ABSTRACT
Martin Hollis’s work was characterised by a core conception of social action, based upon a strict individualism underpinned by a notion of the self acting according to reason. In this paper I want to critique his conception of social action, arguing that it fails to take sufficiently into account any notion (either hermeneutic or naturalist) of social structure. I also want to point to some of the hidden traps in his characterisation of the choices facing social theorists (as represented by his famous two-bytwo matrix). Let me be clear, however, at the outset that I want to critique Hollis’s view of social action precisely because it is one of the most logically coherent and well-worked-out examples of the position in the history of the philosophy of social science. Yet I must also note that it is a view that is often implicit and often has to be deduced from some of the things he does not say. Indeed, in many ways it is a view of social action entirely consistent with the title of one of his books, The Cunningof Reason. By this I simply mean that Hollis wrote very skilfully, setting up alternatives and problems in such a way as to come in at the end of the book or article with a resolution to these alternatives and problems that justified his own view of social action. I must also confess that it has taken me many years both to realise this and to develop this critique, so powerful is his model and his associated reasoning. I am also sure that, were he still alive, he would be able to show with his usual mixture of wit and logic just how mistaken and contradictory I was. It is truly tragic that he is not around to continue that conversation.
