Volume 10 Number 2 1997

Studies of Nothing

Ethnomethodology and the Human Sciences, Graham Button (ed), Cambridge University Press, 1991, 19.95 paperback

Talk and Social Structure: Studies in Ethnomethodology and Conversation Analysis, Deirdre Boden and Don H Zimmerman (eds), Polity Press, 1991, 305 pages, 35 cloth

Erving Goffman, Tom Burns, Routledge, 1992, 386 pages, 12.99 paperback

Andrew Travers
Department of Sociology
University of Exeter

Twenty-five years after Studies in Ethnomethodology, its revolutionary claims remain unchanged. But now, as Garfinkel puts it, there is "a very large corpus" showing "in detail, contrary to the entirety of the social science movement, in incommensurably asymmetrical alternate sociology, the local production and natural reflexive accountability" of what "analysis incarnate in and as ordinary society" could "adequately" be. Revolutionary intent has become retrospective fact.

That "social actions are irreducibly events-in-a-social-order" is still forgotten - Button says in his Introduction - by all except the ethnomethodologists, who find in this forgetting an unavoidable, irremediable reflexivity that is merely a part of the topic of true (ethnomethodological) sociology. So Ethnomethodology and the Human Sciences is an attempt to spell out the ethnomethodological philosophy for non-ethnomethodologists. Separate chapters are devoted to Logic (Coulter), epistemology (Sharrock & Anderson), measurement (Lynch), evidence/inference (Benson & Hughes), the "social actor" (Sharrock & Button), cognition (Coulter), language/culture (Lee), and the moral order of human conduct (Jayyusi). Much is clarified and emphasized by this topic-by-topic exegesis, which has a narrative progression that I shall now try to catch as it really is one of the most well-rounded statements of ethnomethodology to date.

Following Garfinkel's prefatory statement, Coulter and Sharrock and Anderson argue a profoundly anti-positivist ethnomethodological orientation respectively to Logic and to a `scientific' epistemology. Coulter charts the trajectory of Logic from its early Greek formulation as praxis to an "end point in Wittgenstein's Tractatus, Ayer's Language, Truth and Logic and the work of the Vienna Circle". In the heyday of philosophical logical positivism, Logic was (l) formalized in mathematical language, (2) hegemonized as an ideal language, and (3) separated from philosophy. Since then there has been a slow return to Logic as praxis in late Wittgenstein, Ryle, Austin, and Winch: conduct may be analysed through the analysis of the grammar of its concepts. Within this gestalt, Coulter says, it is Garfinkel who sees, for the first time in sociology, the importance of the thoroughgoing indexicality in all accounting practices whereby, defying the demand of positivist `repair', people do succeed in understanding one another very well. That understanding becomes ethnomethodology's topic, and it is construed (by Sacks) as being crucially dependent upon sequence: the process of membership categorization (and recipient design, the consistency rule, inferential adequacy, standardised relational pairing, and the relevance rule) shows how members achieve a sequenced moral categorization through categorial identity predication in opaque contexts where hearings can be generated in equivalence to the hearings of those who are predicated. Thus ethnomethodology reunites Logic and sociology in praxis. And it does this by treating the world as Schutz treated it (Sharrock and Anderson). Schutz argues that the natural attitude of members whose stock of knowledge is a heterogeneous mix of recipes for effective conduct is not opposed to a proper scientific attitude. The proper scientific attitude is simply a more organised version of the natural attitude, but it is radically different to a science of uninvolved demands for semantic clarity and objectivity (Garfinkel's experiments show that both those demands destroy conversation and upset their contexts of inquiry). The indifference of a descriptive (ethnomethodological) programme is not a quietest support of the status quo, however, but a bid for a better description, preferred to the non-phenomenological regimes of doubt that are imposed in the name of the theorist realities that (ethnomethodologists say) inform conventional sociology and which are characterised by Lynch in the next chapter as follows:

Measurement for ethnomethodologists (Lynch) is not a resource for theorising `counts' of pre-selected referentialities but a topic to be understood as a way of producing social order in settings where "vulgar competences" are relied upon and taken for granted "whenever measurements are produced and assessed" (Lynch, p.79). The working equation between "fundamental structure" and "mathematical form" is revealed by ethnomethodology as an "aggregate product of coders' ad hoc procedures" (Lynch, p.88), and so ethnomethodology turns away from foundationalist methodologies that give rise to "principled discussions of validity, reliability, rules of evidence, and decision criteria" (Lynch, p.88) and "the task becomes one of discovering whether and how distinctions between `precise' and `approximate' measures are locally relevanced in various practical activities" (Lynch, p.95). Measurements are social practices through which a society makes itself knowable and are not measures of any societal organisation per se: "A `study' (for it to be ethnomethodological) no longer requires that the analyst devise ways to take the measure of objects or mechanisms in the extant society; instead, it becomes a matter of perspicuously exhibiting the society's hold on the very measures that take account of it" (Lynch, p.105).

Benson and Hughes follow this up with an ethnomethodological deconstruction of variable analysis. Variable analysis, they say, is only possible if researchers and respondents use "unexplicated features of the case/person/event in arriving at some classification for the purposes at hand" (p.119). And when approached through variable analysis people are summarily described in terms that the research format (its methods, procedures, and policies) dictate, and the format is a function of researchers' natural theorizing, not of the phenomena that can only have meaning through members' accomplishments. Ethnomethodology's rigour, by contrast, lies in its election to discover how "inferring", "providing evidence for", and "describing", are features of members' methods: "Instead of orienting to a collection of cases with an eye to producing abstracted generalisations from which the features of the individual case cannot be recovered rigorously, the ethnomethodological objective is to generate formal descriptions of social actions which preserve and display the features of the machinery which produced them. This requirement of `unique adequacy' stipulates the aim of describing in detail members' competences in producing everyday social action" (Benson and Hughes, p.131).

Ethnomethodology, instance by instance, is then - after all - `trans-situational', through revealing features of their local cultures that members orient to in producing their actions. Emphatically, the methods persons use to make sense can only be visibilised by fine description of detail. There is no need to pile up repetitive instances for the sake of `scientifically' warrantable generalisations. Further instances are just further examples. The material itself, in each instance, discloses the structure of social action.

This being so, ethnomethodology neither produces `findings' nor has need of conceptions of the `social actor' and `social structure' (Sharrock and Button). Social settings are self-organising and "the problem of social order . . . (is) . . . completely internal to those sites (of everyday activity)" (Sharrock and Button, p.141). In the manner of Husserl and late Wittgenstein, ethnomethodology attempts to become pre-theoretical, accepting that social order is "built in" (Sacks) to perception's appeals to generality in the identification of actions: "Social actions are irreducibly events-in-a-social-order and they cannot therefore be adequately identified independently of the social order in which they are embedded" (Sharrock and Button, p.158). Description by the ethnomethodologist depends upon his/her interactional competence, not on an external method of observation:

Ethnomethodology is thus a sociology that reiterates commonsense understandings more perspicuously in "a rather different perspective" (Sharrock and Button, p.167). For instance, in the analysis of a tape transcript the aim is to show the relationships between the `thin' captive record and the competence of members to unreflectively find in it a thickly describable social action. And that competence is not a matter of cognition in any computational sense (Coulter). It is "shown by what one can do, in how one behaves . . . in the public satisfaction of circumstantially relevant criteria, not by indicating anything like an internal process" (Coulter, p.185). Cognition phenomena are praxiologically appreciated:

Traditionally, analysts have been led away from this formulation by the big concepts of `language' and `culture', and have tried to find each concept in the other one (Lee). Sacks, however, makes a fresh start by "discovering and describing phenomena that are subsumed under the headings of `language' and `culture'" (Lee, p.197). Sacks' "initial strategy is to invoke the language's own categories for the description of utterances, not to develop an analytic collection of them" (Lee, p.203).

The result is not a set of rules for conversational organization but a list of possible relevances to take into account, developed from a "single event sociology" (Lee, p.217) which holds that people do not learn rules when they learn cultures but instead discover organisations of order when they participate in unique events. Accordingly, the organisation of culture is not an abstract system but "a recoverable, reproducible stock of knowledge and skills available in daily, routine, mundane ways of talking and acting" (Lee, p.225). Discrete realms of being (frames), then, are constituted by moral appraisal in practice, since they are normatively constructed to display (l) facticity, (2) a practical intelligibility of moral standards, and (3) the interactional logic of moral ascriptions (Jayyusi): "(e)thnomethodology reconstitutes them (value/enquiry, value/conduct, is/ought, relativism/objectivity, universalisability/specificity) as topics for sustained analytic elucidation, by which their logic-in-use can be uncovered" (Jayyusi, p.235). Moral reasoning is practically organised and practical reason is morally organized because: "Intelligibility is constituted in practico-moral terms" (Jayyusi, p.241).

This is why, in CA, the generalization that questions expect answers is not "a floating generalisation" but "a matter of moral and practical consequence" (Jayyusi, p.243). This is why the analyst uses his/her moral membership as a resource even while turning it into a topic for further elucidation. But although analysis is indifferent (refusing to collude with some descriptions over/against others), it is also moral, Jayyusi says, in that the analyst "is involved, minimally and unavoidably in laying bare the moral significance of . . . practices as these are made available in our culture" (Jayyusi, p.249).

I shall explain elswhere why I think ethnomethodology and CA have too limited an understanding of their own positions to absorb mainstream sociology, but despite its not attempting (a) to break out of or interrogate the restrictive language game metaphor of Wittgenstein or (b) to examine its own supposition that it is possible to see conduct (as something to be inspected) without any ironic perceptivity whatsoever, or (c) to see the consequences (the impossibility of Schutz's scientific attitude) of Sacks' idea that perception is cultural-theoretical, Ethnomethodology and the Human Sciences is a major contribution to the pioneer, pre-postmodern ethnomethodological programme of Garfinkel and Sacks.

Conversation Analysis in the hands of theorists who do not have the range, insight, and imagination of Garfinkel and Sacks is a shadow of ethnomethodology that is very slowly progressing as an empirical sub-discipline whilst ignoring parallel studies in linguistics, socio-linguistics, kinesics, micro-analysis, ethology, symbolic interactionism, and the massive heterogeneous corpus that goes under the heading of discourse analysis. Hardly any individual papers of CA have become landmarks of behavioural elucidation, and most are worryingly dull though the scholarly accomplishment is nearly always first-rate. An oppressive absence of panoramic vision is the CA hallmark even within its chosen phenomenon. Yet CA writing behaves as though it alone can understand conversation (as a structure) and by fiat it rules out any assumption that there is more to talk than the sequences of the turns of spoken words (such as people's feelings, sensations, perceptions, thoughts, unspoken agreements, sensitivities, and so on). For all its good intentions of scrupulous inquiry CA makes one feel in paper after paper that what is of interest in each paper could be said in about half a page before being elaborated upon in other sociological discourses, and the papers of Boden and Zimmerman's Talk and Social Structure are no exceptions in their catatonic terror of committing the error of talking person-to-person (except for Hopper, whose prose has pace).

Boden and Zimmerman's collection of CA studies starts with four papers that in different ways state the terms for developing ideas of social structure from tape transcripts, and these are followed by seven empirical papers that succeed in gleaning from transcribed interactional speech certain features that build a recognizable context of social structure. (The reader will appreciate that the move from talk to structure is very ethnomethodological in inspiration, and aims to have no truck with any kind of analysis that travels in the opposite direction from an alleged structure to talk heard structurally rather than naturalistically).

Surprisingly, the four preliminary theoretical papers reveal a split between a CA that restricts itself entirely to the description of sequential talk and a CA that accommodates other material. Zimmerman and Boden (Ch.l) and Schegloff (Ch.3) produce the classic CA position, stated with regard to context by Schegloff like this: "When a formulation of the context is proposed, it is ipso facto taken to be somehow relevant and consequential for what occurs in the context. It is the analyst's responsibility either to deliver analytic specifics of that consequentiality or to abjure that characterization of the context"(p.53). Wilson (Ch.2), however, says that there is more to context than can be extracted from sequence. Members orient to context, and this can be seen as a default option recognisable in non-sequential terms:

Mehan (Ch.4) then goes one step further than this. He says that CA ignores a relevant semantics and pragmatics of speech:

This is then demonstrated by the analysis of some institutional sorting practices that, says Mehan, would not be as visible as they are had there been no ethnographic fieldwork with which to contextualise them.

So it would appear that CA has a tendency to divide into those who wish to develop it as an ethnomethodological programme (but without the phenomenological language and without any consideration of a final report's discursive reflexivity) and those who are prepared to employ CA on behalf of fairly pragmatic programmes that latently draw from normal sociological desiderata. This division is further evident in the empirical papers.

The purist chapters are by Heritage and Greatbatch, Maynard, Psathas, Hopper, and Button. The findings in these chapters are just that, findings qua knowledge, to do with conversation primarily, and completely uninterested in sharpening interactional perception, as though sociology is a project that is not about viewing social relations as clearly as possible but is a project that by the shortest possible route through reality ends in the library. Heritage and Greatbatch (Ch.5) analyse the unique turn-taking practices of news interviews as a "core framework"(p.131) that participants actively and collaboratively create each time they do a news interview. Maynard (Ch.7) discovers a rather rigid "perspective-display" series when clinicians try to set up the conditions for eliciting agreement to diagnostic news. Especially when the presumption of the display (that a child has a problem) is not honoured, the talk converges towards an alignment to the clinic's expertise. The mechanics of this suggest a preference for agreement in the structuring of a visible social solidarity, and do not rely on coercive labelling. Psathas (Ch.8) describes how (over the phone) a sequence of talk about how to get from where the speaker is to some place else depends on the route being constructed. The "context-sensitive" (p.214) structure of direction-giving/receiving has a context-free organization that is "recurrent, orderly, and patterned". Also studying phone conversations, Hopper (Ch.9) looks at the `call-waiting' sequences that occur when a subscriber and co-speaker hear a beep that tells both that a third caller is trying to get through. Evidently, the beep takes the same sort of priority in phone conversations that a telephone's ring takes in ordinary conversations. People break off to answer an unknown summons. So "hold-on" is an example of how "speakers may transform . . . interactive practices"(p.228) in adapting to new technologies. Finally, Button looks at conversations that generate "standing" relationship structures between participants, where these occur in the initiations of closings. A conversation between participants in a standing relation can only close if it can be constructed as one in a series, whereby the relationship is testified to, elaborated upon, and invoked (p.272). This suggests that relationships do not exist outside of their constant representation in sequenced chains, and that it would be sociologically pointless, for instance, to ask a person to access his or her memories of another person (mother, daughter, boss, subordinate, etc) in order to learn about the relationship between them, which I think is a very closed position to lock oneself into on a principle.

The two papers that do not properly fit in with CA's programme show respectively a difference between American and Dutch telephone openings (Houtkoop-Steenstra, Ch.10) and an evaluation of how asymmetry between doctors and patients has been approached (Have, Ch.6). In both, the focus is primarily on the macro distinctions (American v Dutch, doctor v patient) though the means to establishing the distinctions is CA.

In Ethnomethodology and the Human Sciences there is the feeling that interactants are ethnomethodological in a fuller sense than the conversationalists of Talk and Social Structure. By this I mean that the move from a general ethnomethodological argument across a wide range of major social scientific topics to an empirical ethnomethodological exploration of conversations seems to diminish members, now only allowed to be analysable when their experience is programmatically reduced, for the most part, to their experience of conversational sequences. The serious moral dimension that may emerge when sequences are confounded or violated (in the eyes of the interactants that is), with attendant emotions of distress and bewilderment (as in Garfinkel's breaching experiments), seems to evaporate from the studies in Boden and Zimmerman, as if it could be perfectly possible to produce a total sociology without any orientation to such. Of this breach of the hegemony of sequence, Goffman is still perhaps the major theoretical key. In Goffman, there is a full account of what happens when things (so far as the interactants are concerned) go wrong and of how things are corrected, and it is an account that does not hesitate to point out the consequences for offenders in language that is not prissy about calling an institution an institution if that is what something has to be called to get some vital interactional meaning across. In other words, Goffman's interactions appear to the reader to be very much more interactive than those of CA and ethnomethodology. This is a difficult idea to convey satisfactorily since it depends both on the reader's own understanding of his or her strongest interactional experiences and on the reader's appreciation of Goffman's attempts to formulate these sociologically in analyses whose descriptiveness is often directly theoretical and ironic and analytic, though couched in vernacular terms (linking member, reader, and analyst) which defeat the ethnomethodologist claim that the work is a series of imposed formats on interactional phenomena.

Burns' Erving Goffman is a firmly written and stringent exegesis of Goffman's thought organized in an intelligent way that makes the whole oeuvre greater than its parts. Some of the parts are missing, however, and, strangely, not even referred to. The reader picking up Burns for an introduction to Goffman would not find out about the existence of Symbols of Class Status (1951), Communication Conduct in an Island Community (1953), The Arrangement Between the Sexes (1977), or Felicity's Condition (1983), which together amount to a book-and-a-half or even two short books. The omission of any discussion of the important Shetland PhD is very odd, especially as the research was carried on while Goffman was at Edinburgh contemporaneously with Burns. Maybe Burns is saving himself for a separate treatment of the Edinburgh Goffman (where somebody must have liked Goffman enough to get the first edition of The Presentation of Self in Everyday Life published by the Edinburgh University Press in 1956)? On the strength of the present performance I for one certainly very much hope that Burns will write about his relationship to Goffman. We need to know more about Goffman's flight from Chicago and the birth of a sociological artist on the outermost fringe of a rationed, fledgling welfare state that was just about to send its conscripted troops half way round the world to die in the Korean War.

Burns does not spread out enough to connect Goffman to any sociology save that of Durkheim, which means that a whole dimension of Goffman's counter-dependent thinking is lost (the parallel track to ethnomethodology, the repudiation of Mead, the positive distaste for role theory and for grand theories, the recoil from deconstruction, the meticulous unravelling of certain serious flaws in the Conversation Analysis programme, the sidestep from literature, the complex relations with his PhD graduates, and the sans fait rien shrugging off of Marxism, political theory, methodology, quantitative methods, and any of the available schools and cults). As a result Goffman's obviously ambivalent attitude to career sociologising is naturalised and removed from the motivational scheme. Nor does Goffman's peculiarly distortive radicalisation of the sociological perspective - on behalf of sociology as a vocation - surface as an ubiquitous narrative motive intertwined with a Garbo-like disgust for limelight that is weirdly coupled to a nearly self-destructing respectability-drive. All those matters await a more ruthless biographer than Burns (not that Burns sets himself up as a biographer - which is a mistake in my view) and one who has the same belief as Goffman that an unreserved dedication to sociological analysis is the only way to stay sane in an insane world.

The Goffman who does arise out of Burns' eighteenth-century Scottish empiricist philosophy is a Goffman whom it is hard to imagine as actually being of direct Ukrainian descent (he was born in Calgary, Canada), as having a west-coast American diction, as affecting to play it cool whilst genuinely being jazz-loving, as growing up in the consumer-led forties just before the invention of teenagers, as living out the latter part of his life in the upper, mandarin tiers of the American academy. One just does not hear Goffman's narrative voice coming through the Burns exegesis. And without that voice you do not have the `attitude' that is nearly always an indispensable ironic component of the meanings that Goffman intends, usually with exquisite prejudice, to deliver terminatively in a manner half way between Chandler's Philip Marlow and Lord Chesterfield. The fatalism of Goffman's essentially spiritual and bleakly hopeless bewailing of modern America does not seem to be heard by Burns, whose Scottish threnody is probably just as bleak, a bagpipe over an urban saxophone on the dark side of Hadrian's Wall.

After some contextualisation and preliminary charting of the Goffman oeuvre (a good idea to use Cooling Out the Mark for this purpose), Burns starts off in Chapter 3 with a discussion of Interaction Ritual, Strategic Interaction, and Encounters in order to establish Goffman as a sociologist of society-as-a-moral-order who flunks a major theoretical approach in favour of serial micro-studies of interaction systems. Chapter 4 works through Relations in Public essay by essay, and an imbalance is detected between Goffman's affirmation of social rules and norms and Goffman's recourse to a "sympathetic understanding"(p.104) of actors' meanings. Clearly Burns would prefer a less closed image of social order than he finds in Relations in Public. Chapter 5 is mostly about Presentation of Self in Everyday Life, which is grouped with Behavior in Public Places, Where the Action Is, and Role Distance. Burns reads these works as being internally stressed by their unfulfilled anticipations of Frame Analysis. After that comes Chapter 6 devoted to Asylums whose topic of madness is widened - favourably to Goffman - in a comparison between Goffman's and Foucault's disciplinary society. The theme of Foucault's "normalization" is then (Chapter 7) taken further as "abnormalisation", again from the background of Asylums but with inclusion of ideas from the famous and much admired Appendix to Relations in Public (Insanity of Place). Chapter 8 (Grading and Discrimination) elaborates the themes of Chapters 6 and 7 within Stigma and Gender Advertisements (I think Burns is exceptionally good on this), and concludes with some high level philosophy that is invoked to fault Goffman's buried definition of the "self":

The rest of the book is a good treatment of Frame Analysis and a resume of Goffman's work on talk. Burns makes no bones about Goffman's view that in a world of frames no frame is more real than any other (that reality is a function of involvement in something that is collaboratively going on). But here an opportunity is lost to explore Goffman's own authorial situation, its groundlessness in anything other than a fragmented and demoralized discipline that can only offer him an empty membership.

It is from the point of view of Goffman's institutional alienation, however, that one can most easily see the profound limitations of the ethnomethodological and CA programmes. From a permanent analytic framelessness, one may immediately note that the EM-CA claims that all interaction is speech, that EM-CA is not ironic, that EM-CA does not rely on members' knowledge, that there is always already order in every interaction, that analysis may not proceed through catechesis and misprision, that interaction can only be properly perceived through a method, that the goal of analysis is mere description, that frame and footing changes can be accounted in terms of conversational sequence are the claims of a position grounded in an idealistic projection of only one possible interaction reality for study, a projection that has no provision for understanding its dependence on its own particular rhetoric. Obviously all these points will have to be unpacked elsewhere. Meanwhile I would say that if a sociologist is interested in understanding face-to-face interaction he or she might first try to discover what he or she thinks about the self of Erving Goffman and the poverty of ethnomethodology in a discipline where studies of interaction are the only studies of society in action.

Copyright Andrew Travers 1997

back HOME