Volume I Issue 1


Tim J. Berard
Kent State University, Ohio, USA

Download this article (255 KB)


Ethnomethodologists have emphasized the pragmatic and contextual nature of description as a variety of social practice, and have suggested the ramifications of this insight for the methodology and philosophy of the social sciences. However, ethnomethodologists have thereby invited difficult questions about the moral and analytic status of their own descriptions. Drawing on Atkinson's study of suicide verdicts and Coulter's writings on schizophrenia, ethnomethodological scholarship is shown to display the possibility and promise of disinterested description, even when the subject matter involves the evaluation of problematic actions and identities. The combination of Wittgensteinian logical grammar and empirical studies of natural language use, suggested by Coulter, is presented as especially relevant and remarkable for purposes of studying social practices including describing, naming, categorizing, classifying, labeling, diagnosing, reaching a verdict, and kindred practices of language use conceived as varieties of practical action.

Keywords: description, ethnomethodology, evaluation, Jeff Coulter, discrimination, labeling, Ludwig Wittgenstein, Peter Winch, schizophrenia, suicide.


It was a hard-won heuristic insight for the human sciences that description can be analyzed as a variety of practical activity, and can be taken as a topic for social inquiry, rather than serving as its unacknowledged analytic resource (see, e.g., Sacks, 1963; c.f. Zimmerman and Pollner, 1970). Ethnomethodological and conversation analytic contributions to the human sciences figure prominently among the various contributions responsible for elevating descriptive practices such that they became a subject in their own right, beyond questions of pedagogy and research methods.i Ethnomethodology and conversation analysis figure especially prominently in exploring the ramifications of this insight for the human sciences. The reception of this heuristic insight in the wider social-scientific community, however, has been very mixed and very problematic.

On the one hand, many social scientists do not yet appreciate the practical and otherwise indexical (contextual, situated, etc.) properties of the commonsense descriptions that they employ as their data, or gloss over in the process of theory building. One significant example of this is the continuing and programmatic tendency, especially in the attenuated research tradition of positivism, to substitute objective expressions for indexical expressions (see, e.g., Garfinkel, 1967a:1-34), as in the employment of operational definitions for research variables, or coding procedures for 'qualitative' data (cf. Atkinson, 1978: 22). In this methodological and analytic sense, the wider social science community has not yet 'got the message' about description as practical activity and topic of inquiry.

On the other hand, many social scientists have endorsed the political, ideological and relativistic implications of this insight for the social sciences, especially as concerns other peoples' versions of social science. This leads to the current state of affairs, in which the descriptive claims of every tradition of social inquiry are susceptible to being dismissed as interested descriptions, and yet such dismissals can in turn be dismissed, on the logic that if all descriptions are interested descriptions, then the interested nature of any particular description hardly needs to be defended. The implicit logic seems to be that all interests are in an important sense equivalent, if not for moral and political purposes, then at least for purposes of entertaining claims about non-partisan scholarship.

These issues seem to me to be central to the curious and problematic status of ethnomethodology and conversation analysis within the human sciences. Ethnomethodology and conversation analysis have suffered professionally from both the reluctance of many social scientists to acknowledge the methodological and analytic significance of description as practical action, and from the facility with which this insight is adopted for ideological purposes. Insofar as ethnomethodological and conversation-analytic claims about description are addressed at all by the wider social-scientific community, they are subject to being dismissed from both directions at once, as unwarranted or exaggerated criticisms of conventional social scientific methods and reasoning, and as ideologically suspect. On the latter point, EM/CA work has been criticized not so much for having a political agenda, as for lacking a political agenda, which to many theorists amounts to a tacit support for the status quo of class, race and gender relations. Ironically, then, it can be the disinterested nature of EM/CA scholarship which leads to the charge of ideological complicity. I suspect that the alleged inability of ethnomethodology and conversation analysis to handle 'macro' phenomena of political relevance is often invoked partly to allude to such complicity, without having to substantiate the thesis (c.f. Berard, 2005a, 2005b).

It is in this professional context that I feel it is professionally appropriate to qualify the initial insight into descriptive practices, by explicitly and programmatically carving out an analytic space in which disinterested description in the service of analytic inquiry is not only an intelligible and worthy principle (cf. Bogen and Lynch, 1990), but also possible to achieve. Without such argument, claims of 'ethnomethodological indifference' might well appear as a boot-strapping type argument, thus facilitating dismissals of ethnomethodological work in the wider community of the human sciences. By 'disinterested,' I do not mean that scholars have no interests outside scholarship, or should have no interests outside scholarship; nor do I mean to deny that scholarship involves its own variety of interests. I use the term 'disinterested description' merely to suggest a variety of description which is not politically or morally driven, but rather driven by scholarly concerns, according to which social inquiry should be governed by principles of empiricism and logic, rather than ideology.

I intend the term 'disinterested' to be consistent with the standard of 'ethnomethodological indifference,' especially insofar as this indifference involves a non-evaluative approach to the descriptive and accounting practices of members (of a culture). Garfinkel and Sacks observe, for example, that "there is no room in the world to definitively propose formulations of activities, identifications, and contexts" (1970: 359). This notion of 'disinterested' is also consistent with Sacks' argument, in his paper on sociological description, that the task of sociology is not to clarify or criticize persons' descriptions of the social world, but to describe them (1963: 7). In remarks on 'unmotivated examination,' Sacks outlined a relevant analytic attitude in which no social phenomenon would be ignored because of the presumed results of its study. Theory is to be driven by analysis of the data, rather than data being selected and findings accomplished according to theoretical problems or agendas. As Sacks argues, and as his methods have arguably now demonstrated, "It is possible that detailed study of small phenomena may give an enormous understanding of the way humans do things and the kinds of objects they use to construct and order their affairs" (Sacks, 1984:24). While I have no objection to either of these terms, as I understand them, 'ethnomethodological indifference' and 'unmotivated examination' have already been used in other contexts to make somewhat different arguments, and have somewhat different and arguably more narrow meanings than I am trying to convey with the term 'disinterested.' This notion of 'disinterested description' is also consistent with, but not identical with, the Wittgensteinian avoidance of criticizing linguistic conventions or prescribing their reform, in favor of describing the logical grammars of concepts, to be addressed below.

I believe it is a fair generalization that ethnomethodology and conversation analysis have been involved in producing disinterested descriptions of their phenomena from the very beginning, and it seems to be the case that the reality of ethnomethodological and conversation-analytic studies has far outpaced the theoretical and programmatic arguments made on their behalf, despite very relevant and edifying ethnomethodological treatments addressing, e.g., the role of morality in different sociological paradigms (Wilson, 1970) and morality as a feature of mundane categorization practices (see esp. Jayyusi, 1984, 1991). Indeed, the claim that description is a practical activity is itself an example of disinterested description, i.e., it wasn't motivated by political or moral concerns, and hasn't even paid off professionally. It has been rewarding only if one evaluates it as an empirical and analytic insight, by which standards it is enormously important and yet enormously under-appreciated. Another way of stating this is to suggest ethnomethodological and conversation-analytic descriptions are in an important sense impractical, and cannot be understood or appreciated if read or heard ironically, as symptoms or indicators of practical interests in, e.g., conserving or critiquing social practices, accounts or explanations, social beliefs, speech, or members' relations with each other, whether individually or in groups.


The professionally distinctive potential of ethnomethodological and conversation-analytic scholarship to produce disinterested descriptions of social phenomena should be apparent to anyone familiar with the relevant literature, but I would argue that nowhere is it more evident than in studies of evaluative categories, where, arguably, disinterested descriptions would be least expected and most difficult to produce. By 'evaluative categories,' I mean categories of social description that are commonly bound up with practices of evaluating persons, actions, institutions, social contexts, etc. Two examples: 'suicide' is an example of an evaluative action category, and 'schizophrenic' is an evaluative identity category. These particular categorizations are not only typically used in an evaluative manner, but can also be quite consequential as a practical matter, both of which characteristics can be illustrated briefly for those who may benefit from a brief reminder about these specific domains of practice.

Claims or verdicts of 'suicide' are evaluative and consequential for a variety of reasons, including the facts that suicide is considered by many an egregious sin, potentially determining the eternal fate of the soul; suicide can in some contexts be used to deny insurance benefits to the family of the deceased; suicide is often thought to reflect negatively on the responsibility of friends, family, or institutional custodians of the deceased, and in the latter cases may raise legal or professional issues of liability. Coroners who cannot adequately defend verdicts of suicide to various interested parties may also find their reputation or their job in jeopardy. Moreover, categorization of a sudden death as a suicide has an additional consequence in that it rules out the applicability of alternative explanations, some of them equally evaluative and consequential, especially the categorization 'homicide.' Categorizations of suicide can therefore raise profoundly important questions of a religious, financial, moral or legal nature.

The category of 'schizophrenia' is equally evaluative and consequential, most clearly in that it is a term of stigmatization, suggesting that a member has a disease of the mind, or a mental illness, which stigmatization can continue even after years without suffering any symptoms of the ostensibly underlying, latent condition. Beyond the stigmatization of the label, psychiatric diagnoses of 'schizophrenia' often serve as grounds for involuntary confinement and treatment regimens. Schizophrenia may also be used, informally or legally, as an excuse or mitigating factor in determining guilt and responsibility for criminal misconduct, and may also be treated as an indicator of dangerousness and recidivism for purposes of sentencing or preventive detention. It may figure as a ground for a divorce, or a custody decision, or other legal interventions in family structure. Diagnoses of schizophrenia can also figure into competency hearings which determine whether an individual himself or herself requires a custodian. 'Schizophrenia,' like 'suicide,' raises profound questions, specifically in this case for the institutions of the family, law, and psychiatric care and control. Thus both categorizations, suicide and schizophrenia, should be understood as typically evaluative and consequential categorizations. Both categories operate in the overlap of moral, legal, and medical varieties of social conventions and social controls, as well as more diffuse cultural conventions and controls.

Not only are categorizations of suicide and schizophrenia typically evaluative and consequential, but so too are studies of coroners' verdicts and psychiatric diagnoses, in the sense that studies of professional decision-making are typically quite practical in nature, reflecting the interests of the funding agencies, the interests of the gatekeepers who can allow or deny access to the sites and subjects of research, and quite commonly reflecting a critical or reformist agenda on the part of the researcher him- or her-self, or perhaps merely the ambition to be 'relevant.' A common concern with suicide rates is the prevalence of erroneous determinations of cause of death, for example, and of course psychiatrists' diagnoses are susceptible to a wide range of generic professional skepticism and case-specific second-guessing.

This is not the case, however, with ethnomethodological forays into these domains of practice, of which I will single out for commentary Max Atkinson's book Discovering Suicide (1978), and Jeff Coulter's publications on schizophrenia, and mental illness or insanity more generally (see, esp., 1991a; 1973; 1983: 148-152). Not only do both authors explicitly disavow any practical interest in criticizing or improving the categorization practices of coroners and psychiatrists, but both authors demonstrate, cumulatively and compellingly, over the course of their respective studies, the possibility of studying evaluative categories in a non-evaluative manner.

I want to briefly summarize what I take to be some of the relevant details of these respective studies, with two goals in mind. First, I would like to further illustrate the non-evaluative nature of these studies, in the interest of making a larger point about the possibility of non-evaluative research in the human studies. Second, I would like to suggest what I take to be a quite promising direction for future studies of evaluative categories, suggested by Coulter's use of ethnomethodological analysis to complement Wittgensteinian inquiries into the logical grammars of natural language concepts, especially concepts of the 'mental,' including in this case the category of 'schizophrenia.'

Atkinson's book, Discovering Suicide, can be described as a cumulative account of a young scholar's conversion to ethnomethodology, a conversion which takes place over the course of his dissertation research, as he struggles simultaneously to find his place in the literature on suicide and make sense of a variety of ethnographic, interview, and documentary data concerning the work of coroners and coroners' officers. Atkinson ultimately accomplishes an ethnomethodological respecification of the suicide question. Constraints of time and space prevent me from addressing the important question of why one might find non-evaluative studies worthwhile, but Atkinson's book does offer an unusual narrative of an earnest researcher driven by his scholarly concerns into a progressive disenchantment with conventional positivisticii and ironic/correctional formulations of the suicide question. One crucial aspect of this disenchantment which Atkinson expresses well is simply that many conventional studies, because they neglect the question of how the relevant data is identified and assembled, don't realize that many types of data are identified and assembled by means of commonsense theories (e.g. commonsense theories of suicide or mental illness) which are uncritically carried over into professional theories, thus rendering any testing of professional theories against such data a variety of circular reasoning, potentially involving self-fulfilling prophesies (1978; 1990: 462-465). But the disenchantment extends as well to the symbolic interactionist alternative offered by Jack Douglas (1967). Concurrently with this progressive disenchantment, Atkinson's analysis communicates a growing appreciation for the empirical and analytic promise of an ethnomethodological and non-evaluative approach to his phenomena. After such a respecification, suicide research revolves largely around the fundamental question of how categorizations of suicide get made, as a matter of practical methods and practical reasoning, given that a number of alternative categorizations are available for describing sudden deaths, including natural death, accident and homicide.

Much of the discussion in Atkinson's book details the practical reasoning involved in the process of establishing practically adequate grounds for reaching one formal verdict and official categorization for particular sudden deaths, given this availability of several verdicts. This is an important point, so I will paraphrase it in two additional ways, with the intention of prefiguring the subsequent development of my argument. One way to put this is to say that coroners and coroners' officers are involved in deciding on a categorization by which particular sudden deaths are accountable - that is, deciding on a formal, public decision about what sudden deaths amount to, what they count as, how they are observable and intelligible and reportable as this, that, or the other culturally known and officially recognized variety of sudden death. Another way to put this is to say that coroners and coroners' officers are involved in a variety of search procedure, where we could say the goal of the search is to establish the case-specific grounds for reaching the final verdict, the appropriateness of which is partly a function of not finding adequate grounds for reaching all other verdicts, adequate grounds for which should be accountably lacking.iii

The details of coroners' work are complex; some of the key grounds for categorizing a sudden death as a suicide, by which a death becomes accountably a suicide, according to Atkinson (1978: 110-147), relate to communications (especially suicide notes, and verbal warnings or threats of suicide), mode of death (e.g. hanging, gassing, overdose), location and circumstances of death (e.g. places where one would not normally go), and the biography of the deceased, under which category would fall medical conditions (such as chronic illness, chronic pain), psychiatric conditions (mental illness, depression), and social conditions (including divorce, loss of a loved one, loss of a job, forced retirement, etc.). The intricate details of such work, however, presuppose a more basic understanding of suicide as the intentional taking of one's own life, in keeping with which the fatality of the incident and the suicidal intent are arguably the key criteria of suicides. Since coroners are only involved after a fatality, their work insofar as suicide verdicts are concerned amounts largely to judging evidence as to the presence or absence of suicidal intent or motive. Significantly, Atkinson discusses such work as involving 'cues' or 'criteria' for inferring intent (1978: 142-143), which will turn out to be a relevant point of comparison to Coulter's work on mental illness.

Atkinson is notably indebted to the more programmatic comments made by Garfinkel (1967b) in the course of his discussion of the work of the Los Angeles Suicide Prevention Center, but Atkinson makes perhaps his most important contribution to the previous work of Garfinkel and Sacks in his careful and extended attention to the question of how coroners practically go about deciding between the available verdicts in specific cases, or the question of how the categorization of sudden deaths actually gets done as a practical and professional matter. Thus, where Garfinkel refers to the work of the Los Angeles Suicide Prevention Center as involving 'common-sense situations of choice' (1967b:171) and refers to the requirement of assigning one of several "titles" of death to each case (1967b:173), Atkinson's book goes further by reporting upon the lived details which illustrate and substantiate Garfinkel's more abstract and programmatic conclusions, specifically reporting upon how common-sense situations of choice, involving choices between various institutionally allowable titles for sudden deaths, are handled as practical matters.

Similarly, Sacks' programmatic piece on 'sociological description' (1963) states the importance of the type of work which Atkinson actually carried out. Most notably, Sacks (1963) writes:

An investigation of how it is that a decision that a suicide occurred is assembled, and an investigation of how an object must be conceived in order to talk of it as 'commiting suicide,' these are the preliminary problems for sociology. Having produced procedural descriptions of the assembly of a suicide classification it may turn out that it is the category and the methodology for applying it that constitutes the interesting sociological problems (p. 8).iv

Note that this is exactly what Atkinson's study subsequently delivers: Atkinson produces procedural descriptions of the assembly of suicide classifications, and in such a way as to argue, with reference to the details of his study as well as to methodological and theoretical considerations, how and why that type of inquiry might actually raise and address the most interesting sociological questions concerning suicide.

One important point of comparison between Atkinson's study of suicide and Coulter's work on schizophrenia is that they both pay serious attention to the criteria used in the practices of categorization, i.e., the criteria for reaching a verdict of suicide or for a diagnosis of schizophrenia. In Coulter's work on Schizophrenia, however, much of the discussion of these criteria appears in the service of an argument to the effect that the category 'schizophrenia' is a polymorph, or a catchall category which can refer to so many and such diverse varieties of speech and conduct as to seriously undermine etiological inquiries, especially etiological inquiries which posit a single physiological determinant of schizophrenia. He suggests that the chronic problems of inconsistent operationalizations and the lack of comparable criteria for 'schizophrenia' across studies reflect an "attempt to regiment this core concept's usage beyond its latitudes of tolerance" (1991a: 163). In an important sense, then, Coulter is suggesting that 'schizophrenia' has been reified into a singular, natural kind, when the diversity of conduct and speech in question and the persistent failures of etiological research both suggest the need to de-reify 'schizophrenia' and instead exhibit the range of properties and uses of this concept, as a tool of practical diagnostic work (cf. 1991a: 168). Coulter asks rather pointedly: "If, as we are told with growing documentation, the category of 'schizophrenia' lacks uniform indicators, clear boundaries, common properties, strict recognition rules, or context-independent diagnostic criteria, what is it for which an etiological search is being mounted?" (1991a: 169). Speech and conduct which can lead to diagnoses of 'schizophrenia' range from catatonic withdrawal, visual and auditory hallucinations, delusions, and disordered speech and thought, and these often appear independently of one another.

Despite the significant differences in emphasis, one of Coulter's primary arguments concerning the constellation of schizophrenia, clinical psychiatry, and research in the tradition of psychopathology, is in a way reminiscent of Atkinson's arguments concerning the constellation of suicide, the professional work of coroners, and suicidology -- namely, that categorization practices are such an important part of the subject matter that we actually need to understood these categorization practices as constitutive of the phenomena being categorized; categorization practices are internally and logically related to the actions and identities in question, rather than being externally related in an empirical and contingent manner. This conclusion in both cases renders the program of causal explanation analytically dubious, at best, and must be understood in the context of the concern of both authors to address how categorizations of suicide and schizophrenia are accomplished, rather than offer deterministic causal explanations for why some members of our culture kill themselves, or speak and behave in ways that are unintelligible. The latter, conventional explanations conceive of human action and belief in terms of causes and effects which are contingently and mechanically related as discrete phenomena, missing the reflexive, mutually constitutive relationships between actions and their grounds, or between delusions and their explanations, following the mundane and holistic logic of what Garfinkel called the 'documentary method of interpretation' (1967a:76-103). That is to say, classifications can be seen as implying (or at least prompting a search for) grounds which would render them accountably adequate, as much as grounds for the adequacy of classifications imply or prompt the classifications. Classifications and grounds both simultaneously confer meaning on each other in a reflexive relationship, and neither make full sense independently of each other, as they are defined partly by their relation to one another and the practices in which they are united.

Coulter's work on schizophrenia does not detail the grounds for diagnoses of schizophrenia to the degree that Atkinson details the grounds for verdicts of suicide. Rather, Coulter's work advances along another front, and constitutes a significant contribution to the naturalistic analysis of classifying practices especially by combining Wittgensteinian insights from analytic and linguistic philosophy with ethnomethodological and conversation-analytic sensibilities and methods.v

A good place to begin the consideration of this combination of analytic traditions is with Peter Winch's famous book, The Idea of a Social Science and its Relation to Philosophy (1958), in which Winch outlines Wittgenstein's relevance to the philosophy of social sciences (c.f. Baccus, 1986).

One of the most relevant elements of Winch's arguments is the observation that "A regularity or uniformity is the constant recurrence of the same kind of event on the same kind of occasion; hence statements of uniformities presuppose judgments of identity" (1958: 83) He continues soon after, "So to investigate the type of regularity studied in a given type of enquiry is to examine the nature of the rule according to which judgements of identity are made in that enquiry" (1958: 84). This argument soon brings him to the observation that such examinations would need to "include learning the criteria according to which [the relevant people] make judgements of identity" (1958: 84).

In the Wittgensteinian philosophy of language, this type of learning amounts to investigations of the logical grammar of concepts, such as investigations into the logical grammar of the concept 'schizophrenia.' Logical grammars of concepts identify rules for their intelligible use. This aspect of Wittgensteinian linguistic philosophy is intimately related to what has been called his 'use theory of meaning,' suggesting that meaning is a function of usage (Winch, 1958: 123). So the concern to specify 'rules of use' for concepts should not be misunderstood as involving the philosophical legislation of correct usage, rather, logico-grammatical investigations are descriptive, explicative inquiries which remind us of the meanings of various natural-language concepts by drawing attention to how these concepts are used in our mundane practices and speech. These investigations identify rule-governed aspects of our language, but in a descriptive rather than prescriptive manner, as is suggested by Winch's notion of philosophy as 'uncommitted enquiry,' in the spirit of Wittgenstein's principle that 'philosophy leaves everything as it was' (Winch, 1958: 102-103). You might note the similarity of these principles with the notions of 'ethnomethodological indifference' 'unmotivated examination,' and 'disinterested description' discussed previously.

A very important question arises at this point, about the compatibility of Wittgensteinian logical grammar with ethnomethodological and conversation-analytic studies, which are largely empirical enterprises. Two brief comments are in order, both of which draw significantly on the work of Coulter. First, there is absolutely no reason why the study of the practical use of concepts in our everyday lives can not be expanded in an empirical direction, enriching the methods of logico-grammatical investigations beyond examples drawn from memory, hypothetical scenarios, and the like. In this respect, ethnomethodological and conversation-analytic studies can directly contribute to and improve upon logico-grammatical investigations. There is nothing in Wittgenstein's works which would preclude such an innovation, which is very much in keeping with an emphasis on the use of concepts in natural languages, tied with practical activities and ultimately rooted in collective forms of life.

Second, the goals of logical grammar/conceptual analysis and ethnomethodology can be formulated as to overlap very significantly. Logical grammar is concerned with describing the conditions for intelligible use of a concept, which are called the 'criterial features' of a concept.vi This concern may appear to be tangential for empirical social science, but it is not. The following discussion in Winch is very instructive: "... if the sociological investigator wants to regard [intellectual or social events] as social events... he has to take seriously the criteria which are applied for distinguishing 'different' kinds of actions and [also for] identifying the 'same' kinds of actions within the way of life he is studying. It is not open to him arbitrarily to impose his own standards from without. In so far as he does so, the events he is studying lose altogether their character as social events" (1958:108). In his continuation of this discussion, Winch then prefigures some of the most important programmatic statements on suicide by Sacks and Atkinson by suggesting that Durkheim's Suicide was quite problematic by these standards (1958: 111).

Ethnomethodology, for its part, is often concerned with describing the accountability of social actions, social identities, and other social phenomena, all of which are only accountable (observable, reportable, describable) as being certain types of action (such as suicide) and certain types of identity (such as schizophrenic). These particular conventional meanings must, necessarily, be represented by or identified as concepts, such as the concepts 'suicide' and 'schizophrenia.' The use of concepts in natural language and social interaction is clearly an empirical phenomenon, but it is also a rather privileged phenomenon, since it is inseparable from the question of methodology and ultimately constitutive of the topics of social order and social structure. This should suggest how central accountability is among the concerns of ethnomethodology. But it should also suggest that intelligibility and accountability can be seen as amounting to much the same thing, indicating a profound overlap between analytic studies of logical grammar and empirical, ethnomethodological studies of accountability. This connection is especially apt in that in both traditions, the concerns with intelligibility and accountability express the basic insight that meaning is rooted in, constrained and enabled by social practices, and especially social practices of describing, naming, formulating, referring, classifying, categorizing, labeling, diagnosing, reaching a verdict as to an appropriate identification of an action or person legal purposes, etc.

This is certainly not the only argument to the effect that ethnomethodology is compatible with, or overlaps with, or profitably extends, central themes in Wittgensteinian ordinary language philosophy. In addition to numerous publications by Coulter, several other scholars have noted parallels or debts, in different but complementary ways (e.g. Heritage, 1978; Lynch, 1992, 1993; Francis, 2003).

John Heritage, for one, sees a point of convergence between Wittgensteinian linguistic philosophy and ethnomethodology in a common interest with the indefiniteness of descriptive resources (1978: 179), related to what Sacks called the 'etcetera problem' facing attempts at complete literal description in the social sciences (1963). An appreciation of the indefinite character of natural language and natural language terms is described by Heritage as a "condition for the theoretical treatment of conflicting descriptions, the flexibility of human descriptive resources and their novel applicability together with related problems concerning personal, cultural and linguistic change, cultural and linguistic relativism etc." (Heritage, 1978:79).

Heritage suggests that Garfinkel's interests in the understanding of natural language "derive from an interest in the ways in which descriptors can be concertedly and intelligibly used by practical language speakers," despite the "inexhaustible describability of states of affairs" in society (Heritage, 1978:85). While Heritage rightly traces this central interest of Garfinkel's ethnomethodology to phenomenological philosophy and gestalt psychology, it also overlaps significantly with the concerns of Wittgensteinian logical grammar, which informed early ethnomethodological scholarship through the influence of Harvey Sacks. Philosophical investigations into the logical grammar of concepts could aptly be paraphrased in the language Heritage uses (above) to describe Garfinkel's ethnomethodological concerns, as investigating the ways in which descriptors (concepts) can be concertedly and intelligibly (grammatically) used by practical language speakers to successfully and competently communicate with other members of a culture and speakers of a natural language, despite the indeterminate and indefinite (fuzzy, vague, muddy, etc) nature of natural language resources. This indefinite character of natural language terms even provides for the basic nature of description as practical activity which is the topic of the present paper; given that there is a practical selection problem of which criteria to orient to and which descriptions to use, "the selection of descriptors may be inspected to locate what the describer is understandably attempting to do with his description" (Heritage, 1978:91). This point is equally consistent with the ethnomethodological understanding of language as a means of practical action and a Wittgensteinian understanding of concepts and words as instruments, figuring practically into various language games (c.f. Coulter, 1991b:27).

Michael Lynch has also addressed similar issues in the course of outlining an ethnomethodological approach in the social studies of science. Lynch observes of Wittgenstein,

In recommending description rather than explanation, Wittgenstein took into account that a description is not a 'word-picture of the facts' and that descriptions 'are instruments for particular uses' [Wittgenstein, Philosophical Investigations, sec. 291]. He did not propose delivering singularly correct descriptions of language use. Instead, he advocated a kind of reflexive investigation, in which philosophy's problems are addressed by 'looking into the workings of our language.' (p. 198-199).

This understanding of Wittgenstein's position on describing language in its workings leads Lynch, then, to a discussion of extending Wittgenstein in an empirical direction. Lynch suggests that empirical cases may play a role in Wittgenstein's project in the way that Garfinkel suggested using empirical investigations not as grounds for inductive theory-building or deductive theory-testing, but as "aids to a sluggish imagination" [Garfinkel (1967a: 38), quoted in Lynch, 1993: 199] in the course of reflective attempts to render the familiar as strange, in order to explicate the properties of mundane, taken-for-granted methods of practical action and practical reasoning. Empirical studies can therefore provide 'perspicuous representation' of the use of our words, which is exactly what Wittgenstein argued is necessary for his ordinary language philosophy (Lynch, 1993: 199).

With these arguments setting the context, I would now like to offer a brief and partial summary of Coulter's use of Wittgensteinian notions of logic, grammar, and 'criteria' alongside and in the course of his ethnomethodological respecifications of mental predicates and mental phenomena, including schizophrenia, mental illness, and insanity.

Let me begin with an extended quotation from one of the most sustained treatments of these topics in Coulter's writings; in Mind in Action, Coulter (1989) is developing programmatic arguments for an 'epistemic sociology,' and in this context he describes the central notion of 'grammars of practical cognition':

Grammars of concepts are rules which not only specify the linguistic frameworks within which words, phrases or types of words or phrases may be used, but also 'what counts as an application of' such expressions' [Stanley Cavell]. Grammars reveal the manifold connections between words and other words, phrases and expressions as these are used by 'masters of natural language', and the manifold connections between kinds of expression and the sorts of circumstances within which and about which they may be used. Since it is the use of words, phrases and other forms of expression which articulate our concepts, and it is our concepts which afford us knowledge of the world... then the (evolving) grammar of our conceptual apparatus establishes (ongoingly) whatever intelligibility the world possesses for us. Exploring the connections and ramifications of concept-use for any domains of human interest within social contexts, practices, and institutional arrangements - the grammars of cognition - becomes the overarching interest of the kind of 'epistemic' sociology I am concerned to exemplify here. It might be argued that such inquiries are essentially 'non-empirical' in the sense that, although appeal is made to empirical... instances of human activities... in documenting parameters of conceptual articulation and the ways in which 'phenomena' of various kinds are thus rendered intelligible, none the less the purpose of such analyses is 'logical' and thus ultimately distinct from a sociological enterprise. If, however, by 'logical' we mean 'grammatical' in Wittgenstein's (extended) sense... then it is clear that we can only arrive at our specifications of 'grammar' by inspecting actual occasions of reasoned use-and-context. We need for such a purpose elaborate 'reminders' when it is a question of examining facets of our commonplace cognition, and such 'reminders' take the form of systematically exemplifying 'data' derived from real-world social events and social processes (p. 49-50).

Coulter goes on to suggest that in this extended sense, grammatical analysis is logical analysis. According to this understanding of logic, logic has to do quite generally with the question of orderliness, which is also a central concern of sociology and ethnomethodology, and so one finds Coulter at times altering the descriptor 'sociological,' putting a hyphen in the middle of it, rendering 'sociological' inquiry into 'socio-logical' inquiry, including inquiry into the social logic of concept deployment in natural language use. This is captured in at least one point as the transformation of 'sociological explanation' into 'socio-logical explication' (1989:5). The explication of grammars for natural language concepts, grammars which specify criterial features for concepts, is clearly one important method of examining the logic of practical action and practical reasoning, as grammars suggest the basic conditions of intelligibility and sense-making, without which neither members nor analysts could provide for the identity and order of any of our experiences.

Another expression of Coulter's respecification of themes from conventional sociology is his characterization of a new sociology of knowledge as involving "a greater interest in the logic of the achievement of 'intelligibility' in social life" (1989: 3). Coulter's vision for the sociology of knowledge, reflecting ethnomethodological, Wittgensteinian, and also Schutzian traditions of inquiry, challenges the orthodox notion in analytic philosophy that the analyst is restricted to 'invented' instances or examples as 'data,' and argues quite compellingly that the study of the practical logic of various conceptual and epistemic domains can benefit enormously by empirical data, and especially data from naturally occurring language use, of the kind emphasized in ethnomethodological studies.

More specific to ethnomethodology, here is a relevant discussion of ethnomethodology as a variety of logical inquiry:

Knowing what people are doing (including oneself) is knowing how to identify what they are doing in the categories of a natural language, which requires knowing how to use those categories in discursive contexts, which includes knowing when to utter them. All of these types of knowledge are logically interrelated. They are all constitutive of human conduct. This integral linkage between what an action is (or what it could possibly, conventionally be) for those engaged in it, and how it could be 'made sense of' for anyone producing or witnessing it (correctly, appropriately, properly, rationally, etc.), forms a central insight of Garfinkel's programme of ethnomethodology (the study of the logic of practical action and practical reasoning in social affairs)vii. As Garfinkel articulates this insight, 'the activities whereby members [of society, of a common language and culture] produce and manage settings of organized everyday affairs are identical with members' procedures for making those settings 'account-able'. The 'account-ability' of conduct and social arrangements (i.e. their intelligibility, describability, recognizability, etc.) is made possible by the very concepts which in turn make possible the production of such conducts and settings. I must satisfy the criteria for the conventional conceptualization ('account-ability' or describability) of my conduct as X-ing if indeed I may properly be said to have X-ed in the first place. And, naturally enough, the same is true for you, he, she, we and they! (1989: 15-16).

Similarly, Coulter refers to Garfinkel's work "on the natural logic of situated practical action and practical reasoning" (1973:ix). Relating to the Wittgensteinian notion of 'criteria,' Coulter offers another characterization of ethnomethodology which is especially relevant here; He argues, "One of the founding tenets of the ethnomethodological enterprise was to analyze human activities in terms which would preserve, without distortion, members' criteria for descriptive adequacy in relation to those activities. It is precisely this sort of constraint upon analytical conceptualization which distinguishes ethnomethodological inquiry, and we fully endorse it as a methodological principle" (Coulter and Parsons, 1991: 263-264).

Coulter's studies of schizophrenia, mental illness, insanity, etc. illustrate these programmatic concerns with domain-specific observations and arguments, amounting to studies of the natural logic of insanity ascription (1973: vii). According to this respecification of the sociology of mental illness, the focus of the field should be on "the conventional procedures and presuppositions involved in any set of recorded instances of talk about psychological status" (1973: 152). In Coulter's most recent work on schizophrenia, he argues that we can avoid the reification of the concept 'schizophrenia' by developing 'sociological elucidations of its grammar of living usage' (1991a: 170).

Coulter's analyses of mental categorizations and predicates therefore draw from both ethnomethodology and linguistic philosophy, in the tradition of Wittgenstein, Ryle, and Austin. Coulter's work can profitably be understood as illustrating what Austin called 'fieldwork in philosophy,' and what Paul ten Have has discussed more recently under the rubric of 'empirical philosophy' (1997).

Let me explicitly relate what I take to be some of the significant aspects of the analytic and methodological developments I have been trying to describe. I began with the argument that ethnomethodological scholarship can and should be disinterested, in a sense similar to Garfinkel's principle of 'ethnomethodological indifference,' and Winch's notion of conceptual analysis as 'uncommitted inquiry.' Contrary to the fashionable skepticism concerning such battered terms as 'objectivity' 'value-freedom' and 'value-neutrality,' ethnomethodological and Wittgensteinian scholars have both been demonstrating the possibility of disinterested description for decades. This is perhaps nowhere so clear as in the case of the empirical studies and logico-grammatical investigations of the accountability and intelligibility of evaluative categorizations, including categorizations 'suicide' and 'schizophrenia.' Other ethnomethodological studies of relevance have addressed contested sexual identity in the case study of 'Agnes' (Garfinkel, 1967a: 116-185), 'normal crimes' (Sudnow, 1965), 'juvenile delinquency' (Cicourel, 1995), plea-bargaining (Maynard, 1984), traffic court hearings (Pollner, 1987), police interrogations (Watson, 1973), and another treatment of mental illness (Smith, 1978). And these are merely some of the better-known 'classical' contributions among a wealth of relevant studies. This scholarship demonstrates the ability to achieve rigorous and disinterested description in domains of practice which are notable for disagreements among experts as well as laypeople.


In the above I have argued for disinterested description and for combining two traditions of inquiry, each descriptive rather than evaluative in purpose. But I would argue as well that evaluative categories, including those that name any number of individual and social problems, can be recognized as an especially important collection of subject matter for such inquiries. Rather than the contentiousness of these subjects being cause for analytic concern, this contentiousness is analytically invaluable, because debates about the applicability of evaluative categories display especially frequently and clearly the social practices of making phenomena accountable and intelligible in situ, practices in which the participants routinely orient to and display the relevance of the grounds or criteria of evaluative categorizations, in the content and process of their arguments and their (other) actions.

In his discussion of the flexibilities of natural language use, Heritage briefly discusses a transcript from a cross-examination of a senior police officer in Northern Ireland who is suspected of being overly polite in attempting to disperse a protestant "mob who had burned and pillaged a catholic area" (1978: 92). In the brief transcript provided by Heritage, it becomes clear that the officer contests the description of the Protestants as a 'mob,' instead characterizing them by the more neutral term 'crowd,' and the officer contests as well the description of himself as having knowledge, at the time, of the true extent of the protestants' violence. With respect to the distinction between the descriptors 'mob' and 'crowd,' Heritage (1978) briefly summarizes the theoretical and methodological significance of this dispute. He observes that:

(1) the contested descriptive distinction between a 'crowd' and a 'mob' is (2) not amenable to definitive (or criterial) definition, but summarizes a variety of different (themselves indefinite) attributes of large social groups; (3) instructs hearers to attend to different dimensions of the state of affairs described; (4) as part of that instruction, instructs hearers to prospectively orient themselves to the possibility of an 'accusation/defence' sequence concerning the adequacy of the police officer's conduct... and (5) could not be other than the object of ... 'persuasive activity' and indeed is hearably the object of such activity (p. 92).

One way in which this incident could be understood is in terms of a negotiation or contest about whether the officer's conduct satisfied the critieral features of professional competence or responsibility, or if on the other hand the officer's conduct satisfies the criterial features of incompetence, or negligence, or some other professional or legal failing. Certainly it can make a very practical difference in terms of an evaluation of professional conduct whether the officer is understood as having been polite with a 'crowd' or having been polite with a 'mob,' just as it can make a very practical difference whether the officer was polite with people he knew had burned and pillaged, or with people he should have known had burned and pillaged, or with people he could not be expected to have known had burned and pillaged. The confrontational nature of this cross-examination reveals two interlocutors describing "the same" incident differently, informed by the disjunctive interests which inform the disjunctive descriptions: the interest in investigating or faulting the conduct of the officer, on the one hand, and the interest in defending the officer's conduct, on the other. Just as violent or riotous behavior is a criterion of a mob, as distinguished from a crowd, and knowledgeability is a criterion of responsibility, as distinguished from innocence or mitigation, the cross-examination can be understood as pivoting around criterial features relevant for establishing whether or not the officer knowingly treated a violent mob with kid gloves.

Another example suggesting the possibility of combining a Wittgensteinian interest in the logical grammar and criterial features of concepts with ethnomethodological and conversation-analytic studies can be drawn from the author's contemporary research on discrimination. This will further illustrate some of the heuristic value of the arguments above and emphasize the profitability of studying evaluative concepts in contexts of arguments. In a precedential U.S. Supreme Court case (McCleskey v. Kemp 107 S. Ct. 1756 [1987]), concerning a complaint of systemic racial discrimination against Blacks in the state of Georgia's capital sentencing system, a finely ordered question and answer sequence between the Supreme Court Justices and the legal counsel for the Black plaintiff on death row can be examined in light of above observations and contentions.

  68. QUESTION:   But the statutory provisions - you don't allege that Georgia's
  69.                      death penalty statutes are unconstitutional?
  70. [PETITION.]: Not facially, Your Honor;
  71.                      they were proved of course in the Gregg case.
  72.                      But they have proven incapable, in fact,
  73.                      of preventing this kind of discrimination
  74.                      which we've documented.
  75.                      And so, like Furman, we've reached a point where the Court,
  76.                      we argue, must say, these procedures, whatever they are,
  77.                      have not worked in the State of Georgia [--]
  78. QUESTION:                                                               [Well,]
  79.                      Mr. Boger, don't you have to show
  80.                      that this particular jury discriminated?
  81. [PETITION.]: Your Honor, I think we have shown
  82.                      that it's more likely than not that this jury did.
  83. QUESTION:   Well, this particular jury was only convened once.
  84.                      And I think you have to show under our cases
  85.                      that this particular jury would have dealt differently
  86.                      with a black defendant who killed a black person
  87. [PETITION.]: Well, Mr. Chief Justice, let me suggest to you why
  88.                      I believe we have made that showing.
  89.                      We of course don't have confessions from the jurors themselves.
  90.                      No one has come forward.
  91.                      But indirectly what we have is a pattern that
  92.                      Professor Baldus documented [--]
  93. QUESTION:                                                 [But]
  94.                      not a pattern on the part of this jury.
  95. [PETITION.]: No, this jury only assembles, as you say, for one decision.
  96.                      Of course, in the Bazemore case,
  97.                      you had a hundred county commissions that had to make
  98.                      judgments about what salaries were going to be paid.
  99.                      And the county commissions were composed of people
100.                      who rotated on and off because of actual politics.
101. QUESTION:   But was there any - did the constitutional holding
102.                      in Bazemore support your position, do you think?
103. [PETITION.]: I believe Bazemore assisted, as did the Batson case,
104.                      which talked about simply having proof that
105.                      makes it more likely than not that discrimination exists.
106.                      If we could show, Mr. Chief Justice, that six out of ten blacks
107.                      who murdered whites are receiving death
108.                      in a racially discriminatory fashion,
109.                      on grounds where if there were white defendants,
110.                      they wouldn't have, we would not be able to show,
111.                      of course, which ones of the six [sic.] they were.
112. QUESTION:   No, but when you're --
113.                      the institution that you're challenging is the jury here.
114.                      And its [sic.] the jury in this defendant's case.
115. [PETITION.]: Well, of course, Your Honor, it's not simply the jury.
116.                      I was responding in terms of your question.
117.                      But Professor Baldus's evidence shows dramatically
118.                      that the prosecutor plays a serious role in this process.
119. QUESTION:   Well, then, do you think your evidence supports a
120.                      finding that this particular prosecutor,
121.                      who prosecuted this case, discriminated as between
122.                      blacks who've killed whites and blacks who've killed blacks?
123. [PETITION.]: Not as between charging, but as between deciding
124.                      who to plead out to a lesser defense
125.                      or permit not to go to trial, and who to move on to penalty.
126. QUESTION:   Okay, you say your evidence supports a finding
127.                      that this particular prosecutor, in doing what you say,
128.                      discriminated in the manner I described?
129. [PETITION.]: I don't believe we have to show that a particular prosecutor,
130.                      as opposed to the prosecutorial office, Your Honor.
131.                      What we have shown is [--]
132. QUESTION:                                     [Well,] but don't you think
133.                      your evidence would support a finding
134.                      as to this particular prosecutor?
135. [PETITION.]: I think we could conclude under your Fernco rationale --
136.                      you said that we looked to all of the rational reasons.
137.                      We assume people act rationally.
138.                      We look to all of the legitimate reasons why one
139.                      would make a decision.
140.                      And if none of the legitimate reasons make the distinction
141.                      that seems to have been made, then we can infer that
142.                      what is at work is an illegitimate consideration.
143.                      In this case, we've shown that there have been 17
144.                      defendants in Fulton County who killed police officers,
145.                      or who were involved in police officer killings.
146.                      Of those 17, in the 1973-1979 period,
147.                      only two even went to a sentencing jury; and of those two,
148.                      one went before a jury having killed a black police officer.
149.                      And he received a life sentence.

Although there is clearly much interactional and legal work being accomplished in this exchange, in the context of this paper we can focus on one particularly relevant aspect of this questioning. The questioning excerpted here can be understood as an extended example of what Sacks referred to as a search procedure, in this case aimed at determining how the plaintiff's counsel would satisfy the criterial requirement that an adequate case for discrimination include an alleged perpetrator. The Justices can be heard or seen to systematically pursue three separate candidates for the role of perpetrator, first the state legislature (lines 68-69) then the trial jury (see the Justices' questions and challenges in the exchange from line 78-118), and then the District Attorney (see the exchange from line 119 to the end). The Justices demonstrate through tenacious and cumulative questioning an orientation to one constant question or theme throughout: the questionable existence of a discriminator in the defendant's legal case.

For his part, the defendant's attorney can be seen to rather skillfully evade formulating one specific person or party as the discriminator, seemingly because the nature of the defendant's evidence is such that it more clearly evidences disparate impact upon Blacks than discriminatory intent on the part of any one person or party in Georgia's capital sentencing system. It can be appreciated in the transcript that the defendant's attorney responds to the question about the constitutionality of the statutory law by acknowledging that the state statutes are constitutional on their face, but by implying that they are unconstitutional in practice (lines 68-77). The Justices pursue the issue of how the statutes are implemented in practice by asking next about the particular jury in the defendant's case (lines 79-80). The defendant's attorney at this point impugns the deliberations of the defendant's particular jury (lines 81-82). When the availability or sufficiency of his evidence against the particular jury is questioned by the Justices in subsequent turns (83-86, 93-94, implicitly in 101-102- which might be considered part of an insertion sequence, and again in 112-114), the attorney first modifies his argument in a probabilistic direction by referring to aggregate data about jury verdicts in capital cases state-wide and a precedent that arguably allows for the relevance of aggregate data (91-92, 95-100, 103-111), and then when the relevance of aggregate data meets with continued skepticism from the Justices (line 114), the attorney also challenges the prosecutor in the particular case, without retracting his allegations about the jury (115-118). When the Justices try to get the attorney to commit to a complaint of discrimination against the specific prosecutor (126-128), the attorney again falls back to a more probabilistic argument based upon aggregate statistics from the prosecutor's office (129-149).

Just as the Justices' tenacious questioning makes observable a continued search for the alleged discriminator in this case, the attorney's repeated equivocations and evasions and repeated appeals to aggregate level statistical data make observable the attorney's understanding that this criterion of discrimination is a weak link in his complaint, and that his arguments may fare best if the discrimination can be seen as a product or a characteristic of the capital sentencing system, comprising multiple juries and multiple prosecutors, rather than as the intention of any particular person or party within it.

Ultimately, the Supreme Court rejected the plaintiff's complaint of racial discrimination because of a failing to show discrimination in his specific case, seemingly due to a failing to show a discriminator in his specific case. This criterial feature of discrimination is therefore a legal standard of proof as well as a praxiological application criterion for the concept 'discrimination.'

The data excerpt above illustrates not just that a discriminator or agent is a criterial feature of discrimination, but also that participants can be seen, empirically, to orient to the question of whether a criterial feature for an evaluative category is present or absent, proved or not proved. The participants clearly oriented to this criterial feature of discrimination in determining whether the petitioner's case was legally accountable or not as a case of unconstitutional racial discrimination in capital sentencing. In this case a conceptual understanding of discrimination and the examination of an empirical case of a discrimination dispute complement each other in that both provide insight as to what is praxiologically involved in making a particular death sentence intelligible and accountable as discrimination.

The two studies mentioned briefly above, concerning the possibility of police negligence and the possibility of discriminatory sentencing, illustrate that disputes can be invaluable data for empirical studies of the accountability (observability, reportability) of instances of evaluative categories as they are demonstrably relevant to and for members in society. Especially in the contexts of arguments, the applicability of an evaluative category such as 'negligence,' 'discrimination' (or 'mental illness' or 'suicide' or 'terrorism' or 'heroic' or 'nagging' or 'over-priced' or 'nutritious' or 'cute,' etc.) will often be contested explicitly, providing invaluable opportunities for the analysis of evaluative categories and their criterial features, as a topic of both conceptual and empirical analysis. Such opportunities allow for a Wittgensteinian study of concepts with reference to their meanings-in-use and their intelligibility, grounded in their uses in various language games and forms of life, as fleshed out empirically by drawing upon the methods of ethnomethodology, conversation analysis, and membership categorization analysis.viii

A final point can be raised with respect to the concept of discrimination and the excerpt provided above, which brings us back full circle to the matter of values and politics. It may be argued that a specific perpetrator is not actually a criterial feature of discrimination, that discrimination can occur on an institutional, systemic, or societal 'level.' Indeed, that is what the plaintiff's counsel attempts in the above case. This argument has most often appeared under the name of 'institutional racism.' Far from being an objection, however, the development of the concept of 'institutional racism' in the sixties and seventies suggests that the earlier, and still conventional concept of "discrimination" has as one of its criterial features the presence and accountability of a perpetrator; this requirement had proved to be a crippling burden of proof for many progressive and radical causes, politically and legally, which is precisely the context in which the concept of 'institutional racism' emerged. This new concept was an answer to the new pragmatic need, to speak of racial inequalities intelligibly and compellingly even when one of the criterial features of 'discrimination' was difficult or impossible to establish. "Institutional racism" is therefore an example of the exception that proves the rule; it is precisely because social critics had difficulty satisfying the criteria of 'racism' and 'racial discrimination' that the new concept, 'institutional racism,' with less stringent application criteria concerning the presence and role of a perpetrator, became so important in the repertoire of social critics. A related development took place in feminist studies of work, with the 'glass ceiling' argument.

Clearly, such analytic attention to the practical use of evaluative concepts will result at times in findings that can be applied for the practical purposes of policy analysis, social criticism, administrative reform, auditing, job evaluations, business consulting, political strategy, litigation, etc. The above analysis and comments certainly have some bearing upon contemporary U.S. civil rights law and civil rights strategies.

As important as these various applications may prove to be, the primary advance for scholarship is simply the identification and exploration of a neglected dimension of controversial topics including suicide, mental illness, and discrimination, and also corruption, plagiarism, drug abuse, adultery, pollution, deficit spending, terrorism, and thousands of other evaluative categories. Such controversial topics have generated many studies and much debate, but much of it lacking the combination of conceptual-analytic leverage and detailed empirical materials necessary for truly compelling analysis. Such analysis is not only possible, but has been illustrated in ethnomethodological studies. This scholarship avoids the powerful temptation to apply controversial and evaluative terms, but promises to explicate the social logics of their meanings and uses, thus making a rare and fundamental contribution to academic social science where this overlaps with profound moral and political questions.


i The majority of ethnomethodological and conversation-analytic studies don't treat description as an explicit topic, but are nevertheless informed by and in a sense speak to an understanding of description as practical activity. Explicit, sustained coverage of description is offered in, e.g., Heap (1980). Ethnomethodological contributions to socio-legal studies have also produced many explicitly relevant papers, including Atkinson (1990), Drew (1998), and Pomerantz (1987). Many relevant studies address descriptive practices by other names, or specific kinds of descriptive practices, including formulations, accounts, glosses, and membership categorization practices.

ii Some more sustained discussions of positivism, informed by ethnomethodology (among other traditions), can be found in Bittner (1973) and McHugh (1970).

iii The notion of a 'search procedure' originates with Sacks, but Coulter uses the term at least once in the more specific manner that I am employing it here, when he refers to 'search procedures' for diagnostically relevant data, relevant to satisfying or ruling out the presence of criteria for mental illness (1979: 147).

iv Although Sacks' paper, "The Search for Help: No One to Turn to" (1967) generally addresses quite different questions about suicide, Sacks does refer to what could be called a ground or criterion for a suicide classification; specifically, he mentions that police and coroners routinely inquire as to whether or not the deceased informed available co-incumbents of relational pair categories for which such notification is expected, such as spouses, family members, and close friends. If the deceased did inform any such people of suicidal tendencies before his sudden death, this is a strong ground for suspecting suicide, and if he did not, this can be a significant obstacle to a suicide classification.

v A very useful overview of the mutual relevance of Wittgensteinian linguistic philosophy and ethnomethodology is provided in Coulter (1991b).

vi For a detailed discussion of logical criteria, see e.g. Hacker (1986:307-335). Somewhat more accessibly, Cavell suggests: "... what we discover in the course of [grammatical] investigations, when we ask 'Under what circumstances, or in what particular cases, do we say...?', are our criteria" (1979:30). Criterial features for natural language concepts are thus less formal and more contextual and pragmatic than the "necessary conditions" of academic logic.

vii Other relevant studies have addressed mental illness (Bittner, 1967), rape (Drew, 1992), an infamous gender-motivated hate crime (Eglin and Hester, 2003), deviance in school (Hester, 1992, 2000), developmental disabilities (Maynard and Gill, 1995), a political scandal/crime (Lynch and Bogen, 1996), gangs (Meehan, 2000), and discrimination (Watson, 1976; Berard, 2002; Berard, 2005b). Some more general discussions address deviance (Sharrock, 1984; McHugh, 1970b), crime (Hester and Eglin, 1992), social problems (Maynard, 1988), labeling theory (Pollner, 1978; Berard, 2003), victims and offenders (Watson, 1976), and motive attributions (Berard, 1998). Many more examples are included in a relevant bibliography on the web: www.timberard.info/_wsn/page5.html

viii Practical descriptions of identity and the reflexive, mutually elaborative relations between identity, actions, beliefs, and other attributes or predicates of persons, is the stuff of membership categorization practices and is treated by membership categorization analysis (MCA). MCA constituted one dimension of Sacks' conversation analysis and is now incorporated into or combined with ethnomethodology in the writings of, e.g., Jayyusi (1984), Hester and Eglin (e.g.1997), and many others, including the author. This growing field is rich with relevant scholarship offering non-evaluative analyses of descriptive practices related to membership categories, including many analyses which trace the relevance of identity descriptions for evaluations, and vice versa.


Thanks to Charles Antaki, Jeff Coulter, and anonymous reviewers from QSR for comments on earlier versions of this paper, one of which was presented at a bi-annual conference of the International Institute for Ethnomethodology and Conversation Analysis (IIEMCA), 4 July, 2003, Manchester, U.K


Atkinson, J. Maxwell (1978) Discovering Suicide: Studies in the Social Organization of Sudden Death. Pittsburgh: University of Pittsburgh     Press.
------. (1990) 'Ethnomethodological Approaches to Socio-Legal Studies.' in Ethnomethodological Sociology, edited by J. Coulter,     Brookfield, Vermont: Edward Elgar Publishing.
Baccus, M.D. (1986) "Sociological indication and the visibility criterion of real world social theorizing." in Ethnomethodological Studies of     Work, edited by Harold Garfinkel, New York: Routledge and Kegan Paul.
Berard, T.J.(1998) "Attributions and Avowals of Motive in the Study of Deviance: Resource or Topic?" Journal for the Theory of Social     Behaviour 28(2), p. 193-213.
------.(2002) " 'Japanese American' Identity and the Problem of Multiple Description: Disjunctive Versions of the Japanese Exclusion     Order" in Language, Interaction and National Identity: Studies in the social organization of national identity in talk-in-interaction, edited     by S. Hester and W. Housley, Ashgate.
------.(2003) "Ethnomethodology as Radical Sociology: An Expansive Appreciation of Melvin Pollner's 'Constitutive and Mundane     Versions of Labeling Theory'" Human Studies 26(4), p. 431-448.
------.(2005a) "Rethinking Practices and Structures." Journal for the Philosophy of the Social Sciences 35(2), p. 196-230.
------.(2005b) "On Multiple Identities and Educational Contexts: Remarks on the study of inequalities and discrimination" Journal of     Language, Identity, and Education 4(1), p. 67-76.
Bittner, Egon (1967) "Police Discretion in Emergency Apprehension of Mentally Ill Persons." Social Problems 14(3), p. 278-292.
------.(1973) "Objectivity and Realism in Sociology." in Phenomenological Sociology: Issues and Applications, edited by G. Psathas, New     York: John Wiley and Sons.
Bogen, David and Michael Lynch (1990) "Social Critique and the Logic of Description: A Response to McHoul /1988/." Journal of     Pragmatics 14, p. 505-521.
Cavell, Stanley (1979) The Claim of Reason: Wittgenstein, Skepticism, Morality, and Tragedy. Revised ed. New York: Oxford University     Press.
Cicourel, Aaron V. (1995) The Social Organization of Juvenile Justice. New Brunswick: Transaction.
Coulter, Jeff (1973) Approaches to Insanity: A Philosophical and Sociological Study. London: Martin Robertson.
------.(1979) The Social Construction of Mind: Studies in Ethnomethodology and Linguistic Philosophy. Totowa, New Jersey: Rowman and     Littlefield.
------.(1983) Rethinking Cognitive Theory. New York: St. Martin's Press.
------.(1989) Mind in Action. Atlantic Highlands, New Jersey: Humanities Press International.
------.(1991a) "The Grammar of Schizophrenia." in What is Schizophrenia? edited by W. Flack, D. Miller and M. Wiener, New York:     Springer-Verlag.
------.(1991b) "Logic: Ethnomethodology and the Logic of Language." in Ethnomethodology and the Human Sciences, edited by G.     Button, New York: Cambridge University Press.
Coulter, Jeff and E.D. Parsons (1991) "The Praxiology of Perception: Visual Orientations and Practical Action." Inquiry 33, p. 251-272.
Douglas, Jack D. (1967) The Social Meanings of Suicide. Princeton, New Jersey: Princeton University Press.
Drew, Paul (1992) "Contested Evidence in Courtroom Examination: The Case of a Trial for Rape." in Talk at Work: Social Interaction in     Institutional Settings, edited by P. Drew and J.Heritage, Cambridge: Cambridge University Press.
------.(1998) "Complaints About Transgressions and Misconduct." Research on Language and Social Interaction 31(3/4), p. 295-325.
Eglin, Peter and Stephen Hester (2003) The Montreal Massacre: A Story of Membership Categorization Analysis. Waterloo, Canada:     Wilfred Laurier University Press.
Francis, David (2003) "Using Wittgenstein to Respecify Constructivism." Paper presented at an American Educational Research     Association conference. Chicago.
Garfinkel, Harold (1967a) Studies in Ethnomethodology. Cambridge, UK: Polity Press.
------.(1967b) "Practical Sociological Reasoning: Some Features in the Work of the Los Angeles Suicide Prevention Center." in Essays     in Self-Destruction, edited by E.S. Schneidman, New York: Science House.
Garfinkel, Harold and Harvey Sacks (1970) "On Formal Properties of Practical Actions." in Theoretical Sociology: Perspectives and     Development, edited by J. McKinney and E. Tiryakian, New York: Appleton-Century-Crofts.
Hacker, P.M.S. (1986) Insight and Illusion: Themes in the Philosophy of Wittgenstein. Revised Edition. Oxford: Clarendon Press.
Heap, James L. "Description in Ethnomethodology." Human Studies 3(1), p. 87-106.
Heritage, John (1978) "Aspects of the Flexibilities of Natural Language Use: A Reply to Philips." Sociology 12, p. 79-103.
Hester, Stephen (1992) "Recognizing References to Deviance in Referral Talk." in Text in Context: Contributions to Ethnomethodology,     edited by G. Watson and R. Seiler, London: Sage.
------.(2000) "The Local Order of Deviance in School: Membership categorization, motives and morality in referral talk." in Local     Educational Order: Ethnomethodological Studies of Knowledge in Action, S. Hester and D. Francis, Philadelphia: John Benjamins.
Hester, Stephen and Peter Eglin, eds. (1992) A Sociology of Crime. New York: Routledge.
------.(1997) Culture in Action: Studies in Membership Categorization Analysis. Washington, D.C.: International Institute for     Ethnomethodology and Conversation Analysis & University Press of America.
Jayyusi, Lena (1984) Categorization and the Moral Order. Boston: Routledge and Kegan Paul.
------.(1991) "Values and Moral Judgment: Communicative Praxis as a Moral Order." in Ethnomethodology and the Human Sciences,     edited by. G. Button, New York: Cambridge University Press.
Lynch, Michael (1992) "Extending Wittgenstein: The Pivotal Move from Epistemology to the Sociology of Science." in Science as Practice     and Culture, edited by A. Pickering, Chicago: University of Chicago Press.
------.(1993) Scientific Practice and Ordinary Action: Ethnomethodology and social studies of science. New York: Cambridge University     Press.
Lynch, Michael, and David Bogen (1996) The Spectacle of History: Speech, Text and Memory at the Iran-Contra Hearings. Durham: Duke     University Press.
McCleskey v. Kemp Oral Arguments (1988) Pp. 961-994 in Landmark Briefs and Arguments of the Supreme Court of the United States:     Constitutional Law, 1986 Term Supplement, Vol. 171, edited by P. Kurland and G. Casper. University Publications of America. [Based on     transcripts at the Supreme Court Library].
Maynard, Douglas W. (1984) Inside Plea Bargaining: The Language of Negotiation. New York: Plenum.
------.(1988) "Language, Interaction, and Social Problems." Social Problems 35(4), p. 311-334.
Maynard, Douglas W. and Virginia Teas Gill (1995) "On 'Labeling' in Actual Interaction: Delivering and Receiving Diagnoses of     Developmental Disabilities." Social Problems 42(1), p. 11-37.
McHugh, Peter (1970a) "On the Failure of Positivism." in Understanding Everyday Life: Toward the Reconstruction of Sociological     Knowledge, edited by J. Douglas, Aldine.
------.(1970b) "A Common-Sense Conception of Deviance." in Deviance and Respectability: The Social Construction of Moral Meanings,     edited by J. Douglas, New York: Basic Books.
Meehan, Albert J. (2000) "The Organizational Career of Gang Statistics: The Politics of Policing Gangs." The Sociological Quarterly 14(3),     p. 337-370.
Pollner, Melvin (1978) "Constitutive and Mundane Versions of Labeling Theory." Human Studies 1, p. 269-288.
------.(1987) Mundane Reason: Reality in Everyday and Sociological Discourse. New York: Cambridge University Press.
Pomerantz, Anita (1987) "Descriptions in Legal Settings." in Talk and Social Organisation. Clevedon, edited by G. Button and J.R.E. Lee,     Philadelphia: Multilingual Matters.
Sacks, Harvey (1963) "Sociological Description." Berkeley Journal of Sociology 8, p. 1-16.
------.(1967) "The Search for Help: No One to Turn to." in Essays in Self-Destruction, edited by E.S. Schneidman, New York: Science     House.
------.(1984) "Notes on Methodology." in Structures of Social Action: Studies in Conversation Analysis, edited by J. Maxwell Atkinson and     J. Heritage, New York: Cambridge University Press.
Sharrock, Wes (1984) "The Social Realities of Deviance." in Applied Sociological Perspectives, edited by R.J. Anderson and W.W.     Sharrock, Boston: George Allen and Unwin.
Smith, Dorothy E. (1978) " 'K is mentally ill' The Anatomy of a Factual Account." Sociology 12, p. 23-53.
Sudnow, David (1965) "Normal Crimes: Sociological Features of the Penal Code in a Public Defender Office." Social Problems 12, p.     255-76.
Ten Have, Paul (1997) "Conversation Analysis as 'Empirical Philosophy.' Paper read at the conference on "Ethnomethodology, an     improbable sociology?" Cerisy-la-Salles, France.
Watson, Rod (1973) "The Presentation of Victim and Motive in Discourse: The Case of Police Interrogations and Interviews."     Victimology 8, p. 31-52.
------.(1974) "The Language of Race Relations: Making Racial Discriminations." Journal of Comparative Sociology 2, p. 89-101.
------.(1976) "Some Conceptual Issues in the Social Identifications of Victims and Offenders." in Victims and Society, edited by E.C.     Viano, Washington, D.C.: Visage.
Wilson, Thomas P. (1971) "Normative and Interpretive Paradigms in Sociology." in Understanding Everyday Life: Toward the     Reconstruction of Sociological Knowledge, edited by J.D. Douglas, London: Routledge and Kegan Paul.
Winch, Peter (1958) The Idea of a Social Science and its Relation to Philosophy. London: Routledge and Kegan Paul.
Zimmerman, Don and Melvin Pollner (1970) "The Everyday World as a Phenomenon." in Ethnomethodological Sociology, edited by J.     Coulter, Brookfield, Vermont: Edward Elgar Publishing.


Berard, Tim J. (2005) "Evaluative Categories of Action and Identity in Non Evaluative Human Studies Research: Examples from     Ethnomethodology". Qualitative Sociology Review, Vol. I Issue 1. Retrieved Month, Year
    (http://www.qualitativesociologyreview.org /ENG/archive_eng.php)

© Qualitative Sociology Review 2005