Styles of Discourse

Nikolas Coupland, Nikolas Coupland

• Publisher: Routledge
• Year: 2016
• ISBN:
  9781315402680
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Linguistics
• Index: Languages & Linguistics

First published in 1988, this book focuses on diversity and discourse, and collects contemporaneous research across a wide range of topics including: description, polemic, narrative analysis, DJ talk, philosophical history, conversation, children's books and nuclear deterrence. The essays demonstrate analyses of discourse in the service of stylistic inquiry, exploring relationships of text and context. This reflects the overall argument that discourse analyses aiming to represent diversity of social context will necessarily approach the task selectively, since all dimensions are of potential relevance to any and every communicative manifestation. Some of contextual dimensions that are addressed include: interpersonal, socio-structural, modal, ideological, and pragmatic.

1
Introduction: Towards a Stylistics of Discourse

Nikolas Coupland

The need to account for discourse processes is pretty much universally recognised across the linguistic sciences, and is rapidly gaining recognition well beyond. On the other hand, there are widely differing views about what should be expected to derive from this endeavour, and therefore about what is involved both theoretically and methodologically. There are clearly different contemporary discourses of discourse analysis. The range of current and past conceptualisations and practices of discourse analysis cannot be comprehensively reviewed here, and this has in any case been recently done (by van Dijk 1985, particularly volumes 1 and 2). But a brief sketch of major current emphases will, I hope, establish a context for the diverse analyses assembled in this book, and allow me to present its collected chapters as showing a broadly agreed, distinctive orientation and enacting its own set of priorities for the analysis of discourse. These particular priorities, as we shall see, require a modification of several of the assumptions underlying more traditional work on discourse, toward~ a greater recognition of diversity, a more direct focus on contextual explanation at various levels, and more conscious integration with broader social scientific (beyond more narrowly linguistic) concerns. I shall present this realignment as the reimposition, in at least certain important respects, of the goals of stylistics (in explanation of the book's title) and work towards a definition of discourse analysis as the contextual interpretation of psycho-sociolinguistic diversity.

In British linguistics circles, discourse analysis tends to be associated with work inspired by Sinclair and Coulthard's analysis of classroom interaction (1975; cf. also Coulthard 1977; Coulthard and Montgomery 1982). The underlying rationale for the essentially structural account this paradigm offers is the need to capture aspects of the organisation of spoken interaction above the sentence level. A loosening of the confines of sentence or clause description is a pervasive theme of all approaches to discourse, both in the sense of operating with 'larger' or 'higherorder' units of expression, and in shifting the focus from formal, syntactic units to functional units, definable in relation to what is achieved in the contexts of their use, and in relation to one another. Analysis takes the form of exhaustively accounting for and labelling structured functional units in texts. The legacy of this model of discourse analysis is considerable. It feeds, for example, the discussion by Montgomery (in Chapter 4 of this volume) of how monologic, less obviously interactional discourse - such as D-J radio-talk - might be accounted for structurally. Again, the structural taxonomy of processes of self-disclosure (by J. Coupland et al., Chapter 9) draws on what has come to be established terminology to identify hierarchical levels of functional organisation - 'moves' and 'acts'. Sinclair and Coulthard's schema has, with minimal modification, been applied to social settings beyond the classroom, for example in Coulthard and Ashby's analysis (mimeo) of doctor-patient inter action, and my own work on travel-agency talk (Coupland 1983). We might predict in fact that the principal functional 'levels' at which Sinclair and Coulthard have shown classroom talk to be organised - transactions, exchanges and moves - will be rele vant in most 'transactional' (as opposed to 'personal', in Gumperz's senses of these terms) encounters and that similar patterings - e.g. of exchange structure - will tend to recur.

The important observation for our current interests, though, is that analyses of this sort are primarily motivated by concerns about descriptive linguistic adequacy rather than aspiration to social explanation. Though contextual/sociological insights will derive from analysis of this sort - leading to rather predictable conclusions, for example, about the distribution of power in the classroom through variable constraints on amounts and types of communicative contribution - the analysis is designed to display linguistic regularity. The classroom setting, as Sinclair and Coulthard acknowledge, is well suited to their analytic procedures (not vice versa) because of the rigid scripting that at least some school settings impose upon spoken interchange; their classroom discourse has 'structure and direction' (1975, p. 5). Though the units of analysis are not formal; the analysis is not essentially different from those of clause-level constituent-analysis, establishing criteria for segmentation and classification. Discourse itself is characterised as a further rank above the clausal rank at which structure is discoverable.

Another major strand of British linguistic studies - systemic linguistics - integrates 'discourse' focally into its theory and description. A moderately uncontroversial representation of the Hallidayan framework (though the precise distribution of categories and delimitation of descriptive planes and levels is a matter of considerable ongoing debate - cf. Halliday 1978, 1984; Halliday and Martin 1981; Greaves and James 1984) will view language as a tri-stratal system, comprising phonology, lexico-grammar and thirdly discourse. Discourse here broadly equates to core areas of the semantic/pragmatic components of alternative descriptive formulations, and is the stratum within which such sub-systems as reference, conjunction, cohesion and conversational structure may be specified. For example in Martin's (MS) account, discourse handles dependency relations between parts of a text. I shall return to Hallidayan accounts of the relationships between language and context and their discussions of register, genre and ideology a little later, where we shall find a discussion far more centrally relevant to the concerns of this book. For now the point is that systemic linguistics again adopts an orientation to discourse which seeks primarily to enrich the descriptive (and in this case also the predictive) adequacy of linguistic theorising. Halliday's theories are sociolinguistic theories to the extent that they work dimensions and categories of the social functioning of language into a systemic account of how language constructs meaning, the drift of argumentation therefore being from sociological to linguistic.

This is an observation that has been made in relation to quite different strands of sociolinguistic research (for example, about the seminal work of Labov 1972 and Trudgill 1974 in the area of dialect variation) - that established paradigms embody essentially linguistic priorities. But as in these cases ( cf. Coupland 1987) so in relation to discourse, we can equally well adopt the converse perspective and embark on the analysis of language-in-context with an eye to the understanding of social processes and context itself. A stronger position, and one that many of the contributions to this volume will endorse, is that the language-to-social organisation schema for discourse analysis is theoretically required as a consequence of prior conclusions reached about the nature and constitution of social organisation itself. Arguing this position is in fact the epistemological corner-stone of sociological approaches to discourse analysis.

It is widely recognised that a major school of sociological theorising— ethnomethodology (cf. Garfinkel 1972, Psathas 1979) — has been influential in establishing a climate wherein discourse analysis (as a broad heading) is seen as an appropriate and even a necessary orientation to analysis and scholarly debate across many disciplines. Ethnomethodology as sociology distances itself from deterministic sociological analyses which it claims unjustifiably endorse the 'real' existence of demographic and institutional categories and 'facts'. Therefore, to take socioeconomic stratification or indeed any social framing or coding as the point of departure for analysis is to misconstrue, it is argued, the nature of social organisation, which is an imputed organisation. However natural such categories are, and more particularly because of their conventionalised naturalness, we need to research whether and by what means they find their existence as natural categories for actors and analysts alike. These researches then become researchable in an inwardly spiralling analysis of methods of categorisation. A cognitive perspective at least begins to live up to our best understanding of the nature of social organisation, as established in and constituted by on-the-ground behaviours — and in particular through social interaction. By this account, it seems unlikely that social 'facts' will be available to linguistic theory-building, but it is to be hoped that delicate analyses of language in use will reveal acceptable truths — perhaps typically ungeneralisable, fleeting and partial — about aspects of social life.

The basic ethnomethodological perspective holds both a positive and a negative implication for discourse analysis: what might be obtainable, and what is not to be trusted. The sociological literature centring on conversation analysis (CA) has begun to show what might be obtained, but is arguably over-constrained by its definition of what is trustworthy. To the extent that CA is committed to an ethnomethodological re-working ab initio of social analysis, it needs to be relatively self-motivating and self-contained. The social structures it can recognise are essentially those demonstrated and created during interaction — structures of social action (cf. Atkinson and Heritage 1984)¹ — most obviously then structures 0/talk. CA derives its data meticulously though not unproblematically (cf. below) through transcription from recorded, naturally occurring interaction, which is clearly 'situated' and contextually relatable. On the other hand, concern about the bogus fixity of contextual categories seems to have led to minimal attention being afforded non-linguistic context in CA, subordinated to the primary status of the data as talk. The structuring and patterning of talk gives analyses their focus, under rubrics like closings, insertion-sequences, step-wise topictransition, oh-prefaced responses, and so on. The focus on autonomous talk tends to drown out what individual protagonists bring to it - variables in their histories, relationships, motives and predispositions - and the effects of talk - satisfaction, dissatisfaction, attitude-change, relational change. Talk is presented as occurring in a rarified social context and sometimes in a world where all but talk is suspended (cf. the discussion in the Appendix to Chapter 9 in this volume).

CA has therefore arrived at a position where its primary input, and a vital one, to the analysis of language in context is in displaying the local mechanics of conversation. There is an irony: CA sociologists have shouldered the burden of pattern-building, which may be put to work doing social explanation (as in several contributions to this volume) in the works of analysts who are less obviously sociologists - linguists, sociolinguists, critics, ethnographers, philosophers, political theorists and social psychologists.

This diversity of disciplinary interest in discourse analysis (although in some of this book's chapters, the disciplines represented are encouragingly difficult to pin-point) is not entirely understandable as a spontaneous epistemological shift away from deterministic, overly empiricist social science, as I have sketched it in relation to ethnomethodology. At least, it is not entirely attributable to the negative implication - the mistrust of other approaches. A more humble explanation is that discourse analysis is, quite simply and within its limitations, a revealing way of doing social science. On the assumption that many or most social relationships, institutions, events and activities are mediated by communicative and specifically linguistic behaviours, it is reasonable to expect that detailed inspection of language-use, its relationship to and constitution of meanings in context, will give access to the dynamic workings of social processes. Discourse analysis does not therefore have to be sustained by a committed interpretivist, relativist epistemology, and be claimed to be the only theoretically enfranchised line of social scientific research. It can appropriately be a complementary practice to other modes of inquiry, over which it will have specific but not total advantages. A consensual view across this volume is that of discourse not as a delimitable technique or method or level of analysis, but as an orientation to social explanation - as an openness to the interpretation of situated communication events where some nonlinguistic dynamic is a candidate for analysis. While this is a modest rationale for discourse analysis, it is a significantly different one from those offered, as we have seen, either by those linguists who see discourse as a further dimension to a linguistic theory or by those sociologists who are reticent to recognise social structures beyond those of talk.

A more ambitious formulation, however, would credit discourse analysis with the same explanatory power but against much higher odds, arguing that it is a nobler enterprise than an alternative focus on old questions. For some (cf. in particular Williams (Chapter 10), Norris and Whitehouse (Chapter 11) and Hunt (Chapter 7) in this collection), it is through the identification of broad cçultural/ sub-cultural discursive formations, and only through this, that the operation of societal power-codes can be read and conceivably broken. This tradition of discourse analysis (epitomised in the writings of Foucault, e.g. 1971, 1980 and overviewed by Kress 1985) argues a systematic relation between discourse, power and knowledge wherein self-justifying modes of discourse set the boundaries of and sustain ideologies - principally those of the dominant social groups and institutions. In consequence, critical analyses present the sole route to contextualising such discourses and possibly recovering rights over what may be said, what may be thought, what may be known. Thus, legal and nationalist discourse of 'ethnicity' and 'language' are argued (Williams) to dictate the progression of a tribunal hearing of a case alleging discrimination. Establishment discourses restrict definitions of 'literature' and 'children's books' (Hunt) in a way that preconditions the type of reading material offered to children, and critical response to it. In super-power heads-of-state pronouncements, the US employs a rhetorical mode premissed on quite different discursive principles (Norris and Whitehouse) from that of the USSR and sustains a position on nuclear deterrence which, these authors argue, cannot be rationally grounded.

In the chapters as a whole there is a pervasive theme of discourse analysis as a forensic detective activity, as a means of accessing covert but essential characteristics - of either individual communicative events (e.g. a poem), conversational modes (e.g. narrative/ anecdote, self-disclosure), generic categories (e.g. various contexts of prayer, news-reading, D-J talk) or social institutions (e.g. social work). Working outwards from selective reading of texts - spoken or written - analyses are constructed to reveal processes salient to the constitution of these dimensions of social context. Certainly there are different levels of ambition here, and claims of varying strength are made by different contributors. The stronger the claim and the greater the explanatory leap from text to social implications, the more space remains (for critical readers) to give modified or alternative accounts. The analyses in fact mix textual description with interpretation in very different measures, and some contributors are at pains to mark the points at which their analyses cross from linguistic characterisation to more obvi...