The Archeology of Knowledge Michel Foucault contents part I introduction 3

6. THE FORMATION OF STRATEGIES

Such discourses as economics, medicine, grammar, the science of liv-ing beings give rise to certain organizations of concepts, certain regroupings of objects, certain types of enunciation, which form, according to their degree of coherence, rigour, and stability, themes or theories: the theme, in eighteenth-century grammar, of an original language (longue) from which all others derive, and of which all others carry within themselves a sometimes decipherable memory; a theory, in nineteenth-century philology, of a kinship between all the Indo-European languages, and of an archaic idiom that served as a common starting-point; a theme, in the eighteenth century, of an evolution of the species deploying in time the continuity of nature, and explaining the present gaps in the taxonomic table; a theory, propounded by the Physiocrats, of a circulation of wealth on the basis of agricultural pro­duction. Whatever their formal. level may be, I shall call these themes and theories ^`strategies^'. The problem is to discover how they are dis­tributed in history. Is it necessity that links them together, makes them invisible, calls them to their right places one after another, and makes of them successive solutions to one and the same problem? Or chance encounters between ideas of different origin, influences, discoveries,
((72))

speculative climates, theoretical models that the patience or genius of individuals arranges into more or less well-constituted wholes? Or can one find a regularity between them and define the common system of their formation?

As for the analysis of these strategies, I can hardly enter into great detail. The reason is simple enough: in the various discursive domains, which I have tried to sketch out — rather hesitantly no doubt, and, especially at the beginning, with inadequate method­ological control — the problem was to describe in each case the discursive formation in all its dimensions, and according to its own characteristics: it was necessary therefore to describe each time the rules for the formation of objects, modalities of statement, concepts, and theoretical choices. But it turned out that the difficult point of the analysis, and the one that demanded greatest attention, was not the same in each case. In Madness and Civilization, I was dealing with a discursive formation whose theoretical points of choice were fairly easy to locate, whose conceptual systems were relatively uncomplex and few in number, and whose enunciative rules were fairly homo­geneous and repetitive; on the other hand, the problem lay in the emergence of a whole group of highly complex, interwoven objects; it was necessary above all to describe the formation of these objects, in order to locate in its specificity the whole of psychiatric discourse. In Naissance de la clinique, the essential point of the research was the way in which, at the end of the eighteenth and the beginning of the nineteenth century, the enunciative forms of medical discourse had been modified; the analysis was concerned therefore less with the formation of conceptual systems, or the formation of theoretical choices, than with the status, the institutional siting, the situation, and the modes of insertion used by the discoursing subject. Lastly, in The Order of Things, my attention was concentrated mainly on the net-works of concepts and their rules of formation (identical or different) as they could be located in General Grammar, Natural History, and the Analysis of Wealth. The place, and the implications, of the stra­tegic choices were indicated (whether, for example, in the case of Linnaeus and Buffon, or the Physiocrats and the Utilitarists); but I did little more than locate them, and my analysis scarcely touched on their formation. Let us say that a fuller analysis of theoretical choices

((73))

must be left until a later study, in which I shall be able to give it my whole attention.

For the moment, the most that I can do is to indicate the directions in which the research will proceed. These might be summarized thus:

I . Determine the possible points of diffraction of discourse. These points are characterized in the first instance as points of incompatibility: two objects, or two types of enunciation, or two concepts may appear, in the same discursive formation, without being able to enter — under pain of manifest contradiction or inconsequence — the same series of statements. They are then characterized as points of equivalence: the two incompatible elements are formed in the same way and on the basis of the same rules; the conditions of their appearance are identical; they are situated at the same level; and instead of constituting a mere defect of coherence, they form an alternative: even if, chronologically speak-ing, they do not appear at the same time, even if they do not have the same importance, and if they were not equally represented in the popu­lation of effective statements, they appear in the form of ^'either . . . or^'. Lastly, they are characterized as link points of systematization: on the basis of each of these equivalent, yet incompatible elements, a coherent series of objects, forms of statement, and concepts has been derived (with, in each series, possible new points of incompatibility). In other words, the dispersions studied at previous levels do not simply constitute gaps, non-identities, discontinuous series; they come to form discursive sub-groups — those very sub-groups that are usually regarded as being of major importance, as if they were the immediate unity and raw material out of which larger discursive groups (^'theories^', ^'concep­tions^', ^'themes^') are formed. For example, one does not consider, in an analysis of this kind, that the Analysis of Wealth, in the eighteenth century, was the result (by way of simultaneous composition or chronological succession) of several different conceptions of coinage, of the exchange of objects of need, of the formation of value and prices, or of ground rent; one does not consider that it is made up of the ideas of Cantillon, taking up from those of Petty, of Law^'s experi­ence reflected by various theoreticians in turn, and of the Physiocratic system opposing Utilitarist conceptions. One describes it rather as a unity of distribution that opens a field of possible options, and enables

((74))

various mutually exclusive architectures to appear side by side or in turn.
2. But all the possible alternatives are not in fact realized: there are a good many partial groups, regional compatibilities, and coherent architectures that might have emerged, yet did not do so. In order to account for the choices that were made out of all those that could have been made (and those alone), one must describe the specific author­ities that guided one^'s choice. Well to the fore is the role played by the discourse being studied in relation to those that are contemporary with it or related to it. One must study therefore the economy of the discursive constellation to which it belongs. It may in fact play the role of a formal system of which other discourses are applications with various seman­tic fields; it may, on the other hand, he that of a concrete model that must be applied to other discourses at a higher level of abstraction (thus General Grammar, in the seventeenth and eighteenth centuries, appears as a particular model of the general theory of signs and repre­sentation). The discourse under study may also be in a relation of analogy, opposition, or complementarity with certain other discourses (there is, for example, a relation of analogy, in the Classical period, between the Analysis of Wealth and Natural History; the first is to the representation of need and desire what the second is to the representa­tion of perceptions and judgements; one may also note that Natural History and General Grammar are opposed to one another in the same way as a theory of natural characters and a theory of conventional signs; both, in turn, are opposed to the Analysis of Wealth just as the study of qualitative signs is opposed to that of the quantitative signs of measurement; each, in fact, develops one of the three complementary roles of the representative sign: designation, classification, exchange). Lastly, one may describe between several discourses relations of mutual delimitation, each giving the other the distinctive marks of its singular­ity by the differentiation of its domain of application (as in the case of psychiatry and organic medicine which were virtually not dis­tinguished from one another before the end of the eighteenth century, and which established from that moment a gap that has since character­ized them). This whole group of relations forms a principle of determination that permits or excludes, within a given discourse, a

((75))

certain number of statements: these are conceptual systematizations, enunciative series, groups and organizations of objects that might have been possible (and of which nothing can justify the absence at the level of their own rules of formation), but which are excluded by a dis-cursive constellation at a higher level and in a broader space. A dis-cursive formation does not occupy therefore all the possible volume that is opened up to it of right by the systems of formation of its objects, its enunciations, and its concepts; it is essentially incomplete, owing to the system of formation of its strategic choices. Hence the fact that, taken up again, placed, and interpreted in a new constellation, a given discursive formation may reveal new possibilities (thus in the present distribution of scientific discourses, the Grammar of Port-Royal or the taxonomy of Linnaeus may free elements that, in relation to them, are both intrinsic and new); but we are not dealing with a silent content that has remained implicit, that has been said and yet not said, and which constitutes beneath manifest statements a sort of sub-discourse that is more fundamental, and which is now emerging at last into the light of day; what we are dealing with is a modification in the principle of exclusion and the principle of the possibility of choices; a modification that is due to an insertion in a new discursive constellation.
3. The determination of the theoretical choices that were actually made is also dependent upon another authority. This authority is char­acterized first by the function that the discourse under study must carry out in a field of non-discursive practices. Thus General Grammar played a role in pedagogic practice; in a much more obvious, and much more important way, the Analysis of Wealth played a role not only in the political and economic decisions of governments, but in the scarcely conceptualized, scarcely theoretized, daily practice of emergent capital-ism, and in the social and political struggles that characterized the Classical period. This authority also involves the rules and processes of appropriation of discourse: for in our societies (and no doubt in many others) the property of discourse — in the sense of the right to speak, ability to understand, licit and immediate access to the corpus of already formulated statements, and the capacity to invest this discourse in decisions, institutions, or practices — is in fact confined (sometimes

((76))

with the addition of legal sanctions) to a particular group of individuals; in the bourgeois societies that we have known since the sixteenth century, economic discourse has never been a common dis-course (no more than medical or literary discourse, though in a differ­ent way). Lastly, this authority is characterized by the possible positions of desire in relation to discourse: discourse may in fact be the place for a phan­tasmatic representation, an element of symbolization, a form of the forbidden, an instrument of derived satisfaction (this possibility of being in relation with desire is not simply the fact of the poetic, fic­tional, or imaginary practice of discourse: the discourses on wealth, on language (langage), on nature, on madness, on life and death, and many others, perhaps, that are much more abstract, may occupy very specific positions in relation to desire). In any case, the analysis of this authority must show that neither the relation of discourse to desire, nor the processes of its appropriation, nor its role among non-discursive prac­tices is extrinsic to its unity, its characterization, and the laws of its formation. They are not disturbing elements which, superposing themselves upon its pure, neutral, atemporal, silent form, suppress its true voice and emit in its place a travestied discourse, but, on the contrary, its formative elements.
A discursive formation will be individualized if one can define the system of formation of the different strategies that are deployed in it; in other words, if one can show how they all derive (in spite of their sometimes extreme diversity, and in spite of their dispersion in time) from the same set of relations. For example, the Analysis of Wealth in the seventeenth and eighteenth centuries is characterized by the system that could form both Colbert^'s mercantilism and Cantillon^'s ^'neo-mercantilism^'; Law^'s strategy and that of Paris-Duverney; the Physio­cratic option and the Utilitarist option. And one will have defined this system if one can describe how the points of diffraction of economic discourse derive from one another, regulate one another, and are involved with one another (how a point of choice about prices derives from a decision about the concept of value); how the choices made depend on the general constellation in which economic discourse fig­ures (the choice in favour of coinage-sign is linked to the place occupied by the Analysis of Wealth, beside the theory of language (langage), the

((77))

analysis of representations, mathesis, and the science of order) ; how these choices are linked to the function carried out by economic dis-course in the practice of emergent capitalism, the process of appropri­ation of which it is the object on the part of the bourgeoisie, the role that it can play in the realization of interests and desires. Economic discourse, in the Classical period, is defined by a certain constant way of relating possibilities of systematization interior to a discourse, other discourses that are exterior to it, and a whole non-discursive field of practices, appropriation, interests, and desires.

It should be noted that the strategies thus described are not rooted, anterior to discourse, in the silent depths of a choice that is both preliminary and fundamental. All these groups of discourses that are to be described are not the expression of a world-view that has been coined in the form of words, nor the hypocritical translation of an interest masquerading under the pretext of a theory: the Natural His-tory of the Classical period is more than a confrontation, in the limbo that precedes manifest history, between a (Linnaean) view of a static, ordered, compartmented universe that is subjected from its very beginnings to the classificatory table, and the still confused perception of a nature that is the heir to time, with all the weight of its accidents, and open to the possibility of an evolution; similarly, the Analysis of Wealth is more than. the conflict of interest between a bourgeoisie that has become a land-owning class, expressing its economic or political demands through the Physiocrats, and a commercial bourgeoisie that demands protectionist or liberal measures through the Utilitarists. Nei­ther the Analysis of Wealth, nor Natural History, if one questions them at the level of their existence, their unity, their permanence, an.d their transformations, may be regarded as the sum of these various options. On the contrary, these options must be described as systematically different ways of treating objects of discourse (of delimiting them, regrouping or separating them, linking them together and making them derive from one another), of arranging forms of enunciation (of choosing them, placing them, constituting series, composing them into great rhetorical unities), of manipulating concepts (of giving them rules for their use, inserting them into regional coherences, and thus constituting conceptual architectures). These options are not seeds of discourse (in which discourses are determined in advance and

((78))

prefigured in a quasi-microscopic form); they are regulated ways (and describable as such) of practising the possibilities of discourse.

But these strategies must not be analysed either as secondary elem­ents that are superposed on a discursive rationality that is, of right, independent of them.. There is not (or, at least, as far as the historical description whose possibility we are tracing here is concerned) a sort of ideal discourse that is both ultimate and timeless, and which choices, extrinsic in origin, have perverted, disturbed, suppressed, or thrust towards a possibly distant future; one must not suppose for example that it holds on nature or on the economy two superposed and intermingled discourses: one that proceeds slowly, accumulating its acquisitions and gradually achieving completion (a true discourse, but one that exists in its pure state only at the teleological confines of history); the other forever disintegrating, recommenced, in perpetual rupture with itself, composed of heterogeneous fragments (a discourse of opinion that history, in the course of time, throws back into the past). There is no natural taxonomy that has been exact, fixism excepted; there is no economy of exchange and use that has been true, without the preferences and illusions of a mercantile bourgeoisie. Classical taxonomy or the Analysis of Wealth, in the form in which they actually existed, and constituted historical figures, involve, in an articulated but indissociable system, objects, statements, concepts, and theoretical choices. And just as one must not relate the formation of objects either to words or to things, nor that of statements either to the pure form of knowledge or to the psychological subject, nor that of concepts either to the structure of ideality or to the succession of ideas, one must not relate the formation of theoretical choices either to a fundamental project or to the secondary play of opinions.