Title: The Order of Discourse
Author: Michel Foucault
Date: Dec. 2, 1970
Source: Untying the text: a post-structuralist reader
Notes: Translated by Ian McLeod

      I

      II

      III

      IV

      V

      VI

      VII

      VIII

I

I wish I could have slipped surreptitiously into this discourse which I must present today, and into the ones I shall have to give here, perhaps for many years to come. I should have prepreferred to be enveloped by speech, and carried away well beyond all possible beginnings, rather than have to begin it myself. I should have preferred to become aware that a nameless voice was already speaking long before me, so that I should only have needed to join in, to continue the sentence it had started and lodge myself, without really being noticed, in its interstices, as if it had signalled to me by pausing, for an instant, in suspense. Thus there would be no beginning, and instead of being the one from whom discourse proceeded, I should be at the mercy of its chance unfolding, a slender gap, the point of its possible disappearance. I should have liked there to be a voice behind me which had begun to speak a very long time before, doubling in advance everything I am going to say, a voice which would say: ‘You must go on, I can’t go on, you must go on, I’ll go on, you must say words, as long as there are any, until they find me, until they say me, strange pain, strange sin, you must go on, perhaps it’s done already, perhaps they have said me already, perhaps they have carried me to the threshold of my story, before the door that opens on my story, that would surprise me, if it opens.’ I think a good many people have a similar desire to be freed from the obligation to begin, a similar desire to be on the other side of discourse from the outset, without having to consider from the outside what might be strange, frightening, and perhaps maleficent about it. To this very common wish, the institution’s reply is ironic, since it solemnises beginnings, surrounds them with a circle of attention and silence, and imposes ritualised forms on them, as if to make them more easily recognisable from a distance. Desire says: ‘I should not like to have to enter this risky order of discourse; I should not like to be involved in its peremptoriness and decisiveness; I should like it to be all around me like a calm, deep transparence, infinitely open, where others would fit in with my expectations, and from which truths would emerge one by one; I should only have to let myself be carried, within it and by it, like a happy wreck.’ The institution replies: ‘You should not be afraid of beginnings; we are all here in order to show you that discourse belongs to the order of laws, that we have long been looking after its appearances; that a place has been made ready for it, a place which honours it but disarms it; and that if discourse may sometimes have some power, nevertheless it is from us and us alone that it gets it.’ But perhaps this institution and this desire are nothing but two contrary replies to the same anxiety: anxiety about what discourse is in its material reality as a thing pronounced or written; anxiety about this transitory existence which admittedly is destined to be effaced, but according to a time-scale which is not ours; anxiety at feeling beneath this activity (despite its greyness and ordinariness) powers and dangers that are hard to imagine; anxiety at suspecting the struggles, victories, injuries, dominations and enslavements, through so many words even though long usage has worn away their roughness. What, then, is so perilous in the fact that people speak, and that their discourse proliferates to infinity? Where is the danger in that?

II

Here is the hypothesis which I would like to put forward tonight in order to fix the terrain — or perhaps the very provisional theatre — of the work I am doing: that in every society the production of discourse is at once controlled, selected, organised and redistributed by a certain number of procedures whose role is to ward off its powers and dangers, to gain mastery over its chance events, to evade its ponderous, formidable materiality. In a society like ours, the procedures of exclusion are well known. The most obvious and familiar is the prohibition. We know quite well that we do not have the right to say everything, that we cannot speak of just anything in any circumstances whatever, and that not everyone has the right to speak of anything whatever. In the taboo on the object of speech, and the ritual of the circumstances of speech, and the privileged or exclusive right of the speaking subject, we have the play of three types of prohibition which intersect, reinforce or compensate for each other, forming a complex grid which changes constantly. I will merely note that at the present time the regions where the grid is tightest, where the black squares are most numerous, are those of sexuality and politics; as if discourse, far from being that transparent or neutral element in which sexuality is disarmed and politics pacified, is in fact one of the places where sexuality and politics exercise in a privileged way some of their most formidable powers. It does not matter that discourse appears to be of little account, because the prohibitions that surround it very soon reveal its link with desire and with power. There is nothing surprising about that, since, as psychoanalysis has shown, discourse is not simply that which manifests (or hides) desire — it is also the object of desire; and since, as history constantly teaches us, discourse is not simply that which translates struggles or systems of domination, but is the thing for which and by which there is struggle, discourse is the power which is to be seized. There exists in our society another principle of exclusion, not another prohibition but a division and a rejection. I refer to the opposition between reason and madness.2 Since the depths of the Middle Ages, the madman has been the one whose discourse cannot have the same currency as others. His word may be considered null and void, having neither truth nor importance, worthless as evidence in law, inadmissible in the authentification of deeds or contracts, incapable even of bringing about the trans-substantiation of bread into body at Mass. On the other hand, strange powers not held by any other may be attributed to the madman’s speech: the power of uttering a hidden truth, of telling the future, of seeing in all naivety what the others’ wisdom cannot perceive. It is curious to note that for centuries in Europe the speech of the madman was either not heard at all or else taken for the word of truth. It either fell into the void, being rejected as soon as it was proffered, or else people deciphered in it a rationality, naive or crafty, which they regarded as more rational than that of the sane. In any event, whether excluded, or secretly invested with reason, the madman’s speech, strictly, did not exist. It was through his words that his madness was recognised; they were the place where the division between reason and madness was exercised, but they were never recorded or listened to. No doctor before the end of the eighteenth century had ever thought of finding out what was said, or how and why it was said, in this speech which nonetheless determined the difference. This whole immense discourse of the madman was taken for mere noise, and he was only symbolically allowed to speak, in the theatre, where he would step forward, disarmed and reconciled, because there he played the role of truth in a mask. You will tell me that all this is finished today or is coming to an end; that the madman’s speech is no longer on the other side of the divide; that it is no longer null and void; on the contrary, it puts us on the alert; that we now look for a meaning in it, for the outline or the ruins of some oeuvre; and that we have even gone so far as to come across this speech of madness in what we articulate ourselves, in that slight stumbling by which we lose track of what we are saying. But all this attention to the speech of madness does not prove that the old division is no longer operative. You have only to think of the whole framework of knowledge through which we decipher that speech, and of the whole network of institutions which permit someone — a doctor or a psychoanalyst — to listen to it, and which at the same time permit the patient to bring along his poor words or, in desperation, to withhold them. You have only to think of all this to become suspicious that the division, far from being effaced, is working differently, along other lines, through new institutions, and with effects that are not at all the same. And even if the doctor’s role were only that of lending an ear to a speech that is free at last, he still does this listening in the context of the same division. He is listening to a discourse which is invested with desire, and which — for its greater exaltation or its greater anguish — thinks it is loaded with terrible powers. If the silence of reason is required for the curing of monsters, it is enough for that silence to be on the alert, and it is in this that the division remains. It is perhaps risky to consider the opposition between true and false as a third system of exclusion, along with those just mentioned. How could one reasonably compare the constraint of truth with divisions like those, which are arbitrary to start with or which at least are organised around historical contingencies; which are not only modifiable but in perpetual displacement; which are supported by a whole system of institutions which impose them and renew them; and which act in a constraining and sometimes violent way? Certainly, when viewed from the level of a proposition, on the inside of a discourse, the division between true and false is neither arbitrary nor modifiable nor institutional nor violent. But when we view things on a different scale, when we ask the question of what this will to truth has been and constantly is, across our discourses, this will to truth which has crossed so many centuries of our history; what is, in its very general form, the type of division which governs our will to know (notre volonte de savoir), then what we see taking shape is perhaps something like a system of exclusion, a historical, modifiable, and institutionally constraining system. There is no doubt that this division is historically constituted. For the Greek poets of the sixth century BC, the true discourse (in the strong and valorised sense of the word), the discourse which inspired respect and terror, and to which one had to submit because it ruled, was the one pronounced by men who spoke as of right and according to the required ritual; the discourse which dispensed justice and gave everyone his share; the discourse which in prophesying the future not only announced what was going to happen but helped to make it happen, carrying men’s minds along with it and thus weaving itself into the fabric of destiny. Yet already a century later the highest truth no longer resided in what discourse was or did, but in what it said: a day came when truth was displaced from the ritualised, efficacious and just act of enunciation, towards the utterance itself, its meaning, its form, its object, its relation to its reference. Between Hesiod and Plato a certain division was established, separating true discourse from false discourse: a new division because henceforth the true discourse is no longer precious and desirable, since it is no longer the one linked to the exercise of power. The sophist is banished. This historical division probably gave our will to know its general form. However, it has never stopped shifting: sometimes the great mutations in scientific thought can perhaps be read as the consequences of a discovery, but they can also be read as the appearance of new forms in the will to truth. There is doubtless a will to truth in the nineteenth century which differs from the will to know characteristic of Classical culture in the forms it deploys, in the domains of objects to which it addresses itself, and in the techniques on which it is based. To go back a little further: at the turn of the sixteenth century (and particularly in England), there appeared a will to know which, anticipating its actual contents, sketched out schemas of possible, observable, measurable, classifiable objects; a will to know which imposed on the knowing subject, and in some sense prior to all experience, a certain position, a certain gaze and a certain function (to see rather than to read, to verify rather than to make commentaries on); a will to know which was prescribed (but in a more general manner than by any specific instrument) by the technical level where knowledges had to be invested in order to be verifiable and useful. It was just as if, starting from the great Platonic division, the will to truth had its own history, which is not that of constraining truths: the history of the range of objects to be known, of the functions and positions of the knowing subject, of the material, technical, and instrumental investments of knowledge. This will to truth, like the other systems of exclusion, rests on an institutional support: it is both reinforced and renewed by whole strata of practices, such as pedagogy, of course; and the system of books, publishing, libraries; learned societies in the past and laboratories now. But it is also renewed, no doubt more profoundly, by the way in which knowledge is put to work, valorised, distributed, and in a sense attributed, in a society. Let us recall at this point, and only symbolically, the old Greek principle: though arithmetic may well be the concern of democratic cities, because it teaches about the relations of equality, geometry alone must be taught in oligarchies, since it demonstrates the proportions within inequality. Finally, I believe that this will to truth — leaning in this way on a support and an institutional distribution — tends to exert a sort of pressure and something like a power of constraint (I am still speaking of our own society) on other discourses. I am thinking of the way in which for centuries Western literature sought to ground itself on the natural, the ‘vraisemblable’, on sincerity, on science as well — in short, on ‘true’ discourse. I am thinking likewise of the manner in which economic practices, codified as precepts or recipes and ultimately as morality, have sought since the sixteenth century to ground themselves, rationalise themselves, and justify themselves in a theory of wealth and production. I am also thinking of the way in which a body as prescriptive as the penal system sought its bases or its justification, at first of course in a theory of justice, then, since the nineteenth century, in a sociological, psychological, medical, and psychiatric knowledge: it is as if even the word of the law could no longer be authorised, in our society, except by a discourse of truth. Of the three great systems of exclusion which forge discourse — the forbidden speech, the division of madness and the will to truth. I have spoken of the third at greatest length. The fact is that it is towards this third system that the other two have been drifting constantly for centuries. The third system increasingly attempts to assimilate the others, both in order to modify them and to provide them with a foundation. The first two are constantly becoming more fragile and more uncertain, to the extent that they are now invaded by the will to truth, which for its part constantly grows stronger, deeper, and more implacable. And yet we speak of the will to truth no doubt least of all. It is as if, for us, the will to truth and its vicissitudes were masked by truth itself in its necessary unfolding. The reason is perhaps this: although since the Greeks ‘true’ discourse is no longer the discourse that answers to the demands of desire, or the discourse which exercises power, what is at stake in the will to truth, in the will to utter this ‘true’ discourse, if not desire and power? ‘True’ discourse, freed from desire and power by the necessity of its form, cannot recognise the will to truth which pervades it;3 and the will to truth, having imposed itself on us for a very long time, is such that the truth it wants cannot fail to mask it. Thus all that appears to our eyes is a truth conceived as a richness, a fecundity, a gentle and insidiously universal force, and in contrast we are unaware of the will to truth, that prodigious machinery designed to exclude. All those who, from time to time in our history, have tried to dodge this will to truth and to put it into question against truth, at the very point where truth undertakes to justify the prohibition and to define madness, all of them, from Nietzsche to Artaud and Bataille, must now serve as the (no doubt lofty) signs for our daily work.

III

There are, of course, many other procedures for controlling and delimiting discourse. Those of which I have spoken up to now operate in a sense from the exterior. They function as systems of exclusion. They have to do with the part of discourse which puts power and desire at stake. I believe we can isolate another group: internal procedures, since discourses themselves exercise their own control; procedures which function rather as principles of classification, of ordering, of distribution, as if this time another dimension of discourse had to be mastered: that of events and chance. In the first place, commentary. I suppose — but without being very certain — that there is scarcely a society without its major narratives, which are recounted, repeated, and varied; formulae, texts, and ritualised sets of discourses which are recited in welldefined circumstances; things said once and preserved because it is suspected that behind them there is a secret or a treasure. In short, we may suspect that there is in all societies, with great consistency, a kind of gradation among discourses: those which are said in the ordinary course of days and exchanges, and which vanish as soon as they have been pronounced; and those which give rise to a certain number of new speech-acts which take them up, transform them or speak of them, in short, those discourses which, over and above their formulation, are said indefinitely, remain said, and are to be said again. We know them in our own cultural system: they are religious or juridical texts, but also those texts (curious ones, when we consider their status) which are called ‘literary1; and to a certain extent, scientific texts. This differentiation is certainly neither stable, nor constant, nor absolute. There is not, on the one side, the category of fundamental or creative discourses, given for all time, and on the other, the mass of discourses which repeat, gloss, and comment. Plenty of major texts become blurred and disappear, and sometimes commentaries move into the primary position. But though its points of application may change, the function remains; and the principle of a differentiation is continuously put back in play. The radical effacement of this gradation can only ever be play, utopia, or anguish. The Borges-style play of a commentary which is nothing but the solemn and expected reappearance word for word of the text that is commented on; or the play of a criticism that would speak forever of a work which does not exist. The lyrical dream of a discourse which is reborn absolutely new and innocent at every point, and which reappears constantly in all freshness, derived from things, feelings or thoughts. The anguish of that patient of Janet’s for whom the least utterance was gospel truth, concealing inexhaustible treasures of meaning and worthy to be repeated, re-commenced, and commented on indefinitely: ‘When I think,’ he would say when reading or listening, ‘when I think of this sentence which like the others will go off into eternity, and which I have perhaps not yet fully understood .‘4 But who can fail to see that this would be to annul one of the terms of the relation each time, and not to do away with the relation itself? It is a relation which is constantly changing with time; which takes multiple and divergent forms in a given epoch. The juridical exegesis is very different from the religious commentary (and this has been the case for a very long time). One and the same literary work can give rise simultaneously to very distinct types of discourse: the ‘Odyssey’ as a primary text is repeated, in the same period, in the translation by Berard, and in the endless ‘explications de texte’, and in Joyce’s ‘Ulysses’. For the moment I want to do no more than indicate that, in what is broadly called commentary, the hierarchy between primary and secondary text plays two roles which are in solidarity with each other. On the one hand it allows the (endless) construction of new discourses: the dominance of the primary text, its permanence, its status as a discourse which can always be re-actualised, the multiple or hidden meaning with which it is credited, the essential reticence and richness which is attributed to it, all this is the basis for an open possibility of speaking. But on the other hand the commentary’s only role, whatever the techniques used, is to say at last what was silently articulated ‘beyond’, in the text. By a paradox which it always displaces but never escapes, the commentary must say for the first time what had, nonetheless, already been said, and must tirelessly repeat what had, however, never been said. The infinite rippling of commentaries is worked from the inside by the dream of a repetition in disguise at its horizon there is perhaps nothing but what was at its point of departure — mere recitation. Commentary exorcises the chance element of discourse by giving it its due; it allows us to say something other than the text itself, but on condition that it is this text itself which is said, and in a sense completed. The open multiplicity, the element of chance, are transferred, by the principle of commentary, from what might risk being said, on to the number, the form, the mask, and the circumstances of the repetition. The new thing here lies not in what is said but in the event of its return. I believe there exists another principle of rarefaction of a discourse, complementary to the first, to a certain extent: the author. Not, of course, in the sense of the speaking individual who pronounced or wrote a text, but in the sense of a principle of grouping of discourses, conceived as the unity and origin of their meanings, as the focus of their coherence. This principle is not everywhere at work, nor in a constant manner: there exist all around us plenty of discourses which circulate without deriving their meaning or their efficacity from an author to whom they could be attributed: everyday remarks, which are effaced immediately; decrees or contracts which require signatories but no author; technical instructions which are transmitted anonymously But in the domains where it is the rule to attribute things to an author — literature, philosophy, science — it is quite evident that this attribution does not always play the same role. In the order of scientific discourse, it was indispensable, during the Middle Ages, that a text should be attributed to an author, since this was an index of truthfulness. A proposition was considered as drawing even its scientific value from its author. Since the seventeenth century, this function has steadily been eroded in scientific discourse: it now functions only to give a name to a theorem, an effect, an example, a syndrome. On the other hand, in the order of literary discourse, starting from the same epoch, the function of the author has steadily grown stronger: all those tales, poems, dramas or comedies which were allowed to circulate in the Middle Ages in at least a relative anonymity are now asked (and obliged to say) where they come from, who wrote them. The author is asked to account for the unity of the texts which are placed under his name. He is asked to reveal or at least carry authentification of the hidden meaning which traverses them. He is asked to connect them to his lived experiences, to the real history which saw their birth. The author is what gives the disturbing language of fiction its unities, its nodes of coherence, its insertion in the real. I know that I will be told: ‘But you are speaking there of the author as he is reinvented after the event by criticism, after he is dead and there is nothing left except for a tangled mass of scribblings; in those circumstances a little order surely has to be introduced into all that, by imagining a project, a coherence, a thematic structure that is demanded of the consciousness or the life of an author who is indeed perhaps a trifle fictitious. But that does not mean he did not exist, this real author, who bursts into the midst of all these worn-out words, bringing to them his genius or his disorder.’ It would of course, be absurd to deny the existence of the individual who writes and invents. But I believe that — at least since a certain epoch — the individual who sets out to write a text on the horizon of which a possible oeuvre is prowling, takes upon himself the function of the author: what he writes and what he does not write, what he sketches out, even by way of provisional drafts, as an outline of the oeuvre, and what he lets fall by way of commonplace remarks — this whole play of differences is prescribed by the author-function, as he receives it from his epoch, or as he modifies it in his turn. He may well overturn the traditional image of the author; nevertheless, it is from some new author-position that he will cut out, from everything he could say and from all that he does say every day at any moment, the still trembling outline of his oeuvre. The commentary-principle limits the chance-element in discourse by the play of an identity which would take the form of repetition and sameness. The author-principle limits this same element of chance by the play of an identity which has the form of individuality and the self. We must also recognise another principle of limitation in what is called, not sciences but ‘disciplines’: a principle which is itself relative and mobile; which permits construction, but within narrow confines. The organisation of disciplines is just as much opposed to the principle of commentary as to that of the author. It is opposed to the principle of the author because a discipline is defined by a domain of objects, a set of methods, a corpus of propositions considered to be true, a play of rules and definitions, of techniques and instruments: all this constitutes a sort of anonymous system at the disposal of anyone who wants to or is able to use it, without their meaning or validity being linked to the one who happened to be their inventor. But the principle of a discipline is also opposed to that of commentary: in a discipline, unlike a commentary, what is supposed at the outset is not a meaning which has to be rediscovered, nor an identity which has to be repeated, but the requisites for the construction of new statements. For there to be a discipline, there must be the possibility of formulating new propositions, ad infinitum. But there is more; there is more, no doubt, in order for there to be less: a discipline is not the sum of all that can be truthfully said about something; it is not even the set of all that can be accepted about the same data in virtue of some principle of coherence or systematicity. Medicine is not constituted by the total of what can be truthfully said about illness; botany cannot be defined by the sum of all the truths concerning plants. There are two reasons for this: first of all, botany and medicine are made up of errors as well as truths, like any other discipline — errors which are not residues or foreign bodies but which have positive functions, a historical efficacity, and a role that is often indissociable from that of the truths. And besides, for a proposition to belong to botany or pathology, it has to fulfil certain conditions, in a sense stricter and more complex than pure and simple truth: but in any case, other conditions. It must address itself to a determinate plane of objects: from the end of the seventeenth century, for example, for a proposition to be ‘botanical’ it had to deal with the visible structure of the plant, the system of its close and distant resemblances or the mechanism of its fluids; it could no longer retain its symbolic value, as was the case in the sixteenth century, nor the set of virtues and properties which were accorded to it in antiquity. But without belonging to a discipline, a proposition must use conceptual or technical instruments of a well-defined type; from the nineteenth century, a proposition was no longer medical — it fell ‘outside medicine’ and acquired the status of an individual phantasm or popular imagery — if it used notions that were at the same time metaphorical, qualitative, and substantial (like those of engorgement, of overheated liquids or of dried-out solids). In contrast it could and had to make use of notions that were equally metaphorical but based on another model, a functional and physiological one (that of the irritation, inflammation, or degeneration of the tissues). Still further: in order to be part of a discipline, a proposition has to be able to be inscribed on a certain type of theoretical horizon: suffice it to recall that the search for the primitive language, which was a perfectly acceptable theme up to the eighteenth century, was sufficient, in the second half of the nineteenth century, to make any discourse fall into — I hesitate to say error — chimera and reverie, into pure and simple linguistic monstrosity. Within its own limits, each discipline recognises true and false propositions; but it pushes back a whole teratology of knowledge beyond its margins. The exterior of a science is both more and less populated than is often believed: there is of course immediate experience, the imaginary themes which endlessly carry and renew immemorial beliefs; but perhaps there are no errors in the strict sense, for error can only arise and be decided inside a definite practice; on the other hand, there are monsters on the prowl whose form changes with the history of knowledge. In short, a proposition must fulfil complex and heavy requirements to be able to belong to the grouping of a discipline; before it can be called true or false, it must be ‘in the true’, as Canguilhem would say. People have often wondered how the botanists or biologists of the nineteenth century managed not to see that what Mendel was saying was true. But it was because Mendel was speaking of objects, applying methods, and placing himself on a theoretical horizon which were alien to the biology of his time. Naudin, before him, had of course posited the thesis that hereditary traits are discrete; yet, no matter how new or strange this principle was, it was able to fit into the discourse of biology, at least as an enigma. What Mendel did was to constitute the hereditary trait as an absolutely new biological object, thanks to a kind of filtering which had never been used before: he detached the trait from the species, and from the sex which transmits it; the field in which he observed it being the infinitely open series of the generations, where it appears and disappears according to statistical regularities. This was a new object which called for new conceptual instruments and new theoretical foundations. Mendel spoke the truth, but he was not ‘within the true’ of the biological discourse of his time: it was not according to such rules that biological objects and concepts were formed. It needed a complete change of scale, the deployment of a whole new range of objects in biology for Mendel to enter into the true and for his propositions to appear (in large measure) correct. Mendel was a true monster, which meant that science could not speak of him; whereas about thirty years earlier, at the height of the nineteenth century, Scheiden, for example, who denied plant sexuality, but in accordance with the rules of biological discourse, was merely formulating a disciplined error. It is always possible that one might speak the truth in the space of a wild exteriority, but one is ‘in the true’ only by obeying the rules of a discursive ‘policing’ which one has to reactivate in each of one’s discourses. The discipline is a principle of control over the production of discourse. The discipline fixes limits for discourse by the action of an identity which takes the form of a permanent re-actuation of the rules. We are accustomed to see in an author’s fecundity, in the multiplicity of the commentaries, and in the development of a discipline so many infinite resources for the creation of discourses. Perhaps so, but they are nonetheless principles of constraint; it is very likely impossible to account for their positive and multiplicatory role if we do not take into consideration their restrictive and constraining function.

IV

There is, I believe, a third group of procedures which permit the control of discourses. This time it is not a matter of mastering their powers or averting the unpredictability of their appearance, but of determining the condition of their application, of imposing a certain number of rules on the individuals who hold them, and thus of not permitting everyone to have access to them. There is a rarefaction, this time, of the speaking subjects; none shall enter the order of discourse if he does not satisfy certain requirements or if he is not, from the outset, qualified to do so. To be more precise: not all the regions of discourse are equally open and penetrable; some of them are largely forbidden (they are differentiated and differentiating), while others seem to be almost open to all winds and put at the disposal of every speaking subject, without prior restrictions. In this regard I should like to recount an anecdote which is so beautiful that one trembles at the thought that it might be true. It gathers into a single figure all the constraints of discourse: those which limit its powers, those which master its aleatory appearances, those which carry out the selection among speaking subjects. At the beginning of the seventeenth century, the Shogun heard tell that the Europeans’ superiority in matters of navigation, commerce, politics, and military skill was due to their knowledge of mathematics. He desired to get hold of so precious a knowledge. As he had been told of an English sailor who possessed the secret of these miraculous discourses, he summoned him to his palace and kept him there. Alone with him, he took lessons. He learned mathematics. He retained power, and lived to a great old age. It was not until the nineteenth century that there were Japanese mathematicians. But the anecdote does not stop there: it has its European side too. The story has it that this English sailor, Will Adams, was an autodidact, a carpenter who had learnt geometry in the course of working in a shipyard. Should we see this story as the expression of one of the great myths of European culture? The universal communication of knowledge and the infinite free exchange of discourses in Europe, against the monopolised and secret knowledge of Oriental tyranny? This idea, of course, does not stand up to examination. Exchange and communication are positive figures working inside complex systems of restriction, and probably would not be able to function independently of them. The most superficial and visible of these systems of restriction is constituted by what can be gathered under the name of ritual. Ritual defines the qualification which must be possessed by individuals who speak (and who must occupy such-and-such a position and formulate suchand-such a type of statement, in the play of a dialogue, of interrogation or recitation); it defines the gestures, behaviour, circumstances, and the whole set of signs which must accompany discourse; finally, it fixes the supposed or imposed efficacity of the words, their effect on those to whom they are addressed, and the limits of their constraining value. Religious, judicial, therapeutic , and in large measure also political discourses can scarcely be dissociated from this deployment of a ritual which determines both the particular properties and the stipulated roles of the speaking subjects. A somewhat different way of functioning is that of the ‘societies of discourse’, which function to preserve or produce discourses, but in order to make them circulate in a closed space, distributing them only according to strict rules, and without the holders being dispossessed by this distribution. An archaic model for this is provided by the groups of rhapsodists who possessed the knowledge of the poems to be recited or potentially to be varied and transformed. But though the object of this knowledge was after all a ritual recitation, the knowledge was protected, defended and preserved within a definite group by the often very complex exercises of memory which it implied. To pass an apprenticeship in it allowed one to enter both a group and a secret which the act of recitation showed but did not divulge; the roles of speaker and listener were not interchangeable. There are hardly any such ‘societies of discourse’ now, with their ambiguous play of the secret and its divulgation. But this should not deceive us: even in the order of ‘true’ discourse, even in the order of discourse that is published and free from all ritual, there are still forms of appropriation of secrets, and non-interchangeable roles. It may well be that the act of writing as it is institutionalised today, in the book, the publishingsystem and the person of the writer, takes place in a ‘society of discourse’, which though diffuse is certainly constraining. The difference between the writer and any other speaking or writing subject (a difference constantly stressed by the writer himself), the intransitive nature (according to him) of his discourse, the fundamental singularity which he has been ascribing for so long to ‘writing’, the dissymmetry that is asserted between ‘creation’ and any use of the linguistic system — all this shows the existence of a certain ‘society of discourse’, and tends moreover to bring back its play of practices. But there are many others still, functioning according to entirely different schemas of exclusivity and disclosure: e.g., technical or scientific secrets, or the forms of diffusion and circulation of medical discourse, or those who have appropriated the discourse of politics or economics. At first glance, the ‘doctrines’ (religious, political, philosophical) seem to constitute the reverse of a ‘society of discourse’, in which the number of speaking individuals tended to be limited even if it was not fixed; between those individuals, the discourse could circulate and be transmitted. Doctrine, on the contrary, tends to be diffused, and it is by the holding in common of one and the same discursive ensemble that individuals (as many as one cares to imagine) define their reciprocal allegiance. In appearance, the only prerequisite is the recognition of the same truths and the acceptance of a certain rule of (more or less flexible) conformity with the validated discourses. If doctrines were nothing more than this, they would not be so very different from scientific disciplines, and the discursive control would apply only to the form or the content of the statement, not to the speaking subject. But doctrinal allegiance puts in question both the statement and the speaking subject, the one by the other. It puts the speaking subject in question through and on the basis of the statement, as is proved by the procedures of exclusion and the mechanisms of rejection which come into action when a speaking subject has formulated one or several unassimilable statements; heresy and orthodoxy do not derive from a fanatical exaggeration of the doctrinal mechanisms, but rather belong fundamentally to them. And conversely the doctrine puts the statements in question on the basis of the speaking subjects, to the extent that the doctrine always stands as the sign, manifestation and instrument of a prior adherence to a class, a social status, a race, a nationality, an interest, a revolt, a resistance or an acceptance. Doctrine binds individuals to certain types of enunciation and consequently forbids them all others; but it uses, in return, certain types of enunciation to bind individuals amongst themselves, and to differentiate them by that very fact from all others. Doctrine brings about a double subjection: of the speaking subjects to discourses, and of discourses to the (at least virtual) group of speaking individuals. On a much broader scale, we are obliged to recognise large cleavages in what might be called the social appropriation of discourses. Although education may well be, by right, the instrument thanks to which any individual in a society like ours can have access to any kind of discourse whatever, this does not prevent it from following, as is well known, in its distribution, in what it allows and what it prevents, the lines marked out by social distances, oppositions and struggles. Any system of education is a political way of maintaining or modifying the appropriation of discourses, along with the knowledges and powers which they carry. I am well aware that it is very abstract to separate speechrituals, societies of discourse, doctrinal groups and social appropriations, as I have just done. Most of the time, they are linked to each other and constitute kinds of great edifices which ensure the distribution of speaking subjects into the different types of discourse and the appropriation of discourses to certain categories of subject. Let us say, in a word, that those are the major procedures of subjection used by discourse. What, after all, is an education system, other than a ritualisation of speech, a qualification and a fixing of the roles for speaking subjects, the constitution of a doctrinal group, however diffuse, a distribution and an appropriation of discourse with its powers and knowledges? What is ‘ecriture’ (the writing of the ‘writers’) other than a similar system of subjection, which perhaps takes slightly different forms, but forms whose main rhythms are analogous? Does not the judicial system, does not the institutional system of medicine likewise constitute, in some of their aspects at least, similar systems of subjection of and by discourse?

V

I wonder whether a certain number of themes in philosophy have not come to correspond to these activities of limitation and exclusion , and perhaps also to reinforce them. They correspond to them first of all by proposing an ideal truth as the law of discourse and an immanent rationality as the principle of their unfolding, and they re-introduce an ethic of knowledge, which promises to give the truth only to the desire for truth itself and only to the power of thinking it. Then they reinforce the limitations and exclusions by a denial of the specific reality of discourse in general. Ever since the sophists’ tricks and influence were excluded and since their paradoxes have been more or less safely muzzled, it seems that Western thought has taken care to ensure that discourse should occupy the smallest possible space between thought and speech. Western thought seems to have made sure that the act of discoursing should appear to be no more than a certain bridging (apport) between thinking and speaking — a thought dressed in its signs and made visible by means of words, or conversely the very structures of language put into action and producing a meaning-effect. This very ancient elision of the reality of discourse in philosophical thought has taken many forms in the course of history. We have seen it again quite recently in the guise of several familiar themes. Perhaps the idea of the founding subject is a way of eliding the reality of discourse. The founding subject, indeed, is given the task of directly animating the empty forms of language with his aims; it is he who in moving through the density and inertia of empty things grasps by intuition the meaning lying deposited within them; it is likewise the founding subject who founds horizons of meaning beyond time which history will henceforth only have to elucidate and where propositions, sciences and deductive ensembles will find their ultimate grounding. In his relation to meaning, the founding subject has at his disposal signs, marks, traces, letters. But he does not need to pass via the singular instance of discourse in order to manifest them. The opposing theme, that of originating experience, plays an analogous role. It supposes that at the very basis of experience, even before it could be grasped in the form of a cogito, there were prior significations — in a sense, already said — wandering around in the world, arranging it all around us and opening it up from the outset to a sort of primitive recognition. Thus a primordial complicity with the world is supposed to be the foundation of our possibility of speaking of it, in it, of indicating it and naming it, of judging it and ultimately of knowing it in the form of truth. If there is discourse, then, what can it legitimately be other than a discreet reading? Things are already murmuring meanings which our language has only to pick up; and this language, right from its most rudimentary project, was already speaking to us of a being of which it is like the skeleton. The idea of universal mediation is yet another way, I believe, of eliding the reality of discourse, and despite appearances to the contrary. For it would seem at first glance that by rediscovering everywhere the movement of a logos which elevates particularities to the status of concepts and allows immediate consciousness to unfurl in the end the whole rationality of the world, one puts discourse itself at the centre of one’s speculation. But this logos, in fact, is only a discourse that has already been held, or rather it is things themselves, and events, which imperceptibly turn themselves into discourse as they unfold the secret of their own essence. Thus discourse is little more than the gleaming of a truth in the process of being born to its own gaze; and when everything finally can take the form of discourse, when everything can be said and when discourse can be spoken about everything, it is because all things, having manifested and exchanged their meaning, can go back into the silent interiority of their consciousness of self. Thus in a philosophy of the founding subject, in a philosophy of originary experience, and in a philosophy of universal mediation alike, discourse is no more than a play, of writing in the first case, of reading in the second, and of exchange in the third, and this exchange, this reading, this writing never put anything at stake except signs. In this way, discourse is annulled in its reality and put at the disposal of the signifier. What civilisation has ever appeared to be more respectful of discourse than ours? Where has it ever been more honoured, or better honoured? Where has it ever been, seemingly, more radically liberated from its constraints, and universalised? Yet it seems to me that beneath this apparent veneration of discourse, under this apparent logophilia, a certain fear is hidden. It is just as if prohibitions, barriers, thresholds and limits had been set up in order to master, at least partly, the great proliferation of discourse, in order to remove from its richness the most dangerous part, and in order to organise its disorder according to figures which dodge what is most uncontrollable about it. It is as if we had tried to efface all trace of its irruption into the activity of thought and language. No doubt there is in our society, and, I imagine, in all others, but following a different outline and different rhythms, a profound logophobia, a sort of mute terror against these events, against this mass of things said, against the surging-up of all these statements, against all that could be violent, discontinuous, pugnacious, disorderly as well, and perilous about them — against this great incessant and disordered buzzing of discourse. And if we want to — I would not say, efface this fear, but — analyse it in its conditions, its action and its effects, we must, I believe, resolve to take three decisions which our thinking today tends to resist and which correspond to the three groups of functions which I have just mentioned: we must call into question our will to truth, restore to discourse its character as an event, and finally throw off the sovereignty of the signifier.

VI

These are the tasks, or rather some of the themes, which govern the work I should like to do here in the coming years. We can see at once certain methodological requirements which they imply. First of all, a principle of reversal: where tradition sees the source of discourses, the principle of their swarming abundance and of their continuity, in those figures which seem to play a positive role, e.g. , those of the author, the discipline, the will to truth, we must rather recognise the negative action of a cutting-up and a rarefaction of discourse. But once we have noticed these principles of rarefaction, once we have ceased to consider them as a fundamental and creative instance, what do we discover underneath them? Must we admit the virtual plenitude of a world of uninterrupted discourses? This is where we have to bring other methodological principles into play. A principle of discontinuity, then: the fact that there are systems of rarefaction does not mean that beneath them or beyond them there reigns a vast unlimited discourse, continuous and silent, which is quelled and repressed by them, and which we have the task of raising up by restoring the power of speech to it. We must not imagine that there is a great unsaid or a great unthought which runs throughout the world and intertwines with all its forms and all its events, and which we would have to articulate or to think at last. Discourses must be treated as discontinuous practices, which cross each other, are sometimes juxtaposed with one another, but can just as well exclude or be unaware of each other. A principle of specificity: we must not resolve discourse into a play of pre-existing significations; we must not imagine that the world turns towards us a legible face which we would have only to decipher; the world is not the accomplice of our knowledge; there is no prediscursive providence which disposes the world in our favour. We must conceive discourse as a violence which we do to things, or in any case as a practice which we impose on them; and it is in this practice that the events of discourse find the principle of their regularity. The fourth rule is that of exteriority: we must not go from discourse towards its interior, hidden nucleus, towards the heart of a thought or a signification supposed to be manifested in it; but, on the basis of discourse itself, its appearance and its regularity, go towards its external conditions of possibility, towards what gives rise to the aleatory series of these events, and fixes its limits. Four notions, then, must serve as the regulating principle of the analysis: the event, the series, the regularity, the condition of possibility. Term for term we find the notion of event opposed to that of creation, series opposed to unity, regularity opposed to originality, and condition of possibility opposed to signification. These other four notions (signification, originality, unity, creation) have in a general way dominated the traditional history of ideas, where by common agreement one sought the point of creation, the unity of a work, an epoch or a theme, the mark of individual originality, and the infinite treasure of buried significations. I will add only two remarks. One concerns history. It is often entered to the credit of contemporary history that it removed the privileges once accorded to the singular event and revealed the structures of longer duration. That is so. However, I am not sure that the work of these historians was exactly done in this direction. Or rather I do not think there is an inverse ratio between noticing the event and analysing the long durations. On the contrary, it seems to be by pushing to its extreme the fine grain of the event, by stretching the resolution-power of historical analysis as far as official price-lists (les mercuriales), title deeds, parish registers, harbour archives examined year by year and week by week, that these historians saw — beyond the battles, decrees, dynasties or assemblies — the outline of massive phenomena with a range of a hundred or many hundreds of years. History as practised today does not turn away from events; on the contrary, it is constantly enlarging their field, discovering new layers of them, shallower or deeper. It is constantly isolating new sets of them, in which they are sometimes numerous, dense and interchangeable, sometimes rare and decisive: from the almost daily variations in price to inflations over a hundred years. But the important thing is that history does not consider an event without defining the series of which it is part, without specifying the mode of analysis from which that series derives, without seeking to find out the regularity of phenomena and the limits of probability of their emergence, without inquiring into the variations, bends and angles of the graph, without wanting to determine the conditions on which they depend. Of course, history has for a long time no longer sought to understand events by the action of causes and effects in the formless unity of a great becoming, vaguely homogeneous or ruthlessly hierarchised; but this change was not made in order to rediscover prior structures, alien and hostile to the event. It was made in order to establish diverse series, intertwined and often divergent but not autonomous, which enable us to circumscribe the ‘place’ of the event, the margins of its chance variability, and the conditions of its appearance. The fundamental notions which we now require are no longer those of consciousness and continuity (with their correlative problems of freedom and causality), nor any longer those of sign and structure. They are those of the event and the series, along with the play of the notions which are linked to them: regularity, dimension of chance (alea), discontinuity, dependence, transformation; it is by means of a set of notions like this that my projected analysis of discourses is articulated, not on the traditional thematics which the philosophers of yesterday still take for ‘living’ history, but on the effective work of historians. Yet it is also in this regard that this analysis poses philosophical, or theoretical, problems, and very likely formidable ones. If discourses must be treated first of all as sets of discursive events, what status must be given to that notion of event which was so rarely taken into consideration by philosophers? Naturally the event is neither substance nor accident, neither quality nor process; the event is not of the order of bodies. And yet it is not something immaterial either; it is always at the level of materiality that it takes effect, that it is effect; it has its locus and it consists in the relation, the coexistence, the dispersion, the overlapping, the accumulation, and the selection of material elements. It is not the act or the property of a body; it is produced as an effect of, and within, a dispersion of matter. Let us say that the philosophy of the event should move in the at first sight paradoxical direction of a materialism of the incorporeal. Furthermore, if discursive events must be treated along the lines of homogeneous series which, however, are discontinuous in relation to each other, what status must be given to this discontinuity? It is of course not a matter of the succession of instants in time, nor of the plurality of different thinking subjects. It is a question of caesurae which break up the instant and disperse the subject into a plurality of possible positions and functions. This kind of discontinuity strikes and invalidates the smallest units that were traditionally recognised and which are the hardest to contest: the instant and the subject. Beneath them, and independently of them, we must conceive relations between these discontinuous series which are not of the order of succession (or simultaneity) within one (or several) consciousnesses; we must elaborate — outside of the philosophies of the subject and of time — a theory of discontinuous systematicities. Finally, though it is true that these discontinuous discursive series each have, within certain limits, their regularity, it is undoubtedly no longer possible to establish links of mechanical causality or of ideal necessity between the elements which constitute them. We must accept the introduction of the alea as a category in the production of events. There once more we feel the absence of a theory enabling us to think the relations between chance and thought. The result is that the narrow gap which is to be set to work in the history of ideas, and which consists of dealing not with the representations which might be behind discourse, but with discourses as regular and distinct series of events — this narrow gap looks, I’m afraid, like a small (and perhaps odious) piece of machinery which would enable us to introduce chance, the discontinuous, and materiality at the very roots of thought. This is a triple peril which a certain form of history tries to exorcise by narrating the continuous unravelling of an ideal necessity. They are three notions that should allow us to connect the history of systems of thought to the practice of historians. And they are three directions which the work of theoretical elaboration will have to follow.

VII

The analyses which I propose to make, following these principles and making this horizon my line of reference, will fall into two sets. On the one hand the ‘critical’ section, which will put into practice the principle of reversal: trying to grasp the forms of exclusion, of limitation, of appropriation of which I was speaking just now; showing how they are formed, in response to what needs, how they have been modified and displaced, what constraint they have effectively exerted, to what extent they have been evaded. On the other hand there is the ‘genealogical’ set, which puts the other three principles to work: how did series of discourses come to be formed, across the grain of, in spite of, or with the aid of these systems of constraints; what was the specific norm of each one, and what were their conditions of appearance, growth, variation. First, the critical set. A first group of analyses might deal with what I have designated as functions of exclusion. I formerly studied one of them, in respect of one determinate period: the divide between madness and reason in the classical epoch. Later, I might try to analyse a system of prohibition of language, the one concerning sexuality from the sixteenth to the nineteenth century. The aim would be to see not how this interdiction has been progressively and fortunately effaced, but how it has been displaced and re-articulated from a practice of confession in which the forbidden behaviour was named, classified, hierarchised in the most explicit way, up to the appearance, at first very timid and belated, of sexual thematics in nineteenth-century medicine and psychiatry; of course these are still only somewhat symbolic orientation-points, but one could already wager that the rhythms are not the ones we think, and the prohibitions have not always occupied the place that we imagine. In the immediate future, I should like to apply myself to the third system of exclusion; this I envisage in two ways. On the one hand, I want to try to discover how this choice of truth, inside which we are caught but which we ceaselessly renew, was made — but also how it was repeated, renewed, and displaced. I will consider first the epoch of the Sophists at its beginning, with Socrates, or at least with Platonic philosophy, to see how efficacious discourse, ritual discourse, discourse loaded with powers and perils, gradually came to conform to a division between true and false discourse. Then I will consider the turn of the sixteenth century, at the time when there appears, especially in England, a science of the gaze, of observation, of the established fact, a certain natural philosophy, no doubt inseparable from the setting-up of new political structures, and, inseparable, too, from religious ideology; this was without a doubt a new form of the will to know. Finally, the third orientation-point will be the beginning of the nineteenth century, with its great acts that founded modern science, the formation of an industrial society and the positivist ideology which accompanied it. These will be my three cross-sections in the morphology of our will to know, three stages of our philistinism. I would also like to take up the same question again, but from a quite different angle: to measure the effect of a discourse with scientific claims — a medical, psychiatric, and also sociological discourse — on that set of practices and prescriptive discourses constituted by the penal system. The starting point and basic material for this analysis will be the study of psychiatric expertise and its role in penal practices. Still looking at it from this critical perspective, but at another level, the procedures of limitation of discourses should be analysed. I indicated several of these just now: the principle of the author, of commentary, of the discipline. A certain number of studies can be envisaged from this perspective. I am thinking, for example, of an analysis of the history of medicine from the sixteenth to the nineteenth century. The objective would be not so much to pinpoint the discoveries made or the concepts put to work, but to grasp how, in the construction of medical discourse, and also in the whole institution that supports, transmits and reinforces it, the principle of the author, of the commentary, and of the discipline were used. The analysis would seek to find out how the principle of the great author operated: Hippocrates and Galen, of course, but also Paracelsus, Sydenham, or Boerhaave. It would seek to find out how the practice of the aphorism and the commentary were carried on, even late into the nineteenth century, and how they gradually gave place to the practice of the case, of the collection of cases, of the clinical apprenticeship using a concrete case. It would seek to discover, finally, according to what model medicine tried to constitute itself as a discipline, leaning at first on natural history, then on anatomy and biology. One could also consider the way in which literary criticism and literary history in the eighteenth and nineteenth centuries constituted the person of the author and the figure of the oeuvre, using, modifying, and displacing the procedures of religious exegesis, biblical criticism, hagiography, historical or legendary ‘lives’, autobiography, and memoirs. One day we will also have to study the role played by Freud in psychoanalytic knowledge, which is surely very different from that of Newton in physics (and of all founders of disciplines), and also very different from the role that can be played by an author in the field of philosophical discourse (even if, like Kant, he is at the origin of a different way of philosophising). So there are some projects for the critical side of the task, for the analysis of the instances of discursive control. As for the genealogical aspect, it will concern the effective formation of discourse either within the limits of this control, or outside them, or more often on both sides of the boundary at once. The critical task will be to analyse the processes of rarefaction, but also of regrouping and unification of discourses; genealogy will study their formation, at once dispersed, discontinuous, and regular. In truth these two tasks are never completely separable: there are not, on one side, the forms of rejection, exclusion, regrouping and attribution, and then on the other side, at a deeper level, the spontaneous surging-up of discourses which, immediately before or after their manifestation, are submitted to selection and control. The regular formation of discourse can incorporate the procedures of control, in certain conditions and to a certain extent (that is what happens, for instance, when a discipline takes on the form and status of a scientific discourse); and conversely the figures of control can take shape within a discursive formation (as is the case with literary criticism as the discourse that constitutes the author): so much so that any critical task, putting in question the instances of control, must at the same time analyse the discursive regularities through which they are formed; and any genealogical description must take into account the limits which operate in real formations. The difference between the critical and the genealogical enterprise is not so much a difference of object or domain, but of point of attack, perspective, and delimitation. Earlier on I mentioned one possible study, that of the taboos which affect the discourse of sexuality. It would be difficult, and in any case abstract, to carry out this study without analysing at the same time the sets of discourses — literary, religious or ethical, biological or medical, juridical too — where sexuality is discussed, and where it is named, described, metaphorised, explained, judged. We are very far from having constituted a unitary and regular discourse of sexuality; perhaps we never will, and perhaps it is not in this direction that we are going. No matter. The taboos do not have the same form and do not function in the same way in literary discourse and in medical discourse, in that of psychiatry or in that of the direction of conscience. Conversely, these different discursive regularities do not have the same way of reinforcing, evading, or displacing the taboos. So the study can be done only according to pluralities of series in which there are taboos at work which are at least partly different in each. One could also consider the series of discourses which in the sixteenth and seventeeth centuries dealt with wealth and poverty, money, production, commerce. We are dealing there with sets of very heterogeneous statements, formulated by the rich and the poor, the learned and the ignorant, protestants and catholics, officers of the king, traders or moralists. Each one has its own form of regularity, likewise its own systems of constraint. None of them exactly prefigures that other form of discursive regularity which will later take on the air of a discipline and which will be called ‘the analysis of wealth’, then ‘political economy’. Yet it is on the basis of this series that a new regularity was formed, taking up or excluding, justifying or brushing aside this one or that one of their utterances. We can also conceive of a study which would deal with the discourses concerning heredity, such as we can find them, up to the beginning of the twentieth century, scattered and dispersed through various disciplines, observations, techniques and formulae. The task would then be to show by what play of articulation these series in the end recomposed themselves, in the epistemologically coherent and institutionally recognised figure of genetics. This is the work that has just been done by Francois Jacob with a brilliance and an erudition which could not be equalled. Thus the critical and the genealogical descriptions must alternate, and complement each other, each supporting the other by turns. The critical portion of the analysis applies to the systems that envelop discourse, and tries to identify and grasp these principles of sanctioning, exclusion, and scarcity of discourse. Let us say, playing on words, that it practises a studied casualness. The genealogical portion, on the other hand, applies to the series where discourse is effectively formed: it tries to grasp it in its power of affirmation, by which I mean not so much a power which would be opposed to that of denying, but rather the power to constitute domains of objects, in respect of which one can affirm or deny true or false propositions. Let us call these domains of objects positivities, and let us say, again playing on words, that if the critical style is that of studious casualness, the genealogical mood will be that of a happy positivism. In any event, one thing at least has to be emphasised: discourse analysis understood like this does not reveal the universality of a meaning, but brings to light the action of imposed scarcity, with a fundamental power of affirmation. Scarcity and affirmation; ultimately, scarcity of affirmation, and not the continuous generosity of meaning, and not the monarchy of the signifier. And now, let those with gaps in their vocabulary say — if they find the term more convenient than meaningful — that all this is structuralism.

VIII

I know that but for the aid of certain models and supports I would not have been able to undertake these researches which I have tried to sketch out for you. I believe I am greatly indebted to Georges Dumezl, since it was he who urged me to work, at an age when I still thought that to write was a pleasure. But I also owe a great deal to his work. May he forgive me if I have stretched the meaning or departed from the rigour of those texts which are his and which dominate us today. It was he who taught me to analyse the internal economy of a discourse in a manner quite different from the methods of traditional exegesis or linguistic formalism. It was he who taught me to observe the system of functional correlations between discourses by the play of comparisons from one to the other. It was he who taught me how to describe the transformations of a discourse and its relations to institutions. If I have tried to apply this method to discourses quite different from legendary or mythical narratives, it was probably because I had in front of me the works of the historians of science, especially Georges Canguilhem. It is to him that I owe the insight that the history of science is not necessarily caught in an alternative: either to chronicle discoveries or to describe the ideas and opinions that border science on the side of its indeterminate genesis or on the side of its later expulsions, but that it was possible and necessary to write the history of science as a set of theoretical models and conceptual instruments which is both coherent and transformable. But I consider that my greatest debt is to Jean Hyppolite. I am well aware that in the eyes of many his work belongs under the aegis of Hegel, and that our entire epoch, whether in logic or epistemology, whether in Marx or Nietzsche, is trying to escape from Hegel: and what I have tried to say just now about discourse is very unfaithful to the Hegelian logos. But to make a real escape from Hegel presupposes an exact appreciation of what it costs to detach ourselves from him. It presupposes a knowledge of how close Hegel has come to us, perhaps insidiously. It presupposes a knowledge of what is still Hegelian in that which allows us to think against Hegel; and an ability to gauge how much our resources against him are perhaps still a ruse which he is using against us, and at the end of which he is waiting for us, immobile and elsewhere. If so many of us are indebted to Jean Hyppolite, it is because he tirelessly explored, for us and ahead of us, this path by which one gets away from Hegel, establishes a distance, and by which one ends up being drawn back to him, but otherwise, and then constrained to leave him once again. First of all Jean Hyppolite took the trouble to give a presence to the great and somewhat ghostly shadow of Hegel which had been on the prowl since the nineteenth century and with which people used to wrestle obscurely. It was by means of a translation (of the ‘Phenomenology of Mind’) that he gave Hegel this presence. And the proof that Hegel himself is well and truly present in this French text is the fact that even Germans have consulted it so as to understand better what, for a moment at least, was going on in the German version. Jean Hyppolite sought and followed all the ways out of this text, as if his concern was: can we still philosophise where Hegel is no longer possible? Can a philosophy still exist and yet not be Hegelian? Are the non-Hegelian elements in our thought also necessarily non-philosophical? And is the anti-philosophical necessarily non-Hegelian? So that he was not merely trying to give a meticulous historical description of this presence of Hegel: he wanted to make it into one of modernity’s schemata of experience (is it possible to think science, history, politics and everyday suffering in the Hegelian mode?); and conversely he wanted to make our modernity the test of Hegelianism and thereby of philosophy. For him the relation to Hegel was the site of an experiment , a confrontation from which he was never sure that philosophy would emerge victorious. He did not use the Hegelian system as a reassuring universe; he saw in it the extreme risk taken by philosophy. Hence, I believe, the displacements he carried out, not so much within Hegelian philosophy but upon it, and upon philosophy as Hegel conceived it. Hence also a whole inversion of themes. Instead of conceiving philosophy as the totality at last capable of thinking itself and grasping itself in the movement of the concept, Jean Hyppolite made it into a task without end set against an infinite horizon: always up early, his philosophy was never ready to finish itself. A task without end, and consequently a task forever re-commenced, given over to the form and the paradox of repetition: philosophy as the inaccessible thought of the totality was for Jean Hyppolite the most repeatable thing in the extreme irregularity of experience; it was what is given and taken away as a question endlessly taken up again in life, in death, in memory. In this way he transformed the Hegelian theme of the closure on to the consciousness of self into a theme of repetitive interrogation. But philosophy, being repetition, was not ulterior to the concept; it did not have to pursue the edifice of abstraction, it had always to hold itself back, break with its acquired generalities and put itself back in contact with nonphilosophy. It had to approach most closely not the thing that completes it but the thing that precedes it, that is not yet awakened to its disquiet. It had to take up the singularity of history, the regional rationalities of science, the depth of memory within consciousness — not in order to reduce them but in order to think them. Thus there appears the theme of a philosophy that is present, disquieted, mobile all along its line of contact with non-philosophy, yet existing only by means of non-philosophy and revealing the meaning it has for us. If philosophy is in this repeated contact with non-philosophy, what is the beginning of philosophy? Is philosophy already there, secretly present in what is not itself, starting to formulate itself half-aloud in the murmur of things? But then perhaps philosophical discourse no longer has a raison d’etre; or must it begin from a foundation that is at once arbitrary and absolute? In this way the Hegelian theme of the movement proper to the immediate is replaced by that of the foundation of philosophical discourse and its formal structure. And finally the last displacement that Jean Hyppolite carried out on Hegelian philosophy: if philosophy must begin as an absolute discourse, what about history? And what is this beginning which begins with a single individual, in a society, in a social class, and in the midst of struggles? These five displacements, leading to the extreme edge of Hegelian philosophy, and no doubt pushing it over on to the other side of its own limits, summon up one by one the great figures of modern philosophy, whom Hyppolite never ceased confronting with Hegel: Marx with the questions of history, Fichte with the problem of the absolute beginning of philosophy, Bergson with the theme of contact with the non-philosophical, Kierkegaard with the problem of repetition and truth, Husserl with the theme of philosophy as an infinite task linked to the history of our rationality. And beyond these philosophical figures we perceive all the domains of knowledge that Jean Hyppolite invoked around his own questions: psychoanalysis with the strange logic of desire; mathematics and the formalisation of discourse; informationtheory and its application in the analysis of living beings; in short, all those domains about which one can ask the question of a logic and an existence which never stop tying and untying their bonds. I believe that Hyppolite’s work, articulated in several major books, but invested even more in his researches, in his teaching, in his perpetual attention, in his constant alertness and generosity , in his responsibilities which were apparently administrative and pedagogic but in reality doubly political, came upon and formulated the most fundamental problems of our epoch. There are many of us who owe him an infinite debt. It is because I have no doubt borrowed from him the meaning and possibility of what I am doing, and because he very often gave me illumination when I was working in the dark, that I wanted to place my work under his sign, and that I wanted to conclude this presentation of my plans by evoking him. It is in his direction, towards this lack — in which I feel both his absence and my own inadequacy — that my questionings are now converging. Since I owe him so much, I can well see that in choosing to invite me to teach here, you are in large part paying homage to him. I am grateful to you, profoundly grateful, for the honour that you have done me, but I am no less grateful for the part he plays in this choice. Though I do not feel equal to the task of succeeding him, I know that, on the other hand, if such a happiness could have been granted us tonight, he would have encouraged me by his indulgence. And now I understand better why I found it so difficult to begin just now. I know now whose voice it was that I would have liked to precede me, to carry me, to invite me to speak, to lodge itself in my own discourse. I know what was so terrifying about beginning to speak, since I was doing so in this place where I once listened to him, and where he is no longer here to hear me.