Methodological reflections on discourse analysis

Part of the difficulty in establishing methods for discourse analysis (DA) is that DA is not singular. Distinct forms of analysis are collected under this label, invoking different understandings of discourse, drawing on different disciplines and canons, and specifying different methodologies. DA is most commonly understood as referring to language. Analyses thus often deploy the term discourse to refer to ‘extended samples of spoken dialogue’ (Fairclough, 1992: 3) such as conversations and so prompt analysis of, among other things, turn taking or the structure of conversational openings and closings (e.g., Atkinson and Drew, 1979; Coulthard, 1992)