Semiotic Investigations: Towards an Effective Semiotics
Alec McHoul
Part One
The Limits and Media of Semiosis
Chapter 4
Signs and Not-signs
Why have the various forms of semiotics tended to neglect the problems and complexities I have proposed? A simple answer would be that most types of semiotics have been unashamedly ahistorical; they have concentrated on the so-called "synchronic" aspects of language and other "sign systems." In this formulation of the problem, the synchronic is opposed, in a binary way, with the diachronic (the study of signs in terms of their historical development). But surely, we could argue, this binary is at least problematic, if not false. In the case of any given, empirical, use of a sign, the synchronic is always and necessarily bisected by the diachronic. In terms of the idea of meaning as use, it makes no sense to carve up a particular use into two separate lines of development (syntactic and historical). It is only for the purpose of a highly theoreticist investigation that the difference can be made in the first place - not to mention the fact that, following the bisection, only one aspect will tend to be taken into account.
But there may be another answer to the question. This answer will involve a paradox - a paradox in terms of how I write what is to follow. That is, in order to show the dehistoricization involved in previous forms of semiotics, I will have to use their own terms and assumptions which, if I am right, will necessarily be dehistoricized ones. But this is a risk which must be taken. And it is one which displays my point: I cannot simply undo the history of semiotics itself, as a discipline. Its own signs are historically located, albeit in an imaginary space of the ahistorical. And this cannot simply be transgressed by an act of will: so that any semiotic critique of ahistoricism in semiotics must be an immanent or inhabitational critique.
My answer to the question of semiotics' neglect of historical differences and complexities has to do with its most basic assumption. Almost by definition, every type of semiotic investigation has begun with the following idea: for a sign to mean, it must bear a relation to something else. We might be tempted to think that this "something else," this thing-outside-the-sign, would be something like a context. Then this notion of context, it would seem, could easily slip into notions of history, socio-political complexity, local specificity and so on. But this has not been the case. The thing-outside-the-sign (which I will henceforth call the "not-sign") has, by and large, been thought of by semioticians as being an object or a referent. At least, this is the starting point for many semiotic theories, even if they later proceed to bracket questions of object and referent by concentrating on the internal structure(s) of the sign. So we could simply say that the mistake in semiotics to date has been the misconstrual of the not-sign as being an object or something very like an object: something autonomous and fixed, rather like the objects which preoccupy physical scientists.
But this would leave out of account some more recent developments in semiotics. For example, it would not apply to those semioticians who have argued that a sign, if it refers at all, can refer only to another sign. The idea begins with Peirce, perhaps - though he also believed that the highly relativistic consequences of this view (the not-sign as other-sign) could be solved by saying that the endless chain of sign-to-sign reference exists only in principle - while, in practice, what a sign is for, its temporary end (or "telos"), will effectively close off the chain. As it were, on the Peircian account: a sign's meaning may not be like an initial or antecedent cause; but it is something like a final cause, or a point where it happens, empirically, to close. The meaning does not make the sign happen; it does not propel it into existence from some imaginary root-point firmly fixed in ultimate reality. But it at least finishes its happening; it pulls it along, from an end-point (a function), to that end-point.
The consequences of this are quite shocking. Either (1) we maintain a simple semiotic view where the meaning of a sign is its object-like referent or (2) we think of the not-sign only arriving ultimately at the functional telos of the sign-to-sign chain or (3) we suppose that sign-to-sign reference goes on endlessly beyond any not-sign or (4), using a weak version of position (3), we say that we can never know where the chain may end, that it's potentially indefinite. None of these positions seems to me to be very satisfactory. Each seems to maintain at least a vestige of the not-sign being a kind of object. The argument is clear in case (1). In case (2) the teleological terminus at least suggests a kind of meeting between sign and object - especially if we allow the word "object" to mean "goal." In the third and fourth cases, the argument is much less clear. Here it seems as if these theorists could think of the not-sign as only ever being possibly an object and (being relativists) they have wanted to do away with all fixed objects - so they have tried to do away with all not-signs. And this, in a peculiar and perverse way, valorizes the idea of the not-sign as object. Whether one values it (case 1, and possibly case 2) or whether one rejects it (cases 3 and 4), the same underlying position holds: if there is a not-sign (though there may not be), then it's an object.
All I want to suggest is this: that we can keep the not-sign (to continue with the shorthand) but we must think of it radically, as something other than an object.1 In short: it's the idea of the object that blocks the path to a dynamic, historical and complex idea of the sign. Instead, I want to begin to think of the not-sign itself as a relation or connection - but a complex one, quite unlike the kinds of relation or connection to be found in Cultural Studies (for example, the connection between culture and economy) which we examined in chapter 3.
There is a reason for this. In the tradition of Saussure, especially as he is read by the early Barthes, the sign-object relation is introduced and quickly forgotten. The notion of object or referent is assumed to be in place and, quite quickly, a different kind of medi(t)ation is put in place - though this is much less true of Anglo-American semioticians such as Peirce or Ogden and Richards, influenced by the logicist tradition of Frege and Russell. According to the Saussurian position, the semiotician is supposed to forget the referent and turn instead to the internality of the sign. The sign is accordingly split into two parts: the signifier and the signified. The signifier is usually held to be the material vehicle of the sign: a sound or an imprint on paper for example. The signified is, by definition, not an object or referent. Rather, it is an idea: what the signifier calls to mind, its mental (rather than objective) equivalent. So this too is a forgetting of the domain "outside the sign" - at least in its initial state. Though we should be fair to Barthes in particular here, noting that his complexification of the sign - whereby it expands into meta-signs and connotations - at least allows for the possibility of a kind of "contextual" build-up. But no matter how far this goes, it goes on within the domain of the sign (or, because of Barthes' expansion, the sign-system) itself. Eventually, all questions of the relation between the sign and the not-sign become irrelevant. All of the contextual action, as it were, takes place on the side of the sign. Yet the referent remains as a necessary, if vestigial, part of the theory. It remains, as it were, repressed. And like all repressions, it threatens to return at any moment.
But within the larger picture that's emerging here, we can notice a certain possibility: that a sign has both internal and external relations, relations which seem never to arise alone, in their purity but which, instead, as we saw in the case of photography, infect one another in the space of the frame or parergon. This is the case even if we look beyond the particular split into signifier and signified (as the primary internal relation) or the split into sign and object (as the primary external relation). In order to explore this question, I simply want to focus on these two relations in the abstract for the present, without naming them or fixing them in a specific way. So for the sake of making a beginning only, let us say that: for anything to mean it must bear a relation to itself (R1) and also a relation to something else (R2).
Now there's a certain absurdity here - a breach of grammar even. For we can see how the second relation (R2) is a relation in a quite straightforward sense. But in the first one (R1), the word "relation" could only be used in an overtly metaphorical way. The phrase "a thing's relation to itself" is prima facie nonsensical, as Wittgenstein well knew since it made him laugh. But in another sense, we know that some things can have relations to themselves. Human beings are typical examples: we have self-images, self-esteem, self-worth, self-loathing, and so on.2 The idea of a relation to oneself appears to be a feature of the things which we call "conscious." So, within my metaphor, I would want to say that R1 marks the "consciousness" of the sign, the practice: its consciousness to (or of) itself.
Lest I be thought to be indulging in the grossest kind of metaphysics here, I want to clear this up. I am taking the term "consciousness" from its early use by Giambattista Vico. In his New Science, Vico separated "coscienza" from "scienza."3 "Coscienza" is a form of knowledge - it is knowledge of "il certo" (roughly: the certain). But this "certain," for Vico, is like our use of the term when we say "certain men," "certain houses," and so forth. It does not mean definite knowledge - more the idea of particularity: particular events, customs, laws, institutions. "Coscienza" is the domain of the observable: what we can see around us as rather specific things. Its opposite, "scienza" is also a form of knowledge: knowledge of what Vico calls "il vero," the true. This truth consists of universal and eternal principles. It is a Platonic domain of underlying sempiternal structures which pervade all times and places. "Scienza" claims to know ideality, that which is not historically specific - and is the sort of thing Wittgenstein refers to when (as we saw in chapter 3) he writes of "the craving for generality" and "the contemptuous attitude towards the particular case."4 The opposite is the case for "coscienza." So when I refer to a sign's relation to itself (R1), I want to refer to its here-and-now specificity: this sign in this place at this time doing this job, and so on. (This is what might be meant by the term "the materiality of consciousness.") To locate a sign's internal relation as deriving from coscienza, from its particular materiality, means that an effective semiotics can avoid the problem of wanting to find, in and as that relation, a sign's "primitive," "original," or "pure" meaning. That is, it retains for the sign a very definite empirical particularism which steers between the ideas of both singular immanent meanings and meanings-in-general.
So we can see at once that this "internal" relation of the sign is by no means a calculable or algorithmic one. It is not fixed and stable. It is therefore unlike the necessary structural relation of signifier to signified (that relation would be very strongly associated with "scienza" and "il vero," for Vico). In traditional semiotic terms, R1 is the point of intersection of the synchronic axis of the sign with its diachronic axis - the point that the specific sign occupies: what it does just here and now, as a specific practice of a specific community (synchrony), as opposed to there and then (diachrony). If a sign's internal relation were to be a point (if we could somehow conceive of it as a point) within an event, within "what happens" (to return to Nancy), that point itself would never be an infinitely small one; it would still occupy a space. No points are spaceless - all points point, or make points within and as part of a field. In short, R1 consists of everything which is specific to a particular sign as it is used (and such that it is never not used). This relation, R1, to take the simplest of cases would not be identical in the case of the sign, "!", as it is used just there and as it is used just here! And even this simplest of cases could produce, I suspect, a complex analysis.5
By contrast, the relation R2 would connect the specific sign to all and any possible not-signs. It might, therefore, connect one sign to another, since one sign may be another's not-sign. It might, depending on the circumstances, connect the first exclamation mark (above) with the second. One has to say "might" here, because specific circumstances and relevances which cannot be known in advance will always determine the relevant connections and relations - and "specific circumstances" must refer to what a community is using a sign for. Thus, even with two uses of (perhaps) "the same" sign, we can see that the relation R1 will necessarily produce differences while the relation R2 may (or may not, depending on the circumstances) account for the relation between those differences - or else it might, for example, constitute a relation between one of the signs and some totally different not-sign. Hence the not-sign is always and necessarily what is not just-and-specifically-this-sign. Hence to restate the case: it can be another sign, though it is not necessarily so. R2 will be formed wherever and whenever one sign connects with something else in whatever way it happens to do so by virtue of the practices of a specific community of sign-users. The not-sign, to use Schutz's distinction, is always a concept of the second order: it is not available to analysis (and still less, for "theory" to decide as a general matter) prior to its construction, first of all, in and by a community's practical (that is, effective) activities.6 The point of an effective semiotics would be - among other things - a descriptive investigation of whatever a community might or might not use in this way. The not-sign, then, is always an empirical (rather than an in-principle) matter - and such that any empirical investigation may actually find it to be absent in any specific case.
So we can already see that a sharp division between "internal" and "external" relations is becoming difficult to sustain. R1 has simply to do with the fact that nothing is atomic, but always complex and therefore opens up the question of the relations between its parts. But these parts may not necessarily be contained within its borders. For as we have already seen in the case of the two exclamation marks, the "particular" differences which mark off one from the other are literally that: they mark off one from the other - they are not intrinsic properties in any ultimate sense. Hence both R1 and R2 are only stresses, tensions or modalities.
The first mistake in semiotics is to assume that either R1 or R2 will furnish "the meaning" of the sign. On a referentialist view, R2 is thought to provide the exclusive meaning of the sign. Vice versa for sign-internal theories in the tradition of Saussure and the early Barthes.
The second mistake in semiotics is to assume that either of these relations must have definite and universal structural forms. There is no reason to think that R2 must be a sign-object relation. It is simply whatever relation exists - as a matter of use - between just-this-sign and its relevant not-signs (including, again, other signs). Equally, there is no reason to think that R1 must consist of the relation between signifier and signified. It is whatever is specific to, and pertinently different from, just-this-sign and, hence, it cannot be totally divorced from the sign's R2.
So I would say - if I had to state my position in such formal terms - that the meaning of a sign is the relation (or exists in the possible space of the relation) between R1 and R2. Let us call this R3 - which should be recognizable now as the space of a sign's community-historical framing. And since neither R1 or R2 can be fixed or given in advance, neither is the meaning, R3. This meaning which we are calling R3 is something like the solidus (or any other mark, or space) which we must write when separating or connecting R1 and R2 to form a whole, albeit a split one: R1/R2. In this sense, being split and whole, being an unprespecifiable effect of two unprespecifiable relations, it is not unlike what Derrida, throughout his work, calls the "mise en abyme." However, according to the position I have outlined here: because we are never talking of signs in general (for example, we would not necessarily want to be discussing exclamation marks in general, just the two in question), the "abyme" can never reach infinity - for there is no effective infinity. Like other abysses, it has a bottom: though where this is can never be known in advance of falling - falling into communication, perhaps. The bottom is not necessarily Peirce's teleology: for the "function" of a sign in motion may be an effect of its motion. Neither is it necessarily an object: for an object is only one kind of not-this-sign (and, on some theories, not a particularly unified matter at all).
While all of this indeterminacy will be of little comfort to formalist semioticians, it appears to me to be the only way of bringing the radical idea of meaning-as-use into semiotics. If meaning is use: this does not mean that only R2 will be affected. Use is not merely a gloss for something external to whatever has meaning - a kind of separate context, outside the frame of the sign. It pervades the so-called "internal" and "external" relations of the sign (and hence the frame itself, the relation between those relations) - finally making their borders and boundaries indistinct. If the concept of "use" provides the idea of a frame, this is not an inert boundary which simply exists in order to make some kind of pure separation; it is a frame in which work can be done, (a) frame-work. Yes, we can say, such-and-such is definitely outside and beyond the sign in question: an innumerable amount of pertinent and impertinent "others" can then be pointed to. And yes, such-and-such is definitely a part of it: look for example at the bar across the top of that letter "T". But for a whole range of features, they exist on, in, and as, the frame-work. They do framing work. And how they do so is never going to depend on pure, analyzable, abstractable "properties" of signs in general - nor upon an utterly sign-less and fixed reality outside them. How they do so is going to depend on how they are worked in specific communities. This answers, perhaps, the theoretic questions of semiotics - too formally and neatly maybe - with a very firm indeterminacy. But it leaves all the work still to be done. A chaotic object of this kind is much more difficult to grasp than a linear one.
I began this investigation by taking a risk - employing a more-or-less standard semiotic vocabulary to try to unpick traditional semiotics itself. Another way of naming this risk would be to say that my "preservation" of any form of not-sign retains a too-strongly realist possibility; that not-signs seem to be stipulated by the very duality I attack; that they are ultimately noumena in disguise.7 What can be said to try to remove such lingering doubts - as much my own as anyone else's?
Like every dog, every sign has its day - or hour, minute or second. What it is very specifically not at that time may (or may not) be pertinent to how it is used, how it works, how it means. And here, "pertinent" must always mean "pertinent for a specific community." Then, after all of this, there is a tendency for theorists to take this possibility - and it is only a possibility, even locally (for there may pertinently be no not-sign in play) - and turn it into a universal principle, or a principle of universal availability. When this happens, we get an obsessive insistence on binarity; so that someone could say, "If x is a sign, y will be the (only) not-sign that (always) gives it its meaning." Then y is turned into a fixed entity, an object beyond doubt, the pure external presence which x properly names. It comes to occupy the space of "the primitive meaning, the original, and always sensory and material, figure."8 This is how it comes to look noumenal. Why do we do this - why crave this generality?
On the other hand, if any not-sign is just as local as "its" sign, this mistake cannot be made. We can no longer find a proper, permanent or singular not-sign for every sign. The double mistake in semiotics is precisely this: to think of all not-signs as proper, permanent or singular, or else to give up the whole business of them. It is this methodological binary (much more than that between sign and not-sign) that I object to. If we discover a community which clearly requires particular (locally constructed) not-signs as part and parcel of its use of signs, then it is going to be just as problematic to ignore this (perhaps by telling them they are wrong and that there can be no noumena) as it is to fix this particular and local usage into a universal scheme.
The space outside a sign is always there - for a sign never covers the entirety of the universe - and so some part of this space is always potentially part of its use, its framing. That potential may or may not be put into practice by a community: but if it is, a very particular not-sign is made pertinent to the case. But any particular item which is subsequently found in that space (and which is held to be relevant to it by some community of users) should not be assumed to be always and immutably its opposite and therefore its "meaning." It should certainly not be assumed to be its equivalent in the ultimately real world - the noumenon to the sign-as-phenomenon, for example, or its transcendental signified.
What we find as relevant to a particular sign depends on local circumstances. The local circumstances may (or indeed may not) include what we (as analysts or "members") see as absent from them. This is all that the concept of the not-sign proposes: that what is absent from, as well as what is present in, particular circumstances may be relevant to how a sign is used. No more than that, and with no a priori ideas as to what may constitute presence or absence in general.
Let me rephrase this in terms of a theoretical paradox. If there are not-signs, we (analysts) risk them becoming fixed, not-signs-in-general. They become the absence (or other potential presence) from which the supposed presence of the sign is guaranteed its meaning. They become, for example, noumenal. But if there are no not signs (a position which some theorists have adopted because of the problem just mentioned), the sign itself seems, as I have said, without limit. It becomes a plenum; it fills up the space of presence all on its own, existing without boundary, frame or limits. Neither of these positions is satisfactory in terms of an effective semiotics. Any theory which fixes not-signs into universals as well as any theory without any conception of not-signs will be incomplete. This is why it is essential to rethink the not-sign as (a) a movable and flexible absence (when it is, empirically, in place; when it happens to be made pertinent to a specific community practice) and (b) not necessarily in place at all on some occasions, even as an absence. The concept of the not-sign is the mark of the possibility, in practice, of an other-than-just-this-sign-here which is practically pertinent to it.9
The onus on us now is, therefore, to begin to depart from wondering (or wandering) about the general forms of semiosis and to return to a particular case. So in the next chapter, I want to examine the case of reading as a practical activity, as a means of localized semiosis (by explicit contrast with theoretical accounts of its essence or general form). What counts as reading on any specific occasion will, as we shall see, depend on (a) whether it is thought of as even having a "not-reading" and - if it is thought of this way - (b) which specific not-reading, from a broad range of candidates, happens to be in place or to be invoked as pertinent for quite practical purposes - for example, in this case, the purposes of teaching a class of schoolchildren to read.
=> chapter 5
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/4.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.