The properties of indexical expressions are ordered properties, and…that they are ordered is an ongoing, practical accomplishment of every actual occasion of commonplace speech and conduct" Garfinkel and Sacks (1970:341). suicides, others not. Thus, the procedure used in such classification, "the fact of the fact," was of capital importance. Durkheim's findings were built on "facts" of uncertain provenance (Douglas 1971). Cicourel (1968) studied how criminal statistics were generated through the interpretive procedures of officials in the criminal justice system. Wieder (1974) describes how behavior in a halfway house was categorized in accordance with a prisoners' code and how, through that categorization, the behaviors were made visible as recognizable forms of social action.Categories are not only administered after the fact. Ethnomethodologists are also sensitive to the work that goes into making an entity categorizable in a particular way.The foundational work in this connection is Garfinkel's (1967) study of Agnes, a transsexual. Garfinkel described in detail Agnes' methods for presenting herself as a recognizable female. There is some tension between these two approaches to categories.Is social order produced by categorizing and otherwise accounting for inchoate behavior, or is behavior constructed in fine detail so as to be categorizable and otherwise accountable? In either case, categories are a product of members' methods for producing and recognizing phenomena. We are directed to look at the "work" that goes into this production and recognition.Another essential grounding for the work presented in this book is sequential conversation analysis (as opposed to categorical analysis, Sacks' other major interest).Conversation analysis (CA) is the study of the organization of talk-in-interaction. We run here into a terminological problem: does CA include Sacksian category analysis or only the study of the sequential organization of talk-in-interaction? When I want to be specific, I will refer to sequential analysis (SA) and category analysis (CtA). I use CA more generally to include both, not only because SA and CtA are both associated with Sacks but also because they share features in common. As with ethnomethodology, the CA focus is on members' methods in the production of social order. CA is characterized by a set of methodological techniques and perspectives: primarily, recording of ''natural'' data, participant orientation, attention to detail (''order at all points,'' as Sacks from another possible formulation, and also from not doing it at all (1992, Vol. 1: 516).Sacks is making two major points about formulations. One, of course, is that they are inevitably a way of doing, not merely saying, something, a notion familiar to us from speech act theory, although Sacks' notion of ''doing'' is broader. But he makes another crucial point with ''non-alternatively'' and ''differentiated from another possible formulation,'' namely, that a formulation is a choice from among a number of alternative ways of identi...