Semiotic Investigations: Towards an Effective Semiotics
Alec McHoul
Part Two
From Formalism and Ethnomethodology to Ethics
Chapter 6
Reflexivity, Problems and Solutions
I want to begin Part Two of the book in a way which seems at first sight to be almost a contradiction of the way I began Part One. There I argued against semiotic formalism on the grounds that it de-effectivizes semiosis, removes it from what Wittgenstein called "the stream of life" and attempts, instead, to totalize the diverse processes of semiosis by finding a general rubric for their operation - on the model, to some extent, of the ways in which linguistics attempts to find a general grammar behind the extremely diverse uses of language.1 Now I want to ask whether it is possible, after all this, to construct a formalism for effective semiotics itself: one which will account for the diversities and differences of principles upon which actual sign-using communities practice semiosis in all its variety. But this formalism is not for its own sake. On the contrary, it acts as a summary of what we have found so far. It is nothing but an assembled reminder of a set of positions already adopted. And, in addition, this formalism is designed to account for and introduce a further analytic concept, Garfinkel's concept of reflexivity, upon which the arguments and analyses of Part Two partly depend. Why reflexivity? What is its position in developing an effective semiotics?
In his Studies in Ethnomethodology, Garfinkel rarely refers to social practices in terms of signs - and he never considers questions of semiosis. Still less does he see ethnomethodology as in any way related to semiotics. However, in one crucial passage, he does refer to the stultifying effects of thinking about communication (or "usage") in terms of signs and referents:
Although it may at first appear strange to do so, suppose we drop the assumption that in order to describe usage as a feature of a community of understandings we must at the outset know what the substantive common understandings consist of. With it, drop the assumption's accompanying theory of signs, according to which a "sign" and a "referent" are respectively properties of something said and something talked about, and which in this fashion proposes sign and referent to be related as corresponding contents. By dropping such a theory of signs we drop as well, thereby, the possibility that an invoked shared agreement on substantive matters explains a usage.2
Instead of such a semiotic position, Garfinkel proposes that what he calls "usage" can in fact be explained in terms of community members' knowledge(s) of how signs are used. They do not share agreements "on substantive matters" but rather on "methodological" ones: hence the term "ethnomethodology." Communities, on this account, are assemblages of common ways of getting things done or (to use a terminology from chapter 3) of making things happen. Garfinkel's crucial theoretical insight involves, then, the substitution of analyzing how something means for the (more traditional) notion of what it means.
Perhaps the most central of Garfinkel's concepts for describing this essential effectivity of semiosis (to use a term which continues to distort his own position) is reflexivity. According to this principle, how a sign (or for Garfinkel, "usage") means is always reflexively related to the type of situation in which it is used. The sign and the situation-type mutually elaborate each other. For example, if I say "Open your books to page sixty," this may make sense by virtue of its being said in (and, very importantly, as part of) a classroom lesson and such that the one saying it is audibly the teacher in this situation. In its turn, the fact that this is a classroom lesson is (in part) a sensible and mandatory understanding of the situation for all concerned by virtue of the fact that someone (audible as a teacher) can say, in it, such things as "Open your books to page sixty." Lessons are made out of such utterances - that is what they are; and that they are lessons provides the grounds of understanding for the utterances in them. Upon such mutually reflexive (or "incarnate") practices does the very possibility of society depend, in Garfinkel's social theory. Society is irremediably practical; it is constructed from (and as) practical activity and practical activity is irremediably reflexive. In acting in particular ways, that is, members not only rely on their community's methods for doing so, they also display that it is these methods (and not others) that they are acting upon. Social practices are meaningful not because of their "contents" but because they display the grounds of their methodic production and understanding. To "understand" a sign, then, is to be able to see in it (as an activity) what its productional methods must be.
We can begin the elaboration of our effective-semiotic formalism, then, by abstracting from Garfinkel's position and adapting it for our own ends: every sign, S, is part of an activity, a. But, after Garfinkel, it is more than just "part of" an activity. Anything which might be construed as part of a sign and which is not relevant to its deployment in an activity (though it is hard to imagine what this might actually be) is trivial. Or at least, it is irrelevant to effective semiotics. Hence the term "sign" can almost be replaced by the term "activity." "Almost," but not quite, because of the following caveat: since a community's activities are methodic - that is, since they are not just any activities ("behaviors") but must display their intelligibility as part of a specific form of life - we can say that a sign is a methodic activity. Its internal relations are relations between activities and the methods, m, which communities use to produce them - to get them done and hence to understand them.
Activities are local (here-and-now, synchronic) accomplishments, even though the methods used to produce them may be deployable from locality to locality within a community (and perhaps even between some communities). Therefore they have an historical (there-and-then, diachronic) trajectory.
In being a methodic activity, ma, every sign is not "aimless"; it is a solution, s, to a socio-logical problem which a community has - albeit a routine problem, faced continually, day in, day out. A sign's external relations, R2, are the ways in which it acts as such a solution.3 Hence we can derive the following formulae:
(1) S =def ma "A sign is a methodic activity"
(2) m R1 a "Methods bear a relation to activities" or:
(2.1) R1 = m/a "A sign's internal relations are the relations between its constitutive methods and activities"
(3) ma R2 s "Methodic activities bear a relation to solutions to socio-logical problems" or:
(3.1) R2 = ma/s "A sign's external relations are the relations between its methodic activities and the ways its acts as a solution to socio-logical problems" or, since (1):
(3.2) S R2 s "A sign bears a relation to the ways it acts as a solution to socio-logical problems" or:
(3.3) R2 = S/s "A sign's external relations are the relations between the sign and the ways it acts as a solution to socio-logical problems"
(4) R3 = R1/R2 "There is a relation between the internal and external relations of a sign"
(5) R3 = fS "The relation between the internal and external relations of a sign is the meaning of the sign, fS"4
Nick Hartland has argued that R1 and R2 are "reflexive" in the sense elaborated by Garfinkel and Sacks.5 This is important for the field of ethnomethodology since it solves a long-standing problem in that domain of inquiry - a problem with some bearing on our current investigation. Traditionally, and as we have seen, ethnomethodologists (for example, Wieder) argued that there is a reflexive relation between utterances and the types of social events of which they form a part.6 By "reflexive" they meant (to repeat) that each is used to elaborate the other. So, in Wieder's example, when a convict answers a guard's question by saying "You know I won't snitch," the utterance derives its sense by being part of a convict-guard interaction in a halfway house (a type of detention center for drug offenders). That is, it relates to an elaborate "code" of conduct which convicts assume to be in place when guards are addressing them such that, in this case, it can be assumed that the guard's query is not "casual" or an inquiry into the well-being of the convict, for the convict's sake; rather the question is taken - retrospectively, after its production by the guard - to be a request or demand that the convict give the guard some information which may be detrimental to the convict or his fellow inmates. Hence: what the utterance means, here and now, depends upon the situational specifics of its being uttered. But the reverse is also true, according to Wieder's use of reflexivity. The sense of the convict-guard interaction is given by virtue of the fact that it is done precisely in these terms, with these words and others like them. It would not be a "convict-guard interaction" if it did not include such responses to inquiries - it would be another order of affairs entirely. Hence the type of event lets you know what the utterance means or does; and the type of utterance lets you know what the event is. The two are reflexive, precisely in Garfinkel's sense. And it is by virtue of this that we can say that everyday activities provide for and display their own "intelligibility."
This type of general reflexivity (which ethnomethodologists take to be an omnipresent feature of everyday activities) is an important discovery about everyday life in general. It shows that the meaning of everyday utterances is not guaranteed by general structural principles outside and beyond the events which they compose and which mutually compose them. But, as it were, once discovered, it leads to very little else than its multiple and chronic rediscovery in exactly the same form, everywhere.7 It becomes more like a principle of ethnomethodological inquiry than something which can generate analytic findings about the specificity of everyday utterances. In short: it can turn into the very thing which it sought to avoid: a general principle for guaranteeing meaning, albeit at the level of how meaning is achieved rather than of what signs mean. And, for this reason, empirical studies of reflexivity have been very low on the ethnomethodological agenda.8 Indeed, we must now suspect that the problem with Garfinkel's original conception of reflexivity is not that it relates utterances to activities but that it relates specific utterances to types of activities such that the idea of a "type" carries the same implications as grammatical or formal semiotic typologies.9
Hartland's refinement and reconstruction of the concept of reflexivity restores to it the possibility of its being used to find specifics (as well as "types" or general methods). Firstly, he returns to a basic ethnomethodological premise (and one which is entailed by the principle of reflexivity): namely, that the activities that persons can be observed to engage in in everyday life are methodic; they are "produced" by methods (including typifications but not reducible to them alone) which are specific to the occasions of the production and understanding of those activities. The problem with this version of things, however, is that - on the face of it - it makes the notion of "social order" seem redundant. For if methods were utterly unique to occasions then, as it were, the social order would have to be made up all over again, time and time again, from occasion to occasion. Each event would be singular, constituted by a radical difference from all other events. In early ethnomethodology, the Schutzian concept of typification solved this problem to some extent. But as a better attempt at a solution, stemming from about 1970, Garfinkel and Sacks began to promote the idea of "transferability." According to this principle, the transfer of methods from occasion to occasion is, itself, every social member's occasioned accomplishment.10 Hence this (as it were) meta-accomplishment guarantees social order. But this requires a further reliance on the famous Sacksian assertion (or perhaps "paradox"): that members' methods are both context-sensitive and context-independent. Hartland solves this problem by saying, simply and brilliantly, that R1 itself is reflexive. That is: methods are relatively stable from occasion to occasion. They constitute a kind of bedrock of social and local-historical order. But the activities they generate are occasion-specific.
As he shows, in his study of magistrates' courts, there are generally available methods for giving descriptions of persons (here: magistrates giving descriptions of accused persons), but the actual descriptions they produce are highly unique to these specific describers, described persons and circumstances of description.11 However, the methods reflexively depend in a crucial sense on the descriptions they generate - and vice versa. In actual practice, they are always methods for doing specifically this (and not some other) kind of description when viewed synchronically. However, they have the diachronic property of being able to be used elsewhere and at other times.
We can illustrate this by looking at two of Hartland's examples. In both cases, the magistrates find there to be a degree of control which their respective defendants should have had with respect to driving along the highway. There is a methodic principle (as Hartland shows) that magistrates should find the areas and fields over which defendants ought properly to have control but over which, in the case of the "crime" in question, they did not happen to exercise control. This lack, effectively, is the crime - even when the particular lack in question does not involve a breach of some specific legal code (though it can involve such a breach in some cases). However, in Hartland's first example, it is the driver's lack of control of his own emotional relations to other drivers that is in question - he has run into the back of another vehicle because of his frustration at the slow speed of that vehicle. In the second example, it is the driver's lack of control of his own vehicle that is in question - he has not allowed for poor road conditions and has "accidentally" run into another car. The words uttered by the magistrates in each case are almost identical: they amount to the fact that the driver should have, and could have, exercised more "control." They depend on exactly the same methods for giving descriptions of felonious drivers. But each description fits the situational specifics of the case in its own unique way - there is a crucial difference between the two types of control, as (described) social objects. Hence broadly-based methods are reflexively bound to unique activities (in this case descriptions).
This is Hartland's account of R1 as a reflexive relation. However, he then goes on to argue that there is another type of reflexivity: that between methodic activities (in this case, the produced descriptions) and "socio-logical problems." Socio-logical problems are not to be thought of as troubles: or as the sorts of things which perplex people and create anxiety. Instead, they are the most ordinary things in the world and stem from Sacks's methodological heuristic. Sacks argued that it can be sociologically productive to view social practices as solutions to problems.12 In several of Hartland's analyzed cases - and again I truncate his detailed analyses - there appear to be no good reasons as to why magistrates should not make prison sentences part of all of their dispositions: they have to find reasons for not doing so; that is, for giving fines, suspending drivers' licenses and so on rather than handing out prison sentences once someone is found guilty (that is, preponderantly, lacking in control). And, as Hartland shows in an overwhelming number of cases, the magistrates find that prison is "relevant" to defendants (where this finding of prison to be relevant is itself a case-specific methodic accomplishment). And the key to this resides in a locally relevant assessment of the degree and type of "control" which the defendant is deemed to have had (but did not exercise). Hence the methods and activities of magistrates (for example, methods of describing and actual descriptions) furnish solutions to this specific socio-logical problem: why not order a prison sentence? And the fact that they act as solutions furnishes the sense of their descriptions. The two are reflexively related (R2).
Another example of a socio-logical problem and its solution (analyzed empirically in the next investigation) is: how is it that people can become sexually competent in societies where that competence is expectably practiced in private and to the exclusion of non-competent observers? In chapter 7, I suggest that there are secondary systems of elaboration which accomplish this: from voyeurism to dirty jokes - a whole range of ways of finding that events can be "sexualized" such that pre-competent sexual practitioners can "check" their knowledges as a community, one against the other.
Taking just one instance: a piece of sexual graffiti can be understood insofar as it can act as a solution to this routine problem; and the problem is constituted as a problem (at least in part) because the graffito (an activity for which there are specific methods of accomplishment which provide for its intelligibility as a "competent" instance of sexual graffiti) exists as a solution to it. Each provides for the sense of the other. Otherwise: we would be hard-pressed to understand the function of, say, genitalia drawn on a toilet wall and, if we were unable to find such things (in fact, quite a large range of things) then we might assume that there were no such problems with sexual competence. This is not to deny that drawings of genitalia cannot be used in other ways - even when inscribed on toilet walls; and, in turn, it is not to deny that they are the only solutions to the problem. But they can be used in such a way - that is, when they are so used. And this last phrase does not involve a tautology: that they are (sometimes) so used is documentable with respect to their users' own formulations of them as specifically those kinds of objects.
One interesting feature of Hartland's second level of reflexivity, then, is that it begins to approach the phenomena encountered in more "macroscopic" social and political investigations than is usually the case in ethnomethodology. The term "macroscopic" should be shown here under erasure: for it can easily buy us into an argument whose very terms may - as a consequence of work in critical ethnomethodology - be spurious. That is, it is routine in non-ethnomethodological sociology to assume a rough division between micro and macro social orders and events.13 According to this division, the "micro" social domain consists of such things as language, face-to-face interaction, intersubjective meanings, social-psychological phenomena (such as intentions and motives) and so on. By contrast, the "macro" domain supposedly consists of such entities as patriarchy, class, the state, the economy, power, and so forth. I do not want to present arguments against this model of affairs: for there may be no arguments for or against it. Rather it appears to be a basic assumption of certain kinds of sociological theorizing - "naturally" basic to both "macro" and "micro" schools, whether in the name of asserting distinctiveness or assuring synthesis.
What I want to ask instead is whether it may be possible for an effective semiotics to avoid making this assumption in the first place: dissolving the problem rather than, firstly, inaugurating it and, only then, solving it. What Hartland's reconstruction of reflexivity accomplishes appears to be precisely this type of dissolution.
Firstly, it prompts us to examine everyday semioses as both producing and produced: producing the sense of socio-logical problems, and produced as solutions to them; such that anything that might count as a social object is not taken as fixed and simply waiting (like a "natural" object) to have its properties examined by a neutral observer. Rather, any social object is a problem-solution doublet, unfixed, in process, and always available to new and different creative treatments (lay and professional).
Secondly, it suggests that so called "micro" phenomena are not merely sociological flotsam and jetsam (or technically, "epiphenomena") whose apparent patterns are the "observable" effects of "deeper" social currents which, in themselves, remain mysteriously unavailable to the untrained eye. This mistake stems fundamentally from the micro/macro division itself: and against it, we are prompted to consider everyday activities as material practices which (as methods) constitute, and (as activities) are constituted by, the very stuff and doing of social order, disorder, structure, anarchy, or whatever it is that is being accomplished in and as a form of life.
Thirdly, it gives no special privilege to what are traditionally conceived of as "micro" phenomena; that is, it avoids having to make an argument - common in phenomenological sociology - about the "production" of "actual" social objects in "consciousness." Instead, it is able to consider (if it must) the domain of consciousness as a material one - not especially privileged - and existing as a relation between methods and activities.
Fourthly, it refuses to reify occasion-independent methods and occasion-specific activities into overall general social categories, pervading all of society (or all of a society). Insofar as it might wish to address questions of the state, of ideology, of power in general, of capitalism and so forth, it must do so tentatively, with the terms under erasure, as standing for massive blocs of socio-logical problems and solutions whose connections and relations are yet to be demonstrated. These "objects" of "scientific" sociology can be seen, that is, as solutions for problems posed within, and specifically within, the form of life of "scientific" sociology. They need to be treated as such, in the first place, rather than as definite and fixed objects which must inform social theory from the outset.
In the next investigation, I want to introduce a social-theoretical base for effective semiotics and, as part of this, I want to investigate the domain of socio-logical problems and solutions. That base preserves the time-honored tripartite sociological distinction between theory, methodology and research. The theoretical part of the investigation offers some remarks on two quite theoretically distinct sociologies, those of Foucault and Garfinkel. By displaying some family resemblances between them, it aims to generate basic study policies for the exploration of everyday forms of semiosis. The methodological part works from those basic policies and deals with the specific socio-logical problem mentioned above: it asks what can be done when competent sexual conduct is formally required but when "training" is only made available indirectly, if at all. The analytic part of the investigation examines an empirical case of a possible solution to that problem. It deals with a case of lecturing in which notions of moral and immoral sexual comportment and talk are, respectively, recommended and discouraged "by the way," "secondarily," or "as an aside" to the main business of the lecture.
=> chapter 7
This page incorporates material from Garry Gillard's Freotopia website, that he started in 2014 and the contents of which he donated to Wikimedia Australia in 2024. The content was originally hosted at freotopia.org/people/alecmchoul/seminv/6.html, and has been edited since it was imported here (see page history). The donated data is also preserved in the Internet Archive's collection.