Discovering order

The polarisation of the experimental and observational traditions in linguistics has tended to obscure the common origins of both in intuitions. In this article I explore one form of observational work - conversation analysis - by examining its perceived limitations and the reasons for its insistence on recorded interactions. Its capacity to capture the temporal production and interpretation of utterances is what makes for its distinctive contribution to linguistics, allowing us to discover order in the organisation of talk that escapes introspection. The analysis of data extracts and the examination of case studies impels us to recognise what the investigation of single utterances and utterance pairs cannot: the importance of sequential placement to the understanding of utterances and the centrality of action in language use.