The Archeology of Knowledge Michel Foucault contents part I introduction 3

2. THE ENUNCIATIVE FUNCTION

It is useless therefore to look for the statement among unitary groups of signs. The statement is neither a syntagma, nor a rule of construc­tion, nor a canonic form of succession and permutation; it is that which enables such groups of signs to exist, and enables these rules or forms to become manifest. But although it enables them to exist, it does so in a special way — a way that must not be confused with the existence of signs as elements of a language (longue), or with the material existence of those marks that occupy a fragment of space or last for a variable length of time. It is this special mode of existence, characteristic of every series of signs, providing it is stated, that we must now examine.
(a) So let us take once again the example of those signs made or drawn in a defined materiality, and grouped in a particular way, which may or may not be arbitrary, but which, in any case, is not grammat­ical: the keyboard of a typewriter, or a handful of printer^'s characters. All that is required is that the signs be given, that I copy them on to a sheet of paper (in the same order in which they appear, but without producing a word) for a statement to emerge: the statement of the letters of the alphabet in an order that makes the typing of them easier, and the statement of a random group of letters. What has happened,

((100))

then, that a statement should have been made? What can the second group possess that is not possessed by the first? Reduplication, the fact that it is a copy? Certainly not, since the keyboards of typewriters all copy a certain model and are not, by that very fact, statements. The intervention of a subject? This answer is inadequate for two reasons: it is not enough that the reiteration of a series be due to the initiative of an individual for it to be transformed, by that very fact, into a state-ment; and, in any case, the problem does not lie in the cause or origin of the reduplication, but in the special relation between the two identi­cal series. The second series is not a statement because and only because a hi-univocal relation can be established between each of its elements in the first series (this relation characterizes either the fact of duplication if it is simply a copy, or the exactitude of the statement if one has in fact crossed the threshold of enunciation; but it does not allow us to define this threshold and the very fact of the statement). A series of signs will become a statement on condition that it possesses ^`something else^' (which may be strangely similar to it, and almost identical as in the example chosen), a specific relation that concerns itself — and not its cause, or its elements.

It may be objected that there is nothing enigmatic about this rela­tion; that, on the contrary, it is a very familiar one, which is constantly being analysed: that, in fact, it concerns the relation of the signifier (significant) to the signified (signifie), of the name to what it designates; the relation of the sentence to its meaning; the relation of the prop­osition to its referent (referent). But I believe that one can show that the relation of the statement to what it states is not superposable on any of these relations.

The statement, even if reduced to a nominal syntagma ('The boat!^'), even if it is reduced to a proper noun ('Peter!'), does not have the same relation with what it states as the name with what it designates or signifies. The name or noun is a linguistic element that may occupy different places in grammatical groups: its meaning is defined by its rules of use (whether these concern individuals who may be validly designated by it, or syntactical structures in which it may correctly participate); a noun is defined by its possibility of recurrence. A state-ment exists outside any possibility of reappearing; and the relation that it possesses with what it states is not identical with a group of rules of

((101))

use. It is a very special relation: and if in these conditions an identical formulation reappears, with the same words, substantially the same names — in fact, exactly the same sentence — it is not necessarily the same statement.

Nor should the relation between a statement and what it states be confused with the relation between a proposition and its referent. We know that logicians say that a proposition like The golden mountain is in California^' cannot be verified because it has no referent: its negation is therefore neither more nor less true than its affirmation. Should we say similarly that a statement refers to nothing if the proposition, to which it lends existence, has no referent? Rather the reverse. We should say not that the absence of a referent brings with it the absence of a correlate for the statement, but that it is the correlate of the statement — that to which it refers, not only what is said, but also what it speaks of, its ^'theme^' — which makes it possible to say whether or not the prop­osition has a referent: it alone decides this in a definitive way. Let us suppose in fact that the formulation ^'The golden mountain is in Cali­fornia^' is found not in a geography book, nor in a travel book, but in a novel, or in some fictional context or other, one could still accord it a value of truth or error (according to whether the imaginary world to which it refers does or does not authorize such a geological and geo-graphical fantasy). We must know to what the statement refers, what is its space of correlations, if we are to say whether a proposition has or has not a referent. ^'The present king of France is bald^' lacks a referent only if one supposes that the statement refers to the world of con-temporary historical information. The relation of the proposition to the referent cannot serve as a model or as a law for the relation of the statement to what it states. The latter relation not only does not belong to the same level as the former, but it is anterior to it.

Nor is it superposable to the relation that may exist between a sen­tence and its meaning. The gap between these two forms of relation appears clearly in the case of two famous sentences that are meaning-less, in spite of their perfectly correct grammatical structure (as in the example: ^'Colourless green ideas sleep furiously'). In fact, to say that a sentence like this is meaningless presupposes that one has already excluded a number of possibilities — that it describes a dream, that it is part of a poetic text, that it is a coded message, that it is spoken by a

((102))

drug addict — and that one assumes it to be a certain type of statements that must refer, in a very definite way, to some visible reality. The relation of a sentence with its meaning resides within a specific, well-stabilized enunciative relation. Moreover, even if these sentences are taken at an enunciative level at which they are meaningless, they are not, as statements, deprived of correlations: there are those that enable one to say, for example, that ideas are never either coloured or colour-less, and therefore that the sentence is meaningless (and these correl­ations concern a level of reality in which ideas are invisible, and in which colours can he seen, etc.); there are also those correlations that validate the sentence in question as a mention of a type of correct syntactical organization that was also meaningless (and these correl­ations concern the level of the language (langue), with its laws and properties). A sentence cannot be non-significant; it refers to something, by virtue of the fact that it is a statement.

How, then, can we define this relation that characterizes the state-ment as statement — a relation that seems to be implicitly presupposed by the sentence or the proposition, and which is anterior to it? How can we disentangle it from those relations of meaning or those values of truth, with which it is usually confused? Any statement, as simple a statement as one can imagine, does not have as its correlate an individual or a particular object that is designated by this or that word in the sentence: in the case of a statement like ^`The golden mountain is in California^', the correlate is not the formation, real or imaginary, possible or absurd, that is designated by the nominal syntagma that serves as the subject. But nor is the correlate of the statement a state of things or a relation capable of verifying the proposition (in the example chosen, this would be the spatial inclusion of a particular mountain in a par­ticular region). On the other hand, what might he defined as the correlate of the statement is a group of domains in which such objects may appear and to which such relations may be assigned: it would, for example, be a domain of material objects possessing a certain number of observable physical properties, relations of perceptible size — or, on the contrary, it would he domain of fictitious objects, endowed with arbitrary properties (even if they have a certain constancy and a certain coherence), without any authority of experimental or perceptive veri­fication; it would be a domain of spatial and geographical localizations,

((103))

with coordinates, distances, relations of proximity and of inclusion — or, on the contrary, a domain of symbolic appurtenances and secret kinships; it would be a domain of objects that exist at the same moment and on the same time-scale as the statement is formulated, or it would be a domain of objects that belongs to a quite different pres­ent — that indicated and constituted by the statement itself, and not that to which the statement also belongs. A statement is not confronted (face to face, as it were) by a correlate — or the absence of a correlate — as a proposition has (or has not) a referent, or as a proper noun designates someone (or no one). It is linked rather to a ^'referential^' that is made up not of ^'things^', ^'facts^', ^'realities^', or ^'beings^', but of laws of possibil­ity, rules of existence for the objects that are named, designated, or described within it, and for the relations that are affirmed or denied in it. The referential of the statement forms the place, the condition, the field of emergence, the authority to differentiate between individuals or objects, states of things and relations that are brought into play by the statement itself; it defines the possibilities of appearance and delimitation of that which gives meaning to the sentence, a value as truth to the proposition. It is this group that characterizes the enunciative level of the formulation, in contrast to its grammatical and logical levels: through the relation with these various domains of possibility the statement makes of a syntagma, or a series of symbols, a sentence to which one may or may not ascribe a meaning, a proposition that may or may not be accorded a value as truth.

One can see in any case that the description of this enunciative level can be performed neither by a formal analysis, nor by a semantic investigation, nor by verification, but by the analysis of the relations between the statement and the spaces of differentiation, in which the statement itself reveals the differences.
(b) A statement also differs from any series of linguistic elements by virtue of the fact that it possesses a particular relation with a subject. We must now define the nature of this relation, and, above all, distinguish it from other relations with which it might be confused.

We must not, in fact, reduce the subject of the statement to the first-person grammatical elements that are present within the sentence. First because the subject of th.e sentence is not within the linguistic

((104))

syntagma; secondly because a statement that does not involve a first person nevertheless has a subject; lastly and above all, all statements that have a fixed grammatical form (whether in the first or second person) do not have the same type of relation with the subject of the statement. It is easy to see that this relation is not the same in a state-ment of the type ^'Night is falling', and ^'Every effect has a cause^'; while in the case of a statement of the type ^'Longtemps, je me Buis couche de bonne heure^' (^'For a long time I used to go to bed early^'), the relation to the enunciating subject is not the same if one hears it spoken in the course of a conversation, and if one reads it at the beginning of Proust^'s A la Recherche du temps perdu.

Is not this subject exterior to the sentence quite simply the indi­vidual who spoke or wrote those words? As we know, there can be no signs without someone, or at least something, to emit them. For a series of signs to exist, there must — in accordance with the system of causality — be an ^'author^' or a transmitting authority. But this ^`author^' is not identical with the subject of the statement; and the relation of production that he has with the formulation is not superposable to the relation that unites the enunciating subject and what he states. Let us ignore the over-simple case of a group of signs that have been materi­ally fashioned or traced: their production implies an author even though there is neither a statement nor a subject of a statement. One might also mention, by way of showing the dissociation between the transmitter of signs and the subject of a statement, the case of a text read by a third person, or that of an actor speaking his part. But these are extreme cases. Generally speaking, it would seem, at first sight at least, that the subject of the statement is precisely he who has produced the various elements, with the intention of conveying meaning. Yet things are not so simple. In a novel, we know that the author of the formulation is that real individual whose name appears on the title page of the book (we are still faced with the problem of the dialogue, and sentences purporting to express the thoughts of a character; we are still faced with the problem of texts published under a pseudonym: and we know all the difficulties that these duplications raise for practi­tioners of interpretative analysis when they wish to relate these formu­lations, en bloc, to the author of the text, to what he wanted to say, to what he thought, in short, to that great silent, hidden, uniform

((105))

discourse on which they build that whole pyramid of different levels); but, even apart from those authorities of formulation that are not iden­tical with the individual/author, the statements of the novel do not have the same subject when they provide, as if from the outside, the historical and spatial setting of th.e story, when they describe things as they would be seen by an anonymous, invisible, neutral individual who moves magically among the characters of the novel, or when they provide, as if by an immediate, internal decipherment, the verbal version of what is silently experienced by a character. Although the author is the same in each case, although he attributes them to no one other than himself, although he does not invent a supplementary link between what he is himself and the text that one is reading, these statements do not presuppose the same characteristics for the enunciat­ing subject; they do not imply the same relation between this subject and what is being stated.

It might be said that the often quoted example of the fictional text has no conclusive validity; or rather that it questions the very essence of literature, and not the status the subject of statements in general. According to this view, it is in the nature of literature that the author should appear to be absent, conceal himself within it, delegate his authority, or divide himself up; and one should not draw a general conclusion from this dissociation that the subject of the statement is distinct in everything — in nature, status, function, and identity — from the author of the formulation. Yet this gap is not confined to literature alone. It is absolutely general in so far as the subject of the statement is a particular function, but is not necessarily the same from one state-ment to another; in so far as it is an empty function, that can be filled by virtually any individual when he formulates the statement_; and in so far as one and the same individual may occupy in turn, in the same series of statements, different positions, and assume the role of different subjects. Take the example of a mathematical treatise. In the sentence in the preface in which one explains why this treatise was written, in what circumstances, in response to what unsolved prob­lems, or with what pedagogical aim in view, using what methods, after what attempts and failures, the position of the enunciative subject can be occupied only by the author, or authors, of the formulation: the conditions of individualization of the subject are in fact very strict, very

((106))

numerous, and authorize in this case only one possible subject. On the other hand, if in the main body of the treatise, one meets a proposition like ^'Two quantities equal to a third quantity are equal to each other^', the subject of the statement is the absolutely neutral position, indiffer­ent to time, space, and circumstances, identical in any linguistic sys­tem, and in any code of writing or symbolization, that any individual may occupy when affirming such a proposition. Moreover, sentences like ^'We have already shown that . . .' necessarily involve statements of precise contextual conditions that were not implied by the preceding formulation: the position is then fixed within a domain constituted by a finite group of statements; it is localized in a series of enunciative events that must already have occurred; it is established in a demonstra­tive time whose earlier stages are never lost, and which do not need therefore to be begun again and repeated identically to be made pres­ent once more (a mention is enough to reactivate them in their ori­ginal validity); it is determined by the prior existence of a number of effective operations that need not have been performed by one and the same individual (he who is speaking now), but which rightfully belong to the enunciating subject, which are at his disposal, and of which he may avail himself when necessary. The subject of such a statement will be defined by these requisites and possibilities taken together; and he will not be described as an individual who has really carried out certain operations, who lives in an unbroken, never forgot-ten time, who has interiorized, in the horizon of his consciousness, a whole group of true propositions, and who retains, in the living pres­ent of his thought, their potential reappearance (this is merely, in the case of individuals, the psychological, ^'lived^' aspect of their position as enunciating subjects).

Similarly, one might describe the specific position of the enunciat­ing subject in sentences like 'I call straight any series of points that .. or ^'Let there be a finite series of any elements^'; in each case the position of the subject is linked to the existence of an operation that is both determined and present; in each case, the subject of the statement is also the subject of the operation (he who establishes the definition of a straight line is also he who states it; he who posits the existence of a finite series is also, and at the same time, he who states it); and in each case, the subject links, by means of this operation and the statement in

((107))

which it is embodied, his statement as his own law). There is a differ­ence however: in the first case, what is stated is a convention of language (langage) — of that language that the enunciating subject must use, and within. which he is defined: the enunciating subject and what is stated are therefore at the same level (whereas for a formal analysis a statement like this one implies the difference of level proper to meta-language); in the second case, on the other hand, the enunciating subject brings into existence outside himself an object that belongs to a previously defined domain, whose laws of possibility have already been articulated, and whose characteristics precede the enunciation that posits it. We saw above that the position of the enunciating subject is not always identical in the affirmation of a true proposition; we now see that it is also not identical when an operation is carried out within the statement itself.

So the subject of the statement should not he regarded as identical with the author of the formulation — either in substance, or in function. He is not in fact the cause, origin, or starting-point of the phenomenon of the written or spoken articulation of a sentence; nor is it that mean­ingful intention which, silently anticipating words, orders them like the visible body of its intuition; it is not the constant, motionless, unchanging focus of a series of operations that are manifested, in turn, on the surface of discourse through the statements. It is a particular, vacant place that may in fact be filled by different individuals; but, instead of being defined once and for all, and maintaining itself as such throughout a text, a book, or an oeuvre, this place varies — or rather i.t is variable enough to he able either to persevere, unchanging, through several sentences, or to alter with each one. It is a dimension that characterizes a whole formulation qua statement. It is one of the charac­teristics proper to the enunciative function and enables one to describe it. If a proposition, a sentence, a group of signs can be called ^`state-ment^', it is not therefore because, one day, someone happened to speak them or put them into some concrete form of writing; it is because the position of the subject can be assigned. To describe a formulation qua statement does not consist in analysing the relations between the author and what he says (or wanted to say, or said without wanting to) ; but in determining what position can and must be occupied by any individual if he is to be the subject of it.

((108))

(c) The third characteristic of the enunciative function: it cannot operate without the existence of an associated domain. This makes the statement something other, something more, than a mere collection of signs, which, in order to exist, need only a material base — a writing surface, sound, malleable material, the hollowed incision of a trace. But this also, and above all, distinguishes it from the sentence and the proposition.

Take a group of words or symbols. In order to decide whether they constitute a grammatical unit like the sentence or a logical unit like the proposition, it is necessary, and enough, to determine the rules accord-ing to which it was constructed. ^'Peter arrived yesterday^' forms a sen­tence, but ^'Yesterday arrived Peter^' does not; A + B = C + D constitutes a proposition, but ABC + = D does not. Only an examination of the elements and of their distribution, in reference to the system — natural or artificial — of the language (longue) enables us to distinguish between what is and what is not a proposition, between what is a sentence and what is merely an accumulation of words. Moreover, this examination is enough to determine to what type of grammatical structure the sentence in question belongs (affirmative sentence, in the past tense, containing a nominal subject, etc.), or to what type of proposition the series of signs in question belongs (an equivalence between two add­itions). One can even conceive of a sentence or a proposition that is ^'self-determining^', that requires no other sentence or proposition to serve as a context, no other associated sentences or propositions: that such a sentence or proposition would, in such conditions, be useless and unusable, does not mean that one would not be able to recognize it, even in its singularity.

One could no doubt make a number of objections to this. One might say, for example, that a proposition can be established and individual­ized as such only if one knows the system of axioms that it obeys; do not those definitions, those rules, those conventions of writing form an associated field inseparable from the proposition (similarly, the rules of grammar, implicitly at work in the competence of the subject, are necessary if one is to recognize a sentence, and a sentence of a certain type)? It should be noted however that this group — actual or potential — does not belong to the same level as the proposition or the sentence: but that it has a bearing on their possible elements, succession, and

((109))

distribution. The group is not associated with them: it is presupposed by them. One might also object that many (non-tautological.) proposi­tions cannot be verified on the basis of their rules of construction alone, and that recourse to the referent is needed if one is to decide whether they are true or false: but true or false, a proposition. remains a proposition, and it is not recourse to the referent that decides whether or not it is a proposition. The same goes for sentences; in many cases, they can yield their meaning only in relation to the context (whether they contain ^`deictic^' elements that refer to a concrete situation; or make use of first — or second-person pronouns that designate the speaking subject and hi.s interlocutors; or make use of pronominal elements or connecting particles that refer to earlier or later sentences); but the fact that its meaning cannot be completed does not prevent the sentence from being grammatically complete and autonomous. Cer­tainly, one is not very sure what a group of words like ^'I^'ll tell you that tomorrow^' means; in any case, one can neither date this ^'tomorrow^', nor name the interlocutors, nor guess what is to be said. Nevertheless, it is a perfectly delimited sentence, obeying the rules of construction of the language (longue) in which it is written. Lastly, one might object that, without a context, it is sometimes difficult to define the structure of a sentence (^'I shall never know if h.e is dead^' may be construed: ^'I shall never know whether or n.ot he is dead^' or ^'I shall never be informed of his death when this even occurs^'). But this ambiguity is perfectly definable, simultaneous possibilities can be posited that belong to the structure proper of the sentence. Generally speaking, one can say that a sentence or a proposition — even when isolated, even divorced from the natural context that could throw light on to its meaning, even. freed or cut off from all the elements to which, implicitly or not, it refers — always remains a sentence or a proposition and can. always be recognized as such.

On the other hand, the enunciative function — and this shows that it is not simply a construction of previously existing elements — cannot operate on a sentence or proposition in isolation. It is not enough to say a sentence, i.t is not even enough to say it in a particular relation to a field of objects or in a particular relation to a subject, for a statement to exist: it must be related to a whole adjacent field. Or rather, since this is not some additional relation. that is superimposed on the others, one

((110))

cannot say a sentence, one cannot transform it into a statement, unless a collateral space is brought into operation. A statement always has borders peopled by other statements. These borders are not what is usually meant by ^'context^' — real or verbal — that is, all the situational or linguistic elements, taken together, that motivate a formulation and determine its meaning. They are distinct from such a ^'context^' precisely in so far as they make it possible: the contextual relation between one sentence and those before an.d after it is not the same in the case of a novel and in that of a treatise in physics; the contextual relation between a formulation and the objective environment is not the same in a conversation and in the account of an experiment. It is against the background of a more general relation between the formulations, against the background of a whole verbal network, that the effect of context may he determined. Nor are these borders identical with the various texts and sentences that the subject may be conscious of when he speaks; again they are more extensive than such a psychological setting; and to a certain extent they determine that setting, for accord-ing to the position, status, and role of one formulation among others — according to whether it belongs to the field of literature or as an isolated remark, whether it is part of a narrative or the account of a demonstration — the way in which other statements are present in the mind of the subject will not he the same: neither the same level, nor the same form of linguistic experience, of verbal memory, of reference to what has already been said, is operating in each case. The psychological halo of a formulation is controlled from afar by the arrangement of the enunciative field.

The associated field that turns a sentence or a series of signs into a statement, and which provides them with a particular context, a spe­cific representative content, forms a complex web. It is made up first of all by the series of other formulations within which the statement appears and forms one element (the network of spoken formulations that make up a conversation, the architecture of a demonstration, hound on the one side by its premises and on the other by its conclu­sion, the series of affirmations that make up a narrative). The associated field is also made up of all the formulations to which the statement refers (implicitly or not), either by repeating them, modifying them, or adapting them, or by opposing them, or by commenting on them;

((111))
there can be no statement that in one way or another does not reactual­ize others (ritual elements in a narrative; previously accepted proposi­tions in a demonstration; conventional sentences in a conversation). The associated field is also made up of all the formulations whose subsequent possibility is determined by the statement, and which may follow the statement as its consequence, its natural successor, or its conversational retort (an order does not open up the same enunciative possibilities as the propositions of an axiomatic or the beginning of a narrative). Lastly, the associated field is made up of all the formulations whose status the statement in question shares, among which it takes its place without regard to linear order, with which it will fade away, or with which, on the contrary, it will be valued, preserved, sacralized, and offered, as a possible object, to a future discourse (a statement is not dissociable from the status that it may receive as ^`literature^', or as an unimportant remark that is barely worthy of being forgotten, or as a scientific truth valid for all time, or as prophetic words, etc.). Generally speaking, one can say that a sequence of linguistic elements is a state-ment only if it is immersed in an enunciative field, in which it then appears as a unique element.

The statement is not the direct projection on to the plane of language (langage) of a particular situation or a group of representations. It is not simply the manipulation by a speaking subject of a number of elements and linguistic rules. At the very outset, from the very root, the state-ment is divided up into an enunciative field in which it has a place and a status, which arranges for its possible relations with the past, and which opens up for it a possible future. Every statement is specified in this way: there is no statement in general, no free, neutral, independent statement; but a statement always belongs to a series or a whole, always plays a role among other statements, deriv­ing support from them and distinguishing itself from them: it is always part of a network of statements, in which it has a role, however minimal it may be, to play. Whereas grammatical con­struction needs only elements and rules in order to operate; whereas one might just conceive of a language (longue) — an artificial one, of course — whose only purpose is the construction of a single sentence; whereas the alphabet, the rules of construction and trans-formation of a formal system being given, one can perfectly well

((112))

define the first proposition of this language (langage), the same can-not be said of the statement. There is no statement that does not presuppose others; there is no statement that is not surrounded by a field of coexistences, effects of series and succession, a distribution of functions and roles. If one can speak of a statement, it is because a sentence (a proposition) figures at a definite point, with a specific position, in an enunciative network that extends beyond it.

Against this background of enunciative coexistence, there stand out, at an autonomous and describable level, the grammatical relations between sentences, the logical relations between propositions, the metalinguistic relations between an object language and one that defines the rules, the rhetorical relations between groups (or elements) of sentences. It is permissible, of course, to analyse all these relations without taking as one^'s theme the enunciative field itself, that is, the domain of coexistence in which the enunciative function operates. But they can exist and are analysable only to the extent that these sentences have been ^'enunciated^'; in other words, to the extent that they are deployed in an enunciative field that allows them to follow one another, order one another, coexist with one another, and play roles in relation to one another. Far from being the principle of individualiza­tion of groups of ^'signifiers^' (the meaningful ^`atom^', the minimum on the basis of which there is meaning), the statement is that which situates these meaningful units in a space in which they breed and multiply.
(d) Lastly, for a sequence of linguistic elements to be regarded and analysed as a statement, it must fulfil a fourth condition: it must have a material existence. Could one speak of a statement if a voice had not articulated it, if a surface did not bear its signs, if it had not become embodied in a sense-perceptible element, and if it had not left some trace — if only for an instant — in someone's memory or in some space? Could one speak of a statement as an ideal, silent figure? The statement is always given through some material medium, even if that medium is concealed, even if it is doomed to vanish as soon as it appears. And the statement not only needs this materiality; its materiality is not given to it, in addition, once all its determinations have been fixed: it is partly made up of this materiality. Even if a sentence is composed of the same

((113))

words, bears exactly the same meaning, and preserves the same syn-tactical and semantic identity, it does not constitute the same statement if it is spoken by someone in the course of a conversation, or printed in a novel; if it was written one day centuries ago, and if it now reappears in an oral formulation. The coordinates and the material status of the statement are part of its intrinsic characteristics. That is an obvious fact. Or almost. For as soon as one examines it a little more closely, things begin to blur and the problems increase.

Of course, it is tempting to say that if a statement is characterized, partly at least, by its material status, and if its identity is susceptible to a modification of that status, the same can be said of sentences and propositions: the materiality of signs is not, in fact, entirely indifferent to grammar or even to logic. We know what theoretical problems are presented to logic by the material constancy of the symbols used (how to define the identity of a symbol through the various substances in which it may be embodied and the variations of form that it can tolerate? How to recognize it and make certain that it is the same, if it must be defined as ^'a concrete physical form?^'); we know too what problems are presented to logic by the very notion of a series of sym-bols (what do ^'precede^' and ^'follow^' mean? Come ^'before^' and ^'after^'? In what space is such an order situated?). Much better known still are the relations of materiality and the language (longue) — the role of writ-ing and the alphabet, the fact that neither the same syntax, nor the same vocabulary operate in a written text and in a conversation, in a news-paper and in a book, in a letter and on a poster; moreover, there are series of words that form perfectly individualized and acceptable sen­tences if they feature as newspaper headlines, and which, nevertheless, in the course of a conversation, could never stand as meaningful sen­tences. Yet the materiality plays a much more important role in the statement: it is not simply a principle of variation, a modification of the criteria of recognition, or a determination of linguistic sub-groups. It is constitutive of the statement itself: a statement must have a substance, a support, a place, and a date. And when these requisites change, it too changes identity. At this point, a host of questions arises: Does the same sentence repeated very loudly and very softly form one or more statements? When one learns a text by heart, does each recitation con­stitute a statement, or should one regard it as a repetition of the same

((114))

statement? A sentence is faithfully translated into a foreign language: two distinct statements or one? And in a collective recitation — a prayer or a lesson — how many statements are produced? How can one establish the identity of the statement through all these various forms, repetitions, and transcriptions?

The problem is no doubt obscured by the fact that there is often a confusion of different levels. To begin with, we must set aside the multiplicity of enunciations. We will say that an enunciation takes place whenever a group of signs is emitted. Each of these articulations has its spatiotemporal individuality. Two people may say the same thing at the same time, but since there are two people there will he two distinct enunciations. The same person may repeat the same sentence several times; this will produce the same number of enunciations distinct in time. The enunciation is an unrepeatable event; it has a situated and dated uniqueness that is irreducible. Yet this uniqueness allows of a number of constants — grammatical, semantic, logical — by which one can, by neutralizing the moment of enunciation and the coordinates that individualize it, recognize the general form of a sentence, a mean-ing, a proposition. The time and place of the enunciation, and the material support that it uses, then become, very largely at least, indif­ferent: and what stands out is a form that is endlessly repeatable, and which may give rise to the most dispersed enunciations. But the state-ment itself cannot be reduced to this pure event of enunciation, for, despite its materiality, it cannot be repeated: it would not be difficult to say that the same sentence spoken by two people in slightly different circumstances constitute only one statement. And yet the statement cannot he reduced to a grammatical or logical form because, to a greater degree than that form, and in a different way, it is susceptible to differences of material, substance, time, and place. What, then, is that materiality proper to the statement, and which permits certain special types of repetition? How is it that one can speak of the same statement when there are several distinct enunciations of it, yet must speak of several statements when one can recognize identical forms, structures, rules of construction, and intentions? What, then, is this rule of repeatable materiality that characterizes the statement?

This may not he a perceptible, qualitative materiality, expressed in the form of colour, sound, or solidity, and divided up by the same

((115))

spatiotemporal observation as the perceptual space. Let us take a very simple example: a text reproduced several times, the successive edi­tions of a book, or, better still, the different copies of the same printing, do not give rise to the same number of distinct statements: in all the editions of Les Fleurs du mal (variants and rejected versions apart) , we find the same set of statements; yet neither the characters, nor the ink, nor the paper, nor even the placing of the text and the positions of the signs, are the same: the whole texture of the materiality has changed. But in this case these ^'small^' differences are not important enough to alter the identity of the statement and to bring about another: they are all neutralized in the general element — material, of course, but also institutional and economic — of the ^'book^': a book, however many copies or editions are made of it, however many different substances it may use, is a locus of exact equivalence for the statements — for them it is an authority that permits repetition without any change of identity. We see from this first example that the materiality of the statement is not defined by the space occupied or the date of its formulation; but rather by its status as a thing or object. A status that is never definitive, but modifiable, relative, and always susceptible of being questioned: we know for example that, for literary historians, the edition of a book published with the agreement of the author does not have the same status as posthumous editions, that the statements in it have a unique value, that they are not one of the manifestations of one and the same whole, that they are that by relation to which there is and must be repetition. Similarly, between the text of a Constitution, or a will, or a religious revelation, and all the manuscripts or printed copies that reproduce them exactly, with the same writing, in the same characters, and on similar substances, one cannot say that there is an equivalence: on the one hand there are the statements themselves, and on the other their reproduction. The statement cannot be identified with a fragment of matter; but its identity varies with a complex set of material institutions.

For a statement may be the same, whether written on a sheet of paper or published in a book; it may be the same spoken, printed on a poster, or reproduced on a tape-recorder; on the other hand, when a novelist speaks a sentence in daily life, then reproduces the same sen­tence in the manuscript that he is writing, attributing it to one of his

((116))

characters, or even allowing it to be spoken by that anonymous voice that passes for that of the author, one cannot say that it is the same statement in each case. The rule of materiality that statements necessar­ily obey is therefore of the order of the institution rather than of the spatio-temporal localization; it defines possibilities of reinscription and transcription (hut also thresholds and limits), rather than limited and perishable individualities.

The identity of a statement is subjected to a second group of condi­tions and limits: those that are imposed by all the other statements among which it figures, by the domain in which it can be used or applied, by the role and functions that it can perform. The affirmation that the earth is round or that species evolve does not constitute the same statement before and after Copernicus, before and after Darwin; it is not, for such simple formulations, that the meaning of the words has changed; what changed was the relation of these affirmations to other propositions, their conditions of use and reinvestment, the field of experience, of possible verifications, of problems to be resolved, to which they can be referred. The sentence ^'dreams fulfil desires^' may have been repeated throughout the centuries; it is not the same state-ment in Plato and in Freud. The schemata of use, the rules of applica­tion, the constellations in which. they can play a part, their strategic potentialities constitute for statements a field of stabilization that makes it possible, despite all the differences of enunciation, to repeat them in their identity; but this same field may also, beneath the most manifest semantic, grammatical, or formal identities, define a threshold beyond which there can be no further equivalence, and the appear­ance of a new statement must be recognized. But it is possible, no doubt, to go further: there are cases in which one may consider that there is only one statement, even though the words, the syntax, and the language (langue) itself are not identical. Such cases are a speech and its simultaneous translation; a scientific text in English and its French version; a notice printed in three columns in three different languages: there are not, in such cases, the same number of state­ments as there are languages used, but a single group of statements in different linguistic forms. Better still: a given piece of information may be retransmitted with other words, with a simplified syntax, or in an agreed code; if the information content and the uses to which it

((117))

could be put are the same, one can say that it is the same statement in each case.

Here too, we are concerned not with a criterion of individualization for the statement, but rather with its principle of variation: it is some-times more diverse than the structure of the sentence (and its identity is then finer, more fragile, more easily modifiable than that of a seman­tic or grammatical whole), sometimes more constant than that struc­ture (and its identity is then broader, more stable, more susceptible to variations). Moreover, not only can this identity of the statement not be situated once and for all in relation to that of the sentence, but it is itself relative and oscillates according to the use that is made of the statement and the way in which it is handled. When one uses a statement in such a way as to reveal its grammatical structure, its rhetorical configuration, or the connotations that it may carry, it is obvious that one cannot regard it as being identical in its original language (longue) and in a translation. On the other hand, if it is intended as part of a procedure of experimental verification, then text and translation constitute a single enunciative whole. Or again, at a certain scale of macro-history, one may consider that an affirmation like ^'species evolve^' forms the same statement in Darwin and in Simpson; at a finer level, and considering more limited fields of use (^'neo-Darwinism^' as opposed to the Darwin­ian system itself), we are presented with two different statements. The constancy of the statement, the preservation of its identity through the unique events of the enunciations, its duplications through the identity of the forms, constitute the function of the field of use in which it is placed.

The statement, then, must not be treated as an event that occurred in a particular time and place, and that the most one can do is recall it — and celebrate it from afar off — in an act of memory. But neither is it an ideal form that can be actualized in any body, at any time, in any circumstances, and in any material conditions. Too repeatable to be entirely identifiable with the spatio-temporal coordinates of its birth (it is more than the place and date of its appearance), too bound up with what surrounds it and supports it to be as free as a pure form (it is more than a law of construction governing a group of elements), it is endowed with a certain modifiable heaviness, a weight relative to the field in which it is placed, a constancy that allows of various uses, a

((118))

temporal permanence that does not have the inertia of a mere trace or mark, and which does not sleep on its own past. Whereas an enunci­ation may be begun again or re-evoked, and a (linguistic or logical) form may be reactualized, the statement may be repeated — but always in strict conditions.

This repeatable materiality that characterizes the enunciative function reveals the statement as a specific and paradoxical object, but also as one of those objects that men produce, manipulate, use, transform, exchange, combine, decompose and recompose, and possibly destroy. Instead of being something said once and for all — and lost in the past like the result of a battle, a geological catastrophe, or the death of a king — the statement, as it emerges in its materiality, appears with a status, enters various networks and various fields of use, is subjected to trans­ferences or modifications, is integrated into operations and strategies in which its identity is maintained or effaced. Thus the statement circu­lates, is used, disappears, allows or prevents the realization of a desire, serves or resists various interests, participates in challenge and struggle, and becomes a theme of appropriation or rivalry.
((119))