If the analytical value of some terms derives from their descriptive precision and specificity of meaning, other words — such as discourse — owe their utility to multiple layers of meaning and their ability to stimulate ambiguity. Anthropological discourse about ‘discourse’ expanded markedly in volume beginning in the 1970s. The term entered the discipline from two directions: it is part of the language of both descriptive linguistics and cultural studies. Beyond a common understanding that discourse involves the communication of meaning, the term has divergent uses in these two fields. Moreover, within linguistics and cultural studies, as within anthropology, discursive analysis signifies several different sorts of methodological enterprise.
Within linguistics, discourse once labelled utterances longer than the sentence or clause. As linguists move to incorporate contextual factors into their analyses, the term ‘discourse’ has shifted, slightly, to label the set of utterances that constitutes a speech event. The tape recorder (and, for ethnographers, notably the cassette recorder) facilitated the development of discourse analysis. With the increasing sophistication of digitization technologies, discourse analysis has become increasing detailed, incorporating finegrained transcription of speech events that captures changing rates of conversational speed, notes overlapped talking, measures the length of silences and also remarks prosodic features such as emphasis, intonational flow, loudness and other vocal qualities of utterances.
Linguists study these structural elements of discursive flow, focusing, for example, on how speakers introduce and control topics, on interruption, conversational ‘housekeeping’ devices that maintain discursive interaction, on markers that define and separate units within discourse, and so forth. Anthropologists, in general, are more concerned with what discourse structuring might reveal about culture at large: ‘In every moment of talk, people are experiencing and producing their cultures, their roles, their personalities’ (Moerman 1988: xi). The sequential organization of discourse, and conversational features such as overlapping patterns, breaks, silences, repairs and the like, can inform an understanding of both individual intention and cultural order. The genealogy of this technique of paying very close attention to discursive form, often also called ‘conversational analysis’, also traces back to the ethnomethodology of the 1960s and 1970s.
A second sort of discourse analysis, associated with cultural studies, takes discourse more globally to refer to particular areas of language use. This approach blurs together three levels of meaning: discourse is the act of talking or writing itself; it is a body of knowledge content; and it is a set of conditions and procedures that regulate how people appropriately may communicate and use that knowledge. Rather than the elemental structures of conversational interaction, this second approach to discourse pursues the connections between orders of communication, knowledge and power.
Michel Foucault in large part pioneered discourse analysis of this sort. In The Archaeology of Knowledge, he set forth a programme for the ‘pure description of discursive events’ that sought to answer the question: ‘How is it that one particular statement appeared rather than another [in a field of discourse]?’ (1972: 270; see also 1981). Foucault’s genealogies of European discourses of madness and sexuality have stimulated many other analyses of the ways in which patterned cultural discourses maintain both particular ways of knowing the world and a network of power relations among those who know. (Bauman and Briggs [1990] summarize a selection of this work.) Anthropologists influenced by Foucault have proposed the term ‘discourse’, with its implicit connotations of power and possible contestation, as an alternative to traditional anthropological notions of culture (Lutz and Abu-Lughod 1990). The spread of the term reflects anthropology’s increasing engagement with oral and written texts as important data for cultural interpretation.
