are less frequent in the spring (see Figure 6) and BM, CM and DM do not speak less in the spring (see Figure 4 of Chapter 2), so the K area changes over time are not a result of this difference either. The women, AF and BF, speak more on Kl topics than they do on K4 (they like talking about evenings out and what members of their families are doing), whereas for the men, it is the reverse. The fact that BF speaks very slightly less in the spring and AM speaks a little more then may have contributed marginally to the overall increase in K4 topics in the spring. However, the spring term is the one in which all recordees speak for about the same time, and BF, AM and DM are not talking in isolation: they are taking part in conversations with other members of the six, on K4 topics. Another difference is that DM refers three times as much to shared interpersonal knowledge than CM does (see Figure 9), but DM can not be held responsible for the increase in these sections in the spring because CM speaks more and more over time. In conclusion, the findings about the predominance of K4 and its increase in the spring and the increase in shared interpersonal knowledge sections throughout the year can be taken as typical of the six and generalisable to all.
Knowledge areas 49
3.4 CONCLUSION This chapter has shown that knowledge about the course itself is the background to nearly half of the dialogues throughout the year, but that course dialogues do not increase over time as they seemed on casual observation. The chapter has suggested that course events dictate the choice of topic more than interaction over time does, course topics being twice as likely to occur than any other topic in the stressful spring term when the pressure of exams and project deadlines increases. What do increase with interaction over time are dialogues assuming shared interpersonal knowledge. This is especially the case in topics assuming general knowledge of the world outside the course, such as home and family situation and activities. As for the function of referring to common knowledge, Brown and Levinson (1978) state that to claim opinions, attitudes and knowledge in common with the hearer, the speaker may assert common ground, and that this is a positive politeness strategy. It would appear, therefore, that the students choose K4 and shared interpersonal topics for discussion with their colleagues in order to establish and maintain ingroup membership, by pointing to the knowledge that they as a group share, to the context that binds them.
This page intentionally left blank
4
GRAMMAR OF THE INGROUP CODE
4.1 INTRODUCTION Overview Chapter 3 explained that the model of K areas was devised so that the language used in each K area could be compared, and the effect of increasing common knowledge over time on language could be examined. This chapter begins the study of the language. It examines the connection between speakers and context, between language and the surrounding reality. This study is principally one of reference, the "relation that holds between speakers /.../ and what they are talking about on particular occasions" (Lyons, 198la: 220), the "function whereby speakers indicate via the use of linguistic expression, the entities they are talking about" (Brown and Yule, 1983: 205). The study, thus concerned with the contextual sensitivity of language use, takes into account the fact that utterances are both contextshaped, speakers' contributions to an ongoing sequence of actions being understood only by reference to the context in which they participate, and contextrenewing, each utterance forming the immediate context for the next action in a sequence and contributing to the contextual framework for the next action (Drew and Heritage, 1992: 1618). Extralinguistic knowledge of the world is interactively and socially constructed (Duranti and Goodwin, 1992: 230).
52 Analysing the language of discourse communities Speakers use verbal and nonverbal signs to relate what is said to knowledge gained through experience, in order to retrieve the presuppositions that they rely on to maintain conversational involvement and assess what is intended (Gumperz in Duranti and Goodwin, 1992: 230). This study focuses on contextualisation cues (Gumperz, 1982), or features of linguistic behaviour that indicate those aspects of the context that are relevant to what the speaker means and enable interactants to make inferences about one another's intentions and goals. Contextualisation cues operate on the levels of prosody, paralinguistic signs, code choice, choice of lexical forms or formulaic expressions and choice of grammar. CookGumperz and Gumperz define them as "the use of code or styleswitching strategies, of formulaic utterances and other lexical options, which affect the inferential process by recalling interpretative schemata or suggesting how message parts can be linked to create a thematic whole" (in Grimshaw, 1994: 381). This study is limited to the grammatical and lexical cues. In normal communication, "much of what is intended to be communicated can be implicit and takenforgranted" (Grimshaw, 1994: 319). Gumperz (1982: 131) suggests that the meaning of contextualisation cues can be especially implicit when used by members of social groups: "exclusive interaction with individuals of similar background leads to reliance on unverbalised and contextbound presuppositions in communication". Firth (1957) and Bernstein (1971) observe that the restricted code of social groups contains contextdependent language based on unspoken assumptions not available to the outsider. Restricted code typically contains a great deal of exophoric reference depending on both a context of common experience and a context of culture (Halliday and Hasan, 1976: 3436). This chapter begins the exploration of the rule of implicitness that the students seem to follow when they refer to aspects of their shared reality. As Drew and Heritage (1992: 22) note, in institutional talk, there are "inferential frameworks and procedures that are particular to specific institutional contexts". They observe that in nonformal institutional settings, patterns have less uniformity than formal; this chapter aims to show that even in the nonformal setting of the commonroom, there are rules. It examines the way that implicitness develops over time and analyses its distinct features. The students' language (see Figure 1) includes an 'ingroup code' of implicit grammatical and lexical cues, as well as 'other implicit features' at the clause, utterance and exchange level. The 'implicit contexualisation cues' are discussed in this chapter and in Chapter 5; the other features are examined in Chapter 6. Figure 1: The implicit language of the in-group Ingroup code of implicit contextualisation cues: grammatical explicit nonanaphoric definite reference implicit reference lexical special course nouns and general words Other implicit features: clausal ellipsis implicitness at exchange level conversational implicature
Grammar of the ingroup code 53 Grammatical Reference Categories The grammar of the noun phrase is central to the code because it amounts to a mode of social interaction (Ochs et al. 1996); for the turntaking system to function, interactants must share an understanding of the grammatical resources of their particular form of the language (Ford, 1993: 11). The central focus is on the explicitness of nounphrase reference. Interest is concentrated on non-anaphoric rather than anaphoric reference, and on definite rather than indefinite reference, because the anaphoric indefinite reference contributes less to the exclusivity of the ingroup conversation. Nonanaphoric definite reference is examined in terms of explicitness, to discover whether a new referent is referred to very explicitly with a posthead modifier, explicitly with no posthead modifier, or implicitly with just a general noun or pronoun. A detailed study of the grammar of verb forms might have been relevant to an analysis of ingroup language but it is beyond the scope of this study. The implicitness of verbs is analysed from a lexical point of view in Chapter 5, which defines verbs such as "to write" as explicit, being contentful, and general verbs such as "to do" as implicit, being a noncontentful. Clausal ellipsis is examined in Chapter 6. The analysis of anaphoric and nonanaphoric reference used in the study is based primarily on Halliday and Hasan's (1976) definition of exophora and endophora. They explain that endophora is textual and cohesive, the referent being retrievable either in the preceding text (anaphora) or in the text that follows (cataphora). The interpretation of the endophoric referring expression lies therefore within the text, and that of the exophoric referring expression lies outside the text. Exophoric reference can be either situational or homophoric. Situational reference "links the language with the context of situation" (Halliday and Hasan, 1976: 37) and includes deixis (Levinson, 1983) in which the referring expression points out to the immediate context of the utterance. Homophoric reference is in the context of culture in which "the referent is identifiable on extralinguistic grounds no matter what the situation" (Halliday and Hasan, 1976: 71). Blakemore (1992) calls this reference the retrieval of a mental representation from memory. Exophoric referring expressions can therefore be understood in the wider context of intertextuality and common knowledge. Hawkins (1978) says that "the larger situation" use can be based on specific knowledge (eg: A: "I'm going to the store" when A knows which store B goes to) or on general knowledge (eg: A: "I wonder where the city hall is", when both A and B know every town has a city hall). It would be simplistic to suggest that endophoric and exophoric reference can be easily distinguished, however. As Halliday and Hasan admit, "The line between exophoric and anaphoric is not always very sharp." (1976: 18). There is an overlap between the two. The first cause of overlap comes from a characteristic of endophora: endophoric reference also needs the hearer to have some background knowledge. An endophoric reference item almost always points to both a presupposed item within the text and the referent itself that lies in the world outside the text. As Lyons (1977: 660) says, an anaphoric pronoun "refers to the referent of the antecedent expression with which it is correlated". There is a "bridging" between anaphoric reference and the antecedent by assumptions which are not actually present in the preceding utterance, but which are constructed by inferences based on what the listener or reader knows, and guided by the principles of relevance (Clark and Clark, 1977). Anaphora that relies partially on inference is known as associative anaphora (Hawkins, 1978). Venneman explains that the listener or reader understands the full meaning of a proposition by referring to a "presuppositional pool", that contains information "constituted from general knowledge, from the situative context
54 Analysing the language of discourse communities of the discourse, and from the completed part of the discourse itself" (1975: 314). In the present analysis, associative anaphora is included in the category of anaphora. The second cause of overlap between endophoric and exophoric reference comes from a characteristic of exophora. Exophoric reference also needs the hearer to have some knowledge of the preceding text and texts, and the likely presuppositional pool. The hearer needs knowledge of the preceding text because speakers following the principle of relevance rarely introduce a noun phrase that has no connection at all with the preceding discourse. The hearer also needs knowledge gained in previous conversations. This is intertextual (de Beaugrande, 1981) knowledge, in which the presupposed item is in previous texts, taken for granted as shared and in the minds of speakers. Since the previous texts are part of the context of culture, the borderline between text and context becomes blurred. However, to say that all conversations between the same speakers constitute one text, and that exophoric reference can be seen as cohesive and anaphoric, would seem somewhat extreme. As Lyons says, "not all of the intersubjective knowledge that is exploited in the interpretation of texts derives from what has been previously mentioned" (1977: 673). In this study, intertextual exophora is included in the category of exophora. Each dialogue is considered a separate text, since it cannot usually be established whether an exophoric referent is intertextual or not. The term "cataphoric" is dropped altogether in the study, because in the data there are no cases of the presupposed item coming further ahead in the text than the posthead dependent immediately following the referring expression. Some forms of reference can never refer forward cohesively past the noun phrase of the referring item. As Halliday and Hasan point out, the definite article "can only refer to a modifying element within the same nominal group as itself." (1976: 71). The postmodifying element can be a prepositional phrase, a relative clause, another noun phrase, etc., as in "The ascent of Mount Everest" and "The people who predicted a dry the summer" (ibid.), or "the fact that he was married", "the opera 'Carmen'", and "the man to do it" (Huddleston, 1988: 93). The idea of dropping the category "cataphora" and using just "anaphora" is not new. Lyons favours using the term "anaphoric" to cover "both normal backwardlooking anaphoric reference and the less normal forwardlooking or anticipatory, anaphoric reference" (1977: 659). Since the term "cataphoric" is dropped and only "anaphora" remains from endophoric reference, exophora can be known as "nonanaphora". Levinson (1983), considering deixis, refers simply to anaphoric and nonanaphoric usages, as does Blakemore (1992). Figure 2: Analysis model for all grammatical reference Anaphoric NonAnaphoric referring to the referring to the textually unknown textually known situational / homophoric / intertextual and usually the Known Introductory cognitively known pointing to the introducing the inc. associative cognitively known cognitively unknown anaphora The definition of the nonanaphoric referring expression used in this study (see Figure 2), is that which points outside the text to a referent not referred to in the preceding text. The referent is either present in the immediate situation (situational reference), in the background knowledge
Grammar of the ingroup code 55 (generalised, homophoric and contextual reference) or in previous texts (intertextual). When speakers use nonanaphoric reference, the referent is textually unknown, and they assume that either the referent is cognitively known (in the listener's mind) or that the referent is cognitively unknown (not in the listener's mind therefore, and needing introductory reference). The traditional definition that states that nonanaphoric reference is only that which is totally new, in the sense of being completely unrelated to anything at all in the preceding text, is stretched. Referring expressions are classified as nonanaphoric if their meaning is not immediately interpretable from something concrete in the text. Example (1) is one of nonanaphoric reference. BF has been explaining how the tutorial task had been divided up: (1) 04124 BF Has anybody done their syntax? ((!)) 04125 DM // I did it yesterday. 04126 DM Oh that's what I was doing yeah. 04127 BM // No I haven't looked at it yet. -> 04128 BM Oh you were doing it in the library weren't you? 04129 BM You're well ahead aren't you? The library is mentioned for the first time and has not been implied in the preceding text, except perhaps in the sense that a student may sit somewhere in the university to "do their syntax". Definite reference is of interest in this study because it depends critically upon mutual knowledge, beliefs and suppositions (Clark and Murphy, 1982). Definite reference is made with definite noun phrases (proper nouns, definite articles, demonstrative determiners, demonstrative pronouns and adverbs, personal and possessive pronouns) which show that speakers assume that hearers can identify the referent using their textual knowledge or their contextual knowledge. Definite referring expressions are subjectively involved in pragmatic use, varying according to where and when they are uttered and by whom (Lyons, 1981b: 168). Speakers base their audience design on the adequate background of the hearers: "When speakers use definite references, they assume their addressees can immediately identify the individuals referred to" (Clark, 1997: 5757). In (2), the interlocutors, talking about experiences with bus drivers, are expected to know "the green buses" from their specific knowledge of larger situation of Edinburgh: (2) 04057 DM Demented bus driver. 04058 BF // (hehhehheh) 04059 AF // (hehhehheh) 04060 AF You get the occasional one. (2.5) > 04061 AF The wild ones are on the on the green buses actually. 04062 AF They go at a hell of a rate. Schiflrin (1994: 199) notes that the maxims of quantity and relevance may work together to constrain referring items, in terms of definiteness and explicitness. She says that "Definiteness is concerned with S's intentions and assumptions about what H can be expected to know. Explicitness is partially motivated by S's cooperative intentions, i.e. information presented to enable H to identify the intended referent." She affirms that the quantity of information conveyed in a particular referring term is as important as the relevance of that information. She explains that whereas definite forms "indicate S's intentions to refer to a single entity that can be specifically identified by S, and that S expects H to be able to identify from whatever clues (textual, contextual) are available", explicitness "has to do with the presentation of information that actually enables H to correctly identify a referent, i.e. the lexical cues that allow H to single out whom (or what) S intends to differentiate from other potential referents." (ibid.). As Coulthard (1977) says, the speaker must make a contribution sufficiently explicit to be clear, and
56 Analysing the language of discourse communities "membership" his listener each time the topic changes. This view of definiteness and explicitness being interwoven is central to the analysis in this book. Some linguists have considered degrees of explicitness at the mode and register level, but they risk oversimplifying and overgeneralising the case. Chafe (1982) describes the written mode as explicit and contextfree, and the spoken mode as implicit and contextdependent. LeckieTarry (1995: 133) says that "the lowest degree of explicitness [is] to be found in the registers of casual conversational and the highest degree of explicitness to be found in the registers of formal written discourse." Tannen (1982) finds that implicitness depends on register and genre. Biber (1988) finds that explicit references, defined as those that can be decoded without recourse to situation, occur more in academic prose than in popular lore texts such as informative texts found in popular magazines. This book examines the explicitness of a spoken mode that is heavily influenced by the written mode: the students talk about the concepts that they have read about in articles, the theories that they are going to write about in the exam. Their language contains a mixture of the formal academic register and the informal commonroom chat register. The distinction between the two is not easy nor indeed useful to make. Those linguists who look at degrees of explicitness at the utterance level remain in abstractions and cooperation principles. Blakemore (1992) analyses explicitness in terms of information conveyed either explicitly or implicitly, noting that the right amount of information is given to make the relevance between two utterances clear enough for the hearer to interpret the meaning using his contextual resources. Grice (1975) covers the issue of explicitness with his maxim of quantity: the speaker should provide the hearer with enough information to be able to understand. According to Grice, the speaker can imply more than his words say by flouting these maxims and using conversational implicature, what Brown and Levinson (1978) call going "off record". This does come closer to the approach taken in this book: Chapter 6 contains an examination of implicitness over utterances and behind utterances, based on cooperative maxims. The degrees of explicitness of referring expressions have been given scant attention by linguists, and this is at the centre of the present study. Warren (1993) mentions only proforms and general nouns as exponents of inexplicit language. Most of those who do examine explicitness at the level of the noun phrase treat it as a gradable quality on a cline from explicit to implicit, yet the analysis remains superficial. McCarthy and Carter (1994: 10) simply say, "terms such as implicit and explicit are not absolutes". Hasan (1984: 125) establishes a cline from the explicit cataphoric, to the less explicit anaphoric, to the more implicit exophoric, in which the referent is a concrete element in the context of situation, to the most implicit exophoric in which the referent is in the common past experience of the participants. Halliday and Hasan (1985) refer to all personal pronouns, demonstratives, possessive pronouns and substitution as "implicit encoding devices", yet this study shows that some are more implicit than others. Schiffrin (1994) outlines the exponents of degrees explicitness of referring expressions in greater depth but still only mentions them in passing. She equates "explicit noun phrases" with "'lexically informative' full noun phrases", and lists general nouns, personal pronouns and zero personal pronouns as "less explicit" referring items. She is mostly interested in the difference between the two being the difference between the first mention of a referent (often indefinite) and the second mention of it (often definite). Her description of how these crosscut on a scale of degrees of explicitness is quite useful:
Grammar of the ingroup code 57 "I can use all of the following (and more) definite descriptions to refer to the same person: my husband, Louis, Dr.Scavo, the man Hive with. These are all more explicit than the indefinites an adult I live with, someone I met in college. The definite he, however, is less explicit than the indefinites just given, /.../ but more explicit than zero anaphora." (p.199) Her scale combining definite and indefinite reference and adding detailed distinctions within each of these categories is unfortunately impractical for building a model to analyse data. It is also incomplete in that she does not examine the explicitness of noun phrases in terms of cohesion; a pronoun can surely be either implicit or explicit (see below). She does not examine the effect of different determiners in the noun phrase (e.g.: "the pen", "that pen") and she does not explore degrees of explicitness in verbs (e.g.: "to work", "to do"). Finally, Halliday and Hasan do link implicitness of referring expressions with impenetrability, as this study does: "Exophorically interpreted implicit devices create an opaque link between the text and its context so far as speakers outside the context are concerned" (1985: 767). The model used in this study contains different degrees of explicitness (superexplicit, explicit and implicit) for both anaphoric and non anaphoric reference. See Figure 3 for examples of how the degrees of explicitness fit in with definite reference. Figure 3; Degrees of explicitness SuperExplicit the/that pen that I bought the pen with a nib Explicit the/that pen John Implicit the/that thing that I bought the/that thing/person that/there it/him The term 'superexplicit' reference is used to mean that in which the referent referred to in the noun head is further identified by means of a postmodifier to add extra information. The post modifier, such as a prepositional phrase, infinitival clause, relative clause, participial clause, adjectival phrase or noun phrase in apposition, is defining and restrictive but it is not obligatory (Huddleston, 1988). This would have been classified as cataphoric in Halliday and Hasan's framework but in the present model both anaphoric and nonanaphoric reference can be explicit or otherwise. The term 'superexplicit' is preferred to another such as 'modified explicit', 'modified noun phrase' or 'noun plus posthead' because it makes it clear that this type of noun phrase is further along the scale of explicitness (at the other end from the implicit) than the explicit, or bald noun head. Thus "the pen that I bought" has more content, gives more information about the referent than the explicit "the pen", just as the "explicit" form carries more content than the "implicit" "it" or "the thing". In superexplicit reference, the speaker further identifies the referent for the hearer, as in (3), in which AF makes her identification of "the street" more explicit with "we go down": (3) 21038 AF The City Cafe on Blair Street. (2.5) 21039 AF Just above Cowgate. 21040 AF You know when we go down to Wilkie House. 21041 DM Aha. > 21042 AF Well the street we go down the City Cafe is in that. 21043 AF It's a nice place. (3)
58 Analysing the language of discourse communities This is Hawkins' (1978) definite reference category of unavailable use, which introduces new information unknown to the listener, using a relative clause or a prehead modifier. The term 'explicit' reference is used to mean that in which the referent referred to in the noun head is not further identified with a posthead modifier. This "bald" noun phrase can be understood by those knowing the text and/or context without the additional help of an explanatory or identifying expression indicating exactly which of all the possible referents with that name is being referred to. In (4), the students are about to go over a tutorial task together and "that list" is introduced with no further identification: (4 ) 03032 NM (heh) We're already ten minutes late so. ((5)) 03033 BM Urn. —> 03034 BM What we decided was wewe got did you get that em list em (2) the em the sh-topic sheet from XX on on Monday? 03035 FF No I had it I had it read out on the phone to me last night. Here, even the modifying expression "the topic sheet" would make it more explicit only to an in group member, because it is another example of bald "explicit" reference. FF obviously knows what is being referred to: she refers to it too, with "it". Explicit reference includes both common nouns and proper nouns. The proper noun has as much denotation as a personal pronoun or a general noun, but it is in the explicit category in that it identifies the referent as explicitly as 'explicit' referring expressions do, and because it identifies the referent more explicitly than 'implicit' referring expressions do. "Louis" is as explicit as "my husband", and less generalisable than "that man" or "he". Excerpt (5) is from a discussion about a lecture handout and the answers that students wrote in their exams: (5) —» 12146 CM Cos I'm sure it said Structuralism Saussure and then the next week it said Structuralism Bloomfield. 12147 AM Ah that's it yeah. ((1)) > 12148 CM And I did qualify it with that Lyons what Lyons said about (0.5) being an American who was trying // to 12149 AM // Mm. Mm. It would have been more generalisable if CM's words had been "what he said" instead of "what Lyons said", which might denote as much as "what the linguist said" to an outsider. MM seems to remember who Lyons is and what he said. Anaphoric pronouns that cohere with a contentful noun with explicit reference (e.g.: "a blue pen") are also considered explicit; pronouns are seen as implicit only if they are used nonanaphorically or if they are cohesive with an nonanaphoric implicit referring expression such as a general noun (e.g.: "a thing") or another nonanaphoric pronoun. The explicit reference category includes the noun head followed by a posthead complement such as a prepositional phrase, a content clause, or an infinitival clause, as in "his refusal to go" or "the need for us to help her" (Huddleston, 1988: 93). In this study, the noun with an posthead complement is considered to be in the same category as explicit "bald" reference because the obligatory complement depends on and is selected by the noun head. Whereas in superexplicit reference, the posthead modifier adds extra nonobligatory information to further identify the referring expression, the posthead complement is intrinsically connected to the noun head. In example (6) of an explicit referring expression containing a complement, a student suggests how to modify a project if the discussion of a particular point is longer than the introduction stated it would be:
Grammar of the ingroup code 59 (6)
17041 CM
//And then you do and then you do a search for "briefly" and change to (0.5) change to "extensively" or something if you actually do discuss it in depth. Here, "a search" selects "for", because of the verb "to search for"; the sentence would have lost its meaning if "for 'briefly'" had been omitted. One major type of complement in the explicit category is the posthead partitive genitive phrase: that which completes the noun phrase by identifying the whole, the mass or the class of which the preceding noun is a part or a member, such as "on the back of American comics" and "95% of the people". Example (7) contains an explicit referring expression with a partitive genitive phrase. AF is talking about her cold: (7) 21029 AF Yes the worst was actually in the middle of the week when I was planning to work very hard. Here, "the week" is the whole and "the middle" is the part. The noun in the posthead partitive genitive phrase functions semantically as the head noun; "the middle of the week" is more "week" than "middle". The term 'implicit' reference is used to mean that reference, usually 'bald' (i.e. with no posthead modifier), in which the referent is not named by the referring expression, but is rather referred to with a vague expression. The implicit referring expression is a general noun or a personal, demonstrative or indefinite pronoun. "That thing", "a thing", "the person" or "people" have as much contentfulness as "that", "something", "he" or "they". Implicit reference also includes colloquial forms of general noun and vague noun clauses such as "the what'saname" and "what you said you'd do". These are WilkesGibbs' (in Clark, 1993) dummy noun phrases "whatsit" and "thingamabob". General nouns are classified as implicit whether they have a posthead modifier or not. The posthead modifier after a general noun does not add enough to make it as explicit as a contentful noun. In (8) from K4, if AM had not limited the reference of "stuff' to something that had to be read, presumably an article or a handout, he might not have communicated his idea at all: (8) »14059 AM That's the stuff we we read last week which we should've read this week. (0.5) 14060 AM So read it again. 14061 AM I found it. 14062 AM It's in there. 14063 NF Chapter Six. NF seems to know what the "stuff1 is. A further discussion of the general noun will follow in Chapter 5. It is included in this grammatical chapter as well as in the lexical chapter because the general noun bridges the lexical and the grammatical boundary. As Halliday and Hasan say, the general noun is a "borderline case between a lexical item (member of an open set) and a grammatical item (member of a closed system)" (1976:274). Implicit reference can be nonanaphoric or anaphoric. In the case of implicit nonanaphoric reference, the referent can only be guessed at if the listener has the requisite background knowledge of the world outside the text. In (9), DM and CM had been talking about handing work in on time; in unit 15045, DM suddenly shifts topic using implicit reference to something that has not been mentioned before: (9) 15041 CM Can't remember the last time I handed in anything late. 15042 DM (hehheh // hehheh) 15043 CM // Usually it's three months early.
60 Analysing the language of discourse communities 15044 DM (heh heh) Right. (8) > 15045 DM So I typed that thing up again after you'd gone. 15046 CM Oh yeah. Here, the general noun "thing" carries very little more semantic information than the pronoun "it" would have done, yet CM appears to be able to identify the referent. In the case of implicit anaphoric reference, the referring expression is cohesive with a presupposed item in the text; the presupposed item itself is sometimes implicit and so does not point out clearly to a referent or bring the hearer any closer to the referent's identification than the referring expression itself. In (10), CM is describing Japanese beaches: (10) 19109 CM But they they're paranoid about their their islands er dissolving into the ocean so they've done all these cement they're called er (1.5) like the // er 19110 BM // What like 19111 CM Nono. 19112 CM They're huge like the size of this room. 19113 CM One is the size of this room. —» 19114 CM And they've got thousands of these things stuck out there with the islands. —> 19115 CM So you might get a stretch of about half a mile of nice sandy beaches and then there's this huge what-you-call-it breaking the surf. Here there is cohesion but CM has not named the presupposed item for the "things", the "what youcallit"s to be cohesive with or the "they" to refer back to.
Analysis Of Grammatical Reference Warren (1993: 3841) makes the point that in conversations, utterances are produced at a lower level of explicitness than in other discourse types. He claims that fluctuation in the level of explicitness is a product of changes in the immediate context, and shows that participants who can both see what is being referred to and have knowledge about entities in their shared physical setting are less explicit than those who cannot. The researcher in the present study felt that fluctuation in explicitness is a product of changes in shared cultural context, in assumed background knowledge. Casually observing the students interacting in the common room, she sensed that over the duration of the course, the form of reference became less explicit, the referring expressions containing less information as precise identification of referents becomes less necessary. She felt that the explicitness varied according to the topic, and that implicitness was greatest in K4 topics. The first stage of the analysis was to tag every noun head and verb to indicate both grammatical and lexical information. The grammatical reference coding tag usually consisted of three letters and one number. The first letter indicated phora type (anaphora, nonanaphora). The second letter indicated the reference type: definite reference tags included demonstrative reference (definite article and demonstratives, adverbs), personal reference (personal pronouns and possessive adjectives), and comparative reference (comparative adjectives and adverbs); all indefinite noun phrases were tagged the same, whether the determiners were "a", zero articles, "some", "no", "every", "much" or "a lot of. The third letter indicated the grammatical form of the reference or the part of speech (pronoun, adjective, adverb, ellipsis). The number at the end of the letter combination indicated the degree of explicitness (superexplicit, explicit and implicit). Ellipsis and
Grammar of the ingroup code 61 substitution of parts of the noun phrase was not examined because nonanaphoric ellipsis proved exceedingly rare. Ellipsis at the clause or utterance level (unfinished sentences) was tagged but analysed along with the "other implicit features". In the tagging of definite reference types, not every noun phrase was analysed. Nouns with demonstrative determiners and definite articles were not tagged when they occurred in adverbial phrases. Phrases of contrast, such as "on the other hand", and addition, such as "in the sense that", were not analysed as they are fixed expressions whose meaning varies little with context. The exception to the rule about adverbial phrases was time expressions such as "this week" or "at the end of this week" in which demonstratives and definite articles were tagged, because they were not fixed general expressions of chronological sequencing, but pointed out to specific moments, with a pragmatic meaning. In (11), speakers have been talking about how unsatisfactory it is to do temporary teaching work: (11) 05094 BM But on the other hand it's it's (0.5) it's it would be useful to do that work (0.5) for the future like for example for the the summer. 05095 BM Get yourself known around places. 05096 BF Mmm. Here, "for the the summer" only has meaning for them in the light of the course timetable. The personal pronoun "it" in generalised empty reference was not tagged, as in (12) in which AF enters the common room and declares: (12) 15097 AF God it's hot in here. because "it" is simply providing a subject for the verb. Nor was "it" tagged in empty reference in fixed expressions with a meaning such as "already", "I agree". In (13), AM explains why he left his group of friends after the pub: (13) 11113 AM I thought I'd had too much to drink as it is. The generalised pronoun "they" was tagged, on the other hand, because it has a more personal meaning. In (14) from a conversation about importing a computer: (14) 06007 CM You're buying a computer? 06008 NF Yeah. > 06009 NF They sent me this. 06010 AM And they persuaded= "they" could be identified as the computer firm personnel/sales manager; AM picks up the same "they" to begin another question. Personal pronouns in tag questions were not tagged because the pronoun does not carry any meaning. In all the data, there were no third person possessive pronouns or nonanaphoric third person possessive adjectives. The possessives category was maintained in the study, however, in order to study the instances of implicit anaphoric expressions. Working with spontaneous recordings meant devising a system to accommodate the "messiness" of spoken discourse: stutters, repetition, false starts and errors of speech. In the case of stutters, as in: (15) 06091 NF Yeah but this (this) is just for er for January 1992. the repeated demonstratives, definite articles, personal pronouns, etc. were counted as one instance. Repetition that happened over more than one discourse unit, was tagged as two separate instances, even when they constituted a repetition of exactly the same words with exactly the same meaning. The second mention of the demonstrative, definite article, personal
62 Analysing the language of discourse communities pronoun, etc. was labelled the same as the first; it was not treated as an item of cohesion with the first. In false starts, as in: (16) 16079 AM (They) the third consonant is the least remembered or something when you when you're when you only get it orally. none of the incomplete constituents were tagged, partly because the false start loses its meaning and significance being an incomplete expression, and partly because this clearly makes it impossible to tag it anaphoric/nonanaphoric or explicit/implicit. Errors in reference type, as in: (17) 17037 AF Can't you sort of say later on em these subjects will be revised briefly? 17038 BM (hehheh) > 17039 NF Oh I do (those). were treated as if they referred as intended, and are tagged as if they were the right word. In (17), "those" was tagged as if it had been "that". Before the data was analysed, this part of the code was given to three subjects for intercoder reliability testing, and they scored 68%, 38% and 63%. The reason why it is possible for a coder to get such a low score as 38% may be that this part of the code is so complicated that coders either failed to understand all the intricacies, or they were quite simply discouraged by the complexity and did not give it the attention that it requires. The analysis of the tagged data was straightforward. Once the tokens of each feature within each of the four K areas were counted, the density of each tag in each day was calculated, by finding the percentage of each one out of the total number of words in each day in each K area file. Then, the average of these percentages for each of the tags in each dialogue was calculated for each of the three terms, to discover the overall changes over time and the difference between noncourse dialogues and K4 dialogues. The average proportion of anaphoric and nonanaphoric reference was calculated, as well as the average density of superexplicit, explicit and implicit referring items in each term. Finally, a qualitative study of the function of each reference type was made, to bring out differences in terms of context and specificity of reference.
4.2 CHANGES IN GRAMMATICAL REFERENCE Non-Anaphoric Definite Reference Three general observations should be made before a discussion of nonanaphoric definite reference. Firstly, the overall density of all pronouns and adverbs out of all words does not change over time, remaining at 8% throughout the year. Any increase in certain types of pronoun cannot therefore be said to be a reflection of a general increase in all pronouns and adverbs. Secondly, the analysis of all nonanaphoric reference, regardless of whether it is definite or indefinite reference, shows that it is consistently denser in K4 than in noncourse K area dialogues. The percentage out of all reference is 33% in noncourse K areas and 43% in K4 in the autumn term, 36% in noncourse K areas and 51% for K4 in the spring term, and 37% in noncourse K areas and 51% in K4 51% in the summer term. This suggests that students talking on course topics are more likely to introduce new referents with little connection with the preceding discourse than when they are talking on noncourse topics. The third general observation to be made is that an analysis of all definite reference, regardless of whether it is anaphoric or nonanaphoric reference, does not show a consistent increase in the density over
Grammar of the ingroup code 63 time. The percentage out of all noun phrases is 74% in the autumn term, 80% in the spring term and 74% in the summer term. This peak in the spring term occurs when K4 topics occupy 57% of dialogues. The analysis of nonanaphoric definite reference, taking all degrees of explicitness together, shows that the density out of all reference has a similar peak in the spring term although it does not return quite to the autumn level in the summer term (25%, 29%, 27%). It would seem that the longer the students interact as a group, the more they can refer to referents that are not in the preceding text, using definite referring expressions that assume that the referents can be identified. Figure 4 shows clearly that K4 has much more nonanaphoric definite reference than noncourse K area dialogues. The peak in nonanaphoric definite reference in all K areas in the spring term is caused therefore by the characteristics of K4 which predominates then. Figure 4: Non-anaphoric definite reference out of all reference
Degrees Of Explicitness An examination of the degrees of explicitness, taking all reference together (anaphoric and non anaphoric, definite and indefinite) and all the data together, reveals that explicit reference is the unmarked form of referring and that there is a slight increase in implicit reference in the spring term, which then keeps at the same level in the summer (see Figure 5). A x2 test of significance was carried out on the proportions of tokens of noun phrases with implicit reference in each of the three terms. The value of %2 was 2.065, significant at the 0.005 level, and showed that the difference between the proportions of implicit reference and nonimplicit reference in each of the three terms is significant. The more experiences the students share, and the more they know of each other, the more they use implicit reference. Figure 6 shows the changes over time in degrees of explicitness in the different K areas. Not only does K4 have a higher density of implicit reference than noncourse K areas but this increases dramatically in the summer term, at the expense of superexplicit and explicit reference. Nearly half of K4's reference in the summer term is implicit. A %2 test of significance was carried out on the proportions of tokens of implicit reference in course and noncourse K area topics. The value of x2 was 102.32, significant at the 0.005 level. It showed that, taking all terms together, the difference between the proportions of implicit reference and nonimplicit reference in course topics and noncourse topics is significant.
64 Analysing the language of discourse communities Figure 5: Degrees of explicitness out of all reference in all K areas
Figure 6: Density of degrees of explicitness out of all reference In Non-course K areas
Grammar of the ingroup code 65 Just within non-anaphoric definite reference now, an examination of the degrees of explicitness, taking all the data together, reveals that the percentage of super-explicit reference increases in the summer term, (the autumn term: 3%, the spring term: 2% and the summer term: 11%). This occurs mainly in noncourse K areas (see Figure 7). It is likely that since the summer term is a time when noncourse talk is about future activities after the course (K3), superexplicit language is required to introduce new referents in case they are not known by each individual. K4 has consistently a lower density of superexplicit nonanaphoric definite reference than noncourse K areas. Course topic referents referred to for the first time are less likely than noncourse topic referents to need additional information to identify the referent. It is not, therefore, so much increasing knowledge over time that influences the density of superexplicit reference as the K area. In superexplicit reference, the posthead dependents in each K area type seem to differ functionally. In noncourse K areas, they tend to have the function of informing, explaining and instructing. In (18), the posthead has a restrictive function. BF is explaining that her partner has been made warden of a university hall of residence: (18) 25064 BF He's been made into the warden of our house. 25065 BF We've got a brilliant university house. 25066 NF Ah and he's the warden. 25067 BF Yep. -> 25068 BF The girl who was doing it has dropped out of university. She has to further identify the girl so that NF knows who she is talking about. In K4, postheads tend to be nonrestrictive, occurring in utterances that express an opinion or add a comment. In (19), AF gives her view of an article; AM agrees. (19) 16032 AF I think I find a real loss actually of not having read the (0.5) Fay and Cutler article which seems to be underpinning this. 16033 AM Yes that's it. Again just within nonanaphoric definite reference and taking all the data together, the percentage of explicit reference shows a dramatic decrease in the summer term (the autumn term: 67%, the spring term: 68% and the summer term: 46%). That is to say, students refer less in an explicit way to contextually new referents assuming listener knowledge of them only once the course has been well established for a time. The analysis of noncourse K areas separately from K4 dialogues shows that this decrease occurs in both K area types and that the density of explicit non anaphoric definite reference is lower in K4 topics. When students are talking on a course topic they are less likely to use an explicit noun than they are on a noncourse topic (see Figure 8). Moving on finally to implicit nonanaphoric definite reference, analysis shows that taking all K areas together, the density rises sharply from 31% in the autumn and spring terms to 50% in the summer term. The density of both K4 and noncourse K areas doubles in the summer term (see Figure 9). The rise in implicitness in noncourse K areas in the summer term can be explained by the fact that this is the period in which K3 topics on doing a PhD and applying for the IALS scholarship occur. These are the topics that are closest to K4. In K4, the course elements only need to be referred to vaguely because they are in the forefront of all the students' minds, even though the components may change according to events on the course calendar.
66 Analysing the language of discourse communities Figure 7: Percentage of superexplicit reference out of all nonanaphoric definite reference
Figure 8: Percentage of explicit reference out of all nonanaphoric definite reference
Figure 9: Percentage of implicit reference out of all nonanaphoric definite reference
Grammar of the ingroup code 67 The reference types in implicit nonanaphoric definite reference were more often demonstratives and personal pronouns ("that thing", "this", "there" and "she") than definite articles occurring with general nouns ("the thing"). The main difference between noncourse K areas and K4 in this respect is that K4 has a higher density of nonanaphoric personal pronouns. Example (20), in which CM suddenly changes the subject and uses a nonanaphoric pronoun, is therefore characteristic of K4: (20) 15159 DM I'm going to give out a questionnaire. 15160 DM And I'll give you one as well. 15161 DM Sometime this week I hope t tomorrow I'll get them all done. 15162 AF What your core project? 15163 DM Yeah. ((0.5)) > 15164 CM Did he like did he like the idea? 15165 AF That's very energetic. > 15166 DM Well you know what he's like. 15167 DM It's difficult to tell isn't it? CM knows that he can ask about the tutor's opinion on the progress of the project without giving him a name. DM knows who he is talking about, and refers to the tutor as "he", too. The implicitness is interactive.
Shared Interpersonal Knowledge There are few instances of nonanaphoric definite reference in shared interpersonal knowledge sections. Cases of superexplicit reference are exceedingly rare, which seems to suggest that if speakers talk about themselves or the interlocutor, then the referents are guaranteed to be in the conscious mind of the hearers and need no extra explanation for them to be recalled. Implicit reference is not a feature of shared interpersonal knowledge sections either but, when it does occur, the reference seems more obscure and exclusive. In example (21), BM, midconversation about computers, apparently suddenly notices that FF is in the common room: (21) 26102 CM // Youyou don't have the processor to do it. 26103 BM Bubut why don't I? 26104 CM Cos (0.5) in technical terms you own an // EightyEight. 26105 BM // Say. > 26106 BM You shouldn't be here. 26107 FF I got the job. 26108 BM How come she got back so early? (0.5) > 26109 BM That was yesterday. 26110 CM In technical (1) terms you own a what's called an eightyeight six processor and you need an eightythree eightysix processor which is two models newer and better than yours. The demonstrative adverb "here" and pronoun "that" have no connection with the preceding text; BM and FF need no more than an implicit reference to what happened. The biggest group of implicit reference is in K4, in which the demonstrative determiner with the general noun is the most frequent reference type. In (22), there is no indication in the preceding text about journal articles of what is being referred to in unit (20050); this unit is the first of a topic shift: (22) 20046 DM There are a few things in ELTJ (2.5) about it. (0.5) 20047 DM ((sniffs)) which are you know. 20048 DM There's not a lot though. (0.5)
68 Analysing the language of discourse communities 20049 DM It's quite nice. » 20050 BM Did you see em X about that MSc thing? 20051 DM That's right. 20052 DM I'll go up and see if I can see it. DM can identify it, as it is, referring to it with an implicit "it". Nonanaphoric personal pronouns are dense in K4 shared interpersonal sections, especially in the spring term. They show knowledge of interlocutors' attitudes towards the members of staff, referred to covertly, as in (20). It may be that the cassetterecorder had an effect in the students' choice of reference, when they were expressing attitudes towards members of staff, using a non anaphoric personal pronoun or a general noun rather than using a proper name. In (23), "she" is the only referring expression that DM uses to refer to the member of staff responsible for the psycholinguistics lecture that they are waiting to go to: (23) 15109 DM What's that? 15110 DM Psycholinguistics? 15111 AF Mhm. 15112 AF I have difficulty getting my brain going first thing in the morning. > 15113 DM She certainly fillsfills it up doesn't she? > 15114 DM She's got lots of things to tell you I'm sure. 15115 AF Yeah. (6) ((yawns)) (3) AF knows who "she" is; they may mention no names because on the cassette recorder or because it is quite simply unnecessary. What is certain is that the nonanaphoric personal pronoun and the general noun also occur in contexts in which negative evaluations are not being made about members of staff and when identities do not need to be hidden. They are not, therefore, solely a product of the presence of the recorder.
4.3 REFERENTS The discussion in this section centres round the context issue of whether the referents in the non anaphoric definite reference are introduced as new or as assumedtobeknown, and whether they are in the immediate situation or in the background knowledge. It also centres round the specificity issue of whether the referents are specific or nonspecific/generic. Analysis suggests the both context and specificity are influenced more by K area than by increasing knowledge over time. Explicit and implicit reference in noncourse K area dialogues are compared with that of K4 ones, in order to discover whether the context and the specificity of K4 dialogues add to their exclusivity.
Context In non-course K areas, explicit nonanaphoric definite reference tends to occur in introductory reference in narratives. (24) is from a tale about a hypnotist show: (24) 02005 FF Did the guy volunteer or did he?= 02006 MM Yeah the guy volunteered. 02007 MM I'm not saying he wasn't hypnotised.
Grammar of the ingroup code 69 —> 02008 MM
It's just that I met someone afterwards and they said that what was happening was things like em (1) like at oneone point they got these glasses. 02009 MM OK. —» 02010 MM And they put these glasses on these blokes. 02011 MM And the blokes acted as if the audience was naked. 02012 BM (hehheh) /.../ 02018 MM I mean you just had to see their faces. 02019 MM I mean just such little horrid smug expressions. 02020 MM Like this ((!)) 02021 BM (hehheh//heh) This use of the explicit proximal demonstrative "these glasses" (unit 02008) and "these blokes" (unit 02010) is very close to the indefinite: they mean "a pair of glasses", "some blokes". The demonstrative heightens the dramatic tone and increases the interest. Another explicit reference example is (25), which comes from a story about weekend activities; the drama is brought closer to the audience by an introductory narrative demonstratives in "this enormous eight hour stretch" (22163) and "this big bloody mountain" (22165): (25) 22158 DM Yeah we did a lot of hill walking. (0.5) 22159 DM Wegotback(l)er(2) 22160 DM Michelle and I got home she looked at her knees. (0.5) 22161 DM They were like this. 22162 DM Swollen up like this. —> 22163 DM Cos we did this enormous eight hour stretch. 22164 AF Uhm. -> 22165 DM Up this big bloody mountain. 22166 DM // And back down the other side. 22167 AF // Yes. Hearers know that they are not expected to know these referents. Examples (24) and (25) contain instances of implicit reference, nonanaphoric demonstrative determiners and pronouns used, this time, in introductory situational reference in narratives. In the hypnotist narrative, MM says that the people wearing glasses had smug expressions: 02020 MM Likethis((l)) and presumably imitates their faces. In the hillwalking narrative, DM describes his wife's knees and seems to be gesticulating: 22161 DM They were like this. 22162 DM Swollen up like this. Nonanaphoric demonstrative adverbs also seem to be mainly situational. K2 has several instances in a dialogue in which BM asks CM to explain what can be done with a certain computing package and the students are presumably pointing to a sketch on a piece of paper in front of them, or it may be a laptop: (26) 26019 BM Cos when it's you know have you ever seen a those sort of a lesson plan? 26020 CM Aha = » 26021 BM Here you have most of the body of the information here (0.5) in the middle of the page or towards // the right 26022 CM // Yeah. Any person present in the conversation could identify these introductory narrative and situational referents; they are clearly more explicit than the demonstrative that depends on background
70 Analysing the language of discourse communities knowledge. Noncourse K area demonstrative adverbs can of course also require hearers to have background contextual knowledge. The background knowledge is often current news issues and events, as in (27), spoken at the beginning of a period of press invasion of the royal family's privacy: hearers need to be familiar with an incident between the Princess Ferguson and an official: (27) 24019 DM Fergey? 24020 BF (hehheh)(l) 24021 BF A royal. ((2)) » 24022 MM Not a royal now. ((2)) 24023 BF Urn? 24024 MM Not a royal any more. ((2)) 24025 BF Was she ever? "Now" means "now that she has been caught on camera having her toes sucked". In K4 dialogues, on the other hand, the majority of demonstrative determiners in explicit non anaphoric reference depend heavily on the background knowledge, assuming the exact circumstances or conditions of a referent to be known. In (28), speakers are comparing notes on an article that they had to read for a tutorial: (28) >16012 AM I knew it was em the er em then again when when we went away after that tutorial we we took the paper we took er em we just sat and went through it together and basically line by line. /.../ 16020 AF I couldn't actually follow (0.5) an argument as such in it. —> 16021 AF I thought she was bit of a bore and hedging and surveying and so on and I couldn't see the point of this semantic field boundary. The reference here is interactive and intertextual: "That tutorial" seems to mean "That semantics tutorial that we went to this week"; "this semantic field boundary" seems to mean "this semantic field boundary that we read about", but this modification is not necessary. K4's implicit demonstrative determiners and pronouns are the most obviously intertextual of reference types. When the demonstrative determiner modifies nonanaphoric general nouns, they emphasise the implicitness of these spontaneous dialogues. Example (9) above: 15045 DM So I typed that thing up again after you'd gone. is a case in point. The great majority of K4 demonstrative determiners with general nouns could not be understood without the background knowledge. The K4 nonanaphoric demonstrative pronouns also depend frequently on background knowledge of the course. In (29), the pronoun occurs in the first discourse unit of a new topic about core course material revised for the exam, after NF's story about almost missing a lecture: (29) 10047 NF Then I realised I had lectures! 10048 CM Aha. 10049 DM And you had to run back. 10050 NF (heh heh heh) Yeah. (3) 10051 NF Ah. ((3)) » 10052 CM When I first started studying I thought I'm going to learn seven or eight of these really well. 10053 CM Then I started with one. 10054 CM I went for the Language and Linguistics. 10055 DM Aha.
Grammar of the ingroup code 71 DM has no problem following CM switching topic with no explicit reference. The meaning of K4's nonanaphoric demonstrative adverbs are also mostly dependent on background knowledge. Most of them are concerned with time: what distance there is between the present moment and a project deadline, the end of a term, the date of the portfolio. In (30), speakers discuss a colleague's project progress: (30) 15016 CM He's nearly finished writing. > 15017 DM Well so he's nearly there. 15018 CM Yeah. 15019 CM Almost finished. (1.5) The "there" seems to imply "the point at which he can hand in his project". The deadline or the date in question is assumed to be in the minds of the speakers; it is not mentioned. An outsider would be excluded.
Specificity In non-course K area dialogues, reference tends to be generic, nonspecific. As far as explicit reference is concerned, Kl has most nonspecific; it contains generic reference to things that many of us have, use or go to, such as "the radio", "the bank" and "the pub". In (31), DM explains what happened when there was no 9 a.m. class: (31) 04006 AF You mean you didn't get up earlier and make the most of the extra time? ((0.5)) > 04007 DM No II thought no I actually woke up at about er quarter to eight and I could've got on the bus I could have made it and I thought oh bloody hell I'll stay in the house. 04008 AF Mmm. ((1.5)) Here, the speaker may have in mind one specific "the 8.30 a.m. number 87 bus going south, that I always get", but he does not seem to expect that to be of importance to AF. In implicit reference, when nonanaphoric general nouns occur with definite articles it is usually again with nonspecific referents. In (32), the speaker most likely does not have in mind specific "things", when talking about what Melvyn Bragg, the television interviewer, said about world history on the chat show: (32) 11082 NM Doom gloom. (1) 11083 NM Noone really challenged him. 11084 NM On this you know. —> 11085 NM Cos I mean since the Romans basically the same things are getting worse and worse. 11086 AM (hehheh) AM either knows the sort of "things" that he means or does not care. Nonanaphoric demonstrative adverbs are often with a nonspecific meaning of "wherever"; that is to say, the speaker has no particular location in mind and nor is it important, it would seem. In (33), speakers are discussing holidays in the Mediterranean and saying how polluted the sea is: (33) 19038 CM 111 stick to the pool. 19039 NF But Malaga is not so bad. 19040 NF Actually (0.5) // maybe 19041 BM // Ah that's right. 19042 CM Do you go and take a picture there and then go back to the pool. 19043 NF Yeah, (heh)
72 Analysing the language of discourse communities 19044 BM That's smart. Presumably "there" could mean "anywhere on the coast" or "somewhere outside the hotel". Non anaphoric personal pronouns are also mainly generalised or empty; any outsider would understand them. In (34), BF entertains listeners by dramatising her Hogmanay experiences: (34) 08080 BF Cos we all met up er in the town centre at ten o'clock. (0.5) > 08081 BF And when we got there (0.5) they'd stopped serving. ((!)) 08082 AM Ten o'clock? "They" can be assumed to mean "pub bar stafi"; most British adults would understand that. As far as K4 reference is concerned, explicit reference with a definite article tends to be specific, as in (35). DM shifts topic away from tutorial tasks and weekend commitments, to a problem with a lecture: (35) 04131 DM I'm going away this weekend. 04132 DM So had to do it. 04133 BF Yah. 04134 BF I've got to do it as // well. 04135 DM // He made a mistake. 04136 DM And I wanted to show one of you somebody actually. —> 04137 DM I think he made a mistake yesterday when he was putting them up when the diagram was up I think he got one wrong. ((0.5)) 04138 BF Oh I didn't copy them all down. Only ingroup members would know which diagram this was specifically. In (36), speakers would know which "project" specifically is being referred to: (36) 22035 MM I've come for Silvia to give me my book. (1) 22036 MM And you? > 22037 MM How's the er project? 22038 DM Just reading each others' now at the moment. In implicit reference, K4 definite articles also tend to accompany general nouns referring to specific referents. In (37), the tutorial task sheet that BM is looking for is not named as such: (37) 07031 BM //(heh heh heh) So who've I been divided up with? 07032 MM WithX. 07033 MM And she's not here. 07034 MM So you've got the whole damn thing to do. // (heh heh) 07035 BF // (heh heh) » 07036 BM But I haven't got the thingymajog in my em= 07037 BF Are you shsure about that? K4 is radically different from noncourse K areas in that it has a much higher density of non anaphoric nonmodified general course nouns with definite reference and specific referents. In (38), BM starts a new topic with a vague expression referring to something specific: (38) 20024 CM // Just a week // to go. 20025 BM // Oh God right yeah. 20026 CM So that's the end of that. (0.5) 20027 CM Definitely staying resident in Edinburgh till then. ((3.5)) 20028 BM How's it how's your going? » 20029 BM I haven't given you your thing back. 20030 BM Do you still want it back? 20031 DM Yeah.
Grammar of the ingroup code 73 K4 demonstrative adverbs tend to have specific reference too, as in (37) above: 07033 MM And she's not here. in which only ingroup members would know whether "here" meant "here in this pretutorial", "here in the department today" or "here in Edinburgh." On the occasions that K4 nonanaphoric personal pronouns do have nonspecific referents, they refer to "people in general" or "members of staff1 in the Institute or the Department, or the linguists that wrote the papers or books that they are reading; they can be generalised. Speakers in (39) discuss what questions "they", the authorities that set exams, could ask in the exam: (39) 08022 AM But things like this linguistics as well. 08023 AM You know I don't mimind. 08024 CM You still got enough time for that? ((0.5)) > 08025 AM There's not a lot of things they can ask. 08026 AM Cos we haven't actually done it that deeply have we? In (40), speakers discuss what sort of answers "they", those that mark the exams and give the grades, expect in the portfolio and give each other advice: (40) 13162 AM Don't do any reading for it though. (2) » 13163 AM So they expect quite long answers. 13164 AM I mean not like like the exam. 13165 AM Three hours. In (41), speakers discuss when "they", those that control the programme, will publish their decision about moving a deadline: (41) 23019 DM Oh we've got three weeks now (2). > 23020 DM When are they going to tell us whether it is or not? = 23021 BF After the (1) portfolio yeah. More often, K4 nonanaphoric personal pronouns are specific and refer to elements of the course such as projects and dissertations or to members of staff, as in (42), in which students discuss how they used diagrams that a member of staff provided, in an exam answer: (42) 12178 NF How many diagrams did you draw? 12179 AM S:: I did drew one for each urn phonetic sound so it was seven. 12180 NF I did too. 12181 NF Somebody said six because they think the two T are the same and the four others were only consonants. 12185 NF I did seven too. 12190 CM Each your own creation seven times? » 12191 AM No because we had the er ones she gave //us. 12192 AM I just traced that. 12193 CM //Yeah. It can be assumed that "she" is the phonetics lecturer. The specific use of singular inanimate non anaphoric personal pronoun in K4 refers to elements of the course: handouts, tutorials, projects, deadlines. Hence: (43) 15053 CM I forgot to answer your questionnaire. 15054 DM Oh that's all right. 15055 DM // That's OK. 15056 DM That's fine. > 15057 DM It's still the end of this week. 15058 CM //I'm afraid I er. (0.5) 15059 CM OK.
74 Analysing the language of discourse communities The "it" could be a handingin deadline, or it could be a tutorial. DM and CM know.
4.4 FURTHER DIMENSIONS It has been suggested in this chapter that implicitness could lead to exclusivity, to making the dialogues impenetrable to outsiders. A test was devised to determine whether this was so and also to arrive at a clearer definition of "the outsider". The test examined all aspects of implicit language (the ingroup code and the other features). This chapter describes how the whole test was set up but concentrates only on the results as far as the grammatical cues are concerned. Chapter 5 describes the lexical cues in the test, and Chapter 6 the other implicit features. This section ends with a comment about how the recordees themselves, during triangulation, showed that they understood their own grammatical reference, in retrospection. Subjects listened to four dialogue fragments, taken from the beginning of each of the three terms. The fragments lasted for 2025 discourse units and each had only one K area. The number of dialogues from each K area represents the proportion of total time in that K area throughout the course: there are two noncourse K area ones (dialogues 1 and 4), and two course K area ones (dialogues 2 and 3). Subjects wrote answers to one global question aimed to test topic knowledge ("What is the main topic of this dialogue; what are the people talking about?") and thus show whether the dialogue is impenetrable. Then they wrote answers to five specific questions aimed to test reference knowledge: each focused on a nonanaphoric definite noun phrase with a proper or common noun (including general), a demonstrative or personal pronoun, or a demonstrative adverb, as well as lexical features and other features of implicitness. Appendix III contains one of the four dialogues fragments and its questions to serve as an example. There were three groups of subjects. Group A consisted of twelve people living in the South of England. Although the majority did not have knowledge of Edinburgh, they were considered to have noncourse K area 1, knowledge of the world. They were not English teachers. Group B consisted of seventeen English teachers, who worked in IALS but knew nothing of any DAL MSc course. They had noncourse K areas 1, 2, knowledge of language teaching and study, and 3, knowledge of EU, DAL and IALS. Group C consisted of twelve people with experience of an MSc course; they were exMSc students who had done different Edinburgh University DAL courses in years prior to or since the 19912 course. They had all K areas but lacked the shared interpersonal knowledge and knowledge of the particular 199192 group. Quantitative analysis showed that the topic knowledge and reference knowledge increase as in group knowledge increases (see Appendix II), and this was statistically significant. A ttest shows that this difference is significant (t = 2.68, df = 19, p = 0.015). Analysis also showed a pointby serial correlation between topic knowledge scores and reference knowledge scores for each dialogue, not taking Group C into account since they got all topic questions right. The value of the correlation coefficient for the four dialogues together is 0.590, which is significant at df = 28, alpha = 0.05 (onetailed test). The closer the hearers are to the 19912 MSc course, the more they will understand both topic and referring expressions. In the course K area dialogues there is a strong association between topic knowledge and reference knowledge: understanding the cues is more vital to the understanding of the topic than it is in noncourse topics. Course topics are
Grammar of the ingroup code 75 more impenetrable, therefore, than noncourse topics for an outsider to the DAL MSc discourse community. Implicitness and impenetrability are associated. A qualitative analysis of the answers to the reference questions showed differences between the groups. Subjects in Groups A and B misinterpreted a wide range of referring items: non anaphoric personal pronouns, demonstrative pronouns and adverbs, and general nouns. The most frequent wrongly answered questions were ones asking about nonanaphoric demonstrative adverbs. In one dialogue, for example, when AM says "It is probably a bit late now" to study Chomsky, all subjects could say that "now" meant "at this moment", but they could not guess that "now" implied that the exam was too close to start preparing such a large new topic. Personal pronouns "it" and "they" caused many more wrong answers for Group A subjects than they did for Group B. Group C wrong answers were limited to the questions that required shared interpersonal knowledge: proper nouns referring to people. They could appreciate the full implications of pronouns and added details showing their insights. In answer to a question about the referent in the line "Imagine doing another two years of this", one subject did not simply answer "studying Applied Linguistics" but put feelingly "Toiling away studying on a postgrad, course, mounds of reading, essays, lectures, seminars, limited social life, etc. etc." In the triangulation questionnaire given to the six recordees themselves, the questions focusing on the recordees' ability to decode the implicit contextualisation cues showed that the recordees got on average 97% of the questions about reference right. Interestingly, the only question that all recordees got wrong was the one that referred to the words in one dialogue (see (38) above): 20029 BM I haven't given you your thing back. and asked what suggestions the recordees had about what the 'thing1 might be. Even BM himself could not remember what he was talking about. This confirms that when the general noun is non anaphoric, the referent needs to be in the mind of the interlocutors at the time of hearing. In order to confirm that the use of grammatical contexualisation cues was not part of the idiolect of one or two particular recordees, a calculation was made of the percentage of the units containing implicit nonanaphoric definite reference out of each one's total number of units. It was found that AM's conversations had 13%, BM's 12%, CM's 15%, DM's 12%, AF's 15% and BF's 14%. This shows that the grammatical features of the ingroup code are typical of all six recordees and generalisable to all NS members of the year.
4.5 CONCLUSION This chapter has described the grammatical contextualisation cues of the ingroup code: explicit nonanaphoric definite noun phrases such as "the student", "that book" and "John", and implicit nonanaphoric definite noun phrases with the general noun as in "the person" and "that thing", the demonstrative pronoun and adverb as in "that" and "there", and the personal noun as in "he". Increasing knowledge over time is associated with the changes in reference. Taking all K areas together, the density of all nonanaphoric reference increases slightly in the spring term and drops slightly in the summer term. The main change in language over time is the increase in implicit reference, especially in nonanaphoric definite reference and above all in K4. There is an overall increase in general nouns with demonstrative determiners, demonstrative pronouns and adverbs, and personal pronouns.
76 Analysing the language of discourse communities
K area seems to be a stronger influence on the form of grammatical reference than increasing knowledge with interaction over time. K4 dialogues contain a much higher density of non anaphoric definite reference than noncourse K area dialogues do. This chapter has shown that the grammatical implicit contextualisation cues are typical of K4. When the MSc students are talking on course topics and they mention something unrelated to the preceding dialogue, they are likely to assume that their interlocutors have enough background knowledge to be able to identify the referent without needing a posthead dependent. There are marked differences between the K areas in terms of the context and the specificity of nonanaphoric definite reference. In noncourse K areas, reference tends to be situational and nonspecific. K4 referents tend to be specific and in the background knowledge. This increases the implicitness of K4 dialogues. Outsiders could understand noncourse dialogues better either because they can see the referent or because it has generic reference and therefore does not have to be identified precisely. Not so with K4 dialogues: speakers have one specific referent in mind and assume that hearers can find it in the background knowledge of the course. In the autumn term, students can take for granted that the subject of tutorial tasks and background reading will be in most people's minds most of the time. At the beginning of the spring term, students know that their colleagues must be thinking about the examination and the first project. By the summer term, the students can assume that they are all thinking about the subjects that they choose to concentrate on for the portfolio and the subject to do the summer dissertation on. As common knowledge of the course and shared interpersonal knowledge grow over time, so referents are referred to increasingly in an implicit and vague way. As far as the function of implicit contextualisation cues are concerned, they are a reflection of common knowledge and an indicator of ingroup membership. Students use terms with highly contextualised meaning and implicit reference because nothing more is needed in order for colleagues sharing the same knowledge to identify the referents. Using implicit contextualisation cues is convenient for them because they obey the cooperative maxim of quantity: give no more information than is needed (Grice, 1975). Students obey the law of least effort: "man only makes as much effort as is needed to reach the goals that he has set himself1 (Martinet, 1970: 177). It may be, however, that the students make a conscious choice to refer to referents in their shared context in an implicit way in order to claim intimacy, as a strategy for claiming ingroup membership. To speak the language of the group is to be accepted as a member of the group. Brown and Levinson (1978: 110) say that ingroup language is an "ingroup identity marker" that a speaker uses to claim ingroup membership with the hearer. They say that the use of pronouns where the referent has not been made explicit is typical of positive politeness. Tannen (1989: 23) says that "the more work ... hearers do to supply meaning, the deeper their understanding and the greater their sense of involvement with both text and author". Labov (1972) and Gumperz (1982) say that the function of social group codes is to bring about group cohesion by members showing their superiority over another group and actively excluding outsiders, using their context dependent language. MSc students do not intend to exclude any outsiders, and yet the impenetrability test has shown that outsiders can be excluded because of the implicit language.
5
LEXIS OF THE INGROUP CODE
5.1 INTRODUCTION Overview This chapter continues the study of the implicit language. Chapter 4 outlined the grammatical implicit contextualisation cues of the ingroup code; this chapter explores the lexical ones. Gumperz describes the use of contextualisation cues as consisting of "code or styleswitching strategies, of formulaic utterances and other lexical options which affect the inferential process by recalling interpretative schemata or suggesting how message parts can be linked to create a thematic whole." (CookGumperz and Gumperz, in Grimshaw, 1994: 381). The importance of lexis in discourse analysis has only recently been recognised. McCarthy and Carter (1994: 1045) point out that "in spoken discourse analysis, vocabulary has often been relegated to a minor role in the features that combine to create the higherorder patterns of texttypes and genres /.../ vocabulary choice is just as discourse sensitive as the grammatical choices". Lexical analysis has a place in the present study of reference because it forms part of the study of the relationship between the ingroup's common background knowledge and its language. The relationship between a community's knowledge and its mental or communal lexicon (Clark, 1997: 580) has been explored by many linguists. Clark explains that "With contextual expressions,
78 Analysing the language of discourse communities reference to the speaker's and addressee's common ground is mandatory" (1993: 337), and that the sense may be conventional within one community. Brown and Levinson (1978) argue that the use of ingroup codes language, dialect, jargon, local terminology assumes that interlocutors understand and share the associations of that code. Swales (1990: 32) includes lexical choice as part of his definition of the academic discourse community, explaining that "somewhere down the line, ... understanding the rationale of and facility with appropriate genres will develop, control of technical vocabulary in both oral and written contexts will emerge". When Kreckel (1981: 36) states that the language of university students consists of "a multitude of ingroup codes, discipline specific and social group specific ... taking discipline or groupspecific knowledge for granted", she refers to both grammar and lexis. Lexical analysis also has a place in the present study because it can be analysed from the point of view of the implicitness. The most important assumption is that the choice of lexeme can contribute to the implicit nature of the ingroup code because the lexemes carry different degrees of explicit information. Implicit and vague lexis has been touched on in the literature; the gradability of implicitness has not. The lexical categories devised for the analysis of the data that are at the more implicit end of the cline are based in part on categories described in the literature. The study does not deal with vagueness in the sense of Lakoffs (1972: 195) words "whose meaning implicitly involves fuzziness" (eg: "sort of), Crystal and Davy's (1975) collectives (eg: "bags of) and number approximations (eg: "about 30"), Kennedy's (1987) vague approximation devices of quantification, or Channell's (1994) vague additives (eg: "around ten") and vague implicature (eg: "1,500 died"). It deals with vagueness in the sense of Ullman's (1962) generic words ("bird") referring to classes of things, Crystal and Davy' dummy nouns expressing total vagueness (eg: "thing"), and Channell's vague "placeholder" words (eg: "thingy" and "whatsisname"). General words are at the most noncontentful end of the cline because general nouns are on the borderline between a lexical item and the personal pronoun (Halliday and Hasan, 1976: 274); this is why they feature as both a grammatical and a lexical cue in this study. Likewise, general "do" verbs are on the borderline between the lexical verb and the substitute. Channell (1994) has explored the most implicit lexis in depth. She affirms that an expression or word is vague if a) it can be contrasted with another word or expression which appears to render the same proposition, if b) it is "purposely and unabashedly vague", or if c) the meaning "arises from intrinsic uncertainty". Although her study concentrates on approximators and vague quantifiers, she does devote some attention to words and expressions that refer vaguely to categories when she studies what she calls "tags" (referred to in Chapter 6 as "vague fillers") such as "or something", "and things" and "and so on". She says, "Hearers and readers need to draw on pragmatic information in order to identify the intended vague category. They use in particular: a) the surrounding linguistic context; b) the purpose of the text or conversation; and c) their world knowledge." (Channell, 1994: 143). Her analysis of vague expressions "shows that their meanings are themselves vague", that "speakers share knowledge of how to understand them" and that "it is apparently impossible to describe their meanings independently of consideration of context and inference" (p. 1968). She states that hearers can go beyond the given information because they share pragmatic assumptions about likely category members.
Lexis of the ingroup code 79 Fodor (1977) claims that every word is vague in some way, because for any noun, one's knowledge of the referent may be extensive or quite fragmentary. This chapter discusses the gradability of contentfulness from proper nouns and technical terms, to generalised nouns with implicit pragmatic meaning, to general nouns and verbs. The chapter explores the coexistence of the formal, disciplinespecific technical vocabulary along with the casual, vague form of reference that is the main interest. LeckieTarry (1995: 127) says that on the cline of explicitness, words which are the most precise are the most explicit, for example technical terms and terms with a high degree of modification. Yet it was hypothesised in the present study that the implicit language would contain elements of both written and spoken language, and that technical terms have implicit meaning for the group. Some proper nouns have a metonymical meaning that only ingroup members would understand. Some common nouns in course topics have a generalised range of meaning and acquire uniqueness or a limited pragmatic meaning in the context of the MSc course. Before the discussion of the categories of analysis, it should be pointed out that the analysis of lexis did not take into account lexical cohesion. Cohesion was analysed from only the grammatical point of view despite Hoey's (1991: 9) assertion that lexical cohesion is "the single most important form of cohesive tie". He would have us tag only the lexically cohesive, when analysing texts, on the grounds that "it is the lexical links that dominate cohesive organisation" (ibid: 74). Hoey claims that the function of demonstrative modifiers is to draw attention to the giveness of some stretch of language and that "if that stretch of language proclaims its own giveness by being a lexical repetition or paraphrase, then to note the modifying demonstrative as a separate repetition would be double accounting". In this study, the tagging of cohesion was limited to the grammatical analysis because it was combined with the categories for definiteness and explicitness, and because the coding system tagged the different types of modifiers and also included pronouns.
Lexical Reference Categories For the purposes of analysis, all nouns were divided into two macrocategories: noncourse nouns and course nouns (see Figure 1). Course nouns are those which relate to activities and experiences in the course, ones that depend on K4 to identify the course referent. Noncourse nouns can feature in K4 dialogues, just as course nouns can slip into dialogues that are predominantly noncourse K area dependent. The course nouns were again divided into two groups: 'special course nouns' and 'coursebycontext nouns'. Under the 'special course nouns', there are two subdivisions: common nouns and proper. There are four common noun categories 'technical', 'unique', 'limited range' and 'general'. Nouns were tagged 'technical' if they were intrinsically specialised terms independent of the context of the course. These were terms from linguistics and language teaching theory, such as "diglossia", "lexical syllabus" and (1) 10028 CM I think there's there's going to be no morphology and no and no Xbar. Nouns were tagged 'unique' special course common nouns if they were generalised nouns with specific pragmatic meaning, used to refer to intrinsic components of the course, used in department organisation, mentioned in the course handbook, etc. of which there was only one, such as "dissertation" and "the examiner". The term 'unique' means unique in the context of the course. (2) 13143 CM The portfolio is on is in May. ((5))
80 Analysing the language of discourse communities 'Limited range' special course common nouns are generalised nouns with specific pragmatic meaning that constitute an intrinsic class of components in the course. Limited range nouns are those that have a limited range of meanings in the context of the course. They are count nouns whose meaning is not expressed explicitly since they constitute the second noun of a twoword phrasal expression (Huddleston, 1988: 103) in which the first word (usually a nominal prehead modifier) is understood. Examples are "your book" belonging to the limited range of nouns "books" and meaning "your syntax book", and "the class" belonging to the limited range of nouns "classes" and meaning "the semantics class". In (3), "the paper" belongs to the limited range "papers" and means "the SLA project paper": (3) 20020 CM SLA class next week and the paper's due in next Friday. Clark (1993) would call these nouns "indirect nouns", which he explains are contextual expressions with shifting senses dependent on time, place and circumstances, such as "one water" which could denote a tub, a drop, a teaspoon or a person who ordered water. These nouns are intrinsically courserelated, as distinct from nouns such as "mistake" and "diagram" that were tagged as coursebycontext. Figure 1; Analysis model for all nouns Non-Course Nouns Course Nouns Special Course-by-context Common Proper Common Proper Common general technical actual unique metonymical limited range general The special course 'general noun' category is structurally the same as the general non course noun. General nouns were divided into two minicategories: discourse general nouns and lexical general nouns, in order to keep the former out of the analysis. The term 'discourse general nouns' refers to general nouns in expressions that do not have a referent but that serve a function in the discourse. The function may be introducing a new idea or suggestion, or a contrastive point of view, in expressions such as "The thing is..." meaning "The following is important". In (4) for example, CM introduces his views on how to buy a computer: (4) 06011 CM The best thing to do is use a chemaster card or something. Another discourse general noun function is that of introducing an example, or holding on to a turn while formulating thoughts. For instance, the expression "things like", means "for example", as in (5) from the hypnotist narrative: (5) 02008 MM It's just that I met someone afterwards and they said that what was happening was things like em (1) like at oneone point they got these glasses. The last of the discourse general noun functions is that of hedging and softening, or simply filling the end of a sentence, in expressions such as "and things like that" and "and stuff' that mean "etc.", or simply "um". Take (6) from K4, in which DM tries to formulate his feelings about exam essay questions: (6) 10095 DM Well that's the other reason I won't do an essay question. 10096 DM Just the sort of thing er physically memory=
Lexis of the ingroup code 81 Under the 'special course proper noun1 heading there are two subdivisions: the 'actual intended use' of the proper noun and the 'metonymical use'. In the case of the 'actual' use, the proper noun is used to refer to the writers, schools, etc. themselves. The proper noun may be the name of writers, such as "Chomsky" and "Lyons": (7) 07090 BF Conclusion drawn by researchers and by Fasold. Other proper nouns are: sciences and schools of thought, such as "Descriptivism" and "Structuralism", books and journal titles, such as "ELTJ" and "Observation in the Classroom", parts of Edinburgh University, such as "King's Buildings" and "David Hume Tower", members of staf£ such as "XX" and "XY", and students, such as "X" and "MM". Included in this group are the names of students' family members: spouses, partners and children, such as "Dave" and "Michelle", because course members could be expected to know them. The other special proper noun subdivision is for the 'metonymicar use, in which the noun refers elliptically to something other than the course or linguist named, something other than the literal meaning of the words used. As with the limited range nouns, these nouns can be seen as one word of a twoword phrasal expression. In the case of the metonymical use, it is the second word that is understood. Proper nouns with metonymical use are MSc course names, such as "Syntax" meaning "Syntax course exam question", and "Language and Linguistics" meaning "Language and Linguistics option tutorial". Linguists' names are used metonymically too, for example "Channell" meaning "Channell's paper". In (8), (8) 13001 AM It's a real text book. 13002 AM It's not like Brown and Miller. "Brown and Miller" means not the linguists themselves but the book that they wrote. As Gibbs (1987) says, context plays a crucial role in the interpretation of metaphoric language. The second main group of course common nouns is that of the 'coursebycontext' nouns. These are not intrinsically courserelated but become courserelated by their use in course topics. To take an example: the referents of "discussion" and "this week" are courserelated when the "discussion" is one in a tutorial and "this week" means "this week in classes". In (9), "pressure" means not "pressure of life in general" but "pressure from the staff and work in the course": (9) 07011 BM There's a lot of pressure isn't it? Keeping the coursebycontext nouns separate from the other common course nouns allowed the intrinsically course common nouns to be studied in isolation. Including these terms that are not intrinsically specialised, within the special terms category would have meant clouding the analysis of special terms. Courseby context nouns have a category of their own within the course nouns category and separate from the noncourse nouns, because an understanding of their meaning depends on an understanding of the course nouns. Their case is similar to that of the technical terms that exist independently of the course and yet are in the course noun category. The second macrogroup is 'noncourse nouns', or nouns with referents outside the course (see Figure 1). These too are divided up into 'common' (e.g.: "budgie", "government") and 'proper' nouns. The proper noun category contains countries, towns and cities (e.g.: "Japan", "Manchester"), streets and buildings (e.g.: "Leith Walk", "Cowgate"), famous people (e.g.: "Julia Roberts", "Ballesteros"), names of products, companies and institutions (e.g.: "IBM", "The Education Dept"), world incidents (e.g.: "The First World War", "The Gulf War"), and times of year, months and days of the week (e.g.: "Christmas", "Monday"). Noncourse common nouns and coursebycontext nouns are not divided up into subcategories 'technical', 'unique' and
82 Analysing the language of discourse communities 'limited range' as special course common nouns are. Technical noncourse nouns are related to computing and occur mainly in K2 and K3. There is not a separate category for them because since the study focuses on ingroup words, technical words that can be part of any person's common knowledge of wordprocessing are not of as much interest as those specialised technical words that are related to linguistics. Computing technical nouns occur in only two dialogues, whereas linguistics technical nouns are spread throughout all the recordings. Coursebycontext nouns cannot be technical, by definition. If courserelated nouns are technical, they are classed as technical special course nouns. The verbs are divided into two categories: the 'all verbs' category and the 'general "do" verb' category (see Figure 2). The analysis of "do" verbs was based on Halliday and Hasan's distinctions between verbal substitute "do", lexical verb "do", general verbs "do", "make" and "have", proverbs "do" and "happen", and verbal operator "do" (1976: 117129). Halliday and Hasan rightly observe that these "are all related to each other and form a continuum, or at least a cluster, of meanings that shade into one another at the edges, yielding various indeterminate instances". The verbs "Do", "make", "have" and "happen" verbs are grouped in either the 'general "do" verb' category (if they are noncontentful verbs that depend pragmatically for their meaning on the text), or the 'all verbs' category (if they are verbs that have contentful meaning independently of any text). Of these categories, the 'general "do" verb' category, which contains more than Halliday and Hasan's (ibid.) general verbs, is the main interest in this study. Figure 2; Analysis model for all verbs All Verbs All non"do" verbs Be and Fillers such as "I mean" and "you know" Proverbs "do" in endophoric contexts General verbs "do"/"make"/"have" in fixed expressions General "Do" verbs Proverbs "do'V'happen" in exophoric contexts. General verbs "do"/"make"/"have" not in fixed expressions "Do" verbs in the all verb category, which consist of the proverb "do" in endophoric contexts and the general verb "do" in fixed expressions, are not of interest here because their meaning is not highly contextualised or implicit. The proverb "do" (and "happen") is considered 'all verb' when it functions in combination with reference items such as "it" and "that" as a carrier for anaphoric reference items and also when it occurs in cataphoric contexts. (10) contains anaphoric reference with proverb "do"; the students are discussing the way to write an introduction to a project: (10) 17037 AF Can't you sort of say later on em these subjects will be revised briefly? /.../ 17040 NF I did that. (11) is an example of proverb "do" in a cataphoric context: (11) 02110 MM And so what was happening was that it really got the audience going. This proverb "do" is in the 'all verb' category because it means the same as the verb that it links with. Included in this category too are the general verbs which "express a process in a nominalised form, by means of an empty verb plus its object" (Halliday and Hasan, 1976: 109), in fixed expressions such as "make a mistake" and "have a conversation" because these have a conventionalised meaning as a unit such as "to err" and "to converse", regardless of context.
Lexis of the ingroup code 83 The verbs in the 'general "do" verb' category are the empty, noncontentful ones with no meaning out of context. They are the proverb "do" in exophoric contexts and the lexical use of Halliday and Hasan's general verb "do" that is not in fixed expressions. The proverbs "do" and "happen" here refer to the whole prepositional element and stand for an unidentified or unspecified process, as in "What was she doing?", "Nothing's happening" and "What am I to do?". The understanding of the proverb "do" standing for an unidentified or unspecified process depends on knowledge of the context. In the following example: What're you doing? (12) 09027 AM 09028 AM LPD?= the listener knows that the speaker is asking "What option courses are you taking?" and not "What are you doing sitting there with your file open?" or "What are you planning to do with your family over the Easter break?" This is what Clark (1993) calls proact verbs "do", whose meaning depends on time, place and circumstances. The lexical verb "do" is that which functions as a transitive verb, a lexical item "with generalised meanings" (ibid.: 125) such as "do a book" that can mean "read a book" or "write a book", or "do a job" which can mean "complete a job", depending on the context. The ability to express the meaning with another verb depends on appropriate knowledge of the context, as in, "do a subject" which means "answer a question on the subject", "write a project on the subject" etc.
Analysis Of Lexical Reference During casual observation of the MSc students interacting, the researcher gained the impression that at the beginning of the spring term, they used proper names and terminology a great deal, probably as a result of reading and discussing thinkers and their thoughts, and preparing for exams and projects. It appeared that in the summer term, the names and terms were still there but that there were more vague nominal and verbal expressions as well. The researcher also felt that dialogues based on K4 had more general words and course special terms and proper names than language in dialogues based on noncourse K areas. She wanted to discover whether the overall increase in implicitness in grammar and the higher density of grammatical implicitness in K4, was reflected in a similar pattern in implicitness in lexis. To prepare for the data analysis, each noun was given a number tag that indicated whether it was courserelated or not courserelated, and whether it was proper or common. Prepositionand noun phrases in fixed expressions such as "in terms of and "in some senses" were tagged as adverbial adjuncts. Each verb was given a number tag to indicate whether it was an 'all verb' or a 'general "do" verb'. Verbandnoun expressions such as "have a look" and "bear relation to" were tagged as 'all verbs'. All components of the verb were classified as one verb: thus "I forgot to change it" and "I keep thinking" were tagged once as "change" and "think". Some "Do" verbs were not analysed or put into any category because they are grammatical: these are the verbal operator "do" and the substitute "do". The verbal operator "do" is "a purely grammatical element whose function is to express simple present or past tense" in interrogatives, negatives, marked positives and short answers (Halliday and Hasan, 1976: 127). Verbal operators "do" serving as prompts or backchannels are not included either, such as in the following: (13) 02046 AM And they had to come down this aisle (0.5) in the middle of the audience kind of. ((!)) 02047 BM And they did.
84 Analysing the language of discourse communities The verbal substitute "do" as in "She's never sung in England. She has done in France" is grammatical, being necessarily anaphoric. The following is an example from a dialogue about diagrams drawn in an exam answer: (14) 12179 AM S:: I did drew one for each urn phonetic sound so it was seven. 12180 NF I did too. The categorisation of lexical items was affected by the fact that the spoken data is interactive and unplanned. Rephrasings by one speaker and incomplete verbs were not analysed, in order to concentrate on verbs with completed and analysable meaning. When speakers repeat their own words as a stutter or in a false start, as they hesitate, think what they are going to say and reformulate their ideas, it was counted as one token of the noun. On the other hand, when the same noun or verb was repeated over several discourse units, by different speakers repeating their interlocutors' words to show solidarity, check comprehension and negotiate meaning, each occurrence of a noun was counted. In (15) for example, 'project1 was counted as three tokens: (15)—» 29079 DM Yeah but this is for your dissertation or (0.5) are you talking about a project? 29080 AF Yeah. > 29081 DM You're talking about a project. > 29082 AF I'm talking about a project. In the data analysis, the percentage of each noun category out of all nouns, and the percentage of each verb category out of all verbs was found in each K area in each term. Nouns and verbs were subjected to a qualitative analysis to determine the difference between the K areas in terms of tangibility. The three people who tested the noun part of the code for intercoder reliability got the results 82%, 77% and 60%, which suggests that this part of the code could be used by any other coder and produce similar results to the ones in this study.
5.2 CHANGES IN LEXICAL REFERENCE Nouns When the hypotheses about course proper names and technical terms were tested that in the spring term students used more, and that K4 dialogues had more than noncourse dialogues it was found that they do increase slightly in the spring term but then they decrease in the summer term (the autumn term: 1%, the spring term: 3% and the summer term: 2%), and that K4 dialogues do have, not surprisingly, a greater density than noncourse K areas (see Figure 3). Taking all the K areas together, then, these nouns do not continue increasing or even level off as it seemed on observation. However, analysis of the noncourse K areas separately from the K4 shows that the density of nongeneral special course nouns does increase in both macroK area types, and most dramatically in K4. The overall drop in course nouns in the summer term can be explained by the fact that K4 dialogues occupy 57% of the data in the spring term but only 27% in the summer term. It must be concluded that the main influence on the density of course nouns is not increasing knowledge over time, but the K area and the course events in the spring term. Figure 3: Density of all course special nouns out of all nouns
Lexis of the ingroup code 85
The analysis was continued further by examining each type of special course noun in each K area. As far as noncourse K areas are concerned, the only ones are proper names with their actual use. In Kl and K2, the proper names are principally those of course members' families, course members and staff. In (16), from Kl, AF describes an outing with her son: (16) 21033 AF Actually I was up at the City Cafe. /.../—> 21035 AF Heaping plates of chips into Julian. 21036 AF They do very good chips there actually. 21037 DM Where's that? 21038 AF The City Cafe on Blair Street. (2.5) Here, DM may not know The City Cafe, but he does not ask who Julian is. This dialogue comes towards the end of the course; AF has talked about him before to DM. At the very beginning of the year, in a Kl discussion about busroutes, she brings him into the picture: (17) 04038 AF It's very confusing where we live cos just near Julian's school the thirty four and the thirty five both go in both directions. 04039 AF If you get the wrong one it branches off. // (heh heh) 04040 DM // Right. // Mhm. DM either knows at that point who he is, or deduces from the context of "where we live" and "school" that he is AF's son. K3, being closest to K4, has not only proper names of course members and their families but also names of members of staff and university buildings. In (18), students discuss what DAL members of staff have said about the IALS scholarship and job hunting: (18) 25011 BF And she was saying (1) why don't I apply for an IALS grant and just teach there part time. 25012 CM Yeah. ((!)) 25013 BF I hadn't even considered it. ((4)) /..../ 25033 CM As YY said 1 said I said I'm just going to get my reference to IALS and he said well that'll all change when you start looking for jobs you know. ((3)) 25034 NF Mm. (heh) 25035 NF No way. ((4)) 25036 BF If I could find something I was interested in for two years say part time. Here, all three students either mention IALS or react to "YY'"s comment without question, which suggests that the referents are in their common knowledge. The density of proper nouns in
86 Analysing the language of discourse communities noncourse K dialogues increases over time, most likely as a result of the increasing shared interpersonal knowledge of students' families, members of staff and university buildings. They are densest in Kl dialogues: discourse units with proper nouns comprise about 10% of all shared interpersonal knowledge discourse units in K2, K3 and K4, but 50% of shared interpersonal knowledge units in Kl. (19) is an example of a shared interpersonal Kl dialogue; it demonstrates the density of the proper nouns, as well as showing how the names and the places run through the conversation with an ease that reflects the shared experience. BF, AM and NF discuss an evening that they had started together and then continued separated into groups: (19)> 11127 BF We went (0.5) back to no we went to the PGSU. 11128 BF After that. (1) 11129 BF I stayed there till about eleven. 11130 BF I don't know about the others. 11131 BF NF was there. 11132 BF Did you stay latish? /.../ 11135 NF Yesyes em. (1) 11136 NF We went to a disco afterwards. 11137 BF Oh there you are. 11138AM You went to a disco! /.../ 11142 BF Did em DM and Michelle go to the disco? /.../ 11144 NF Yes. // (heh) 11145 BF // Cos they were going down to Worthing the next day. K4 has a high density of lexical items with implicit meaning. The fact that 43% of all conversations have course topics and that 62% of all K4 nouns are special course nouns means that just over a quarter of the nouns in all conversations are special course nouns. The density of the unique terms is lowest (0% in the autumn, 2% in the spring and 0% in the summer), because the number of nouns possible in the unique group is small anyway. The most dense are the limited range common nouns and the proper nouns (see Figure 4). Just as K4 has a high density of implicit grammatical reference, so it also has a high density of lexical items which, when analysed from a pragmatic point of view, could also be seen to have implicit meaning, only accessible to those with knowledge of the course. Proper nouns seem to be responsible for the overall increase in the density of special course nouns over the three terms. Within the category of proper nouns with their actual use, the densest of course nouns in K4, the most frequent are those which refer to linguists, books and journals, and sciences and schools of thought. An outsider to the course might be familiar with these but not know their particular significance within the context of the course. Ingroup members know what course option the linguist or school of thought is connected with, and what course task the books and journals were recommended for. In (20), BF asks about a book that appears to be in front of the speakers. Here she has to specify that she means the book by Lyons before her colleagues can give her the answer that she needs. She appears to be looking at a book: (20)> 14001 BF Semantics One. > 14002 BF Is there another Semantics? (1) > 14003 BF Is there another book//called Semantics by Lyons? 14004 NF // Yes. > 14005 DM //YeahTwo. 14006 NF // Two One and Two.
Lexis of the ingroup code 87 Hitches in communication are discussed in the next chapter. Figure 4: Density of nongeneral special course noun types out of K4 nouns
One of the most implicit ways of referring to course components is to use the proper name of the component without mentioning the component itself. This is the metonymical proper noun, the third biggest group, densest in the spring term. Students use course names to refer to an exam question or a lecture. In (21), students refer presumably to lectures or tutorials for the courses that they name: (21) 09003 NF Are you going to Stylistics? 09004 CM No. 09005 NF Oh. /.../ 09008 CM Oh I've got (1) Second Language Acquisition and Language Programme Design. 09009 NF Ah. ((12)) They can refer elliptically to a tutorial homework task by using the name of the course for which the task was set as in: (22)> 04124 BF Has anybody done their Syntax? ((1)) 04125 DM // I did it yesterday. 04126 DM Oh that's what I was doing yeah. > 04127 BM // No I haven't looked at it yet. Here, DM knows that BF is not enquiring about a tutorial or lecture. In (23), they seem to be referring elliptically first to exam questions and then to revision of materials for the exam: (23)» 12001 AM So you didn't do the the Grammar or the Semantics? 12002 CM No. > 12003 CM I studied the Semantics quite a bit and then I didn't do it. 12004 AM Yeah. 12005 AM same thing here. 12006 AM I just you I mean first question was it something like father of forty forty? 12007 AM That was the grammar one. ((!)) 12008 CM Oh yeah. 12009 CM I didn't even look at it.
88 Analysing the language of discourse communities AM can then go straight back into his analysis of the exam question and CM knows that he is not talking about revision any more. Students use the names of linguists to refer implicitly to their theories or papers. In (24), they are not talking about the linguist himself but, on the first occasion, about revision of course material on his theories for the exam, and on the second, about information given in course lectures about his theories : (24)> 08031 AM Though though I haven't I haven't done any Chomsky. 08032 AM Probably a bit late // now. » 08033 CM // Chomsky doubles up in Psycholinguistics. In (25), they are not talking about Tolwood as a person; only an ingroup member would know if they refer to a book, an article, a theory, etc.: (25) > 15204 DM I didn't like the Tolwood at all. 15205 DM I thought it was a bit (0.5) a bit much really. ((0.5)) 15206 BM It wasn't er it wasn't reflected anywhere. Limited range nouns, referring to components of the course in such a general way that the full meaning is implicit, constitute the next biggest group of course nouns in K4. They are only ever qualified by a demonstrative or possessive determiner. The course component may be an organisational one, such as a "term" or a "group" as in (26) in which "group" refers to the "pre tutorial group for discussion of tutorial tasks amongst students before the tutorial itself with a member of staff1, of which there are a limited number: (26) 03006 NF I'm coming ((0.5)) 03007 NM Mhmwithus? 03008 NF I think so. 03009 NM // With us? » 03010 BM // No no you're in X's group I think. 03011 NF Oh yes sorry. 03012 NF I keep thinking you're X. The course component may be a piece of work, such as an "exercise" or a "project" as in (27), in which the "project" is the third project in the course and specifically the one that BM is doing: (27)> 17002 CM How's your project going? 17003 BM ((unintelligible: whispered)) I tried to write the introduction over the weekend. 17004 CM Yeah I always write those last. 17005 NF It took me a whole afternoon to write a few sentences. 17006 NF And then I didn't like it. (heh heh) 17007 AF Oh it's awful when you start considering what you're writing. From this exchange, it is evident that they are all familiar with the "project" at hand. All the coursebycontext nouns cluster around CM's limited range noun: BM mentions his "introduction", NF her "sentences", and AF refers vaguely to the same with "what you're writing". The course component may also be something that has been suggested to the students for reading, such as a "book" or a "paper". In (28), BM offers DM an "article", which, it would seem, they have both read and both need because they appear to be able to evaluate and apply it to their project: (28) »15194 BM // What about the article? 15195 BM I wasn't really using it. 15196 DM I could quote from it actually. 15197 DM I definitely I might do that.
Lexis of the ingroup code 89 15198 BM Do you think it's pretty easy? 15199 BM It's pretty it's scurrilous isn't it? 15200 DM I think it is awful. In none of these examples is it necessary for the students to specify which "group", which "project" or which "article" is being referred to: the exact identity of the referent is implicit and ingroup members know what the range of possible referents is limited to. Technical terms, which occupy a small percentage of all nouns in K4, are used in the context of the tasks set in the course. In (29), AF is telling DM about her project. She uses terms with which they are both obviously at ease: (29) >29030 AF Well the thing I'm most interested in is pidgins and Creoles but I don't know that it's practical for a project. 29031 AF Um. 29032 DM You might get a trip somewhere 29033 AF (hehheh) » 29034 DM Where are you going to find pidgins and Creoles? 29035 AF Well I don't know. ((2)) 29036 DM There's a lot been written on it. Although outsiders to the MSc might know the term if they were linguists or language teachers, they could not know exactly what had been said in the course about these technical matters, so they would not know exactly what these technical terms meant to the speakers concerned. There were no hypotheses about coursebycontext nouns and noncourse nouns, but examples of these were studied to see whether they could be said to contribute to the implicitness of the in group language. The main difference between coursebycontext nouns and noncourse common nouns and is that while the former refer mostly to second order entities (Lyons, 1977), or events and processes that occur, the latter refer to predominantly concrete, first order entities, or physical persons and things that exist. This adds an element of intangibility to K4 dialogues. Whereas outsiders might easily imagine a physical thing and thus understand the conversation in Kl, in order for them to imagine the sort of event or process that occurs in K4 they would have to be part of the ongoing experience of the course. A few examples from each K area will demonstrate the different degrees of abstractness and intangibility. Within noncourse K nouns conatin different degrees. Kl nouns are nearly all concrete and tangible. Because many Kl dialogues tell of the speaker's daily life, nouns refer to "the doorbell", "a babysitter", etc. Since many Kl dialogues are narratives, the nouns identify concrete entities and participants in the stories such as "buses" and "the lady upstairs". The nouns related to entertainment and outings are necessarily physical: "plates of chips", "pubs", etc. Descriptions of outings and experiences in the world require mention of the physical surroundings such as "islands" and "sand". The physical descriptions are often parts of the body: "head", "knees", etc. Narratives and descriptions require reference to actions and events, which are a little more abstract. Nominalisations of verbs, such as "sunbathing" and "hillclimbing" abound. Narratives and descriptions require reference to moments in time, such as "hour" and "month", which might be classed as second order entities in that they are conceptual. K2 has a lower proportion of nouns referring to first order entities than Kl does, yet the nouns are still mostly tangible. First order entities exist in the realms of schools and work, such as "certificates" and "mastercard". Some of them are less tangible technical entities related to computing: "spreadsheet", "indents",
90 Analysing the language of discourse communities etc. The second order entities tend to be changes in states of affairs such as "inflation" and "innovation", and observable events such as "interview" and "courses". In K3, the proportion of concrete first order nouns and not so concrete second order ones is about 50:50. Nouns refer to first order tangible objects in DAL and IALS: the "coffee machine" and the "photocopier", and parts of daily life in the university: "queues" and "tea". There is a large group of nouns that straddle first and second order entities, in the sense that they are materialisations of intellectual activity, in connection with the IALS scholarship and higher degrees in DAL: "proposal" and "MLitt". In (30), students discuss the IALS deal: (30)» 25022 NF You work for the scholarship scholarship? 25023 CM Well you have to work for it. 25024 CM You have to work four hundred eighty hours. 25025 NF Ah. > 25026 CM They get tuition for the salary. Themes of work, time and money abound here; the nouns "job", "salary" and "years" occur frequently. In K3, there is a small group of third order abstract propositions that express attitudes and evaluations, such as "truth" and "danger". In (31), DM warns BF not to try for the MLitt scholarship if she does not want to, in case she finds herself having to do it: (31) 28046 BF Then I thought I don't want to do it. 28047 DM Yeah. 28048 DM It's a real shame to go for it if you're not interested isn't it really. 28049 BF Yeah. Coursebycontext nouns in K4 are generally abstract second order entities. On the borderline between the nonphysical and the physical is the materialisation of the intellectual in writing. Thus students talk about "introduction" and "words" in projects and exam answers that they have written. By far the largest group of coursebycontext nouns are those that refer to abstract second order entities processes and states. A large number of these are related to time: "hours" and "weeks" as they form part of the course experience. Most coursebycontext nouns relate to intellectual activity, e.g.: abstract nouns such as "consideration" and "perspective", and to the relationship between thinkers, e.g.: "links" and "contacts". In (32), the exact meaning of what is being talked about is "fuzzied" by the abstract nouns, quite appropriate to the discussion of theorists and their theories: (32) 12084 CM Well I think Lyons said that (0.5) that there that Descriptivism er was a particularly American version of Structuralism. 12085 CM And not really related // in any 12086 AM // Well that cos yeah. 12087 AM Co // to me 12088 CM // Developed independently I think. 12089 AM Yeah. (0.5) 12090 AM Yeah. -> 12091 AM Because (0.5) I mean they had the same kind of divide didn't they? > 12092 AM same axis. > 12093 AM // They had the er mental and the ... 12094 CM // Yeah. (0.5) There is, in addition, a large group of abstract, third order entity nouns which are emotive, expressing attitudes and emotions: "pressure" and "mood". In (33), students look forward with trepidation to the summer dissertation:
Lexis of the ingroup code 91 (33)
21109 DM
» 21110 AF 21111 DM
Been trying to think of something that might stretch to twenty thousand words. ((2)) And that you might be able to find endurance for. And is not excruciatingly boring.
There appears, then, to exist a cline of tangibility along which the K areas can be placed, as far as their nouns are concerned: at the most tangible end is Kl, whose nouns nearly all refer to first order entities, and at the most intangible end is K4, whose nouns nearly all refer to second order entities and which contains a substantial number that refer to third order entities. The suggestion being made here is not that abstractness can be directly related to implicitness, but that abstract nouns depend more for their meaning on the context than concrete nouns do. Understanding the meaning of a word such as "consideration" depends on knowledge of setting, participants and topic, just as understanding the meaning of "endurance" depends on knowledge of the quality of the situation endured and the personality and attitude of the endurer. This point is obviously somewhat debatable and needs more investigation than it can be given here.
Verbs Casual observation had prompted the hypothesis that in the summer term, there would be more vague verbal expressions and that K4 dialogues would have more general verbs than noncourse K area dialogues. The density of general "do" verbs out of all verbs taking all K areas together does increase slightly: in the autumn term it is 5%, in the spring term 6% and in the summer term 7%, and K4 does have a much higher density of "do" verbs (see Figure 5). The fact that the overall density increases while K4's density drops dramatically is yet again explained by the fact that K4 occupies much less time in the summer term than it did in the spring term. K4's general "do" verbs frequently collocate with both lexical (special course nouns) and grammatical implicit contextualisation cues (nonanaphoric pronouns), taking them as a direct object (see Figure 6) and this adds to the implicitness of a sentence that is already implicit, because the meaning of the "do" verb depends on the meaning of the special course noun or nonanaphoric pronoun. K4's decrease in general "do" verbs over time may be explained by the collocation with limited range nouns and metonymical proper names, as in "read books in preparation for", "revise" and "answer an exam question on" (see below). By the summer term, the exams are over, and with them discussions about revision and how well they did. Students are on different option courses and there is less comparing of notes and working together. 48% of course general "do" verbs collocate with special course nouns. 18% have limited range nouns as objects, and speakers can slip from meaning to meaning of "do" within the same exchange and still understand each other. In (34), AM uses "do" to mean "answer an exam question on" and then "write", and CM follows the switch: (34)» 12221 AM No I'd1 s I thought I had to best to do the data questions first then you know // how much you've got for the essay. 12222 CM Then you know how // much time yeah. > 12223 AM And I knew I was going to do one essay so. 12224 CM I wanted to have a break between the two essays // so thought 12225 AM // Ah yeah.
92 Analysing the language of discourse communities In (35), the first "do", used by AF, accompanies a technical term and means "studying in order to write an answer to the project question on", and the second, used by BM, accompanies a limited range noun and means "taking", "going to the classes". (35)—> 20001 AF Are you doing discourse for your core project? 20002 BM No but maybe I should. (0.5) 20003 BM Maybe that would explain what it is about. > 20004 BM I'm doing my option on discourse. 20005 NF How do you decide // what you want to? 20006 AF / / I wonder why you really bother. None of his interlocutors ask him what he is talking about. Figure 5: Density of general "do" verbs out of all verbs
Figure 6: Proportion of general "do" verbs with special course terms and non-anaphoric pronouns out of all general "do" verbs in K4
14% of general "do" verbs have metonymical proper names as direct objects; the verbs can have one of a whole series of meanings depending on the context. In (22) above, in the "Has anybody done their Syntax?" and "I did it yesterday", the "done" and "did" mean "worked on the tutorial questions for". In (24), in the "I haven't I haven't done any Chomsky", "done" means not "worked
Lexis of the ingroup code 93 on the tutorial questions for" but "revised Chomsky's theories and all the material on him for the exam question". In (23), in the "So you didn't do the the Grammar or the Semantics?", the do" means neither "worked on" nor "revised" but "write an answer to the exam question". In (36), "do" has yet another meaning in addition to "worked on", "revised" and "answer exam questions"; it is to "take the option course": (36)» 15135 BM You do you do Language Planning don't you? 15136 DM Yeah. 15137 DM I've stopped doing that though. DM understands BMs meaning and uses the same verb himself. To uninitiated outsiders, unit 15135 could have been taken to be an enquiry as to whether he was a language planner by profession, as in "do town planning" or "do language testing". 12% of K4 general "do" verbs combine with general nouns; the implicitness is double since neither the verb nor the noun could be easily understood by an outsider. In (37), students discuss what seems to be parts of a tutorial task, each one referred to as a "thing"; presumably "doing" the thing means "answering the questions" in a particular part of the task: (37) 03073 BM I did em the literary // thing. 03074 BM The literary language one. 03075 AM // I did the I did the business one. Again, AM uses BM's general "do" verb himself in his response. In (38), students compare how each of them have revised for the exam and the "people" would appear to be linguists and other thinkers; thus to "do" the "people" must mean to "revise", "prepare to answer questions on" them. Both DM and CM use both general "do" verb and general noun. This interactive vagueness would seem to be a way of emphasising shared common ground and thus showing solidarity: (38) 10066 DM So did you do everything? 10067 DM // Or sort of choose a few people? 10068 CM // Em. (0.5) 10069 CM No I've done all the people. 10070 DM // Really? The cotext is vague in this example: witness the vague reference of the indefinite pronoun "everything" in unit 10066 and the hedged "sort of in 10067 implicitly involving fuzziness (Lakoff, 1972: 195). 5% of course general "do" verbs have nonanaphoric personal pronouns as direct objects. (39) contains an example: the excerpt begins at the point that DM is talking about ideas for a project that he showed to his tutor XX; the "questionnaire" seems to have had something to do with it: (39) 15045 DM So I typed that thing up again after you'd gone. 15046 CM Oh yeah. 15047 DM But I did get it finished. 15048 DM Went to see XX (0.25) with it. 15049 CM How did it go? » 15050 CM // Ah nuts! I forgot to do it. 15051 DM // He said it was OK 15052 CM Gee. 15053 CM I forgot to answer your questionnaire. Whereas DM seems to be talking about writing the questionnaire in 15045 (the outsider can only guess this since it is referred to by the general noun "thing"), CM's "do" in 15050 is not about
94 Analysing the language of discourse communities designing it but about filling it in. He has to clarify this by using the more contentful verb "answer" in unit 15053. CM's "it" in unit 15050 remains vague because the referent is not named, even if it coheres with the equally obscure general noun "thing"; that the speaker has to clarify its meaning by using the more contentful noun "questionnaire" in 15053. A brief inspection of the most frequent verbs in the 'all verb' category shows that, as in the case of noncourse common nouns and coursebycontext nouns in which K areas differ along a cline of tangibility, K4 verbs tend to be more abstract than noncourse K area verbs. K4 verbs reflect the mental activity of the course. Verbs such as "judge" and "define" contribute to the general abstractness and intangibility of the K4 lexical reference, as do "consist of and "believe" in the following: (40) > 12115 CM They said Structuralism consists of Saussure. » 12116 CM Then they told us what Saussure believed.
5.3
FURTHER DIMENSIONS
It was explained in Chapter 4 that a test was devised to determine whether implicitness could lead to exclusivity, to making the dialogues impenetrable to outsiders. Chapter 4 described the results as far as the grammatical cues are concerned. Subjects in Group A (nonEnglish teachers) and Group B (English teachers who knew nothing of DAL MSc courses) misinterpreted non anaphoric personal pronouns, demonstrative pronouns and adverbs. Group C subjects (exMSc students from previous EU DAL MSc courses) could appreciate the full implications of all grammatical reference. This chapter describes the results as concerns the lexical cues. Subjects in Groups A and B misinterpreted a wide range of lexical items: limited range nouns, proper nouns, and metonymical proper nouns, and general nouns. The most frequent wrongly answered questions were ones asking about limited range nouns and the metonymical proper nouns. These occur in the dialogues quoted from K4. In Groups A and B, but especially in Group A, the lack of understanding of both of these affected the understanding of the whole topic. In one dialogue, for example, subjects who did not know that the limited range noun "group" refers specifically to the "pretutorial study group", and simply suggested vaguely that its referent is "a group of students", could not appreciate what the topic of the conversation was. In this dialogue, the topic is "Who is in whose tutorial group"; the students are settling down to prepare to work on their tutorial task. Subjects suggested that they are discussing "a planned trip somewhere" or "filling in time ... not much content". In another dialogue, Group A subjects did not know that metonymical proper nouns "Language and Linguistics" and "Psycholinguistics" refer to the courses that had these names rather than the field of study itself, and they did not know that the term "Chomsky" refers not to the man himself but to his theories that students had studied. This seemed to make it difficult for subjects to appreciate that the students are talking about course material and that the topic is "What AM and CM studied for the exam". Thus subjects who left the metonymical proper noun questions blank, wrote comments such as "haven't the foggiest idea" or went to great lengths trying to define the sciences, writing for example "modern ways and approaches to researching information for the study of how and what people say", then suggested that the topic was "Chomsky" or
Lexis of the ingroup code 95 "Psycholinguistics, whatever they are". That is to say, these lexical cues were associated with impenetrability. General nouns and proper nouns with their usual use caused fewer problems. In the case of the proper nouns, neither Group A nor Group B subjects could identify the people were who were referred to by name. They differed in that whereas Group B subjects could identify institutions, Group A subjects could not. In one dialogue, for example, most Group B subjects knew "the Council" it was "The British Council", but Group A subjects offered suggestions such as "UNESCO" or asked "could it be overseas council?". The general noun "things" gave a few Group A subjects problems. One subject thought that the "things" in one dialogue were not points in an exam question but "assignment topics". Group C wrong answers were limited to the questions that required shared interpersonal knowledge: proper nouns referring to people. Getting such answers wrong did not affect the comprehension of the topic in the way that not understanding the metonymical proper nouns and limited range nouns did for Groups A and B. Group C subjects showed an appreciation of the fiill implications of K4 nouns: they rightly identified the "groups" as "pretutorial groups", and "Language and Linguistics" and "Psycholinguistics" as "core courses", for example. As far as the recordees themselves are concerned, in the triangulation questionnaire, they all got all the answers on the lexis of their dialogues correct. There was no doubt about the reference. Moreover, once again, it can be said that the features of the ingroup code are not part of one person's idiolect but a characteristic of the ingroup code of all. An analysis of each recordee's conversations throughout the data revealed that they had a similar density of course common and proper nouns and general nouns and verbs: AM had 21%, BM 15%, CM 19%, DM 18%, AF 16% and BF 13%. This adds weight to the claim that the lexical contextualisation cues are typical of the six, the whole MSc group, and indeed of the whole DAL discourse community.
5.4 CONCLUSION This chapter has shown that some nouns have a meaning contextually determined by the course and that this adds to implicit language of the MSc ingroup, making their conversations impenetrable to those without knowledge of EU DAL MScs. These nouns are the general "do" verbs and special course nouns: technical (eg: "diglossia"), unique (eg: "examiner"), limited range (eg: "book"), general nouns (eg: "thing"), proper nouns with actual use (eg: "Chomsky") and their metonymical use (eg: "your Chomsky"). As far as nouns are concerned, the quantitative analysis shows that the most frequent are the limited range, and proper nouns with actual and metonymical use, and the qualitative analysis shows that the limited range and the metonymical proper nouns add to the implicitness of K4 dialogues because their precise referents are not named but implied. There is an increase in nongeneral special nouns throughout the year especially in K4 dialogues in which they are most frequent. K4 also differs from the noncourse K areas in that whereas most noncourse common nouns refer to physical, first order entities, most K4 coursebycontext nouns refer to abstract, second and third order entities, adding a dimension of intangibility. As concerns verbs, K4 has more general "do" verbs than other K areas. General "do" verbs would only be understood by those with knowledge of the course, especially when they take as objects special course nouns and nonanaphoric pronouns, which also depend on
96 Analysing the language of discourse communities knowledge of the course for their meaning. K4 also has a number of abstract verbs, which may contribute to the intangibility of the language. Most lexical choice has a social function. Brown and Levinson (1978) list ingroup code and jargon amongst their "ingroup identity markers" that speakers use to claim ingroup membership with the hearer. Specialised vocabulary from a permanent collocational environment can be used by members of ingroups to express solidarity. Atkinson (1981) discovered that medical students learned medical jargon and ingroup language as part of their socialisation process. The ability to use the terminology correctly was one that developed over time but the conscious choice to use it in casual conversations was related to the desire for ingroup membership. Widdowson (1993) suggests that each social group "closes off' the group through its choice of lexis. As far as the function of lexical implicit contextualisation cues in this study is concerned, it is possible that speakers consciously choose to use technical terms and proper nouns to claim ingroup membership by using the terminology of the specialism. There may be an element of "We know how to use terms and names because we're Applied Linguists, aren't we?" General words, and vague ingroup expressions can also have the socially cohesive function of expressing inthesameboatness. As Halliday and Hasan (1976: 276) say, general nouns express "interpersonal meaning" and "a particular attitude on the part of the speaker". Channell (1994) lists several purposes that general nouns can be put to: avoiding offence or pretentiousness, deliberately withholding information, avoiding showing uncertainty or a lexical gap, and protecting oneself or somebody/something else. This study does not examine the function of general nouns in this detail. It may be, however, that the students make a conscious choice to speak vaguely using metonymical proper nouns, limited range nouns and general words with their implicit meaning, in order to claim ingroup membership. Here, it may be a case of "We won't bother to name these things too carefully: we know what we're talking about, don't we?" The lack of precision also suggests that there is a desire to establish a relaxed, informal relationship between speakers. Chapter 7 contains further discussion of the function of implicit contextualisation cues.
6
IMPLICITNESS OVER UTTERANCES
6.1 INTRODUCTION This is the third and last chapter about the form of the 'implicit language'. Chapters 4 and 5 described the 'ingroup code', or the implicit use of language at the noun phrase and verb level. They showed that, throughout the course, there is an increase in nonanaphoric definite reference and implicit nonanaphoric definite reference with proper nouns, general nouns, demonstrative pronouns and adverbs and third person pronouns. They also showed that when students are talking on course topics, the density of technical terms, unique and limited range nouns, and general "do" verbs is greater, and that the density increases over time. This chapter contains a description of the 'other implicit features', or the implicit use of language at the utterance and exchange level. Taking into account Goodwin's observation that "sentences in natural conversation emerge as the products of a process of interaction between speaker and hearer and /.../ they mutually construct the turn at talk." (1979: 98), the chapter goes outside the boundaries of the noun and verb phrase and looks at how speakers share the implicit use of language between and behind utterances. The features of implicitness reinforce each other textually and stretch together over whole sections of exchange. The assumption of this part of the study was that if implicitness at the noun and verb phrase level increases, especially in course topics, then implicitness at the exchange level is also likely to. The chapter falls into three parts. The first examines clausal ellipsis, or unspoken meaning at the end of utterances. It contains a description of unfinished sentences and shows how hearers have
98 Analysing the language of discourse communities to supply the missing words for themselves. It also analyses the effect of ingroup code in unfinished sentences. The second part examines humorous conversational implicature. This is a case of unspoken meaning behind utterances; hearers have to understand a meaning other than the literal one of the words used. It also looks at ingroup code in humorous utterances. The third part of this chapter explains how implicitness can extend over several utterances. It examines the effect of implicit language on topic shifts, and looks at minor communication breakdowns caused by it.
6.2 CLAUSAL ELLIPSIS The Categories Clausal ellipsis is the term used in the study to cover three types of unfinished sentences, or incomplete sentences at the end of a speaker's turn: that in which the sentence is completed by the hearer or responded to as if it were complete, that which the speaker ends with a hanging conjunction such as "so" and "but", and that which the speaker ends with a Vague filler'. The sentence completed by the hearer is described by Sacks as "an utterance which is a whole sentence followed by an utterance which is not a whole sentence, but a part sentence that is, however, syntactically coherent with the prior, and the two of them constitute a sentence." (1992: 652). Sacks says that the second utterance is often a prepositional phrase or an appendor clause. (1) is an example from the data. AM and CM are checking that they have both understood the theories of linguists that they had to learn for the exam: (1) 12091 AM Because (0.5) I mean they had the same kind of divide didn't they? 12092 AM same axis. » 12093 AM // They had the er mental and the ... 12094 CM // Yeah. (0.5) 12095 CM What you what you know and what you do. 12096 AM Yeah. 12097 AM It was just that Saussure concentrated on the mental and Bloomfield concentrated on the on the // physical. Here, CM's words overlap AM's; he provides a rephrasing of AM's words and adds the second hah0 of what he expects that he was going to say. As WilkesGibbs (1993) says, people coordinate in the making of definite reference and the collaborative process requires actions by both speakers and interlocutors. The sentence completed by the hearer could be expressed in her terms as a 'proxy noun phrase': the speaker hesitates and the addressee offers a noun phrase that they think is appropriate. The second type of unfinished sentence, that which the speaker ends with a hanging conjunction such as "so" and "but" can imply the type of clause that is understood. The existence of a contrasting situation is implied with "but", and the existence of a resulting situation is implied with "so". (2) shows that it is not always obvious, however, whether there is actually an implied situation at all. CM and AM are discussing how to they planned their time in the exam: (2) 12222 CM Then you know how // much time yeah. —» 12223 AM And I knew I was going to do one essay so.... 12224 CM I wanted to have a break between the two essays // so I thought //
Implicitness over utterances 99
In the case of the third type of unfinished sentence, the speaker ends their turn with a Vague filler', an expression with little informational content such as "and so on", "and all that", "and things" and "or something". Channell (1994) calls these items "tags" and Jefferson (1990) calls them "generalised list completers". They "cue the listener to interpret the preceding element as an illustrative example of some more general case" (Dines 1980) and depend partly on common knowledge of the "more general case". In (3), AF explains why DM can go out at the weekend more easily than she can: (3) 21014 AF Yes but you don't have to find a babysitter. 21015 DM Mm. 21016 DM Mm. 21017 AF And you've got somebody there to go out with straight away. 21018 AF You don't have to sort of phone and make arrangements. » 21019 AF And things. 21020 AF And find out that = 21021 DM And find someone yeah. = As it can be seen by this example, sentences ending with a vague expression are only a little more complete than unfinished sentences with a hanging conjunction but they assume that the listener knows what more concrete reference the filler could be replaced with. Not all unfinished sentences and fillers were analysed. Unfinished sentences occurring within the turn of a speaker, or sentences that are interrupted by an utterance not predicting a continuation or responding to them, were not analysed in the study because these sentences are not left unfinished by the speaker thinking that the hearer has enough information to understand the rest. They are unfinished either because the speaker is reformulating their ideas or because the hearer is not interested enough or patient enough to let the speaker finish talking. Fillers that are hesitation phenomena and hedgers such as "I think" and "you know" were not studied because these could not be said to carry a pragmatic implicit meaning. The unfinished sentences selected for study have implicit meaning. In the case of the unfinished sentence that the hearer completes, the hearer predicts what the speaker is going to say because they have the relevant common background knowledge. With the unfinished sentence left hanging with a conjunction, the speaker possibly invites the hearer to join them in thinking about the implications of what they have said, using background knowledge of the context. The speaker using a vague filler such as "or something" might imagine that the hearer can join them in thinking about the implications of what the "something" could be, using knowledge of the context.
Changes In Clausal Ellipsis It was predicted that the density of these three would increase over time, just as implicitness at the noun phrase and verb level does. The density of the features was calculated not out of all words but out of all discourse units, since these generally only occur once in a discourse unit. A general impression was taken of how the features combine with implicit contextualisation cues. Figure 1 (on a scale of maximum 8% so that all features in this chapter can be compared on the same scale) shows that in noncourse K areas the density decreases over time but that in K4, on
100 Analysing the language of discourse communities the contrary, they are absent in the autumn and grow in the spring and summer. It would seem that unfinished sentences completed by the hearer only feature in K4 in the spring and summer because that is when the background knowledge about the course has accumulated. The predicted increase of clausal ellipsis occurred in K4 alone, therefore. Figure 1: Clausal ellipsis In non-course K areas
In K4
Each of the three features was examined taking noncourse and K4 area dialogues separately. In K4, unfinished sentences could only be completed by ingroup members. In (4) from a K4 dialogue about an option course, DM is talking about X, Y and Z, students who went to a tutorial, and when he hesitates, CM knows that he is going to talk about whether X has chosen to do a project about the subject of the course in question: (4) 15006 CM Just X and I were there. > 15007 DM X's doing it for er... > 15008 CM Yeah project. 15009 DM Yeah. ((!)) 15010 CM And er I think Y and Z. 15011 CM I know Z is doing // it. 15012 DM // Z is definitely doing it.
Implicitness over utterances 101 It seems that they are saying the same sentence, that their minds are "on the same wavelength", that they are having the same ideas without having to express them fully. The unfinished sentences with a hanging conjunction are typical of the noncourse K areas alone. No specific information is implied in the unfinished part of the sentence; it could actually be that the speaker has no further ideas in mind. In (5), from a conversation about whether BF should apply for the IALS scholarship, the effect of the "so" might be to soften the negative statement and to imply "therefore it is not worth talking about this any more". DM understands: (5) —» 28032 BF I don't really want to carry on studying so... 28033 DM // (hehhehheh)(l) 28034 DM Are you having second thoughts about your future then? The unfinished sentences that the speaker ends with vague fillers maintain a stable density throughout the year in noncourse K areas, but they only occur in K4 dialogues in the spring and summer. Again, K4 vague fillers add to the implicitness because only a hearer with knowledge of the course could understand the implications. Extract (6) comes from a conversation about the linguistic fields covered in the course that they were revising for the exam: (6) 10001 CM What are you which ones are you concentrating on? 10002 DM On the grammar // syntax. 10003 CM // Grammar sure. 10004 DM Er the grammar syntax and semantics. 10005 CM Semantics. /.../> 10009 CM You're not bothering with going through Bloomfield and all those kind of things?((0.5)) 10010 DM I don't want to write an essay. Here only a course member, who would know what else was covered in the course in which Bloomfield featured, could imagine what "all those kinds of things" are. An analysis of implicit contextualisation cues in sentences containing clausal ellipsis shows that K4 unfinished sentences contain technical terms, course limited range nouns and proper nouns, and to a lesser extent nonanaphoric pronouns. The combination of the cues and clausal ellipsis makes the implicitness seem double. Unfinished sentences with hearer response can cooccur with special course nouns, as in (7) about what the students wrote in answer to the linguistics question in the exam. CM refers to something that a linguist said and leaves the sentence hanging, confident that AM knows how it ends: (7)» 12148 CM And I did qualify it with that Lyons what Lyons said about (0.5) being an American who was trying // to... > 12149 AM // Mm. Mm. » 12150 AM Yeah. ((4)) 12151 CM It did it did throw me a bit. (0.5) 12152 CM I was expecting them more on just the men. 12153 AM But I mean that's what I ended up writing. AM responds sympathetically, as if he could have completed the sentence. When vague fillers occur with special course nouns, only ingroup members could imagine exactly what might be implied. Example (6) above has the metonymical proper name "Bloomfield", and then a vague filler with a general noun, which helps to explain its exclusiveness. In (8), MM enters what seems to be a crowded common room with no empty seats. DM is sitting with AF:
102 Analysing the language of discourse communities (8)
29095 DM Are you wanting here? 29096 MM Sort of. > 29097 MM Are you waiting for a lecture or something? 29098 MM Do you want to go for coffee? 29099 DM Well no thank you. 29100 DM YouAF? 29101 AF I can't no. 29102 AF I'm meeting X. 29103 MM Oh it's OK. DM seems to know what other limited range nouns apart from "a lecture" MM refers to. He turns to AF to see whether she has a "something" and it emerges that she has.
6.3 CONVERSATIONAL IMPLICATURE The Categories In Chapter 2, in the discussion of markers of intimacy taken to indicate that the group is forming, all forms of humour were analysed. Humour was divided in two categories, the first being 'non flouting' humour, in which speakers talk about funny situations or use funny language, and the second 'flouting' humour, in which speakers humorously flout the maxims of cooperation. Because this chapter is on implicitness over utterances, it focuses on flouting humour, with its dependence on background knowledge. The chapter describes implicitness in the speaker's intention, implicitness in terms of an entirely different meaning than that of the words used. Flouting humour in the data is not related to all cooperative maxims. The maxims of relevance and quantity are not flouted on their own in humorous exchanges unless the maxim of manner, which states that the speaker should be perspicuous and avoid ambiguity, is also involved. It is the maxim of quality, which says that the speaker should not wittingly lead the hearer into erroneous beliefs and assumptions, that is the one most involved in humorous utterances. When this maxim is flouted for comic effect, the outsider has a problem with truth. Although the language may be clear, hearers need some knowledge of the speaker and their attitudes and opinions, in order to appreciate that the speaker is "not telling the truth". The outsider who does pick up the clues that an utterance is ironic or exaggerated then has the problem of determining how far from the truth the attitude expressed is. As it was explained in Chapter 2, flouting humour can be public or private. Public flouting humour, consisting of ambiguous allusions, irony and hyperbole about a third party or a situation, can be used by familiars and strangers alike, and a noningroup member would be able to appreciate it, unless it is based on K4. Private flouting humour, consisting ambiguous allusions, banter and irony about the interlocutor, can only be used by familiars, and a noningroup member would not normally be able to appreciate it because hearers need shared interpersonal knowledge of the interlocutor's general attitudes and present mood. With private flouting humour, the outsider has an even greater problem with determining whether to take the words literally or not. Leech brings out the two levels of meaning, stating that "While irony is an apparently friendly way of being offensive (mockpoliteness), the type of verbal behaviour known
Implicitness over utterances 103 as "banter1 is an offensive way of being friendly (mock impoliteness)" (1983: 144). Dews, Kaplan and Winner say that "speakers choose irony over literal language in order to be funny, to soften the edge of an insult /.../ and to avoid damaging their relationship with the addressee." (1995: 347). Banter is a more risky enterprise than interpersonal irony. If the hearer takes irony at its face value, nothing is lost because the speaker was seen to be being friendly; if the hearer takes banter at its face value, they may interpret it as an insult or criticism. To embark on the act of threatening the interlocutor's face, there must be a certain level of trust and an awareness that the conditions are right for it.
Changes In Humour Flouting humour was studied in order to find whether the nature of the conversational implicature varies with increasing knowledge over time and according to K area, and whether in group code is used in humour. Since it had been found that grammatical and lexical forms used implicitly increase with interaction over time, it seemed reasonable to assume that flouting humour, implicitness in the speaker's attitude, also increases. Since K4 dialogues had been discovered to contain more of these grammatical and lexical forms, it was felt that they might also contain more implicit humour. Finally, it was felt that ingroup code would be used in flouting humour because it seemed that it might increase the ingroupness of humour. For the analysis, the percentage of discourse units containing all types of humour out of all discourse units was calculated within each K area and within each term, as was the percentage containing implicit contextualisation cues. In addition, a qualitative analysis of flouting humour was made in order to contribute to the general picture of the implicitness of ingroup language, and a comparison of noncourse K area and K4 humour was made. As it was seen in Chapter 2, the density of discourse units containing all forms of humour taken together drops in the spring term and returns to the level of the autumn term in the summer term. The analysis of each type of humour shows that flouting humour does not increase with knowledge over time as it was expected to. Figure 2 shows that flouting humour is present from the autumn term, before the code is formed. It also shows that K4 does not contain more flouting humour than noncourse K areas, so the second hypothesis is not confirmed either. In all the data, nonflouting and public flouting humour remain relatively stable throughout, but private flouting humour decreases in the spring and is therefore the one responsible for the spring term's overall decrease in humour. This decrease is all the more dramatic because the density of private flouting humour is very much greater than the other types of humour in the autumn and summer terms. Figure 2 shows more specifically that it is K4's density of private flouting humour that drops so low in the spring term. This means that private flouting humour in the spring term is acceptable so long as it is on a topic unrelated to the course. K4 dialogues occupy 57% of the stressful spring term, so private flouting humour is relatively unusual then. Although students might tell funny anecdotes or be ironic about a situation or third person to relieve the anxiety of the spring term, they are less likely to risk teasing each other about their participation and progress in the course then. As private flouting humour is the greatest source of humour and the most implicit form, it was subjected to qualitative analysis to discover the characteristics.
104 Analysing the language of discourse communities Figure 2: Density of each type of humour In all the data
In non-course K areas
In K4
Implicitness over utterances 10' The most implicit type is the ambiguous allusion: based on privileged knowledge of events in th< hearer's life, it combines the flouting of the maxims of quantity and manner to refer to the even without mentioning it. Witness (9) in which DM says that, unlike BF, he did not go out this weekend: (9) 14021 DM More than I did this weekend I'm telling you. 14022 BF You had friends didn't you? —> 14023 BF I can imagine why you wouldn't want to. 14024 DM Resting. > 14025 BF Yeah. > 14026 BF Sure, (hehheh) BF's words border on the innuendo: she uses no referring expressions at all to point to what she "can imagine". A whole section of text is "missing". Interpersonal irony takes the form of overpolite and overexuberant expressions. It can be an un necessarily polite request for permission as in (10), taken from a session in which students go over tutorial questions together, and AM asks his colleagues if they would mind waiting for him: (10) 03025 AM I'll just go and nip down and get some coffee. » 03026 AM If that's all right? 03027 NM Yes. 03028 NM That's all right. > 03029 BM Well I don't know actually. 03030 BM // We have we are all ready er > 03031 NM // Probably start without you. Hiding under the deference is possibly the implication that ingroup members can behave informally. BM responds with banter and a mock refusal to grant permission, and NM joins him. Here what is implicit is the attitudinal element of "you know we don't mean it". Banter falls into two main categories. The first is banter with which the speaker pretends to reprimand or accuse the interlocutor and flouts the politeness principles of not disagreeing and not impinging. In (11) for example, when DM tells the tale of how his wife's ankles swelled up after a mountaineering expedition, AF affectionately uses a mock accusation of neglect: (11) 22173 DM No no she wasn't in pain at the time. 22174 DM But after it's = > 22175 AF Not that you'd notice = 22176 DM she walked. ((7)) The implicitness here lies in the fact that DM knows that under AF's teasing words is warmth and approval. The second type is banter with which the speaker pretends to misinterpret the interlocutor's meaning and flouts the politeness principle of presupposing knowledge of the hearer's knowledge and values. In (12) DM again comes under mock attack for his treatment of his wife, this time from BF: (12) 23040 AF Oh I've got no money. 23041 AF I've got to go to the bank. 23042 DM same here. /.../> 23066 BF No money? > 23067 BF What do you mean you've got no money? (2) > 23068 BF You've a working wife. 23069 DM That's why I have no money.
106 Analysing the language of discourse communities On the surface, BF seems to misunderstand "I've got no money" to means not "none in my pocket" but "none at all", and to accuse him of lying since he is living off his working wife while he studies. Under the mock attack is a friendly attitude. DM responds positively by continuing the joke. Public flouting humour and nonflouting humour carry different degrees of implicitness. Public flouting humour is implicit in that it too depends on the hearer's knowledge of the speaker's attitude and personality. Course members humorously use hyperbole to describe one member of staff as "fanatic" and another as "a maniac", to describe writing the dissertation as risking becoming "excruciatingly boring", and to assess their ideas for a project as "totally uninformed" and "basically stupid". The hearer who lacks information about events and the students' attitudes, cannot gauge how close to the truth the description is. Nonflouting humour is not based on indirect meaning, but the background knowledge itself in K4 dialogues can make the nonflouting humour exclusive. When speakers play with the language, mix registers and express witty thoughts in a noncourse K area, the meaning is transparent, whereas in a K4 dialogue, it may not be. The third hypothesis that K4 humour would contain implicit contextualisation cues was confirmed. 43% of K4 nonflouting humour contains implicit contextualistion cues. This is typical of K4 dialogues: analysis of nonanaphoric definite reference showed that they have on average a density of 36%, and analysis of special course lexis showed an average of 43%. Although the source of the humour is not usually the implicit cue in itself, the appreciation of a joke might be hindered if the referent is unknown. On the other hand, 54% of all K4 discourse units with public flouting humour contain implicit contextualisation cues and this is greater than the typical cue density. In (13), an example of K4 irony, a student evaluates a member of staffs teaching style positively in order to imply that she does not like it: (13) 29051 NF I'd better go to my class. ((2)) 29052 AF You did this last week didn't you? // (heh) 29053 NF // Yeah. —> 29054 NF I really like the teacher very much. 29055 AF (heh heh) 29056 DM It's very relaxing. (1) The presence of the nonanaphoric definite limited range noun, combined with ignorance as to the teacher's identity and NF's attitude, might make it difficult for an outsider to guess how ironic she was being. Ambiguous allusions to ingroup information are especially implicit because they are heavily laden with cues. In (14), BM asks what happened in DMs meeting with his tutor: (14) 20041 BM Did he direct you to any books or anything about your or have you?= > 20042 DM No his book. » 20043 BM His book right. > 20044 BM // (heh heh) His book right. 20045 DM // (heh heh heh) (3) The whole injoke is merely hinted at. Presumably, this tutor had suggested once too often that students read his book. The course limited range noun "book" increases the implicitness of the humour: both students seem to know the tutor's book. It could be that the joke needs the implicit reference, that the implicitness intensifies the humour. Only 11% of private flouting humour discourse units contain implicit contextualisation cues, which is not surprising since it is based on shared interpersonal knowledge, in which there are few cues.
Implicitness over utterances 107
6.4 TOPIC SHIFTS AND HITCHES This part of the chapter shows how implicit language on the topic shift affects the stretches of discourse that follow. Chapter 2 defined topics as "what speakers say about something", and examined topic shifts in order to demonstrate that they followed patterns that suggested that the group had formed. This chapter examines topic shifts in order to find whether implicit language occurs on them and adds to the implicitness of a whole exchange. This chapter also explores the issue of whether the implicit language can be used interactively by interlocutors over stretches of discourse, from exchange to exchange and never break down. It examines the relationship between implicit language and "hitches", or interruptions in the flow of conversation, and asks whether the language can ever be so implicit that even ingroup members have to check the meaning.
Changes In Topic Shifts Since the comprehension of the whole topic depends largely on the comprehension of the discourse unit introducing the topic, one would expect an unmarked topic shift to contain non anaphoric, indefinite, explicit, introductory reference. One would expect nonanaphoric reference because if the topic is not relevant to the previous topic, there cannot be referential or lexical ties. Indefinite reference would be expected because, with a new topic, the referent has not already been mentioned and is most likely not in the mind of the hearer. It is justifiable, too, to expect explicit and even superexplicit reference because enough information has to be supplied in order for the hearer to identify the referent. A new topic may require introductory reference because, as Coulthard (1977) says that the speaker must "membership" the listener each time the topic changes. The reason for looking at topic shifts in this study was to test the notion that implicit language occurs on the very discourse unit that introduces the new topic, and can make the whole exchange dependent on the cues implicit. It was supposed that implicit language on shifts would be more frequent over time and densest in K4 dialogues, just as it is in throughout the data. The analysis centred on implicit contextualisation cues, as opposed to clausal ellipsis and conversational implicature. Taking all the data together, the percentage of discourse units at shifts containing the cues, out of all discourse units at shifts, increases dramatically in the summer term (autumn term 19%, spring term 17% and summer term 24%), confirming the hypothesis that, as far as the ingroup code is concerned, implicit language on shifts is denser over time. Figure 3 shows that K4 dialogues have twice as many topic shifts containing cues as noncourse K area dialogues, and that main topic shifts with cues are almost four times more likely in K4. This confirms part of the hypothesis that shifts with implicit language are densest in K4 dialogues. Since the density of cues on subtopic shifts is almost the same in K4 as in noncourse dialogues, the discussion of the qualitative analysis focuses here on main topic shifts, where the difference is so striking.
108 Analysing the language of discourse communities Figure 3: Percentage of DUs on shifts with implicit contextualisation cues out of all DUs on shifts All K Area Main topic shift Subtopic shift within same K area to new K area in same K area Noncourse K4
13 47
14 52
17 20
15 30
47% of main topic shifts to K4 contain implicit contextualisation cues, then. This again is almost the same as the overall density of cues in K4 dialogues. The outsider is left guessing what they are talking about for the rest of the topic, if the main shift contains a nonanaphoric personal pronoun as in (15). Up till 23018, the conversation had revolved around the topic of why BF has not applied for the IALS scholarship yet. In 23019, DM suddenly mentions what appears to be the time left before a course deadline: (15) 23014 DM Have you applied for anything? ((!)) 23015 BF No. 23016 BF Not for the moment. 23017 DM Oh you're waiting for Dave // aren't you? 23018 BF // Yeah. Yeah. ((4)) > 23019 DM Oh we've got three weeks now (2). » 23020 DM When are they going to tell us whether it is or not? = 23021 BF After the (1) portfolio yeah. 23022 DM Have you told him we need to know soon? ((2)) 23023 AM Eryeah. 23024 AM Em. (2) 23025 AM Er. 23026 AM // I asked1 asked for the end of term. 23027 BF // Why don't you go and see him? 23028 AM And he said said said that in in effect yeah. (0.5) 23029 AM But 111 go I'll go and see him. 23030 DM Yes sure. 23031 DM No that'll be brilliant yeah. 23032 AM But1 mean I mean I asked him and he said possibly even more. 23033 AM An // extra week. 23034 DM // O::h. DM does not need to say what the deadline is for, and in the second unit of the shift he refers to vaguely with "they", "it" and nominal ellipsis to something specific that only hearers can identify. Not only would the outsider not know what the deadline is for, but because they have missed the topic and therefore the context, they would also be unable to guess to the referent of the "him" and "he" that threads through 23022, 23027, 23028, 23029 and 23032, and is used interactively by all three speakers 52% of main topic shifts within K4 have implicit contextualisation cues. This is considerably higher than the overall density of cues. Again the cues can be nonanaphoric personal pronouns as in (16) in which DM changes the topic from one about why he had done a task in advance to one about what an unnamed lecturer wrote on the blackboard: (16) 04131 DM I'm going away this weekend.
Implicitness over utterances 109 04132 DM 04133 BF 04134 BF —> 04135 DM » 04136 DM 04137 DM
So had to do it. Yah. I've got to do it as // well. // He made a mistake. And I wanted to show one of you somebody actually. I think he made a mistake yesterday when he was putting them up when the diagram was up I think he got one wrong. ((0.5)) 04138 BF Oh I didn't copy them all down. 04139 BF I didn't have time. 04140 BF I was trying to listen rather than copy. 04141 DM I wanted to ask him about that. The implicit pronoun "he" on the topic shift again makes the rest of the topic impenetrable, as "he" and "him" thread through the next units (04137 and 04141) and the impenetrability is compounded by the fact that a new nonanaphoric pronoun appears within the topic. The "them" of 04137 and 04138 cannot refer to the singular definite nonanaphoric "the diagram", and then the "one" with nominal ellipsis ("one of them") is cohesive with the implicit pronoun. When K4 topic shifts with cues also assume shared interpersonal knowledge, the implicitness is even greater. In (17), there is a general "do" verb, a vague noun phrase and a limited range noun as well as evidence of privileged knowledge. BM changes the subject from AF's topic about how she sees fewer students this term, to a topic about what DM's doing for his project: (17) 15156 DM Mm. 15157 AF You kind of expect to see all your mates in the classroom and you look around and think what a lot of skivers. ((3)) —> 15158 BM Are you er (0.5) are going to do what-you-thought-you'd-do about your project? 15159 DM I'm going to give out a questionnaire. 15160 DM And I'll give you one as well. 15161 DM Sometime this week I hope t tomorrow I'll get them all done. 15162 AF What your core project? 15163 DM Yeah. ((0.5)) In unit 15162 AF shows that this is so vague and private that it produces a hitch in communication even for an ingroup member. Turning now to clausal ellipsis and conversational implicature on topic shifts, analysis revealed that a very small proportion of shifts contain them. However implicit a main topic shift's referring items may be, the idea has usually to be completely expressed: the vagueness cannot generally be at utterance level. There is no clausal ellipsis on any main topic shifts and only 2% on subtopic shifts. Only 4% of main topic shifts to a new K area have flouting humour, and it is always public and never in K4. In (18), a Kl example, AM enters the room spilling his coffee, and opening a main topic with a hyperbole: (18)—»03058 AM ((Returns)) I don't know I'm not doing very well with liquids today.((whispered)) 03059 BM // (hehheh) 03060 AM // (hehheh) 03061 BM Did you have some kind of accident?
110 Analysing the language of discourse communities 03062 AM
Well at six thirty this morning (0.5) downstairs (0.5) comes up and rings the door bell. 03063 BM Aha. 03064 AM I've got er I've got water dripping on my head. 03065 AM Have you got a leak in your flat? (0.5) 03066 AM And I've got a leak in my flat. 03067 AM So yeah liquids and me are just not like that at the moment. This implicature is easily understood, partly because the common knowledge is situational, and partly because the understanding of the narrative does not depend on comprehension of the hyperbole in the "abstract" (Labov, 1972). None of the main topic shifts within the same K area contain conversational implicature and only 2% of subtopic shifts do. That is to say, the part of the implicit language that features most on topic shifts is implicitness at the reference level but not at the utterance level, possibly because it would risk a breakdown in communication.
Changes In Hitches The term 'hitch1 is used to mean a temporary halt in the flow of communication, a minor breakdown. This can be brought about either by 1) the speaker's miscalculation of the knowledge of the hearer and a misjudgement as to the degree of explicitness needed, or 2) the hearer's inattention or inability to hear. In the first of these, the hitch has a linguistic, referential cause: it occurs when what the speaker assumes to be known does not correspond to what the hearer knows and has in their mind at the time. This is what Grice (1975) calls "infringing" rather than "violating" a cooperative maxim: the speaker has no intention of generating an implicature or of deceiving, as they do with a violation. A hitch can be caused by an infringement of the maxim of a) quantity, when the right amount of explicit reference is not provided, and b) relevance, when the speaker does not take into account the hearer's knowledge and goals or make clear the "why this now and to me?". As Cook observes, "a successful communicator is a person who correctly assesses the state of knowledge of his or her interlocutor" (1989: 64). Wardhaugh makes the point that implicitness can cause misunderstanding: "We may be led astray by the imprecision of certain very common words, words whose meaning can be determined only in relation to the contexts in which they are used. But since those contexts may never be made fully explicit or may be differently viewed by speaker and listener, here too the result can be misunderstanding" (1985: 37). These imprecise "very common words" with contexts not "made fully explicit" are at the centre of the present study. Hitches with linguistic, referential causes are more of interest to the study than ones caused by the hearer's inattention, because they are relevant to the question of knowledge and implicitness. In the study, the category of hitches with a referential cause does not include any caused by failed conversational implicature, speech acts or macrofunctions. Wardhaugh (1985) lists examples of such causes: the speaker thinks that someone has agreed when he has not or that he has answered the question when he has not, or the hearer does not understand the speaker's the illocutionary force or takes something seriously when it is intended as a joke. There are no examples of "misfires", or failures in performative speech acts that occur when one or more of the felicity conditions are not satisfied. Tannen (1991) says that misunderstandings occur between men and women because they speak with different macrofunctions. She says that whereas men do
Implicitness over utterances 111 "reporttalk", or "public" speaking to negotiate status, inform and perform, women do "rapport talk", or "private" speaking to negotiate relationships, interact and establish connections. Gender difference does not appear to cause hitches in the data. The second type of hitch, that which is caused by the hearer's inattention or inability to hear, has a sociofunctional cause. This type of hitch is not so relevant to the study since it occurs despite the fact that enough explicit information has been provided. Whereas in the case of hitches with referential causes, the information is not there in the text, with sociofunctional hitches, the information is there. There are two categories of sociofunctional hitch. The first is related to the context of situation: the hearer cannot hear or concentrate because of the surrounding noise or because of people coming and going, carrying out activities around them, and even interrupting the conversation. The distracted hearer is unsure of the topic, cannot remember the information given, or is unable to make the connection between an anaphoric pronoun and the preceding presupposed item. In (19), a conversation about how they are progressing, DM's "it" of unit 15061 clearly refers to the questionnaire of unit 15053: (19) 15053 CM I forgot to answer your questionnaire. 15054 DM Oh that's all right. 15055 DM // That's OK. 15056 DM That's fine. 15057 DM It's still the end of this week. 15058 CM // I'm afraid I er. (0.5) 15059 CM OK. 15060 DM I've got to give it the I haven't given anybody else one anyway so. (0.5) » 15061 DM In fact he told me to re-revise it. » 15062 CM What was? —» 15063 DM The questionnaire. 15064 CM Oh. 15065 CM Right. This sort of hitch is not mainly caused by speaker implicitness but by the hearer not paying full attention. The second sociofunctional cause of a hitch is not so much a lack of communication as an interruption in the flow with the hearer asking for more information than they need for comprehension, or predicting wrongly the speaker's next words. In (20), FF is explaining the history of Scottish national traditions to Canadian CM, who dares to be flippant about it: (20) 13104 FF Christmas wasn't an official holiday. 13105 CM What! (0.5) 13106 CM Why? 13107 CM That's that's a (0.5) Catholic = 13108 FF Well. 13109 FF It's it's a it's a weird reaction against = > 13110 CM the workers. > 13111 FF Well no it wasn't. > 13112 FF Because you had New Year off. This is not a case of clausal ellipsis in the form of an unfinished sentence completed by the hearer, as FF did not leave the sentence unfinished expecting CM to know how it ended: she was
112 Analysing the language of discourse communities interrupted by him. This sort of hitch is not caused by implicitness so much as by the inattentive hearer's overenthusiastic positive politeness (Brown and Levinson, 1978). In the study of linguistic referential hitches, the 'repairs' of the hitches were also analysed. Repairs occur when the hearer makes a clarification check or gathers further information and the speaker supplies the missing information or makes the reference more explicit. Just as a hitch can occur because of insufficient information being provided, a repair can consist of the supplying of that information. As Fox et al. (1996) say, repair and syntax are interdependent. In (21), BM has been talking about a friend's interview for a TEFL job in Edinburgh and says that the questions were too numerous and basic. Edinburgh resident AF gives her an opinion: (21) —» 05046 AF I know I think most of them most of the English schools here are cowboy outfits anyway. > 05047 NF Inin Edinburgh or in this country? »05048 AF In Edinburgh, (hehheh). 05049 BM She said the school was absolutely beautiful. 05050 BF Was it? 05051 AF Oh really? The nonanaphoric demonstrative adverb "here" has to be made more explicit for NF to understand, and "In Edinburgh" provides the repair. The repair can also be what WilkesGibbs (1993) calls 'othercorrected noun phrases', in which the addressee makes the repair of the speaker's hitch. When the text was being tagged, however many discourse units were covered between the hitch and the repair, the sequence only counted as one hitch and repair. The hitch consists of the unit (or units) that causes the breakdown. The repair can consist of two parts: the unit that contains a request for clarification and the one (or ones) that contains the clarification. The repair usually begins in the discourse unit immediately following the hitch, and it may extend over several discourse units until it has achieved its objective. Although most hitches are repaired, some are not, either because comprehension of that unit is not essential to the conversation, or because there are too many participants in the conversation and not enough time to merit interrupting the exchange. The aim of the study was to discover whether hitches can be caused by an inappropriate use of the implicit language. Kreckel believes that "the better people know each other, the more they expect complete understanding" (1981: 40). The study explores whether the increased expectation of knowledge makes it more difficult to calculate how explicitly to speak, whether as K4 and shared knowledge grow and the ingroup code becomes denser, the number of hitches increase. It was hypothesised that K4 would have more hitches caused by cues than noncourse K areas. Hitches were also examined to see what proportion contained clausal ellipsis and conversational implicature. The percentage of discourse units with hitches out of all units was calculated for each K area and the average for each term was found. The analysis shows that the percentage of discourse units with all types of hitch out of all discourse units remains stable at about 1% throughout the year, not confirming the hypothesis that the number of hitches would increase. Figure 4 shows that K4 has slightly more hitches than the other K areas; its increased frequency in the spring term may be a result of the stress then causing inattention and impairing performance (Thomas, 1995: 74). There are only five hitches in
Implicitness over utterances 113 all the data that contain shared interpersonal knowledge; reference to shared knowledge is likely to be understood by the hearer. Thus it can be seen that the frequency of hitch does not depend on knowledge over time or the K area, but on course events. Further analysis showed that 21% of all hitches are caused by hearers who are NNS and not familiar with the cultural background of Britain, as discussed in Chapter 2. 27% have socio functional causes. Most hitches (52%) are referential. Figure 4 shows that K4 hitches are highest in the spring shown. It is worth noting that, whereas 70% of all referential hitches are caused by ingroup code, 100% of K4's referential hitches contain ingroup code. That is to say, the code does not always work efficiently even for ingroup members. Very occasionally it can be so vague and ambiguous that they have to check whether they have understood. Figure 4: Percentage of all hitches out of all DUs in each K area
The largest group of ingroup code hitches are those caused by lexical implicit contextualisation cues. The cause can be, for example, the metonymical use of the proper course names, as in (22), in which students discuss why some books are better than others: (22) 13001 AM It's a real text book. 13002 AM It's not like Brown and Miller. 13003 CM Figure thirteen. 13004 CM See figure thirteen. 13005 AM (hehhehheh) 13006 CM Figure twelve. 13007 CM Fifigure fourteen. 13008 AM // (hehehheh) 13009 CM // No figure thirteen. (7) > 13010 CM You when you read over that Syntax didn't it seem very simple? —> 13011 CM After you read it again now? » 13012 AM Er you mean the book? > 13013 CM Yeah the book. 13014 AM Yeah. Here, the discussion before unit 13010's "that Syntax" has been about "Brown and Miller", another metonymical proper noun to refer to their book. AM may wonder if CM now refers with "that Syntax" to something else such as "tutorial task", "an option's material for the exam", "exam
114 Analysing the language of discourse communities question" or "all that has been studied from lecture notes and articles and books". The repair contains a limited range noun supplying the missing second word of the phrase "Syntax book". The same problem happens the other way. Speakers can use limited range nouns and hearers request a prehead modifier. In (23), students are talking about what they had to prepare for the tutorial. BM is not sure whether he missed picking up the "thing", presumably a sheet with the task on. Since he does not have certain knowledge of the task, he cannot understand the limited range noun with the nonanaphoric definite reference in "the whole task" (unit 07018): (23) 07014 BM Er what ha1 didn't pick my thing up. 07015 BM Did we have a thing to pick up? 07016 BF Yeah. —> 07017 BF You've got a task to do with one (0.5) with one other person. » 07018 BF It's been divi the task the whole task has been divided up between the group. 07019 NF Yes. > 07020 BM What this one? 07021 BF // No. 07022 BM // No. > 07023 BM What are we talking about? > 07024 BF To // day's. 07025 BM // Today's. 07026 BM No who's (0.5) ? 07027 BM I haven't got one of those. He goes through other members of the limited range noun category "task", and only when BF supplies him with the prehead modifier "Today's" in her repair and most likely shows him the sheet, can he identify it and refer to it with "one of those" in the situational context. Hitches caused by grammatical cues are far fewer than those caused by lexical ones. (21) above ("I know I think most of them most of the English schools here are cowboy outfits anyway") contains an example of a hitch with a grammatical cause. The following Kl example has another; it comes from the same dialogue as (20) above ("Christmas wasn't an official holiday"). AM does not know whether implicit nonanaphoric demonstrative adverb "here" refers to the MSc course, the university, Edinburgh or Scotland: (24) 13091 FF I don't think you'll get Easter Monday holiday here. 13092 CM You don't! > 13093 AM You mean us or people generally? ((3)) » 13094 CM It's here. 13095 CM // The term begins the twentieth 13096 FF // You don't get it's the (1) 13097 FF It's EEnglish. 13098 FF Everyone gets a holiday but EEdinburgh er gets a = CM's repair is not entirely clear because he just repeats the demonstrative adverb "here"; even the explanation of "the term". FF seems to suggest that "here" is Edinburgh. There are no hitches caused by other features of implicit language. Instances of conversational implicature occur in utterances that are in hitches, but it is not the conversational implicature itself that is the cause of the hitch. In (25), for example,
Implicitness over utterances 115 (25) 04060 AF You get the occasional one. (2.5) > 04061 AF The wild ones are on the on the green buses actually. 04062 AF They go at a hell of a rate. 04063 AF Don't stop for anything. 04064 AF If you got to // want to get somewhere > 04065 BF // Limited stop buses are they? —> 04066 AF No just the no they're just em the the SMT buses 04067 BF // Oh. —> 04068 AF // the green ones that go into Saint Andrew's Square. 04069 AF The big ones and the little ones and the (0.5) 04070 AF I d1 don't know. 04071 AF I think they're on some sort of productivity bonus or something. 04072 BF (hehheh) it is not the hyperbole of "the wild ones" that causes the breakdown in unit 04061; it is "the green buses" and the lack of cultural information. An exact identification of referents is not needed for an understanding of the anecdotes.
6.5 FURTHER DIMENSIONS Both the triangulation dinner and the impenetrability test focused principally on implicit contextualisation cues, but there were a few questions about conversational implicature. At the triangulation dinner, all recordees understood the flouting humour of their dialogues, except DM, who missed the irony in two dialogues. One of them had these lines: (26) 20001 AF Are you doing discourse for your core project? 20002 BM No but maybe I should. (0.5) 20003 BM Maybe that would explain what it is about. 20004 BM I'm doing my option on discourse. and the question was, "BM says, "No but maybe I should. Maybe that would explain what it is about.1 Is he serious?". DM missed the fact that BM was being ironic. AM, for example said, "He knows that it would be foolish to start a project that he knew nothing about", and AF said, "If he's doing it as an option he must have understood the core course". It would seem, therefore, that even the participants of the dialogue can occasionally have difficulty in assessing the truth value of the interlocutor's words. In the impenetrability test, the group C subjects were the only ones who read between the lines and appeared to appreciate the humorous conversational implicatures. Moreover, they volunteered comments about irony and banter in answer to questions that were not specifically aimed at that. (27) contains parts of one of the dialogues: (27) 23001 BF I told you my other halfs (0.5) thinking of going for a PhD? 23002 DM Oh you might as well do one as well then. 23003 BF Get out. ((!)) 23004 DM No? 23005 BF I don't want to do one. 23006 DM No I didn't think you did. /.../ 23012 BF But we've just heard that (0.5) as well that er (0.5) that he applied for a job in the Sudan and they've taken up references. 23013 BF Find out this week (2.5) so:: ((2))
116 Analysing the language of discourse communities 23014 DM Have you applied for anything? ((!)) 23015 BF No. 23016 BF Not for the moment. 23017 DM Oh you're waiting for X // aren't you? 23018 BF // Yeah. Yeah. ((4)) In this case, one added a little note at the bottom of her sheet after all her answers, : "Do BF and DM fancy each other? Yes. From the way she complains about X, and he is supportive and warm and yet sounds sarcastic (jealous) in line 12" [unit 23017]. It could be that the closer the subjects in the impenetrability test are to the ingroup, the more they feel in a position to add opinions about feelings; nobody in Groups A and B made such a comment. On the other hand, only an ingroup member would know that DM is supportive with everyone, that BF jokes with all the men, and that this dialogue took place after they had all spent almost a year together. Although flouting humour may not prevent an outsider from understanding a conversation or identifying the referent, it can exclude the outsider from a finer appreciation of what is going on under the surface, of the subtleties of attitude. The instances of clausal ellipsis, conversational implicature and hitches in the data were examined to check whether the results could be said to be typical of the six recordees and generalisable to the group. Figure 5 shows for each feature what proportion of the total number of discourse units were spoken by each recordee. It was found that the individualities of the recordees did not skew the results; the qualitative analysis revealed no more than hints of personality traits. For example, the six use similar amounts of clausal ellipsis; CM does use three times as much as DM, but this is simply because CM has a habit of ending his sentences with "or something", whereas DM never ends his sentences with vague fillers. Figure 5; The percentage of features for each recordee Out of the total in each case AF BF AM BM CM clausal ellipsis flouting humour hitches and repairs
15 16 20
18 23 25
20 11 18
13 11 8
26 7 18
DM 8 32 11
The recordees also use similar amounts of humorous conversational implicature. There are slight differences, but again these differences do not influence results to any great extent. DM is the most frequent user of private flouting humour. Thus a typical exchange is (28), in which he is ironic about BF's achievement: (28) 14031 BF I beat the Sandwich rush. 14032 DM Oh yeah! (heh heh) » 14033 DM That's quite clever. (5) » 14034 DM On the ball as ever. 14035 BF When it comes to food yeah that's about it. and (29), in which he indulges in banter with AF. He gives up his seat to MM with (29)> 29106 DM Shit! » 29107 DM I can't sit here and er be chatting to AF all day. 29108 AF (heh heh) 29109 DM I'm afraid. 29110 AF Oh well.
Implicitness over utterances 117
All the recordees cause hitches and repair them. What varies from person to person is the reason for the hitch and the attitude towards repairing. AF, the Scot, causes a number of linguistic referential hitches by referring to places in Edinburgh and wrongly assuming that her interlocutors have the background knowledge, as in (21) above about English schools in Edinburgh, and (25) about the green buses. In (30), she talks about her day out with her son forgetting that DM is not from Edinburgh: (30)»21033 AF Actually I was up at the City Cafe. 21034 DM Mm? 21035 AF Heaping plates of chips into Julian. 21036 AF They do very good chips there actually. > 21037 DM Where's that? » 21038 AF The City Cafe on Blair Street. (2.5) BF's characteristic is to give a careful and fully explicit repair to any hitch. (31) contains a referential hitch caused by a lack of cultural knowledge. They are talking about a novel of Melvyn Bragg's serialised on the TV: (3 !)> 11060 BF Maybe Melvyn Bragg's like that. 11061AM Maybe. ((2)) 11062 BF I saw him on the South Bank Show last night. (1) 11063 BF With a trumpet player. 11064 BF And they were doing a profile of him. 11065 BF And I was sort of= > 11066 NM Who was it? —> 11067 BF He was a guy who wrote this dirty book. » 11068 BF Melvyn Bragg? 11069 NM Melvyn Bragg. AM and CM tend to be hasty and unclear with their repairs. (24) above about Easter "here", showed CM's unhelpful repair.
6.6 CONCLUSION Clausal ellipsis adds to the implicitness of student dialogues. Although these features do not become denser throughout the data, they do increase in K4 dialogues. When units containing clausal ellipsis also contain technical terms, proper nouns and nonanaphoric pronouns, this increases the implicitness K4 language. It may be that clausal ellipsis is simply a reflection of in group membership and a product of assumed common knowledge. On the other hand, ingroup members may use these informal structures consciously to create a sense of involvement by pointing to the fact that they share common knowledge. As Goodwin (1996) says, by starting to speak before X has finished, Y demonstrates that she is so in tune with X that she is willing to commit herself to a position without yet having heard X's. It is a collaborative display of mutual appreciation. Implicit humour does not increase with increasing knowledge over time; it is course events that cause changes in the density. K4 has more private flouting humour, and the dramatic decrease in density in the spring when the pressure is on causes the overall level of density to fall. All humour contains implicitness to a certain degree and private flouting humour contains the most. The
118 Analysing the language of discourse communities function of humour is to entertain and to make a claim for ingroupness. Nash (1985: 920) affirms "We share our humour with those who have shared our history and who understand our way of interpreting experience", and as Brown and Levinson (1978: 124) say, "since jokes are based on mutual shared background knowledge and values, jokes may be used to stress that shared background or those shared values". Fillmore (1994: 294) lists the following as the function of humour in general: "asking for support, giving support to present or absent others, establishing solidarity and consensus among the members of the group /.../ appealing for empathy /.../ and tapping group solidarity". It could be that flouting humour, depending for its meaning on unspoken words, consists of a stronger claim for ingroupness than nonflouting. Blakemore (1992: 170) says, "the speaker who leaves his attitude implicit suggests that the hearer and the speaker share assumptions about what, for example, is ridiculous or absurd". Private flouting humour is a particularly strong in group identity marker because it depends on shared interpersonal knowledge and mutual trust. Brown and Levinson (1978: 229) suggest that banter is a way of asserting intimacy, and Leech (1983: 144) says that underpoliteness can have the effect of "establishing or maintaining a bond of familiarity". Johnson (1993) found that "aggressive humorous backchat" in malebonding footballtalk stressed similarity between speakers and created ingroup norms. Eggins and Slade claim that "ties and solidarity in many of the conversations are created through teasing each other" (1997: 124). They go on to say, "humorous devices such as teasing, telling dirty jokes or funny stories, and using hyperbole enable interactants to negotiate attitudes and alignments, and provide a resource for indicating degrees of'otherness' and 'inness'" (ibid: 155). In the case of topic shifts and hitches, it was found that shifts containing implicit contextualisation cues increase dramatically in the summer term. Approximately half of K4 shifts contain cues and this can be said to cause the pervading implicitness of K4 exchanges. When the ingroup code occurs on the very unit that shifts the topic, this may be an economy measure, a timesaving device, or it may be that students want to underline the sharedness of their knowledge and make a claim for intimacy. The density of hitches remains low and stable throughout. Only in K4 does the level peak in the stressful spring, when the ingroup code itself breaks down very occasionally, as the lexical contextualisation cues become so vague and ambiguous that even ingroup members have to check whether they have understood the right referents. Ingroup members know to repair a code hitch with another minimal piece of code. The function of repairs is to ensure communication and by extension to help social cohesion.
7
FUNCTION
7.1 INTRODUCTION This chapter provides the functional background to the implicit language. Fairclough (1995: 5) says that all speakers have background knowledge about orderly interactions: knowledge of language codes, of principles and norms of language use, of situation, and of the world, and that knowledge implies ways of doing which serve the interests of some social group. This chapter discusses the principles and norms of use in the student common room, and the ways of interacting that serve the needs of the academic discourse community. The chapter falls into three parts. The first deals with the students' purpose in speaking to one another and discusses why they talk about the course. It describes the macrofunctions, exploring the changes that occur throughout the course. The second part deals with the microfunction of their interactions, describing the general attitude that students find it acceptable to express and the particular speech acts that the common room prescribes, and examining the changes in the social rules over time. The last part of the chapter draws all these threads together to describe the function of implicit language, by examining the macrofunction and speech acts involved in utterances containing it. The final aim is to discuss whether they use implicit language simply as an unconscious reflection of ingroup membership or whether it is a conscious strategy to claim ingroup membership.
120 Analysing the language if discourse communities
7.2 MACROFUNCTIONS The Categories This study explores the students' adherence to the positive politeness principle (Brown and Levinson, 1978; Leech, 1983) which states that the speakers should claim common knowledge in order to claim common ground. The study takes into account analyses of how involvement is created in interaction (Gumperz, 1982) and how social occasions create expectations for a display of involvement (Goffinan, 1971). The students' interactions are examined to discover whether they claim common ground to show involvement or to test the normality of their actions and attitudes. Particularly relevant are Ochs et al's (1996: 37) investigations of "ways in which talk and interaction both organize and are organized by institutions, relationships and culturally specified environments", and ways of "relating strategies for engaging in verbal interaction to the socialisation, maintenance and transformation of social realities such as family, the school, work or community political structures". Function is a nebulous and subjective aspect of conversations to analyse. The analyst is often obliged to draw conclusions based on intuition. This is easier while the speakers are still relative strangers, but, as Brown and Yule (1983: 64) say, "For the discourse analyst /.../ the more personal and particular the occasion for the participants, the more limited and circumspect he must be in his interpretation". Fortunately, the researcher's own experience of DAL M.Sc. courses helped to render her assessment of function not entirely guesswork. Language is multifunctional and there is a cline from the transactional to the interactional; for convenience in the study, all talk was coded as interactional except on the occasions when it was purely transactional. Transactional language has been defined as that which is concerned with the intentional transmission of factual, prepositional information (Lyons, 1977), the expression of content, practical discourse designed to accomplish specific purposes (Berry, 1981: 132), and "talk for getting business done in the world /.../ in order to produce some change in the situation that pertains" (McCarthy, 1991: 135). Pure transactional language, should it exist, is instrumental. An exchange was categorised as transactional when immediate action was needed in the present, as in (1). The students are at their first day welcoming buffet: (1) 01053 MM I think I might go and have another bun. 01054 AM I thought I was going to get another one. > 01055 BM Could you get me a tuna and sweet corn one please? » 01056 AM Me as well. It was also categorised as transactional if the information was needed for a certain action in the future, as in (2), in which CM gives advice about buying a computer: (2) 06003 AM She said howhow can I be sure that when I send send them my cheque they won't they won't just take the cheque and not send me the computer. /.../ » 06007 CM You're buying a computer? 06008 NF Yeah. /.../» 06011 CM The best thing to do is use a che- master card or something. 06012 NF I don't have a master card.
Function 121
The interactional part of the cline closest to the transactional end is the function of communicating information that interlocutors believe might be worth storing for future reference. As Sperber and Wilson (1986) claim, human minds have a long term aim of increasing their knowledge of the world. It is not always clear, however, to what extent the transmission of information is really aimed at informing the hearer and to what extent it is simply to be sociable in an interesting way. As Wardhaugh (1985: 177) points out, "Often, we just want to be sure that everyone has access to the same information that we have, so we check up on what information is available rather than pass on anything that is particularly novel". Next to the function of increasing world knowledge on the cline, is the social function of normalitytesting, of communicating information about one's experiences and reactions to them in order to check whether they are acceptable. Argyle and Henderson (1985: 24) call this "comparing notes, sharing experiences /.../ to build a shared cognitive world, to put our private experiences into words and compare them with those of others". Berger and Luckman (1966) refer to it as reality maintaining, and say that interlocutors maintain reality by talking through elements of experience and giving them a place in the world. Schachter (1959: 84) refers to affiliative behaviour as providing "a means of checking opinions against those of others". Others express the normalitytesting function in terms of a psychological necessity: Stubbs (1983: 46) says that it is a way of "reliving experiences, releasing tension or 'getting something off one's chest'" and Duck (1991) says that friends are "benchmarks" to guide our attitudes and keep us emotionally stable. Further along the cline is the predominantly interactional function, concerned with the expression of social relations. This is Halliday and Hasan's (1976: 267) interpersonal function, concerned with "the social, expressive and conative functions of language, with expressing the speaker's 'angle'; his attitudes and judgements, his encoding of the role relationships in the situation, and his motive for saying anything at all." This social discourse, in which participants are "interacting for the sake of interacting" (Berry, 1981: 132), is "the lubrication of the social wheels" confirming and consolidating relationships and expressing solidarity (McCarthy, 1991: 136). It has to do with emotion and how speakers "indicate friendliness, cooperation or hostility, or annoyance, pain or pleasure" (Yule, 1996: 6). Eggins and Slade (1997: 4850) say that interpersonal meanings actually motivate and structure casual conversations. At the extreme interactional end of the cline is what Malinowski (1923) called phatic communion, "a type of speech in which ties of union are created by a mere exchange of words" (in Laver and Hutcheson, 1972: 149). This is speech with insignificant informational content whose function is to keep the channel of communication open between speakers. Tannen (1991: 102) calls it small talk crucial to maintaining "a sense of camaraderie when there is nothing special to say". Leech (1983) argues for a Phatic Maxim, which can be formulated as "avoid silence" or "keep talking". The changes over time in the function of the interaction are of interest in this study. Some sociolinguists have attempted to look at the way function varies with the degrees of intimacy but often conclusions have been based on hunches rather than systematic studies. Goldberg (1983: 33) feels that "conversational interaction between familiars is /.../ more often a matter of the reaffirmation of social relationships than the exchange of unknown information".
122 Analysing the language if discourse communities Wardhaugh (1985: 54) has the same feeling: "much of the language between closely bonded people is phatic rather than informative". Once again, these are descriptions of product rather than of process. They are snapshots of a particular time in the relationship; they do not trace the route from one level to another.
Changes In Macro-Function Studies of conversations in pubs and bars that have revealed that very little information is typically exchanged, and that the social functions are more prominent (Trask, 1999: 98). It was expected that the students would spend most of their time in interactional talk because the common room is a place to relax between lectures. For the coding of the data, the macro function was established for a whole topic, as with the K area coding, rather than on an individual discourse unit level. The percentage of discourse units with an interactional function out of all units was calculated, and the proportion of each macrofunction type out of all units in each K area in each term was found in order to discover the students' purpose in speaking to one another and find out why they talk about the course. Analysis confirmed that throughout the year on average 62% of their conversations in the common room seem to have an interactional function (see Figure 1). When the interactional / transactional balance was examined in each of the three terms to test Goldberg and Wardhaugh's hunches that "familiars" and "closely bonded people" are more likely to use interactional language than transactional, or more specifically, to test the researcher's hypothesis that as the students get to know each other the function of their talk is increasingly interactional, it was discovered that as the course progresses, the interactional function does increase dramatically from the autumn term to the spring, only dropping slightly in the summer term (see Figure 1). Figure 1: Percentage of interactional units out of all units within K area types
Presumably, transactional talk occurs mostly in the autumn term because all the students have the same courses, and exactly the same assignments and reading material. K4 dialogues then are concerned with negotiations about tasks: who is whose tutorial group, what a tutorial task consisted of, who will write out the questionnaire that they are working on together and so on. In the rest of the course, there is less to be negotiated, since they have chosen different
Function 123 options and are working more on their own. Transactional talk occurs in the autumn term because students are still finding their way about the department, the university, and indeed, around Edinburgh. Noncourse K area exchanges then are concerned with questions such as where the key to the photocopier is, whether there is Microsoft Word in the David Hume Tower, and whether the green buses are limited stop buses. By the spring, they have got all their ways of studying and their support mechanisms established. On the other hand, it could be said they that indulge in more interactional talk in the spring and summer in order to show solidarity with their colleagues and claim ingroup membership. It may be too that this increase is bound in with the increase in shared interpersonal knowledge. The more they know about each other, the more they are able to pass the time socialising. In answer to the question of their reason for talking about the course, it was hypothesised that K4 conversations would have more often a function of maintaining group cohesion than non course K area conversations do. Figure 1 shows that K4 conversations have on average 20% more interactional talk than noncourse K area ones do. It would appear that when the students choose to speak on K4 topics, they assert ingroup membership by pointing to the particular knowledge that they share as a group, referring to the course context that binds them all together. It may be that K4 topics alleviate the pressure and stress through expressions of inthesameboatness. In K4 dialogues, it seems that they benefit from asserting in group membership by testing the normality of their progress in and attitudes towards the course, as well as gaining knowledge from each other that they can use in their tasks, assignments and exams. This is especially the case in the spring term, when nearly all talk is interactional, the density reaching as much as 97% in K4 conversations, and 57% of conversations assume K4. In (3) just before the exam in the spring term, for example, the speakers are indulging in a friendly exchange of information about what colleagues have revised for the exam, but the main function seems to be to obtain information about how they are progressing, to check how they measure up in comparison and how much work is necessary: (3) > 10001 CM What are you - which ones are you concentrating on? 10002 DM On the grammar // syntax. 10003 CM // Grammar sure. /.../—> 10009 CM You're not bothering with going through Bloomfield and all those kindofthings?((0.5)) 10010 DM I don't want to write an essay. In (4), CM asks DM how far he has advanced with the writing of his project; his main objective may be to test the normality of the stage that he is at himself: (4) ^15020 CM But I suppose you're moderately efficient. (1) » 15021 CM But I'm not near an outline. 15022 DM Oh God no I'm on my reading. 15023 DM We've still got a long time. 15024 CM Yeah we've got a month. 15025 DM That's a lot. 15026 CM That's all right. 15027 DM I'm not sweating at all! This issue was explored further through a more detailed analysis of the language used in microfunctions, and the social rules of the common room. Bell (in Firth, 1994: 54) says that
124 Analysing the language if discourse communities the workplace has a distinctive "normative order", and that this will "constrain the type of cooperation, and have an impact on the process of negotiation itself". The next section explores what type of cooperation is permitted and what normative order is dictated and describes the way that the conventions about expressions of attitude change over time.
7.3 SPEECH ACTS The Categories The speech act analysis framework used in the present study was inspired by the ideas of Brown and Levinson (1978) who, in describing positive politeness strategies, list what amounts to speech acts used in chatting and establishing social cohesion. Discussing strategies for claiming common ground, they say that the speaker may claim common points of view, opinions, attitudes, knowledge and empathy with the hearer, by using ingroup identity markers, by seeking agreement and avoiding disagreement. Their framework as it stands was not used to analyse speech acts, however, because it lacks a global systematicity and their categories are not mutually exclusive. The positive politeness model is a good starting point, for as Carlson says, "the intrinsic context for understanding speech acts is mutual knowledge or beliefs that is, common ground" (1993: 73), but a system was needed that would make it possible to code the data in a measurable way. The model for examining microfunction in the study is derived from, but quite different from, that of Speech Act Theory (Austin, 1962; Searle, 1969) and that proposed by Systemic Linguistics (Berry, 1989; Fawcett, 1973; Halliday and Hasan, 1976; Hoey, 1983). Speech Act theorists see language as being used to perform acts and they list the communicative acts: directives, commissives, expressives, etc. They begin with function and then move on to establishing linguistic realisations of the functions, which is the opposite of the bottomup approach used in this study. Systemic models of microfunctions contain a set of some 50 speech acts, with one linguistic form for each of these. They are a priori models that propose algorithms of act sequences that can be predicted in all verbal interactions. The drawback here is that one form can have several functions, and one function several forms. Neither of these models suits the purposes of a description of casual conversations because "chat" is not primarily necessitated by a practical task (Cook, 1989). The students are not "doing" anything in the sense of physically carrying out a joint activity together with a visible end result; they are not "requesting", "persuading" and "promising" etc. in order to get something done. They are simply sitting in the common room together, "asserting", "informing" and maybe negotiating agreement (Firth, 1994), in order to say something as they pass the time together. Even the theory of indirect speech acts is not entirely relevant, because it is mainly concerned with indirect ways of "requesting", "persuading" and "promising" etc. in order to get something done. The microfunction model used was born of ethnomethodological observation of the data, taking into account the context of situation and of culture, as well as the cotext. Speech act theorists deal less with actual utterances than with a closed set of options of utterance types.
Function 125 As Goodwin and Duranti say (1992), whereas in speech act theory, context usually does not go beyond speaker and hearer, in the ethnographic approach, several other aspects of the speech event are taken into consideration, such as spatial and temporal dimensions. Drew and Heritage point out that Searle used idealised sentences with no context, but that sentences and utterances are "shaped to occur in particular sequential and social contexts" (1992: 12). The model used considers the relationship between speech acts: how one act predicts a series of preferred responses. One of the aspects of the speech act that this model highlights is the expression of attitude. As Carlson (1993: 73) observes, understanding what a speaker meant "consists largely in trying to recognize the attitudes the speaker intended his listeners to recognize the speech acts he performed." The model can be seen in the light of Eggins and Slade's (1997) description of "interpersonal semantics", in which talk with attitudinal colouring is known as "appraisal", and involves social evaluations of reality and expressions of emotional states and judgements on the ethics and morality of others, and is closely related to the semantics of involvement, which they say involves informal and technical ingroup language. The speech acts showing attitude amount to what Goodwin and Goodwin (in Duranti & Goodwin, 1992) call an "assessment action", in which speakers evaluate a person or event in an "affect display". They would see the sections of dialogues showing attitude as an "assessment activity", which involves several participants producing assessment actions and coparticipating in the evaluative loading, matching the affect display and making an alignment toward words that is congruent with the assessment. The literature on the socially cohesive function of expressions of attitude and emotion fails to consider in depth how these change over time. Goffinan (1971: 326) says that the speaker "will often present initially a front of diffidence and composure, suppressing any show of feeling until he has found out what kind of line the others will be ready to support for him." Others have suggested that the difference between the language of "newly related" speakers and that of "intimate" speakers is that the intimate's communication is more emotional, more personalised, and contains more positive and negative evaluation remarks. Neither of these studies traces the development of these expressions of feeling and evaluation as speakers pass from being newly related to intimate. The model of speech acts is designed to suit the characteristics of the casual conversations. The 50plus speech acts described in the literature are taken, along with the features of real life spontaneous conversations such as phatic fillers and backchannels, and regrouped together under nine general headings according to the object of discussion and the attitude being displayed (see Figure 2). Speakers discuss either themselves, the interlocutor or a third party; they display either a neutral, a positive or a negative attitude. Having nine categories rather than 50 makes the analysis easier, as speech acts are even more slippery and subjective to analyse than macrofunctions, being a smaller category. Moreover, as Drew and Heritage (1992: 12) point out, it is difficult to interpret how a particular utterance squares with the speaker's intentions. This model relates the microfunction and illocutionary force to the interactional macrofunction and perlocutionary act of claiming ingroup membership. The expression of a positive and negative attitude is usually interactional in that it is way of testing the normality of the speakers' views and reactions: speakers expect a reaction to the opinion expressed, whether agreeing or disagreeing.
126 Analysing the language if discourse communities Figure 2: Speech acts of expressions of attitudes Neutral Positive
Negative
To self
"I work well"
"I work badly"
To interlocutors/ "You work" communication "I see"
"You work well" "I agree"
"You work badly" "I don't agree"
To third party/ situation
"He works well" "It is good"
"He works badly" "It is bad"
"I work"
"He works" "It is this"
The category 'attitude to self, means the attitude of speakers to the themselves. This was labelled 'neutral to self when the words had the illocutionary force of 'inform', 'explain', 'answer question' and 'narrate', or 'express own intentions, plans, wants, needs and desires', as in line 01036 of (5) from the first day that the students met, in which MM tells BM about his teaching experience, without adding an evaluation or colour of any sort: (5) 01035 BM So how about yourself? 01036 MM Em I taught in Japan like CM did a couple of years eh south coast for years (0.5) and couple of months here and there Portugal, Italy, was in Saudi Arabia for three months but. Discourse units were labelled 'positive to self when they contained speech acts 'express satisfaction with self; 'evaluate self positively'; 'show self in a positive light'; 'reassure/console self. They were labelled 'negative to self when there were speech acts 'express dissatisfaction with self; 'evaluate self negatively'; 'show self in a poor light'; 'minimise praise of self; 'criticise self; 'excuse/justify self; 'apologise'. The category 'attitude to interlocutors or communication' means the attitude of speakers to hearers or to the conversation itself. The main justification for having one category for both was that the attitude to interlocutors is often reflected in the attitude to the communication. Thus, those who echo their interlocutors' words, or prompt them to continue, show solidarity with the interlocutor; and those who challenge the felicity conditions of interlocutors or draw attention to the fact that they have not gauged their background knowledge correctly could be said to be showing a negative attitude to his interlocutor. In the category 'neutral to interlocutors or communication', the speech acts tended to be 'greet1 and 'say goodbye', or 'ask after health'. The speakers' neutral attitude to the communication included acts that fill the conversation phatically, such as with a "I mean", "you know", 'checking and showing comprehension', 'catching interlocutor's attention1 and 'taking the turn', 'repeating himself, 'laughing at own joke', 'adding an afterthought that adds no new information1 and 'relinquishing the turn', 'backchannelling' and 'acknowledging1. The expression of an entirely neutral attitude to interlocutors is rare in that even an enquiry after interlocutors' health could be said to be an expression of interest in them. 'Questions' and 'requests' and 'commands' were nevertheless categorised as neutral when there was no marked overt lexicalised expression of empathy or disagreement. Showing a neutral attitude to interlocutors implies 'making a request' or 'command', as in (6), from a dialogue in which CM says that he is going to miss the last few days of class and asks DM a favour:
Function 127 (6) > 27205 CM 27206 CM 27207 DM
Can I ask you to get the notes for me for the last few classes. (2) For // the // Sure. I
The category 'positive to interlocutors or the communication' included such speech acts as 'evaluate the interlocutor positively', 'reassure', 'console', 'encourage', 'congratulate', 'praise', 'sympathise', 'empathise with the interlocutor's feelings about his situation or himself, 'establish the normality of interlocutor's situation', as in (7) in which DM, looking over BF's project that she is about to hand in, reassures her that he had the same problem with the conclusion, that there it is normal: (7) 22074 DM Is that the conclusion? 22075 BF Yeah. > 22076 DM Don't worry. » 22077 DM The conclusion isn't it's not (1) it's not (2) > 22078 DM I couldn't get anything like that. 22079 BF OK. Showing a positive attitude also means 'agreeing with interlocutors', 'accepting, 'admitting' 'permitting' and 'thanking', as well as 'advising', 'warning', and 'suggesting interlocutor's action'. It means 'offering goods and action', 'promising' and 'inviting', and 'suggesting joint action1, as well as 'echoing the interlocutors' words', 'prompting them to continue', "predicting their drift', and 'finishing their sentence'. Speakers who show a 'negative to interlocutors or the communication' 'disagree' with them, 'mention opposite point of view' or situation, 'challenge the factual content or truth value of what they say', and 'decline', 'refuse1, 'deny' and 'defend', as in (8), in which BM teases AM that he has a negative attitude to having to wait for him to get a coffee before they can start going over their task: (8) 03025 AM I'll just go and nip down and get some coffee. 03026 AM If that's all right? 03027 NM Yes. 03028 NM That's all right. > 03029 BM Well I don't know actually. 03030 BM // We have we are all ready er 03031 NM // Probably start without you. Finally, there was the category 'attitude to the situation or a third party'. It could be a 'neutral': 'informing', 'narrating', 'explaining', 'answering questions', as in (9) in which BF talks about a programme that she had seen with Melvyn Bragg the TV interviewer. She offers no evaluation: (9) 11062 BF I saw him on the South Bank Show last night. (1) 11063 BF With a trumpet player. 11064 BF And they were doing a profile of him. The attitude 'positive to the situation or the third party' could be 'inform', 'explain', 'express satisfaction with the situation or the third party', 'evaluate the situation or the third party positively', as in (10) in which BM describes the warden's house that her partner has been offered: (10) 04102 BF And the house is really nice so. (0.5)
128 Analysing the language if discourse communities 04103 BF I reckon it's all right. ((5)) The category 'negative to the situation or the third party' included 'express dissatisfaction with the situation or the third party'; 'evaluate the situation or the third party negatively'; 'express fear and apprehension of the situation or the third party1. In (11), which AF criticises the writer of an article that they had had to read: (11) 16021 AF I thought she was bit of a bore and hedging and surveying and so on and I couldn't see the point of this semantic field boundary. 16022 AF I found it extremely fiddley. To check the coding system, intercoder reliability tests were carried out but the results were not all satisfactory: 70, 50 and 50. One reason for this is that, given the multifunctional nature of speech acts, even a broad coding system cannot avoid situations in which there are two microfunctions and one coder thinks that one predominates and another feels the other predominates. Another reason for the disagreement is that coding speech acts is subjective and each person's interpretation of the implications of one excerpt can vary according to their own personality, attitude and perception of the world. The low reliability of this part of the code does not undermine the validity of this part of the study because of the size of the corpus (26,000 words; 3,605 discourse units). If an intercoder had been asked to take 3,605 decisions, it is likely that the majority of the codings would have corresponded to those of the researcher. For the analysis, speech acts were labelled at their face value: if they were spoken tongue in cheek, they are analysed as at their face value in the speech act field of the spreadsheet, and labelled as humorous in the macrofunction field to show that there was irony, humour, etc. In (8) above, the apparently negative attitude towards waiting for the interlocutor to get a coffee is an illustration of this. This double labelling system meant that findings were not disturbed by speech acts being taken at their face value. Once each discourse unit was classified in one of the nine categories, the density of each category was calculated for each of the K areas in each term. Then each of the nonneutral speech acts was analysed qualitatively, in order to explore the general attitude that students find it acceptable to express and the particular speech acts that the common room prescribes, and find how the rules change over time.
Changes In Speech Acts Influenced by Goffman and Taylor's theories that displays of feeling and evaluative remarks are markers of intimates, the researcher expected the density of positive and negative, or non neutral, speech acts to increase throughout the course. Analysis showed that on average 23% of all discourse units in each term contain language expressing a nonneutral attitude, and that the density increases in the spring and remains at that level through the summer term (see Figure 3). Speech acts expressing emotion seem generally acceptable once the students come back after Christmas. An analysis of each individual nonneutral speech act type pointed to the particular social rules about the expression of attitude that the students seem to follow. In both K4 and noncourse K area dialogues, the predominant attitude is 'positive to interlocutors and communication', constituting almost half of all nonneutral speech acts (see Figure 4). A longitudinal analysis of this speech act, taking both K areas together, shows that there is a steady increase throughout the year (see Figure 5).
Function 129 Figure 3: Percentage of units with non-neutral speech act types out of all units
Figure 4: Percentage of each non-neutral speech act out of all non-neutral units K4 noncourse K Positive to self 1 8 Negative to self 3 14 Positive to interlocutors/communication Negative to interlocutors/communication
43 19
43 7
Positive to third party/situation Negative to third party/situation
13 21
5 24
Figure 5: Percentage of speech acts with positive attitude to interlocutor and negative attitude to situation and self out of all units
Statistical analysis of the difference between the number of speech acts expressing a 'positive attitude to interlocutors and the communication', and those not expressing it, in the three terms, gave a value of x2 of 35.899, significant at < 0.005 level. This increase in is not by
130 Analysing the language if discourse communities chance then. The increase in language expressing a positive attitude to the interlocutor and the communication may be indicative of an increased need to express inthesameboatness, since it reflects the increase in the interactional macrofunction. Thus the most important of the particular social rules seems to be "Express a positive attitude to your interlocutors and the communication, whatever topic you are talking about, and especially in the spring and summer terms". A more detailed analysis of expressions of a 'positive attitude to interlocutors' shows that the main speech acts in this category are 'empathise', 'console', 'encourage' and 'advise'. The students empathise with each other using expressions such as "I've done that" and "same here". In (12) about whether to do a higher degree immediately or leave it till later; DM empathises with CM's attitude and approves of his decision: (12) 27069 CM Not doing any more schooling. 27070 CM It's either now (2) or never for me. /.../ 27077 DM Yeah. 27078 CM But = > 27079 DM But as as I feel at the moment I'm the same as you. 27080 DM Finish. /.../ 27098 DM No if you want to do it badly enough. » 27099 DM You're doing the right thing. 27100 DM // Do it now yeah. 27101 CM // Yeah. Another response to the worried colleague is to console him; to tell him "don't worry", that his progress seems fine. This happens mostly in K4. When in (13), BM is worried because he does not have the papers for a tutorial, BF teases him in order to calm him in a nonfacethreatening manner: (13) 07023 BM What are we talking about? 07024 BF To // day's. 07025 BM // Today's. 07026 BM No who's (0.5) ? 07027 BM I haven't got one of those. 07028 BM That's // no » 07029 BF // Stop panicking. 07030 BF What's wrong with you boy? // (hehhehheh) In (14), CM says that he is going to miss the last week of class and seeks the approval of his colleagues; DM reassures him by minimising the problem: (14) 27119 CM I wasn't going to go back but then the inlaws decided to take my wife. 27120 DM The last week. > 27121 DM It's no big deal. » 27122 DM You can miss that anyway. 27123 CM I only have two classes. 27124 CM So 27125 DM Yeah. » 27126 DM Er it's no big deal. 27127 DM By the time you get to the last week you'll have done most of what you need to you'll know what you need to know for the portfolio. > 27128 DM If you're answering questions anyway (3) shouldn't be a problem ((2))
Function 131 On occasions, students do more than encourage; they praise or congratulate their colleagues. This happens more in K4 than in other K areas: students tell each other they have done well, or have good arguments and ideas. In (15), DM has just read through BF's project and praises it, using understatement and congratulations: (15) 22086 DM There's nothing (1) there's nothing startling that I can see missing from this (2) > 22087 DM and it's very readable. (1) In the spring and summer terms, students who feel close to each other do not always show empathy; they trust each other enough to risk threatening each other's face by offering overt advice and warnings. Thus, in (16), when DM fears that he will have difficulty finding a dissertation topic, AF's response is not "don't worry", but plain advice about tackling the problem, and DM does not seem to object. (16) 21109 DM Been trying to think of something that might stretch to twenty thousand words. ((2)) /.../ 21111 DM And is not excruciatingly boring. /.../> 21115 AF ((Blows nose)) Well you might have to just face that and cope with it. ((sniffs)) 21116 DM Mm. ((2)) The analysis of the other most frequent nonneutral speech acts, 'negative attitude to self and 'negative to third party or situation' (see Figure 5), took into account the possibility that the increase in all nonneutral speech acts in the spring is a reflection of the increase in pressure from the course then. An impressionistic examination of emotive vocabulary shows that in K4 dialogues at that time, there is a large group of abstract, third order entity nouns which seem emotive: "rubbish", "pressure" and "mood", verbs that show the students' anxious feelings: "panic", "worry" and "sweat" and adjectives associated with stress, such as "scared", "worried" and "bothered". Aware of this characteristic of lexis in course topics, the researcher expected K4 dialogues to contain more negative speech acts than noncourse K area ones. Analysis showed that the density of acts expressing an attitude 'negative to third party' were the second highest out of all nonneutral speech acts, and the density was slightly higher in K4 than in noncourse K area. On the other hand, acts expressing attitudes 'negative to self (see Figure 4), which are the third densest in K4 dialogues, are almost five times denser in K4 than in noncourse K areas. Statistical analysis of the difference between K4 and noncourse K area in terms of the number of speech acts expressing a negative attitude either to self or to a third party gave a value of %2 of 41.492, significant at < 0.005 level, suggesting that the difference between K4 and noncourse K area dialogues is significant as concerns these negative speech acts. A longitudinal analysis showed that 'negative attitude to self and 'negative to third party or situation' peak in the spring and then drop in the summer. Since K4 dialogues have more of these two and K4 predominates in the spring, it is reasonable to suppose that this explains the overall peak in the spring. Thus the second most important social rule seems to be "Express a negative attitude to yourself and to the situation, especially if you are talking about the course, and especially in the spring term".
132 Analysing the language if discourse communities The detailed qualitative analysis of expressions of a 'negative attitude to a situation or third party1 showed that, in K4 dialogues, students express a negative attitude to the work, to the books and articles and to the lecturers. They complain that the revision for the exam is "a lot of work", that 3,000 words for the project is "lot" to write, that there is "a lot of pressure", and that the main problem is "time". They evaluate theories, fields of thought and material that needs to be learnt for the spring exam as "rubbish", as "fuzzy", and they use expletives, as in (17): (17) 08016 AM GramGrammar Two. 08017 AM But it I mean it's it's just (0.5) 08018 AM I don't know. 08019 AM It's just I don't know. » 08020 AM Bloody tosh, isn't it? > 08021 CM Well it's it's a bit abstract. ((!)) CM shows sympathy with this, even though his negative evaluation is more toned down. The students complain that books are too "theoretical", that articles are "scurrilous" or "fiddley". (18) contains an example of total condemnation: (18) 14087 NF // Andandandthen what about this one? —> 14088 NF It's completely use- useless. 14089 AM You've read that have // you? 14090 NF // Yeah. Here AM accepts the condemnation and ask for an explanation. The students also evaluate lecturers in a negative fashion. Negative evaluations of teaching staff possibly feature less in the recordings than they do "in real life": this is the one area where the effects of the cassette recorder are in evidence. (19) contains an exaggeratedly negative description of lecturer; when the students remember the recorder, CM does not reciprocate but humorously changes the description to a positive one: (19) 12198 AM I think you need to I mean you know what she's like. > 12199 AM She's really fanatic. 12200AM So that if you don't // get it it 12201 NF // (heh heh) (0.5) (heh heh heh) ((0.5)) 12202 AM Aha. 12203 CM But a good teacher all the same. 12204 AM A wonderful teacher yes. (heh heh heh) (0.5) In these isolated cases, the cassette recorder turns into a "hearer" as a member of staff could listen to the recording, so the students are in fact obeying the rule "Express a positive attitude to your interlocutors and the communication, etc." Detailed analysis of expressions of a 'negative attitude to self in K4 shows that the acts in this category are mainly 'criticise yourself, 'express worry', and 'minimise your progress'. Students criticise themselves and their work, possibly in the hopes that the interlocutor will contradict them or reassure them of the normality of the situation. They say that their "handwriting is awful", that they are "no good at memory things", that they "never get things done on time", that they cannot "get [their] brains going", that their project is "just what everybody says" and that they cannot "come up with" a point themselves. In (20), BM shows himself in a negative light as he describes how he did the tutorial task: (20)> 04111 BM And I reached the stage where I'd no idea. 04112 BF Yeah. ((!))
Function 133 04113 BM
And also when I when I was looking back I realised I actually hadn't read the first part again which gave me a real big clue for one of the first. > 04114 BM It was really stupid. 04115 BF Yeah. 04116 BM But em (1.5) I didn't do any reading. (0.5) 04117 BM I just had one book which wasn't on the list which was too too long really to give // me any conclusions. -* 04118 BF // No I- I haven't read anything specific for it.(2.5) BF's reaction is not, in unit 04115, to agree or disagree that BM is "really stupid", but to wait for the moment when she can empathise by showing herself in a negative light (unit 04118). Students express worry about themselves as in (21): (21) —>29063 AF I'm not very good at thinking up ideas. 29064 AF Especially in Linguistics. There's such a fuzzy field isn't it? 29065 AF 29066 DM It's very amorphous er. Again, DM waits until the focus of the negative feeling is off AF and onto Linguistics, before he responds by echoing her sentiments. If speakers feel that they have alarmed interlocutors about their good progress, they minimise the significance of it by playing it down, obeying the modesty maxim, as in (22) 29010 AF I've got a couple of totally uninformed sort of basically stupid ideas for a project. /.../ 29029 DM Well what are these yeah what are these ideas then? ((1.5)) In this example, the lines between 29011 and 29028 are an insert sequence from another student asking for change for the drinks machine. When DM brings the subject back to AF's ideas, he has emptied it of the negative evaluation of herself. Another way of playing down progress is to justify it in some way, as in (23): Has anybody done their syntax? ((!)) (23) 04124 BF 04125 DM // I did it yesterday. 04126 DM Oh that's what I was doing yeah. 04127 BM // No I haven't looked at it yet. 04128 BM Oh you were doing it in the library weren't you? 04129 BM You're well ahead aren't you? 04130 BF Can you —> 04131 DM I'm going away this weekend. > 04132 DM So had to do it. 04133 BF Yah. 04134 BF I've got to do it as // well. BM is evidently worried that DM is "well ahead", so DM blames his progress on circumstances out of his control. BF shows solidarity with DM by paralleling his situation with a suggestion of some outside obligation of her own. Conversely, the common room prohibits a show of positive attitude to self and negative attitude to the interlocutor, especially in K4 (see Figure 4). However, it would seem that if the rule "Express a negative attitude to yourself and to the situation, etc." is broken, then the rule "Express a positive attitude to your interlocutors and the communication, etc." can also be
134 Analysing the language if discourse communities broken. In (24), CM seems quite pleased with himself, and his refusal to empathise with the tension appears to unnerve AM and FF: (24) 13126 CM Do you realise that we have from er March nineteen until (0.5) April the twentieth? 13127 CM With nothing to do? > 13128 AM Nothing // to do? » 13129 FF / / W e got a project to hand in at // em (heh heh) » 13130 AM // Yes! 13131 CM The project the project is due on the thirtieth of March. 13132 AM The twentyseventh. 13133 AM Thirtieth. (1) 13134 AM And then that's the first or the second one ? 13135 FF The core project's the twentyfirst. 13136 CM Thirtieth of March and the twentyseventh of April. > 13137 AM All right so one month later is the em // first course project. > 13138 FF // And then (0.5) and then another the month after that so there's quite a lot to do. » 13139 AM Right. AM and FF disagree with him openly, forcibly and at length.
7.4 FUNCTION OF IMPLICIT LANGUAGE Macro-Function If one is to be guided by Brown and Levinson's (1978) suggestion that using ingroup identity markers is a way of claiming ingroup membership, by Giles' (1979: 46) Accommodation Theory that "individuals shift their speech styles to become more like that of those with whom they are interacting", and Tannen's (1984) view that strategies of rapport and involvement are expressed in conversational style, one can suppose that the students use an ingroup style of reference to cohere with the rest of the group. It was hypothesised that the ingroup code and other implicit features occur more in speech with an interactional function than with a transactional function. Analysis revealed that the ingroup code is indeed used most with an interactional function: 62% of all discourse units are interactional, but 90% of the discourse units that contain ingroup code are interactional. In addition, the increase in interactional utterances coincides with the increase in nonanaphoric definite reference, implicit reference and course nouns and general words in K4. This coincidence is obviously no 'proof that they use it in order to cohere with the rest of the group. 87% of clausal ellipsis and 95% of humorous conversational implicature have an interactional function too, but this is to be expected since ellipsis invites interlocutor participation and humour is aimed at interlocutor entertainment. There does seem to be an argument for saying that if implicit language is used mainly in utterances that have a socially cohesive function, the language may be more than a reflection of ingroupness; it may also be a strategy for asserting ingroupness.
Function 135 Speech Acts Taking into account the fact that nonneutral speech acts increase over time as do grammatical and lexical contextualisation cues, the researcher was interested to know which of the non neutral speech acts would contain most contextualisation cues, what attitudes the code was most used to express. When the percentage of cues in each of the nonneutral speech act types was calculated, it was found that expressions of 'positive attitude to self, 'negative attitude to self and 'negative attitude to third party' have the highest densities (see Figure 6). These last two predominated in K4 dialogues. Using these speech acts with the most implicit contextualisation cues seems to be a socially acceptable way of talking about oneself and the course. Figure 6: Density of units with implicit contextualisation cues out of all DUs in each positive and negative speech act type
This now suggests that ingroup membership requires that members express a negative attitude to themselves and to the situation, especially if they are talking about the course, using implicit contextualisation cues. Thus, when CM shows himself in a positive light in the course context of revision progress for the exam, he uses a general "do" verb and general noun "people", echoing DM's "people": (25) 10066 DM So did you do everything? 10067 DM // Or sort of choose a few people? 10068 CM // Em. (0.5) » 10069 CM No I've done all the people. 10070 DM // Really? thereby mitigating the violation of the rule, by minimising the importance of his accomplishment through the vagueness. It may be that if members want to evaluate themselves positively and still be seen as one of the group, they have to use more ingroup code than ever in order to counterbalance the functional rule breaking. Obeying the social norm "Express a negative attitude to yourself and to the situation, etc." using implicit contextualisation cues strengthens the claim for ingroup identity. In (26), BF is 'negative to self, criticising her own project, and uses the course limited range noun "paper",
136 Analysing the language if discourse communities the course proper names "Language and Linguistics", "Newmeyer" and "Lyons", and the non anaphoric pronoun "this": It's just my my paper is on Language and Linguistics is just what (26) 17051 BF everybody says. 17052 CM Yeah. 17053 BF According to Newmeyer we have this. 17054 BF According to Lyons this. (2) The social norm of sharing a 'negative attitude to a situation1 is even more bonding if the course is referred to vaguely with an nonanaphoric personal pronoun, as in (27) in which AM may be referring to finishing his project, or the end of an option course or even the end of the MSc course itself: (27) 08047 AM 111 be glad when it's finished. ((!)) 08048 CM Yeah // it's 08049 AM // I'm not really into it. ((4)) Using an nonanaphoric personal pronoun in an expression of negative attitude towards an invited speaker is an ingroup claim as it contributes to the exclusivity of the communication: (28) 13010 CM You when you read over that syntax didn't it seem very simple? 13016 CM You know when I first looked at it I thought what's this. 13017 CM // I'll never get this stuff. 13028 AM Do you remember when he came to talk about it? 13029 CM Yeah. 13030 AM He could he couldn't seem to explain it simply but (0.5) you can. Clausal ellipsis features very much less than the ingroup code does, in the nonneutral speech acts. It only occurs in three utterances expressing a negative attitude to a third party or situation. The unfinished sentence inviting interlocutor completion is a rare way of implying a criticism. In (29), AM and BF discuss the hero of a televised novel: (29)> 11052 AM Urn he's a bit of a ... —> 11053 BF Well (1) sort of like a willy with on two legs. 11054 AM Urn. 11055 AM He's a bit of a wimp in the film. 11056 BF He's a bit sort of wimpy in the book. Here it seems that AM invites BF to join in the criticism, and that BF completes AM's thoughts correctly but not using the exact words that he would have chosen. There are only five units with clausal ellipsis containing a positive attitude to the interlocutor, and they are mainly unfinished sentences with a vague filler, which occur in CMs helpful advicegiving conversations and the vague ending may serve to soften the advice. The nonneutral speech acts involved in humorous conversational implicature are mainly positive to interlocutor (16% of units with flouting humour) and negative to third party (10%). This is natural since implicit humour is usually aimed at showing a positive attitude towards the interlocutor. In (30), to take an example, (30) 15148 BM We'll have to go out sometime. 15149 DM Yeah. > 15150 BM Before we forget each other's faces. 15151 BM // (heh heh) It's true. 15152 DM //(heh heh heh)
Function 137 BM exaggerates amusingly the urgency of their need to meet up socially. Flouting humour is often used to criticise something indirectly, it is hardly surprising that it occurs in utterances showing a negative attitude to a third party. The answer to the question of why they use implicit language can be found in the quantitative and qualitative analysis, but the final conclusions must remain subjective, based on common sense in the light of the findings. Implicit contextualisation cues increase at the same time as the nonneutral speech acts increase. They occur most densely in speech acts expressing a 'negative attitude to self and a 'negative attitude to the situation1, which are themselves the most frequent nonneutral speech acts, after 'positive to interlocutor'. This may mean that in group code does more than reflect ingroup membership. If it is possible to use implicit language consciously, using it may constitute a claim for ingroup membership. Common sense suggests that only certain features of implicit language lend themselves to conscious use. It is unlikely that students could consciously choose to use implicit grammatical contextualisation cues of nonanaphoric definite reference and implicitness, although the non anaphoric personal pronoun is sometimes used to avoid mentioning a name. It is unlikely that speakers are aware of their unfinished sentences with conjunctions and fillers. It may be that these elements of implicit language mainly reflect ingroup membership. On the other hand, it may be that they consciously choose to use the lexical contextualisation cues of technical terms and metonymical proper names. They may even intentionally use limited range course nouns, general words and unfinished sentences inviting interlocutor completion. Implicit humorous utterances are a result of conscious choice: the speaker decides not to speak seriously and explicitly in order to entertain and maybe to be accepted as a group member. These elements of implicit language not only reflect ingroup membership, they can be used to show ingroup membership. However, the fact that the students use elements of implicit language consciously does not necessarily mean for certain that their motive is to be sociable, and to claim and assert ingroup membership. There is now no way of testing whether students were aware of the way they talked and used the cues in order to claim membership, or whether they were just unconsciously complying with the Cooperative Principle of Quantity of giving as much information and explicitness as needed.
7.5 FURTHER DIMENSIONS Triangulation demonstrated that the recordees felt that the macrofunction of their conversations was towards the phatic end of the transactional/interactional cline. In the questionnaire, most said that they spoke to ingroup members to "enjoy their company", to "have a laugh", "to pass the time" and to "see how they fared compared to others". In discussion, they insisted that their main aim in talking with friends had been "just to pass the time". They seemed unaware of how much time they spent talking in a way that could be interpreted as "seeing how they fared compared to others". As far as speech acts are concerned, in the discussion, the male recordees protested that they never spoke to any of their colleagues on topics that showed their feelings, and the female recordees believed that a discussion of feelings was reserved for those that they felt particularly close to. Analysis of speech acts has shown, however, that by the spring term, all
138 Analysing the language if discourse communities six of them could not help showing feelings, emotions, attitudes and opinions in what they said to everyone. It does appear that there would have been more expressions of negative attitude to third parties, had it not been for the recorder: in the questionnaire, BF said that at first she was "more careful who [she] talked about" and CM said that the recorder made him "careful what [he said] about whom". The study of macrofunctions showed general tendencies regardless of personality, as recordees engaged in equal amounts of interactional talk. Differences emerge in transactional talk, as Figure 7 shows, in that CM engages in more transactional talk than AF and DM do, making CM a "typical man" in Tannen's (1991) eyes, since transactional talk corresponds to her "reporttalk" to inform and perform. CM*s transactional talk is mostly informing: advising people about how to buy a computer with a credit card, what application to use for tables, etc. The study of the predominant nonneutral speech acts also shows that the distribution of units was generally equal for each of the recordees. Only in the case of 'positive attitude to the interlocutor and the communication' is there a slight imbalance, CM and DM being responsible for over half of the expressions between them. There is a noticeable difference between the exact speech acts used in these expressions of CM and DM. Figure 7; The percentage of features for each recordee Out of the total in each case AF BF AM BM Macrofunction: transactional Nonneutral speech acts: positive to interlocutor negative to situation negative to self
CM
DM
4
11
16
29
34
6
11 20 14
11 11 14
12 19 20
9 19 12
28 14 19
29 17 21
Just as DM has little transactional talk, he is a warm solidaritygiver, guaranteed to give reassurance to colleagues who are fretting. (7) and (15) above, from a conversation in which he reassures BF that she should not be concerned about her project with "Don't worry. I couldn't get anything like that" and "it's very readable", is therefore typical of him. In (12), above it is DM who empathises with CM's wanting to finish his studies, with "I feel at the moment I'm the same as you. Finish." A characteristic of DM's collaborative attitude towards the communication is his backchannelling; this is what Tannen (1994) would call 'cooperative overlapping'. (31) is from the conversation about why AF cannot go out as much as DM: (31) 21011 AF I mostly don't you know. » 21012 DM Yeah. (1.5) 21013 DM I mostly do. ((2.5)) 21014 AF Yes but you don't have to find a babysitter. > 21015 DM Mm. > 21016 DM Mm. 21017 AF And you've got somebody there to go out with straight away. 21018 AF You don't have to sort of phone and make arrangements. 21019 AF And things. 21020 AF And find out that = > 21021 DM And find someone yeah. =
Function 139 21022 AF > 21023 DM » 21024 DM
Your friend hasn't got a babysitter either even if you have. Right. Right. ((10))
CM's expressions of a positive attitude to the interlocutor take the form of advicegiving and offering of congratulations to those who, during a normalitytesting conversation, he discovers are progressing well: (32) 08009 AM I did a lot of the grammar recently. » 08010 CM Oh you have // excellent. 08011 AM // Of the grammar. On the other hand, CM is the only one who occasionally denies the reassurance that his fretting colleagues seem to want. In (33), in the spring term the week before the exam, AM is worried about his linguistics and psycholinguistics questions: (33) 08022 AM But things like this linguistics as well. 08023 AM You know I don't mimind it. » 08024 CM You still got enough time for that? ((0.5)) 08025 AM There's not a lot of things they can ask. 08026 AM Cos we haven't actually done it that deeply have we? /.../ 08031 AM // Though though I haven't I haven't done any Chomsky. (1) 08032 AM Probably a bit late // now. > 08033 CM // Chomsky doubles up in Psycholinguistics. > 08034 CM There'll be a question on him in that. 08035 AM And I've I've haven't done anything about the Psycholinguistics. 08036 AM I haven't done Chomsky. 08037 AM Which is (0.5) probably very stupid but you // know. > 08038 CM // Avoiding Chomsky in Linguistics is:: procrastination I think. 08039 AM We::ll. 08040 AM No em avoiding it in terms of em (1.5) in terms of er I mean obviously in relation to other things you've got to= AM describes himself in a negative way in units 18032 and 18037, but CM does not respond by echoing his sentiments. He escalates AM's worries, leaving him to console himself in 0802626, and almost incoherently in 08040.
7.6 CONCLUSION Speech with an interactional function predominates and increases with interaction over time. The students' purpose in speaking to one another seems to be to establish and maintain social relationships, and they appear to talk about the course in order to test their opinions and progress against that of the others and relieve anxieties by receiving solidarity from the group members. The general social rules that they find it acceptable to follow are a) "Express a positive attitude to your interlocutors and the communication, whatever topic you are talking about, and especially in the spring and summer terms", and b) "Express a negative attitude to yourself and to the situation, especially if you are talking about the course, and especially in the spring term". The particular speech acts that the common room prescribes are express a negative attitude to the work, books, articles and lecturers; criticise yourself, express worry and minimise your progress; and empathise, console, encourage and advise colleagues. Expressing a negative attitude towards self or a situation can be risky with strangers in that the
140 Analysing the language if discourse communities speaker cannot predict whether he will offend his interlocutors. It is with familiars a marker of intimacy. The students may use implicit language to claim ingroup membership. Implicit language occurs mostly in interactional discourse and it is likely that they use the lexis of the ingroup code (technical terms, proper nouns and general words) and the humorous conversational implicature consciously to point implicitly to shared knowledge and thus assert the ingroup membership, while the grammar (e.g.: nonanaphoric definite reference and clausal ellipsis) simply reflects membership. What can be affirmed with certainty is that reference reflecting in groupness is used in utterances that seem to have the function of claiming ingroupness. As Tannen (1989: 23) says, "the more work /.../ hearers do to supply meaning, the deeper their understanding and the greater their sense of involvement with both text and author".
8
FURTHER STUDY
This final chapter begins with a summary of the components of implicit language and the definition of ingroup members, a summary of the main influences on changes in language and function, and conclusions about the function of implicit language. It then discusses the generalisability of the findings and suggests further avenues of exploration open to students and experts in linguistics. It ends with suggestions as to the applicability of the findings to the fields of English Language Teaching, Clinical Pragmatics and Forensic Linguistics.
8.1 SUMMARY Implicit Language And In-Group Membership The impenetrability tests relate the degrees of outsiderness to degrees of understanding. They show that the closer the hearers are to the 19912 MSc course, the more they understand the topic and referring expressions. The tests confirm, too, the link between implicitness and impenetrability. Hearers only understand the topic of all conversations if they have both K4 and shared interpersonal knowledge and can identify the referents of the implicit language. Figure 1 lists the components of the implicit language and Figure 2 shows the layers of the language accessible to different social groups, depending on the degree of outsiderness.
142 Analysing the language of discourse communities Figure 1: The implicit language of the in-group The ingroup code of implicit contextualisation cues: grammatical explicit nonanaphoric definite reference eg: "the project" implicit nonanaphoric definite reference general nouns eg: "the thing" demonstrative pronouns eg: "that" demonstrative adverbs eg: "now" third person personal pronouns eg: "she" lexical technical terms of ELT and Applied Linguistics. eg: "interlanguage" course proper nouns actual use eg: "Chomsky" metonymical use eg: "[their] Chomsky" unique course noun eg: "exam" limited range course nouns. eg: "project" eg: "done [their Chomsky]" general "do" verbs Implicitness over utterances: clausal ellipsis unfinished sentences with interlocutor completion eg: "the mental and the..." unfinished sentences with conjunctions and vague fillers eg: "...or something" humorous conversational implicature public flouting eg: "His book right!" eg: "Not that you'd notice!" private flouting Figure 2: Layers of implicit language
The real outsiders are those who have no knowledge of Teaching English as a Foreign Language or Applied Linguistics, and do not know Edinburgh. Subjects in Group A in the impenetrability test fell into this category. Slightly less of an outsider is any person who has studied and/or works in the field of Teaching English as a Foreign Language and Applied Linguistics (Group B subjects), because they belong to the world discourse community of English Language Teachers and Applied Linguists and should therefore be able to understand the technical terms and the proper nouns with their actual use, which are associated with those fields.
Further Study 143
Any person who has taught on, has studied for or is doing an MSc in DAL of Edinburgh University, as was the case of Group C subjects, belongs to the discourse community of DAL MSc students and staff, and should be able to understand the DAL MSc discourse community implicit language in common room casual conversations. In addition to the technical terms and the proper nouns with their actual use, which they understand as members of the world discourse community, they can understand elements of the ingroup code such as unique and limited range course nouns. It is possible that, in the context of the course, they might also be able to guess at the sort of entity that is being referred to with other elements of the ingroup code: the non anaphoric general nouns, demonstrative adverbs and pronouns and third person personal pronouns, the metonymical use of course proper nouns and general "do" verbs. They might even be able to guess at the possible meaning implied in clausal ellipsis and public flouting humour. The term 'ingroup members' applies to all NS members of one particular MSc course. Only in group members can understand the way that the DAL MSc discourse community implicit language is used by their own particular ingroup. In addition to the technical terms and the proper nouns of the world discourse community, and the unique and limited range course nouns of the DAL MSc students and staff discourse community, they know exactly which entities are referred to with the nonanaphoric general nouns, demonstrative adverbs and pronouns, the third person personal pronouns, the metonymical use of course proper nouns, general "do" verbs, clausal ellipsis and public flouting humour. They can identify the specific reference of explicit nonanaphoric definite reference and know the special names for staff members and options. In possession of shared interpersonal knowledge, they also recognise the names of student members and their families and understand the implied meaning in private flouting humour. It can be assumed that the recordees are members of the ingroup because some of the markers of intimacy discovered by linguists have been found to increase steadily throughout the course and can be associated with the increasing ingroup knowledge. If informality and the expression of emotion can be taken as intimacy markers, then the increase in initial ellipsis and nonneutral speech acts throughout the course shows that the group is forming. If abrupt topic shifts are a feature of high involvement, then the steady increase in main topic shifts throughout the course are evidence of group involvement. If humour based on common knowledge is a marker that stresses solidarity, then its presence here can be taken as further evidence. It appears that the group has formed by the spring term, and it is reasonable to claim that any changes in form or function that occur in the spring or summer terms can be associated with the formation of the group.
Changes In Form And Function The point was made in the introduction to this book that the ingroup's language is context shaped, as speakers' "contributions to an ongoing sequence of actions cannot be adequately understood except by reference to the context in which they participate" (Drew and Heritage, 1992: 18), and contextrenewing: "Since every utterance will itself form the immediate context for some next action in a sequence, it will inevitably contribute to the contextual framework in terms of which the next action will be understood" (ibid.). The book has shown that the talk in
144 Analysing the language of discourse communities the common room both organises and is organised by EU's DAL, relationships within the in group and the culturally specific environment of the university and Edinburgh (Ochs, 1996: 3). The main assumption in this study was that with interaction over time, the group's common knowledge of the course and shared interpersonal knowledge would increase, and that there would be changes in the language. Analysis showed that some features do increase in the spring term and continue increasing in the summer term, but that events on the course calender also have a great influence on the way that the students talk (See Figure 3), so that some aspects of language change in the spring term but then the tendency is reversed in the summer term. Still other features do not show any remarkable changes from one period to another, but rather vary according to the K area. Some aspects only increase over time in K4. Figure 3; Changes in form and function Features that increase steadily throughout the course: sections of dialogue based on shared interpersonal knowledge implicit grammatical reference nongeneral course nouns and general "do" verbs in K4 clausal ellipsis in K4 interactional function speech acts expressing a positive attitude to the interlocutor and the communication Features that increase in the spring and decrease in the summer: dialogues based on K4 nonanaphoric definite reference lexical contexualisation cues hitches in K4 speech acts expressing a negative attitude to self and the situation in K4 Features that are densest in K4 dialogues: nonanaphoric definite reference implicit grammatical reference lexical contextualisation cues private flouting humour topic shifts and hitches with implicit contextualisation cues Some changes occur throughout the data regardless of K area. There is a steady increase in the density of utterances with a predominantly interactional function, and with speech acts expressing nonneutral attitudes, and especially a positive attitude to the interlocutor and the communication. There is also throughout the course a dramatic increase in dialogue sections based on shared interpersonal knowledge. The only language feature that increases consistently regardless of K area is implicit grammatical reference: general nouns, demonstrative pronouns and adverbs, and third person pronouns. Yet these are possibly the most distinctive of all the elements of the implicit language. These forms of reference are the most empty semantically and the most exclusive to those without the background knowledge. Thus it appears that, as the group forms, speaking on a topic related to the interlocutor's life becomes increasingly acceptable, and that, whatever the students are talking about, using the vaguest of referring expressions as well as demonstrations of group solidarity becomes more and more the expected behaviour. These aspects of common room chat may be related.
Further Study 145 Other changes occur consistently throughout the course only in K4 dialogues. Here, there is a steady increase in implicit nonanaphoric definite reference, course limited range nouns, metonymical proper names and general "do" verbs. K4 clausal ellipsis increases sharply, and when it contains ingroup code, utterances becomes doubly implicit. Thus, when the students are talking about the course, the further into the year they go, the more they assume that hearers do not need explicit or complete reference because they have the specific referents in their minds as they are interacting. Some changes occur only in the spring term, presumably affected by the course events then. That is when K4 dialogues predominate and this characteristic of the spring explains most changes, especially the peak in nonanaphoric definite reference and lexical contextualisation cues. By then, the "groupusing" function is strong, and students can consciously claim ingroup membership by talking on K4 topics in order to use the group to test their progress and relieve their anxieties. There is a peak in speech acts expressing a negative attitude to a third party or situation and to self, and a trough in private flouting humour, possibly because of the pressures. Students may consciously choose to sound negative and not tease in order to gain the acceptance of the group. The spring sees a peak in hitches in K4 dialogues in the spring. The ingroup code and stress do not mix: sometimes communication breaks down. Thus the rule seems to be that, when the pressure is on, students are expected to talk about the course, using the ingroup code, and be generally serious and negative. Some features are not affected by increased interaction over time but vary according to the topic. Shared interpersonal knowledge sections of all K area dialogues have their own characteristics. There are predictably few instances of superexplicit reference, as referents are guaranteed to need no precise identification. Implicit language is also rare in these sections, but when it is used, it underlines the sharedness by making the language more implicit. Outside the shared interpersonal sections, differences between noncourse K areas and K4 can be felt in all three terms. K4 language always has more implicit nonanaphoric definite reference, special course lexis, more topic shifts containing these and more hitches caused by them. One of the biggest differences between K areas is in the context of noun phrases with nonanaphoric demonstrative reference. In noncourse K areas they tend to occur with a narrative or a situational function and are nonspecific, whereas in K4 they tend to depend on background knowledge and have specific referents. As far as lexical reference is concerned, it is always the K area that influences the choice of lexeme. As concerns humorous conversational implicature, K4 has more private flouting humour and fewer nonflouting funny stories. It is K4's limited range nouns and metonymical proper nouns heavily dependent on the context for their meaning, its generally intangible noncourse nouns and verbs, and its humour in which the truth value could be in doubt for the outsider that makes K4 dialogues more inaccessible than noncourse K area dialogues.
Function Of Implicit Language Implicit language is used mostly in exchanges with an interactional function, and the increase in interactional utterances coincides with the increase in implicit nonanaphoric definite reference, course nouns and general words in K4. The density of the ingroup code increases at the same time as that of nonneutral speech acts increase; it occurs mostly in speech acts expressing a
146 Analysing the language of discourse communities negative attitude to self and the situation. Using these speech acts with ingroup code seems to be a socially acceptable way of talking about oneself and the course. It could be that if implicit language is used in utterances with a socially cohesive function, using it is a high involvement strategy for asserting ingroupness. Only certain features of implicit language are likely to be chosen consciously. It is improbable that students use nonanaphoric definite reference intentionally, although they may use a non anaphoric personal pronoun to hide an identity. It is also improbable that they are aware of their unfinished sentences with conjunctions and fillers. These elements of implicit language can be said simply to reflect ingroup membership. On the other hand, the students may consciously choose to use technical terms and metonymical proper names, and even limited range course nouns, general words and unfinished sentence inviting interlocutor completion. Implicit humorous utterances must be chosen consciously. These elements of implicit language not only reflect ingroup membership, they can be used to show ingroup membership. However, even though it seems that some elements of implicit language are chosen consciously, there is no way of proving that the motive behind the choice is to be sociable, and to claim and assert ingroup membership.
8.2 RECOMMENDATIONS FOR FURTHER RESEARCH This section explores further expansions of the model of implicit language, the generalisability to other discourse communities, and other areas that the implicit language model could be taken into. The model could be expanded by adding a phonological or paralinguistic dimension, and the NNS use could be examined. The model may be generalisable to other student communities or nonacademic discourse communities. Aspects of implicit language that remain to be examined in depth are the relationships between implicitness and situation, gender, race, class and age. It should be emphasised that there is not one example, hypothesis or idea suggested in this section that is based on research or even on real hunches. They are intended rather to be 'provocations' to stimulate the imagination of readers and point to the type of direction that further study could take.
Expansion Of The Model As it was explained in the introduction to the book, this study does not include an analysis of phonological features because it was felt that it this would constitute a whole new study and need specialised equipment. A study of the same data from the point of view of stress and intonation would relate to the description of implicitness, since stress patterns highlight known information or given knowledge, and intonation patterns carry a whole subtext of implied meanings. It might be useful to analyse the relationship between meanings implied by phonological patterns and meanings implied by implicit language. For example, the researcher could analyse whether items of the ingroup code were mainly stressed or unstressed, whether clausal ellipsis and humorous utterances had a fallrise intonation implying an unspoken text, or whether the stress pattern was partly to blame for hitches. There may be changes over time and differences according to topic that analysis would reveal.
Further Study 147 The study could be repeated on another MSc group using a video camera instead of the audio cassette recorder, in order to add a paralinguistic dimension. The video camera was not used here because it was felt that it would be intrusive, and that it would not contribute to the central focus on implicitness. However, facial expressions and hand gestures can add unspoken information; they could be said to constitute a whole system of implied meaning. The researcher who wished to pursue this aspect might be able to distinguish between conventional expressions and gestures, and those which seem to have a particular meaning for the ingroup. They might find, for example, that nonanaphoric personal and demonstrative pronouns are accompanied by a gesture in the direction of the referent, but that as the course went on these were fewer because the referent was assumed as known. This study has excluded NNSs, as it was explained in Chapter 2, because of cultural, linguistic and functional differences, which were observed in the parts of dialogue that were not analysed because they contained mainly NNS, and in the dialogues that were analysed with NNS on the periphery. NNS were kept out of the study because their lack of cultural background knowledge of Britain would have complicated the assumed K area model; their different cultural habits as regards explicitness and their need for explanations of neologisms such as "thingymajog" would have affected the implicit language model; their inability to appreciate always when utterances are intended to be funny or serious would have unbalanced the flouting humour part of the model; and finally their own culture's politeness, cooperative principles and speech act realisations would have influenced the functional side of the analysis. However, these differences constitute a fascinating area to explore in its own right. A study could be carried out to see which parts of implicit language NNS understand and which parts they use successfully. If NNS do use implicit language, the study might show that it grows at a different rate from the NSs' model. Casual observation of DAL MSc courses in general suggests that there is a tendency for the NSs to make a great effort in the autumn term to interact with as many as possible of their NNS colleagues. By the spring term, with the strain of the course, this enthusiasm has abated somewhat; they no longer seem to have the time or the energy to make a great effort. It would be useful for the designing of EAP courses, to study NS interactions with NNSs to discover whether it is their inability to speak vaguely and informally about their studies that keeps them apart. It may simply be that NNS fail to grasp the topic exactly and make acceptable contributions. In the spring and summer terms, NSs seem to limit their social interaction to other NSs and a small number of overseas students who have somehow managed to integrate themselves into the NS group. It would be helpful to study whether these overseas students do better in the course as a result of this.
Generalisability Of The Model This study has the same limitations as all case studies have, in that there can be no guarantee that the particular is representative of the general. Generalisability cannot be proved in case studies or in any psychosociological study involving human beings. Reasonable conclusions drawn from common sense can be made, however. This study has assumed that the findings from the analysis of the six recordees' conversations can be taken as representative of all their conversations, and indeed representative of all NS conversations in the 199192 MSc course.
148 Analysing the language of discourse communities Their individualities do not cause particular tendencies to develop in the data that could make the findings not generalisable to any group of NSs: the conversations of all six contain similar densities of both grammatical and lexical cues, as well as clausal ellipsis and flouting humour. They all talk mainly about Kl and K4 topics, and obey the rules of nonneutral speech acts. The fact that one recordee uses slightly more banter than the others or that one is a little better at giving solidarity than another does not unbalance the general picture. The findings are felt to be generalisable because of the size of the database. Claims have been made in this book about the generalisability of the 199192 DAL MSc discourse community's implicit language to the discourse community of all DAL MSc students and staff. Indeed, whenever the researcher sat in the MSc common room in 1993, 1994 and 1995, and observed students' casual conversations informally, she was aware of the abounding "thing"s, nonanaphoric "he"s and "it's, "I've done my Chomsky and all that stuff's, and injokes in any year's group. This suggests that subsequent years of MSc students have used a similar in group code. Here is an area that could be investigated. The situational and cultural context of the MSc will not change substantially despite the 1999 merger with the Department of Linguistics, so the model of analysis could be applied to future generations of linguistics postgraduates to discover similarities and differences. Care should be taken not to mix PhD postgraduates with MSc taught course postgraduates in any study since they will have a different code. The PhD community is not so tightly knit: they meet colleagues less frequently, sharing the general stress but each at different stages of the process and living in the separate worlds of their particular research topic. On the other hand, a comparison between PhD student groups and MSc student groups would be a valid study. It could be that PhD students have the same technical terms and proper nouns but lack the non anaphoric definite reference, unfinished sentences and injokes, for example. The generalisability could be taken out of the DAL common room; herein lie other researchable areas. It could be investigated exactly what DAL MSc community members have in common with other members of the world discourse community of Teachers of English as a Foreign Language and Applied Linguists, when they sit to chat informally about their specialism. The model could be tested in other linguistics departments and indeed language schools around the world. It may also be the case that the findings about the DAL MSc would be typical of an MSc department in any subject, and possibly typical of any undergraduate student group. Each group would be likely to have the technical terms and proper nouns of the specialism, the common nouns of the course, and in addition the generic model of grammatical implicit contextualisation cues typical of any familiars who have interacted over time. Those who set out to analyse the generalisability of the model to MSc discourse communities of different specialisms should take into account that they may be introducing situational and cultural variables that could affect the extent to which the whole of the model will be transferable. It could be that different course design causes different changes over time: a lighter timetable of deadlines more spread out through the year with no peaks of stress, or days of working jointly in laboratories causing the group to reach this study's spring term stage after a month only, etc. If these departments or schools do not have a common room, the researcher may have to follow members to a cafeteria or to someone's house to record. A building outside the university might not dictate, and might not even permit, conversations about the course.
Further Study 149 Despite the different variables, it is probable that researchers will find certain elements of implicit language growing over time, others being affected by course events and others varying according to the topic. The generalisation could be carried further afield, to nonacademic discourse communities. Research into the casual conversations of any discourse community could be undertaken using the model so long as they meet frequently in informal circumstances and tend to talk about their common interest. These discourse communities could be people united by a common activity such as a team of cleaners from a work place or a group of animal rights protesters. Here there may be more implicit grammatical cues than lexical, because they might have the markers of familiarity but not the theory of an international community. They could be people united by common interests who meet frequently in order to discuss their mutual passion such as football team supporters, science fiction buffs or computer nerds. Here there may be more lexical cues than implicit grammatical ones, because they might share the terminology and specialist knowledge but not know each other very well on a personal level. Studies could be made of people united by neither a common activity nor a common interest but by frequent physical proximity in circumstances in which they chat informally about entities in a wellestablished common background knowledge. These people may form a social group more than a discourse community, since they are unlikely to fulfil Swales' (1990) criteria of being jointly engaged in fulfilling common goals, using established intercommunication mechanisms, a particular genre or specific lexis. These groups are regular cronies in a local pub or neighbours who always meet over the garden fence. Investigation might show that even these people have a particular implicit language, with its proper nouns as in: "When Lil's husband got demobbed, I said /.../ Now Albert's coming back, make yourself a bit smart" (Eliot, 1963) and its nonanaphoric definite specific reference as in "Did he buy that thing in the end?" A researcher may find that technical terms are lacking and there is not enough common knowledge for clausal ellipsis, but that banter, especially amongst pubcronies, assumes greater importance. The language of the family and couples could be analysed along similar lines. Family conversations would presumably be based more on shared interpersonal knowledge than the students' chat. There may be much less phatic talk and more transactional talk. As the woman said the marriage counsellor, "Our small talk is getting smaller and smaller": the more the knowledge known to be shared, the less the need to refer to it. The only problem involved here may be gaining access to couple conversations without coming up against a very strong form of the observer's paradox: a couple might use more implicit language with a cassette recorder trespassing on their intimacy than they would without, so as to hide meaning from the intruder. Shalom (1997: 187) has found that the personal advertisement in lonely hearts columns "resonates with ambiguity" and that "the most 'stereotypical' lexis is imbued with a vagueness" because the intimacy is public. It may be that family member and couple talk contains more clausal ellipsis than friend and colleague talk, and that it is carried over whole exchanges, as in: A: "Where did you put the er...?" B: "Oh, it's on the er..." A researcher might find a higher incidence of hitches caused by grammatical implicit contextualisation cues, as in: A: "Did he do it, then?"
150 Analysing the language of discourse communities B: "Did who do what? Hello?" The limitation of studying families and couples is that the researcher will be unable to study the growth of the implicit language as the relationship will most likely be well established, and most couples will be loath to take a cassetterecorder with them on their first date. Finally, work could be done to discover whether the implicit language model is generalisable to other languages. Every language must have ways of reflecting group common knowledge and asserting ingroup membership. Technical terms and proper nouns are most likely used in the same way in most languages. What is not certain is whether every language uses nonanaphoric demonstrative and personal pronouns, general nouns, metonymical proper nouns, limited range nouns, clausal ellipsis and flouting humour. If parallels and equivalents could be drawn up between languages, this could help language teaching, as well as business and other relationships between countries.
Social Dimensions Of The Model Outside the issue of the generalisability of the implicit language model to different discourse communities and social groups, is the question of what social variables in the context or in the features and characteristics of the speakers themselves in the wider social context are associated with implicit language. Cutting (1998) has looked into the influence of different contexts on implicit language and its function, using as a database a random selection of dialogues from CANCODE (The Cambridge and Nottingham Corpus of Discourse in English, which is part of the Cambridge International Corpus of Cambridge University Press). She examined the extent to which implicit language is used by strangers and familiars in a wide range of contexts: informal, such as the family meal, the gossip session and the chat over a joint task, and formal, such as tutorials, interviews and service encounters. She analysed the function of the implicit language: excluding the outsider, giving little importance to something to appear friendly and avoid being offensive or pretentious, giving little importance to something to be supercilious or critical, being evasive and deliberately withholding information, and compensating for the fact that a word has been forgotten. Cutting (ibid.) found that the most frequent feature of the implicit language in all contexts was the grammatical contextualisation cue and the least frequent was the proper noun. Thus, for example, even in a service encounter about setting up a mortgage, the client uses a general noun and a nonanaphoric personal pronoun: "but I'm doing a PhD and there's various things they want". However, analysis did show that the context that contained the highest density of implicit language was the informal conversation between friends and the lowest was in the formal setting of tutorials, lectures and debates. The following example comes from a gossip session between friends: A: "You know when we were in the pub that time with Stuart, and John was saying to Stuart what you said. I told you that Stuart told Ed, didn't I?" B: "I1 shouldn't have told him all the stuff I did, cos I told him so much." The most frequent function of implicit language was that of giving little importance to the referent, either to be friendly or to be critical. To take an example, when three friends are talking informally as they organise the sale of secondhand goods, they show a friendly attitude by being vague about each one's task, "Cos I'm thinking if we got that then Alison could start doing her stuff." Similarly, when friends are shopping together, they show their disgust of the prices by reducing the value of the referent, "seven pounds for that little thing?" This study of context
Further Study 151 contains three variables relating to implicit language: degree of formality, depth of relationship and function. Other studies could be carried out that take one of the variables and look into it in more depth. Little work has been done on the relationship between implicit language and gender. Tannen (1994: 2334) touches on aspects of it, in her discussion of gender and discourse. She has found that North Americans associate indirectness with female style, and that men and women tend to interpret ambiguity differently. This is a field of study that offers a wide scope of possible research projects. The question of whether women use more implicitness than men could be investigated, as could, for example, the different uses that men and women put implicit language to. Other fertile areas awaiting analysis are the relationship between implicit language and race, class and age. Examples of studies that could be done are a comparison between the casual conversations of British whites and Asians born in Britain, or between working class and middle class language, or the language of teenagers and the language of senior citizens. The assumption being made here is that there will be a difference in degrees of implicitness between each of these crosssections of society. It may be that there is not. Bernstein's (1971) theories about the working class using restricted code and the middle class using elaborated code could be challenged by showing that both classes use the same amount of implicit language. On the other hand, the choice of lexis tends to reflect the level of education and the life experience, events and circumstances: some may use more technical vocabulary and others more general words. It may be for example, that teenagers use more grammatical contextualisation cues for speed of communication, whereas senior citizens use more explicit language because they compensate for their lack of appreciation of hearer knowledge or the hearer's possible loss of information through deafness. Finally, the relationship between implicit language and power needs analysis. It is possible, for instance, that using implicit language gives power, or at least that those with status and dominant social roles, because of situation, gender, race, class or age, use more implicit language than those without. Fairclough (1989) says that the discourse associated with social institutions is opaque in that people are unaware of the underlying meaning and power (p.4041). He points out that texts with power can carry implicit assumptions that they impose upon the reader or listener by making them bring the assumptions into the process of interpretation (p.8283). Wodak (1996) explains that disorders of discourse, or confusion caused by unfamiliar jargon and the structure of the discourse resulting from "gaps between distinct and insufficiently coincident cognitive worlds", can separate "insiders from outsiders, members of institutions from clients of those institutions, and elites from the normal citizen uninitiated in the arcana of bureaucratic language and life" (p.2). She says that the technical terminology doctors employ is often impenetrable and intimidating to patients (ibid.), and that in the media those who lack the relevant background knowledge can be prevented from understanding by imprecise references and pronominalisations, and by intertextuality being only barely explicit (p.101129). If implicit language is used to assert ingroup membership, and asserting membership of a dominant social group is a way of asserting power in order to control or exclude, then implicit language can be said to give power. The casual conversation is said to be one of equal status; exchanges of unequal status could be analysed to find the uses that implicit language is put to. The disorders
152 Analysing the language of discourse communities of discourse and hitches in communication caused by implicitness constitute another area of analysis open to exploration.
8.3 APPLICATIONS Teaching English As A Foreign Language The influence of psychological, sociological and linguistic theory on language teaching has been felt since the 1960's. With the growth of computational research and corpus linguistics in the 1980's, the findings of linguistic analyses of real language and corpora were incorporated into language course books. As McCarthy and Carter say, "By studying the natural patterns of everyday situations, we can come to a much better understanding of what each one demands of participants in terms of cultural and linguistic behaviour. With such information, we are better placed to design syllabuses and materials, evaluate existing ones and, perhaps most important of all, to understand the interpersonal and intercultural areas of language learning that are most sensitive to subtle differences in the manipulation of interpersonal grammars." (1994: 123) Discoveries about the grammar of spoken discourse emerged in the early 1990's, but as Eggins and Slade say, "there is still a paucity of adequate materials for teaching casual conversation to learners of English as a second or foreign language" (1997: 8). EAP course books tend to ignore the social aspect of the students' lives. Lynch's Study Listening (1983) and Lynch and Anderson's Study Speaking (1992), Flowerdew's Academic Listening (1994) and Espeseth's Academic Listening Encounters (1999), for example, train students to listen and speak in formal academic situations, to take notes in lectures, give presentations in seminars and so on. They do not prepare them to sit in the common room and interact socially with their colleagues, referring in an informal way to aspects and components of their academic life. EFL books rarely prepare learners to chat in order to pass the time, keeping the conversation going in groups in which "nothing is happening" at the time. Even Abbs, Cook and Underwood's Realistic English Dialogues (1979), which centres on a group of people "whose social and professional lives interact", contains mostly dialogues and exercises in which they are doing things, such as organising parties, waiting at busstops, and "something is happening". Wong's You Said It! (1999) does concentrate on studentcentred authentic conversations but focuses more on listening and speaking strategies than on the characteristics of spoken language. Carter and McCarthy's Exploring Spoken English (1997) prepares students to appreciate the characteristics of conversational English but does not contain exercises that train them to use it. EFL course books tend to introduce dialogues and activities of a public nature, which require the explicit description of events and expression of opinions. Few course books contain dialogues of a private or intimate nature, containing implicit, vague language. Students are rarely trained to guess what is being talked about by piecing together vague expressions and using the context. If they were trained to recognise the features of ingroup code, they might appreciate when the cause of their lack of comprehension is because of their own linguistic or cultural gaps, and when it is because of the implicit language. They might take comfort from knowing that "lack of precision is one of the most important features of the vocabulary of informal conversation" (Crystal and Davy, 1975: 1112), that native speakers of English need to "draw on pragmatic
Further Study 153 information in order to identify the intended vague category" (Channell, 1994: 143) and that even they can be excluded by implicit language, although they possess the appropriate world knowledge. In addition, they might be helped by being made aware of the function of imprecision. Crystal and Davy (1975) stress that "the use of lexical vagueness is undoubtedly a main sign of social and personal relaxation. /.../ the 'choice' of the vague lexical item is conducive to maintaining the informal atmosphere of the situation". They could also be told that, as Channell (1994) says, imprecision can be a way of being polite, avoiding being offensive, derogatory and pretentious. In order to test whether EAP students can be trained to recognise ingroup code and appreciate its function, Cutting (1999) carried out an experiment on two groups of students, in their normal class time, in the University of Sunderland. The first group consisted of ten toplevel international students on an EAP module, and the second, nine intermediatelevel students on an English for Business Studies module. They were all told what the features of ingroup code were and helped to find them in a dialogue from the MSc common room data. They were then asked to choose a partner and go with a cassette recorder and record any group of NS students, in the cafeterias, common rooms, libraries or halls of residence, who appeared to have known each for some time and were talking informally. Next, they were trained to transcribe and were set a 1,000word project on their transcriptions, as part of their assessed coursework. The projects had to contain a description of the features of ingroup code found in their dialogues and a discussion of whether the ingroup code showed how well the interlocutors knew each other. First drafts were checked by the lecturer, to give pointers for the improvement of linguistic accuracy, style and overall organisation; second drafts were assessed on form and content. At the end of the course, they were asked to fill in a questionnaire with eight open questions, asking them to analyse a new MSc common room dialogue and offer their impressions of the whole exercise. The answers to the questionnaire showed that they could learn to recognise the code and were mostly convinced about the theory that ingroup code was evidence of ingroupness. In answer to the question, "What were your thoughts and feelings when you saw the first dialogue at the beginning of the course?", most said that they did not understand what the speakers were talking about. A few noted that it was a conversation typical of students. In answer to the second question, "What are your thoughts and feelings on seeing this second one?", many students made the point that they still did not understand the dialogue but that they realised that the markers they had seen in the first dialogue were here. When asked "Can you see any markers of intimacy as described in your project?", most students could pick out the nonanaphoric pronouns, general nouns, proper nouns, and a few pointed to the general verbs. A few added other markers of intimacy such as "hesitations, short sentences" and "interruptions, laughing". This seems to suggest that all levels of EAP students can be trained to appreciate ingroup code features. Question four asked, "Would you draw any conclusions about their relationship?", and nearly all of them said the speakers seemed to know each other or were friends. Some of the higher level EAP students did not allow themselves to be so easily persuaded, saying that they could only say with certainty that the speakers were colleagues. The fifth question, "Do you think an outsider might find it difficult to understand what they are talking about?" provoked a unanimous "yes"; half of them blamed the difficulty on the implicitness of the language, and half on the lack of knowledge of the context. The next question was: "To what extent do you agree with the theory of the markers of intimacy?" and many said that they did. One identified personally with the scenario:
154 Analysing the language of discourse communities "I agree at least partly because it is very often the case within my group of friends. I think we developed something like a group code which is probably difficult to understand for outsiders." Those who claimed not to agree said that the presence of ingroup code did not "prove" that the speakers were ingroup members. A few of the EAP students said there are other markers, such as tone, intonation and jokes. The penultimate question asked: "Do you think that studying this language closely has helped you in your socialising with British students?" and a slight majority said "no" in both groups. This question threw up responses that underlined the isolation of the international student: "As I don't have British students as friends, I can't see if this has helped me or not with them." However, those who felt that it had helped them offered answers such as: "if I don't understand a conversation between two persons that's because they have a closed talking." One can conclude that EAP students can be made aware of the function of ingroup code, and that they can reassured that it is not always their own language competence level that makes conversations difficult to understand. The exercise does not immediately improve integration with British students, but this may be because they are unaware of the changes that have occurred in them or because it takes time to realise whether the knowledge about ingroup code has improved their social lives. The answers to the last question, "What is your assessment of the whole exercise of doing the project on this aspect of language? Would you recommend it for future EAP/EBS groups?", were most encouraging. Nearly all of them said "yes", and gave explanations such as these: "I found this project really thrilling since I had never done such a study on language." "I do think that many students will have as much fun as I did doing the recording." "It is a very good and innovative project and I strongly recommend it to be continued and to the other groups." The enthusiasm of their responses here ("really thrilling", "fun", "very good and innovative" and "strongly recommend") and other comments that it had been "interesting and helpful" led the researcher to believe that this was more than the routine polite comments to the teacher at the end of a course. The next stage is to determine whether international students want to produce implicit language, whether they want to sound more like native speakers of English in order to make their own language more acceptable for their colleagues and claim ingroup membership. Foreign language learners tend to look for the most explicit way of expressing themselves in order to ensure that they communicate their meaning, and generally speaking EAP books aim to teach vocabulary with content. The result is that learners often sound "bookish and pedantic" because they do not know how to use vague expressions (Brown, 1979). Brown says that one aspect of acquiring a second language is "learning to be imprecise". Cook (1989) suggests the use of exercises that oblige the student to evaluate the known information and remove superfluous information from passages. This implies removing complete phrases and words from sentences. Controlled exercises could be devised to train students to remove content from words and leave less contentful words in their place, and therefore remove explicitness and leave implicitness. They could then be given free practice in using the discourse community implicit language, talking in nontransactional conversations about their course subjects.
Further Study 155 However, it should be kept in mind that not all students may want to sound quite to British as they would if they used this form of speaking. They may feel that they wish to preserve more of their own cultural identity. It is possible that they have different ways of expressing ingroupness in their own cultures that they will prefer to reflect. Beebe points out: "Second language learners may never attain nativelike proficiency to the best of their ability because they may find that the reward of being fluent in the target language is not worth the cost in lost identification and solidarity with their own native language group." (1988: 63) It is because of this that in the design of lessons based on the vague, care should be taken to determine whether students would rather simply be made aware of the language and thus only listen to it and pick out the implicit features and functions.
Clinical Pragmatics The findings in this study about the function of implicit language of reflecting and possibly claiming ingroup membership might also be of some use to those working in clinical pragmatics. A great deal of work has been done in this field on the speaker's compliance with the maxims of quantity and relevance and his ability to refer effectively, and the listener's interpretation of speech acts. The goal of clinical pragmatics is described by Smith and Leinonen (1992) as that of characterising "clients' communicative behaviour and ability with a view to diagnosis and remediation by considering not only the role of the client in communication but also of the context of situation and of those interacting with the client." (p.44). Clinical pragmatics treats clients with communicative disabilities because of semanticpragmatic disorders, by developing their pragmatic knowledge. Smith and Leinonen say, "Since communicative performance demands the integration of complex knowledge and a variety of skills, the client's confidence to experiment with imperfect abilities is crucial to the acquisition and mobilisation of communicative competence." (ibid.: 252). The part of clinical pragmatics that this study relates to is that which treats clients who have difficulty taking into account the hearer's knowledge and who misinterpret the speakers' meaning. As Smith and Leinonen say, people who have trouble communicating may not be performing speech acts successfully, "because the right things are not being said in the right way at the right time and because the implications of what is said by either party are not fully understood by at least one of the participants" (p.149). Bryan (1988) found that adults with right hemisphere damage had difficulties in speech act comprehension and comprehension of non literal meanings, an inability to use contextual information to derive meaning and appreciate metaphors, humour, etc. Clinical pragmatics deals with cases of speakers whose referring items are too implicit. Smith and Leinonen describe the case of a pragmatically dysfunctioning boy whose utterances lacked the theme part of information structure corresponding to the unknown. McTear (1985) cites the case of a boy with a "failure to use ellipsis, difficulty in making factual and especially causal information clear, ... failure to take his partner's knowledge into account or failure to differentiate between his own and other people's knowledge" (in Smith and Leinonen 1992: 158). Hassibi and Breuer (1980) found psychotic children who did not realise that words must have shared meaning to be understood and used pronouns without having provided the
156 Analysing the language of discourse communities reference. Wodak (1996: 145146) found that in psychotherapy sessions, psychotic patients spoke with "private meanings and private language". Each one's language had particular associations and their connotations could only be understood in the context of their life history and expectations. Schizophrenics had private symbols and structures "similar to the mechanisms of dream texts". It may be that dialogues such as the ones in the data could be used to demonstrate to "clients" with semanticpragmatic disorders how speakers who share a great deal of common knowledge, because they are members of an ingroup, take into account the common knowledge and refer to entities in a vague way and yet identify them sufficiently to guarantee communication. Ingroup conversations could also be discussed by clients and therapists in terms of the nonliteral meanings of the humorous conversational implicature. Clients could be made to see the social function of implicit language, and how using it can make them more accepted as one of the group or at least one of a pair.
Forensic Linguistics The findings in this study about the implicit and vague nature of the language of people who share common knowledge could also help those involved in forensic discourse analysis. This discipline is a relatively new one (the first British conference of Forensic Linguistics was held in 1992) and the methodology of forensic discourse analysis is still being developed. In forensic linguistics, linguists are asked to analyse the language of police transcripts and either say whether it is authentic or fabricated, or determine the intentions of the interlocutors. They can also be asked to analyse taperecordings of suspects' conversations and determine the intentions and speech acts involved. The importance of understanding implicit, vague language and recognising the speaker's meaning and intention has been referred to in the literature on this subject. Coulthard (1992) says that forensic discourse analysis was born as a result of linguists being asked to check whether the written records of witness and suspect interviews and statements have been tampered with by whoever owned the recordings, in most cases the police, and may die as the police forces turn to taperecording or videotaping interviews. He finds that the fabricator of interview and statement records makes three kinds of mistakes: psycholinguistic ones, ones of quantity and ones of discourse structure. The psycholinguistic mistake is to copy a version of one verbatim record to another word for word, overlooking the fact that speakers do not remember the exact words that they or anyone else has used. The discourse structure mistake is to reproduce the interview with rapid topicshifts, omitting the characteristic topiclinked followup questions. The quantity mistake is to break the quantity maxim and make nominal groups and clauses "overexplicit". A quantity mistake consists of fabricating discourse in which a speaker gives information that he is known to know, or in which a speaker uses more prehead or posthead modifiers than is necessary or relevant, as in "Walker was carrying two white plastic carrier bags. Walker gave me one of the white plastic carrier bags." The fabricator is "creating his text with the overhearer, in this case the court, in mind", forgetting that conversations usually contain
Further Study 157 utterances shaped for a specific addressee with common knowledge and are consequently "frequently opaque and at times incomprehensible to an overhearer" (p.249). Shuy (1993) examines transcripts of recordings that have not been tampered with but are being used to accuse a suspect of committing a language crime, such as threatening, offering a bribe and soliciting. He finds that a linguist is needed to analyse the discourse, because laymen often have a number of misconceptions about language. These misconceptions are: believing that meaning is found in individual words, that all people in a conversation understand the same things by their words and that people say what they mean and intend. With regard to the misconception about speaker's meaning and intention, he suggests that a linguist very familiar with the taperecording is in the best position to analyse speakers' compliance with the maxim of relevance and speech acts such as promising and admitting. The linguist can determine how these are affected by factors such as setting and role relationships and decide whether a language crime has been committed. Discussing the second of these misconceptions, Shuy cites what he calls "vague reference", as in "The man came today" used by a wife to a husband and "incomplete sentences or ambiguously uttered ones" in which third person personal pronouns are "the major culprit" (p. 16). When pronouns are "imperfectly referenced", says Shuy, even though the interlocutors share knowledge of the world, feelings, beliefs and concerns, the listener may infer the wrong meaning and the speaker may not realise that he has been misunderstood. He suggests that often only a linguist who has listened to the tape 50 times and can "walk in the shoes" of the participants, can detect such ambiguities and thus say whether Y accepts X's bribe, agrees to his illegal plan, etc. These findings can be stated in the terms of the present study. Coulthard is citing the repeated superexplicit noun phrase as an unnatural feature of any conversation, once the topic has been established and enough background information provided. Shuy is quoting the use of an implicit nonanaphoric definite noun phrase with an unmodified general noun and implicit nonanaphoric personal pronouns, as markers of interlocutors who share knowledge of the world, feelings, etc. and as sources of communication hitch. The model of implicit contextualisation cues and other pointers to common knowledge which are markers of intimacy could be used in forensic discourse analysis. The model could be used to detect tampering with transcribed records of interviews taking place well into the questioning, i.e. once the police officer can be assumed to be familiar with most of the referents. The model could also be used in the analysis of conversations between suspects. It could help towards determining whether suspects were speaking for the first time or whether they had known each other for a length of time before the conversation and could possibly be accused of working in cahoots. The model may help towards determining whether suspects were conspiring. If the language was more vague and implicit than required, they could be said to have the intention to make the communication impenetrable to any overhearer, and to have something to hide. The "giveaway" might be a hitch in communication because of the language being too implicit for even the intended hearer.
8.4 CONCLUSION This book has described a longitudinal study of the formation of the implicit language of an academic discourse community. The study has shown that the knowledge area assumed has as
158 Analysing the language of discourse communities great an influence on the form of the language as does increasing knowledge with interaction over time. It has been argued that to use implicit language is to assert the ingroup membership needed for solidarity and normalitytesting. The model of implicit language is hoped to contribute to the field of applied linguistics, to provide a model and a stimulus for further investigation, and to be useful in the fields of Teaching English as a Foreign Language, clinical pragmatics and forensic linguistics.
APPENDIX I QUESTIONNAIRE FOR THE Six RECORDEES
INTRODUCTION Thank you very much for agreeing to give up some of your time to helping me. Imagine that you were in the AL common room, for whatever reason, on four occasions this year and overheard these four fragments of conversation. Please try to answer the questions. First, kindly indicate which of these five categories you belong to: 1 I am a university graduate presently resident in Edinburgh. I am not a language teacher, have not studied anything related to linguistics and know nothing about EU AL M.Sc. courses. 2 I am presently resident in Edinburgh, am a language teacher/linguist but know nothing about EU AL M.Sc. courses. 3 I am presently resident in Edinburgh, am a language teacher/linguist, have general information about EU AL M.Sc. courses, but know nothing specifically about the 9192 course. 4 I am a 9192 EU AL M.Sc. student. I am not on the recordings. 5 I am one of the 9192 EU AL M.Sc. students who is on the recordings. 6 I am none of these. NOTE: Please put any general comments you may have at the end of the questionnaire.
DIALOGUE ONE Text MM
So you were in em (0.5) AM In Sudan. MM Sudan. AM Mmm. I've been there for three years but um one year em promoted one year and that included two years. MM Was that with the Council? AM Nono. Em (1) That's for the Sudanese government.(0.5) And then er (0.5) I actually applied for a job in Gabbitas through Gabbitas. CM Yeah they they do a lot of private em secondary school don't they. AM So I was with them. But.
160 Analysing the language of discourse communities (2)
BM AM BM AM BM
CM BM AM
BM AM BM MM
BM MM CM MM BM MM BM MM BM MM AM
MM AM BM AM
What did you say? Advertised through what. Gabbitas and Thring is urn an educational er appointments agency. (1) Yeah, through them we'd intended to go to Argentina er (4) but that fell through. I actually did (2) have a job set up in Argentina. Same thing happened. Fell through. It was inflation. And then it came back on again. Oh. By that stage I thought no way. (ha ha) Cos of inflation. You know they said it was under control. Well, I was looking at it later and I thought well= =You fancy getting it clear. Yeah II don't fancy this. Oh so you know because my wife had to change her name. (2) Because apparently the Argentinean authorities remust have you must have the same name on your passport as your birth certificate. And my wife had changed her name because her parents had (0.5) divorced and em her mother had married again. (0.5) But she'd always been known by one name. So we changed it back and in between that of course em we'd been married at that stage so now she has three names. (Hehhehhehhehheh) and then of course we didn't get the job. (13) So how about yourself MM? Em I taught in Japan like CM did a couple of years eh south coast for years (0.5) and couple of months here and there Portugal, Italy, was in Saudi Arabia for three months but. Well did you like it. It was interesting. Three months is enough though. Were you on a compound or were you. Yeah. Mainly in a compound. Teaching military telecommunications. (3) That was some time ago was it? About two and a half three years ago yeah. Ah right. (4) A lot of the Sudanese in Saudi. I mean a lot of the (0.5) a lot of the clerical work (0.5) a lot of the stuff like I mean say midmanagement work are done by Palestinians. Yes= ^Egyptians Sudanese Somalis. I think that was one of the problems with Kuwaitis. (2) Half their work force about a million people was made up of (0.5) foreign Arabs. I think the Sudanese quite a lot of Sudanese go to em Saudi Kuwait and places like that. They bring in a lot of foreign currency that way. (3) I think I might go and have another bun. I thought I was going to get another one. Could you get me a tuna and sweet corn one please. Me as well.
Appendices 161 Questions 1
2
How long would you think these people have been friends? just met 13 weeks 12 months 23 months 34 months 45 months 56 months 13 years What was it about the dialogue that suggested to you the length of acquaintance
3
What title would you give to this dialogue to show what the people are talking abou! what they are doing, what the main topic is?
4
Could you have joined in this dialogue, had you been invited? Why or why not?
5
X asks AM, "Was that with the Council?" What is the Council?
6
AM says, "And then of course we didn't get the job." Why would he say "of course"
7
AM asks, "So how about yourself, X?" What does AM want X to tell him about
8
After saying that it was interesting in Saudi Arabia, X says, "Three months is enoug though." What do you feel he means by this?
9
BM says, "I think the Sudanese quite a lot of Sudanese go to em Saudi, Kuwait an places like that." What would you understand by "places like that?"
[There were three other dialogues and questions along the lines of these. Dialogue One serve an example.]
GENERAL QUESTIONS ABOUT YOUR CONVERSATIONS Relationships Look at the list of students' names and put a circle round those you spoke to for at least 101 minutes three times a week, and underline those you spoke to about 25 minutes about once week. [These have been deleted to preserve their anonymity]
Function Of Talk Why do you talk? 1 to pass the time 2 because you enjoy the company 3 to see how you fare compared with others 4 to relieve your anxieties 5 to have a laugh
162 Analysing the language of discourse communities 6 other reasons with the people you circled: with the people you underlined:
Topics What do you talk about? 1 the world 2 Edinburgh 3 everyday life 4 linguistics 5 the course 6 the staff 7 other students 8 your personal life 9 your feelings 10 other with the people you circled: with the people you underlined:
Language Which do you use? 1 informal language 2 slang and obscenities 3 emotional language 4 specialised AL terms and names with the people you circled: with the people you underlined:
Conversation Flow Which three of the following characteristics do you like best in people? Those who 1 speak for a long time on one topic 2 switch from topic to topic 3 speak fast without pauses 4 speak slowly with pauses 5 sometimes speak along with you 6 speak only when you've finished your turn
Recording Did my recording affect your speech? If so, how?
Appendices 163
APPENDIX II IMPENETRABILITY TEST SCORES
Figure 1: Mean of correct answers for topic knowledge and reference knowledge questions Subjects reference knowledge topic knowledge (total 20 questions) (total 4 questions) Group A Group B Group C
11 14 17
2 3 4
Figure 2; Correlation between topic knowledge scores and reference knowledge scores Group B Total Group A Dialogue 1 Dialogue 2 Dialogue 3 Dialogue 4 Dialogue 1, 2, 3 and 4
0.409 0.524 0.602 0.121 0.487
0.236 0.096 0.390 0.631 0.300
0.098 0.135 0.482* 0.217 0.590*
This page intentionally left blank
Appendices 165
APPENDIX III QUESTIONNAIRE FOR THE IMPENETRABILITY TESTS
INTRODUCTION Thank you very much for agreeing to give up some of your time to helping me. Imagine that you were in the AL common room, for whatever reason, on SIX occasions in the academic year 199192 and overheard these six small fragments of conversation. Please try to answer the questions. First, I need you to tick whichever of these five categories that apply to you: 1 I am a university graduate presently resident in Edinburgh. 2 I am a language teacher/linguist. 3 I have information about / experience of one or several EU AL MSc courses. 4 I am a 9192 EU AL MSc student. 5 I know I am one of the six people who were recorded for this research. 6 I am none of these
DIALOGUE ONE Text MM
So you were in em (0.5) AM In Sudan. MM Sudan. AM Mmm. I've been there for three years but um one year em promoted one year and that included two years. MM Was that with the Council? AM Nono. Em (1) That's for the Sudanese government.(0.5) And then er (0.5) I actually applied for a job in Gabbitas through Gabbitas. CM Yeah they they do a lot of private em secondary school don't they. AM So I was with them. But. (2) BM What did you say? Advertised through what. AM Gabbitas and Thring is um an educational er appointments agency. (1) Yeah, through them we'd intended to go to Argentina er (4) but that fell through.
166 Analysing the language of discourse communities BM
AM
I actually did (2) have a job set up in Argentina. Same thing happened. Fell through. It was inflation. And then it came back on again. Oh.
Questions 1
What is the main topic of this dialogue; what are the people talking about?
2.1
What is "the Council"?
2.2
What do you imagine the "job" might have been in "And then er I actually applied for a job in Gabbitas through Gabbitas." ?
2.3
Who exactly does "they" in "they do a lot of private em secondary school" refer to?
2.4
What does "it" in "It was inflation" refer to?
2.5
What does "it" in "And then it came back on again" refer to?
[There were five other dialogues and questions along the lines of these. Dialogue One serves an example.]
REFERENCES
Abbs, B., U. Cook, and M. Underwood (1979). Realistic English Dialogues. Oxford University Press, Oxford. Altman, I. and D. A. Taylor (1973). Social Penetration: The Development Of Interpersonal Relationships. Holt, Reinhart and Winston, New York. Argyle, M., A. Furnham, and J. A. Graham (1981). Social Situations. Press Syndicate of the University of Cambridge, Cambridge. Argyle, M. and M. Henderson (1985). The Anatomy Of Relationships. Heineman, London. Atkinson, P. (1981). The Clinical Experience. Gower Publishing Company, Hampshire. Austin, J. L. (1962). How To Do Things With Words. Oxford University Press, London. Beebe, L. M. (1988). Five sociolinguistic approaches to second language acquisition. In: Issues On Second Language Acquisition. Multiple Perspectives (L. M. Beebe, ed.), Newbury House, Rowley, Massachusetts. Bell, D. V. J. (1994). Negotiation in the workplace: the view from a political linguist. In: The Discourse Of Negotiation: Studies Of Language In The Workplace (A. Firth, ed.) Elsevier Science Ltd., Oxford. Berger, C. R. and J. J. Bradac (1982). Language And Social Knowledge. Edward Arnold Ltd., London. Berger, P. L. and T. Luckman (1966). The Social Construction Of Reality. Doubleday, New York. Bernstein, B. (1971). Class, Codes And Control. Vol.1. Routledge and Kegan Paul, London. Berry, M. (1981). Systemic linguistics and discourse analysis: a multilayered approach to exchange structure. In: Studies In Discourse Analysis (M. Coulthard and M. Montgomery), Routledge and Kegan Paul, London. Berry, M. (1989). An Introduction To Systemic Linguistics: 1 Structures And Systems. University of Nottingham Press, Nottingham. Biber, D. (1988). Variation Across Speech And Writing. Cambridge University Press, Cambridge. Blakemore, D. (1992). Understanding Utterances: An Introduction To Pragmatics. Blackwell Publishers, Oxford. Boden, D. (1994). Agendas and arrangements: everyday negotiations in meetings In: The Discourse Of Negotiation: Studies Of Language In The Workplace (A. Firth, ed.) Elsevier Science Ltd., Oxford. Brazil, D. and M. Coulthard (1980). Discourse Intonation And Language Teaching. Longman, London. Brown, E. and J. Miller (1980). Syntax: A Linguistic Introduction To Sentence Structure. Hutcheson University Library, London. Brown, G. and G. Yule (1983). Discourse Analysis. Cambridge University Press, Cambridge. Brown, J. (1979). Vocabulary: learning to be imprecise. Mod. Eng. Teacher, 7/1, 257. Brown, P. and S. Levinson (1978). Politeness. Cambridge University Press, Cambridge.
168 Analysing the language of discourse communities Bryan, K. L. (1988). Assessment of language disorders after right hemisphere damage. Br. J. Of Disorders Of Communication, 23, 111125. Carlson, L. (1983). Dialogue Games. Reidel Publishing Company, Dordecht, The Netherlands. Carlson, T. B. (1993). Context for comprehension. In: Arenas Of Language Use (H. H. Clark, ed.), University of Chicago Press. Chafe, W. L. (1976). Giveness, contrastiveness, definiteness, subjects, topics and points of view. In: Subject And Topic (C. N. Li, ed.) Academic Press, New York. Chafe, W. L. (1982). Integration and Involvement in speaking, writing, and oral literature. In: Spoken And Written Language: Exploring Orality And Literacy (D. Tannen, ed.), pp. 26172. Ablex, Norwood NJ. Channell, J. (1994). Vague Language. Oxford University Press, Oxford. Carter, R. and M. McCarthy (1997). Exploring Spoken English. Cambridge University press, Cambridge. Clark, H. and G. Murphy, G. (1982). Audience design in meaning and reference. In: Language And Comprehension (J. LeNy and W. Kintsch, eds.), North Holland Publishing Company, Amsterdam. Clark, H. H. and Clark, E. V. (1977). Psychology And Language. Harcourt, Brace, Jovanovich, New York. Clark, H. H. (1993). Arenas Of Language Use. University of Chicago Press, Chicago. Clark, H. H. (1997). Dogmas of understanding. Discourse Processes, 23, 567598. Cook, G. (1989). Discourse. Oxford University Press, Oxford. CookGumperz, J. and J. J. Gumperz (1994). The politics of a conversation: conversational inference in discussion. In: What's Going On Here?: Complementary Studies of Professional Talk (A. D. Grimshaw, ed.), Ablex, New Jersey. Coulthard, M. (1977). An Introduction To Discourse Analysis. Longman, London. Coulthard, M. (1992). Forensic discourse analysis. In: Advances In Spoken Discourse Analysis (M. Coulthard, ed.), Routledge, London. Coulthard, M. and M. Montgomery (eds.) (1981). Studies In Discourse Analysis. Routledge and Kegan Paul, London. Craig, R. T. and K. Tracy (eds.) (1983). Conversational Coherence. Sage Publications, California. Crystal, D. and D. Davy, D. (1975). Advanced Conversational English. Longman, London. Cutting, J. (1991). Discourse Analysis: Topics And Intimacy Levels At The Bun Lunch. Unpublished MSc dissertation. Dept. of Applied Linguistics. University of Edinburgh. Cutting, J. (1998). The function of inexplicit language in 'CANCODE' casual conversations. Unpublished paper at Sociolinguistics Symposium 12. University of London. Cutting, J. (1999). Can we train international students to use academic ingroup code? Unpublished paper at IATEFL 99. University of Edinburgh. de Beaugrande, R. and W. Dressier (1981). Introduction To Text Linguistics. Longman, London. Dews, S., J. Kaplan and E. Winner (1995). Why not say it directly? The social functions of irony. Discourse Processes, 19, 347 367. Dines, E. (1980). Variation in Discourse and "stuff like that". Lang, in Soc., 9, 1331. Drew, P. and J. Heritage (1992). Talk At Work (Interaction In Institutional Settings). Cambridge University Press, Cambridge. Duck, S. (1991). Friends For Life. Harvester Wheatsheaf, Hertfordshire.
References 169 Duranti, A. and C. Goodwin (eds.) (1992) Rethinking Context: Language as an interactive phenomenon: Studies in the social and cultural foundations of language No.ll. Cambridge University Press, Cambridge. Edmondson, W. (1981). Spoken Discourse. Longman, London. Eggins, S. and D. Slade (1997). Analysing Casual Conversation. Cassell, London. Eliot, T. S. (1963). Collected Poems. Faber and Faber Ltd., London. Espeseth, M. (1999). Academic Listening Encounters. Cambridge University Press, Cambridge. Fairclough, N. (1989). Language And Power. Longman, London. Fairclough, N. (1995). Critical Discourse Analysis: The critical Study Of Language. Longman, London. Fawcett, R. (1973). Systemic Functional Grammar In A Cognitive Model Of Language. University College of London: mimeographed. Fillmore, C. J. (1994). Humor in academic discourse. In: What's Going On Here?: Complementary Studies of Professional Talk (A. D. Grimshaw, ed.), Ablex, New Jersey. Firth, A. (1994). The Discourse Of Negotiation: Studies Of Language In The Workplace. Elsevier Science Ltd., Oxford. Firth, J. R. (1957). Papers In Linguistics. 1934-51. Oxford University Press, Oxford. Flowerdew, T. (1994). Academic Listening. Cambridge University Press, Cambridge. Fodor, J. D. (1977). Semantics: Theories Of Meaning In Generative Grammar. Harvester, Hassocks. Ford, C. E. (1993). Grammar In Interaction. Cambridge University Press, Cambridge. Fox, B. A., M. Hayashi and R. Jasperson (1996). Resources and repair: a crosslinguistic study of syntax and repair. In: Interaction and Grammar (Studies in Interactional Sociolinguistics 13) (E. Ochs, E. A. Schegloff and S. Thompson, eds.), Cambridge University Press, Cambridge. Garfinkel, H. (1967). Studies In Ethnomethodology. Prentice Hall, Englewood Cliffs, NJ. Giles, H. (1977). Language, Ethnicity And Intergroup Relations. Academic Press Inc., London. Giles, H. and P. F. Powesland (1975). Social Style And Social Evaluation. Academic Press, London. Giles, H. and R. St.Clair (1979). Language And Social Psychology. Basil Blackwell, Oxford. Goffman, E. (1963). Behaviour In Public Places. Free Press, New York. Goffinan, E. (1971). Relations In Public. Penguin, London. Goldberg, J. A. (1983). A move towards describing conversational coherence. In: Conversational Coherence (R. T. Craig and K. Tracy, eds.), Sage Publications, California. GoldmanEisler, F. (1968). Psycholinguistics: Experiments In Spontaneous Speech. Academic Press, New York. Goodwin, C. (1979). The interactive construction of a sentence in natural conversation. In: Everyday Language: Studies in Ethnomethodology (G. Psathas, ed.). Irvington Publishers Inc., New York. Goodwin, C. (1996). Transparent vision In: Interaction and Grammar (Studies in Interactional Sociolinguistics 13) (E. Ochs, E. A. Schegloff and S. Thompson, eds.), Cambridge University Press, Cambridge. Goodwin, C. and A. Duranti (1992) An introduction. In: Rethinking Context: Language As An Interactive Phenomenon: Studies In The Social And Cultural Foundations Of Language No.ll. (A. Duranti and C. Goodwin, eds.). Cambridge University Press, Cambridge. Goodwin, C. and M. H. Goodwin (1992). Assessments and the construction of context. In: Rethinking Context: Language As An Interactive Phenomenon: Studies In The Social
170 Analysing the language of discourse communities And Cultural Foundations Of Language No. 11. (A. Duranti and C. Goodwin, eds.). Cambridge University Press, Cambridge. Graesser, A. C. and L. F. Clark (1985). Structures and Procedures of Implicit Knowledge (Advances in Discourse Processes XVII). Ablex Publishing Corporation, New Jersey. Grice, H. P. (1975). Logic and conversation. In: Pragmatics (Syntax And Semantics, Vol. 9) (P. Cole and J. Morgan, eds.), Academic Press, New York. Grimshaw, A. D. (ed.) (1994). What's Going on Here?: Complementary Studies of Professional Talk. Ablex, New Jersey. Gumperz, J. (1982). Discourse Strategies. Cambridge University Press, Cambridge. Gumperz, J. (1992). Contextualisation and understanding. In: Rethinking Context: Language as an interactive phenomenon: Studies in the social and cultural foundations of language No. 11. (A. Duranti, A. and C. Goodwin, eds.), Cambridge University Press, Cambridge. Halliday, M. A. K. and R. Hasan (1976). Cohesion In English. Longman, London. Halliday, M. A. K. and R. Hasan (1985). Language, Context, And Text: Aspects Of Language In A Social-Semiotic Perspective. Oxford University Press, Oxford. Hasan, R. (1984). Ways of saying and ways of meaning. In: The Semiotics Of Culture And Language (R. Fawcett, M. A. K. Halliday, A. Makkai and S. M. Lamb, eds.), Francis Pinter, London. Hassibi, M. and H. Jnr. Breuer (1980). Disordered Thinking And Communication In Child. Plenum Press, New York. Hawkins, J. A. (1978). Definiteness And Indeflniteness. CroomHelm, London. Heritage, J. (1984). Garfmkel And Ethnomethodology. Basil Blackwell, Oxford. Hinde, R. A. (1987). Individual, Relationships And Culture. Cambridge University Press, Cambridge. Hoey, M. P. (1983). On The Surface Of Discourse. George Allen and Unwin, London. Hoey, M. P. (1991). Patterns Of Lexis In Text. Oxford University Press, Oxford. Huddleston, R. (1988). English Grammar: An Outline. Cambridge University Press, Cambridge. Hurford, J. R. and B. Heasley (1983). Semantics: A Coursebook. Cambridge University Press, Cambridge. Hymes, D. (1972). Models of the interaction of language and social life. In: Directions In Sociolinguistics: The Ethnography Of Communication (J. Gumperz and D. Hymes, eds.), pp. 3571. Holt, Rinehart and Winston, New York. Jefferson, G. (1978). Sequential aspects of storytelling in conversation. In: Studies In The Organisation Of Conversational Interaction (J. Schenkein, ed.), pp. 4338. Free Press 29, New York Jefferson, G. (1990). List construction as a task and resource. In: Interaction Competence (G. Psathas, ed.), International Institute for Ethnomethodology and Conversation Analysis and University Press of America, Washington DC. Jefferson, G., H. Sacks, and E. A. Schegloff (1987). Notes on laughter in pursuit of intimacy. In: Talk And Social Organization (G. Button and J. R. E. Lee, eds.), Multilingual Matters, Clevedon, England. Johnson, S. (1993). Male gossip. Unpublished paper in 26th BAAL Annual General Meeting, University of Salford. Kanfer, F. H. (1960). Verbal rate, eyeblink and content in structured psychiatric interviews. J. Of Abnormal & Soc. Psych., 61, 341347. Kasl, S. V. and G. F. Mah (1965). The relationship of disturbances and hesitations in onnntaruaniic erw»i=»fVi tr» anviptv .J Of Ppmnnfll
SnC. PffVch..
1. 425-33.
References 171 Kelly, H. H., E. Bersheid, A. Christensen, et aL (1983). Close Relationships. W. H. Freeman and Company, New York. Kennedy, G. (1987). Quantification and the use of English: a case study of one aspect of the learner's task. Applied Lings, 8/3, 26486. Kreckel, M. (1981). Communicative Acts And Shared Knowledge In Natural Discourse. Academic Press, London. Labov, W. (1972). Introduction. In: Sociolinguistic Patterns, pp. xiiixviii. University of Pennsylvania Press, Philadelphia. Lakoff, G. (1972). Hedges: a study in meaning criteria and the logic of fuzzy concepts. In: Chicago Linguistic Society Papers. Chicago Linguistic Society, Chicago. Laver, J. and S. Hutcheson (1972) Communication In Face To Face Interaction. Penguin, Middlesex. Lawton, D. (1965). Social class language differences in individual interview. Mimeographed article. Institute of Education, University of London. LeckieTarry, H. (1995). Language And Context: A Functional Linguistic Theory Of Register. Pinter, London. Leech, G. (1983). Principles Of Pragmatics. Longman, New York. Levinson, S. (1983). Pragmatics. Cambridge University Press, Cambridge. Levy, D. (1979). Communicative goals and strategies: between discourse and syntax. In: Syntax And Semantics (T. Givon, ed.), Vol.12, pp. 183210. Academic Press, London. Li, C. N. (ed.) (1976). Subject And Topic. Academic Press, New York. Lynch, T. (1983). Study Listening. Cambridge University Press, Cambridge. Lynch, T and K. Anderson (1992). Study Speaking. Cambridge University Press, Cambridge. Lyons, J. (1977). Semantics. Cambridge University Press, Cambridge. Lyons, J. (198 la). Language, Meaning And Context. Fontana, London. Lyons, J. (1981b). Language And Linguistics: An Introduction. Cambridge University Press, Cambridge. Malinowski, B. (1923). Phatic communion. In: Communication In Face To Face Interaction (J. Laver and S. Hutcheson), Penguin, Middlesex. Martinet, A. (1970). Elements De Linguistique Generale. Libraire Armand Colin, Paris. Matarazzo, J. D. A. (1970). Interviewee speech behaviour under different content conditions. J. Of Applied Psych., 54, 1526. McCarthy, M. (1991). Discourse Analysis For Language Teachers. Cambridge University Press, Cambridge. McCarthy, M. and R. Carter (1994). Language As Discourse: Perspectives For Language Teaching. Longman, London. McTear, M. (1985). Pragmatic disorders: a case study of conversational disability. Br. J. Of Disorders Of Communication, 20/2, 12942. Milroy, L. (1987). Observing And Analysing Natural Language. Basil Blackwell, Oxford. Moscovici, S. (1967). Communication processes and properties of language. Advances Of Experimental Social Psychology. Nash, W. (1985). The Language Of Humour. Longman, New York. Ochs, E., E. A. Schegloff and S. Thompson (eds.) (1996). Interaction and Grammar (Studies in Interactional Sociolinguistics 13). Cambridge University Press, Cambridge. Orestrom, B. (1983). Turntaking in English conversations. PhD dissertation. Lund, Sweden. Osgood, C. E. and E. G. Walker (1959). Motivation and language behaviour: a content analvsis of suicide notes. J. Of Abnormal And Soc. Psvch.. 59. 5867.
172 Analysing the language of discourse communities Palmer, F. R. (1981). Semantics (2nd.edition). Cambridge University Press, Cambridge. Planalp. S. and A. Benson, A. (1992). Friends' and acquaintances' conversations 1: perceived differences. InJ. OfSoc. And Personal Relationships, 9, 483506. Plum, G. A. (1988). Text And Contextual Conditioning In Spoken English: A Genre-based Approach. PhD thesis. University of Surrey. Prince, E. F. (1981). Toward a taxonomy of givennew information. In: Radical Pragmatics (E. Cole, ed.), Academic Press, New York. Quirk, R., R. Greenbaum, G. Leech, J. and Svartvik (1985). A Comprehensive Grammar Of The English Language. Longman, New York. Ratner, S. C. and R. E. Rice (1963). The effect of the listener on the speaking interaction. Psych. Rev., 13, 26568. Sacks, H. (1972). Lecture Notes. School of Social Science, University of California at Irvine. Sacks, H. (1992). Lectures on Conversation: Volume I. Blackwell, Oxford. Sadock, J. (1985). Whither radical pragmatics? In Meaning, Form, And Use In Context: Linguistic Applications (Georgetown University Round Table on Languages and Linguistics) (D. Schiffrin, ed.), pp. 13949. Georgetown University Press, Washington, DC. Schachter, S. (1959). The Psychology Of Affiliation: Experimental Studies Of The Sources Of Gregariousness. Stanford University, California. Schegloff, E. (1968). Notes on a conversational practice: Formulating place. In: Studies In Social Interaction (D. Sundow, ed.), pp. 75119. Free Press, New York. Scherer, K. B. and H. Giles (1979). Social Markers In Speech. Cambridge University Press, Cambridge. Schiffrin, D. (1994). Approaches To Discourse. Blackwell Publishers, Oxford. Schober, M. F. (1993). Understanding by addressees and overhearers. In: Arenas Of Language Use (H. H. Clark). University of Chicago Press, Chicago Searle, J. R. (1969). Speech Acts. Cambridge University Press, Cambridge. Shalom, C. (1997). That great supermarket of desire: attributes of the desired other in personal advertisements. In: Language And Desire (K. Harvey and C. Shalom, eds.), pp. 185203. Routledge, London. Shuy, R. W. (1993). Language Crimes. Blackwell, Oxford. Sigman, S. J. (1983). Some constraints placed on conversational topics. In: Conversational Coherence (R. T. Craig and K. Tracy, eds.) Sage Publications, California. Sinclair, J. and M. Coulthard, M. (1975). Towards An Analysis Of Discourse: The English Used By Teachers And Pupils. Oxford University Press, London. Slade, D. (1995). Gossip: two complementary perspectives on the analysis of casual conversation in English. ARAL (Special Edition on Spoken Discourse Studies), pp. 4783. Smith, B. R. and E. Leinonen (1992). Clinical Pragmatics. Chapman and Hall, London. Sperber, D. and D. Wilson (1986). Relevance. Basil Blackwell, Oxford. Stubbs, M. (1983). Discourse Analysis. Basil Blackwell, Oxford. Swales, J. (1990). Genre Analysis. Cambridge University Press, Cambridge. Tannen, D. (1982). Oral and literate strategies in spoken and written discourse. Language, 58, 1 20 Tannen, D. (1984). Conversational Style: Analyzing Talk Among Friends. Ablex Publishing Corporation, New Jersey. Tannen, D. (1989). Talking Voices. Cambridge University Press, Cambridge.
References 173 Tannen, D. (1991). You Just Don't Understand. Women And Men In Communication. Virago Press Ltd., London. Tannen, D. (1994). Gender And Discourse. Oxford University Press, Oxford. Thomas, J. (1995). Meaning In Interaction: An Introduction To Pragmatics. Longman, London. Trask, R. L. (1995) Language: The Basics. London: Routledge. Trask, R. L. (1999). Key Concepts In Language And Linguistics. Routledge, London. Tyler, S. A. (1978). The Said And The Unsaid. Academic Press, New York. Ullman, S. (1962). Semantics. Blackwell, Oxford. Venneman, T. (1975). Topic, sentence accent, and ellipsis: a proposal for their formal treatment. In: Formal Semantics Of Natural Language (Keenan, E. L., ed.), Cambridge University Press, Cambridge. Wardhaugh, R. (1985). How Conversation Works. Basil Blackwell, Oxford. Warren, M. (1993). Inexplicitness: a feature of naturalness in conversation. In: Text And Technology: In Honour Of John Sinclair (M. Baker, G. Francis and E. TogniniBonelli, eds.), pp. 3753. John Benjamins, Amsterdam. Widdowson, H. G. (1993). Peter Strevens memorial lecture. Unpublished paper in IATEFL 27th. International Annual Conference, University College of Swansea. WilkesGibbs, D. (1993). Referring as a collaborative process. In: Arenas Of Language Use. (H. H. Clark, ed.), University of Chicago Press, Chicago. Wodak, R. (1996). Disorders Of Discourse. Longman, London. Wong, M. S. (1999), You Said It! Cambridge University Press, Cambridge. Yule, G. (1985). A Study Of Language. Cambridge University Press, Cambridge. Yule, G. (1996). Pragmatics. Oxford University Press.
This page intentionally left blank
175
AUTHOR INDEX Abbs, B 152 Altman, I. 10 Anderson, K. 152 Argyle, M. 18,2223, 121 Atkinson, P. 96 Austin, J. L. 7, 124 Beebe, L. M. 154155 Bell, D. V. J. 34, 40, 123124 Benson, A. 10, 36, 37 Berger, C. R. 10, 37 Berger, P. L. 121 Bernstein, B. 9, 52, 151 Berry, M. 7, 120, 121, 124 Biber, D. 56 Blakemore, D. 53, 54, 56, 118 Boden, D. 4, 34 Bradac, J. J. 10, 37 Brazil, D. 6 Breuer, H. 155 Brown, E. 11 Brown, G. 7, 27, 28, 51, 120 Brown, J. 154 Brown, P. 6, 9, 3536, 49, 56, 76, 78, 96, 112, 118, 120, 124, 134 Bryan, K. L. 155 Carlson, T. B. 33, 124, 125 Carter, R. 56, 77, 152 Chafe, W. 35, 56 Channell, J. 78, 96, 99, 152, 153 Clark, E. V. 53 Clark, H. 55 Clark, H. H. 33, 34, 35, 37, 42, 53,55, 59, 7778, 83 Clark, L. F. 37 Cook, G. 7, 20, 35, 110, 124, 154 Cook, U. 152 CookGumperz, J. 52, 77 Coulthard, M. 6, 8, 35, 5556, 107, 156, 157 Crystal, D. 78, 152153 Cutting, J. 150, 153 Davy, D. 78, 152153 de Beaugrande, R. 4, 54 Dews, S. 103 Dines, E. 99
176
Drew, P. 4, 6, 7, 34, 51, 52, 125, 143144 Duck, S. 18,30,31, 121 Duranti, A. 3334, 51, 52, 125 Edmondson, W. 10 Eggins, S. 6, 7, 8, 22, 118, 121, 125, 152 Eliot, T. S. 149 Espeseth, M. 152 Fairclough, N. 119, 151 Fawcett, R. 7, 124 Fillmore, C. J. 24, 118 Firth, A. 7, 15, 34, 40, 123, 124 Firth, J. R. 9, 52 Flowerdew, T. 152 Fodor, J. D. 79 Ford, C. E. 3, 8, 53 Fox,B. A. 112 Garfinkel, H. 3, 8 Giles, H. 10,22,23,36, 134 Goffman, E. 3, 6, 7, 10, 120, 125, 128 Goldberg, J. A. 121, 122 GoldmanEisler, F. 36 Goodwin, C. 89, 3334, 51, 52, 97, 117, 125 Goodwin, M. H. 125 Graesser, A. C. 37 Grice, H. P. 3, 5, 24, 28, 56, 76, 110 Grimshaw, A. D. 24, 52, 77 Gumperz, J. 34, 56, 9, 21, 52, 76, 77, 120 Halliday, M. A. K., 5, 7,52, 53, 54, 56, 57, 59, 78, 82, 83, 96, 121, 124 Hasan, R. 5, 7, 52, 53, 54, 56, 57, 59, 78, 82, 83, 96, 121, 124 Hassibi, M. 155 Hawkins, J. A. 53, 58 Heasley, B. 5 Henderson, M. 18, 121 Heritage, J. 4, 6, 7, 3435, 51, 52, 125, 143144 Hinde, R. A. 18, 19 Hoey, M. P. 7, 79, 124 Huddleston, R. 54, 57,58, 80 Hurford, J. R. 5 Hutcheson, S. 121 Hymes, D. 4 Jefferson, G. 3, 8, 10, 99 Johnson, S. 118 Kanfer, F. H. 36 Kaplan, J. 103
177
Kasl, S. V. 23 Kelly, H. H. 10 Kennedy, G. 78 Kreckel, M. 9, 3637, 78, 112 Labov, W. 3, 6, 8, 76, 110 Lakoff, G. 78, 93 Laver,J. 121 Lawton, D. 36 LeckieTarry, H. 56, 79 Leech, G. 3, 6, 28, 102103, 118, 120, 121 Leinonen, E. 155 Levinson, S. 3, 5, 6, 9, 19, 21, 35-36, 49, 53, 54, 56, 76, 78, 96, 112, 118, 120, 124, 134 Levy, D. 10 Luckman, T. 121 Lynch, T. 152 Lyons, J. 51,53, 54, 55, 89, 120 Mahl, G. F. 23 Malinowski, B. 33, 121 Martinet, A. 76 Matarazzo, J. D. A. 36 McCarthy, M. 6, 27, 56, 77, 120, 121, 152 McTear, M. 155 Miller, J. 11 Milroy, L. 11 Moscovici, S. 36 Murphy, G. 55 Nash,W. 118 Ochs, E. 4, 8, 34, 53, 120, 144 Orestrom, B. 19 Osgood, C. E. 23 Palmer, F. R. 5 Planalp, S. 10, 36, 37 Plum, G. A. 8 Powesland, P. F. 36 Prince, E. F. 35 Quirk, R. 22 Ratner, S. C. 36 Rice, R. E. 36 Sacks, H. 3, 8, 10, 35, 98 Sadock, J. 3 Schachter, S. 121
178
Schegloff, E. 3, 8, 10 Scherer, K. B. 23 Schiffrin, D. 3, 35, 55, 5657 Schober, M. F. 42 Searle, J. R7, 124, 125 Shalom, C. 149 Shuy,R.W. 156157 Sigman, S. J. 35 Sinclair, J. 8 Slade, D. 6, 7, 8, 22, 118, 121, 125, 152 Smith, B. R. 155 Sperber, D. 27, 34, 121 St. Clair, R. 10, 22 Stubbs, M. 27, 121 Swales,!. 1, 19,78, 149 Tannen, D. 3, 4, 6, 9, 10,27,56,76, 110111, 121, 134, 138, 140, 151 Taylor, D. A. 10, 128 Thomas,!. 112 Trask,R. L. 21, 122 Tyler, S. A. 9 Ullman, S. 78 Underwood, M. 152 Venneman, T. 3, 35, 5354 Walker, E. G. 23 Wardhaugh, R. 27, 35, 36, 110, 121, 122 Warren, M. 56, 60 Widdowson, H. G. 96 WilkesGibbs, D. 59, 98, 112 Wilson, D. 27, 34, 121 Winner, E. 103 Wodak,R. 151, 155 Wong, M.S. 152 Yule, G. 7, 27, 28, 51, 120, 121
179
SUBJECT INDEX academic discourse communities, 12, 15, 157 see also MSc InGroup Language Study accommodation theory, 10 anaphoric and nonanaphoric reference, 5355, 56, 57, 58, 5960, 6263, 107,145 attitude, 125, 126128, 129130, 131, 132134, 135140, 144, 145, 146 Birmingham School, 8 Cambridge and Nottingham Corpus of Discourse in English (CANCODE), part of the Cambridge International corpus, 150 cataphora, 53, 54, 56, 57 clinical pragmatics, 155156 cohesion, 5, 79 context and knowledge, 3336, 6871, 76 approach, 3, 4 definition, 3 psychological constructs, 3 contextualisation cues, 56, 8, 14, 15, 77, 135, 144, 145, 150, 151 see also implicit contextualisation cues conversational analysis, 3, 4, 6, 8, 15, 3436 conversation organisation analysis, 89 definiteness, 5, 5556 nonanaphoric, 6263, 65, 67, 6871, 76, 97, 106, 137, 144, 145, 146, 157 deixis, 5, 53, 54 discourse analysis approach, 39, 156157 ellipsis, 22, 23, 31,6061, 134 clausal, 97102, 109110, 116, 117, 136, 142, 144, 148 endophora, 5354 English native speakers (NS), 20, 21, 31, 147, 148 nonnative speakers (NNS), 2021, 147 exophora, 5354, 56 explicitness, 53, 5559, 60, 6971, 72, 75, 78, 79 degrees, 6367 forensic linguistics, 156157 functional analysis, 67, 15 genre, 8 grammar
180
and context, 52 effect of familiarity, 2, 143144 in interaction, 89 reference, 5, 14, 15, 51 reference categories, 5360 hitches, 107, 110115, 116117, 144, 145, 157 humour, 5, 9, 10, 14, 15, 18,2427,31, 102106, 115118, 134, 136137, 142, 143, 144, 145, 146, 147, 148, 156 implicit contextualisation cues, 52, 75, 76, 91, 101, 106, 107, 108, 113, 115, 118, 137, 142, 144, 149, 157 see also contextualisation cues implicit language model, 11 expansion, 146147 generalisability, 147150 social dimensions, 150151 implicitness function, 119140, 141146, 150 grammar, 5175 lexis, 7795 utterance, 97118 inference, 5 informality, 2223, 31, 143 ingroup language characteristics, 67, 811, 141146 development, 2, 4, 10 function, 2, 4, 5, 6, 7, 76, 119140, 143146 studies, 911,7778, 149157 turntaking, 8 see also academic discourse communities; MSc InGroup Language Study; shared knowledge ingroup markers, 1718, 96 see also humour; informality; topic shifts interactional sociolinguistics, 34, 56, 15 intertextuality, 4, 15 knowledge and context, 3336 categories, 3643 see also shared knowledge language and context, 51, 125 function, 121122, 123, 124 interactional, 120, 121, 122, 123, 134, 137, 138, 144, 145146 transactional, 120121, 122123, 134, 137, 138 linguistic analysis, 56, 15, 156 lexical analysis, 7778 lexis, 7796, 145
181
reference categories, 7991, 131 MSc InGroup Language Study approach, 15, 2021, 37 assumptions, 11 characteristics of recordees, 2, 18, 20, 3031, 142143, 148 course description, 1920 data analysis, 11, 14,31,43,48,6075,8396,99118, 120140, 141146 impenetrability test, 163166 Karea model, 3747, 51, 147 method of data collection, 1112, 13 reliability, 11,62, 128 triangulation questionnaire, 17, 18, 30, 31, 4748, 75, 95, 115, 137, 159162 see also implicit language model negotiation discourse, 15 nouns, 7991, 9496, 97, 98, 101, 114, 117, 142, 143, 144, 145, 150, 157 obscenities, 22, 23, 31 phatic communion, 121122 politeness principles, 67, 15, 120, 124 pragmatics, 3, 5, 6, 15, 155156 presupposition, 5, 35, 52, 5354, 60 relationships psychology, 1819 shared knowledge, 1 , 2 , 3 , 4 , 6 , 15,3349,55,6768, 113, 117118, 121, 141, 144, 145, 157 slang, 22, 23, 31 social groups language, 2, 34, 52, 150151 see also academic discourse communities; MSc InGroup Language Study specificity, 7174, 76 speech acts, 135137, 155, 157 categories, 124128 changes, 128134, 144, 146 theory, 7, 124125 systemic linguistics, 7 Teaching English as a Foreign Language, 152155 topic definition, 2728 effect on conversation, 36 relevance, 28 topic shifts, 8, 10, 11, 13, 14, 15,2730,31, 107110, 118, 143, 144, 145 utterance, 97118, 144, 145
182
vagueness, 7879, 96, 99, 109, 144145, 156, 157 variation analysis, 3,6, 15 verbs, 8283, 9194, 9596, 97, 142, 143, 144, 145 vocabulary, see lexis