Albert and De Ruiter provide an introduction to the Conversation Analytic approach to ‘repair’: the ways in which people detect and deal with troubles in speaking, hearing and understanding in conversation. They explain the basic turn‐taking structures involved, provide examples, explain recent developments in the field and highlight some important points of contact and contrast with work in the Cognitive Sciences.

Over the lifetime of Conversation Analysis (CA), scholars have discovered many systems of action organisation (machineries) describing how conversational turns occur, what actions are expected, and how intersubjectivity in conversation is maintained. However, when John Heritage proposed a new machinery that examines the knowledge orientation of participants in interactions, a debate broke out between conversation analysts in which Michael Lynch and his colleagues in radical ethnomethodology descend upon on epistemics. The controversy begins with Lynch accusing Heritage of cognitivism and the (...) extra-situational nature of epistemics, while research on ethnomethodology and conversation analysis has traditionally focused on situated action. The discussion of epistemics points to an internal tension in CA as to where the boundaries of situations lie and what, therefore, can be the focus of CA. This article reactualises the problem of situationality in CA by analysing the arguments in the debate on epistemics. The authors show that epistemics and the debates surrounding it constitute a serious test for CA, revealing a conceptual problem that has hitherto been obscured - the relation and potential hierarchy of different machineries. Turning to the origins of the concept of situationality in the writings of Goffman and Sacks, the authors demonstrate that for opposing sides, the localisation of phenomena within situations is an analytical decision about what can be seen in empirical data. In contrast, distinguishing between the position of the analyst and the participant in the interaction shifts the analyst's attention to how the machineries become relevant to the interactants, that is, how their omnirelevance is realised. The authors argue that this is a more productive formulation of the question than that of the boundaries of the situation. (shrink)