Review The Archaeology Of Knowledge

The Archaeology of Knowledge Review The Archaeology of Knowledge, published in 1969, is Foucault's methodological reflection on his major works in the 1960s: Madness and Civilization, The Birth of the Clinic, and The Order of Things. In it he attempts a systematic theory of historiography (or at least he wants to establish the possibility of such a theory) based on new developments in the history profession. AK is a difficult text, and Foucault will ultimately modify a key thesis of archaeology as well as make archaeology a subordinate rather than independent method. We will trace AK's consistent anti-phenomenological stance as well as its affinities with structuralism; our major focus will be on the way Foucault portrays the relation of discursive and non-discursive practices. As we will see, the approach adopted here--that discursive practices integrate non-discursive practices in producing an intelligible order of discourse independent of any causal action by nondiscursive practices--will be revised in the work of the 70s as the interplay of power/knowledge. In other words, AK reveals that archaeology has a single, fixed virtual/transcendental realm governing the production of discourse. Introduction Two trends in writing history Foucault begins by noting two trends in contemporary historiography. In history proper, as exemplified in the work of Fernand Braudel and the Annales school, the focus is on the abiding patterns of "material civilization" (trade routes, crop distribution and the like) that went to make up the longue durée in Europe. But in the histories of ideas, science, philosophy, thought, and literature, the focus is on different types of ruptures and discontinuities, of "transformations that serve as new foundations" (5). Foucault mentions Bachelard, Canguilhem, Serres, Guéroult, Althusser, and structuralist literary criticism here as examples of a historiography that concentrates not on tradition but on limits. In our terms, then, discontinuity is a bifurcation, a switch from one virtual pattern to another. The questioning of the document Now we cannot think that Foucault is only interested in discontinuity: the changed foci of both types of historiography are mere surface effects of a deeper change, "the questioning of the document", or in other words, the rejection of treating historical evidence as "the language of a voice since reduced to silence," (6) which interpretation will reawaken. Historiography is no longer a "collective consciousness" that tries to refresh its memory. Rather, contemporary historiography is the transformation of the document into the monument, so that contemporary historiography "aspires to the condition of archaeology, to the instrinsic description of the document" (7). In other words, contemporary historiography will reject the handed down unities of anthropological historiography and strive to constitute its own "unities, totalities, series, relations" (7). The term "series" is particulary important to understand. Foucault says "the problem now is to constitute series: a) to define the elements proper to each series; b) to fix its boundaries; c) to reveal its own specific types of relations, to formulate its laws; d) and, beyond this, to describe the relations between different series, thus constituting series of series, or 'tables'..." (7-8). Let's constitute a series: (a little web surfing came up with these tidbits [a rather mysterious unit of information!]: the hand (only used to measure horses), the foot (the length of a foot), the yard (the distance between Henry I's [1068-1135] nose and the tip of his outstretched arm), the furlong (220 yards, the length of a furrow in a field), the mile (derived from the Roman "mille passus" or 1000 double steps; originally 5000

feet as in the Roman definition [1 "passus" = 5 feet]; later, it took 5280 feet to accomodate exactly 8 furlongs). Let's take up each of Foucault's requirements in turn. a) We define the elements of this series as "traditional English distance measurements." It is a multiplicity or assemblage, a hetereogenous constellation that achieves a certain consistency enabling it to survive for a while (it is now under attack by the metric constellation). We can see them as a subset of the virtual realm of all possible distance measurements. b) To fix its boundaries we would have to specify a time period in which these were in common use. c) To reveal the relations between these elements, the law of the series, we would figure out why these were selected, why these virtual singularities were actualized together (the connective synthesis, to use D/G terms). Thus we would have to consider how for the first three units we have the human body as basis (or more precisely, the body of an adult noble male taken as the standard), then an agricultural term related to the man/horse/plow/manor machine [thus a convergence of the pure body series with complex series linking other singularities], and then a military term [itself articulated with the body series]. Thus here we are dealing with conjunctive syntheses that makes the human body series converge with the complex agricultural and military series [these are already, as we have noted, connective and conjunctive]. We would then have to ponder the redefinition of the military term by the agricultural term [a form of capture and re-interpretation, as studied in genealogy, or content and expression to use D/G terms], by taking into account the English/Norman social dichotomy. d) To construct a "table" that determines the relations between series would be to map out all the virtual realm. i) We would map out the possible conjunctive syntheses that determine convergent series (for instance, we could pick the time measurement series and then consider its relations with the distance series in the velocity table: how many miles could a Roman legion march in an hour/day/week?); ii) we would also map out the possible disjunctive syntheses that affirm the coexistence of divergent series rather than excluding them (for instance, we would examine the "schizophrenia" of a society that uses both English and metric measurements). The constitution of series, rather than seeking the relations of pre-given series (kings, popes, battles, treaties), is the first of four consequences of the questioning of the document. The others are: 1) the prominence of the notion of discontinuity; 2) the move from a total to a general history: total history is centralized on the principle or form of a society (e.g. how everything in England at a certain period of time is "feudal") while general history aims to discover relations between divergent series, thus "dispersing" a society (along its "lines of flight" as Deleuze might say) [NB: totality and generality are also important terms in the post-phenomenological movement, especially for Levinas and Derrida]; 3) and a set of methodological problems, including a) that of the reference of historians to "events, institutions, practices" (10) [the problem of power/knowledge], and b) the problem of determining the relations that characterize a group. Among these relations, which include "numerical or logical relations; functional, causal or analogical relations", we also find that of "the relation of 'signifier' (signifiant) to the 'signified' (signifié)" (11). What's important to notice here is the way Foucault puts the signifier/signified relation as only one of a number of possible relations for constituting groups for study in historiography. This is a major departure from structuralist practice, for which this pair was the privileged model. The inclusion of "functional and causal" relations will be an important clue to the way Foucault eventually works out a notion of language as a trigger of real production (as does Deleuze) rather than as a non-referential web of signs always slipping away from presence. After brief remarks on the relation of historical methodology with the "philosophy of history," and with structuralism, Foucault traces the origin of the "discontinuist" mutation in historiography to Marx. The difficulty in accomplishing this transformation is due to a reluctance to "conceive of the Other in the time of our own thought," so that continuity in history is the shelter of consciousness and promises the [re-] appropriation of historical difference, that is, the giving of meaning to history, the projects of Kant and Hegel (and Marx, according to some). [We have seen Foucault trace the "retreat and return of the origin" in OT.] The various reactions, misreadings, and ignoring of the new history (especially regarding Marx and Nietzsche) are just the cries mourning the death of history tied to the subject and his constitution of

edifying historical narratives of progress (Hegel) or decline (Heidegger). Foucault now situates his project in AK relative to the new historiography. After "imperfect sketches" in his previous works, Foucault is now attempting a systematic treatment. To avoid misunderstandings, Foucault makes three observations: AK is not a transfer of structuralist methods to history but the attempt to uncover the principles and consequences of an autochthonous transformation in historiography that resonates but is not identical with structuralism; Foucault will not use cultural totalities to impose on history a structuralist analysis [recall that the positing of a totality is a key requisite of structuralist practice]; AK poses a dual relation with F's previous works: a) a general formulation of the tools used therein; b) a definition of a rigorously non-anthropological method [F's previous works had precisely delimited the historical possibility--the archaeology--of the anthropologism of modern thought]. Specifically, F shows the way AK corrects the problems of his previous works. MC flirted with an "anonymous and general subject of history" [recall the talk about the "experience of madness in the Classical Age" and so on]; BC's emphasis on structure threatened to obscure the [historical, institutional] specificity of the question of the birth of the clinic; OT's "absence of methodological signposting" gave the impression of dealing with cultural totalities [but OT's sins are not only those of omission but also comission, as there are plenty of passages concerning the "entire episteme of Western culture" (54). To correct these flaws, AK will be self-reflective (if not ironically so); defined by the "exteriority of its vicinity" (how it is not like its rival theories); and a definition of the "blank space from which I speak" (but this seems, as Dreyfus and Rabinow note, to involve Foucault in the anthropological doubles--or at least the first two--analyzed in OT). Foucault ends the Introduction with a small rehearsed dialogue between himself and a critic. The critic complains that F is only setting the stage for another chameleon act in his next book, in which he will appear in yet another guise (as in fact he will, as a genealogist, 6 years later with DP). Foucault replies that the process of writing a book is for him not a matter of transferring information, of telling others what he already knows, but of entering a "labyrinth ... in which I can lose myself ..." Here the relation of writing to identity is posed in terms of losing identity. In Deleuzean terms, Foucault is here positing writing as the exploration of the virtual ("the labyrinth") by a nomadic subject. This attempt at freeing oneself from one's usual patterns (one's identity) is of course what the "bureaucrats" and the "police" seek to regulate with their concern with one's "identification papers." The question we should ask ourselves is whether we can take the risk of losing our identity (learning to think otherwise, meeting the Other, constructing a plane of immanence by taking some lines of flight) when we read and write on Foucault. The Discursive Regularities The Unities of Discourse Here Foucault indicates the unities of discourse assumed by standard history of ideas that must be suspended so that the regularities of the dispersion of discourse are revealed: tradition, influence, development and evolution, spirit, book, oeuvre. Foucault indicates that it is the later two unities that must be suspended above all. Both the material and discursive unity of the book are questionable: the material unity because of the diversity of genres (collections, anthologies, volumes in series, etc.) and the discursive unity because "it is caught up in a system of references to other books, other texts, other sentences: it is a node in a network." Thus the unity of the book is "variable and relative ... [to a] complex field of discourse" (23). The oeuvre is even more complex, Foucault writes, because it presupposes a notion of the unity of an author's intention. However, Foucault shows-- using arguments he will later examine in more detail in "What is an Author?"--there are a number of variables at work here that expose the historical arbitrariness or at least questionableness of this interpretation.

After warning the audience not to search for a secret origin to discourse or for what is "already-said / never-said," Foucault states his positive aim: the reception of discourse in its "sudden irruption ... that punctuality ... that temporal dispersion ... as and when it occurs" (25). We now can grasp "a population of events in the space of discourse in general" (27). On this basis one now sees the project of "a pure description of discursive events as the horizon for the search for the unities that form within it" (27). In our terms, Foucault is proposing constructing the map of the virtual realm of discourse as it has appeared in his target domains. The suspension of pre-given unities ensures the grasp of actual discourse in its proper dispersion; the map of the virtual (that which accounts for the dispersion of those "events in the space of discourse in general") will give the rules for the dispersion of actual statements. Foucault's project is to be distinguished from linguistic analysis, which generates the system of possible statements, "an infinite number of performances." But Foucault is not after rules for infinite possibilities that may or may not ever have been or ever will be actualized, but the rules for the finite discursive performances that have actually happened. (This anticipates what Foucault will later in AK call the "principle of rarity"). Discursive analysis is also to be distinguished from history of thought, which is always "allegorical," that is, that treats the linguistic sign as the marker for another event, the conscious or unconscious activity of a speaker. History of ideas makes discourse into a document whose explanation is sought by reconstructing another discourse (the infinity of commentary). By contrast, Foucault's project is to grasp the discursive field in the "specificity of its occurrence" so that one can "determine its conditions of existence." (Contrast with the Kantian conditions of possibility of experience). Here again we see the demand to survey accurately the actual field so that the map of the virtual (conditions of existence) we construct on its basis does not include smuggled-in unities but is grasped in its own (virtual) dispersion. We must avoid the uncritical retrojection of produced unities onto their conditioning field (the Nietzschean / Deleuzean critique of the transcendental subject). The virtual is difference in itself, as Deleuze would say: the conditioning must not resemble that which it conditions. Foucault's term for such difference is "discontinuity." The "strange event" of the statement is where discontinuity "emerges": the point of actualization of the virtual. This discontinuity emerges in the irreducibility of the statement to a possible parole for a langue, or to the interpretable meaning of a subject, the twin projects of linguistic analysis and history of ideas. The discontinuity of the statement appears in three registers, each of which indicates the resonance of Foucault's work at this time with that of Derrida: 1) the statement is linked to speech and writing (i.e., to a subject), yet it leaves a trace (a "residual existence") in memory or matter; 2) the statement is unique, yet repeatable, transformable, able to be reactivated (see especially Derrida's Introduction to Edmund Husserl's "Origin of Geometry"); 3) the statement is linked to situations and consequences, but also to the field of statements within which it occurs. (This difficulties Foucault encounters with the archaeological approach to this latter point is of course what will provoke the development of genealogy.) Discursive Formations After noting that he will defer the definition of "statement," "event," and "discourse," Foucault specifies four types of unities of discourse: object, enunciative style, concept, strategies. Each of these has to be seen in the rules of dispersion that govern their production. Definitions are offered of "discursive formation" (system of dispersion [=divergent series] of objects, enunciative styles, concepts, strategies) and of "rules of formation" (conditions of existence of statements [=virtual field]). The formation of objects Foucault begins by noting three sets of factors: surfaces of emergence, authorities of delimitation, grids of specification, then continues by noting the inadequacies of this simple listing, since objects are not just handed over to discourse; Foucault needs to know the relations among these factors. The question is

about the status of the formation of new objects. Foucault argues against both the practice of standard history of science and reductionist materialism. Against narratives of the history of science as approximations to a rational method, Foucault argues that new objects cannot simply be discoveries; Foucault wants to know what made them possible [transcendental rules as condition of existence = virtual field]. Similarly, against reductionist accounts, Foucault argues that new objects are not just effects of institutional change, as these are by definition non-discursive and Foucault wants an archaeology that is a pure analysis of discourse. Here again we see the problem that will give rise to genealogy. The isolation of dicourse from institutions, we claim, limits Foucault to a fixed virtual realm that cannot account for discontinuity as event of switching virtual regions; only genealogy and a multifactorial model of co-evolution and coupled landscapes can account for production and novelty within the virtual. Thus for archaeology, the problem of new objects must be accounted for by considering the relations among the three factors of surfaces of emergence, authorities of delimitation, and grids of specification. While Foucault is clear that the historical conditions of new objects are positive, not just limiting (even a fixed virtual field can account for some novelty by re-shuffling of possibilities), the problem of the relation of discursive and non-discursive practices remains insoluble for archaeology, although Foucault makes a brave attempt at this point. He begins by claiming that discursive relations are established in field of non-discursive relations; these enable objects to appear in field of exteriority, but are not present in the object. Thus discursive relations must be distinguished from: primary relations (those between non-discursive factors) can establish relations of dependence for discursive relations; secondary relations in discourse: e.g, what psychiatry says re: family/criminality. In other words, at this point, discursive relations are at the "limit" of discourse (the point of actualization of the virtual). Foucault now summarizes his points: discursive relations form set of immanent and defining rules; this is not a history of the referent: not "the things themselves" and thus anti-phenomenology; nor an analysis of meaning and thus anti-cultural history. Thus discursive formations are not a question of "words and things" (indicating the irony of the title, Les Mots et les choses.)