1Definitions of the real say a lot about the discursive formation1 within which they have been articulated. Of what else might they speak? To what else might they bear witness? The real is discovered and recovered, it is invented, it has shape given to it, according to the terms constitutive of the innumerable acts of discourse. But may we add the specification: according to the terms constitutive of the discourse which it (the real) authorizes? In analysing the protocol of realism, Roland Barthes spoke of the effet de réel2 as an effect governed by a set of codes and procedures, to be analysed as such. Meanwhile we persist in our aspiration to something else, probing the slightest shadow or blur within discourse as a possible trace of a heterogenous, non-discursive source, as if the real were the designation of whatever is in excess of a discursive formation, of what is not to be numbered among its effects. In which case the conclusion to any critical inquiry — that all this is merely an effect of the real — leaves room for the supposition of a real force that is beyond discourse. Thus in speaking of an effect of the real (by which we are not deceived), we leave open the possibility of a real that is not merely a produced or an induced effect, that, on the contrary, is incontrovertible and absolute in its evidence, as it bears downwards and inwards on our body and our speech, like a moment of diabolic or divine possession. (If we cling to this possibility, it may be from a reluctance to resign ourselves to having to live by a set of discursive rules which are not unconditionally and transcendentally given, that, on the contrary, are of our own predictable human making: to wish the real to be more than the mere effect of our discourse may be an abuse of the grammatical option of nominalisation, by which we make for ourselves a fetish, the Real.) In other words, our engagement with the notion of the real is a variation on the Pascaban wager: we hedge our bets, acting our parts as the disenchanted decoders of discursive effects, yet remaining open to the possibility of the real as that which comes, the real as adventurus, beyond our scheming and invention: as disruptive of our establishments as the visitor in Pasolini's Theorem.

2Thus our invocations of the real bear witness doubly (and are doubly suspect): to the discursive formation whose protocol conditions our speech acts; to the naivety and obstinacy of our aspiration to something else. Such a dichotomy is symptomatic of a discursive formation in which regret, pathos, sentiment, are major traits. The produced effect and the graceful afflux are mutually supportive; they collaborate despite their apparent opposition. Michel Foucault has taught us to identify the collusion of contraries, to see that a discursive formation can orchestrate the semblance of opposition and option, its strategic system allowing for a play of collaboration and antagonism between the component members.

3To invoke Henry James, whose work may be considered the exemplary achievement of the logic of representation, enables us to advance an entirely different definition of the real, harder-edged, less conducive to pathos. The Jamesian definition of the real as "the things we cannot possibly not know"3 is the concomitant of a definition of romance as the illusion of a detachment from restrictions, so that the intentional adjustment between consciousness and the world may be brought about by way of the latter's compliance with a desire that is sovereign.4 Chapter forty-two of The Portrait of a Lady fits in with such a definition, being the dramatisation of a moment of reflux in which the world presses inwards and the enterprising imagination has no option but to retreat, as the subjective experience comes up against an absolute limit forcing it to turn back on itself, till it is led to reflect critically on its earlier outward venture. Here it is the weight and impress of the unmoveable outside which has forced consciousness into compliance, and not the contrary. Clearly, in the case of James, what we are dealing with is a novelist's definition, one which presupposes the prior moments of a narrative movement, whose end is the imposed recognition of all that is in excess of the naiveties of a consciousness, or of the surface play and levity of a discursive configuration. The Jamesian narrative moves towards a final state in which consciousness will at last be squared.

4There are other lapidary definitions which we may cite. Again they can be understood as moments of capitulation, as terminations of a movement. But perhaps it is always the same ending, as the restriction which until then has been evaded is at last endured. The real is then definable as the excipit of all the enterprises of narrative. I shall quote the following passage, by Louis Marin:

Passion du signifiant qui serait interminable voyage du sens, son douloureux roman d'apprentissage dans le monde, le réel, c'est-à-dire l'insensé, comme aussi bien dans les systèmes où le réel est toujours déjà articulé. (Marin, n. d., 52)5

5How are we to read the word "comme"? As a connection, or as an opposition, in this shorthand version of Hegelian phenomenology which, no longer equipped with the power to state that the real is the rational, can scarcely go beyond the proffering of an exclamation? Being powerless to adjust anything at all — the real is here the undefined senseless, a something, or the Thing — one can only declare one's bewilderment. The systematized real, we may infer, is not the senseless real. If the latter is an influx in excess of our intentions (an excess, whether gift or catastrophe, beyond our faculties of adjustment, beyond those forms and shapes of accommodation which Nabokov credits with a merely provisional validity6), the systematised real is the measured real, cut and framed to the aperture of an individual consciousness (to speak a Jamesian and phenomenological idiom), or (to speak a Foucaldian idiom) a production of the discursive formation which, at a given moment, ensures that the real is never encountered as the senseless. If the real is the senseless, the senseless is what we are powerless to make anything of: recalcitrant, devoid of plasticity, its emblem is stone, the stony silence of one petrified and aghast, the stumbling block against which the charms of our language can "do" — the Jamesian doing — absolutely nothing. Leaving room only for a look of bewilderment, in accordance with the habitual topos of the real as an epiphany beyond words. Faced with the real that one is powerless to penetrate, consciousness itself is in turn petrified. This sublime real is however a cultural commonplace, designed and fitted out for us within discourses which shape both our desires and what is beyond them.

6I doubt whether the reinvention of the real is to be understood as a will to recover the senseless: it is the senseless which, in the event of an involuntary collapse of our systematisation, might find us out. Reinvent, a transitive verb, suggests a will to redispose things — the latin res — to test anew our capacity to make the real pliant, subject to our designs. A reading of Foucault's Les mots et les choses weans us away from any urge to hear the reinvention of the real as a rallying-cry or a manifesto-type utterance, like the catchcry of phenomenology: Back to things. The interminable resystematisation of the real — a guarantee against any exposure to the senseless — is by now a process that is forever in progress, being governed by a logic of systemics which is in no way contrary to a logic of process which is the concomitant of the plasticity of the real, and of the pragmatic, inductive efficacy of discourse. Foucaldien biopower7 we may thus regard as the product of an evertighter knotting or meshing between discourse and facticity, with the latter no longer appearing otherwise than in the guise of an effect already reworked and reinvented, according to its utmost potential for plasticity.

Periodisations

7How many epistemes can we identify in the Foucaldian scansion of Les mots et les choses? Renaissance, âge classique, modernité, postmodernité8: Foucault's archeology does not radically overturn the usual division between epochs. We note it is a sequence of four moments, and not the more usual ternary sequence. If there are four moments, the last is to be understood as an ultimate reunion, polymorphic and processive, and as a strange or sinister repetition of the episteme of the Renaissance. The drama of disunion can thus be understood as a drama played out only in the two intermediary moments: the classical age of discourse, and its successor, the age of the transcendental subject, when the depths are dug more deeply, the durations of time extended, and language escapes from the mode of being of discourse. Only within these two intermediary epistemes can the will to know and to enounce the real be exercised.

8For the knowledge of the real implies an intentional adjustment: a gap, and the possibility of bridging it; the search for the real followed by the recovery of the real. A euphoria of signifying, an unchecked effusiveness of meaning, on the contrary renders void any postulate of the apprehension within the chamber of consciousness of the real as that which is absent — as the ens beyond oneself — but which one might make sense of and encompass. Caught up in what is a continual process of reinvention, the real is not a motive to be reinvented. Unless, that is, it be recovered and reinvented as the senseless. In Les mots et les choses Foucault rendered highly problematic the possibility of any access to the senseless. Subsequently, in L'Archéologie du savoir, he was to repeat these procedures of conceptual problematisation, as if to ensure that no path offering access to a real (as a donnée independent; of designs, independent of a discursive formation) remained open. Derrida's critique of Histoire de la. folie à l'âge classique had implied the candour of a primitive desire on Foucault s part to enounce or to rediscover a senseless condition, folie, perhaps a real outside, prior to the discourses abusively projected upon it9, as purely real as the Mallarméean flower absent from all bouquets: the singular real thing as hapax. The critique thus closed off any possible attainment within discourse of the senseless real — the real of folly or the folly of the real —, with the result that Foucault, in writing out his subsequent œuvre, would appear to have battened down the hatches, as he took stock of the ubiquity of a discursive power endowed with an ever-greater capacity for the invagination of the outside. Having written a treatise on representation whose title, Les mots et les choses, is a repetition of the scholastic question of the relation between dictus / sermo and res (thus establishing the question of the real as that of the relation established across a maintained gap, one which allows for the discursive achievement of an extensional and referential outreach, and leaves room also for the acknowledged ex-istence of the thing referred to), such a hypothesis — to which the title of the book lent credit — that of a power of nomination exercised upon the state of a (systematised) world, was subsequently to be cast aside by Foucault, as an over-convenient conciliation of the claims of articulation and of separation, so that from then on, what remained to be conceptualized in his œuvre (reflexively, as a discursive inquiry upon the status of discourse) was a state of absolute incorporation, of which biopower is the supreme, final achievement.

9It may be legitimate, from a vantage-point located within the episteme of representation, to speak of the real. If Foucault rarely resorts to the term, the reason is that the sequence of epistemes is a movement towards the end and suspension of representation: it is from a vantage-point beyond representation that Les mots et les choses came to be written, as a treatiste on representation. Hence its paradoxical status, as a purely reflexive language, but which retains a residual extralinguistic design. Such a final ending-point — a parody of an excipit, since from now on there is no venturing-out into the non-discursive beyond — can exhibit neither the triumphant recovery of the real, nor the pathos of its loss or of its recovery as the senseless other. The proliferation of induced and produced effects, which the consummation of biopower gives rise to, leaves no room for any identification by the intentional consciousness of the real as a claim pressing down upon a speech act that is the announcement or programme of its reinvention to come; or, on the contrary, for the lucid recognition that the real, in all its pure senselessness, is not to be reinvented or tampered with. Biopower is an anthropocratic nightmare, akin to the Heideggerian Gestell.10 What it achieves is the goal of pure romance, insofar as nothing is any longer external to it, or unaffected by it. It also is the mirror-image of that theocratic perfection proper to the first of the epistemes delineated in Les mots et les choses: in the earlier episteme language is lodged in the world like a pearl in an oyster; in the later one the consummately formatted world is criss-crossed by capillary veinlets and discursive micro-networks, so that not the slightest patch of senselessness is left intact. In his eloquent formula, Louis Marin continues to lend credence to the idea of a noble confrontation with the senseless real, as the double which (out) flanks the systematised real. Foucault's writing, after Histoire de la folie, is a preemptive warning that the senseless, what is outside the syntaxonomy of language, fails to show up for any (no longer noble) melodramatic face-to-face with the thing, or with the Real Thing. The presentable effect of the Other, the effect of senselessness, is everywhere distributed (in discourse): no longer will it show up otherwise than in accordance with the rules of discourse, the rules of a power that has at last achieved its final condition, being both unquestionable and unidentifiable, while remaining a capacity to take in and tc take hold of everything. In other words — but the emphasis and pathos are displaced, they are utterly groundless in this last formation—a transcendental horizon of reinvention excludes any recovery of the real as the senseless, or as an object of our research. Henceforth the real is everywhere interned and remodelled, in process and in progress.

Naming

10The theory of language proper to the first of the epistemes analysed in Les mots et les choses is a theory of the name: to speak is to name. This theory of nomination is not that of a well-kept register of designations, in which things reflect and duplicate themselves: such a coded registration is the prerogative of the classical age, of the episteme of representation. Prior to this, the sixteenth-century episteme is qualifiable in the following manner:

Au XVIème siècle, le langage réel n'est pas un ensemble de signes indépendants, uniforme et lisse, où les choses viendraient se refléter comme dans un miroir pour y énoncer une à une leur vérité singulière. Il est plutôt chose opaque, mystérieuse, renfermée sur elle-même, masse fragmentée et de point en point énigmatique, qui se mêle ici ou là aux figures du monde, et s'enchevêtre à elles: tant et si bien que, toutes ensemble, elles forment un réseau de marques où chacune peut jouer, et joue en effet, par rapport à toutes les autres, le rôle de contenu ou de signe, de secret ou d'indication. Dans son être brut et historique du XVIème siècle, le langage n'est pas un système arbitraire; il est déposé dans le monde et il en fait partie à la fois parce que les choses cachent et manifestent leur énigme comme un langage, et parce que les mots se proposent aux hommes comme des choses à déchiffrer. (49-50)

11Two points may be noted. First, there is no break between a knowledge of signs, a semiology, and the substantial knowledge of the world, a natural philosophy. The first is not merely a heuristic instrument resorted to for the knowledge of the latter. Language and world are intertwined, they form a single textual fabric. Both participate in an enterprise of ciphering and deciphering, being part of a cryptic revelation. Second, we may note the frequency, here and throughout the book, of the terms loger, se loger, déposer (51). A sign, whether as an element of language or an element of nature, does not simply indicate and advert to a referent outside of it. (Unless, that is, we envisage the entire universe as one vast sign indicative of its Creator.) Every sign participates synecdochically in the truth it both reveals and hides: language and world are alive with the creative force which originally conceived them, whose effects they figure and manifest in time. There is no difference between an effect and a sign. Thus it is perfectly logical that there should be cases of illicit magic, endeavours to dislodge the encrypted truth by way of an illicit penetration towards the hermetic deposit, in order to use the discovered power for one's own abusive purposes. At all times it is a case of the analysis of signs, and not of the knowledge of the real. Text and world are both of them signifiers, since the treasure of a truth has been lodged and deposited in them: the traces of the Creator's cosmic act of conception.

12In this first moment of his treatise, Foucault resurrects the seamless garment of a significant universe, comparable to, though far more robustly textured than the universe of Baudelaire's Correspondances:

La Nature est un temple où de vivants piliers
Laissent parfois sortir de confuses paroles;
L'homme y passe à travers des forêts de symboles

13Are we then to opt for a ternary sequence: the magic of the ciphered world of the Renaissance, then a classical taxonomy, and finally the romantic — compensatory, restorative — imagination? We shall remain faithful to the hypothesis of four moments. First a semaphore world, pervasively significant, where vision and reading come together, nature being considered as a sequence of legenda. Here the real is not a pertinent category. Or it is of no more pertinence than for the early medieval texts examined by Auerbach in Mimesis,11 since the category only assumes its full analytic pertinence with the retreat, or at least the temporary suspension, of figurative value, since it is this moment which leaves room for the exposure of our intentional designs to the senseless idiocy of a world which signifies nothing apart from its stolid muteness. Devoid of pertinence until the arrival on the threshold of a modern world of representation, whose paradigmatic genre will be the bildungsroman, a narrative of enlightening disappointment. It is here that Don Quixote is called up by Foucault, as the victim of his own misreading, a misreading symptomatic of the rupture between the hermeneutics of legenda and the visibilia registered by the eye, in this new order where names are no longer "deposited upon that which they designate" (51). The hero of narrative repeatedly learns (Isabel Archer included) that values are not securely readable on the face of things, being subject to a continual process of disenchanting revision, with the result that the real is less to be defined as an object of investigation and as a possible capture than as a moment of unsettlement, désarroi, a syncopation in the romanesque movement, whose emblem which might be the noble rider unseated.12 The roman réaliste, as the dysphoric literary form proper to the episteme of representation, associates in its name the projective movement of signification — the romanesque as a conquering adventure, its hero buoyantly convinced that meaning has been deposited in the world, left waiting for him — and the sobering realization that the deposit is void, the coinage counterfeit. There is of course a convention which allows for the movement to be arrested in the euphoric instant of the successful projection and/or discovery of a deposit of meaning. The Rimbaldian moment of voyance exemplifies this. But it is only a moment, the true terminal-point of the discovery of the real being th g exposure of the supposed act of figuring and visionary penetration to a donnée in which nothing has been left waiting for those who are able to see and to read. The real is thus the rupture of the figurative, tropological movement: as such it tends to be dreadfully anti-revolutionary.13

14The series of epistemes had started out under divine tutelage: a theory of language built upon nomination is, whether openly or covertly, theological, being the amplification of two affirmations: By the word of the lord were the heavens made (Psalms, 33:6) and The heavens declare the glory of God/ And the firmament sheweth his handywork (Psalms, 19:1). The second of these evokes the response which creation addresses to its Creator. The category of the real, I repeat, is here entirely absent. Absent from the Foucaldian reinvention of the Renaissance, first of the book's four epistemes, the category of the real achieves its (provisional) presence and effectiveness in the second and third moments. In addressing ourselves to the sequence of epistemes, in historicising the logic of an archeology, we are of course going against the grain of the argument. An episteme is a configuration to be examined in terms of the coherence of its internal organisation, and not by way of a comparison with other strategic systems, nor by way of the scholastic definition of truth as a correspondance to a state of things, a definition which acknowledges both the gap and the articulation between res and intellectus. The opening of Foucault's book, where the author admits to his laughter on reading through the extravagant Borgesian classification of the Chinese emperor's animals, can be read as a fable cautioning us, on the threshold of this archeology, that the arrangement proper to any discursive disposition has a merely internal, pragmatic rationale, that all classifications are essentially groundless, and the passage from any one disposition to another devoid of either an overriding or an underlying motivation.

15However one's experience as a reader of the book does involve a sense that the sequence of epistemes is in fact rigorously orientated, though not as a strict rectilinear sequence, rather as a looping-back or a return. The fourth episteme repeats and replays the first — perhaps interminably —, while the logic of the sequence prior to this final turningback is an affair of lodgement-dislodgement. If his archeology of human sciences proposes a sequence of epistemes which we apprehend as looping backwards on itself, Foucault thus reveals his kinship with the Yeats of A Vision, who, like Foucault, was sensitive to the Nietzschean model of a movement turning back on itself, in lieu of a progressive line, or the triadic resolution of opposing theses. Such a closure, as a line is forced backwards into the pull of its own magnetic force, whether in the case of Yeats, Foucault, or the Joyce of Finnegan's Wake, enacts a paradigm whose coefficient of derealization is multiplied far beyond that of any other conceptual system. The real, when it is reeled back into the charmed circle — the Yeatsian gyre — is a provisional moment of désarroi, never the unrelieved exposure to a senseless remainder.

Representing

16It is here that we encounter disunion, as the consequence of a disengagement:

Don Quichotte dessine le négatif du monde de la Renaissance; l'écriture a cessé d'être la prose du monde; les ressemblances et les signes ont dénoué leur vieille entente; les similitudes déçoivent, tournent à la vision et au délire; les choses demeurent obstinément dans leur identité ironique: elles ne sont plus que ce qu'elles sont; les mots errent à l'aventure, sans contenu, sans ressemblance pour les remplir; ils ne marquent plus les choses; ils dorment entre les feuillets des livres au milieu de la poussière. (Foucault, 1966, 61-62)

17"The negative of the world of the Renaissance" (and for Foucault the Renaissance is essentially neo-Platonist and anti-Thomist), we take as a useful qualification of the real, if we call up the diverse connotations of renaissance: that the real allows for no possibility of rebirth, Foucault's qualification being suggestive of the disappointment of all such hope, proposing as it does a variation on the Weberian Entzauberung, intimating a condition of unsignifying obstinacy (idiocy).

18The dislodgment of language does not initially imply its condemnation to a quixotic wandering. Retreat and dislodgement can be imagined as a process akin to the separation in Genesis between the air above and the waters of the deep, the result of which is the formation of a secure ground for man. Here the dislodgement is productive of the classical order, which is in truth the linkage of two orders, the form of discourse and the substance of the world, the one being the adequate regulation of the other, the disintrication having made possible a relation between separates, there being no relation except as an articulation of congruous, but distinct entities. The term disunion is thus somewhat ambiguous. The word receives no authorization from the Foucaldian definition of the order of discourse. It is nonetheless in tune with a certain pathos in the writing, present through the topos of an untying, and with the recurrent evocation of the twin personae of the madman and the poet, "tout proche par sa symétrie" (63), both of whom persist in the task of deciphering resemblances, with the poet being defined in the following manner:

Celui qui, au-dessous des différences nommées et quotidiennement prévues, retrouve les parentés enfouies des choses, leurs similitudes dispersées. Sous les signes établis, et malgré eux, il entend un autre discours, plus profond, qui rappelle le temps où les mots scintillaient dans la ressemblance universelle des choses. (63)

19The poetic word is deemed to possess a glow, a brightness and force of incarnation, which the order of discourse has slacked over, but not extinguished. A language whose property is scintillation is the attribute of a semiotic regime which operates by participation: a sign that is radiant like a light is one which participates in and cannot fail to signify its source, its sun. Meaning is secured by the consubstantiality between the (transcendental) signified and its sublunary signifier. A meaning which scintillates is fully motivated, a meaning epiphanic and apodictic. It is only if we start out from this instance of a scintillating meaning, the emanation of its dynamic source, and take this as our paradigm of secure nomination, that the composite order of discourse shows up as a state of relative disunion. If we bracket this episteme of scintillation and decipherable resemblance, the order of discourse can be taken as a case of articulate union, whose condition is the separation and conjunction of a discursive formation and a given world, the latter envisaged as the theme of propositions which can lay reasonable claim to their status as acts of a representational knowledge which effects a conceptual realignment of what is given to one's vision, producing an ideally recomposed order from which the accidents interjected by time have been elided (see Foucault, 1966, 95-107). Foucault's diagnosis, elsewhere in his writings, of what is involved in the departure from the classical age provides us with an implicit, oblique definition of what the latter had achieved: "Le langage échappe au mode d'être du discours — c'est-à-dire à la grande dynastie de la représentation" (Foucault, 1986, 12).

20Thus there is a mode of being of language which lies outside the "dynasty of representation" — another qualification for the order of discourse. Representation is a term to which Foucault will give no more substantial definition than this, the most developed definition of the term in Les mots et les choses being a series of quasi-tautological phrases:

Représenter est à entendre au sens strict: le langage représente la pensée, comme la pensée se représente ellemême. Il n'y a pas, pour constituer le langage, ou pour l'animer de l'intérieur, un acte essentiel et primitif de signification, mais seulement, au cœur de la représentation, ce pouvoir qu'elle détient de se représenter, c'est-à-dire de s'analyser en se juxtaposant, partie par partie, sous le regard de la réflexion, et de se déléguer elle-même dans un substitut qui la prolonge.... Les représentations ne s'enracinent pas dans un monde auquel elles emprunteraient leurs sens; elles s'ouvrent d'elles-mêmes sur un espace qui leur est propre et dont la nervure interne donne lieu au sens. (92)

21Representation, through which the empire of discourse exercises its power, is an act of ordering: "language analyses" (131). In the passage quoted above, we note the abundance of reflexive verbs: se représenter, s'analyser, se juxtaposer, se déléguer (a paradoxical splitting of self), s'enraciner. This is evidence of the effort to conceptualise the act of ordering without postulating any legatory, mandating power above, or any underlying material or object of analysis located beneath or outside discourse. Representations do not root or lodge themselves in the world. The discursive order does nonetheless constitute an order of propositions whose fiduciary value is that of a world recomposed. It would appear that the only knowledge proper to this order is the knowledge of its own component elements, of its pacified dynasty: the order is consummate, so much so that its presentation, if we subscribe to the argument of the above-quoted passage, can only take the form of an autoanalysis, secure against any risk of a venture outwards, or any collision with the senseless. If the real is what we "cannot possibly not know", we can argue that the dynasty of representation does achieve a knowledge of the real, since it knows itself, and can only ever know itself. But an empire of solipsism, immune to any threat of resistance or aggression: is that, ultimately, the only possible reading of the Jamesian definition of the real?

Supersession

22Chapter seven of Foucault's book is entitled "Les limites de la représentation". It inaugurates the second part of the book. The hinge upon which the book turns is the shift from one episteme to another:

Discontinuité symétrique de celle qui avait brisé, au début du XVIème siècle, la pensée de la Renaissance; alors, les grandes figures circulaires où s'enfermait la similitude s'étaient disloquées et ouvertes pour que le tableau des identités puisse se déployer; et ce tableau maintenant va se défaire à son tour, le savoir se logeant dans un espace nouveau. (229)

23The chapter-heading suggests the existence of a location that is beyond or beneath the table of representation. While a book-title such as La pensée du dehors accredits the first-mentioned location, numerous passages in Les mots et les choses adopt a romantic topology opposing a surface order to a movement in the depths in order to qualify the disposition of the "modern" episteme which takes over from the dynasty of representation. The transition from one episteme to another is flawless. There is merely the recognition of the newly established edifice: knowledge housed (se logeant) in a new space. No interlude, euphoric or sinister, prior to the new accommodation of knowledge. Housing, unhousing, rehousing: the sequence repeats, in slightly different words, the Deleuzean sequence territorialisation, déterritorialisation, reterritorialisation. The changes of decor are impeccable. It is not here, as a "new space" establishes itself, that we are in danger of finding the real, if what we are seeking after is the real as the senseless. Derrida's suggestion that Histoire de la folie implied the postulate of an intact force beneath or outside discourse, prior to any discursive arraignment, has no cogency here, since the analysis of the post-classical episteme — work, life, language — shows it to be in no sense vulnerable to the accusation that it involves the substantialization of a "pure" force. The "modern" episteme sets up a game conducted on two complementary levels, each of which implies the other: what is beneath is neither the hidden truth nor the real, of which the surface appearance is the deceptive mask. The episteme operates through the articulation and coimplication of two levels:

Classer ne sera donc plus référer le visible à lui-même, en chargeant l'un de ses éléments de représenter les autres; ce sera, dans un mouvement qui fait pivoter l'analyse, rapporter le visible à l'invisible, comme à sa raison profonde, puis remonter de cette secrète architecture vers les signes manifestes qui en sont donnés à la surface des corps. (242)

24The following passage suggests the intricacy of this reciprocal relation:

Le travail, la vie et le langage apparaissent comme autant de " transcendantaux " qui rendent possible la connaissance objective des êtres vivants, des lois de la production, des formes du langage. En leur être, ils sont hors connaissance, mais ils sont, par cela même, conditions de connaissance; ils correspondent à la découverte par Kant d'un champ transcendantal et pourtant ils en diffèrent sur deux points essentiels; ils se logent du côté de l'objet, et en quelque sorte au-delà; comme l'Idée dans la Dialectique transcendantale, ils totalisent les phénomènes et disent la cohérence a priori des multiplicités empiriques; mais ils les fondent dans un être dont la réalité énigmatique constitue avant toute connaissance l'ordre et le lien de ce qu'elle a à connaître. (257)

25What then is this "enigmatic reality", the reality of a being that is "outside knowledge, while rendering knowledge possible", since it is the condition of a knowledge which, as a knowledge of the manifest, can only be a knowledge of the surface effects conditioned by their hidden ground? Again it is an affair of reinvention and rearticulation. What is present before the eye is referred to what is hidden. The hidden is not however a truth to which one accedes, being merely the condition that is epistemologically posited, so that the appearance can be made amenable to reason. Our usual assumption is that reality is either knowable or unknowable, but that in either case it is the object of an intentional strategy. Here the "enigmatic reality" is merely the condition of a knowable effect. Any knowledge of this "enigmatic reality" can be only a reflexive knowledge, by which the episteme knows itself. Again we note the predominance of the reflexive verbs, a symptom, perhaps, that the destiny of all Foucaldian epistemes is to allow only for a knowledge that is a self-knowledge, for an access to an "enigmatic reality" which has already been incorporated by the episteme as the condition of any instance of empirical, local knowledge. In short, one is always only dealing with the regulated effects of an episteme. The invocation of the real thus involves an untenable promise, since it designates a senseless something that is in re and extra dictum, whereas any known effects can only be referred to their place of production in an episteme, in one of the series of endlessly productive configurations of what exists.

26For the real is a theoretico-practical hybrid: the only real is a reformed real, the only encounter is with what exists. Socialisme réel is really existing socialism; the real world is the existing world, that is to say, a discursive formation which is the result of an accumulation of collective action, with all its impressive, preemptive and restrictive weight. To say that this is the real, when it is only what exists, is to abuse of the property we mentioned in our introduction: the capacity of any nominalized usage of the word real to absolutize its claim, and thus designate a given, or a destiny. The result is that what is to be analysed, and which is a complex, humanly wrought effect — effectivité, Wirklichkeit, facticity — illegitimately masquerades as an ineluctable given: "the things we cannot possibly not know". Such a definition is acceptable only if we add: the things which we cannot not know, given the conditions of possibility of an effect of reality, in the prevailing discursive and pragmatic configuration.

27It is the latter which is given. Not randomly, but through the long-term impress of historical determination: that which exists — and which is not the real — is the code-name resorted to in order to ward off any intimation that within some other different discursive and pragmatic formation, things might have presented and impressed themselves otherwise. So long as facticity can consort as the real, an episteme can take over from its predecessor without any risk of turbulence. Order (the dynasty of representation or one more discreet) reigns. Never will any sudden draught of air from the outside be felt. If we persist in conceiving of the category of a real extra dictum, this can only be as a Utopian gesture towards a non-place beyond the space of our establishment, as a gesture to something senseless. The evaporation of what habitually veils things, the withdrawal of the usual discursive formations, through the forceful agency of a poetic defamiliarisation, probably in the vicinity of a death: that is the condition of the Proustian and the Tolstoyan disclosure of the real — for the narrator in the courtyard of the Guermantes, for Prince Andrei on the battlefield in War and Peace — as the world declares itself and impresses itself as being intensely present, before consciousness takes leave of it for good. Manifestly the confr ontation with the real has to do with death. Our emblems of the real are our emblems of death, both being figured as an unscaleable barrier.

28Thus we have good reason to keep separate the categories of the real and of the existing. We can point to the ideological and self-serving assimilation between the existing as a (pseudo) ineluctable and the ineluctability of death, or of an irreversible lived time, limitations which (until now) have been absolute. Foucaldian biopower we take as the hypothesis of an ultimate knotting of what exists, so as to produce an unquestionable power over life and death. The hypothesis is that of an inductive, processive power no longer confronted with its outer limit, so that never more will it come up against its ending-point, since it has incorporated everything, in the manner of the Pauline Christ triumphant, who "must reign until He has put all things under his feet". There now reigns, the hypothesis intimates, a power whose empire is no longer provisional, no longer susceptible of being bracketed, which no longer needs to bow to the rare, incoercible force of these two moments: donation and withdrawal.

29We might end here, with Tolstoy and Proust, with the intimation of mortality as the condition of openness and exposure to the real. But Foucault does not authorize our invocation of such moments of opening. Such moments are emblematic of a modern literature, though not necessarily of the literature which Foucault celebrates — Roussel, Sade, Lautréamont. They are evocative of the literature which Fredric Jameson (1991, 154-80), places at the heart of a modern experience centred on temporality, and which he considers to have now been superseded by a postmodern configuration whose nervure, sinew, is spatiality. The prolonged elucidation in Les mots et les choses of the notion of finitude may not be cited in order to suggest a link-up with Proust and Tolstoy: Foucaldian finitude is not an exposure to something beyond us, it is not the postulate of an infinite: the infinity of the world without us, or the infinity of God. On the contrary, it designates the ever-more effective experience of the perfect adequation between the reach of a consciousness, or the structuring power of an episteme, and whatever can be known and produced. In other words, it designates the finitude of an "empirico-transcendental doublet" (Foucault, 1966, 329) for which the conditions of possibility of an act of knowledge and power match with and mesh with whatever is effectively produced and known. It is an ironic gloss on the Jamesian axiom: the reality we cannot possibly not know is the existing state of affairs, whatever is interminably induced and produced from the finitude of our empirico-transcendental hybrid. Let us quote the following passage, where the indefinite implies the radical foreclosure of any possible infinite ·.

La finitude de l'homme sera définie — une fois pour toutes, c'est-à-dire pour un temps indéfini...

Plus l'homme s'installe au cœur du monde, plus il avance dans la possession de la nature, plus fortement aussi il est pressé par la finitude, plus il s'approche de sa propre mort. L'Histoire ne permet pas à l'homme de s'évader de ses limites initiales — sauf en apparence, et si on donne à limite le sens le plus superficiel; mais si on considère la finitude fondamentale de l'homme, on s'aperçoit que sa situation anthropologique ne cesse de dramatiser toujours davantage son Histoire, de la rendre plus périlleuse, et de l'approcher pour ainsi dire de sa propre impossibilité. Au moment où elle touche à de tels confins, l'Histoire ne peut plus que s'arrêter, vibrer un instant sur son axe, et s'immobiliser pour toujours. (271)

30Interminably we come up against the same: le même. The (Mallarméean) emblem of such a condition would be the deck of cards, shuffled to produce the impression of the new, but within the strict limits of a rule-governed combination. Foucault breaks with the pathos of finitude, in order to think in key with the episteme that is now consolidating its hold. The adequation and perfect fit of his critical enterprise is a measure both of the appositeness and the inevitable limitation of such an enterprise of critical formulation, to which, according to its own definition of the notion of discursive formation, any breakthrough to or invocation of an outside, of the real as that which is beyond one's designs, from which the (utopian) critique of what exists might therefore be enounced, is prohibited. The involution of Foucaldian thought is the proof of its consummate espousal of that reflexive epistemic inventiveness which it is trying to analyse.

31I was mistaken, taken in by the surface effects, when I registered evidence of a disunion of words and things. Foucault wrote a book that is a polyptych: two central panels, discourse, modernity. To the left of these, the Renaissance realm of resemblance, where words are lodged in things, and both words and the things of the world are signs. To the right, an absolute reincorporation of everything that exists, in the round of the same. Disunion was temporary. Only during this temporary separation was the category of the real in any sense pertinent. Since the real did posit a gap, since the discovery or invention of the real did posit an articulation of separated terms, and — perhaps — the discretion of an act of speech involving a deference towards something which preceded and exceded it. The real, that to which an act of speech defers, was only provisionally utterable.