The Archeology of Knowledge Michel Foucault contents part I introduction 3

5 THE FORMATION OF CONCEPTS

Perhaps the family of concepts that emerges in the work of Linnaeus (hut also in that of Ricardo, and in the Grammaire de Port-Royal) may be organized into a coherent whole. Perhaps one might be able to restore the deductive architecture that it forms. In any case, the experiment is worth attempting — and it has been attempted several times. On the other hand, if one takes a broader scale, and chooses as guide-lines such disciplines as grammar, or economics, or the study of living beings, the set of concepts that emerges does not obey such rigorous conditions; their history is not the stone-by-stone construction of an edifice. Should this dispersion be left in its apparent disorder? Or should it be seen as a succession of conceptual systems, each possessing its own organization, and being articulated only against the perman­ence of problems, the continuity of tradition, or the mechanism of influences? Could a law not be found that would account for the successive or simultaneous emergence of disparate concepts? Could a system of occurrence not be found between them that was not a logical systematicity? Rather than wishing to replace concepts in a virtual deductive edifice, one would have to describe the organization of the field of statements where they appeared and circulated.

((63))

(a) This organization involves firstly forms of succession. And among them, the various orderings of enunciative series (whether the order of inferences, successive implications, and demonstrative reasonings; or the order of descriptions, the schemata of generalization or progressive specification to which. they are subject, the spatial distributions that they cover; or the order of the descriptive accounts, and the way in which the events of the time are distributed in the linear succession of the statements); the various types of dependence of the statements (which are not always either identical or superposable on the manifest succes­sions of the series of statements: this is the case in the dependences of hypothesis/verification, assertion/critique, general law/particular application; the various rhetorical schemata according to which groups of statements may he combined, (how descriptions, deductions, def­initions, whose succession characterizes the architecture of a text, are linked together). Take, for example, the case of Natural History in the Classical period: it does not use the same concepts as in the sixteenth century; certain of the older concepts (genus, species, signs) are used in different ways; new concepts (like that of structure) appear; and others (like that of organism) are formed later. But what was altered in the seventeenth century, and was to govern the appearance and recur­rence of concepts, for the whole of Natural History, was the general arrangement of the statements, their successive arrangement in particu­lar wholes; it was the way in which one wrote down what one observed and, by means of a series of statements, recreated a perceptual process; it was the relation and interplay of subordinations between describing, articulating into distinctive features, characterizing, and classifying; it was the reciprocal position of particular observations and general principles; it was the system of dependence between what one learnt, what one saw, what one deduced, what one accepted as prob-able, and what one postulated. In the seventeenth and eighteenth cen­turies, Natural History was not simply a form of knowledge that gave a new definition to concepts like ^'genus^' or ^'character^', and which intro­duced new concepts like that of ^'natural classification^' or ^'mammal^'; above all, it was a set of rules for arranging statements in series, an obligatory set of schemata of dependence, of order, and of successions, in which the recurrent elements that may have value as concepts were distributed.

((64))

(b) The configuration of the enunciative field also involves forms of coexistence. These outline first a field of presence (by which is understood all statements formulated elsewhere and taken up in a discourse, acknow­ledged to be truthful, involving exact description, well-founded rea­soning, or necessary presupposition); we must also give our attention to those that are criticized, discussed, and judged, as well as those that are rejected or excluded); in this field of presence, the relations estab­lished may be of the order of experimental verification, logical valid­ation, mere repetition, acceptance justified by tradition and authority, commentary, a search for hidden meanings, the analysis of error; these relations may be explicit (and sometimes formulated in types of specialized statements: references, critical discussions), or implicit and present in ordinary statements. Again, it is easy to see that the field of presence of Natural History in the Classical period does not obey the same forms, or the same criteria of choice, or the same principles of exclusion, as in the period when Aldrovandi was collecting in one and the same text everything that had been seen, observed, recounted, passed on innumerable times by word of mouth, and even imagined by the poets, on the subject of monsters. Distinct from this field of pres­ence one may also describe a field of concomitance (this includes statements that concern quite different domains of objects, and belong to quite different domains of objects, and belong to quite different types of discourse, but which are active among the statements studied here, either because they serve as analogical confirmation, or because they serve as a general principle and as premises accepted by a reasoning, or because they serve as models that can be transferred to other contents, or because they function as a higher authority than that to which at least certain propositions are presented and subjected): thus the field of concomitance of the Natural History of the period of Linnaeus and Buffon is defined by a number of relations with cosmology, the history of the earth, philosophy, theology, scripture and biblical exegesis, mathematics (in the very general form of a science of order); and all these relations distinguish it from both the discourse of the sixteenth-century naturalists and that of the nineteenth-century biologists. Lastly, the enunciative field involves what might be called a field of memory (statements that are no longer accepted or discussed, and which con­sequently no longer define either a body of truth or a domain of
((65))

validity, but in. relation to which relations of filiation, genesis, trans-formation, continuity, and historical discontinuity can be established): thus the field of memory of Natural History, since Tournefort, seems particularly restricted and impoverished in its forms when compared with the broad, cumulative, and very specific field of memory pos­sessed by nineteenth- and twentieth-century biology; on the other hand, it seems much better defined and better articulated than the field of memory surrounding the history of plants and animals in the Renaissance: for at that time it could scarcely be distinguished from the field of presence; they had the same extension and the same form, and involved the same relations.

(c) Lastly, we may define the procedures of intervention that may be legit­imately applied to statements. These procedures are not in fact the same for all discursive formations; those that are used (to the exclu­sion of all others), the relations that link them and the unity thus created make it possible to specify each one. These procedures may appear: in techniques of rewriting (like those, for example, that enabled the naturalists of the Classical period to rewrite linear descriptions in classificatory tables that have neither the same laws nor the same configuration as the lists and groups of kinship established in the Middle Ages and during the Renaissance); in methods of transcribing statements (articulated in the natural language) according to a more or less formalized and artificial language (the project, and to a certain extent the realization, of such a language is to be found in Linnaeus and Adanson); the modes of translating quantitative statements into quali­tative formulations and vice versa (the establishment of relations between purely perceptual measurements and descriptions); the means used to increase the approximation of statements and to refine their exactitude (structural analysis according to the form, number, arrangement, and size of the elements has made it possible, since Tournefort, to achieve a closer and above all more constant approxi­mation of descriptive statements); the way in which one delimits once again — by extension or restriction. — the domain of validity of state­ments (the enunciation of structural characters was restricted in the period between Tournefort and Linnaeus, then enlarged in that between Buffon and Jussieu); the way in which one transfers a type of

((66))

statement from one field of application to another (like the transfer­ence from vegetal characterization to animal taxonomy; or from the description of superficial characters to the internal elements of the organism); the methods of systematizing propositions that already exist, because they have been previously formulated, but in a separated state; or again the methods of redistributing statements that are already linked together, but which one rearranges in a new systematic whole (as Adanson takes up the natural characterizations that had been made before, either by himself or by others, and placed them in a group of artificial descriptions, the schema of which he had previously worked out on the basis of some abstract combinatory).

These elements that I am proposing to analyse are of rather different kinds. Some constitute rules of formal construction, others rhetorical practices; some define the internal configuration of a text, others the modes of relation and interference between different texts; some are characteristic of a particular period, others have a distant origin and far-reaching chronological import. But what properly belongs to a discursive formation and what makes it possible to delimit the group of concepts, disparate as they may be, that are specific to it, is the way in which these different elements are related to one another: the way in which, for example, the ordering of descriptions or accounts is linked to the techniques of rewriting; the way in which the field of memory is linked to the forms of hierarchy and subordination that govern the statements of a text; the way in which the modes of approximation and development of the statements are linked to the modes of criticism, commentary and interpretation of previously formulated statements, etc. It is this group of relations that constitutes a system of conceptual formation.

The description of such a system could not be valid for a direct, immediate description of the concepts themselves. My intention is not to carry out an exhaustive observation of them, to establish the charac­teristics that they may have in common, to undertake a classification of them, to measure their internal coherence, or to test their mutual com­patibility; I do not wish to take as an object of analysis the conceptual architecture of an isolated text, an individual oeuvre, or a science at a particular moment in time. One stands back in relation to this manifest

((67))

set of concepts; and one tries to determine according to what schemata (of series, simultaneous groupings, linear or reciprocal modification) the statements may be linked to one another in a type of discourse; one tries in this way to discover how the recurrent elements of statements can reappear, dissociate, recompose, gain in extension or determin­ation, be taken up into new logical structures, acquire, on the other hand, new semantic contents, and constitute partial organizations among themselves. These schemata make it possible to describe — not the laws of the internal construction of concepts, not their progressive and individual genesis in the mind of man — but their anonymous dispersion through texts, hooks, and oeuvres. A dispersion that character­izes a type of discourse, and which defines, between concepts, forms of deduction, derivation, and coherence, but also of incompatibility, intersection, substitution, exclusion, mutual alteration, displacement, etc. Such an analysis, then, concerns, at a kind of preconceptual level, the field in which concepts can coexist and the rules to which this field is subjected.

In order to define more precisely what I mean by ^'preconceptual^', I shall take the example of the four ^'theoretical schemata^', studied in my book The Order of Things, and which characterize General Grammar in the seventeenth and eighteenth centuries. These four schemata — attribu­tion, articulation, designation, and dervation — do not designate con­cepts that were in fact used by the Classical grammarians; nor do they make it possible to reconstitute, over and above different grammatical works, a sort of more general, more abstract, more impoverished system, but discover, by that very fact, the profound compatibility of these different, apparently opposed systems. They make it possible to describe:
1. How the different grammatical analyses can be ordered and deployed; and what forms of succession are possible between analyses of the noun, analyses of the verb, and analyses of the adjective, those that concern phonetics and those that concern syntax, those that con­cern the original language (longue), and those that project an artificial language (longue). These different orders are laid down by the relations of dependence that may be observed between the theories of attribution, articulation, designation, and derivation.

((68))

2. 
   How General Grammar defines a domain of validity for itself (according to what criteria one may discuss the truth or falsehood of a proposition); how it constitutes a domain of normativity for itself (according to what criteria one may exclude certain statements as being irrelevant to the discourse, or as inessential and marginal, or as non-scientific); how it constitutes a domain of actuality for itself (comprising acquired solutions, defining present problems, situating concepts and affirmations that have fallen into disuse).
   

2. 
   What relations General Grammar has with Mathesis (with Car­tesian and post-Cartesian algebra, with th.e project of a general science of order), with the philosophical analysis of representation and the theory of signs, with Natural History, the problems of characterization and taxonomy, with the Analysis of Wealth and the problems of the arbitrary signs of measurem.en.t and exchange: by marking out these relations one may determine the ways by which the circulation, the transfer and the modification of concepts, the alteration of their form or changes in their field of application, are made possible between one domain and another. The network formed by the four theoretical segments does not define the logical architecture of all the concepts used by grammarians; it outlines the regular space of their formation.
   

2. 
   How the various conceptions of the verb ^'to be^', of the copula, of the verbal radical and the flexional ending (for a theoretical schema of attribution) were simultaneously or successively possible (under the form of alternative choice, modification, or substitution); the various conceptions of the phonetic elements, of the alphabet, of the name, of substantives and adjectives (for a theoretical schema of articulation); the various concepts of proper noun. and common noun, demonstrative, nominal root, syllable or expressive sonority (for the theoretical segment of designation); the various concepts of original and derived language (langage), metaphor and figure, poetic language (langage) (for the theoretical segment of derivation).
   

The ^'preconceptual^' level that we have uncovered refers neither to a horizon of ideality nor to an empirical genesis of abstractions. On the one hand, it is not a horizon of ideality, placed, discovered, or

((69))

established by a founding gesture — and one that is so original that it eludes all chronological insertion; it is not an inexhaustible a priori at the confines of history, set back both because it eludes all beginning, all genetic restitution, and because it could never he contemporary with itself in an explicit totality. In fact one does not pose the question at the level of discourse itself, which is not external translation, but the locus of emergence of concepts; one does not attach the constants of discourse to the ideal structures of the concept, but one describes the conceptual network on the basis of the intrinsic regularities of discourse; one does not subject the multiplicity of statements to the coherence of concepts, and this coherence to the silent recollection of a meta-historical ideality; one establishes the inverse series: one replaces the pure aims of non-contradiction in a complex network of con­ceptual compatibility and incompatibility; and one relates this com­plexity to the rules that characterize a particular discursive practice. By that very fact, it is no longer necessary to appeal to the themes of an endlessly withdrawing origin and and inexhaustible horizon: the organization of a group of rules in the practice of discourse, even if it does not constitute an event so easy to situate as a formulation or a discovery, may he determined, however, in the element of history; and if it is inexhaustible, it is by that very fact that the perfectly describable system that it constitutes takes account of a very considerable set of concepts and a very large number of transformations that affect both these concepts and their relations. Instead of outlining a horizon that rises from the depths of history and maintains itself through history, the ^'preconceptual^' thus described is, on the contrary, at the most ^'superficial^' level (at the level of discourse), the group of rules that in fact operate within it.

Nor is it a genesis of abstractions, trying to rediscover the series of operations that have made it possible to constitute them: overall intu­itions, discoveries of particular cases, the disconnexion of imaginary themes, the encountering of theoretical or technical obstacles, succes-sive borrowings from traditional models, definition of the adequate formal structure, etc. In the analysis proposed here, the rules of forma-tion operate not only in the mind or consciousness of individuals, but in discourse itself; they operate therefore, according to a sort of uni­form anonymity, on all individuals who undertake to speak in this

((70))

discursive field. On the other hand, one does not suppose them to be universally valid for every domain; one always describes them in par­ticular discursive fields, and one does not accord them at the outset indefinite possibilities of extension. The most one can do is to make a systematic comparison, from one region to another, of the rules for the formation of concepts: it is in this way that I have tried to uncover the identities and differences that may be presented by these groups of rules in the General Grammar, the Natural History, and the Analysis of Wealth of the Classical period. These groups of rules are specific enough in each of these domains to characterize a particular, well-individualized discursive formation; but they offer enough analogies for us to see these various formations form a wider discursive grouping at a higher level. In any case, the rules governing the formation of concepts, however generalized the concepts may be, are not the result, laid down in history and deposited in. the depth of collective customs, of operations carried out by individuals; they do not constitute the bare schema of a whole obscure work, in the course of which concepts would be made to emerge through illusions, prejudices, errors, and traditions. The preconceptual field allows the emergence of the dis-cursive regularities and constraints that have made possible the hetero­geneous multiplicity of concepts, and, beyond these the profusion of the themes, beliefs, and representations with which one usually deals when one is writing the history of ideas.

In order to analyse the rules for the formation of objects, one must neither, as we have seen, embody them in things, nor relate them to the domain of words; in order to analyse the formation of enunciative types, one must relate them neither to the knowing subject, nor to a psychological individuality. Similarly, to analyse the formation of con­cepts, one must relate them neither to the horizon of ideality, nor to the empirical progress of ideas.