In his article Goldthorpe utilizes his reading of the particular books under review to formulate a general critique of ethnomethodology as he understands it. My concern here, in the space available, will be not so much to defend the books concerned as to consider some of the central issues which Goldthorpe raises . Firstly, let us remove a couple of niggling remarks made by Goldthorpe concerning the work of ethnomethodologists. The 'privately circulated typescripts' to which he refers were, in most part, pre-publication drafts, available to anyone on request, of articles which have now been published either in the Douglas book or in David Sudnow's 'Studies in Social Interaction'.2 The practice of circulating such drafts for comment from others in the field, prior to their production as finished articles is, as most academics are aware, a common procedure and this rejoinder is the direct result of such a practice by Goldthorpe himself. The remaining unpublished materials comprise three works by Garfinkel,3 and the transcripts of Harvey Sacks' lectures which are of such a voluminous, lengthy and discursive nature that publication of them would be difficult, given the conventional constraints on, for example, journal publication format. However, Sacks is currently preparing a book which is to cover the same 'ground' as his lectures.4 As to the 'divergent' positions taken by ethnomethodologists on certain issues, Goldthorpe is of course correct. Ethnomethodology is organized round a set of concerns and is by no means monolithic in structure. Thus the Filmer book, which shows signs of having been heavily influenced by Cicourel, would not be taken by some ethnomethodologists as representing their position regarding certain issues. If it is the case, as Goldthorpe suggests, that 'it will not do' to characterize conventional sociology as being represented by Lundberg and Homans, then it may presumably also be the case that it will not do to characterize ethnomethodology by a similarly selective reading. What I want to do now, therefore, is to consider how far these initial grounds for suspecting that Goldthorpe may have misconceived ethnomethodology are borne out by the contents of his arguments. At a number of points in the review Goldthorpe implicitly, and sometimes explicitly, seems to adopt the standpoint of holding what might be characterized as a 'correspondence theory' of reality.5 That is, there is held to be a reality 'out there' which is distinct from the perceived and intersubjectively constructed reality of members. Thus we are presented with the world of 'physical states', ontology number one, and the world of ' intelligibilia ' or 'objective ideas' which, whilst perhaps the product of intersubjective phenomena, is, in some strange way distinct from it. This is ontology number three. Ontology number two is the world of mental states, the inner world which Goldthorpe seems, mistakenly, to believe corresponds to the intersubjective empirical world with which ethnomethodology is concerned.6 Ethnomethodology is not interested in 'mental states' in themselves, for empirically how is one to get at them? What it is concerned with is the communicative acts which are the givens of the intersubjective world as it is lived. One of the consequences of holding such a correspondence theory is that the purpose of any schema, set of categorizations or explanation is somehow to arrive at an approximation as to what is 'really there'. Thus the sociologist is to obtain,