The Declarative Model
With the derivation of the declarative sentence we have reached the final moment of the dialectic of linguistic form per se. The chief obstacle to the comprehension of the intentional structure of the declarative is its familiarity to us. We write of the lower forms from the telic perspective of the higher. But having attained the final stage of linguistic evolution, we find ourselves writing about the structures of the declarative sentence in declarative sentences. The paradoxes generated by desire in elementary language, which had led to the evolution of the elementary forms, can, in the declarative sentence, be converted without syntactic innovation into logical paradoxes.
The declarative sentence may be described as a predication about a topic. In its origin, as we saw, this topic was pre-established by the desire of the interlocutor as expressed in an imperative. We may assume that in any declarative sentence, the locutee has an implicit a priori interest in the topic; the topic-comment form makes the objectal focus of this interest clearer. Obviously, however, this interest need not be previously expressed, and certainly not in imperative form. The topic is always substantive, even if it refers to an activity, in which case we may think of it as a verbal noun (like the present participle in English).
Predication is a term rich in philosophical implications. Our use of it is not meant to imply that the evolution of language is telically subordinated to logic. To attempt to define predication at this stage as anything more than saying something about a topic would be self-defeating, because formal dichotomies like substance/accident or being/modification are merely ex post facto formalizations of the topic-predicate relation.
I chose to designate the chief nominal as topic rather than subject to avoid giving currency to the view, expressed in the Grammaire de Port-Royal and elsewhere, that the grammatical agreement between subject and predicate in Western languages possesses a peculiar ontological significance. It seems more likely that the topic-comment form evolved into the subject-predicate form through the tightening by habitual usage of the morphological links between topic and verb. This evolution is not, however, irreversible, for the concomitant reduction of the topic to a merely coordinate rather than superordinate position in the sentence provokes the adjunction of a new, emphatic topic not bound by rules of agreement. Such constructions (e.g., compare “My sister(, she) likes spinach” with “My sister, her teeth are crooked” or “My sister, I don’t like her”) are available in colloquial English, and, we may assume, in all subject-predicate languages. The significance of this plausible cyclical alternation between topic and subject is that it demonstrates the resistance of “natural” language to grammatical formalization, which is also a fortiori a resistance to the rigor of logical formalization. The logical proposition is no doubt a sentence, but the assimilation of sentences to logical propositions, however justified by the inherent potential of the declarative for the formulation of “context-free” de-temporalized models, eliminates from consideration precisely that element of temporal urgency, or in other terms, of significance, without which no linguistic usage can be understood. If grammarians and linguists fail to comprehend the elementary linguistic forms because they see them as degenerate declaratives, logicians and language philosophers misunderstand the higher linguistic forms in viewing them as avant la lettre elements of logical discourse.
The original negative-ostensive form of the declarative does not include a verb. The subject and negative operator are simply “coupled,” without a verb to be—as in Russian, except that whereas the verb is supplied in other tenses, in the primitive case there is no verb to omit. The specific question of the origin of the copula is not of interest to us here. But we cannot avoid the more general one of the origin of the verb. The latter is ubiquitous, and the markers of tense and person which we have designated as the most fundamental grammatical categories are in all languages attracted to it, if not actually contained within it as inflections. It is no doubt the ubiquity of the verb as bearer of these markers has led grammarians to consider it a more fundamental constituent of the sentence than its subject; a verb alone (as in the imperative) may constitute a complete sentence, a noun, never. The grammatical prestige of the verb appears paradoxical in the light of the unquestionably more fundamental character of nouns as the names for persons and palpable objects. But the paradox remains only so long as we consider the declarative as the “natural” sentence form, in which case the very existence of syntactical relations requires the simultaneous genesis of both nouns and verbs. It vanishes in our hypothesis, where nominal forms (whether semantically nominal or “verbal”) are the more primitive, and where the first utterance-forms are precisely lacking in verbs.
The verb is the sign of the declarative because it is the more evolved of the two substantial forms, through an evolution that is at the same time a degeneration, given that pre-declarative verbals were, like nominals, capable of serving as topics in themselves. But whether the topic be nominal or verbal in nature, it is not a verb. A verb can serve only as a predicate; it may in fact be defined simply as a predicate verbal. And whether nominals or verbals constitute the most fundamental lexical elements, words denoting objects are particularly likely to be found as topics, and words denoting actions as predicates.
If we return to our derivation of the negative ostensive, we note that the negative operator, which we consider to be the first predicate, can apply equally well, if not better, to verbal as to nominal imperatives, but that as a predicate it is more readily associated with nominals. This is in both cases because of the performative nature of the imperative verbal as opposed to the objectal nature of the nominal. Thus if one is equally likely to request an object or an action, the action, being commonly available to the imperative addressee independently of any requested object, is, although more likely to encounter interdiction, a far less likely candidate for unavailability than the object. And if indeed an action cannot be performed, merely “negating” it as unavailable conveys no information beyond a simple refusal, whereas the absence of an object is prima facie a verifiable fact. This original superiority of nominals as topics is only relative. But if we suppose the differentiation of predicates along the lines that we have suggested in our locative example, predicates that carry supplementary information beyond that of mere unavailability will apply even more exclusively to nominals, because an absent object exists and can thus be otherwise qualified, whereas an absent performance exists only in its absence. Thus if we assume that the original function of the declarative is to express the non-performance of imperatives, nominals will tend to appear more frequently as topics and verbals as predicates.
In the case of predicate adjectives and nominals (“the scalpel is broken”; “that tree is an X” [and therefore not good for tool-making] , etc.) it is difficult to suppose that the “verb” provided by the copula is anything more than the result of assimilation of this sentence form to the true verbal form, presumably in order to bear the grammatical burden of person and tense already attached to the verb. Which is to say that the existence of the verb as an indispensable formal element of predicates rather than a merely probable one is dependent on its association with these shifter paradigms. But this association should not be looked on as a merely morphological one, as though the evolution of the inflected verb could be described as the attachment to a proto-verbal predicate of proto-adverbial morphemes denoting tense and person. For such an explanation merely begs the essential question of why these morphemes are indeed necessary to the constitution of the declarative, even to the extent that dummy verbs come into existence to bear them. Our answer to this question must be founded on the already-established dominance of the verbal predicate. It is actions, not relations, that are essentially located in time and that differ in nature according as they are performed by speaker or hearer.
Thus it would appear that the morphological, paradigmatic elements of the verb merely formalize the verbal nature of predicates. Yet the passage from the probabilistic dominance we have described to formal dominance must still be explained by an additional element: the linguistic present constituted by the declarative. We have already seen the germ of this present in the imperative, which can be said to possess an incipient tense because it refers to a real time outside of linguistic presence stricto sensu (i.e., the time of the speech act), although referential time is intended not as independent of this presence but as an extension of it. In contrast, the declarative, even in its most primitive form, provides the model of a present independent of linguistic presence, and thus possesses a true tense, even if, as we must assume, the emergence of a paradigm of tenses is a later development. The existence of the verb is thus prior to its grammatical inflections, but this is only so because, even before the existence of tense paradigms, the declarative sentence already possesses an implicit tense. Once this has been more clearly established, the emergence of the verb as the general predicative form will follow, because, as we shall show, tense is an essentially verbal category.
Linguistic Present and Linguistic Presence
As a negative ostensive, the nascent declarative would appear to stand in the same dependent relation toward linguistic presence as its positive ostensive counterpart. The absence of a hammer “takes place” in the same real time and space as its presence: the presence of communication as established by the speaker of the imperative. But the function of the declarative model is very different from that of the ostensive. The latter recreates communal presence centered on a significant new phenomenon; the declarative functions in an already-established linguistic presence to negate the model proposed by the first speaker. The information contained in the declarative acts as a bar to the anticipated fulfillment of the imperative request, and in so doing establishes a barrier between the prolonged linguistic presence within which this fulfillment was awaited and the situation at hand.
The model of reality presented by the negative ostensive can, of course, be acted on, but the model for such action is not given by linguistic form. If the answer to the original imperative be, for example, that the hammer is “over there,” then the first speaker can make use of this information to go and get it. The relation of act to model, however, is now no longer immediate but analytical. The declarative has presented a state-of-affairs, and the realization of the original speaker’s desire within this state-of-affairs is neither dependent on the linguistic presence of the speakers nor, indeed, mediated by the utterance at all. It is the fact that the hammer is over there now that makes the appropriative action possible, not the fact that it is said now to be over there.
The correspondence between the now of the utterance and the now of the being-over-there of the hammer is thus not essential to the declarative model. If the second speaker had said that the hammer was over there yesterday, or that it would be over there tomorrow, his interlocutor might still act (perhaps differently) on this information. This locative predicate here, of course, has evolved beyond the simple negative. But the same considerations apply even in the more primitive case. The original imperative expressed a desire that was to be immediately satisfied through performance. The negative-ostensive reply leaves it to the first speaker whether he will redefine his desire in more realistic terms. The negative-ostensive model refers to the present, but only to annul the relationship between this present and the linguistic presence of the speakers. The imperative was founded on the faith that these two were inseparable, that there was no “present” other than linguistic presence, prolonged sufficiently into the future to permit the presentification of its referent. The negative ostensive reveals the illusion of this faith in the magical powers of the scene of representation.
Thus the declarative has a tense from the beginning, even if at first it be only the present. For the other tenses too are presents, of the past or future: To say the hammer was there yesterday is to say that, yesterday, it was present, yesterday’s present being, from the context-free perspective of the model, just like today’s. This present is that of the declarative model as a whole, yet it is within this model specifically an element of the predicate. The topic is requested by the first speaker and denied by the second, but whether it exists at all in the world, it has a reality in linguistic presence on the scene of representation. Its absence or even its nonexistence is what is said about it by the predicate. The topic is simply a given of the linguistic model, as the topic-comment sentence makes explicit by setting it off in first position, independent of the grammatical dependencies of the rest of the sentence.
Philosophers have been confused by the coordinate subject-predicate form into arguing for and against the “existence” of such things as a round square. In topic-comment form, however, even if we say “a round square cannot even be imagined,” the topical linguistic “existence” of the round square is beyond dispute. The present of the declarative, in which the topic becomes an element of a model of reality, is realized only in the predicate. This predicate need not be verbal in nature. But to the extent that it refers to a present, it temporalizes the topic, which at first presents itself atemporally, as a non-referential linguistic presence. Thus the topic-predicate form expresses a passage from the atemporal to the temporal.
Now insofar as we can distinguish verbals from nominals, the former are names of actions, that is, phenomena that can only be realized in time, whereas nominals exist in the significant memory as atemporal images. But this implies that when the predicate is verbal, the temporality inherent in the verbal-as-such becomes a property of the topic-as predicated-of, or we might say that the verbal predicate verbalizes the topic. In contrast, adjectival and nominal predicates have no verbal with which to verbalize the topic because the adjective or nominal cannot be itself an agent of temporalization. In a sentence like “the hammer (is) broken” or “John (is) sick,” the words “broken” or “sick” express states, not actions, and the now implicit in the declarative is verbalized, not in these words, but in the copula, even if unexpressed. The semantic sources of copulatives in verbs like to stand (Latin stare), to bear (Sanskrit bhū), which denote static “activities” and thus temporalize stasis, lend support to this analysis.
Thus declarative predicates acquire a verbal form as the result of their expression of a linguistic present. The other chief grammatical categories, governance and person, also inchoate in the imperative, are likewise formalized as specifications of the predicate and thus as functions of the verb. Imperative verbals may be associated with objects, but cannot truly be said to govern them because they designate in themselves a desired performance in which the object is presentified. In the declarative, the object becomes an element of the predicate, temporalized by its role in the action denoted by the verb. Thus in a sentence like “John takes (took) the hammer,” the hammer is not present in the model as an object of the speaker’s desire, but as the object of John’s action. The hearer is not merely permitted but required to conceive it as subordinate to this action, because it is this action alone that constitutes the now of the linguistic present.
The category of person is similarly formalized in the declarative model. Here we may pass over the third person, rightly classed by Benveniste as merely the nonpersonal, or if we like “zero-personal,” member of the paradigm. The first-person or second-person topic, upon its temporalization by the predicate, becomes itself situated in the now of the linguistic present. The shifter function of the personal pronouns situates the declarative model relatively to the linguistic presence of the two interlocutors, so that the original speaker, hearing (say) “I” as the topic, must imagine the other speaker in the temporal situation designated by the predicate. Thus the declarative model specifically presents one or the other speaker as a “real” element of the present (such as was only implicit in the imperative model), and because the verb carries in the predicate the tense of the present, it will tend to become “personalized” as well.
In contrast to the asymmetry of the imperative-declarative dialogue, the use of personal pronouns in the declarative reestablishes the symmetry of the ostensive gestures of pointing which were no doubt the most primitive “shifter” forms. In the present of the declarative, “I” and “you” form a paradigm, with the third-person form standing in contrast to both. The asymmetrical speaker-hearer relation is neutralized within the linguistic model; this is the same neutralization as was carried out by the (declarative) present on the asymmetry of (imperative) desire. Artificial languages may be made “context-free” without reference to the linguistic presence in which their messages are conveyed, but in human language this presence can never simply be ignored if the speakers are to remain liberated from prelinguistic violence.
To be continued…