Sociology and ideas about language
Getting things done: language and social interaction
Conversation analysis
The ethnomethodological approach developed to accommodate the research known as conversation analysis, whose aim is to explore how language is used in social action. This work is probably the area in sociology which has paid the closest attention to linguistic detail 'actual occurrences of conduct in interaction'- and how these provide indications of social structures, in the sense of 'what sort of social scene is being constructed' (Schegloff 1991: 48, 46). Core principles in conversation analysis are close attention to authentic empirical data and a reluctance to admit concepts such as social categories or social structures as relevant to the interaction unless this is demonstrably warranted in the utterances of the participants themselves. Conversations themselves display 'patterns ... with robust and consistent properties' (Wooffitt 2001: 58), and it is these 'indigenous features of talk-in-interaction' (Schegloff 1991: 67) which should be appealed to first, while social structures should be considered extrinsic, unless there is empirical evidence that they are relevant enough to participants to affect their talk-in-interaction.
Conversation analysis attracts similar criticisms to those levelled against ethnomethodology. Those most relevant to our concerns here are, firstly, that conversational participants are deemed to have knowledge sufficient to generate the analysis, and, secondly, the denial of the influence of social structures on social life. What is needed, argue the critics, is an approach which considers the relations between irreducible structures and social interaction in a 'balanced manner' (Mouzelis 1995). This would emphasize a conception of structure as providing the deterministic conditions of action, whilst recognizing that 'interaction can be understood to be the product of determinate structures which are irreducible, but which none the less do not vitiate the importance of the interpretive capacities and intentions of social actors' (Layder 1981: 87).
There is a further important consequence to the 'unbalanced' view of structures in ethnomethodology and conversation analysis, namely their empiricism. Several writers (Bauman 1978; Mouzelis 1995) have made this point, but Layder (1990: 61) puts it most forcefully:
I would argue that the idea that social structures do not exist independently of the activities they govern and the agents' conception of what they are doing is only true within the terms of an exclusively empiricist epistemology, and thus an individualist/ voluntarist conception of social structure.
Moreover, the essentially empirical task of understanding human interaction, as Bauman (1978: 189) puts it, 'by throwing light on the procedures through which meanings are produced in practice' engenders a radical moral neutrality. By forcing attention on to the practices of meaning production, rather than the meanings themselves (providing formal knowledge of interpretive procedures and their application as opposed to substantive knowledge of whether people's interpretations are 'true' or 'false'), these approaches avoid the problem of truth. The impoverished view of structures coupled with the empirical concern with how actors accomplish or 'do' social life leaves ethnomethodology and conversation analysis with little to say about any possible correspondence between the meaning actors assign to topics and the intrinsic, extra-discursive features of the topic itself (Bauman 1978).
