2004) ‘Tasting the Witches’ Brew: Foucault and Therapeutic Practices’, Sociology 38(3): 515–32

Interaction is at the heart of most therapeutic practices, yet sociological work on the therapeutic has either tended to ignore this dimension or focused on it at the expense of understanding the wider social context within which such practices unfold. Not surprisingly, Foucauldian work on the therapeutic has, for different rea-sons, also tended not to concern itself with interaction.Yet there are those who argue not only that Foucault might be used in alliance with perspectives concerned with language as constitutive of social reality (Miller, 1997), but that such an alliance might be useful in analysing therapeutic practices (Silverman, 1997). In this article, I think through the theoretical and methodological possibilities and limitations of such an exercise by drawing on qualitative data from a specific set of therapeutic interactions. Bringing together an interest in therapeutic interaction and therapeu-tic government with a sociological interest in who is participating in which thera-pies could, it is argued, address the social context of the therapeutic while at the same time exploring the interaction through which therapy is constituted. KEY WORDS analytic bridging / Foucault / interaction and qualitative data / therapeutic practices here is a lively debate among practitioners and academics (Dryden and Feltham, 1992; Silverman, 1997) about what constitutes therapeutic prac-tice. In a Foucauldian context, therapeutic practices have been broadly taken to be those which constitute subjects as ‘psycho-social beings with needs, motives and problems ’ (Bloor and McIntosh, 1990: 169). It is not surprising that Foucault, who described ‘actual practices ’ as being like a ‘witches ’ brew ’ in comparison with ‘programmes of conduct ’ (Foucault, 1981: 5), did not choos