The Archeology of Knowledge Michel Foucault contents part I introduction 3

Part III The Statement and the Archive

((89))

1. DEFINING THE STATEMENT

I suppose that by now we have accepted the risk; that we are now willing, in order to articulate the great surface of discourse, to posit the existence of those somewhat strange, somewhat distant figures that I have called discursive formations; that we have put to one side, not in a definitive way, but for a time and out of methodological rigour, the traditional unities of the book and the ceuvre; that we have ceased to accept as a principle of unity the laws of constructing discourse (wi.th the formal organization that results), or the situation of the speaking subject (with the context and the psychological nucleus that character­ize it); that we no longer relate discourse to the primary ground of experience, nor to the a priori authority of knowledge; but that we seek the rules of its formation in discourse itself. I suppose that we have agreed to undertake these long inquiries into the system of emergence of objects, the system of the appearance and distribution of enunciative modes, the system of the placing and dispersion of concepts, the system of the deployment of strategic choices. I suppose that we are willing to construct such abstract, problematic unities, instead of welcoming those that presented themselves as being more or less perceptually familiar, if not as self-evident realities.

But what, in fact, have I been speaking about so far? What has been the object of my inquiry? And what did I intend to describe?

((90))

^'Statements^' — both in that discontinuity that frees them from all the forms in which one was so ready to allow them to be caught, and in the general, unlimited, apparently formless field of discourse. But I refrained from providing a preliminary definition of the statement. Nor did I try to construct one as I proceeded in order to justify the naivety of my starting-point. Moreover — and this no doubt is the reason for so much unconcern — I wonder whether I have not changed direction on the way; whether I have not replaced my first quest with another; whether, while analysing ^'objects^' or ^'concepts^', let alone ^'strategies^', I was in fact still speaking of statements; whether the four groups of rules by which I characterized a discursive formation really did define groups of statements. Lastly, instead of gradually reducing the rather fluctuating meaning of the word ^'discourse^', I believe that I have in fact added to its meanings: treating it sometimes as the general domain of all statements, sometimes as an individualizable group of statements, and sometimes as a regulated practice that accounts for a certain number of statements; and have I not allowed this same word ^'discourse^', which should have served as a boundary around the term ^'statement^', to vary as I shifted my analysis or its point of application, as the statement itself faded. from view?

This, then, is the task that now confronts me: to take up the defin­ition of the statement at its very root. And to see whether that defin­ition really was present in my earlier descriptions; to see whether I really was dealing with the statement in my analysis of discursive formations.

On several occasions I have used the term ^'statement^', either to speak of a population of statements (as if I were dealing with individuals or isolated events), or in order to distinguish it from the groups that I called ^'discourses^' (as the part is distinguished from the whole). At first sight, the statement appears as an ultimate, undecomposable element that can be isolated and introduced into a set of relations with other similar elements. A point without a surface, but a point that can be located in planes of division and in specific forms of groupings. A seed that appears on the surface of a tissue of which it is the constituent element. The atom of discourse.

And the problem soon arises: if the statement really is the elem­entary unit of discourse, what does it consist of? What are its distinctive

((91))

features? What boundaries must one accord to it? Is this unity identical with that to which logicians have given the term ^'proposition^', and that which grammarians call a ^'sentence^', or that which ^'analysts^' try to map by the term ^'speech act^'? What place does it occupy among all those unities that the investigation of language (langage) has already revealed? (Even though the theory of these unities is so often incomplete, on account of the difficulty of the problems that they present, and the difficulty in many cases of delimiting them with any degree of rigour.)

I do not think that the necessary and sufficient condition of a state-ment is the presence of a defined propositional structure, or that one can speak of a statement only when there is a proposition. In fact, one can have two perfectly distinct statements, referring to quite different discursive groupings, when one finds only one proposition, possessing only one value, obeying only one group of laws for its construction, and involving the same possibilities of use. ^'No one heard^' and ^'It is true that no one heard^' are indistinguishable from a logical point of view, and cannot be regarded as two different propositions. But in so many statements, these two formations are not equivalent or inter-changeable. They cannot occupy the same place on the plane of dis-course, nor can they belong to exactly the same group of statements. If one finds the formulation ^'No one heard^' in the first line of a novel, we know, until a new order emerges, that it is an observation made either by the author, or by a character (aloud or in the form of an interior monologue); if one finds the second formulation, ^'It is true that no one heard^', one can only be in a group of statements constituting an interior monologue, a silent discussion with. oneself, or a frag­ment of dialogue, a group of questions and answers. In each case, there is the same propositional structure, but there are distinct enunciative characteristics. There may, on the other hand, be complex and doubled propositional forms, or, on the contrary, fragmentary, incomplete propositions, when one is quite obviously dealing with a simple, com­plete, autonomous statement (even if it is part of group of other state­ments): the example ^'The present king of France is bald^' is well known (it can be analysed from a logical point of view only if on.e accepts, in the form of a single statement, two distinct propositions, each of which may be true or false on its own account), or again there is a

((92))

proposition like ^'I am lying^', which can be true only in relation to an assertion on a lower level. The criteria by which one can define the identity of a proposition, distinguish several of them beneath the unity of a formulation, characterize its autonomy or its completion are not valid when one comes to describe the particular unity of a statement.

And what of the sentence? Should we not accept an equivalence between sentence and statement? Wherever there is a grammatically isolable sentence, one can recognize the existence of an independent statement; but, on the other hand, one cannot speak of statement when, beneath the sentence itself, one reaches the level of its constitu­ents. It would he pointless to object, against such an equivalence, that some statements may be composed, outside the canonical form of subject-copula-predicate, of a simple nominal syntagma (^'That man!^') or an adverb ('Absolutely'), or a personal pronoun (^'You!^'). For the grammarians themselves recognize such formulations as independent sentences, even if those formulations have been obtained through a series of transformations on the basis of the subject-predicate schema. Moreover: they recognize as ^'acceptable^' sentences groups of linguistic elements that have not been correctly constructed, providing they are interpretable; on the other hand, they accord the status of grammatical sentences to interpretable groups on condition however that they are correctly formed. With so broad — an.d, in a sense, so lax — a definition of the sentence, it is difficult to see how one is to recognize sentences that are not statements, or statements that are not sentences.

Yet the equivalence is far from being a total one; and it is relatively easy to cite statements that do not correspond to the linguistic structure of sentences. When one finds in a Latin grammar a series of words arranged in a column: amo, amas, amat, one is dealing not with a sen­tence, but with the statement of the different personal inflexions of the present indicative of the verb amare. One may find this example debat­able; one may say that it is a mere artifice of presentation, that this statement is an elliptical, abbreviated sentence, spatialized in a relatively unusual mode, that should he read as the sentence ^'The present indica­tive of the verb amare is amo for the first person^', etc. Other examples, in any case, are less ambiguous: a classificatory table of the botanical species is made up of statements, not sentences (Linnaeus^'s Genera Plantarum is a whole book of statements, in which one can recognize

((93))

only a small number of sentences); a genealogical tree, an accounts book, the calculations of a trade balance are statements; where are the sentences? One can go further: an equation of the nth degree, or the algebraic formula of the law of refraction must be regarded as state­ments: and although they possess a highly rigorous grammaticality (since they are made up of symbols whose meaning is determined by rules of usage, and whose succession is governed by laws of construc­tion), this grammaticality cannot be judged by the same criteria that, in a natural language (Iangue), make it possible to define an acceptable, or interpretable sentence. Lastly, a graph, a growth curve, an age pyramid, a distribution could are all statements: any sentences that may accom­pany them are merely interpretation or commentary; they are in no way an equivalent: this is proved by the fact that, in a great many cases, only an. infinite number of sentences could equal all the elements that are explicitly formulated in this sort of statement. It would not appear to be possible, therefore, to define a statement by the grammatical characteristics of the sentence.

One last possibility remains: at first sight, the most likely of all. Can one not say that there is a statement wherever one can recognize and isolate an act of formulation — something like the speech act referred to by the English analysts? This term does not, of course, refer to the material act of speaking (aloud or to oneself) or of writing (by hand or typewriter); nor does it refer to the intention of the individual who is speaking (the fact that he wants to convince someone else, to be obeyed, to discover the solution to a problem, or to communicate information) ; nor does it refer to the possible result of what he has said (whether he has convinced someone or aroused his suspicion; whether he was listened to and whether his orders were carried out; whether his prayer was heard); what one is referring to is the operation that has been carried out by the formula itself, in its emergence: promise, order, decree, contract, agreement, observation. The speech act is not what took place just prior to the moment when the statement was made (in the author^'s thought or intentions); it is not what might have hap­pened, after the event itself, in. its wake, and the consequences that it gave rise to; it is what occurred by the very fact that a statement was made — and precisely this statement (and no other) in specific circum­stances. Presumably, therefore, one individualization of statements

((94))

refers to the same criteria as the location of acts of formulation: each act is embodied in a statement each statement contains one of those acts. They exist through one another in an exact reciprocal relationship.

Yet such a correlation does not stand up to examination. For one thing, more than a statement is often required to effect a speech act: an oath, a prayer, a contract, a promise, or a demonstration usually require a certain number of distinct formulas or separate sentences: it would be difficult to challenge the right of each of these formulas and sentences to be regarded as a statement on the pretext that they are all imbued with one and the same speech act. In that case, it might be said that the act itself does not remain the same throughout the series of statements; that in a prayer there are as many limited, successive, and juxtaposed acts of prayer as demands formulated by distinct statements; and that in a promise there are as many engagements as sequences that can be individualized into separate statements. But one cannot be satisfied with this answer: first because the act of formulation would no longer serve to define the statement, but, on the contrary, the act of formula-tion would be defined by the statement — which raises problems, and requires criteria of individualization. Moreover, certain speech acts can be regarded as complete in their particular unity only if several state­ments have been made, each in its proper place. These acts are not constituted, therefore, by the series or sum of these statements, by their necessary juxtaposition; they cannot be regarded as being present whole and entire in the least of them, and as renewing themselves with each one. So one cannot establish a hi-univocal relation between the group of statements and that of speech acts either.

When one wishes to individualize statements, one cannot therefore accept unreservedly any of the models borrowed from grammar, logic, or ^'analysis^'. In all three cases, one realizes that the criteria proposed are too numerous and too heavy, that they limit the extent of the statement, and that although the statement sometimes takes on the forms described and adjusts itself to them exactly, it does not always do so: one finds statements lacking in legitimate propositional structure; one finds statements where one cannot recognize a sentence; one finds more statements that one can isolate speech acts. As if the statement were more tenuous, less charged with determinations, less strongly structured, more omnipresent, too, than all these figures; as if it had

((95))

fewer features, and ones less difficult to group together; but as if, by that very fact, it rejected all possibility of describing anything. And this is all the more so, in that it is difficult to see at what level it should be situated, and by what method it should be approached: for all the analyses mentioned above, there is never more than a support, or acci­dental substance: in logical analysis, it is what is left when the prop­ositional structure has been extracted and defined; for grammatical analysis, it is the series of linguistic elements in which one may or may not recognize the form of a sentence; for the analysis of speech acts, it appears as the visible body in which they manifest themselves. In rela­tion to all these descriptive approaches, it plays the role of a residual element, of a mere fact, of irrelevant raw material.

Must we admit in the end that the statement cannot possess a char­acter of its own and that it cannot be adequately defined, in so far as it is, for all analyses of language (langage), the extrinsic material on the basis of which they determine their own object? Must we admit that any series of signs, figures, marks, or traces — whatever their organiza­tion or probability may be — is enough to constitute a statement; and that it is the role of grammar to say whether or not it is a sentence, the role of logic to decide whether or not it contains a propositional form, the role of Analysis to determine what speech act it may embody? In which case, we would have to admit that there is a statement whenever a number of signs are juxtaposed — or even, perhaps — when there is a single sign. The threshold of the statement is the threshold of the existence of signs. Yet even here, things are not so simple, and the meaning of a term like ^`the existence of signs^' requires elucidation. What does one mean when one says that there are signs, and that it is enough for there to be signs for there to he a statement? What special status should be given to that verb to be?

For it is obvious that statements do not exist in the same sense in which a language (longue) exists, and, with that language, a collection of signs defined by their contrasting characteristics and their rules of use; a language in fact is never given in itself, in its totality; it could only he so in a secondary way, in the oblique form of a description that would take it as its object; the signs that make up its elements are forms that are imposed upon statements and control them from within. If there were no statements, the language (longue) would not exist; but no

((96))

statement is indispensable for a language to exist (and one can always posit, in place of any statement, another statement that would in no way modify the language). The language exists only as a system for constructing possible statements; but in another respect, it exists only as a (more or less exhaustive) description obtained from a collection of real statements. Language (longue) and statement are not at the same level of existence; and one cannot say that there are statement in the same way as one says that there are languages (longues). But is it enough, then, that the signs of a language constitute a statement, if they were produced (articulated, drawn, made, traced) in one way or another, if they appeared in a moment of time and in a point in space, if the voice that spoke them or the gesture that formed them gave them the dimen­sions of a material existence? Can the letters of the alphabet written by me haphazardly on to a sheet of paper, as an example of what is not a statement, can the lead characters used for printing books — and one cannot deny their materiality, which has space and volume — can these signs, spread out, visible, manipulable, be reasonably regarded as statements?

When looked at more closely, however, these two examples (the lead characters and the signs that I wrote down on the sheet of paper) are seen to be not quite superposable. This pile of printer^'s characters, which I can hold in my hand, or the letters marked on the keyboard of a typewriter are not statements: at most they are tools with which one can write statements. On the other hand, what are the letters that I write down haphazardly on to a sheet of paper, just as they come to mind, and to show that they cannot, in their disordered state, constitute a statement? What figure do they form? Are they not a table of letters chosen in a contingent way, the statement of an alphabetical series governed by other laws than those of chance? Similarly, the table of random numbers that statisticians sometimes use is a series of numer­ical symbols that are not linked together by any syntactical structure; and yet that series is a statement: that of a group of figures obtained by procedures that eliminate everything that might increase the prob-ability of the succeeding issues. Let us look at the example again: the keyboard of a typewriter is not a statement; but the same series of letters, A, Z, E, R, T, listed in a typewriting manual, is the statement of the alphabetical order adopted by French typewriters. So we are

((97))

presented with a number of negative consequences: a regular linguistic construction is not required in order to form a statement (this state-ment may be made up of a series possessing a minimal probability) ; but neither is it enough to have any material effectuation of linguistic elements, any emergence of signs in time and space, for a statement to appear and to begin to exist. The statement exists therefore neither in the same way as a language (longue) (although it is made up of signs that are definable in their individuality only within a natural or artificial linguistic system), nor in the same way as the object presented to perception (although it is always endowed with a certain materiality, and can always he situated in accordance with spatio-temporal coordinates).

This is not the place to answer the general question of the statement, but the problem can be clarified: the statement is not the same kind of unit as the sentence, the proposition, or the speech act; it cannot be referred therefore to the same criteria; but neither is it the same kind of unit as a material object, with its limits and independence. In its way of being unique (neither entirely linguistic, nor exclusively material), it is indispensable if we want to say whether or not there is a sentence, proposition, or speech act; and whether the sentence is correct (or acceptable, or interpretable), whether the proposition is legitimate and well constructed, whether the speech act fulfils its requirements, and was in fact carried out. We must not seek in the statement a unit that is either long or short, strongly and weakly structured, but one that is caught up, like the others, in a logical, grammatical, locutory nexus. It is not so much one element among others, a division that can he located at a certain level of analysis, as a function that operates vertically in relation to these various units, and which enables one to say of a series of signs whether or not they are present in it. The statement is not therefore a structure (that is, a group of relations between variable elements, thus authorizing a possibly infinite number of concrete models); it is a function of existence that properly belongs to signs and on the basis of which one may then decide, through analysis or intu­ition, whether or not they ^'make sense^', according to what rule they follow one another or are juxtaposed, of what they are the sign, and what sort of act is carried out by their formulation (oral or written). One should not be surprised, then, if one has failed to find structural

((98))

criteria of unity for the statement; this is because it is not in itself a unit, but a function that cuts across a domain of structures and possible unities, and which reveals them, with concrete contents, in time and space.

It is this function that we must now describe as such, that is, in its actual practice, its conditions, the rules that govern it, and the field in which it operates.

((99))