Semiotic Investigations: Towards an Effective Semiotics
Alec McHoul
Part Two
From Formalism and Ethnomethodology to Ethics
Chapter 9
Indexicality's Horizon of Possibility
In this investigation, I want to begin with our first semiotic level, namely intelligibility (R1). In the ethnomethodological literature which I have been drawing on as a basic social theory for semiotics, the meaning of a sign (or, more strictly, a "usage," "utterance" or "expression") is frequently confined to this level of intelligibility. This is achieved via a central ethnomethodological concept which is somewhat broader than the concept of reflexivity and also less easily definable or directly usable in semiotics, namely: indexicality. Indexicality is usually referred to as the context-dependence or context-sensitivity of any utterance.1 The thesis is that the "meaning" (that is, the intelligibility) of an utterance is dependent on, or sensitive to, its context where "context" is to be heard as the highly local or in situ embedding of a stretch of language. Here, the notion of context is such that it is immediate and audio-visually available, on the scene of the utterance itself.2 It is whatever is pertinently on hand for the interpretation of the utterance right here and now - for any particular "right here and now."3 In this restricted sense, the concept of indexicality would not be very important for our investigations since it appears to be confined to the specifics of everyday scenes and, as a concept within a theory of meaning, it appears to take "meaning" to mean "intelligibility" only.
But perhaps in another sense, the concept of indexicality could be useful to effective semiotics. That is: all three of the "levels" of sign-use which we have so far identified appear to have context-sensitive aspects. At the level of R1, "context-sensitivity" appears to mean exactly what ethnomethodologists have meant by "indexicality"; at the level of R2, "context-sensitivity" would mean the sensitivity of particular solutions to their particular community's socio-logical problems; at the level of R3, "context sensitivity" would mean "community-specific." It is, given the above, the last of these that would be most pertinent to our inquiries here. And so, I now want to offer a semiotic reconceptualization of indexicality; for if "indexicality" could mean "community-specificity," then it might (at the least) give us a key into this crucial semiotic concept. It might mean that we could begin to think of communities as collections of "what happens" (to return, again, to Nancy's definition) in the handling of indexicality.
In order to rethink indexicality in this way, it is clear that it has to be taken from being merely (and perhaps trivially) bound up in (and as) an essential constituent of empirically available social encounters and has, instead, to be thought of as a critical rather than just a descriptive concept. Hence, I want to propose a very slight shift in the ethnomethodological concept of indexicality, taking it towards "critique" - that is, towards actionability (R2) and historical meaning (R3). This would appear to be necessary if the concept is to have any value for effective semiotics.
By the term "critique" I want to indicate something like the premise of the irreducible différance of any sign as it (the premise, that is) is currently used in what might be called "critical theory" (CT).4 "Critique" would then mean any form of analysis which worked from this premise - one which can be alternatively expressed as a reaction against axiomatic or algorithmic theories of meaning (such as that of Wittgenstein's Tractatus) and which insists, against such theories, that any intelligible communication is achieved not by virtue of the intrinsic properties of signs (such that, for example, they map directly on to the world) but only by virtue of an essential semiosic instability of any sign in principle. I take this critical potential to be implicit in the concept of indexicality - even though, as I argue, the full potential of the concept has not yet been exploited in ethnomethodology (EM) itself.
I want to argue for a partial rethinking of the concept of indexicality from "within" EM and "between" EM and CT. It could be argued that the abiding effect of Garfinkel's Studies in Ethnomethodology has been, despite the insistence of some practitioners, its impact on social theory. That is, Garfinkel's theoretical solution to the problem of social order is utterly unique - to the point where there are probably now only two forms of social theory, namely social theory and EM. Garfinkel's radical hypothesis is twofold: (a) that the social order is underwritten discursively and (b) that discursive stability ("meaning" and/or "understanding") is achieved via practical actions upon expressions which are, in principle, indeterminate or indexical (rather than stable). Putting these two hypotheses together: EM is founded upon the radical insight that social homogeneity (in practice) is predicated upon discursive heterogeneity (in principle). Every sign's intelligibility is always the effect of muddling through, making do, with unstable socio-discursive conditions. In this sense, Garfinkel's insight partly overlaps (and pre-dates) the founding premise of CT whose locus classicus is Derrida's Of Grammatology.
The premise of the in-principle instability or indexicality of intelligibility, shared by EM and CT, is nevertheless "taken up" in distinct ways within each domain in terms of its conception of social-textual practice. CT takes instability (différance) to be either (a) inherently discoverable in otherwise seemingly definite and closed texts (such as those of Saussure or Rousseau) or (b) celebrated in the case of "naturally" indeterminate texts (such as the postmodern novel). In certain cases, the boundaries between CT and its "object" texts becomes blurred, so that CT enters self-consciously into social-textual practice.5 One of its principal aims is to explore and exploit any possible relation between civil debate and action, on the one hand, and theoretical debate and action on the other.6 This may mean that, in for example the literary domain, the distinction between a fictional work and a critical work becomes - is made - deliberately unclear.
EM, on the other hand, has attempted to show how discursive indeterminacy is worked, in practical actions, by members, in order that it may be "overcome" if not "repaired" in any ultimate sense - so that its "repair" is only ever "for the moment', "for all practical purposes" - thereby generating intelligible communications or "practical understandings." While no EMist has ever argued that EM itself is "outside" or beyond such fundamental methods of social accomplishment (hence leaving a continually possible, if rarely actualized, relation between civil debate and theoretic debate),7 there has nevertheless been, in EM, an "analytic" tradition (as against an interventional tradition) such that EM's aim is to describe social accomplishments in terms of members' methods for handling the open texture (indexicality) of linguistic expressions.
Hence there are two distinct directions in which one can proceed on the basis of the premise of indexicality/différance. The single thing that I want to argue for, at present, is that these two "directions" are not incommensurate, either in principle or in practice since, among other things, they both reject objectivist and formalist versions of how signs can "mean."
One upshot for EM would be that we could not decide a priori whether or not an indexical expression (or sign) is going to be, in practice, worked towards definiteness and stability or whether, on the other hand, it is going to be, equally in practice, worked towards amplifying its essential (its always already given) polyvalency. Indeed, it might be possible to see that this distinction between semiosic identity (discursive stability) and semiosic difference (discursive indeterminacy) does not constitute an "either/or" possibility for any given sign.
Instead, it might be discovered that there is an array of possibilities, across this continuum, with any actual expression's actional "uptake" in respect of the array being highly dependent on contextual particulars. John Heritage puts EM's position on members' actional uptakes of indexicality succinctly in terms of "making definite sense with indefinite resources."8 I want to say that this formulation, which is widespread but by no means universal in EM, makes the a priori assumption that "definite sense" is always what is made, empirically, in actual situations of social-moral choice; and that, instead, we should leave that question open prior to inspecting particular details of those arrangements. It is not impossible to conceive of (or indeed to discover) practical situations where quite indefinite senses are permissible or even required. Cases in point would be some (but not all) instances involving:
_ poetry
_ madness
_ bold hypothesizing (_ la Popper)
_ expressionist art
_ certain popular texts (Total Recall and Twin Peaks, to cite only two examples)
_ postmodern fiction (Calvino, DeLillo, Pynchon, Joyce and others)
_ quantum mechanical indeterminacies
_ psychoanalytic interpretation of certain kinds (such as free association)
_ surrealist "automatic writing"
_ various types of religious action and debate (where meanings are expected to remain specifically unclear)
_ children's games and fantasy play
_ musical improvization
_ dreaming
_ playing "Chinese Whispers"
_ punning
_ irony
_ making double-entendre jokes
_ playing with visual illusions (for fun or else seriously, as in the drawings of Martin Escher)
To this we could add, no doubt, a much larger range of quotidian events. As Wittgenstein puts it: "Is it ... always an advantage to replace an indistinct picture by a sharp one? Isn't the indistinct one often exactly what we need."9 Or in EM terms: some forms of life (law or classical mechanics, for example) may treat "reality disjunctures" as "puzzling events" requiring "solutions," while others may take them as unremarkable, as part of the routine business of life as usual.10
Another way of conceptualizing indexicality has been in terms of a gap or lacuna, a space to be filled in and by specific methods. (Mehan and Wood wrote of members' impulses to "fill in the rules.")11 My premise is no more and no less than this: that there are more things to do "with" a gap than simply to fill it, and that there are manifold practical situations of action where something other than a "filling" takes place. Rather than filling a gap, for example, one may attempt to widen it, alter its shape, leave it in place exactly as it is, create another gap of a different sort in its place or beside it, ignore it, take it for granted and so forth - a potentially endless variety of "types" of practice.
EM's initial theorization of indexicality envisaged a range of possible ways of "filling" gaps and went looking for them in practical, empirical instances. The limitations of this, to me, seem homologous with the limitations of logical and axiomatic theories of meaning. The later Wittgenstein showed these to be unduly restricted by arguing that there is not just one language game but many. He had in mind the propositional attitude which had served as a model for all expressions within the traditions of formal logic. I have in mind the gap-filling attitude (where expressions "make definite sense") which has served as a model for all practical actions within many - but not all - traditions of EM. No doubt gaps do get filled, fairly definite senses do get made for all practical purposes; but there seems to be no reason why this specific variety should be generalized to cover all cases of "gap work" or "sense-making." This is especially so since the metaphor of a gap implies that some sort of firm ground lies to either side - so that community methods need only be designed in order to move from one firm terrain to another via gaps. What such firm terrain could be remains untheorized and unexemplified empirically.
Instead of conceptualizing indexicality as a gap, one could, for example, conceptualize it as a point - a hypothetical, unrealizable (or ideal type) zero degree at which an expression is not yet "put into" social practice (pN). At this point, what we can call its "indexical potential" would be indefinite.
Along one path emerging from pN (where the path represented the putting of the expression into social practice - putting the expression back into the "stream of life," as Wittgenstein has it), the expression would move towards another hypothetical point (p0),12 a point of ultimate closure, where its indexical potential would be so narrowed and honed that a single and final meaning would henceforth be forever invoked along with (or as) that expression.13 Informationally, this point, p0, would represent the lowest possible "entropy" that a sign could have. The path of the sign could be thought of as converging towards p0, representing zero entropy. But the point could never be reached in practice simply by virtue of this paradox: namely that singular and final meanings would have zero information! Or more conventionally: no meaning without some divergence (difference).
This is how, for example, Husserl imagined the Pythagorean Theorem which "exists only once, no matter how often or even in what language it might be expressed. It is identically the same in the 'original language' of Euclid and in all 'translations'."14 By virtue of achieving this hypothetical point, Husserl argues, an expression "keeps its ideal identity throughout all cultural development. It is a condition that allows communication among the generations of investigators no matter how distant and assures the exactitude of translation and the purity of tradition."15
A second path - perhaps in the opposite direction - would carry the expression away from semiotic unity and towards another hypothetical point (p1),16 a place of ultimate dispersal, where its indexical potential would be increased to an ideally infinite degree. Informationally, this point, p1, would represent the highest possible "entropy" that a sign could have (the semiosic equivalent of thermal equilibrium). The path of the sign could be thought of as diverging towards p1, representing maximum entropy. But this point, again, could never be reached in practice by virtue of another paradox: namely that infinitely differentiated meanings would have infinite information - they would be able to say absolutely everything at once. Or more conventionally: no meaning without some convergence (identity).
So at point p0, the expression could mean only one thing; and at p1, it could mean anything at all. Neither of these points can, then, ever be reached empirically - the points represent "tendencies," "impulses," or "habits" towards unattainable extremes. Before an expression reaches either point, it comes up against practical limits. (Moreover, the expression's starting point (pN) is itself an unattainable "limit": for no expression is ever not located, always already, in the "stream of life.") However: I do want to argue for the "lines" or "paths" in this model (one of which is linear and semiosically convergent, the other of which is non-linear and semiosically divergent) - which I will call the paths of an expression's indexical potential.
It should already be clear, then, that I am not proposing just another binary: say, between a totally unlimited and a fully limited semiosis.17 In fact, my second "path" of indexical potential is not a single line at all. Rather it is an array of lines: representing multiple pathways.18 It is an impulse associated with trying to multiply and proliferate an expression's indexical potential. So: in one direction lies an impulse towards linearity and singularity; in the other lies no single "direction" as such at all, but rather an impulse towards "chaos" (in the sense of anti-linearity) and multiplicity.
The ultimate limits-in-principle (p0, p1) of these directions (or one direction and one multiple direction) cannot be practically achieved - perhaps they cannot even be imagined. The two tendencies towards them have limits in practice so that all intelligible signs will have a p-value between 0 and 1. And because they are limits in practice, they will vary with the in situ specifics of any empirical case. The abiding fruitfulness of EM has been its ability - with some exceptions - to locate, with analytic precision, only one (set of) these limits: the set produced by linear-singular methods - those designed to make definite sense. My question (which I am calling "critical" since it insists on the possibility of routine differences and contestivities in the domain of community-historical meanings) is: what does an investigation look like if it accepts the task of delineating both sets of limits instead of just one? That is, to extend the metaphor, unlike the physical universe which, most physicists hold, always tends in one direction (towards maximum entropy), the semiotic universe has no one tendency. There is no equivalent of the second law of thermodynamics in the semiotic universe.
A criticism of EM would be that it, for the most part anyway, conceives of the semiotic universe (society) as always tending towards low indexical potential despite its practical unachievability;19 and a critique of CT would be that it works the other way, imagining most signs to tend towards dispersion despite the attempts of, for example, philosophers and logicians to contain them.20 Hence, for all the insistence (in fact because of it) on difference in CT, it is just as singularizing as EM. In fact, on the physical metaphor, CT holds to a kind of semiotic equivalent of the second law (all signs can mean by virtue of their tendency to high entropy). The effective semiotics I am proposing would hold that movements towards the equivalent of the second law and towards its opposite are empirical community matters and cannot be deemed in advance theoretically or as "natural tendencies." This is another way of putting Schutz's dictum: that the "data" of the social sciences are always of the second order; for, unlike the data of the natural sciences, they are always already "interpreted" in advance by the communities under investigation.21
Even within particular communities (or sub-regions of the semiotic universe), there will be occasions when high-entropic sign-paths will occur and others when low-entropic sign-paths occur, even though there may be community "tendencies" - such that, for example, we would expect a community of Platonists, by and large, to favor low-entropic paths (minimizing difference and maximizing identity) and a community of relativists to favor high-entropic paths (maximizing difference and minimizing identity).22 In this case, EM and CT would themselves be respective examples of such communities.
Practical limits along the linear-singular axis (towards p0) will take the form of such things as "enough is enough for all practical purposes." The question here, for members, is, as it were, how to keep the path relatively singular and linear; how to prevent it, at any point, from breaking up into an array of multiple meanings - and how far the path should be extended in the direction it is travelling (in the extreme case: "as far as possible" or "until we get to the truth"). Practical limits in the other direction (towards p1) will take the form of "how to say more and different." The parallel question here, again for community members, is: how to proliferate lines of the path, how to invent new ones and how to extend currently existing ones as far as possible (again in the extreme case: "as far as possible" or "until truth cannot possibly be an issue for us"). While these can be called different forms of socio-logics, they could also be referred to, in their specificity, as textual technologies or reading/writing technologies. But it is clear that the indexical potential of signs (in principle) can - if we allow it to have (roughly) "closing" or "opening" directionalities23 - lead to a whole array of socio-logical problems and their solutions (in practice). It is in this sense that the concept of indexicality may be of use to effective semiotics. For if we assume that indexical potential can "go either way," we may be able to identify methodic actions of (again, roughly) "opening" and "closing" that potential. One classic study here is that of Latour.24 As he has shown in the case of scientific communities, some "closing" methods can solve the problem of, or at least (since extremes are unattainable) guard against, having a particular sign "mis-used" or "mis-interpreted" by another (perhaps a rival) community or by "revisionists" within one's own.25
If, in everyday practice, a sign's (in-principle) indexicality can be made practically "definite" (EM) or practically "indeterminate" (CT) - to invoke a cumbersome shorthand - then an analysis which empirically investigated this overall horizon of possibility would, in the sense intended here, be critical. It would be a critical semiotics, perhaps, for it would leave open the possibility that not all social/discursive practice actually achieves social order (let alone "consensus") but that some instances have such things as disorder, conflict, contradiction, struggle, antagonism (and so on) not merely as actional achievements but as part of their taken-for-granted background and foundation - that is, in terms of the reading/writing technologies (language-games) they routinely occur in and as.
How this would relate to questions of "politics" in the broader sense is an, as yet, open research question and one which cannot be easily answered at present. However, I would like to speculate that one could arrive at a critical definition of "the political" where that term meant no more and no less than the traditions which come, historically, to inhabit a form of life or community, as that community's routine ways of dealing with the practical limits of signs' indexical potentials.26 In this sense, the question of where and how to limit a sign's potential might be a socio-logical problem - moving the analyst already to the level of actionability (R2). Such limits (the limits of commentary) may even be in contention from community to community and the resultant "border disputes" appear to be marked by the principle which Lyotard calls "the différend" - a principle to be discussed in chapter 13 (below). For now we could speculate that the expression "ethno-politics" might be appropriate to such cases since it acts as a kind of mnemonic, continually referring the analyst to the location of politics in (and as) situated practices.
So, in terms of a practical ethno-political analysis, we might begin to ask how it is that communities of sign users could actually "work" a text's indexical particulars towards point p1: that is, expand its potential towards semiosic indefiniteness. Are there communities - pace ethnomethodology - which actually do this; which require the signs they find in their everyday lives to be "opened up" to multiple meanings or uses? I want to pursue this in the next chapter where I look at a cinephilic reading of a song and dance sequence from a well-known musical film, Singin' in the Rain.
=> chapter 10
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/9.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.