Critical trends in discourse analysis emphasise the connection between discourse and social structure. They locate the critical dimension of analysis in the interplay between discourse and society, and suggest ways in which features of...
moreCritical trends in discourse analysis emphasise the connection between discourse and social structure. They locate the critical dimension of analysis in the interplay between discourse and society, and suggest ways in which features of social structure need to be treated as con text in discourse analysis. For instance, in analysing doctor-patient interaction, the facts that one participant is a doctor and another is a patient, and that this interaction consequently develops in an institutional environment, are crucial elements in understanding the power balance in that interaction. There will be a particular power dynamic because one is a doctor and another is a patient, and because this turns the particular interaction into an instance of an institution alised genre. Critical analysis is thus always and necessarily the an alysis of situated, contextualised, language, and context itself becomes a crucial methodological and theoretical issue in the development of a critical study of language. There is a vast and significant literature on context (see, for exam ple, Auer and Di Luzio, 1992; Duranti and Goodwin 1992; Auer 1995), and the most general way of summarising it is to say that it addresses the way in which linguistic forns -'text' -become part of, get inte grated in, or become constitutive of larger activities in the social World (see also Scollon 2001). o some extent, this is self-evident: lan guage is always produced by someone to someone else, at a particula time and place, with a purpose and so forth. But, given the history of linguistics as the study of an object defined as necessarily non Contextual and autonomous, attention to the context-sensitive dimen S1ons of language was something that required substantial argument. We are beyond such arguments now, fortunately, and we can turn to a whole complex of approaches to text-context relations. We can now accept without having to go into detailed discusion that the way in sounds selfevident, but it is not, it has considerable implications. In order to clarify this, I need to start from John Gumperz's (1982Gumperz's ( , 1992) ) seminal concept of 'contextualisation'. Contextualisation 'comprises all activities by participants which make relevant, maintain, revise, cancel any aspect of context which, in turn, is responsible for the interpretation of an utterance in its particular locus of occurrence' (Auer 1992: 4). Gumperz developed the notion of contextualisation to account for the ways in which people 'make sense' in interactions and, taking on board both broad ethnographic concerns as well as narrower conversation-analytic ones, he observed that people pick up quite a few "unsaid' meanings in such interaction. These are the indexical meanings mentioned in chapter 1: the connections between language form and social and cultural patterns. People detect these indexical meanings because speakers provide verbal and nonverbal, behavioural 'cues' that suggest a fit between utterances and contextual spaces in which they become meaningful: argue that conversational interpretation is cued by empirically detectable signs, contextualization cues, and that the recognition of what these signs are, how they relate to grammatical signs, how they draw on socio-cultural knowledge and how they affect understanding, is essential for creating and sustaining conversational involvement and therefore to communication as such. (Gumperz 1992: 42) The pivot of this process is the utterance itself: 'it is the linguistic form itself which serves to signal the shift in the interaction ' (1992: 43). And the target of contextualisation consists of complexes of presupposable knowledge in which utterances are made coherent (contextualised): it ultimately rests on all understanding is framed understanding, contingent inferences made with respect to presuppositions concerning the nature of the situation, what is to be acconmplished and how it is to be accomplished. (1992: 43-44) Such forms of framing are linguistic and cognitive, to be sure, but also eminently social and cultural. They have a perduring, conventional, normative character: 'we can speak of systems of contextualization con ventions in terms of which individual cues are related.' (1992: 51, italics added). Relatively little work on interaction has focused on this systemic dimension of contextualisation, most analysts having concentrated their efforts on analyses of the contingent conversational deployment of contextualisation practices (but see Gumperz 2003; Eerdmans 2003: I should also note that systemic aspects of contextualisation are cen tral in poetics' as conceived by Jakobson 1960