• 1 There are some exceptions : P. Gillot, Althusser et la psychanalyse, Paris, Presses Universitaires (...)

2In certain cases, the mere absence of a word, given that not all absences are equal, may produce effects in a given utterance or text. The word « ideology », for example, the word to which Althusser’s name is perhaps more frequently attached than any other, does not appear a single time in “The Underground Current”. But before we attempt to assign a meaning to this absence, as if, for example, it represents a rejection of the concept on his part, we must first both acknowledge and confront the uneven and contradictory development of Althusser’s notion of ideology. The discontinuous accounts of ideology in the brief period from “Marxism and Humanism” (1964) to the “Three Notes on a Theory of Discourses” (1966) and finally “Ideology and the Ideological State Apparatuses” (1970) might appear as stages of development in the progressive appropriation of the materiality of ideology. According to such a conception, Althusser’s theory of ideology would be driven by the overcoming of contradictions on the way to the ISAs essay. The earlier conceptions of ideology are thereby transformed into anticipations, weaker and less coherent versions, of the later ones. Such a view, however, deprives the attempts to theorize ideology prior to the ISAs essay of the interest and force proper to them, especially when they conflict with or simply complicate the theses of the 1970 essay. Thus another absence appears, not in Althusser’s work, but around it, in what has been written about it : in the twenty years since its posthumous publication, the “Three Notes” has been generally overlooked, despite the fact that it focuses on ideology, the most celebrated aspect of Althusser’s work in the Anglophone world.1 In part, this text has been overlooked because it derives from Althusser’s attempt to develop a theory of discourses as a way of conceptualizing ideology. “Discourse” for most of those who read Althusser (in the Anglophone world) now has a dated air about it, as if it has outlived its usefulness or was never anything other than a place-holder for other, more refined concepts. But perhaps such a moment, that is, the moment a once ubiquitous concept is shunned and ignored, is precisely when it becomes possible to think about it in a new way.

• 2 P. Macherey, "Langue, discours, idéologie, sujet, sens : de Thomas Herbert à Michel Pêcheux", in Dé (...)

3Accordingly, I want to argue, against chronology, that the discussion of discourse and ideology that took place between Althusser and a number of his colleagues (Balibar, Macherey, Badiou, Duroux, and, above all, Pêcheux) in the period 1966-67 was not only not superseded by Althusser’s later text, but in fact offers “solutions” to some of its most difficult and persistent problems. Thus, the fact that Althusser rather quickly abandoned this project should not deter – or excuse – us from reading the “Three Notes”. Moreover, while Althusser moved away from this line of inquiry, Pêcheux (whose name is seldom seen today in Anglophone studies, even in conjunction with Althusser’s) continued to develop it in ways that confer considerable interest on the work produced in the course of that exchange. As Pierre Macherey has noted, Pêcheux’s work presents significant difficulties to the reader and not only because it develops within the element of linguistics and must thus confront the contradictions proper to theories of syntax and semantics which may, especially in the latter case, differ radically from each other in their assumptions2. A work like Les verités de La Palice (1975) shows the extent to which Pêcheux was willing to pursue different lines of inquiry on the hypothesis that they would converge without ever forcing or inventing this convergence, even when it remained to be discovered. To read his texts today, however, is to see the extent to which his work, not despite but because of its theoretical specificity, constitutes an ongoing dialogue with Althusser which is important for our own time and which begins with the otherwise overlooked discussion of discourse that took place in 1966-1967.

• 3 L. Althusser, "Trois notes sur la théorie des discours", in Écrits sur la psychanalyse, Paris, Stoc (...)

4I want to propose a reading of two texts, Althusser’s “Three Notes” (written in 1966) and Pêcheux’s “Réflexions sur la situation théorique des sciences sociales et, spécialement, de la psychologie sociale”, which appeared in Cahiers pour l’analyse (also in 1966, under the pseudonym Thomas Herbert), to explore the hypothesis that they illuminate a critical but under-analyzed element that haunts the discussion of ideology that took place around Althusser in the sixties.3 The element in question is precisely that of an essential link between ideology and the unconscious, not only because of the attributes they share (transhistoricity, a decentering of the subject, transindividuality, etc.), but because they are unthinkable except in relation to language, not insofar as it represents a formal system (or system of systems) but as discourse (the point of indistinction between system and history which resides precisely in the cracks and fissures that open language to the outside). The identification of this link was certainly a consequence of what Pêcheux called the triple alliance between (Althusserian) Marxism, (Lacanian) psychoanalysis and linguistics, but the theoretical encounter that produced the texts referred to above could not have taken place in the absence of a specific reading of Spinoza, the reading that emerged around Lire “Le Capital” (1965) and Macherey’s Pour une théorie de la production littéraire (1966). As Althusser announced in the opening of Lire “Le Capital”, it was now necessary to develop what Spinoza had posed as a problem: the meaning of the simple and (too) obvious acts “of reading, and in consequence, of writing”.

• 4 M. Pêcheux, "Remontons de Foucault à Spinoza", in L’inquiétude du discours, textes édités par Denis (...)

5Spinoza’s importance to the development of a the theory of discourse or discourses was underscored when Pêcheux, a decade later, in 1977, sought to differentiate his notion of discourse from that developed in Foucault’s Archaeology of Knowledge. He did so by using chapters 7-15 of Spinoza’s Tractatus Theologico-Politicus as a contrast medium. “Remontons de Foucault à Spinoza”, the short text of a presentation Pêcheux gave in Mexico in 1977 and which was only published posthumously, is less significant for its critique of Foucault’s notion of discourse than for its concise, too concise, excavation of a theory of discourse in Spinoza’s discussion of scripture and the discursive forms of classical Hebrew4. Pêcheux’s use of the verb “remonter” is itself notable in that it suggests both a return or a “going back” and, simultaneously, an ascent, and thus a “going back up” from a lower elevation to a higher. Pêcheux has thus offered a slightly different version of the famous theoretical pilgrimages of the time: Lacan’s return to Freud and, of course, Althusser’s return to Marx. To return to Spinoza, then, is to go back up to the vantage point he occupies and to see what can only be seen from there. To continue Pêcheux’s journey, however, may require a slight detour through the Ethics, where the questions of language and perhaps discourse do not appear to occupy the central place they do in the TTP. I refer to one of its most provocative and productive moments, less obvious in its treatment of discourse, undoubtedly because it concerns the very production of obviousness, a moment at which Spinoza explores the act of speaking by proposing what I will call a theory of the decree (decretum) that compels us to speak certain words arranged in certain phrases, and allows us to speak others. More importantly, the decree concerning what we cannot say is determined by a forgetting (of words and phrases) that is itself forgotten and lived by the individual as the freedom to decree for himself what he does and does not say. The passage in question, the scholium to EIII, P2, never ceased to inspire and fascinate Althusser. The essentials of his theory of ideology were derived from it and so too, in certain ways, were the essentials of attempts by Althusser, Pêcheux and others in 1966-1967 to develop a theory of discourse.

• 5 Cicero, De Finibus Bonorum et Malorum, II.96.

7In his attack on the doctrines of the Epicureans in De finibus II, above all, on the notion that good and evil are nothing more than names of what brings us corporeal pleasure or, in opposition, corporeal pain, Cicero cites Epicurus’s final letter, supposedly written on his deathbed to his disciple Hermarchus. In it, Epicurus not only declares that he is near death, but that he is “suffering from diseases of the bladder and the intestines which are of the utmost severity”5. If indeed bodily pain is the summum malum, says Cicero, then Epicurus’s affliction must be complete and unrelieved. But Epicurus denies that this is the case: his pain, he tells us, is in fact “counterbalanced (compensabatur) by the joy (laetitia) which I derive from remembering my theories and discoveries”. But the compensation for the pain and suffering Epicurus describes cannot be “in kind”, according to Cicero, in that the theories and discoveries he remembers have nothing to do with the pleasures of the body. It appears then that on his deathbed Epicurus both admits the existence of a pleasure, and thus a good, separate from and independent of the body and, simultaneously, denies that bodily pain is the greatest evil. Accordingly, Cicero concludes, Epicurus departs simultaneously from life, the life of the body, and from his own doctrine of bodily pleasure as the summum bonum.

• 6 Cicero, De Fin., II.104.

8While Cicero drops the account of Epicurus’s death at this point in his narrative, he will abruptly return to it a few pages later6, as if he had momentarily forgotten his theme and suddenly remembers something that had eluded him in his attempt to reveal the telling discrepancy between Epicurus’s theory and his actual practice. Cicero recalls what the letter to Hermarchus says, or at least implies, about memory. Has not Epicurus simply reaffirmed the old adage that “the Wise Man will not let past blessings fade from memory, and it is a duty to forget past misfortunes (bona praeterita non effluere sapienti, mala meminisse non oportere)?” If so, if according to the testimony contained in his final letter, Epicurus was able to counteract or even diminish present bodily pain of the greatest magnitude through the act of will by which he is able to recall his theories and discoveries, their truth above all, but also the acclaim they produced as their effect, then he has granted the human individual the power, perhaps the absolute power, absolutum imperium, to determine what he remembers and what he forgets. But, Cicero asks, is it indeed “in our power to determine what we remember (in nostrane potestate est, quid meminerimus)”? Did not Themistocles when given the opportunity to learn the art of memory declare that he would “prefer forgetting, for I remember what I don’t want to remember and cannot forget what I want to forget (Nam memini etiam quae nolo, oblivisci non possum quae volo)”? It is not possible here to explore the connections between this section of De finibus and Cicero’s well known advocacy of the ars memoriae in De oratoria; suffice it to say that his postulation of an art of forgetting in the former text serves to remind us that the challenge of an art of memory is not simply that of retrieving what has disappeared into oblivion (or of preventing it from disappearing) but of establishing order in the face of a surfeit of singular memories, as if the voluntary forgetting of some might be necessary to the rational concatenation of others. A problem emerges here that neither Cicero nor most of his early modern readers would or could acknowledge.

• 8 J. Lacan, The Four Fundamental Concepts of Psychoanalysis, New York, Norton, 1978, p. 235
• 9 J. Locke, An Essay Concerning Human Understanding, II.1.2. See E. Balibar, Identity and Difference  (...)

15What Althusser calls external verbal discourse, however, is followed in his list by what he calls “‘internal’ [in quotation marks] verbal discourse (consciousness)”. His use of punctuation, by now familiar to his readers, serves as a kind of shorthand. The term “internal”, unlike “external” in the previous phrase, which Althusser does not put in quotation marks, is thus not to be taken literally, in its spatial sense, as if the human world were divisible into inner and outer, subjective and objective. Althusser’s punctuation merely emphasizes what the repetition of “verbal discourse” already suggests: that what is taken for the inner is in fact a continuation of the outside, as if he had borrowed Lacan’s assertion that in following the silent discourse of thought, one will come back to the surface that is supposed to be its other side8. Althusser’s reference to “consciousness” (a word he will declare “disappeared” less than a page later), enclosed as it is in parentheses, compels us to translate it just as we translated “internal” into the language of pure exteriority. Henceforth, “consciousness” can be understood as the continuation of an external surface that has no inside or underside. Thus, in the idea of reflexivity contained in Locke’s notion of consciousness as a doubling that permits the “perception of what passes in a man’s own mind”9, the adjective “own” essentially marks the mind as an internal space irreducibly separate from all such other internal spaces and the perception of what passes in it, that is, consciousness, pertains to that mind, that self, and that person and no other. The notion of discourse, social and transindividual by its very nature, excludes the idea of an interior, of thought, belief and faith as the products of an autonomous mind’s “own” activity, the legal person whose separation from others serves to guarantee his freedom – the very prejudice Spinoza denounced. Discourse, then, against consciousness, against interiority, against the forms of ideality. The problem may now be posed: how are individuals constituted as subjects and what is the function of discourse in this process?

• 10 L. Althusser, "Three Notes", op. cit., p. 50.

16The fact that the theory that Althusser proposes in the “Three Notes on a Theory of the Discourses” is not a theory of discourse in the singular, but a theory of “discourses”, is significant. While he refers to four discourses, the discourse of the unconscious, ideological discourse, aesthetic discourse and scientific discourse, nothing in his text suggests that these are the only, possible or actual, discourses (in the manner of Lacan’s four discourses in L’envers de la psychanalyse). But there remains a problem : what permits us to speak of a discourse of science and a discourse of the unconscious together, as Althusser does, as if they were modes of a single substance ? The answer, of course, lies in what is common to both and, more, the very element in which they exist : language (by which I mean what is understood as the French “langue”). In fact, at a certain point in the text, when he speaks of the “structure of these different discourses”, and the “nature of the specific signifiers that make up the elements of each of these structures”, the first discourse in a list of five, rather than the four thus far mentioned, is none other than “langue” : “The signifiers of language or morphemes (matter : phonemes)”10. In one sense, the term langue generally denotes the system, the system of three systems (the phonological, the morphological and the syntactical) that furnish the foundation (or base) for the particular langues. Thus, while scientific discourse could certainly exist in the absence of aesthetic discourse and vice versa, no discourse, insofar as it is expressed in a langue, could exist except on the basis of langue (understood as the system of the systems necessarily mobilized in the enunciation of even the simplest utterance) that “generates” it. How can the origin of discourse as such be itself another discourse, one among many, without endangering the very explanatory scheme that assigns langue its priority ?

• 11 Ibid., p. 82 n.
• 12 Ibid.
• 13 F. Gadet-M. Pêcheux, La langue introuvable, Paris, Maspero, 1981.
• 14 L. Althusser, "Three Notes", op. cit., 73.
• 15 See V. Morfino, Plural Temporality: Transindividuality and the Aleatory in Spinoza and Althusser, L (...)
• 16 J.-Cl. Milner, L’amour de la langue, Paris, Éditions du Seuil, 1978.
• 17 F. Gadet-M. Pêcheux, La langue introuvable, op. cit., p. 51.

17This problem occurs to Althusser, but only as an afterthought written in the margins of the manuscript : “The function of language is not at the same level =since there is no function of language ! But only of the discourse for which it provides either constitutive (first-storey) elements (segments) of signifiers”11. Here he appears to endorse the topographical representation of langue, langues, and discourse as a hierarchy of levels (plans) such that langue furnishes the elements to, but must remain outside of, discourse in relation to which it has no (explanatory) function. This model is in certain key respects similar to that of base and superstructure, including the need (which also arises as an afterthought) to defend the “relative” autonomy of the “upper floors” of the structure. But Althusser’s dissatisfaction with such models and more importantly with the notion of emanative causality they incarnate, leads him back towards the idea of a structural causality and the concept of the immanent or absent cause he had begun to develop in Lire le Capital. His addendum concludes with a surprisingly general assertion, but one which allows us to make sense of his notion that langue is a discourse : “Thus there are no functions of language in this sense for language does not exist : only discourses exist”12. This note captures in a clear and succinct manner the movement or movements proper to Althusser’s thought, the struggle to evade capture by the dominant forces in philosophy, the sense of the ever-present traps and snares laid for the unwary, the ruses and diversions that more often than not are successful, but which his heightened awareness and acceptance of the need to think through the struggle allow him, in certain spectacular instances, to evade. And in this war, a war of resistance, survival, the survival of a line of thought, is victory enough. The conclusion of the addendum not only does not follow from the topographical explanation that precedes it, but nullifies that explanation. There can be no topography in which langue would occupy the ground floor, because “langue does not exist”. Further, because it does not exist, the idea of discourse in general, of a discourse of discourses, disappears. If langue, however, is not present “in person”, it is because it has disappeared into its effects, dispersed not in discourse, but in the irreducible plurality of discourses. It has become, as Pêcheux and Gadet put it more than a decade later, “la langue introuvable”13. Linguistics, whose object is precisely langue in the singular and in person, is, according to Althusser, “incapable by itself of producing a theory of the different discourses, and this inability is masked by the claim that it can produce this theory on the pretext that it can furnish a theory of discourse – but no theory of discourse can stand in for, replace or deduce from itself a theory of discourses”14. Moreover, the inability to think not the unity of discourse but its diversity is tied to the existence of langue as a discourse. Langue may be thought of as a discourse in a double sense : because it is the cause that cannot exist outside of its effects, here, discourses, but also because langue, scattered among the discourses, themselves embedded in the material existence of ideologies which are subject to the aleatory and plural temporality of history,15 can never be simply a rule-governed system whose expansion follows a juridical model of consistency based on a sequence of precedents (even if it is grounded in neurobiology). As Pêcheux and Gadet would argue later, in a sense completing or at least extending Althusser’s analyses, whether this system is understood as an “order proper to langue, immanent in the structure of its effects” (the tradition that extends from Port Royal to Hjemslev and Chomsky), or as an order deriving from the accretion of decisions and therefore imposed from the outside (sociolinguistics), an artificial construction that serves the broader political order, that is, whether the foundation of the model is rational or empirical, it can exist only through a systematic repression of what Gadet and Pêcheux will later call, following Jean-Claude Milner16, the “real of langue” : the fissures (failles), gaps, and contradictions that set this order against itself in a perpetual production of equivocity. There exists “in every langue a segment” that can be “both itself and at the same time other through the homophony, homosemy, metaphor, glissement of the lapsus and word play, and the double meanings (double entente) of its discursive effects”17. Thus, the process of the production of discursive effects is simultaneously and necessarily a production of side-effects. It is this fact that makes the historical existence of discourses something other than the actualization of already existing logical possibilities, that is, to use the familiar phrase, a process without a subject or end(s).

• 18 "Three Notes", op. cit., p. 77.

19The question of the subject is at the center of the “Three Notes”, a center not only in conflict with itself but, in its very division, moving in different directions. Althusser begins by defining discourse in part as the arrangement of signifiers according to the rules or laws proper to each of the four discourses (unconscious, aesthetic, scientific and ideological). But what really marks a discourse as discourse and allows Althusser to produce a theory of their difference is that each in its own manner produces the subject proper to it. The subject proper to the discourse of science is an absent subject, a subject that has always just disappeared. Indeed, the absence of the subject is what alone distinguishes it from ideology, into which it would lapse if the subject were present on the scene. The subject of aesthetic discourse always appears in the form of “interposed persons” (always in the plural). The discourse of the unconscious differs from the others in that the signifier that functions as a subject is always a “stand-in” or “lieu-tenant” for a subject never present in person but represented by its emissaries. Only ideological discourse produces a subject form that requires its individual Träger to be present in person. By the end of the “Three Notes”, Althusser will reject the account of the subject proper to each of the four discourses, to argue that it exists only in ideological discourse: since writing the first note, he tells us, “I have come round to thinking that the notion of subject cannot be employed unequivocally not even as an index of each of the discourses. Increasingly, the notion of subject seems to me to pertain to ideological discourse alone, of which it is constitutive. I don’t believe that one can talk about a “subject of science” or a “subject of the unconscious” without playing on words and opening the door to serious theoretical ambiguities”18.

• 19 Cf. Fr. Torterat-A. Thibault, "L’Interpellation : un objet discursif singulier... et pluriel", in C (...)

20The effect of this conclusion is to allow Althusser to comprehend the subject form of ideology not as the pre-existing position proper to each discourse, open to be filled, as in the quasi-formalist scheme of the four discourses he had earlier sketched out, but as a process of subjection/subjectivation and therefore a site of struggle and contestation. Deviating from his own version of the linguistic model, Althusser proposes a new concept : interpellation. “Ideology interpellates the individual by constituting him as a subject (an ideological subject and therefore the subject of its discourse)”. There remains much to say about the verb “interpellate”, its meaning and function both in general and within Althusser’s text. Here, I will only refer to its most immediately relevant meanings and associations. In Latin and in French legal and political language, “interpellate” denotes an act (typically an interruption, unexpected and usually unwelcome) by which the interpellated person is separated from a group and “called upon to answer” for himself and his actions. Interpellation is also understood today as a version of the rhetorical device of apostrophe, in which a speaker directly addresses an absent person or personification as if it were present (as in “my Lord, my Lord, why have you forsaken me ?”).19 In the “Three Notes”, ideology, or more properly ideological discourse, separates every individual from every other individual and in one and the same movement calls upon the individual to answer for himself, furnishing in the process the “raisons” or arguments which, once articulated, confirm the individual’s status as subject and agent. The subject thus called upon is strictly speaking absent until the call renders him present. In this sense, ideological discourse interpellates the individual by “imputing” agency to him and designating him as the uncaused cause of his speech and actions for which he thus bears sole responsibility: these are precisely the defining characteristics of the subject. But this discourse simultaneously supplies him with the “raisons-de-sujet”, the signifiers and utterances whose very materiality allows them produce the effect of subjectivity, the folding of discourse back upon itself to produce an “inside” that is in fact a continuation/extension of the outside.

• 20 Cf. W. Montag, Althusser and His Contemporaries, op. cit., p. 131-140.

21But we cannot fail to note the resistances that prevent Althusser from confronting and overcoming the hesitations and inconsistencies in his argument and therefore from developing his theory of discourses. I have discussed elsewhere the moment, near the end of his text, at which Althusser, as if attempt to negate or diminish what he has written, retreats from the notion of the materiality of discourses by declaring that discourses are not practices.20 Practices produce effects that transform the real, while discourses produce such effects only insofar as these effects pass through a practice or practices, which thus mediate between discourse and reality, the sole means of contact between them. But this is not the only point at which Althusser draws back from the very conclusions to which his arguments lead.

• 21 "Three Notes", op. cit., p. 52.
• 22 Ibid., p. 83.

22Ideology doesn’t simply interpellate individuals as subjects, authors of, and responsible for, their works. It also, he tells us, “recruits” and “requisitions” them as Träger, and thus as supports for the economic base of every social formation. Althusser’s repeated use of these terms as synonyms highlights the contradiction at work here. Both are military terms (although “requisition” has a more limited lexical range than “recruit”) denoting the act by which the military replenishes its ranks or increases the number of soldiers. When Althusser argues that the economic base of every social formation “requires” individuals to fill the positions necessary to its functioning, he uses the related term “requisition”. The verb “requisition” in French as in English signifies a command that as such neither seeks the consent of the individual whose property or person is appropriated nor tolerates any refusal : all are required to heed the order when they are “called up”. While the means of appropriation may vary, the idea of requisitioning assumes the resistance of the proprietor to the loss of property and thus carries with it the ever-present threat of force. In contrast, the ideology which arises from this base precisely in order to assure its persistence (Althusser writes here in a functionalist “shorthand” that too easily replaces the theory for which it was meant to serve as a temporary stand-in) “recruits” the individuals with which it will fulfill the requisition order that emanates from the base. “Recruit” typically presupposes the consent of the individual who is convinced to “enlist”, “sign up for” or join the army, a company, an organization, etc. Indeed, he tells us, perhaps in response to Pêcheux’s idea of the social command, which we will examine in a moment, “ideology is not a command/commandment” nor an “injunction pure and simple” realized in “pure force (there is no such thing as pure force)”21. If ideology interpellates individuals by constituting them as subjects to meet the demand of the economic base, the subjects it constitutes are provided with a guarantee that they are indeed subjects (agents and authors separated from every other agent and author), proprietors of their speech and action and thus responsible for their consequences: their identity. They are thus addressed as identical to themselves and no other, and as such become identifiable. Althusser added the following allegory in the margins of the text: when the police stop (interpellate) an individual in the street for questioning, their first order is inevitably that the individual in question produce proof of his identity, proof whose sole legitimate form is the “identification” that they themselves furnished him with.22

• 23 Ibid., p. 56.

23But this guarantee has an internal as well as external function : by verifying the identity of the interpellated subject, it both requires and permits the individual subject to “recognize” himself as subject, that is, not simply to be, act and feel like a subject but to know that he does so and to know that he knows ; this is precisely Locke’s definition of consciousness. The very act of recognizing (re-cognizing, re-knowing, remembering) oneself requires more than simply an external guarantee of identity : Althusser asserts that the subject must duplicate itself in the form of an omnipotent and omniscient other who knows what he cannot know about himself, including all that he has forgotten, and to whom he is accountable. It is to this internal tribunal that ideology must present its case : in order to win, ideology must use “persuasion” to “convince/convict” the subject to comply with the requirements of the economic base. Althusser has thus forgotten the task of conceptualizing the materiality of the discourses which alone will allow him to escape the trap of subjective interiority. If this were the conclusion of his inquiry, Althusser would have paradoxically produced a theory of the interpellated subject whose interpellation confers upon him the freedom to refuse the demand of the economic base – if he does not find its arguments “persuasive” – and therefore the power to decree whether or not he (his body) will be at the disposal of the economic base that requires it. Fortunately this is not the case: Althusser is drawn out of this impasse by the power of the concept of the unconscious as Lacan has explained it, above all in the Four Fundamental Concepts. It is the notion of the discourse of the unconscious, not the subject of the unconscious, but precisely the gap or hole that interpellation “induces” (as if it induced amnesia) in the very subject it calls into being. We should not forget that “imputation”, the legal attribution of free will and personality to a body whose words and actions remain so opaque that the very existence of legal responsibility depends on individuals being treated “as if” they were free and therefore responsible for “their own” actions, possesses a double materiality: the stubborn resistance of the body to interpretation on the one hand and on the other the rather cynical materiality of a discourse without a referent, a discourse that does not represent the internal state of the accused individual but compensates for its irreducible lack by supplying in prosthetic form the attribution that the law demands. The faculty of consciousness, the perception of perception and the thought about thought, the guarantee that what we think is what we really think, emerged out of the concept of “conscience” which already implied not simply a moral faculty, but an ability unique to the individual to “supervise” (super-videre or oversee) not merely actions, but thoughts and passions. It is possible to salvage from Althusser’s initial but soon rejected discussion of the subject forms proper to each of the four discourses, the idea that interpellation as he understands it is the combined effect of the discourse of ideology and the discourse of the unconscious: “the interpellation of human individuals as ideological subjects produces in them a specific effect, the unconscious effect that permits human individuals to assume the function of ideological subjects”23. If we suppress the functionalist afterthought here (which Althusser will deny is functionalist a page later) and refrain from assigning the unconscious a role in the production of the imputed/interpellated subject, we might say that the unconscious is a necessary effect of the discursivity realized in the act of interpellation, of the discursive materiality that is only one of the modes of matter mobilized in this complex operation. Far from arising to fulfill a social need, the unconscious induced by interpellation is the reproduction of the impossible proper to “the subject of discourse”, the impossibility of its coinciding with itself, of its being present to itself to be known by an act of con-scientia, the thinking about what it thinks and feels. Thus, interpellation postulates a subject whose ability to appropriate himself is guaranteed by the accompanying postulation of an absolute Subject, the Subject of all knowledge who, because he forgets nothing, demands an accounting for everything. At the same time and in the same gesture, the discursive acts by which it is accomplished produce as a side-effect an “abîme” or hole where the conscious subject should discover itself but cannot, as if interpellation adds subjectivity to the body but only at the cost of a subtraction that deprives this subjectivity of what is its own.

• 24 Th. Herbert, "Réflexions", op. cit., p. 145.
• 25 Idem.

24Pêcheux’s essay, which appeared six months before Althusser began to write the “Three Notes”, perhaps, as I have suggested, in part as a response to Pêcheux’s text, suggests a way out of some of the impasses that Althusser encounters in the “Three Notes”. He does so not by abandoning the concept of discourse, as Althusser does in the ISAs essay, to avoid the problems it engendered in the “Three Notes”, but precisely by establishing the materiality proper to it, a materiality that the very concept of language (in the broad sense) works to deny. Pêcheux’s starting point is identical to Althusser’s : the reproduction of a social formation (although he will insist, almost from the moment the ISAs essay appeared in La Pensée, on the formulation “reproduction/transformation” as a preventative measure against functionalism). Unlike Althusser, however, even in this early text, Pêcheux is not only unwilling to use functionalist arguments as a kind of shorthand or place-holder faute de mieux, but refuses to rank functionalism as a secondary enemy with whom one can safely conclude tactical alliances. His sense of the threat of functionalism, especially in the theoretical conjuncture of the pre-1968 period, not only leads him to rectify some of Althusser’s formulations, but also pushes him to theorize the disorder specific to langue and to discourse (in the face of the nearly irresistible force of the formalism of emerging linguistic theories) from the earliest moments of his career. Ideology, he tells us early on in the essay, is not produced by the final cause of social reproduction, but first appears as a “sous-produit” (or by-product) of what he calls “technical practice”, a practice that applies “instruments and the forms of human labor implied by them” on a given “raw material” to obtain certain products.24 This practice exhibits what Pêcheux calls “an external teleological structure” : it seeks “to fulfill a need, a lack, a demand” that comes from without. The “law of the technical response to a social demand is constitutive of technical practice”25. But we should not be misled by the term “constitutive”; because technical ideology “belongs to the same process” as technical practice and is therefore coextensive with it, the very means by which technical practice “responds” to social demand opens the ever-present possibility that the by-product may overpower the product, the side-effect neutralize the intended effect, and a given response to social demand may represent a kind of interference that represses or displaces another response.

• 26 "Transformer les rapports sociaux en re-formulant la demande sociale (demande et aussi commande, au (...)
• 27 Ibid., p. 150.

25Even when Pêcheux moves to political practice and its relation to a given mode of production, he carefully avoids the notion that an economic base produces the means of its own reproduction, as if it were a deliberating (collective) subject acting with an end in view. Instead, he refers to the “immanent cause”26 that makes intelligible the conflictual or even ruptural unity of unequal forces that constitutes every mode of production and submits it not to the law of an internal teleology, but to an aleatory movement of encounter and struggle. Political practice works on “social relations”, but does so in response to the “social demand” that “emanates” from them. “Demand”, however much the social sciences work to deny it, is first of all a word and as such belongs to a discursive field that determines its possible synonyms, and thus its meaning. Even in this early text, Pêcheux treats meaning as a matter of (re)formulation and paraphrase based on a finite chain of possible substitutions. Here, he will propose a synonym for demand that, running the very real risk of remaining illegible, plays on the visibility of a shared root (the Latin mando) to expose the ideological determination of what can and cannot be substituted for “demand”. Thus, political practice seeks to “transform social relations by reformulating social demand (demand and also command in the double sense in which it is to be understood form, this point on) by means of a discourse. In saying this, we do not claim that politics is reducible to discourse, but that every decision, every « measure » in the political sense takes its place in political practice like a sentence in a discourse”27. This is a dense and elliptical passage that suggests a number of lines of inquiry, some of which directly link Pêcheux’s analysis to Spinoza’s use of the term “decree” in the scholium to EIII, P2, and that proposes ideas and concepts whose development Pêcheux defers to other texts, above all, Les vérités de La Palice. His critical intervention here, which constitutes a turning point in the discussion of discourse, is his insistence that every political decision, every measure taken, including the extreme measures that war and revolution require, that is, even the use of force itself, must always also, in order to be effective, take the form of a sentence (une phrase) linked to other sentences in a discourse. Every political “decision” (Spinoza’s decretum) exists in a material form that is simultaneously physical and discursive, without the one modality being reducible to or caused by the other: both are determined in their consubstantial unity by an immanent causality.

• 28 M. Pêcheux-C. Fuchs, "Mises au point et perspectives à propos de l’analyse automatique du discours" (...)

27We may now see why Pêcheux would link demand in this sense to command : not only do both terms have their origins in the Latin verb, mando (meaning to order someone to do something), but their common historical root is visible in the graphic materiality of the words themselves. This fact is a symptom, a trace of the work of “reformulation/paraphrase” that has enforced a separation between these terms that is critical, perhaps necessary, to the existence of what we now call neo-liberalism which must deny its own existence as a “command economy”, a phrase used exclusively to denote those other, “unfree” social and economic systems. To paraphrase social demand as social command (perhaps at the risk of intelligibility) is to show that meanings are imposed by a necessity that is simultaneously internal and external to discourse. Typically, as Pêcheux explained in a later text, the utterances produced by the speaking subject are “displacements within the interior of the formulable” determined by a given discursive formation itself determined by an outside that, because it determines it, “remains strictly unformulable”28. The unformulable outside is nevertheless present in discourse as the barrier that limits what can be said and thought and by that fact renders itself unthinkable. By what he does as much as by what he says, Pêcheux shows that if it is possible to speak of illusion here, it is not located in consciousness or subjectivity, prior to the use of language, but in discourse itself, in the relations between its elements, words and phrases, particularly the relations of substitutability which are normally imposed on individuals without their knowledge or consent.

• 29 M. Pêcheux, Les verités de La Palice, Paris, Maspero, 1975, p. 112.

28How then can a “reformulation/paraphrase” escape the meanings imposed by the discursive formation in which a word or utterance is produced ? The answer, of course, lies in the fact that discursive formations are themselves sites of struggle, struggles over meaning, as well as against subjection. Discourse, as fortune as theorized by Machiavelli, brings the good with the bad and in its variability presents certain openings for an intervention that may overturn an existing discursive regime. If “the meaning of a word, proposition, does not exist in itself” but is constituted by discourse, that is, what Pêcheux would later call a discursive formation, through “the relations of substitution, paraphrase, synonymies, etc., which are operative between linguistic elements”29, it is precisely in this dislocation that the ideological outside is present within discourse, the effect of the equilibrium of forces in the great struggles that both envelope and are enveloped by discourse. In this way, discursive relations and discursive processes are subject to alternating periods of stability in which relations between elements are held in place and to periods of instability which permit new formulations and thus new meanings to emerge. Even in the latter case, however, political practice may reformulate “social demand” to give it a simultaneously discursive and ideological existence in a way that is clearly apologetic: social demand is not simply a demand for what is socially “necessary” (the meaning of which is one of the most important stakes of class struggle), but a demand for what should be, for what is both rational and right. But Pêcheux’s own analysis is itself a “reformulating” of (social) demand made possible by an opening that not only allowed him to give it a new meaning but in doing so make legible and audible the command whose mandate demand executes.

• 30 Ibid., p. 137-138.
• 31 Ibid., p. 105.

31Pêcheux described his work as defined by the “linking of the question of the constitution of meaning to that of the subject”, a linkage “located inside the central thesis itself in the figure of interpellation”30. Citing Althusser, Pêcheux notes that both the question of meaning and the question of the subject are constructed as “obvious” (évidentes), that is, immediately visible and standing before us, “in the way” (Latin: ob-via), unavoidable: 1) every word has a meaning, a notion based on the equally obvious fact of the “transparency of language”31 and 2) all of us, you who are like me, are free, ethical subjects. Thus, it is necessary to account for the effect of the obviousness of the obvious, but also its causes, the means by which is it produced around certain words and phrases and held in place or reiterated. We see this not only in the constraints of synonymy but also in constructions such as the appositive relative clause in which, especially in political, social and economic discourse, the effects of the command are more visible and hence more exposed to contamination by irony: “the United States, the most peace-loving nation on earth, sent more troops to Iraq yesterday” or “X, a well-known terrorist organization, has been operating with impunity and must be taken out”. For Pêcheux, the obvious is a consequence of what may be his own “central thesis” concerning the primacy of forgetting. Words and phrases are “always already” endowed with a meaning or correlated with a thing, just as the individual is “always already” a subject separate from other subjects and uniquely responsible for his own free choices: in the realm of the “always already”, causal processes are obscured by the effect of obviousness that they themselves produce. Pêcheux proposes a visible image of this theoretical fantasy: Escher’s “Drawing Hands”, in which two disembodied hands are captured in the process of drawing each other into existence.

• 32 Ibid., p. 114 n.

32Forgetting in this sense, however, must be understood as entirely non-subjective and thus as inherent in discourse itself. It “does not mean the loss of something once known, as when one speaks of a loss of memory, but the occlusion of the cause of the subject inside its very effect”32. Forgetting in this sense is not a loss at all, certainly not a negation of memory; it is instead a movement, perhaps a pulsation, entirely within discourse, of synonymy and substitution, of what can and cannot be said and thought, including and above all about oneself as a thinking thing. It is thus in the command, the decree that the individual as subject (and the history of the meaning of the term “subject” itself, the reversal of its meaning, the shift from denoting one who is subjected to the will of another to one who is the agent and author of an action, is surely relevant here) emerges as one whose autonomy and freedom are the effects of the decree that they will freely decree their speech and action for which they are causally and morally/legally responsible. It is in this movement, as Spinoza had already noted, that both discursive memory and discursive forgetting, and thus discursive materiality, exist. Such processes are neither functions of a formal system or systems, nor are they grounded in the psychology of a subject who “uses” language, well or badly. To understand the work of forgetting, Pêcheux refers directly to Spinoza (among others) and the argument takes up where the scholium to EIII, P2 leaves off.