From a grammatical standpoint, operators are different from, but similar to, modifiers, traditional grammarians emphasizing the similarities, modern linguists, the differences. Negation is, in imperative as in mature language, the most extreme case because it is the most unambiguously transformational in character, requiring for its application a preexisting linguistic expression. In the imperative, negation (or more properly interdiction) requires for its formulation the designation of a substantial performance that could always in principle be the object of a prescriptive imperative.
Interdiction shares the asymmetrical structure of all imperatives; it does not negate the subordination of the addressee to the desire of the speaker, but only the performance that, formulated in the same words, the latter might in another context have requested. Thus interdiction is meta-representational, inverting the positive relationship between the signified performance and the desired outcome that obtains in the prescriptive form. But the factor that stops interdiction short of the context-free—in our terms, grammatical—functioning of negation in the declarative is its continued dependence on imperative awaiting, which limits its field of operation to the linguistic scene of speaker and hearer.
Thus non-performance remains always, even when it consists in inaction, a real-time fulfillment of this awaiting, just as, conversely, the performance of the forbidden action would constitute a violation of it. The substantial performance is refused, but its shadow haunts the nonperformance and by so doing prevents it from becoming a purely formal operation. In the last analysis, not-running in an interdictive context remains a kind of running because its satisfaction of the interdiction will be measured by its (non-)correspondence to the criteria that define running.
Declarative negation is quite different. If I say “John is not running,” my statement will be judged true or false by those same criteria, but John’s activity is not itself governed by them. Negation here is purely formal, metalinguistic. Whatever else John may be doing, I may if I like interpret his activity as being not-running, thereby liberating my imagination to grasp analogies with areas of experience not intended in John’s action.
Hence, like governance and modification, the operation of negation (and a fortiori the use of other operators such as “again,” “twice,” etc.) is not fully grammaticalized in imperative language. But because interdiction, unlike the other qualifiers of the imperative, involves the use—and hence the lexicalization—of an operator not independently realizable as a performance, it would appear to create a declarative-like distinction between “parts of speech” linked in a formal hierarchy. This would be fatal to our hypothesis, in which the full grammaticalization of the negative can occur only in the context of the declarative.
Because interdiction remains a request for performance, it is difficult to consider its “not” as a potentially substantive element defined outside the linguistic context, such as “fast” or “big,” which add observable qualities to a performance. Yet that this is indeed the case is suggested by our own usage of the expression Don’t! One might object that, like “No” in answer to an interrogative, Don’t! is merely an ellipsis for “Don’t do X!” But although this would be the case when I respond “Please don’t!” to “Should I close the door?” it need not be so.
The woman who exclaims “Don’t!” to a lover’s tentative caresses is a useful example. In claiming that this is not an ellipsis I am not merely playing with words. An ellipsis is the omission of an already present linguistic element, like “(Please) don’t” in reply to “Should I shut the door?” The woman’s “Don’t!” to her would-be lover is another case entirely. Although we might imagine various verbs to fill out the meaning of this expression (“Don’t touch me!”, “Don’t try to seduce me!”), the lack of a specific verb is significant. By failing to offer a linguistic model of her partner’s behavior, the speaker avoids any characterization of its intentionality and thereby forecloses any possible attempt to reinterpret his actions to her satisfaction. Her utterance alludes to a behavioral complex of seduction that “Don’t!” interdicts in toto, rather than “Don’t do that!” which focuses on the specific act performed. And this is only possible because in the imperative, negation is applied not to strictly linguistic models, as in the declarative, but to performances that, although they are indeed normally specified in language, may simply be exemplified in reality.
Thus the operator of interdiction, although specifiable lexically, retains sufficient substantiality to stand alone when its real-world referent is sufficiently evident. And this is not an accident of modern usage, but a necessary consequence of the intentional structure of the imperative, in which the interlocutors’ shared scene of representation is not fully divorced from their real temporal presence. Interdiction is not true negation because it addresses itself to the “will,” not to the “intellect,” to performance and not to a context-free model of performance. And by the same token, it is not wholly devoid of substantiality because it can assume as its content that of the present moment. Taken out of its temporal context, the lexeme of interdiction appears as a pure operator, but imperative language does not permit this purely lexical abstraction. There is no linguistic space available in which to ask the metalinguistic question of what “Don’t!” means because its utterance intends unavoidably its real-time context. Its evolution into a true negative, which is at the same time the genesis of the declarative, must therefore depend on its inappropriate use.
We have been assuming that the hearer of the imperative, knowing the requested object to be unavailable, requires a means of expressing this fact to the speaker. But we cannot consider that on the one hand he possesses the “thought” of the absence of the object, but on the other, he is unable to “express” it and must therefore invent a new form for this purpose. Rather, the “thought” itself is the invention, and if we can specify precisely what it should be, we will find it already expressed. For whether or not thought be deemed possible without language, the desire to express a thought to another, as is the case here, can be formulated only in the terms of whatever system of representation is available.
Thus the imperative did not arise when, desiring to command the presence of an object or action, someone decided to use the ostensive for this purpose in the absence of its referent. Rather, the “thought” of the desire for the object was simply expressed by the (ostensive) means at hand, without deliberation on the change in linguistic convention that it would entail. And the second party’s response, by correcting this “inappropriate” usage and at the same time satisfying the speaker’s desire, led to its reinforcement and to its eventual acceptance as an appropriate linguistic act. Similarly, in the present case the second speaker does not seek the means to express the absence of the object—for if he sought them, he would find them wanting—but expresses it simply as he has formulated it to himself. The model he creates may not be immediately comprehensible to the first speaker. But because he has no doubt been on occasion in the same situation as this interlocutor, he will eventually grasp its meaning.
The creation of new linguistic forms thus passes through a moment of subjectivity in which desire is expressed as faith. In the case of the nascent imperative, the object of this faith was the “ontological” power of the ostensive to compel the presence of its referent. In the case now at hand, the desire of the second speaker is to communicate to his interlocutor the impossibility of carrying out the imperative. For both interlocutors it is the absent object that is the focus of their attention: the first speaker desires its presence, the second knows that this presence cannot be obtained. This object is designated by a linguistic expression, a word or combination of words, which has already been employed by the first speaker as an imperative, and which, if the object were indeed presentable, could be employed anew by the second speaker as an ostensive, as in the “Scalpel!”-“Scalpel” dialogue.
But it is we who have classified these two usages as belonging to two linguistic forms, linked by a historical dialectic. For the speakers who make use of ostensive and imperative utterances in the process of communication, they do not possess, as they do for us, the discrete existence of elements in a paradigm. The forms represent for us two different intentional structures, but these structures cannot themselves be “known” to their early speakers, and are not normally perceived as such even by speakers of mature language. They are conventions of communication, the choice among which is not overtly made but predetermined by the situation in which the speaker finds himself. If, in ostensive language, the word designating the object could be said to “mean” or intend its presence, this was because the usage of the word always accompanied, and specifically designated, the object-as-present. In the significant memory of the members of the community, even at this stage, the word could be said to possess the same lexical signification that its counterparts possess today, that is, it is simply the name of the object. The “ontological faith” that we have attributed to the user of the ostensive is not the product of the signification of the lexeme, but of its meaning, which is nothing but the shared memory of its usage.
The speaker of imperative language possesses two utterance forms that are in effect conventional uses or meanings for the word, not significations of it. “Hammer” simply signifies (a) hammer; it is the usage of the word that is limited to an imperative or ostensive interpretation, and this not because the word could not be pronounced independently of either form, but because its utterance could be given no other meaning. In elementary language, one cannot simply “talk about” a hammer, because there is as yet nothing else to say about it than that it is present or that one wishes it present. To combine the words “big” and “hammer” in an utterance tells us, of course, that the hammer is big, but the meaning of the sentence at this stage is either that the big hammer is here, or that the speaker wishes it delivered. That it is big is not information conveyed by the utterance, because the function of utterances is not to convey context-free information but to designate significant or desired phenomena.
In the “Scalpel!-Scalpel” dialogue, the word designates a specific object of interest to the speakers (to the second through the mediation of the first). Both surgeon and assistant are capable of formulating sentences of indefinitely great complexity concerning the scalpel, but in the given situation such sentences would be inappropriate. Nor is it likely that the speakers would be aware that their dialogue consists of an imperative followed by an ostensive, even assuming that such terminology were familiar to them. Their conversation uses the word simply as a means of communication, the intentional structure created in each utterance being an appropriate model of reality: The surgeon says “Scalpel!” because he wants the scalpel, and his assistant repeats the word to show that it is now available for the surgeon’s use. In both cases the word simply signifies the object; its place in the intentional structure is in each case determined by the context.
We should suppose this to be the case as well for the speakers of imperative language. Thus for the second speaker in an “infelicitous” imperative dialogue, the word used by the first to demand the object is simply its name and not in itself either an imperative or an ostensive. And by the same token, his own (ostensive) reply, were the object in fact available, would repeat this name. Given that it is not, we may assume that he imagines the requested object/performance, but that he understands that the image is at the present moment inactualizable.
This situation bears a certain similarity to that of the original speaker of the imperative, who, imagining an object in its absence, called its name to make it appear. But this role has already been preempted by the first speaker whose ontological faith the second knows to be in the present case unjustified. It is this non-justification that constitutes the object of the latter’s desired communication. Calling the name of the object will not make it appear, whatever the sanctity of the scene of representation. The second speaker’s knowledge demystifies the faith of the first. This knowledge is the negative moment of the declarative, and of higher linguistic form in general. The use of the word must now be divorced from the presence of the object. But from the standpoint of the first speaker, one still implies the other; it is not his usage that is intended to exemplify this disillusion.
The first speaker uses the word as an imperative; the second is aware that his interlocutor awaits the desired object. To fulfill this awaiting is impossible, for all he can produce is its name. But at the same time there is an awaiting that would indeed be fulfilled by the absence of the object: the negative imperative consisting of the name of the object and the operator of interdiction. Uttered in linguistic presence, the name-plus-operator would indeed be an interdiction. In the second speaker’s thought, however, it would simply evoke the image of the object as absent; in other words, it is the name of the object-as-absent. The second speaker is aware that to say the name of what is not will not necessarily make it appear. But to say the name of what is, even if all there is, is the absence of the object, is to use the name not as an imperative but as an ostensive.
This negative ostensive is at the same time a correction of the first speaker’s utterance, which, insofar as it is an imperative, is itself a transformed ostensive, no longer “inappropriate,” no doubt, but deferred. The second speaker’s ostensive, then, in its negativity, is already fulfilled, but at the same time, by presenting in ostensive form the object requested by the first speaker, even if it is presented-as-not-present, his utterance has at least the potentiality of putting an end to the awaiting created by the original imperative.
There is no guarantee, of course, that this communication will be successful, because the ostensive offered in the dialogue is not what was originally expected. But in the genesis of the imperative there was no reason either to assume that the inappropriate use of the ostensive would be automatically rewarded. What is essential is that the new form exist as an intentional structure for the speaker, so that the hearer, who at first finds it inappropriate, may understand its meaning and eventually come to use it when he finds himself in similar circumstances.
The negative ostensive can thus arise as a possible negative “reply” to the imperative. Its acceptance by the first speaker in lieu of the requested object, as opposed to the more violent response that might be expected in a case of inappropriate fulfillment of his request, would constitute a further lowering of the threshold of significance from that which gave rise to the imperative. At that moment it was individual desire that was accepted as a possible source of significance; now it is the unfulfillability of this desire. But this is too negative a formulation: what the negative ostensive presents is simply the state of affairs, not as an object of desire in itself but insofar as it withstands the desire to modify it. Which is to say, that the lexeme of negation/interdiction has become a predicate.
This defeat of desire by reality is, in the sphere of representation, an immensely significant triumph of objectivity. The inappropriate (positive) ostensive opened the domain of linguistic representation to the infinity of desire; the inappropriate negative ostensive, in representing the limitations of desire, permits the dialectic of desire and reality to be comprehended entirely by language, so that linguistic models can henceforth mirror and anticipate the results of our attempts to realize our representations. But this development is predicated on the prior acceptance of the significance of those elements of reality that oppose desire. In urgent situations, these facts in themselves are signs of crisis and must be overcome through action. In those less urgent, the facts, albeit negative, may acquire communal significance in themselves.
Thus if our “Scalpel” dialogue occurs during an operation, the answer “Scalpel-no!” to the doctor’s question is not likely to be of help. If no scalpel is present then a substitute must be found immediately, and the assistant would do better to rush off to seek one than to attempt to “correct” the doctor’s imperative. But were the request made in more leisurely circumstances, say in the course of taking inventory, the negative reply would permit the functional act of reordering the missing item. The key criterion here is the immediacy of the universe of discourse. When the horizon of the interlocutors is limited to the present moment—to the moment of mutual presence—the inappropriate negative ostensive is functionless. Conversely, its functioning makes the universe beyond this presence and its extension through imperative awaiting for the first time a possible source of the significant.