History of Structuralism

Volume 2: The Sign Sets, 1967-Present

Francois Dosse

Translated by Deborah Glassman

University of Minnesota Press


Structuralism is nota new method, it is the awakened and troubled consciousness of modern thought.

Michel Foucault


Preface xiii

Acknowledgments xv

Part I. First Fissures

1. Chomskyism-New Frontiers? 3

2. Derrida or Ultrastructuralisrn 17

3. Derridean Historicization and Its Erasure 3Z

4. Benveniste: The French Exception 42

5. Kristeva and Barthes Reborn 54

6. Durkheim Gets a Second Wind: Pierre Bourdieu 66

7. 1967- 1968: Editorial Effervescence 76

8. Structuralism and/or Marxism 88

9. Media Success: A Criticism-fed Flame 99

Part II. May I968 and Structuralism; or, The Misunderstanding

10. Nanterre-Madness 107

11. jean-Paul Sartre's Revenge 112

12. Lacan: Structures Have Taken to the Streets! 122

13. Institutional Victory: The University Conquered

14. Vincennes: The Structuralist University

15. Journals: Still Going Strong 154

16. The Althusserian Grid: A Must

The Althusserian Grid: A Bust 179

Part III. Structuralism between Scientism, Ethics, and History

18. The Mirage of Formalization

19. From Explosive Literary Mourning to the Pleasure of the Text 200

20. Philosophy and Structure: The Figure of the Other

21. The Reconciliation of History and Structure 227

22. Foucault and the Deconstruction of History :

The Archaeology ofKnowledge 234

23. Foucault and the Deconstruction of History (Il):

Discipline and Punish 247

24. The Golden Age of New History 260

Part IV. The Decline

25. Lost Illusions (1): The Gulag Effect

26. Lost Illusions (Il): Extenuated Scientism

27. Lost Illusions (III): The Return of Ethics

28. From Reproduction to Regulation:

Heirs to Keynes and Althusser, and the Crisis 288

29. A Middle Path: The Habitus 3°1

30. Geography: A Latecomer Discovers Epistemology 312

31. The Subject; or, The Return of the Repressed

32. Michel Foucault: From Biopower to an Aesthetics of the Self 336

33. An Autonomous Subject 35°

34. History Returns 364

35. The Master Thinkers Die

36. The Crisis of Universalist Models and

Disciplinary Retrenchment 39 1

37. Structural Naturalism 397

38. Assimilating the Program 40 8

Part V. Time, Space, the Dialogic

39. Clio in Exile

40. ATopo-Logic 437

41. For a Dialogic 445 Appendix: List of Interviewees 453 Notes 459 Index


Were there many structuralisms or simply one structuralism? By the end of the decades of structuralism's triumph described in the first volume of History of Structuralism, it had become clear that structuralism wove a reality of different logics and individuals resembling a disparate fabric more than a school. However, there were a specific orientation and many dialogues indicating a "structuralist moment." In the mid-sixties, both Louis Althusser and Michel Foucault were trying to bring together the most modern social science research around an effort at philosophical renewal that came to be known as structuralist. In 1966, these efforts reached their apex.

By 1967, cracks started to appear. It became clear that the regroupings of the first period were often artificial, and a general withdrawal of sorts began at this point. Certain of the players sought less-trodden paths in order to avoid the epithet "structuralist." Some even went so far as to deny ever having been a structuralist, with the exception of Claude Levi-Strauss, who pursued his work beyond the pale of the day's fashions.

Paradoxically, while structuralists were distancing themselves from what they considered to be an artificial unity, the media were discovering and aggrandizing this unity. This period of deconstruc-tion, dispersion, and ebb, however, only quite superficially affected the rhythm of structuralist research. Research continued elsewhere, in the university, and obeyed another temporal logic. May 1968 had contributed to structuralism's institutional success, and this played an essential role in assimilating the program that had lost its blazened banner of a counterculture in revolt to become one of the theoretical, but unarticulated, horizons of social science research.


I would like to thank all those who were kind enough to agree to be interviewed. These interviews were entirely transcribed and their contribution was absolutely fundamental to this project of writing this history of French intellectual life. The specifics of the area and current affiliation of each of the interviewees are to be found in the appendix. Those interviewed were Marc Abeles, Alfred Adler, Michel Agli-etta, Jean Allouch, Pierre Ansart, Michel Arrive, Marc Auge, Sylvain Auroux, Kostas Axelos, Georges Balandier, Etienne Balibar, Henri Bartoli, Michel Beaud, Daniel Becquemont, Jean-Marie Benoist, Alain Boissinot, Raymond Boudon, Jacques Bouveresse, Claude Brernond, Hubert Brochier, Louis-jean Calvet, Jean-Claude Chevalier, Jean Clavreul, Claude Conte, Jean-Claude Coquet, Maria Daraki, Jean-Toussaint Desanti, Philippe Descola, Vincent Descombes, jean-Marie Dolmenach, joel Dor, Daniel Dory, Roger-Pol Droit, Jean Dubois, Georges Duby, Oswald Ducrot, Claude Dumezil, Jean Duvignaud, Roger Establet, Francois Ewald, Arlette Farge, jean-Pierre Faye, Pierre Fougeyrollas, Francoise Gadet, Marcel Gauchet, Gerard Genette, jean-Christophe Goddard, Maurice Godelier, Gilles Gaston-Granger, Wladimir Granoff, Andre Green, Algirdas Julien Greimas, Marc Guillaume, Claude Hagege, Philippe Hamon, Andre-Georges Haudri-court, Louis Hay, Paul Henry, Francoise Heritier-Auge, jacques Hoarau, Michel Izard, jean-Luc jamard, Jean Jamin, julia Kristeva, Bernard Laks, Jérôme Lallement, Jean Laplanche, Francine Le Bret,

Serge Leclaire, Dominique Lecourt, Henri Lefebvre, Pierre Legendre, Gennie Lemoine, Claude Lévi-Strauss, jacques Lévy, Alain Lipietz, Rene Lourau, Pierre Macherey, Rene Major, Serge Martin, Andre Martinet, Claude Meillassoux, Charles Melman, Cerard Mendel, Henri Mitterand, juan-David Nasio, Andre Nicolai, Pierre Nora, Claudine Normand, Bertrand Ogilvie, Michelle Perrot, Marcelin Pleynet, Jean Pouillon, joelle Proust, ]acques Ranciere, Alain Renaut, Olivier Revault d' Allonnes, Elisabeth Roudinesco, Nicolas Ruwet, Moustafa Safouan, Georges-Elia Sarfati, Bernard Sichere, Dan Sperber, Joseph Sumpf, Emmanuel Terray, Tzvetan Todorov, Alain Touraine, Paul Valadier, jean-Pierre Vernant, Marc Vernet, Serge Viderman, Pierre Vilar, Francois Wall, and Marina Yaguello.

Others were contacted but were not interviewed: Didier Anzieu, Alain Badiou, Christian Baudelot, Jean Baudrillard, Pierre Bourdieu, Georges Canguilhem, Cornelius Castoriadis, Helene Cixous, Serge Cotte, Antoine Culioli, Gilles Deleuze, ]acques Derrida, Louis Du-mont, julien Freund, Luce lrigaray, Francis Jacques, Christianjambet, Catherine Kerbrat-Oreccioni, Victor Karady, Serge-Christophe Kolm, Claude Lefort, Philippe Lejeune, Emmanuel Levinas, jean-Francois Lyotard, Gerard Miller, jacques-Alain Miller, Jean-Claude Milner, Edgar Morin, Therese Parisot, jean-Claude Passeron, ]ean-Bertrand Pontalis, Paul Ricoeur, Jacqueline de Romilly, Francois Roustang, Michel Serres, Louis-Vincent Thomas.

I would also like to thank all of those whose difficult task it was to read this manuscript in its early stages and whose comments and suggestions made it possible for me to carry out this undertaking: Daniel and Trudi Becquemont, Alain Boissinot, Rene Gelly, Francois Cese, Thierry Paquot, and Pierre Vidal-Naquet.

Finally, for having given me the print runs of a certain number of works of the period, I would like to thank Monique Lulin at Editions du Seuil, Pierre Nora at Editions Gallimard, and Christine Silva at Editions La Decouverte,

Part I

First Fissures


Chomskyism-New Frontiers?

In 1967, Pion published Nicolas Ruwet's doctoral dissertation, Introduction to Generative Grammar.1 in which he presented Chomskyan principles. For Ruwet, as for many linguists, Chomsky was the expression of a radical break with the first structuralist period. Ruwet had discovered Chornsky thanks to an itinerary that took him from Belgium to Paris, where he attended many of the important seminars being held at the time.

Born in 1933, Nicolas Ruwet was first a student in Liege. Dissatisfied with a style of teaching that resembled the pedagogy at the 50r-bonne, he left Belgium in 1959 to come to Paris. "I was vaguely thinking about ethnology, but I was also interested in psychoanalysis. I was a musician at the beginning and I had already read a certain number of works in linguistics including Saussure, Trubetzkoy, and jakobson."? From the outset Ruwet was at the confluence of different disciplines, a good indication of the totalizing structuralist imperative. He left Belgium seeking rigor and in the hope of participating in the scientific adventure that was unfolding.

In Paris, Ruwet went to Emile Benveniste's seminar at the College de France, Andre Martinet's seminar at the Sorbonne, and Claude Levi-Strauss's seminar at Hautes Etudes. "What was going on in Levi-Strauss's seminar particularly excited me, at the beginning when he brought in a long article by Roman jakobson that had just come out in English, entitled 'Linguistics and Poetics.' He was completely carried away by it and read us practically the entire text during the two hours of the class."! In 1962, Ruwet became a member of the poetics program in the Belgian National Foundation for Scientific Research (FNRS): "I was planning to do a thesis on Baudelaire, which I never did."4 In 1963, he wrote the preface to the collected works of jakobson, one of the major publications of the period, published by the Éditions de Minuit as Essays in General Linguistics.s He and his friend Lucien Sebag were both attending Lacan's famous seminar at the time. While on a trip together with Lacan's daughter and other friends in a house that Lacan had rented in Saint-Tropez, Ruwet discovered Chomsky, entirely by accident.

I was alone in the room that Lacan used as a study and there was a little blue book, published by Mouton, lying on his desk. It was Chomsky's Syntactic Structures. I ordered it right away at the end of the vacation, and found it very interesting, but I did not understand a thing. There were still too many pieces missing."

Despite this fortuitous discovery, Ruwet continued to work along the lines defined by Jakobson and Hjelmslev and wrote an article for Eric de Dampierre summarizing the situation in general linguistics in 1964, in which he sang the praises of structuralism.7

The Conversion

In 1964, everything changed. A friend from Liège lent Ruwet a book by Paul Postal which had just come out, Constituent Structure: A Study of Contemporary Models of Syntactic Description, in which Postal presented Chomsky's major ideas. "I read it on the train between Liège and Paris. When I got to the Gare du Nord I was a genera-tivist. In the space of a few hours I had walked my road to Damascus. Everything changed. I had to finish my article for Éric de Dampierre, but I no longer believed in what I was saying."8 Ruwet spent the next three years reading everything published on generative grammar and preparing to write his thesis-which he had initially not planned to publish as a book but only in order to get an official diploma and crown a rather interdisciplinary career, like that of most structuralists. In 1967-68, this book quickly became the breviary of the new generation that was discovering linguistics.

Chomsky was not well known in France at the time. Although Syntactic Structures had come out in 1957, it was only translated in

1969 at Seuil. Thanks to Nicolas Ruwet, who adopted an entirely new approach with respect to what had come before in linguistics, Chomsky became known in France. In December 1966, Ruwet introduced generative grammar in issue 4 of Langages; Chomsky gave him the possibility of working on syntax, which Saussure and Jakobson had ignored. While the search for greater scientificity provided the link between structuralism and Chomskyism, Ruwet saw an advantage in generativism's Popperian conception of science as falsifiable. "The break lies in the possibility of offering hypotheses that can be proved false. This made a deep impression on me."? Generative grammar required a precise and explicit theory, which worked like an algorithm whose operations can be applied mechanically. "Karl Popper clearly showed that it was possible to establish a science on the principle of induction."lO With the double articulation of language on the deep structure of competence and a surface structure of speech, a double universality was postulated. Not only were there established rules and a system, but there were also "a certain number of substantial univer-sals. "11 This quest for universals carried structuralist ambitions even further, ambitions themselves taken from the general principle evoked by Plato in the Sophist (262 B.C.), offering "the material foundations of structural linguistics." 12 Plato had argued that the study of a system of signs presupposed a certain limited number of conditions: determining minimal units, their finite number, their combinability and, finally, that not all combinations are possible.

Although May 1968 would also weaken the structuralist paradigm, as we will see, Chomsky's thinking was in phase with the events of the late sixties. But this was due to a curious misunderstanding. In the first place, Chomsky was known as an American radical who protested the war in Vietnam, thereby embodying the very expression of a critical attitude. But even more, the term "generative" in France was understood "in the sense of that which engendered, fruitful moves. We no longer wanted static structures, and structuralism at that point was associated with conservatism. The term 'generative,' although purely technical, had nothing at all to do with all of that."13 For Chomsky, in contrast, generative grammar meant simply an explicit grammar modeled on the competence of native speakers and it "simply meant the explicit enumeration by means of rules."14 Thanks to these misunderstandings, generative grammar met the generation of protest, which saw in Chomsky's ideas the means of reconciling his-tory, movement, and structure. This misperception was fruitful in many ways, including making generativism known in France.

The Archaeology of Generativism

There was a second misunderstanding. Chomsky's criticism did not address European structuralism. It focused on American structuralism, represented by Leonard Bloomfield and his "distributionist" or Yale School, the dominant form of linguistics in the United States in the fifties. Bloomfield drew his inspiration from behavioral psychology, and considered that it was enough to describe the mechanism of language, to underscore its regularities. These mechanisms were the concern, the meaning of utterances was not. Utterances were to be broken down into their immediate constituents and classified in a distributional order. American linguistics prior to Chomsky was thus essentially descriptive, linear, and based on an assumed transparency between speech acts and their meaning. The systems of opposition emphasized by American structuralism made it above all possible to avoid mentalism. This descriptive, distributional approach was largely inspired by work done in the twenties and that sought to restore the various Amerindian languages. Ethnolinguistics, which Boas and Sapir had been developing on the other side of the Atlantic, removed from Saussureanism, saw linguistics in this light. "Chomsky's rupture has to be understood with respect to the school of American linguistics. The split is clear but there is an undeniable foundation, which is articulation. No theory proposes to analyze sentence structure." 15

American structuralism, or distributionalism, also moved ahead thanks to the work of Zellig Harris, who described its method in 1951.16 Like Bloomfield, Harris argued that meaning and distribution corresponded to each other. He defined the principles of an approach based on the constitution of a representative, homogeneous corpus in order to determine the different morphemes and phonemes by means of successive segmentations. To get to these original structures, Harris defined mechanical rules of calculus and eliminated all traces of subjectivism and context. "Functional notions such as the subject of a sentence, for example, were replaced by complex classes of distribution. "17 All forms of speaker intentionality were relegated to somewhere beyond the scientific field of distributionalism. Harris therefore pushed Bloomfield's logic to its limits, and introduced the notion of transformation in order to reach the study of discursive structures using classes of equivalence. His research led him greater and greater formalism.tf in order to make different discursive manifestations derive from a limited number of elementary sentences generated by fundamental operators. "Everything in this model depends on the assimilation of meaning to objective information and on the position of a weak semantics." 19

The Principles of Generativism

Initially, Chomsky adopted Harris's distributionalism and maintained the explicit character of the approach. But, together with Morris Halle at the Massachusetts Institute of Technology, he quickly oriented his work in a new, "generative" direction. He rejected the distributional-ist imposition of limits to a corpus that did not exhaust the richness of a language. With the intention of going further than a simple description, he sought the more essential level of explanation and therefore denounced economical methods. Initially, he limited his field of study to syntax so as to establish an independent theory and an autonomous grammar. "The end results of this research should have been a theory of linguistic structure in which the descriptive mechanisms used in particular grammars were introduced and studied abstractly without any specific reference to any particular languages. "20 This grammar would take the form of a generative mechanism that revealed possibilities, rather than a corpus serving as the basis for induction.

By its formalism and rejection of meaning, the generative approach upheld the structuralist legacy. "This conception of language is extremely powerful and general. If we adopt it, we consider the speaker to be essentially a machine of the type known in mathematics as the Markov process with a finite number of states. "21 Once the technical hypotheses of the construction of this generative grammar were described in 1957, Chomsky published Aspects of the Theory of Syntax in 1965, in which he described the philosophical dimension of his approach, rooting it historically and theoretically. Seuil published the French translation in 1971. Starting with the observation that every child learns its mother tongue with remarkable speed, Chomsky argued that a child had the potential to learn any language. But rather than concluding that an initial context determined language acquisition, he argued for universal laws that determine languages as well as universals of language. Every individual therefore possesses an innate linguistic competence, to be distinguished from the use made of it in individual linguistic performance in a particular language.

Chomsky's linguistic universalism was therefore the expression of innateness, founded on a notion of human nature irrespective of cultural differences. This goal of universalization was also consonant with the general structuralist program, on the border between nature and culture. The analysis did not begin by describing any particular language, but started with the concept, the construction, in order to end up with reality. "The first object of linguistic theory is an ideal speaker-listener who belongs to a completely homogeneous linguistic community." 22

Chomsky's approach was doubly rooted. Historically, he invoked the European linguistic tradition going back to the grammar of Port-Royal. He used seventeenth-century Cartesian rationalism with the theories of innateness of the period, Cartesian substantialism.P and hoped to scientifically establish this innateness with the help of genetics. In this he echoed Levi-Strauss's goal of reaching mental structures.

"Everything happens as if the speaking subject___had assimilated

into its very thinking matter a coherent system of rules, a genetic code."24

For Chomsky, genetics, on the threshold of technological modernity, made this primary structure accessible. "By adopting the cogni-tivist program, Chomsky and the Cambridge school adopted the following proposition: an idea has the structure of coded information in a computer. "25 Chomsky believed that with generative grammar linguistics could claim to have attained the level of science, in the Galilean sense of the term. He was explicitly scientistic and his model was located in the natural sciences. Taking competence as his fundamental structure, he turned toward "an ontology of structures."26

Is the competence/performance distinction the equivalent of Saus-sure's language/speech dichotomy? Francoise Gadet considered that Chomsky essentially continued along lines drawn by Saussure: "This is a fundamental point where his ideas are consonant with those of Saussure.... Competence can be compared to Saussure's idea of language. "27 Indeed, we can easily discern a strong analogy between these two conceptual couples underlying the positive references Chomsky made to Jakobson even if, as of the early sixties, Saussure's conception of language was considered to be naive. For Nicolas Ruwet, however, underscoring the creative aspect of language in Chomsky implied that "Chomsky's distinction between competence and performance is radically different from Saussure's dichotomy between language and speech."28 Whereas Saussure defined language as a simple taxonomy of elements and limited creation to speech, Chomsky differentiated between two types of creativity: the one changed rules and the other was governed by rules. In the first case, performance gave the proof of creativity, whereas in the second, linguistic competence demonstrated creativity. For Ruwet, this notion fundamentally and radically renewed thinking about language: an infinity of possible sentences that the speaking subject could understand or produce while never having previously uttered or heard them.

Quite subtly, as Chomsky used the old notion of a dehistoricized and decontextualized human nature, structuralism became a structural naturalism. "Every true social science or any revolutionary theory of social change must be based on certain concepts of human nature."29 He reoriented toward a cognitive psychology of which linguistics would be only one element, and in so doing, announced the future paradigm of cognitivism and of neuronal man. By contrast with behaviorism, Chornsky insisted on innateness and its genetic foundations. "It was a question of considering general principles like the properties of a biological given making language acquisition possible.'^ And yet, Chomsky's field of investigation remained strictly linguistic, syntactic, and the inspiration he drew from the biological sciences only played an analogical and essentially methodological role that served to establish the framework for a universal grammar.

For Ruwet, Chomsky offered a way of discovering his road to Damascus and of leaving the sidelines to which structuralism relegated him. For many other linguists, however, there was no significant break between structuralism and generativism. Louis-Jean Calvet remarked: "For me, Chomsky was profoundly structuralist. He is the heir to Saussure. "31 For Calvet, Saussure's legacy lay essentially in his work on language as a scientific object removed from the social realm and concrete sociological or psychological situations. Nonetheless, from a heuristic viewpoint, Calvet credited Chomsky with having furthered the idea of a syntactic model. Similarly, Oswald Ducrot saw in Chomsky a continuation of Saussure: "I never perceived Chomsky as opposed to structuralism. I don't see why the search for a formal system that accounts for all possible utterances would be antistructural-ist. But it is true that, for historical reasons, he greatly threatened many in France who called themselves structuralists."32 Chomsky was as much a stranger to the idea of subject as he was to context, and he articulated his positions with respect to the continental model by referring to Descartes. He seemed to be constructing his generative grammar from the perspective of a European structuralist problematic, and in this respect the linguistics of utterances set the two approaches back to back.

Is Chomskyan Theory Antistructuralist? The tensions quickly ran quite high, however, between Chomsky, his disciples, and a certain number of eminent European representatives of structuralism, particularly Andre Martinet. Martinet had spent about ten years in the United States, from 1946 until July 1955, and was the editorial director of Word, one of the two important linguistic journals with positions radically opposed to the ambient and dominant Bloomfieldism. In the mid-fifties, Chomsky therefore chose to submit his first article on syntax to Andre Martinet. "Chomsky sent me his article at Word. I read it and immediately said, 'Impossible!' This perspective is going to get us into hot water. So from then on I was deemed the great enemy of Chomskyism."33 The polemic was immediately violent. And Martinet did not at all appreciate being relegated during his own lifetime to the ranks of "antiquities" by an often thankless new generation eager to break with the past and the discipline's founding fathers. He reacted by tending to energetically reject any elaboration of structural methods, at the risk of finding himself isolated in his bunker. But above all he was careful to preserve the heritage whose bearer he considered himself to be, by stalwartly resisting the vogue for Chomskyism. "Chomsky represents the heights of a priori assertion when he claims that all languages are basically identical and that a deep structure therefore exists."34

Martinet was caught between a humanist tradition that saw him as a dangerous structuralist respectful of nothing and the development of Chomskyism with its purely formal conception of language, which, precisely in the name of humanist positions, he was reluctant to accept. For Martinet, this was a linguistics of engineers. As the grand heir to phonology and the work of the Prague Circle, he

did not see himself going back to school to be retrained in mathematics and informatics. He chose to leave America ratherthan adopt

an unappealing direction that he also felt was misguided. Of course he did feel a certain bitterness, especially since he was contested by both the extensionalists, who wanted to broaden the structural method, and by those who wanted to formalize it.35

Claude Hagege agreed that generative grammar represented a break with respect to other linguistic traditions, but he considered it negatively to the extent that it radicalized a break from social reality in order to generate its formal models free "of all social and historical interference. "36 The Saussurean dichotomy between language and speech was also antisociological, but Saussure had been influenced by Durkheim; we can see his dichotomy between language and speech as a linguistic reiteration of Durkheim's distinction between systems defined by their social relationships and those produced by individual creativity. Chomsky, on the other hand, "totally betrayed this sociological tradition, which had a long history in both France and Ger-many."3?

Indeed, Chomsky broke with all tradition, particularly that of the comparatists, and could persuade neither Andre-Georges Haudricourt, for whom generativism had essentially negative effects, nor Tzvetan Todorov, who remained strictly loyal to jakobson and Benveniste.

The first structuralists were immersed in the plurality of languages, and could cite examples from Sanskrit, Chinese, Persian, German, or Russian. Chornsky, on the other hand, was the total and complete negation of all that because he always worked in and on English, his native language. Even if he was a good specialist in what he did, his influence was disastrous because it led to an altogether striking sterilization of linguistics.P

But Chomsky theorized this limitation to a native language and turned it into a methodological necessity: only a native speaker could recognize the grammaticality of a sentence from the language in question. Moreover, Chomsky's concern with syntax was perceived at once as a sign of progress, as if a new and long ignored field of analysis was being opened up, but also as a closure, because all other possible approaches-phonetics or semantics, for example-were eliminated.

Because of his notion of innateness and the distinction between surface and deep structure, Chomsky was considered by some to represent a regression. His approach did imply a return, which he made explicit, to the logic of Port-Royal according to which thought shapes itself independently from language, which serves only to communicate it. In other words, he took an essentially instrumental view of language that structuralism, since Saussure, had rightly contested. "It is clear to me that the notion that a thinking human nature or human essence exists, a priori, was an ideology that structuralism condemned, and vigorously rejected."39 In fact, the theoretical foundation that Chomsky laid with his notion of deep structure and human nature took its distance from structuralism in general, and, for example, from the fundamental principle expressed by Benveniste that "the linguist believes that no thought can exist without language. "40

Chomskyism: Structuralism's Second Wind? Despite the structuralists' and functionalists' biases against Chomsky's generative grammar, it undeniably gave a second wind to linguistics in France at the end of the sixties. It made its way thanks to the notion of transformation. In fact, generativism was initially known as transformational grammar.

Jean Dubois was an important promulgator of this model for the French. As early as 1965 he had applied certain aspects of Harris's dis-tributionalism." A French grammarian schooled in the thinking of the classical humanities about dead languages, Dubois turned his sights toward the models being used across the Atlantic. "Bloomfield was my preferred reading. The Americans were also working on languages they did not speak, on Amerindian languages. "42 Dubois's interest was also evident in his work in neurology. For many years, he worked with Dr. Henry Hecaen in a hospital laboratory in Montreal, and later in France. Dubois championed a syncretic position amalgamating the methods of functional structuralism, distributionalism, and genera-tivism. "Because I was involved in making a contemporary French dictionary, I came to use a method that was half structural and half transformational. "43 Dubois's theories translated his institutional situation. He was at a point of confluence between different currents, a professor at Nanterre, the director of the review Langages, as well as a collection of the same name at Larousse, not to mention his activism among the PCF linguists at the CERM.

Dubois's interest in generativism led to a definitive rupture with Martinet, who could not stand the increasingly numerous references to Chomsky, which he interpreted as being part of an effort to challenge him. Dubois dates their disagreement to experiences with Larousse:

Martinet had arranged to create his review and his collection at Larousse. Then, clumsily, because he is a veryhonestman, he undertook parallel negotiations with PUF without saying anything about it. Larousse did not appreciate this, particularly since Martinet preferred PUF because its name included the word 'university.' Martinet was extremely unhappy that the project was taking shape without him at Larousse. In fact, I was in the situation of launching Langages in 1966 without ever considering that I was on the same level as Martinet.f

This vogue for Chomskyism was appealing because it offered the possibility of making structure into something dynamic, and reconciling genesis and structure, even if these were not Chomsky's intentions. The entire generation of linguists, including Julia Kristeva, reacted in this way: "I read Chomsky with great interest because his model was more dynamic than the phonological model. I felt that this could correspond to the vision of meaning in progress that I was beginning to envisage. "45 In order to underscore this dynamization, Kristeva turned to biology and the oppositions between genotype and phenotype, which she imported into linguistics as a mode of articulation between genotext and phenotext. With this distinction she could explain that the text is in fact a phenotype organized according to certain quasi-instinctual processes determined by a genotype. The field of interpretation was also opened up to psychoanalysis. Kristeva was interested in Chomskyism, but she did not really adopt its ideas. She rejected postulates of innateness and the always already there-ne ss of linguistic notions, which seemed to her to be secondary with respect to a certain phenomenology and to Freudian thinking: "I was quite disappointed by our conversations because of his disdain for everything involving stylistics and poetics. So far as he was concerned, these phenomena were little more than decorations. "46

Cognitivism's First Steps

Another aspect of generativism perceived as a sign of clear progress was its ability to formalize and test its hypotheses by verifying their validity, even if, with the development of expert informatics systems, things have gone even further in terms of formalization. Historically, Chomsky was an important moment. "This was the first time that we had been able to define the structure of a linguistic theory or been able to evaluate the different possible explanations that it offered. "47

By pushing linguistics toward ever greater formalization, however, Chomskyism ended up by cutting it off from other social sciences, whereas, in contrast, the first effect, in the sixties, had been to breathe a new dynamism into linguistics, considered the pilot science among other social sciences. Generative grammar did infuse linguistics with its exigency for rigor and concern for explanation, along with a certain continuity of Saussurean thinking about language and its operations. But we might wonder whether generativists did not dig their own graves when such respected linguists as Francoise Gadet admit that generative grammar today "has become entirely unreadable."48 Nonetheless, generativism led to a scientific paradigm. In doing so, it renewed with the first thrust of structuralism and its hope of going beyond the naturelculture split to take the natural sciences as a model, with the cognitive paradigm. joelle Proust discovered Chom-skyism in the mid-sixties, but it was only in the seventies, at Berkeley, where she was immersed in the great flowering of the cognitive sciences, that she embarked on a new path. "At that point, I realized that many of the things I had learned had to be unlearned and re-assimilated differently. "49 She therefore adopted Chomskyism because of its search for the logical, computational, organic structure underlying the observable diversity of cultures. She embraced Chomsky's notion of human nature, which had been his first important theoretical reference and which Louis Althusser had qualified as an ideological notion. "Today, we have to admit that, scientifically speaking, there are universal bases to cognition; some things are shared by all members of our species and can, in principle, be duplicated by a machine. There is no reason to think that reason ends with man."50 This working hypothesis presupposed that man's reason may not be spe-cificto the organic matter that constitutes us, that a system of memory thinks because it calculates in symbols. After that, the only things that count are the relational, formal properties of calculus, while the organic aspect can vary just as different computers can use the same program. "That was the reason that we said that there might be a form of functional equivalence between men and machines."51

Chomsky's thinking was also used in anthropology, particularly thanks to Dan Sperber's double allegiance to Levi-Strauss and then Chomsky. This also ensured Chomsky's momentary success. Sperber sought to synthesize the two by examining the Levi-Straussian paradigm via Chomsky's theses. In 1968, he wrote the article on structuralism in anthropology in the collective work directed by Francois Wahl at Seuil, What Is Struauralismi-? Having dealt with the two preferred realms of structural analysis-kinship systems and mythol-ogy-Sperber addressed the structuralist theory in the same way that Chomsky did when he argued against the inductive and descriptive orientation of structural linguistics. He began with the principle that, contrary to what Levi-Strauss says, structuralism does not claim to be a scientific method so much as a theory, which should be tested as such, as Popper suggests: "After Chomsky had demonstrated that structuralism was a particular theory in linguistics-which, moreover, he considered to be false-and not a scientific method, we are justified in asking whether we are not dealing with a theory in anthropology as well-true or false. "53

Sperber began with this Chomskyian problematization to insist on the internal tension within a Levi-Straussian discourse between his scientific aspiration to reach mental structures and the ability to describe the semantic dimension of myths. Sperber credited Levi-Strauss with having entirely removed the study of myths from the conditions of their communication and with having envisaged them as codes. But if he lauded Levi-Strauss for this, he also criticized him for not having totally left the anthropological tradition but having stopped midway because he needed to construct the theory of the system. He criticized structuralism for continuing to regard myths as depending on a system of symbols. Of course, Levi-Strauss did break with empiricism when he evoked the internal constraints of the human mind, but, according to Sperber, he did not go so far as to construct a scientific method articulating the two levels of a work-levels that Levi-Strauss had discerned in his approach to myths, as, on the one hand, a language engendered by a grammar and, on the other hand, transformational products of other myths. Here Sperber reintroduced Chomsky's distinction between the structure of the mythic mind as competence and its exercise as performance. "I therefore saw that the transformation of myths among themselves did not define a grammar, contrary to what Levi-Strauss seemed to think."54 Levi-Strauss could only realize his revolution by moving toward cognitivism rather than in the direction of his semiological aspirations. "Claude Levi-Strauss's work brings anthropology back to the study of its initial object: human nature."55

The key to constructing a true anthropological science was therefore to be found in the orientation of the human mind. For Sperber, Chomsky offered the tools for a second conversion following the one that had already taken him from Balandier to Levi-Strauss.

Generative grammar was a true scientific revolution proving the structuralist model inadequate and far too simple. But generative grammar in no way aspires to spread to other disciplines. Structural linguistics, paradoxically, aspired to establish a broader discipline, whereas its model did not even work in its initial fieldof language. Its claim to work for the rest of the universe was altogether doubtful.56

Chomsky's scientific exigency was, for Sperber, the possible and necessary dissociation of ethnography as an interpretation of specifics dependent on a literary genre, and anthropology as a possible science of the general. Seen in this light, Levi-Strauss did not break radically enough with the anthropological tradition because he continued to try to house the two realms within a single discipline.

After the high point of the structuralist paradigm in 1966, the introduction of Chomsky's work in France in 1967-68 appeared curiously as both a second wind and a crisis for structuralism. It drastically changed the configuration of semiology, and a rupture took place relegating Lacan's 1964 lecture on Chomsky, in which he criticized his theoretical postulates, to the past. He then reiterated the criticism he had raised as early as 1959 against Jakobson, and reproached Chomsky for having enclosed the subject in a grammatical structure by forgetting that it was not a coherent entity, but a split being. In contrast to the grammatical model, Lacan proposed his formal theory of the signifier.v

If, in 1964, the structuralist model still presented itself as a possible unification of all the various fields of research on communication, in 1967-68, with Chomskyism, a decisive fault line appeared within the very heart of what had till then been seen as a pilot science: linguistics.


Derrida or Ultrastructuralism

What Americans call poststructuralism existed even before the structural paradigm waned. In fact, it was contemporary with its triumph. In 1967, two books came out by the same author, both questioning structuralism from a philosophical perspective. jacques Derrida's Of Grammatologyt and Writing and Difference,2 collections of articles he had been writing since 1963, like the one on Jean Rousset, continued to target the problem of the spatiotemporal split that he perceived in the works of classical philosophy.

Derrida was born on July 15, 1930, in El-Biar, Algeria, in a Jewish milieu, although he was never completely immersed in a truly Jewish culture. "Unfortunately, I do not speak Hebrew. My childhood in Algiers was too colonized and too uprooted."3 And yet he always felt, and cultivated, a certain foreignness with respect to the Western tradition. His exteriority was not, however, based on an Other, on another place, but on a lack, a place that was nowhere and that he had left at age nineteen, an outside space that dimmed any glimmer of a foundation. "The gesture that seeks to find draws itself away from itself. We should be able to formalize the law of this insurmountable separation. It is a game I always play. Identification is a difference from oneself, a difference with oneself. Therefore, with, without, and except oneself.?" Derrida relived in writing his personal experience of loss of time and memory and that which remains as ashes after death. "It's the experience of forgetting, but the forgetting of forgetting, the forgetting of which nothing remains.I" This personal itinerary led Derrida, like many of the philosophers of his generation, to Heiddegger. And the principle of deconstruction fueling his entire undertaking was nothing more than the slightly displaced translation of Heidegger's term, Destruktion.

Derrida, Phenomenologist

Before becoming the deconstructor of critical thinking as represented by structuralism, Derrida had been interested in phenomenology. His first published work was an introduction to Husserl's Origin of Geometrys Phenomenology was in vogue at the time and practically dominated philosophy in France; Sartre and Merleau-Ponty had given it a particular inflection through their interest in lived experience and the perceiving consciousness. Derrida's contribution was original: he did not adopt this orientation, but was interested above all in questions of objectivity and science and, by avoiding the level of internal observation, was more in the line of the German disciples of Husserl. Derrida did not deduce the death of the subject from his examination of the ultimate basis of phenomenology using the enigma of the geometric object, but rather its limits within a more restricted sphere. He spoke about the withdrawal of the principle of foundation, "necessary to appearance itself."? Using Husserl's text, he criticized the double errors of historicism and objectivism. He had already found, in Origin of Geometry, the internal subversion of the hierarchization used to subordinate writing to the voice, a theme developed in all later decon-struction. The notion of "transcendental" was the absolute certainty of this progression toward an origin perceived in its original difference, always to come. "It is also in this way that this writing has, as Husserl said, 'an exemplary meaning.t'"

Derrida then turned to the sign and to language, still using the Husserlian axiom of Logical Search" to emphasize Husserl's distinction between a preexpressive level (indicative sign) and an expressive level (expressive sign) in states of consciousness. The sign was not unitary but doubled. Husserl saw expression as complete externalization and the indication referred to the site of the involuntary. "The indicative sphere remaining outside of expressivity defined in this way limits the failure of this teloS."lO We cannot point to any truth or essence of the sign; the philosopher's task is to describe the conditions of its appearance. The theme of textual indefiniteness, of writing as

abyss, as a veritable cryptic universe, of a past that has never been present, was already here in Derrida: "To conceive as preoriginal and normal what Husserl thought he could isolate as a specific, accidental, dependent, and secondary experience: that of the indefinite drift of signs as errance." 11

Radicalizing Structuralism

At a time when structuralism was calling phenomenology into question in France, Derrida might have found himself aligned with tradition. He therefore managed to "radicalize phenomenology in order to go further even than the structuralist objection."12 Going quickly from a defensive to an offensive position, Derrida systematically deconstructed each structuralist work by locating the traces of a logo-centrism that had not yet been surpassed. He did not adopt a phenom-enological perspective but used Heidegger to wage critical warfare against structuralism. His position was paradoxical for he was at one and the same time inside and outside the structuralist paradigm, "the first in France to raise a certain number of reservations about structuralism, and Derridean deconstruction was a movement that weakened the development of structuralism as it might have evolved."13 But he might just as well have been considered to be the person who pushed the structuralist logic to its limits and toward an even more radical interrogation of all substantification or founding essence, in the sense of eliminating the signified. In this respect, Derrida was a structuralist thinker from the outset, even if he adopted a position of critical distance: "Since we live from the productivity of structuralism, it is too soon to destroy our dream."14 This was only 1963, of course, during the still-glorious period of a promising program; Derrida was effusively praising structuralism, which he considered to be far more than a simple method or new form of thought. Structuralism at the time was a new "adventure of observation, a conversion regarding the manner of questioning any object." 15

Derrida was referring to a veritable epistemological revolution, which could not be ignored, at the same time that he considered structuralism to be neither a simple question of style nor simply a particular moment in the history of thought. "The classical history of ideas cannot define structuralism," 16 even if it unfolded during a period of historical dislocation during which the immanent fervor was reduced to formal concerns. Thus, "the structuralist consciousness is simply the consciousness of thinking the past, I mean of the general fact. A reflection on what has been accomplished, established, constructed.:"? Even if Derrida, like Foucault, systematically avoided partisanship, he clearly abandoned the phenomenological horizon for the bases of the structuralist paradigm. Many structuralist semiologists of the sixties and seventies also drew on his work. "Deconstruction as a method was another name for a type of structuralist approach, which is to say transforming and simplifying a complicated text in order to make it legible, reducing it to oppositions and dysfunctions."18 Unlike the classical structuralists, however, Derrida, like psychoanalysts, was more interested in failure and dysfunction, rather than in the regularities or the invariants of structure. This considering of limits, which was also part of an entire literature of the same order, radicalized the idea of structural structurality by constantly decentering, such that no extrastructural order existed; "everything is structure and all struc-rurality is an infinite game of differences."19 Structure became an endless game of differences, and thinking fell into the abyssal vertigo of a kind of writing that abolishes disciplinary boundaries in order to attain the pure creation of a writer incarnated by the figure of the poet.

This aesthetic concern drew its inspiration from the Mallarmean program and led to muddied boundaries between philosophy and literature, subjected to philosophical problematization anchored on the grounds of undecidability. The corpus of writers was drawn from the nether side of literary history: Antonin Artaud, Georges Bataille, Edmond jabes, and so on. This proximity was consonant with the structuralist slant on questioning language, radicalizing it beyond the separation of genres and traditional classifications in order to address the text through the specific laws of textualiry, "My first desire tended surely toward the point where the literary event traversed and even went beyond philosophy.t'-?

Derrida continued to move ever closer to literature and away from epistemological concerns for pure creation. Glas provides a good example.P In 1965, he was a teacher at the Ecole Nationale Superieure (ENS), and an excellent didact of philosophy. Derrida was not only the first but also the most radical in changing the way philosophical texts were read, using new modes of interpretation taken from linguistics, psychoanalysis, ethnology, and all the leading disciplines of the social sciences. "He was basically a professor who profoundly changed and renewed the reading of philosophical texts, but who ran after his interpretations. There was a somewhat blind side to the energy spent on establishing a practice. His readings raised the problem of their foundations. "22 He could put himself in the text being deconstructed in order to follow its internal thread. For a generation of philosophers, he was "extraordinarily efficient, incorporating by impregnation in order to give philosophy students, who must first of all master rhetoric, the impression of being heard and awaited by the professor. "23


Derrida's strategy was that of deconstruction. This meant both destruction and construction, a strategy that made it possible to recognize the traces of Western metaphysics in the thinking of the other, while introducing a new manner of writing. Such a strategy considered the realm of writing to be autonomous and belonging to textual-ity in general, beyond the generic difference between philosophy and literature. So Derrida was in tune with the new structuralist literary criticism, but he avoided its scientistic categories by aiming to create new and undecidable concepts and thereby rising to "the heights of a creative activity."24 He realized the grand ambition of most structuralists who borrowed language from the social sciences in order to produce a creative, literary work. He was also in harmony with the early twentieth-century formalists of the Prague Circle, who had already tried to make a symbiosis between poetics and philosophical reflection. Derrida was therefore fully in line with structuralist objectives.

Heidegger's work was Derrida's other specifically philosophical source. "What I have attempted to do would not have been possible without the opening of Heidegger's questions, ... without the attention to what Heidegger calls the difference between Being and beings, ontico-ontological difference. "25 In this respect, the entire work of deconstructing meaning attributed to being is directly set in Heidegger's tracks. Each concept was addressed to the limits of its pertinence, to its exhaustion and disappearance, which were supposed to simulate the disappearance of Western metaphysics itself. Deconstruction appeared in all its ambiguity, and was all the more seductive in 1967-68, because it was perceived "both as a structuralist and an antistructural-ist gesture. "26 Derrida could thus win over an entire generation, as illustrated by Tel Quel, to recapture the structuralist legacy, while at the same time putting an end to structural hermeticism. Deconstruction continued to valorize the hidden sphere of the unconscious but it made pluralization and dissemination possible by exploding all references to a structural center or to the unity of any kind of structuring principle. Derrida adopted a real strategy with respect to Western rationality. "The strategy of deconstruction is the ruse that makes it possible to speak, at the same time as there is, finally, nothing more to say."27

Derrida's success also had to do with the context of rupture with the Sorbonne's academicism. Like linguists with respect to dassicallit-erary history, Derrida provided philosophers with a combat strategy that could radically destroy the foundations of metaphysics taught at the Sorbonne. Within the philosophical tradition, he introduced a series of undecidable concepts in order to shake up the foundations and denounce the errors. The subversive dimension of this strategy made it possible to weaken institutional foundations and radicalize the struggle of the structuralists, to broaden its base by linking it to all critical reflec-tion-whether Lacanian, Foucauldian, Chomskyan, or Althusserian-while at the same time recuperating it within philosophy.

In this respect, it was also Derrida who took seriously the challenge of the new social sciences to enrich discourse and the questioning of philosophy. This strategy proclaimed the end of philosophy and recuperated the social sciences for philosophy, while at the same time discovering—even before Barthes's book had come out-textual pleasure: "There is produced a certain textual work that gives great pleasure. "28 The various binary couples-signifierlsignified, nature/ culture, voice/writing, perceptible/intelligible-that compose the very instrument of structural analysis were each put into question, plural-ized, disseminated in an infinite game that peeled, disjoined, and dissected the meaning of words, tracking down every master word, every transcendence. A whole Derridean language destabilized traditional oppositions by bringing undecidables into play as veritable units of simulacrum, organizers of a new, carnivalesque order of reason.

Derrida drew his ambivalent concepts from the philosophical tradition in order to create a boomerang effect. From Plato he took the term Pharmakon, neither remedy nor poison, neither good nor evil. From Rousseau he took supplement, neither some more nor less, from Mallarme hymen, neither confusion nor distinction. These ideas served as so many instruments of deconstruction, and had a point in common. "All cross out opposition from inside and outside."29 Writ-ing therefore could ready its attack on the concept and substitute a seminal stream opening up onto infinity. Deconstruction attacked not only phenomenology by decentering the subject but also the Hegelian dialectic with its notions of unity and identity. "Negation is relegated to the secondary role of a knowledge police____The concept is^ relegated to the exercise of a theological commandment."3o Derrida continued to give philosophy pride of place among the sciences while preparing a possible escape route via purely literary creation but not, à la Heidegger, as redemption. The Heideggerian perspective was radicalized because any possible foundation was eliminated. In its place was a simple errancy that "does not even grant itself a pause to recollect Being,"3! and that preferred the Mallarmean margins. After Saussure, the referent was no longer a linguistic concern, Lacan had slipped the signified beneath the signifier, and Derrida eliminated it in favor of an indefinite signifying chain without any site at which it could be perceived. In so doing, he set in motion a spectacular reversal, and from this point on sought a physicalness of writing.

Deconstructing Foucault

Derrida wanted to deconstruct everything. He began with the structuralists, who were closest to him and whom he considered to have remained prisoners of logocentrism despite themselves. His first target resembled the death of the father because the expiatory victim was none other than Michel Foucault, who had been his professor at the rue d'Ulm, Derrida had become Jean Wahl's assistant at the Sorbonne. When asked to lecture at the College de Philosophie.v he chose to lecture on Foucault's Madness and Ciuilization.v The conference took place on March 4, 1963, with Foucault in the audience, much surprised by his former student's attacks. Derrida published the lecture shortly thereafter in the Revue de rnetapbysique et de morale under Jean Wahl's direction.>' and later in Writing and Difference, a collection of articles published in 1967.

Derrida deconstructed the text, concentrating only on its internal economy. Selecting a minute part of it that he felt revealed the whole, much like a medical biopsy, he set his scalpel to work. Thus Foucault's tremendous doctoral thesis" was criticized on the slim biopsy of his reading of Descartes's position on madness, a mere three of the 673 pages! "The reading of Descartes and of the Cartesian cogito engages the entire problematic of Madness and Civilization."36 Since contesting the validity of the lessons Foucault drew from Descartes's first Meditation implied the whole of his work, we can take the measure of the radicality of a criticism of a work nonetheless considered to be a "book that is admirable in so many respects. "37 But the hour of emancipation had sounded, and with it came the symbolic murder.

First, in his capacity as a radical structuralist, Derrida criticized Foucault for having preserved the idea of subject. Even if his subject was the underside of history, Foucault was wrong to claim that the idea of a subject ran through history-that of madness. "This is what is maddest about his project."38 Foucault was in fact sensitive to this criticism and his later archaeological project erased any viewpoint starting with any type of subject, repressed or not. Next, Derrida considered that it was illusory to imagine being able to place oneself outside the bounds of reason based on the elsewhere of madness as a site of exile. "The irreplaceable, insurmountable imperial grandeur of the rational order ... is such that we cannot argue against it except by using it, we cannot protest reason without reason."39 Where Foucault believed he had created a revolution, he had done little more than slightly disturb things. His demonstration began with the important decision presented as the very condition of history, initially excluding madness from the world of reason before incarcerating it. Foucault attributed this founding act of the classical age to Descartes in his first Meditation in which he ostensibly set the dividing line between two soliloquies that remain forever foreign to one another. This was the big point of contention between Foucault and Derrida, who considered that Descartes had not ostracized madness. On the contrary, for Descartes, "the sleeper or the dreamer is more mad than the madman. "40 Even if the hypothesis of the sly genius calls for total madness, the decisive split between reason and madness is not situated with the cogito since it is valid "even if my thinking is mad through and through. "41 Derrida thus challenged the validity of the binary pair reason/madness (a split that allowed Foucault to exhume the repressed part of Western history) by demonstrating that when Descartes created his cogito, he did not eliminate madness.

Derrida thus considered that Foucault seriously misread Descartes. But he went even further, fundamentally questioning Foucault's method. "Structuralist totalitarianism would enclose the cogito in a manner reminiscent of the violence of the classical age."42 With this, Foucault was accused of having perpetrated the same kind of violence he denounced. We can understand that he did not particularly appreciate his disciple's Parthian arrow, but he took his time before answering, silent and attentive in the audience and no less silent when the text appeared in Writing and Difference in 1967. In 1971, however, he responded in an article initially published in the review Paideia'? and later reprinted in the new edition of Madness and Civilization published by Gallimard in 1972. Foucault called Derrida's argument "remarkable," but he maintained his interpretation of Descartes's text and considered Derrida's hypothesis to be valid only at the price of what it left out, and which allowed him to extirpate all the differences of the text in order to "transform the Cartesian exclusion into inclusion. "44 Moreover, Foucault considered Derrida's reading of Descartes to be not only naive but an application of a traditional system of interpretation characterized by erasing whatever is bothersome. Derrida became the latest representative of this tradition, as Foucault saw it. He did not limit himself to a defensive answer but went on to respond as a master judging his pupil. He reduced Derrida's work to a brilliant stylistic exercise with didactic pretentions:

I won't say that it is a metaphysics, metaphysics, or its closure that are hidden in this textualization of discursive practices. I will go much further and say that this is a small and historically well-determined little lesson that, in a very visible way, shows itself off. Pedagogy teaches the student that there is nothing beyond the text, and that within it, its intentions, its spaces, and in what goes unspoken, reigns the realm of the origin."


In 1965, in the review Critique, Derrida laid out the bases of a new approach that was part of the logy-effect of the times-grammatol-ogy. His theses reached a broader public in 1967 when the Editions de Minuit published Of Grammatology. Beginning with the observation that the problem of language had never before so intensely concerned research in so many different disciplines, and using this to respond, as a philosopher, he advocated historicizing Western civilization's repression of writing and preference for the phone. Grammatology was the "science" of writing restrained by metaphysics, which "shows signs of liberation all over the world, as a result of decisive efforts. "46 The epigraph therefore referred to a scientific ambition, but this ambition was annulled as soon as it was announced since, once all the obstacles had been overcome, "such a science of writing runs the risk of never being established as such and named that way."47 Grammatology was not a discipline like others. "Graphematics or grammatography ought no longer to be presented as sciences."48 Derrida was already in this intermediary zone, in an internal tension between writing and science, within a space of lack, of textual blankness, of inaccessible temporal spacing that dons the figure of a supplement forever eluding presence: "The constitution of a science or a philosophy of writing is a necessary and difficult task. But, thinking about the trace, différante^ or reserve, having arrived at these limits and repeating them ceaselessly, must also point beyond the field of the episteme. "49

Since Plato, the West has incessantly sought the proximate and the specific, since the voice is considered to be a veritable essence, a bearer of meaning and of the signified. The entire historical thread of the West is nothing other than the history of the elimination of writing. The distinctive unit and object of this new science capable of going beyond phonologism was the gramme, the grapheme. Derrida used Hjelmslev's glossematics, the most formal of linguistics. "Derrida considered that Hjelmslev freed the signifier from the signified and made possible a writing that replaces the phonic signifier," 50 The graphic signifier can thus replace the phonic signifier.

Drawing his inspiration from Hjelmslev's Principles of General Grammar, Derrida dissociated the phonologistic principle from that of difference. Glossematics gave him the bases of a formal science of language. Setting meaning aside also meant setting aside sound, and Derrida built his new science on linguistics, in an explicitly Hjelmsle-vian tradition. "Hjelmslev criticizes the idea of a language naturally bound to the substance of phonic expression."51 The split allowing us to give primacy to writing dated from glossematics, alone making it possible to reach the basic literary element of the gramme.

But Derrida was not satisfied with taking up the legacy of the most formal branch of linguistic thought. He wanted to go beyond structuralism, claiming that "glossematics still operates with a popular concept of writing." 52 He introduced temporality, the lack of being, absence based on which writing is understood as trace, unteth-ered from the idea of an origin. Trace referred to understanding the conditions of possibility preceding the existence of the sign, its condition of existence being irreducible to any being-present. Derrida was trying to create a symbiosis between glossematics and an archaeologi-cal approach; his goal was not to restore the contents of thinking, but the conditions that make thinking possible. He set himself outside of and eccentric to Western thought. Grammatology presupposed a maximal autonomy of writing from its genesis, and in this respect entirely embraced the structuralist paradigm that broke with the referential framework. Writing eluded the speaker just as it did the listener and constituted a scientific object, like any other, by virtue of the reit-erability of reading. "All graphemes are of a testamentary essence."53

For Derrida, phonological structuralism had gone as far as it could go, but he avoided generative grammar, preferring a specifically philosophical approach. "He tried to radicalize the fundamental intuition of structuralism____He took the direct route leading from the philosophy of conscious of the early Husserl to the philosophy of language of the late Heidegger." 54 In a Heideggerian perspective, Derrida renounced all ontology. The trace he perceived always eluded itself by a continuous veiling making it impossible to establish any meaning. So he used linguistics, transporting certain notions into philosophy via a scientific ambition foreign to Heidegger, who had always been hostile to scientific pretensions.

Grammatology was both a possible Heideggerian deconstruction of current scientific norms and a possible means of getting beyond the limits of the traditional field of scientificity toward a new scientific rigor free of logocentrist and phonological presuppositions. The critical side of grammatology especially was productive, but, as the prolegomena of a new science, it was quickly forgotten. The development of rationality was to be destabilized not by its hidden face of madness, as Foucault had thought, but from a truly external position. "We want to reach the point of a certain exteriority with respect to the totality of the logocentric period." 55

Beyond Structuralism

Constructing the future of structuralism required criticizing two of its founding fathers-Saussure and Levi-Strauss. Of Grammatology sought to do just this by identifying the phonological and logocentric limits of early structuralism. In Saussure, Derrida revealed an approach that had remained fundamentally a prisoner of the subject present to itself through its speech. He acknowledged that Saussure had broken with the metaphysical tradition by desubstantializing the contents of the signified and its expression but nonetheless felt that Saus-sure had not gone far enough when he sketched out the notion of sign as a founding notion of linguistics, since the "age of the sign is essentially theological."56 Saussurean thought was centered on the word as the unit of meaning and sound. It might have opened up an analysis of writing, but Saussure had refused this perspective by placing it in a quasi-evil exteriority: "Already in the Phaedrus Plato says that the evil of writing comes from without (275 a)."57 Plato repressed writing because it was responsible for ruining memory, and Saussure, who demonstrated the importance of becoming aware of the specific way in which language functioned, began by devalorizing writing and presenting it as the simple reproduction of speech: "Writing obscures language, it is not a guise for language, but a disguise." 58 Writing and speech therefore were subordinate and devalorized one with respect to the other, something Saussure reinforced by incorporating the semio-logical project into psychology.

Nothing justified Saussure's distinction between linguistic and graphic signs, as far as Derrida was concerned. There was even an internal contradiction in Saussure when he suggested the arbitrariness of the sign and at the same time banished writing to the exteriority of language, its antechamber, or even its leper hospital. "One must therefore challenge, in the very name of the arbitrariness of the sign, the Saussurean definition of writing as 'image'-hence as natural symbol-of language."59 For Derrida, to the contrary, writing escaped reality as a trace forever hidden from itself, as foreign as the acoustic image was to the referent or subject. "This deconstruction of presence accomplishes itself through the deconstruction of consciousness, and therefore through the irreducible notion of the trace (Spur). "60

The issue was to deconstruct this notion of the Saussurean sign central to structuralist thinking and to problematize writing. Given the collapse of disciplinary boundaries among those disciplines studying man, grammatology offered a way to consolidate the research that was taking off in all directions. It was presented as the fully accepted realization of structuralist ambitions pushed to their logical limits and receptive to the deconstruction of the One and to the disappearance of man. "Grammatology ought not to be one ofthe sciences ofman because it asks first, as its characteristic question, the question of the name ofman. "61 This science proposed to transcend the social sciences by appropriating their program without trying to use concepts from other disciplines. Derrida's hegemonic bid in fact reiterated the dominant position of philosophy among those disciplines studying man, and if he preferred science to philosophy, this science was not simply to be added to the others but claimed to be free of all limits and boundaries.

Deconstructing Levi-Strauss

Obviously, Levi-Strauss was the other grand master of structuralism whom Derrida took on. For Levi-Strauss, Jakobson's phonological model had been a valid scientific model for all of the social sciences. "Phonology cannot help but play the same role of renewing the social sciences that nuclear physics, for example, played for all the hard sciences.'^ Since Derrida was hunting down the traces of phonologism, he quite naturally took on Levi-Strauss, applying the same method he had used with Foucault. He took a small portion of an immense work-the "Writing Lesson" in Tristes Tropiques, in which Levi-Strauss described the advent of writing among the Nambikwara, which introduced exploitation, perfidy, and different forms of servil-ity-and denounced the repression of writing. He considered Levi-Strauss's observations to constitute the proof that he was no more successful than Saussure had been in fully decentering Western ethno-centrism. It is true that Levi-Strauss belonged to the era of suspicion which had introduced the logic of play in order to escape conscious models, that preferred a signifying chain to a central signified and tried to escape the traditional nature/culture dichotomy. In all of these ways, Derrida's project "is quite clearly consonant with Levi-Strauss's, even if it does not begin, as Levi-Strauss did, by solemnly abjuring the exercise of philosophy."63 Both sought differences between myths. Levi-Strauss saw their reciprocity, whereas Derrida saw them as texts within an intertextual weave. In issue 4 of Les Cahiers pour l'analyse, Derrida argued that Levi-Strauss's social anthropology reactivated Rousseau's thinking, and therefore implied a whole series of categories including genesis, nature, and sign, all of which demonstrated his logocentrism.w "Structuralism would remain a tributary of a philosophy of nature. "65 His article was reprinted in Of Grammatology, where he considered that Levi-Strauss was expressing his Western guilt by opposing innocent nature, full of goodness and beauty, and Western culture, which had breached this ideal reality presented through the equally deforming mirror of Western counter-ethnocentrism. "As always, this archeology is also a teleology and an eschatology; the dream of a full and immediate presence closing history. "66

Defending the philosophical turf that Levi-Strauss had abandoned, Derrida denounced anthropology's empiricism. In response to Levi-Strauss's criticisms of philosophers of consciousness, Derrida answered that none of them, neither Descartes nor Husserl, would have been as naive as Levi-Strauss had been to conclude so hastily in favor of the innocence and original goodness of the Nambikwara. Derrida saw Levi-Strauss's ostensibly ethnocentric-free viewpoint as a reverse ethnocentrism with ethical-political positions accusing the West of being initially responsible, through writing, for the death of innocent speech.

In this respect, Levi-Strauss was like Rousseau, his master, who had warned against writing: "The misuse of books is the death of learning." "We should not read, but rather look." "Reading is the curse of childhood."67 Hoping to get beyond Levi-Straussian structuralism, Derrida adopted certain of his positions minus their Rous-seauian underpinnings, which prevented Levi-Strauss from breaking radically and forced him to borrow all the old conceptual tools and metaphysical dichotomies that he believed he had gone beyond but that caught up with him.

Derrida analyzed these Rousseauian presuppositions by reestablishing the site, the issues, and the link with Rousseau's Essay on the Origins of Language.68 He pointed out Rousseau's classic opposition between voice and writing as reproducing the opposition between presence and absence, between freedom and servitude. Rousseau ended his Essay with this judgment: "But I say that any tongue with which one cannot make oneself understood to the people is a slavish tongue. It is impossible for a people to remain free and speak that tongue. "69 The soft maternal voice contrasts with the pitiless voice of writing. The social shift toward evil comes from a catastrophic moment, a simple and barely perceptible inaugural displacement. "He who willed man to be social, by a touch of the finger, shifted the globe's axis into line with the axis of the universe. I see such a slight movement changing the face of the earth and deciding the vocation of mankind."70 This subtle movement, this little sign, was no less than the hand of God, the divine trace. It opened the age of society, and with it the prohibition of incest: "Before the festival, there was no incest because there was no prohibition of incest. After the festival, there is no longer any incest because it is forbidden."?' This prohibition was the Law determining all laws and was, as it was later for Levi-Strauss, the seam between nature and culture. According to Derrida, Rousseau's description of this substitution of written expression for speech and presence was useful. A prisoner of metaphysics, however, Rousseau could not conceive of this writing as endogenous to and preceding speech: "But Rousseau's dream consisted of making the supplement enter metapysics by force."72 He kept life and death, good and evil, the signifier and the signified in a relationship of exteriority, whereas Derrida intended to shift all of these boundaries.


Derridean Historicization and Its Erasure

In 1966, Jacques Derrida went to the United States for a colloquium on the topic "The Languages of Criticism and the Sciences of Man" at Johns Hopkins University in Baltimore. Roland Barthes, Jacques Lacan, Gerard Genette, jean-Pierre Vernant, Lucien Goldmann, Tzve-tan Todorov, and Nicolas Ruwet also took part. French critical thought, united beneath the structuralist banner, was at its zenith. Americans were fascinated and curious about what was going on in France. Derrida's lecture was symptomatic. He described his double position as a structuralist seeking to go beyond the paradigm while at the same time defending critical thought and criticizing criticism for not going far enough. He set his lecture-"Structure, the Sign, and the Game in the Discourse of the Human Sciences"-within the scope of Levi-Strauss's work in order to deconstruct it. Although he acknowledged that structuralism heralded an important break, he himself intended to open up a game of differences by denying any reference to any kind of center that would put an end to the game of possibilities. Yet, "a structure lacking any center represents the inconceivable itself."! He therefore took on the heart of structural thinking and as a result was considered a poststructuralist in America.

Derrida agreed that Levi-Strauss's Mythologiques represented an entirely positive attempt to jettison all references to a center. "The mytho-logical discourse should itself be mytho-morphic. It should have the shape of what it speaks of."2 But while considering that Levi-

Strauss's thinking had gone in directions that were harmonious with the work of deconstruction, he still criticized Levi-Strauss for lacking dynamism and for neutralizing historicity, which belonged to his structural theme.

Historicizing Structures: Differance

It was true that Levi-Strauss had rightly broken with history as the accomplice of Western metaphysics. In doing so, however, he had run the risk of an equally classical ahistoricism, since he referred to a Rousseauian notion of history. This was one of the major thrusts of Derrida's criticism. With it, he echoed the feeling so present during the latter half of the sixties of wanting to make the structures more dynamic and more historical. This was the sense of the notion of differance that he introduced, and which was the focus of his lecture at the French Philosophical Society on January 27, 1968. Written with an a rather than an e, differance became the most effective instrument of deconstruction. It had both the temporal sense of deferring, of putting something off-"This temporizing is also a temporalization and a spatialization, the becoming-time of space and the becoming-space of time"3-and the more common meaning of differing, pointing to that which is not identical. Derrida could pull these two definitions together with the "a" of differance in order to introduce the idea of temporization, missing in the classical definition of difference. Ideally, this idea let Derrida play the role of undecidable that systematically unveiled any illusion about thinking of being by contrasting that which, in the present of the present, never presents itself. Differance also played on reintroducing the absent movement in the idea of structure. It added a dynamic dimension from within, drawing it infinitely forward. By offering an example of an aurally imperceptible but graphically visible concept, Derrida could reduce the importance of the phonological postulates of structuralism: "Contrary to an important and widely held prejudice, there is no phonetic writing."4

Deconstruction's most important notion made it possible to consider the conditions of possibility of what we call reality, rather than reality itself. There was therefore no concern for any essence or existence; what counted was to open up the deconstructive play of logos as broadly as possible. The term differance also magnificently expressed Derrida's ambiguous position with regard to structuralism. He agreed that Levi-Strauss had found a conception of difference in primitive societies, but at the same time he wanted to radicalize this savage mind even further by freeing it from any empirical reality. He hoped to set all Western metaphysics on fire. The notion of differance or of trace-as simulacrum-of presence was also expressive of the literature of Maurice Blanchot in particular, who privileged the figure of the oxymoron, a rhetorical figure of internal contradiction wherein an identity contains its own effacement.

In the same way that Being was forever absent from being for Heidegger, différance conditioned the existence of positivities without ever being perceptible in them for Derrida. Claiming that "the theme of différance is incompatible with the static, synchronic, taxinomie, ahistoric motifs in the concept of structure,"5 Derrida also set this concept in a continuum with structuralist positions. "The concept of differance even develops the most legitimate principled exigencies of structuralism."6 Beginning with the sign and the signifierlsignified distinction in order to valorize signification at work even within the signified, he argued for a shift of the signified into the realm of the signifier, and thus made it impossible to codify language, opening it quite broadly onto the sphere of literary creativity. "To risk meaning nothing is to start to play, and first to enter in the play of differance."7 To say that nothing remains to be said was deconstruction's perspective on the suspension/suspense of meaning.

Derrida reopened the door to history and the moment, but kept a traditional notion of history. Here, he based himself on Althusser's antihistoricist denunciation in his critique of Hegelianism. History was therefore also to be deconstructed, and if total history was banished to the illusory role of myth and error, it could nonetheless be perceived as plural and partial histories: "there is not one single history, a general history, but rather histories different in their type, rhythm, mode of inscription, intervallic, differentiated histories."8 This multidimensional history made it possible to transcribe a conception of writing and to let movement filter into structure. But the wrinkles in time that unfold this knowledge in fact lead to its disappearance. Deconstructed history led to a foreclosed future. It was nothing more than the unfolding of the simulacrum of a slack and ungraspable present. There was no stopping point in the temporal carnival and even fewer options for shifting from one point to another. With his obvious penchant for making endlessly moving movement, Derrida reinfused the morphologism being used at the time with a certain vitality and relativized the thrust of all philosophical concepts.

Hermeneutical reading became meaningless through deconstruction for hermeneutics was possible only when interpretation was limited. "A generalized interpretative posture is impossible unless it is conceived in a Nietzschean perspective."? Having considered the battle between interpretations to be endless, Derrida questioned the autonomous ontological existence of the text itself. Like Nietzsche, he shifted the original text and its contents from an experimental to an imaginative field. This supposed an initial "erasure"; barely emerging from a netherworld, the text was immediately dissolved. "Generalized intertextuality and the critique of textual closure are themes which only reiterate the Nietzschean paradox. They are a hypercriticism."lO This infinite flow of the order of things makes any effort at under-'standing vain. At the same time, it postulates an originary impotence. All of Derrida's liminary declarations, oral as well as written, "clearly express Achilles' anxiety at being unable to catch up with the tortoise. Since water cannot be stopped in order to grasp the river, reality col-lapses."11 During the debates of the sixties between hermeneutics and structuralism, Derrida sided with structuralism and took a firmer position on the question of eliminating the subject and the referent, while granting them the mobility and lability they had been lacking.

Deconstructing Freud

Derrida's ideas were quite close to Freud's analytic practice, although he did not lean on psychoanalytic theory. Thus, the concept of trace evoked involuntary unconscious manifestations, even though it referred to no identity whatsoever, repressed or not. Some basic ideas were, however, transferred from psychoanalysis to graphematics. From the analyst'S floating attention to Derrida's polyvalent polysemism there is a common ground, and a basis for collaboration and for a possible suture between deconstruction and the acknowledged scien-tificity of psychoanalytic discourse. "Derrida reinscribes what Freud explained by repression within the general economy of the text. "12 Here, the idea of differance was conceived as a means of taking into account the forces of wear that Freud had observed, their modes of inscription in the ever out-of-sync moments of the after-the-fact.

Deconstruction needed Freud. Derrida's deconstruction of him, presented in a lecture at Andre Green's seminar at the Psychoanalytic

Institute and published in Tel Quel in 1966,13 adopted the Freudian break insofar as it questioned the traditional divisions between normal and pathological, and pointed to the illusions of the notion of consciousness. Freud had offered a new conception of temporality. Particularly the notion of after-the-fact cast the originary moment as a supplement, and supported Derrida's argument for differance. The Freudian unconscious eluded this presence of the present that Derrida was tracking; it was always already out of sync, deferred, woven of differences, and forever other with respect to consciousness.

Derrida therefore paid Freud his deepest respects. "His thinking is doubtless the only one that is not exhausted in metaphysics or science."> And he acknowledged that Freud was the only thinker not to have repressed writing, but, on the contrary, to have problematized the stages of its infinite unfolding by means of paths opened by wear and through established resistances.

However, Derrida wanted to go beyond Freud, who was too conservative with respect to his own rupture. Freudian concepts were to be revisited since they "all, and without exception, belong to the history of metaphysics, which is to say to the system of logocentric repression that was organized in order to exclude or debase ... the body of the written trace."15 Derrida did not stop with the psychoanalytic notion of displacement. He completely reinserted everything outside the text, the hors d'oeuvre, into the textual weave itself, without limiting it to any interpretation that would, through displacement, favor certain elements of the trace to the detriment of others in order to recompose a hierarchical explanatory system. Psychoanalysis could not consider itself to be an encompassing science; it could claim no interpretative privilege. And yet, given that its major object was the analysis of dreams, whose singular space offered no tangible boundaries with the nonphonetic space of writing, psychoanalysis was a force to be contended with given the consideration and status conferred on writing: "Freud ... constantly appeals to writing, the spatial synopsis of the pictogram, of the rebus, of the hieroglyph, to nonphonetic writing in general." 16 His interest in Freud set Derrida in tune with an entire generation fascinated by psychoanalysis. At the same time, he could shore up philosophy against too many potential disaffections and conversions.

Derrida's implicit Freudianism led him to Lacan. While their theoretical proximity might have augured positive relations, it led to an extremely violent dialogue. Doubtless their proximity was too great; the results were a fratricidal combat. "I know that at one point Lacan had a somewhat paternal relationship with Derrida. He once said, 'I'm keeping an eye on him,' which meant that he was interested in his work, but in a paternalistic way." 17 It would seem that a purely anecdotal and personal incident contributed to the eruption of an argument between the two, but this was doubtless also due to the confrontation between two hegemonic ambitions. Derrida in philosophy and Lacan in psychoanalysis were both implicitly following a disciplinary logic and combating institutional power. Each also tried to ensure that his renewed discipline reigned over others. His imperialist, annexationist ambitions meant that Lacan, who presented the analytic discourse as the crown of the four possible discourses, kept Derrida under close surveillance. Derrida reciprocated. He had no intention of paying allegiance to Lacan.

Their confrontation, therefore, could only be brutal. For Derrida, the work of deconstruction did not stop at the doors of the unconscious, and Freud and Marx were only so many moments, albeit important ones, in Western metaphysics. "There was an obvious incompatibility between these two terrible wills. Each had a terrifying will to power." 18

Initially subtle hostilities broke out in 1971 on the occasion of an interview Derrida gave Jean-Louis Houdebine and Guy Scarpetta for the review Promise.t? In a long note, Derrida evoked the absence of any reference to Lacan in his past work, complained about Lacan's many aggressions and reappropriations of his work. He went on to criticize Lacanian positions, whose limits he had already seen as early as 1965 when writing Of Grammatology. "Assured of the importance of this problematic in the field of psychoanalysis, I will point out a certain number of major themes that kept it within the critical questions that I was in the process of formulating, and inside the logo-centric, that is phonologistic, field that I undertook to delimit and to shake. "20 Not only did Derrida reduce Lacan's contribution to a simple regional continent of knowledge, but he also raised a certain number of radical critiques that sought to present Lacan's contribution as merely an apparent contribution, as something to be deconstructed.

Derrida's criticisms were organized around four general issues: Lacan was the prisoner of a telos of speech identified with truth; he had uncritically adopted a Hegelian and Heideggerian perspective;

he had been light-handed in his use of Saussurean linguistics by carelessly adopting Saussure's phonologism; and Lacan had not really addressed the issue of writing in Freud, despite his positive return to him. Moreover, Derrida saw Lacan's attention to the signifier as the sign of a new metaphysics that did not dare to call itself such. Finally, he considered that Lacan's style was in the service of "the art of avoidance."21

But Derrida did not stop there. Six months later, he repeated his criticism during the conference at Johns Hopkins by taking up Lacan's proposed reading of Edgar Allan Poe's novella in his Seminar on the Purloined Letter, a speech that was later published in Poetique.22 Der-rida acknowledged that Lacan had taken an important step forward with his criticism of semanticism, considering the letter meaningless and that its author had no control over it. In short, the only important thing to consider was the letter's circulation. "Lacan is therefore attentive to the letter, which is to say to the materiality of the signifier. "23 But if Lacan made it possible to go beyond the referent and the subject, he did not take things to their ultimate conclusion because he brought us back "toward the truth that is not lost. He brought the letter back, showed that the letter returns to its particular space thanks to its particular path. "24 The letter's fate therefore would lead it to its place. And so, despite what he said, Lacan was in fact arguing for a hermeneutics whose ultimate signifieds were the sites of truth and femininity. This story of the purloined letter would thus ultimately lead to the unveiled truth of Marie Bonaparte as the bearer of Freud's work in its letter. As the repository of Freud's authority, she betrayed the letter of his teaching. "The fiction reveals the truth: the manifestation that illustrates itself while hiding itself. "25

Unveiling truth remained tied to the power of the verb and Lacan, in Derrida's view, remained a prisoner of the phonologism he denounced. There was "a structural complicity in Lacan between the pattern of the the veil and the voice, between truth and phonocentrism, phallocentrism, and logocentrism.Ve Behind these theoretical accusations lay Lacan's claim to represent a discourse that would put an end to the philosophy that Derrida targeted. There was no attempt at renewal, therefore, that escaped coming under the deconstructive gun, and the most fruitful disciplines of the period, ethnology and psychoanalysis, both of which relied on the linguistic model, were subject to the critique of deconstruction, which remained the master of the game.

Dissolving the Subject

For Derrida, writing did not belong to or depend on a context; it was independent of subjectivity. Writing traces were purely anonymous and no pragmatic analysis could account for them. Like Foucault and Lacan, Derrida also changed his first name. jackie became jacques at the price of a certain number of connotations from his childhood in the Jewish community in Algeria. But for Derrida, the I, the conscious model, was meaningless and decentering made the positions of structuralism in this area more radical. This position led Derrida to take issue with Anglo-Saxon analytical philosophy.

In Montreal in August 1971, at the International Congress of Francophone Philosophical Societies on the subject of communication, Derrida gave a talk entitled "Signature, event, context," later known simply as SEC, which he published in Marges.17 At the end of this text, Derrida began to debate John Austin's positions on performatives (a performative performs the action it describes: I promise is both an utterance and the act accomplishing the utterance). Derrida insisted on the limits of a theory of linguistic action that could not restore the Freudian slips, misunderstandings, and unspoken aspect of communication. He referred to the absence of the other in writing: "A written sign goes forth in the absence of its receiver. "28 The condition of its legibility was neither the other's presence nor any specific communication but the iterability of what was written. Far from expressing a context, writing was defined as an act of rupture. "This strength of rupture comes from the spacing that constitutes the written sign: it separates other elements of the internal contextual chain. "29 Derrida was interested in the objection of analytical philosophy and in the case of performatives that, as Austin saw it, could not be detached from their referent, unlike constative utterances. He answered that the utterance could only be understood if it answered a code, if it could be iterated. He argued for its autonomy with respect to the specific referential context of ordinary speech. The transparency of meaning was therefore a complete error, according to Derrida, in the case of performatives or constatives.

As a result of all iterability differing and differentiating, Derrida claimed that there was a "nonpresent remainder,"30 since nothing proved that the meaning of a language act was the same when used a second time, for either reader or speaker, whose intention was never entirely adapted to his utterance. John R. Searle, to the contrary, considered that the flexibility of concepts came from their intrinsic properties, which made it possible to perceive their mobility in the particular situations of daily language. Searle only learned of Derrida's text in 1977 when it was published in English in the review Glyph. He decided to defend the principles of Austin's theory, as well as his own theory of illocution: "To defend in particular the pertinence and interest of the fundamental distinction between serious and fictional uses of language, but also to establish the exact meaning and import of such notions as intentionality, repeatability, meaning, the success or failure of an illocutionary act, etc."31 In his answer to Derrida, the author of Speech Acts32 did not contest iterability as a condition of communication, but did not consider that it conflicted with intentionality; indeed, iterability was the presupposition of intentionality. We can easily understand the stakes of the debate. From Derrida's point of view, the issue was to assure the mobility of the game of signification for this or that subjectivity or intentionality so as to allow the infinite chain of repetitions in which "the individual withdraws in order to yield to the universality of the system." 33 Iterability did not operate in any observable way within ordinary discourse but eluded empirical observation and was located at a metalevel, the condition of possibility of discourse.

Loyal to his habitus of cutting off heads, Derrida never stopped trying to demonstrate the inanity of the demonstrations of Saussure, Foucault, Levi-Strauss and Lacan-and little appreciated it if anyone dared to contradict his theses. In 1977, he answered Searle's reply in a particularly bitter polemic on speech acts, during which he punned on his adversary's name, calling Searle a SARL.34 "Poor Searle never recovered. He was very humiliated by this SARL. We have to say that Derrida's irony is rather unusual in intellectual debates in the United States."35 Although this might appear to be little more than anecdotal, it was in fact characteristic of Derrida's identification with philosophy as the reigning discipline that justified his pulling out all stops and hitting below the belt.

Searle had raised a number of objections, nevertheless, that warranted discussion. Several arguments were put forth: that iterability was not just a privilege of writing; that the rupture that seemed to belong to writing between the utterance and its receiver had no relationship with quotability, and, finally, that the fact that writing could be cut off from its author in no way excluded the idea of intentionality. joelle Proust pointed out that Searle and Derrida could never have come to a consensus since Searle's underlying assumptions leaned toward confrontation whereas Derrida's assumptions tended systematically to avoid it: "The second type of procedure characterizing decon-struction involves putting the very nature of what is targeted in the

exchange into question____If we don't safeguard the independence of

the logic, don't we lose the very grounds for a possible consensus?"36 Beyond the form, there were the historical roots underpinning this polemic. The difference in analytical and continental traditions goes back to their divergent sources: Saussure and Frege. Analytical philosophy belongs to the Austro-German tradition and is generally considered to descend from Frege. On the one hand, Saussure ignored the issue of reference in order to raise the question of the scientificity of linguistics; on the other hand, Frege popularized the distinction between meaning and reference, between the meaning of an expression, which is a certain way of accounting for the reference, and the object to which the expression refers. From this perspective, analytical philosophy was always concerned with distinguishing these two levels without losing sight of the problem of reference. By basing itself on Saussure's positions and broadening its own positions beyond linguistics, structuralism defined itself by eliminating this concern, arguing that language referred to nothing other than itself. Frege's analysis of language focused on the level of a thinking of language and of language propositions, and held that only a concrete proposition allowed any real victory in the language game. In this respect, "Derrida's view of the situation through Saussure is pre-Fregian. It is never a question of anything other than words and their meaning. He has no veritable theory of propositions."37 Even if he changed the structuralist perspective by introducing temporality into the structure, Derrida's Saus-sureanism placed him clearly in the structuralist line, which he intended to deconstruct.


Benveniste: The French Exception

After its high point in 1966, structuralism underwent a long and progressive crisis. Generativism took up where structuralism had left off, Derrida's work met with success, and a branch of linguistics concerned with enunciation theory, or human utterances, repressed until that point, was on the rise. Emile Benveniste played a major but rather underground role in this, until 1968. He was an innovator, but from within. Despite his universally recognized notoriety, however, Ben-veniste was initially preaching in the wilderness in an era that had eliminated the speaking subject from language. A Sephardic Jew, born in Aleppo, Émile Benveniste had been destined for a religious career; his father had sent him to a rabbinic school in Marseilles. However, Sylvain Levi, a well-known Indianist at the College de France, recognized Benveniste's exceptional talent and brought him to meet Antoine Meillet, a disciple of Saussure. Ben-veniste therefore began his training as a linguist, following in Antoine Meillet's double tracks as a comparatist and a Saussurean. After an indirect and rather marginal career, he was invited to join the College de France in 1937. And with him, structural linguistics made its way to the heights of scholarly legitimation. When Levi-Strauss appealed to linguistic structuralism to lend support to his anthropological project in 1960, he called upon Benveniste to codirect the review L'Homme.

His position as professor at the College de France did not, however, allow Benveniste to make his ideas widely known. The margin-ality of the institution for shaping the nation's elite, together with the technical nature of linguistics, confined him in splendid isolation. "There were very few people in his courses, a dozen perhaps. It was only after the publication of General Linguistics) in 1966 that there were more, around twenty-five. Benveniste was very myopic and saw no one when he entered the hall. He went straight to his desk and spoke with tremendous aesthetic talent, improvising from his notes," recalled Tzvetan Todorov, who was privy to the master's confidences while nursing him after a debilitating stroke that left him a hemiplegic.2

Despite his isolation, Benveniste's renown was such that the most important linguists were drawn to his courses: Oswald Ducrot, Claude Hagege, jean-Claude Coquet, Marina Yaguello all came. But Benveniste was by nature a loner and remained closed off from others. "Benveniste was a private person, a poor communicator. I took his course at the College de France for three years. He was extremely shy and distanr."? Andre Martinet met him in New York before meeting him again in France, and confirmed this impression: "He came to my house in New York and we became pals. I was the only French linguist with whom Benveniste was friendly, insofar as he could be, because he was uptight.:"

Recognition from beyond the Linguistic Pale Benveniste was an Indo-European specialist and comparatist of a great number of ancient and modern languages. But thanks to his interest in utterances, in speech, the subject came back into the linguistic limelight. Benveniste defined a very distinct path from that taken by Anglo-Saxon pragmatic philosophy, while at the same time engendering a debate with it. "Personally, I owe this linguist more than any other. He was absolutely essential in demonstrating that while the linguistic system remains a system, it has to take utterances into account."5 Particularly precocious in his concern for the speaking subject, Benveniste emphasized as early as 1946 what he considered to be universals. Unlike certain researchers, such as Ramstedt, who worked on Korean, he argued for the indissociability and universality of the relationship between person and verb, regardless of the language: "We seem to know no language in which verbs or verbal forms do not distinguish for person in one way or another."6

Structuralism avoided logic and analytical philosophy; Benveniste, to the contrary, engaged in a dialogue with them. Ten years after his article on verbs, Benveniste linked his analyses to the pragmatic project of Charles Morris. "The utterance containing'!' belongs to this level or type of language that Charles Morris calls pragmatic, including not only signs but those who use them."? However, Morris had worked with Rudolf Carnap and had tried to use pragmatics to supply the missing link in the general science of signs, which already included a syntax in logic and semantics, but which did not define the relationship between signs and their interpreters. "The problem raised by Charles Morris just after the war was quite clear. It was a question of manipulating crowds with signs and from there, of constructing a philosophical theory of action." 8

In I 956, Lacan needed an important linguist in his camp and solicited a piece from Benveniste for the first issue of his review La Psychanalyse. Benveniste, who appealed to Lacan because of his interest in the subject, wrote an article on the function of language in Freud that helped Lacan substantiate his thesis of the unconscious being structured like a language. "Psychoanalysis seems to distinguish itself from all other disciplines primarily in this: the analyst works on what the subject tells him."? Benveniste of course criticized the analogy Freud drew between the way dreams operate, indifferent to contradiction, and the way in which, according to Karl Abel, the most ancient languages ostensibly function. Abel's etymological speculations were without basis, as far as Benveniste was concerned, for he considered that all languages were systems that could not function without this basic principle of contradiction. But this objection to Freud's sources was designed to better emphasize Lacan's ahistoricity and his preference for rhetorical figures and tropes. "The unconscious uses a veritable rhetoric that, like style, has its figures. The old catalog of tropes gives us an appropriate inventory of the two registers of expression." io The dialogue with psychoanalysis clearly offered Benveniste a means to lend weight to his positions taking utterance into account, and in I958 he wrote an article in the Journal de psychologie to once again support Lacan's positions. "It is in language that man constitutes himself as a subject because only language establishes the concept of the ego, in the reality of being. "ii

Unlike the generally held view of the speaking subject, which structuralism had eliminated, Benveniste proposed a distinction between the speaking subject and the subject of what is being spoken. It was only much later that his positions were adopted by linguists. "We might say that before 1970, French linguistics was only slightly or not at all familiar with the utterance as a theoretical whole for which Benveniste was arguing."12

The encounter between Lacan and Benveniste was not fortuitous. Both wanted to free their respective fields from a dependency on history-Freudian philogeneticism for Lacan and historical philology for Benveniste-and both were interested in establishing the scien-tificity of their discourse. When Benveniste presented the historical development of linguistics, he described a succession of three ages: the philosophical age of Greek thinkers on language; the historical age, from the nineteenth century onward, inaugurated by the discovery of Sanskrit; and finally, the structuralist period of the twentieth century based on which "the positive notion of a linguistic fact is replaced by the idea of relationship." 13 This new era was contemporary with social complexification and both Benveniste and Lacan, whose Real/SymboliclImaginary trilogy was to lead to the dominance of the Symbolic, considered that it opened up the vast cultural field of symbolic phenomena. Benveniste did not get the reception from linguists he had hoped for, and therefore, in order to alleviate his isolation, and to use the recognition he enjoyed among philosophers and psycho-analyts, he adopted a strategy of reaching beyond his own discipline and colleagues. He wrote and lectured actively, broadening the audience for his arguments on the relationship between the subject and language.

In addition to the article he contributed to the first issue of La Psychanalyse, he codirected L'Homme beginning in 1960, and he wrote a piece for the first issue of Les Etudes philosophiques in 1963 in which he presented the theses of analytical philosophy at a time when they were being carefully ignored, particularly by linguistics. "Philosophical interpretations of language generally inspire the linguist with a certain fear."14 This article was published after the 1962 colloquium at Royaumont on analytical philosophy which had not really drawn many linguists. Benveniste had given a paper in which he discussed John 1. Austin's work on performatives and his distinction between performatives and constatives. He approved Austin's pragmatism and drew attention to the interest these arguments could have while at the same time recalling that since 1958, he himself had emphasized the subjective forms of linguistic utterance and the distinction that followed between a speech act and simple inforrnation.>

Benveniste's reflection on the subject wâs therefore not an external graft but followed its own rhythm, always more on philosophical grounds for want of interest among linguists. In 1965, he published another article in Diogene, a philosophical review, on the relationship between temporality and subjectivity: "Of the linguistic forms that reveal subjective experience, none is as rich as that those that express time."16 Benveniste discerned two notions of time: the infinite and linear physical time of the world, and chronic time, woven of events. Both temporalities had objective and subjective versions. Since chronic time eludes lived experience, what happens to linguistic time? "Linguistic time is distinguished by its organic tie to the exercise of speech."i7 It is therefore located both in a present that is reinvented each time as a new moment, and as an individual act. It necessarily refers to the subjectivity of the speaker, and to an intersubjectivity insofar as linguistic temporality should respond to the interlocutor's conditions of understanding. Linguistic temporality therefore refers necessarily to intersubjective exchange. "The time of discourse works like a factor of intersubjectivity." 18

It was only in 1970 that Benveniste saw his positions win over linguists. His article on utterances, published that same year in the important linguistic journal Langages, was the sign of this success.t? But the battle was not completely won. The subject was back for reasons that did not really have anything to do with linguistics, but rather with the effects on linguistics of May 1968, with the new questions that had suddenly been raised in the social sciences and that had made it possible for the subject to come back in through the window after having been tossed out the front door.

The Repressed Subject

Until that point, and despite Gallimard's publication of General Linguistics in 1966, Benveniste had been carefully ignored by other French linguists. Claudine Normand confirmed this through a comparative study that allowed her to make a true discovery. She compared notes taken in 1966-67 in Paul Ricoeur's course to notes taken during the same year in Jean Dubois's class. She could thus measure the importance given to Benveniste by a philosopher on the one hand, and by a linguist, Jean Dubois, on the other hand, both of whom were professors at Nanterre at the time.20 Paradoxically, philosophy students at Nanterre were aware of the issues raised by Benveniste thanks to Ricoeur, whereas those in Dubois's class on linguistics were not. Normand concluded that "the philosopher seems armed to understand better and more quickly appreciate the import of certain new linguistic theories than linguists themselves, who are too busy transforming their traditional or recent methods to want to see them change dramatically already. "21

In addition to this case study, her article in Langagest? demonstrated that the different articles written by linguists in the sixties made no mention of Benveniste as the initiator of enunciation theory. Despite his interest in Benveniste's work, Ducrot did not quote him in his What Is Structuralism? ("Structuralism in Linguistics'll.P At the time, julia Kristeva (Julia Joyau at that point) quoted Benveniste in her 1969 work Language, This Unknoumr" but only to support structuralist theses without making any mention of the idea of enunciation. Similarly, Jean Dubois and Luce Irigaray coauthored an article in 1966 entitled "The Verb and the Phrase" in issue 3 of Langages in which they considered the speaking subject, but completely ignored Benveniste.

It was not because Benveniste was unknown that he went unmen-tioned. Structural linguistics deliberately set up roadblocks barring access to the subject. This was the price for the split from psychologism, phenomenology, and hermeneutics, for everyone who embraced the structural paradigm. For both Greimas and Dubois, the subject had to be normalized since it parasitized the scientific object under construction, which was supposed to be "an obiectified or standardized language from which all potentially descriptive elements concerning its analysis have been eliminated. "25 Such an analysis was completely uninterested in everything that was of interest to analytical philosophy. For Benveniste and Ricoeur, all forms of dialogue and the different modalities of subjective expression were important. Normalizing language with Hjelmslev's model of formalization made it possible to construct canonical third person utterances and to eliminate all temporal criteria for a "then," a deliberately vague term that referred to a past as indefinable as it was distant. "This was exactly the reverse of Benveniste's positions. He felt that it was important to establish the positional field of the subject and therefore the I1HerefNow triad that forms the reference point for any speech."26 It was only after 1970 that this view gained any currency, having long been obliterated by structuralism.

This negation had been nourished by a misunderstanding of the contribution of analytical philosophy, and of the great logical philosophers of the turn of the century, including Gottlob Frege, Bertrand Russell, Rudolf Carnap, and Ludwig Wittgenstein, all of whom had been ignored in France, where the German, Nietzsche-Heideggerian tradition was favored. It is true that Gallimard published Wittgenstein's Tractactus logico-philosophicus in 1961, but it did not attract enormous attention except for a short book introducing Wittgenstein by Gilles Gaston-Granger, and later jacques Bouveresse's work, which severely criticized Louis Althusser for having shut French philosophy off from the influence of analytic philosophyP "One day on the way to Althusser's house for lunch, we met Bouveresse, and Althusser said to me, 'you see, Bouveresse no longer says hello because he criticized me for having stopped the French from becoming familiar with analytical philosophy.' It's true that we were unaware of it for a long time. "28

The Vienna Circle at the time and those who gravitated around it, mistakenly called the Anglo-Saxon School, were baptized neo-positivists. This was enough to disqualify them. In the early years of the twentieth century, the interest in the philosophy of language had been left to psychologists whose expertise was quickly considered to have been definitively left behind by the partisans of structuralism.

Then, in the mid-sixties, at the height of structuralism's popularity, Paul Ricoeur became interested and wanted to integrate the philosophy of language into hermeneutics. So Ricoeur became the targeted enemy, particularly for the Althussero-Lacanians, who, in 1965, responded with a particularly virulent piece in Les Temps modernes. In an article by Michel Tort on Ricoeur's work on Freud.i? Tort characterized Paul Ricoeur's enterprise as having the appearance of a simple pedagogical booklet, a manual for the neophyte Freudian, but that underhandedly applied to Freud the implicit treatment of external categories borrowed from hermeneutics. Hermeneutics was criticized in this polemic as running counter to the critical and epistemological concerns of the period: "Ricoeur's phenomenological epistemology is nothing more than a rationalization of an ethical-religious scruple."30 Tort presented hermeneutics as an antiscience, a kind of phrenology of symbols whose only goal was the "sly denial of Freudian thinking. "31 Michel Tort rejected any attempt at an archaeology of the subject, considering that it could only lead to an imaginary speleology limited to "exploring the abyss of its own misunderstanding of the subject,"32 for the decentering of the subject that Freud demonstrated leads to the suppression of an organizing center. Yet Benveniste was closer to Ricoeur than to the Althussero-Lacanians when he conceived the unconscious symbolic as infra- and supralinguistic.

Several other reasons also help to explain the French resistance to the concerns of analytical philosophy. There was the radicality of the structuralist break which based its identity on holding at bay all current definitions of the subject, whether in philosophy with phenomenology or in literary history, with the psychologist vogue of the time. There was also a posthumous fascination with German philosophy. Moreover, work on logic in the French university had always remained very marginal, perhaps for contingent historical reasons since, as Canguilhem had said, French logicians had met tragic fates: Jean Cavailles, a member of the Resistance, was shot by the Germans during the war; Jacques Herbrand had died in a mountain accident on July 27, 1931. Beyond the deaths of the potential leaders of a French logical school, philosophical roots could also account for the different paths taken by Anglo-Saxon countries. "It goes back to the position of English mathematicians with respect to the status of the symbolic. There is a configuration that makes it possible to understand why analytical philosophy developed in England, and it is tied to taking a position on the nature of mathematical objects and on the existence of these mathematical objects, which is quasi-ontological."33 By their metaphysical presuppositions, English mathematicians would have favored an idealist conception of a subject that existed in itself with a simple, quasi-instrumental relationship to language. Since French philosophers wanted to kill Western metaphysics, they were not well disposed toward accepting such an approach.

Benveniste's Heirs

Under such conditions, Benveniste's efforts to include the subject among the theoretical concerns of linguists encountered a certain number of difficulties. But several of his disciples continued after him and, in a more favorable context, were more successful at introducing analytical philosophy. One was Oswald Ducrot, the author of the linguistic part of the anthology What Is Structuralism? published by Seuil in 1968. Ducrot's discovery of analytic philosophy was characteristic of the state of ignorance and disdain of the period. A philosopher by training, Ducrot discovered structuralism while preparing to enter the École des Hautes Etudes Commerciales. "I was also very interested in mathematics and was trying to do something on the philosophy of mathematics. From there I drifted toward that area of mathematics that is the easiest for a philosopher: logic." 34 Oswald Ducrot then worked on formal grammars, which were used a good deal in Chomskyan grammar.

In 1963, Ducrot joined the CNRS to do a thesis on the history of philosophy, on Descartes. Like all CNRS researchers, he read the journals carefully, and it was while doing thesis research and simply compiling articles that he made an important discovery. "Those who had come last had the least interesting journals, journals that all the French philosophers were rejecting, so I found myself in charge of English journals on the philosophy of language. They were completely absorbing and led me toward the philosophy of language rather than toward structuralism." 35 Much later, at the beginning of the seventies, Ducrot introduced pragmatics in France. For him, he was abandoning nothing, but rather was giving a new dimension to structuralism, as his introduction to the French edition of John R. Searle's Speech Acts in 1972 amply demonstrated.

His introduction paid homage to Saussure for having dissociated the linguistic object of study from language, which could not be directly studied, and which contrasted both with the ability to speak and with speech itself. But he disagreed with Saussure's elimination of speech from the realm of scientific analysis. According to Ducrot, if the path leading from Saussure to Austin culminated in a new area, that of performatives, a certain continuity ran between them making it possible to add a specific, new sector to the basic structuralist postulate that would only have a completely marginal position in language. "The value of utterances, even if it does question the Saussurean thesis identifying linguistic activity with individual initiative, does not prevent us from maintaining a good part of this thesis." 36

In considering the linguistic order to be irreducible to any other level of reality and requiring a logic sui generis to explain it, Ducrot was faithful to the Saussurean tradition. His analysis was fundamentally structural; it did not work from any empirical given but rather from the semantic unity that he called the meaning of the utterance.37 Ducrot adopted structuralism's hermetic notion of language, and considered that the seduction of the philosophy of language was a reactivation of Piatonism, "in other words, the idea that, before discussing philosophical problems, we have to agree on what we mean by the words we use. That is what I found completely engrossing and Platonic in Austin."38

There were two currents within the philosophy of language. There was the logical school that began with Rudolf Carnap, who, in his 1928 The Logical Structure of the World, sought to go beyond a simple critique of language in order to achieve a more perfect logic and demonstrate that a system of protocol utterances existed that could be divided in order to establish a basic scientific corpus. Everything that did not conform to the rules establishing this system of utterances was banished to the realm of nonsense, including metaphysical propositions. The entirely semantic and formalist elimination of metaphysics would therefore make it possible to articulate the elements into combinations and compositions giving a satisfactory picture of reality. This logical school was not the branch of the philosophy of language that influenced Oswald Ducrot, but rather the branch that remained inside language, represented by John Austin and Searle.

What later made me take my distance from them was that they saw the study of language as really offering a solution to philosophical problems. I, on the other hand, thought this was less and less true. Moreover, they thought that by studying the meaning of words one could find satisfying concepts for describing ordinary language, which I did not find reasonable since I don't see why language would be the best metalanguage for its own description."

Ducrot also considered that Searle and Austin's notion of the subject, which he saw as a complex, plural entity, was too simplistic.

So, although Oswald Ducrot was the best person to introduce pragmatics in France, his perspective had a structuralist twist and went back to Benveniste, who, since the seventies, influenced a whole current of increasingly numerous enunciation theorists. This was the perspective of Catherine Kerbrat-Orecchioni, who sees herself as one of Benveniste's heirs. Her book on all the indices of subjectivity in language, beyond deictics, includes subjective verbs and lexical forms of subjectivity.w This problem of the subject's place in language and language acts produced an entire school of pragmatics in France, which included Francis Jacques, jean-Claude Pariente, and Francois Reca-nati," for whom pragmatics involved studying "the use of language in discourse and the specific marks that, in language, attest to its discursive vocation. "42

Antoine Culioli and his school, also to be included in this group, were similarly concerned with constructing a theory of utterances based on deep universal patterns. These "mechanisms of production," a complete formal apparatus of utterances, were also part of Ben-veniste's legacy. Culioli, professor in the Department of Linguistic Research at Paris VII, influenced an entire school, including Marina Yaguello, but his work has become so sophisticated that it is unreadable for the neophyte and painful for the specialist. This search for deep structures à la Chomsky supposed the existence of what Culioli called lexis, and that led to predicative relationships. "There are enunciation operations that let deep structures come to the surface in certain languages where the enunciation operations receive what we call a grammaticalization. "43 Unlike the generativist approach, which goes from surface to deep structures on the basis of the native speaker's intuition about the grammaticality of the sentences, letting the speaker differentiate between possible and impossible locutions, Antoine Culi-oli postulated a certain number of operations (modalization, aspect, nominal and verbal determination) that let the speaker organize the verbal relationship and lead it to the surface utterance. "Culioli does not see the utterance as belonging to a corpus but as the discursive confirmation of these operations, which he postulates abstractly. There are utterances that can take very different forms but that still refer back to the same enunciative operations."44

Culioli reflected structuralism's early goals of formal translation, the quest for regularity and the universal based on invariables, the concern for going beyond the individual and the particular in a new field, which had initially been ignored because of the restrictive definition of language that Saussure had given, eliminating speech. "There is no isolated utterance; every utterance is one of many, selected by the speaker from among the many possible equivalents. In a word, every utterance belongs to a family of paraphrastic transformations. On the other hand, no utterance is unmodulated-in other words, is not unique. "45 Claude Hagege, who succeeded Emile Benveniste at the College de France, adopted a less formal perspective, closer in spirit to Benveniste, whose work he also helped bring out of the shadows to which it had been condemned during the era of structural splendor. Did the progressive success of the question of the subject in language contribute to the decline of the structural paradigm or did it offer a second wind with a new field of study? According to Marina Yaguello, it was just as legitimate to consider pragmatics as an adjacent or connected area:

We can consider that linguistics is one, and that there is both a theory of speech acts and a theory of language and that the two are articulated together. But we can just as well argue that we can address speech acts and therefore the illocutionary value of utterances (or when the utterance itself constitutes an act) without at the same time addressing the question of knowing how the utterances are made.46


Kristeva and Barthes Reborn

Julia Kristeva arrived in France on Christmas I965 and quickly began to shake up structural semiotics. Soon after her arrival, having enrolled in Roland Barthes's seminar, she gave a presentation on the important changes in the structuralist paradigm of the second half of the sixties, introducing a new vision, drawn from Russian postformalism and based on the work of Mikhail Bakhtin. Bakhtin's work was unknown in France at the time, but she later introduced it in her preface to Seuil's edition of his works.' The choice was not gratuitous since Kristeva wanted to inject a historical dimension into the structuralist approach and get beyond textual closure in order to make literary texts more broadly comprehensible. Her presentation came at a particularly opportune moment. In I966, structuralism was at its zenith and a certain number of attempts were in the making to weaken its monopoly. Derrida, Chomsky, and Benveniste were all beginning to seriously challenge the ambitions of the first period. Kristeva's presentation was initially published in Critique and when it was printed in Semiotike, Research for a Semanalysis- in I969, it reached a broader audience.

One listener who was particularly seduced by Kristeva's presentation was none other than Roland Barthes himself. Barthes later used Kristeva's ideas, which were new for him, to make a radical shift in his own work. "Bakhtin's approach was interesting because he saw the literary text, be it that of Rabelais or of Dostoyevsky, as polyphonic within the text itself."3

Mikhail Bakhtin

Bakhtin saw the dialogue of literary texts among themselves as fundamental. Saturated by their literary predecessors, they engaged in a polyphonic dialogue that decentered their initial structure. Bakhtin therefore broadened literary-critical study to include the historical fabric in which literary texts were woven. His approach obviously contrasted from the outset with the structuralist position claiming that a literary text was hermetic and that this closure made the work's structure accessible. Bakhtin compared Rabelais's work with the popular culture of the Middle Ages and the Renaissance. Just as Lucien Febvre had already argued that Rabelais was not radically new when read in his context, thereby demonstrating that he could not have been an atheist, Bakhtin deciphered the enigma by looking at Ra-belais's work through its popular sources and categories. He felt that Rabelais drew his inspiration primarily from the comic grotesque of the Carnival, of life turned upside down and of the parody of everyday life: "This was Rabelais's language."4 Bakhtin condemned the mistaken interpretations of Rabelais as the poet of the flesh and of gluttony (Victor Hugo) and those who saw in him the expression of the bourgeois interest in the economic individual; Rabelais's style could only be understood as the translation of a popular, comic culture that he called "grotesque realism."5 Beyond the comic effect lay a whole cosmogony in Rabelais, and his focus on orifices, protuberances, and outgrowths corresponded to corporeal parts that put the individual into contact with the external world.

Kristeva had immediately understood structuralism's historical limitations and intended to palliate these shortcomings with Bakhtin, and lend "dynamism to structuralism."6 The dialogue between texts that she considered fundamental could serve to address the subject, the second element that structuralism had repressed, and reintroduce it as part of the theme of intersubjectivity, much in the manner of Benveniste, But in 1966, things had not yet evolved that far and Kristeva avoided the issue of the subject, preferring to use a new notion that was immediately successful: intertextuality, "It was at that point that I created the gadget called intertextuality."? Even today, Kristeva is invited to lecture on intertextuality in the United States and to write articles to further elucidate and develop this notion.

Mikhail Bakhtin had postulated a translinguistics and had used Rabelais, Swift, and Dostoyevksy to develop his arguments of a polyphonic thread. Kristeva added modern authors, including Joyce, Proust, and Kafka, whom she considered useful for a similar approach-pointing out, however, that representative and fictional dialogue was, with these twentieth-century authors, supplanted by an interior dialogue.

In addition to this intertextual perspective, Kristeva introduced dialogics (criticism as dialogue, the encounter between two voices). More than she realized at the time, this helped to seriously upset structuralism, even if dialogics was still considered to be immanent to structure: "Dialogism is coextensive with the deep structures of discourse."8 It would be a mistake to see in this the return of the classical subject or the notion of the author in dialogics. Kristeva was quite careful to dissolve this notion within the narrative system itself, and, true to the structuralist perspective in this, she felt that the author "becomes anonymous, an absence, a blank, in order to allow the structure to exist as such."? So the author was nothing other than the expression of this vacuum, and in his place was the intertextual dialogue in which he dissolved as he appeared.

Kristeva distinguished two types of narrative: the monologic story that includes descriptive, representative, historical, and scientific modes in which "the subject assumes the role of i (God), and submits to this role by the same gesture, "10 and the dialogic story, which expresses itself particularly through the form of the carnival, the Menip-pean satire and the polyphonic modern novel. In order to make the modernity of dialogism clear, Kristeva saw in it not only a new method for literary analysis that was richer than binarism, but also "the basis for the intellectual structure of our period."l1 With dialogism, the Hegelian dialectic could be reconsidered, it could be absorbed in a concept that was not an opposition that implied getting beyond, but a harmony based on a simple disruption that made it possible to concentrate on the work of transformation. "Dialogism puts philosophical problems in language, and more precisely, in language as a correlationship of texts, like writing-reading."12 A person of letters enjoyed a certain hegemony and could analyze a range of fields, including philosophy. By opening the text up onto an environment that was more than referential and contextual and that included the surrounding literary universe of prior, contemporary, or future texts, new analytical perspectives became possible. The contemporary writer in particular could dialectize his position as author-reader differently, and even include his reading within his own writing.

A Turning Point for Roland Barthes

Roland Barthes was intrigued by all forms of renewal, and by youth in general. And with Kristeva's presentation, a bell tolled for the scientis-tic ambitions so carefully laid out in Elements of Semiology and in Critique et Vérité. Barthes was engaged in a true exchange with his students and knew how to give and take. Extremely encouraging, he was always attentive to and encouraging of student presentations. "Roland played a very important role for me. He was the only person I knew who was able to read others, and, for a professor, this was enormously important because in general professors read themselves."13

Barthes's own work also reflected his interest in intertextuality. In 1970, he published 5/Z, the trace of his 1968 and 1969 seminars at the École Pratique des Hautes Etudes, inspired by Kristeva's presentation of 1966. This was a major turning point; Barthes deconstructed his own conceptual grid and gave freer rein to his literary intuition. After the discourse on method came writing, the expression of sensibilities and infinite, undefinable meaning. "Like Sollers, Roland was first and foremost a man of letters. You might say that he used methods, for, as Buddha says, 'If you want to cross the river, take a pile of bits of wood, make yourself a raft, and then throw it back into the river."'14 From the beginning of 5/Z, Barthes took his distance with what he later considered to be the illusory reduction of all the stories "to a single structure. "15 Not only did he consider this structuralist ambition overblown, but he also considered structuralism to be tainted with a questionable perspective because this Sisyphean effort led to the negation of differences between texts.

Given this new concern to make difference the goal rather than the means of the analysis as it was being used in phonological bina-rism at the time, we can easily see Kristeva's influence on Barthes and the entire Tel Quel group. Derrida in particular had adopted Kristeva's notion of intertextuality. Even before 5/Z came out, Barthes remarked in an interview with Raymond Bellour: "We can speak about intertextuality with respect to literature, but not of intersubjectiv-ity."16 These were Kristeva's very terms. In 1970, Barthes confided to Bellour the names of those to whom he was indebted but who went intentionally unmentioned in S/Z to better suggest that the whole of the work was a quote. "1 removed the name of my creditors: Lacan, Julia Kristeva, Sollers, Derrida, Deleuze, and Serres, among others." 17

Beginning with 5/Z, the whole of the deconstructionist problematic was at work in Barthes in his concern to pluralize and exacerbate differences, to set them into play outside the signified in an infinity in which they dissolved and gave way to the "whiteness of writing." Derrida's concerns were quite clear in this new Barthesian discourse that marked a turning point. Barthes addressed the question of the Saussurean sign: "Now we must take our battle even further and try to crack the very idea of the sign, not signs with signifiers and signi-fieds. We might call this operation semioclastic. "18 Behind his determination to fissure the foundations of Western discourse, of course, lay Derrida's deconstruction of Western logocentrism. The goals were not altogether synonymous, however, even if in both cases writing was the issue. For Barthes, the question was literature, whereas for Derrida it was philosophy. And yet, when Barthes said that "the writing of the writer depends essentially on a criterion of indeterminability,"19 we cannot but recall Derrida's undecidables, which were supposed to deconstruct Western metaphysics. At the end of the sixties, in the increasingly frequent interviews of 1970-71 explaining his reconversion, Barthes recognized quite explicitly that there had been a shift and a break.

His explanation confirmed his extreme sensitivity to his environment. "The causes for this change (for it is more a matter of mutation than of evolution) are to be sought in recent French history-why not?-and then also in the intertextual, by which 1 mean the texts surrounding me, accompanying me, which precede and follow me and with which, of course, 1 communicate. "20 His allusions to May 1968 were clear, and the philosophical guarantee of Derridean deconstruc-tion let Barthes clearly express his wish to write Literature and to freely express his subjectivity and his difference, freed of codes and other formal systems. Describing his wishes for the seventies, Barthes expressed the desire to work within the signifier, or to write in what he called "the romanesque less the novel."21 That was what he began to do with 5/Z, which he considered to be very important in his personal itinerary thanks to the "formulators," the researchers surrounding him who "taught me things, who opened my eyes, who persuaded me."22The other reason for the shift came from the object of study itself.

5/Z, the first microanalysis Barthes had undertaken, focused on

Balzac's 1830 novella Sarrasine. Barthes described the interplay of five codes that allowed him to look at the internal plurality of Balzac's writing. He observed things differently here, following the text step by step and constantly setting writing against reading. Above all, he wanted to create a new form of writing/reading that was supposed to be the product of the notion of intertextuality. Kristeva's influence was quite clear: examining a process as it unfolded, and substituting structuration for structure: "To rediscover what Julia Kristeva calls a productivity. "23 Barthes perceived this productive horizon in the very unfolding of writing/reading, infinite and forever open. Balzac's text and its dissolution in contemporary languages and codes clearly expressed his desire for a limitless writing having nothing to do with the search for a system of single or multiple causes leading to a closed explanation of the text and to its definitive interpretation. "There is never an end to the text. Lansonians-" consider that the book stops with the author and his sources. Intertextuality makes authors anonymous and considers that the text goes on infinitely."25

For Barthes, the active/author, passive/reader relationship needed to be redefined by readers rewriting the written text, or a plural text allowing for many possible voices and paths. The five different codes used to allow the polyphony of Sarrasine to reverberate included the atemporal semic, cultural, and symbolic codes and the temporally irreversible hermeneutic and proairetic codes. The apparently rigorous method, drawn from a strict system of coding, radically broke with the first period of structuralism: "For the plural text, there can be no narrative structure, grammar, or logic to the story. "26

The ambitions for narrative structures laid out in 1966 in Communications were set to rest. There were no other interpretations except those at the level of multiple meanings, but no total, hermetic text. Intuitive sensitivity triumphed beneath the rigidity of codes that were used but were carefully hierarchized according to the principle of taste. They made no claim to scientificity. "There are good and bad codes. "27 The proairetic had its place on the axis of insignificance. At the other extreme, the symbolic code was entirely positive, including everything that seemed intuitively interesting to him. Although implicit, "This hierarchy flows as it were from itself. "28 At the top was the symbolic, which depended on the pure signifier, on nonlogic, and on the power of textual plurality to which Barthes aspired. This symbolic code was so important in analyzing Balzac's novella that Ray-rnond Bellour saw in it the sign of the use of an underlying matrix of production structuring the text.

Three symbolisms-gold, meaning, and sexuality-set the dynamics of the text into motion and referred respectively to Marx, Aristotle, and Freud. Balzac's novella unfolded during the Restoration. The author severely criticized the new bourgeoisie, which had just come to power thanks to its wealth. By contrast with the dignity of the nobility's gold, theirs lacked true roots, and was unconnected to the land. The second part of the story focused on Zambinella, a castrato for the mistaken love of whom the sculptor Sarrasine, who believed Zarnbi-nella to be a woman, was assassinated. By shifting the symbolic code, Barthes drew a parallel between the two parts of the story: the satire of the gold-owning parvenus without a past paralleled the theme of the castrato which referred to a woman who was not a woman.

This interpretation, like Barthes's previous interpretation of Racine, borrowed heavily from psychoanalysis, and more specifically from Lacan, who, together with Kristeva and Derrida, was one of the major inspirations for Barthes's analyses. "My use of psychoanalytic language is, like the use of any other idiolect, playful, citational."29 Barthes's work on the letter, which is set off by the book's title, S/Z, comes from the play of meanings that unfold in the impossible relationship between SarraZine and Zambinella. In the first place, Barthes pointed out that a French ear would expect to read Sarrazine, whereas the Z

gets tossed out. "Z is the letter of mutilation____It cuts, bars, creates a

zebra stripe. From a Balzacian point of view, this Z, like the one in Balzac's name, is the letter of deviation."30 Moreover, Z is the first letter in Zarnbinella, the castrato,

such that by this spelling mistake within the very heart of his name, at the center of his body, Sarrasine understands the Zambellian Z in its true nature, which is the wound of the lack. In addition, Sand Z graphically reverse one another, as if they were the same letter seen from opposite sides of the mirror. Sarrasine contemplates his own castration in Zambinella. "31

We can imagine Barthes's extreme pleasure in constructing his interpretation, which, by simply explaining the title S/Z, made it possible to take seriously the importance of the letter in the unconscious, according to Lacan, the importance of graphic writing and its repression by phonologisrns, as Derrida saw it, and finally, the reintroduction of the Saussurean bar reinterpreted by Lacan between Sand Z, a bar that formed the screen, the true censure, the wall of hallucination.

The Empty Sign: Japan

In 1970, Barthes also published Empire ofSigns,32 which reconfirmed his new tack. This book, in which Barthes freely described his Japan using fragments, offers a posttheoretical counterpoint to the pre-theoretical Mythologies. Barthes abandoned the conceptual adventure, and although he was extremely caustic and critical about the signs of daily life in the West, he looked at those in the Orient through Chimene's eyes. What fascinated him above all, and in this there was a continuity between his two periods, was that the Japan he discovered and wrote about was a Japan that had rid itself of all full meaning.

Barthes experienced the intense pleasure of fully entering, for the first time, into a signifier finally freed of any signified, a world of empty signs, emptied of meaning and of all the forms of rigidification known to the West, and which he had denounced in Mythologies. He did not abandon his critical perspective but used the Orient to protest Western values, indirectly. "Like many of us, I profoundly reject my civilization, to the point of nausea. This book expresses my absolute demand for a total alterity, which has become a necessity for me."33For want of the possibility of getting beyond Western reality by using its internal contradictions, Barthes rejected everything about the Western universe, contrasting it with a binary utopia. One finds here the structuralist theme of the closure of history and the progressive elimination of the referent and then of the signified. "In our Western world, in our culture, in our language and languages, we must engage in a battle to the death, a historical battle with the signified."34 The imaginary voyage to Japan to which Barthes invited his readers in 1970 reiterated his quest for the loss of meaning, known as satori in Zen, to let an infinite game of signifiers unfold. Everything was thus perceived, down to the slightest detail of daily life, as the illustration of this distance with respect to signs. Speech was empty and perception essentially graphic. Food, for example, was decentered: Japan has a veritable cult of raw food that is honored to the point of preparing food in front of the eater, in order to "consecrate by spectacle the death of what is being honored." 35 Entirely visual, composed of many fragments without any order of ingestion, Japanese food has no center but gives the eater free rein to use his chopsticks as he chooses.

Everything is fragmented and multiple in Barthes's Japan, in contrast to the West where everything is structured, focused, and ordered. This was also true for Japanese art; Barthes contrasted the Western tendency to transform an impression into a description based on a full subject with a subjectless haiku, which never describes but simply links signifiers without any demonstrative goal but with the trace of the pleasure of writing. Haiku serves no purpose and lends itself to no commentary. "It is that, it is thus, says the haiku, it is so ... but the haiku's flash illuminates, reveals nothing."36 Barthes was fascinated by what he had long repressed and that nonetheless constituted his veritable being, the freedom of the writer writing, the capacity to detach himself from any didactic discourse and to give intuition full rein. In 1970, Barthes therefore returned to literature after a detour via linguistics. His path was symptomatic of an entire generation of structuralists who adopted the discourse of the social sciences without fully renouncing the unavowed vocation of the writer. During the sixties, some of the important contemporary novels were "scientific" works.

Paragrams; or, The Veiled Return of the Subject Julia Kristeva used intertextuality to go beyond structuralist closure. But she also opened up a second new direction for research involving a dynamic subjectivity that was not that of the classical subject but that of the subject as Lacan understood it, the subject of desire. In 1966, thanks to Jean Starobinski's publication of some excerpts, Kris-teva discovered what was called the second Saussure, the Saussure of the anagrams. She immediately saw the correlation between this quest for the proper name underlying Saussure's apparent text and the analytic approach as Lacan formalized it.

What became clear to me was that their writer is sort of influenced by the game of phonemes and syllables. This power gives a kind of consonantal and phonic regularity; repetitions and alliterations that can be stabilized in a proper name and that would be the obsessional proper name to which the individual is unconsciously tiedfor sexual or morbidreasons.37

This research referred to another dynamic of the unconscious structure that Kristeva explored in a programmatic text, "For a Semiology of Paragrams, " in which she outlined a new science at a time when they were proliferating: grammatology in 1965 and paragrammatics in 1966. Taking Saussure's work on anagrams, Kristeva claimed that he had been mistaken in looking for a single word or a particular anagram each time, whereas an entire underlying thematic chain slipped beneath the apparent text. "There is this insistence on paralogies, of meaning other than the explicit meaning."38 Based on this, she proposed to read Mallarrne and Lautreamont differently. The paragram became a form of destruction of the writing of the other and of dissolving frozen meaning. "After destroying man, the paragram destroys the name." 39 This outline of a new direction for research bespoke the scientism of the period. Kristeva evaluated the reminiscences in the paragrammatic space on the basis of a solid alliance between semantics and mathematics. "The effort required to understand the logic of paragrams in an abstract way is the only means of getting beyond vulgar psychologism or sociologism. "40

But beyond the scientistic mask, Kristeva judiciously looked to the subject, which until then had been obliterated. The paragrarn-matic quest echoed the logic of the unconscious in everything that it had stored up, or "engrammed" like a signifier. This quest pointed to personal history, woven of souvenirs, reading, and different influences located at levels other than that of the language of communication, which limited, and by definition defined, the numbers of codes being used.

Taking this transversal path between linguistics and psychoanalysis, which Kristeva later called semanalysis, eventually led her to abandon literature and become a psychoanalyst. This type of reading already had the advantage of getting beyond neutrality toward subjectivity and setting up resonances with the literary critic's unconscious. This newly ascribed importance of subjectivity opened up the path toward literary writing and therefore gave Barthes the necessary scientific balast he needed to let his creative desires flourish. Kristeva, on the other hand, remained in the realm of science. Psychoanalysis provided the necessary conceptual grid for going further in her quest for the subject and in unveiling its mode of existence. "I felt slightly daunted for having put forward this personal subjectivity, particularly because French is a foreign language for me."41 So Kristeva stayed within a theorizing discourse longer than did Barthes.

In order to take both analytic paths into account, she later distinguished between semiotics and symbolic. Symbolic referred to the simple denotation of a coded exchange; semiotics opened onto a "secret continent of language in which the perfumes, colors, and sounds echo each other and all refer to a childhood experience and to the unconscious. "42 Defined this way, semiotics in fact took up the semana-lytical project defined in 1969, which had already proposed a criticism of the sign able to deobjectify its object and to conceive of it based on a shard offering "a vertical cut unlimited by either beginning or end. "43 Kristeva's semiotics drew on two important renewals that were under way-Marx by Althusser and Freud by Lacan. Yet she sought her legitimation at the very roots of Western culture, based on which we can understand its temporal density. "I am referring to the Timaeus, a text in which Plato speaks about a modality of signification that he calls the cbora, or, in other words, a receptacle."44 Plato considered this modality of common sense to precede the One, and ascribed to it natural connotations of a nourishing and mobile receptacle. Kristeva used this dialogue to examine prelanguage, preceding linguistic signs, linked more closely to the relationship between the future speaker and its mother. "I tried to propose that notion of a semi-otic Chora that goes back to a translinguistic and more archaic model of signification. "45 She differentiated herself in this from Derrida's absolute deconstruction, which had nonetheless been very important in her critique of the sign. Her interest in psychoanalytic discourse led her, to the contrary, to interpret and therefore to be attentive to the meaning revealed by psychoanalytic attention, to a truth, however provisional.

This receptivity to psychoanalysis and to subjectivity let Barthes free himself from a certain number of constraints. In 1971, he admitted that

the big problem for me, in any case, was to trick the signified, to trick the law, the father, and the repressed____Wherever it is possible to undertake a paragrammatic effort or a certain paragrammatic tracingofmy own text, I feel at ease. If I everreally had the opportunity to criticize my own work, I would center everything on "para-grammatism. "46

Barthes embodied the sensitivities and ambitions of his period's avant-garde, and he moved in a new direction, thanks in part to this reorientation suggested by Kristeva's 1966 work, toward intertextuality and paragrams. This shift can also be seen as the literary expression of Benveniste's questions about enunciation. "Linguists who have a theo-retical framework (Jakobson, Chomsky, and Benveniste) are men who raise the question of enunciation and not only the problem of the mes-sage."47Once linguistics raised this issue, it entered into contact with psychoanalytic thinking, with Lacan's work, as we have already seen in the case of Benveniste.

This shift was part of a new intellectual climate that gave priority to the search for the subject of desire in its different modes of expression. The same quest was taking place in literature: Philippe Sollers published Dramaw in 1965, reflecting on the use of pronouns, or, in other words, on signs in utterances. But beyond this propitious context in which the second Barthes was "born" there was above all an internal echo leading him to reembrace subjectivity and allowing his unacknowledged aspiration to flourish. Intuition gained ground as writing became freer and better able to openly express the pleasure it procured.


Durkheim Gets a Second Wind: Pierre Bourdieu

At the beginning of the century, sociology had barely emerged from philosophy. Durkheimians had only half succeeded in their emancipation and had entirely failed in their attempt to unify the social sciences around the concept of a social morphology. But they were able to take advantage of the postwar growth of the social sciences and to establish themselves more and more securely in the university, although their institutional success did not mask their failure in terms of scholarly legitimation. Although they could determine their own curriculum, philosophers and historians in particular, but also the younger disciplines such as anthropology, with their dearly defined ambition and rigor, disdained sociology as a minor discipline, which they relegated to a secondary role, a disciplinary refuge for empiricism, with limited and essentially instrumental goals.

Philosophers Challenged

When Pierre Bourdieu came to sociology, his theoretical goals, his hegemonic will, his way of problematizing sociology as an institution gave new life to Durkheimian ambitions. This second wind was possible because Bourdieu had assimilated the structuralist program-at least initially, for, like many others, he later took his distance from it. During the sixties, however, he proposed a structural-Durkheimian method that tended to consolidate Durkheimian positions in order to redynamize them and to reunify sociology, which was being fractured into many ideological families. "His structuralism was extraordinarily enriching; it was the great contemporary sociological oeuvre."!

Bourdieu, like Durkheim, challenged philosophers, although he never really abandoned the philosophical problematization or accepted a rupture between sociology and ethnology. "In order to escape relativity, even somewhat, we must absolutely give up any pretension to absolute knowledge and remove the crown from the philosopher king."2 Not that Bourdieu abandoned the philosophical debate, for within his own work he continued to dialogue with Kant, Heidegger, Wittgenstein, and Austin, which was altogether unique in sociology. "It seems to me that there was always an amorous attitude of rupture with philosophy."3 Bourdieu challenged philosophy on its own grounds, amassing all of sociology's statistical tools, methods, concepts, and verification procedures. He could therefore calculate the advantages of a position that was at once philosophical and scientific: "Sociology long ago abandoned the grand theories of social philosophy____So why not say that it's a science if it is one?"4

With Bourdieu, sociological analysis questioned the philosopher's position by establishing a correlation between the contents of philosophical discourse and philosophy's institutional position in the university. There was a great deal of work devoted to objectifying philosophical discourse by studying what validated and legitimated it in the very conditions of its pronouncement. Bourdieu considered sociology to be in a privileged position to do this, and therefore had to be heard in the chorus ofdiscourses about the social sciences. In this way, sociology offered a liberating perspective: "Sociology frees by freeing from the illusion of freedom."! Sociology made it possible to formulate the most extravagant ambitions in order to unify the social sciences, which Durkheim had hoped to do, around the structural-Durkheimian paradigm. In order to accomplish this, Bourdieu introduced structuralism into sociology-a delicate introduction of a paradigm that sought to unlock what was hidden, occulted, and unspoken in a discipline that valorized testimony, interviews, and statistics, or, in other words, the realm ofthe visible.

A Miraculous Recovery

Bourdieu's relationship to his object of study was one of radical denial. His vast system served to demonstrate the power ofreproductive structures, the weakness of mobility, the futility of events, and the ways in which an individual is defined by his or her roots. And yet Bourdieu is the living negation of his own theses; his personal itinerary belies institutional omnipotence and unbearable determinism. This contradiction in itself reveals a very particular role played out between Bourdieu and himself, a quasi-therapeutic confrontation between the individual at the height of scholarly legitimation and the man who has never hidden his roots and is increasingly uneasy in his academic and worldly success. "Curiously, my insertion in the social world should have been easier and easier but in fact it is becoming more and more difficult."6

Bourdieu recognizes himself as a miraculous escapee. Edgar Morin dubbed him-polemically-bourdivin. 7 All of his work can be seen in this transition from the religious (miraculous events) to the scientific (sociology) in order to attempt to rationalize a success that is as spectacular as it is unlikely, statistically speaking. "I am in a universe where 1 should not be. I should have been eliminated forty times over____In two hundred years at the College de France, about i percent of the members have been in my category,"! This is not self-flattery on Bourdieu's part, for he comes from a popular, rural, exploited, and marginal background.

Born in a small village in Southwest France, his father was a petty civil servant who had come into public service rather late in life, a significant promotion for someone who had been a sharecropper for more than thirty years. "My childhood was marked by the experience of social inequality, and of domination."? Besides distinguishing himself at school, Bourdieu also stood out for remaining faithful to his first revolt, contrary to the development of other prodigious children from the popular classes, who tended rather to become integrated in order to free themselves of their roots and to recognize the validity and naturalness of the criteria that allowed them to liberate themselves from their milieu. Bourdieu went to khâgne at Louis-le-Grand High School in Paris in 1950-51. Later, he was admitted to the École Normale Superieure (ENS) on the rue d'Ulrn. This was not a happy period for him, even if this philosophical training attracted considerable recognition. "I felt paralyzed by a sort of infamy.... 1 felt terrifically bad."!" He was cut off both from the milieu of other students, whose pastimes he considered futile, as well as from his roots. Every time he went home to Mont-de-Marsan, it seemed farther and farther away, which Bourdieu found unbearable.

In order to express his alterity, Bourdieu chose a conceptual track, and tried to describe the mechanisms of domination. This led him to choose his objects of study from society itself. For students in kbdgne classes, Iean-Paul Sartre incarnated the philosophical ideal of the period. Sartre was the obvious and only model because he could cover the whole gamut of intellectual activities. He was as talented a writer as he was a critic and philosopher. In the early fifties, in the eyes of the ENS philosophy students, any philosopher worthy of the name was "forbidden to break the rules by becoming attached to certain objects, particularly those dealt with by specialists in the sciences of man."l1 Bourdieu felt disconnected from this ideal pattern, which he found both "fascinating and repulsive."12 He was socially isolated at the ENS. Very early on, he became interested in philosophers like Martial Gueroult and Jules Vuillemin who were oriented more toward episte-mology and the history ofphilosophy and ofscience, which he saw as a possible alternative. He felt close to them and to their positions in the intellectual and philosophical context of the period because they also came from popular and provincial milieus. At that point he was going to undertake his first research project on the phenomenology ofaffective life, which would have allowed him to apply philosophical thinking to biology-a concrete, scientific realm. But he ultimately chose ethnology, preferring a specific field of investigation and a method that claimed to be scientific. "The new prestige that Levi-Strauss had given to this science [ethnology] doubtless helped me a great deal." 13

Shortly after the ENS, Bourdieu went to Algeria, which in 1957 was in the throes of a war of liberation. An assistant at the University of Algiers, he discovered not only a subject of study, but an existential proximity to and sympathy with the Algerian people, which led him to undertake a research project. He become a sociologist in order to account for colonial reality in Algerian society, and published his first book, Sociology of Algeria. 14 In the same vein, he also focused on the situation of Algerian workers.P But Bourdieu also became an ethnologist in Algeria. Interested in Kabyle society, he studied marriage laws, kinship rules, and symbolic systems. There was a continuum between sociology and ethnology, and he simultaneously carried out his research at these two levels. "During Durkheim's time, the ethnology/ sociology distinction did not exist."16 At the time, the field and its methodology were not considered to be more than temporary detours with respect to philosophy, with which Bourdieu never really broke, except institutionally: "I only admitted to myself that I was an ethnologist very late. I thought I would do ethnology on a temporary basis and then come back to philosophy."17

Bourdieu's Structuralism

Until the seventies, the theoretical perspective of Bourdieu's work was structuralist. He himself quite precisely dates his last "felicitous structuralist" work from 1963.18Not that he abandoned structuralism. In 1969, he published an article in which he contemplated how and under what conditions the structuralist method could be extended to sociology." Much later, and despite a certain critical distance from the structuralist paradigm, he paid homage to a method that had made it possible to introduce a mode of relational thinking into the social sciences, and to break positively with a substantialist mode of thinking.s? On the occasion of a broadcast on Levi-Strauss in 1988, Bourdieu again acknowledged that many aspects of his book, Distinc-tion,21 grew out of a structuralist approach, particularly the basis of the whole analysis tending to demonstrate that to exist symbolically is to differ: "Distinction bespoke a typically structuralist equation between meaning and difference."22 Again in 1988, Bourdieu stated that he and Levi-Strauss had similar modes of thought and that the obvious differences between their work had less to do with their common theoretical framework than with their fields-ethnology for the one, sociology for the other. Bourdieu had to work on a differentiated society and had to consider its different levels-symbolic, economic, social-which meant that the effects of the same structural method yielded different results. Bourdieu therefore built his oeuvre over a long period of time within the structural paradigm: "I needed a very long time to really break with some of the fundamental postulates of structuralism.... In becoming a sociologist, I needed to get out of ethnology as a social world, so that certain unthinkable questions could become possible. "23 This led him to enclose objects of analysis within a system of essentially static determination in which events and historicity were reduced to meaninglessness: "This was typically a system for which there were no events. "24 The will to given precedence to oppositions in a relation set in the present led to a valorization of spatial, topological determinations at the expense of other considerations. This method made it possible to reveal certain logics, but it also led to a certain reductionism when studying contentless confrontations and genealogies.

It was within this logic of desubstantialization that Bourdieu presented the BarthesIPicard polemic over Racine in 1965, in Homo aca-demicus.o He reduced the quarrel between the Ancients and the Modems to what was in fact a complicity between the two protagonists, a circularity in the arguments of the adversaries, and the simulacrum of theoretical combat. This was simply an epistemological couple "between the consecrated oblats of the great priesthood and the little modernist heresiarchs,"26 brought together, in fact, by their structural complicity. So there would be nothing gained by looking at the arguments of each camp, by comparing their methods, "in the very content of their respective positions,"27 for everything exactly reproduced the positions of the Sorbonne's literary studies and those of the social sciences at the École Pratique des Hautes Etudes. On the other side, there was Barthes and the others, who were marginal with respect to the university. In the other camp were the canonized bearers of legitimacy, the die-hard defenders of tradition, such as Frederic Deloffre in 1968. Neither camp did much more than repeat, this time somewhat farcically, the battle that had been going on since the end of the nineteenth century between the new Sorbonne of Emile Durkheim, Gustave Lanson, and Ernest Lavisse and the old Sorbonne of the worldly critics. Racine was hostage to an underlying power struggle and the social success of structuralism could only be explained as a magic potion for getting jobs for an entire generation whose numbers were increasing, of professors and students engaged in new disciplines, "by allowing them to position themselves on the playing field of 'science.' "28

This dimension of things brought to light by Bourdieu did have its merit as there was indeed much in this quarrel having to do with institutional positions and power. But it was particularly reductive to limit the nature of the confrontation to its social aspect, in so cavalier a manner and in the name of a social topology, and to simply discount the arguments. Everything becomes little more than a simple structural game of different places in which all changes in the rules of play or all desire for real historical change can only be dissolved. Bour-dieu's work displays certain characteristics of structuralism, applied to sociology, including denying all pertinence and emptying all content from the semiological arguments of the sixties that tried to make a decisive epistemological break with structuralism. For Bourdieu, social actors, even those who considered themselves to be most free of social determination, were, unbeknownst to themselves, moved by forces that acted upon and reified them. It was up to the sociologist to restore these objective conditions of discursive practice in order to reach a causal level in which subjects were absent, or present only by their illusions. According to Raymond Boudon, "there is an exaggeration here of constraints, and the absurd idea that these constraints come from the social totality and from its ostensible concern for reproducing itself. "29

This meant that Bourdieu adopted the paradox of most of the structuralists, who were leftist intellectuals working for change, developing theoretical critical weapons to advance a progressive struggle, while being simultaneously seduced by a paradigm that stifled all desires for change and in so doing announced the end of history. In exchange, it offered the guarantees of scientificiry and the possibility of understanding the reified social realm in order to get a firm grasp on its totality. "This was the phase of despair, and it was not wanting for beauty. But this despair had more to do with defeated optimism than with any true rational pessimism.v-?

The young Bourdieu who introduced structuralism into sociology maintained the theme of the absent subject subjected to its social destiny, outside of which it does little more than fool itself with words that mask its failures. The only comprehensible mechanism in the system was concrete material interests belonging to the process of objec-tivation by which the subject is revealed in a truth that does not belong to it.

Structures of Reproduction

Bourdieu's work in sociology immediately became significant and influential, even if he considered it to be only secondary at the time, the simple expression of militant necessity. Bourdieu's more fundamental concerns still tended toward kinship systems and ritual systems-in other words, ethnology. But he wanted to react to what he considered to be a particularly simplistic ideology on the rise, which consisted in saying that all students constituted a social class unto themselves. He therefore decided to present a more scientific view, as a sociologist. In I964, together with jean-Claude Passeron, he published The Inheritors:» Both authors attacked the mystifying aspect of Jules Ferry's egalitarian discourse on the public school, which claimed that every individual had an equal opportunity to realize his or her potential. In this respect, while this work remained true to structuralism by demonstrating the inescapable logic of the reproduction of systems, it became a real critical weapon against the school system and one of the central issues of May 1968. Bourdieu and Passeron in fact plainly showed that behind the facade of institutional pseudoneutrality, schools were fundamentally selection machines, albeit in the name of purely academic criteria, whose function was to reproduce existing social relationships. Real social selection was hidden. "For the most underprivileged classes, it is a pure and simple question of elimina-tion."32 Those who did manage to get to the university had two possible relationships to knowledge. Either they inherited culture and were relatively removed from scholarly knowledge, or they were the sons and daughters of the petite bourgeoisie, who "more strongly embraced scholarly values. "33

The structuralist paradigm was palpable in this view of the school world. The truth about the school system and its ironclad logic was to be found in the hidden side of things. For Bourdieu and Passeron, all efforts or pedagogical thinking that did not serve to occult the reproductive function of the teacher came to naught. "The most routine professor fulfills his objective function in spite of himself."34 There was no freedom, therefore, no possibility for the agents of the system to act, and those who were marginalized and excluded could do nothing other than turn to the sociologist as a therapist because he could at least explain their case to them, even if the conditions were incurable. "For want of changing the classification of the badly classified, he would give them 'the possibility of guiltlessly and painlessly accepting the situation."'35 Everyone had his or her place, teacher and student alike, whatever the content of the discourse or the singularity of the behavior; the acceptance or rejection of dominant knowledge was implacably recuperated by the reproduction machine. Escape was impossible since the most radical protest reinforced the system's capacity to classify. "How could we avoid seeing that the revolt against the school system and the efforts at escaping its constraints by taking up very different causes ultimately and indirectly served the university'S pur-pose?"36All escape routes were sealed off.

This book was remarkable for the paradox that it raised to a paroxysm and that clearly expressed the general situation of structuralism in the sixties. On the one hand, it was possible to see critical thinking advance and to provide it with solid tools, but on the other, these weapons were stilled by the fundamental impossibility of any change. The revolt against the rule became a means of internalizing the rule. In 1964, this structuralist vision of the world of schools and university provided the future movement of May 1968 with serious arguments. At the same time, it denied, if not the possibility, then at least the significance and impact of May 1968. Once again the event and history were negated and static systems of classification given priority.

There was one important bit of progress, however, in terms of theory. The symbolic dimension was taken seriously and Bourdieu, as a sociologist of Marxist economism, managed to escape the mechanistic vulgate. In this respect, his contribution resembled the work of the Althusserians in the new importance given to superstructures. "At the beginning, I argued over this with Bourdieu at Lille, telling him that he gave too much importance to symbolic capital. I have to admit that he was right."3? But, like the Althusserians invoking the autonomy of the modes of production, Bourdieu also spoke about the autonomy of the field of cultural production, in which each subgroup was governed by its own rules, leading to fights about internal classification in every field. With this notion, Bourdieu could escape the mechanism that amounted to making every form of discourse a reflection of class position within society at large, by hypothesizing the autonomy of the symbolic and its logic. But this autonomy was limited. Just as the economy was ultimately the determining factor for the Althusserians, so it was for Bourdieu. According to Alain Caille, this reductionism was made by analogy to the notion of material interest, the true matrix of Bourdieu's theory. "There was clearly a generalized economism that was no longer substantialist economism."38

Bourdieu rejected vulgar causal economism, replacing it with the idea of a total system that went beyond the dichotomy between economic and noneconomic. In this way, motivations stemming from material interests could be discerned just as much as those based on the most apparently gratuitous activities and least tied to anything economic. Reasoning essentially by analogy, Bourdieu constructed his own staggered, "generalized political economy,"39 based on economic capital, social capital, and symbolic capital, each involved in complementary and autonomous relations. Where Marx saw class struggle as the motor force of history, Bourdieu saw the struggle for classifications as the motor force of the logic of social space. The historical dialectic became dissolved into the synchrony and fixed stratification of the different fields and of the game of placement they made possible, according to a logic similar to that of material interests. "The same always engenders the same. "40

A Concern for Style

Bourdieu was concerned with style: he had not laid literature to rest. Although he had chosen to work in the human sciences, he also considered himself to be a writer like the other structuralists. "What interested me the most in Bourdieu is the work on the text, how little by little he unveils while hiding or hides while unveiling____First he begins like a novelist."41 Like a writer, Bourdieu's thinking works from analogy. Like a novelist, his observations and their commentary are more engaging than the raw material of sociological research. In this respect, scientific discourse serves as a base for telling his story by telling others' stories; his work unfolds by what goes unspoken, in the margins, notes, and exergues. "His writing inevitably evokes Balzac. We can but recall Rastignac or Lucien de Rubempre while reading the analyses explaining how the acquisition of a solid social and cultural capital can opportunely palliate the initial deficiencies of economic capital. "42

Bourdieu often invoked another subjective analyst of society, a great writer whose work was so vast and so perfect that it tended to dissuade all rival literary vocations: Marcel Proust. He also referred to Flaubert, that other debunker of the petite bourgeoisie. Pierre Encreve thought that Bourdieu should be compared with Rousseau, given his militancy and determination to free people of their chains.

A writer and sociologist who shifted the boundaries of economy, sociology, ethnology, and philosophy, Bourdieu belonged above all to that unclassifiable French critical thought rallied around the sign and method of structuralist thinking, even if, as we shall see, he would take his distance during the seventies and eighties and become increasingly critical of certain orientations of this way of thinking.


I967-I968: Editorial Effervescence

If the year 1966 was the high point for structuralism, a series of questions arose immediately on the heels of success. Not that the decline was palpable, however-on the contrary. It was actually in 1967-68 that the media had felt the shock waves of structuralism's success and a broader public joined the structuralist vogue, as if it were adopting a panacea. Structuralism was on the lips of anyone who was anyone, a phenomenon identified with modern thought itself, the means to felicitously unite almost all of the major thinkers of the period. Only the cabarets, which had tipped their hats to the existentialist era with a jazz tune, had yet to give a playful dimension to the phenomenon, for the rockers and Hello Pals' were not truly part of the structuralist festivities.

Levi-Strauss, Foucault, Lacan, Barthes, and Althusser had each enjoyed individual public acclaim. This time, structuralism provided a means of transcending the singularity and talent of each of these writers, leaving the public speechless for not having made the link before.

But while structuralism was being trumpeted, its foundations were being fissured, and the will to go beyond or radicalize the phenomenon was already in place. This lag simply translated the different rhythms of research, colloquiums, special issues of journals, and media attention. One sign of the end of the epic era and beginning of a new period of decreasing returns was that increasing numbers of publications began to come out claiming to take the pulse of the phenome-non. These journals presented structuralism didactically so as to spread the word. They did, of course, contribute to the movement's success, but at the same time, the authors of the books being discussed became increasingly mistrustful of their transformation into part of a passing fashion. All forms of structuralist labeling were generally shunned; no one wanted to be a victim of the wave when it crashed. On the one hand, there was the ephemeral aspect of this type of collective taste, and, on the other hand, increasingly radical and numerous critics were voicing their dissent from within the structuralist camp itself.

Alice in Structuralism Land

Everyone was mobilizing in the publishing houses. In 1967, Seghers put out Keys for Structuralism- by jean-Marie Auzias, which described the different pieces of the movement. This work was didactic ("This book is intended for teachers"]! and peremptory ("Structuralism is not an imperialism! It aspires to be scientific, and is!")." There was a run on the book; it was barely out before it was already out of print, even if a recalcitrant Francois Chatelet announced that he preferred to leave these keys "under the doormat."5

Privat published jean-Bapriste Pages's Understanding Structuralism» in 1967 and, the following year, Structuralism on Trial,' At PUF, Jean Piaget was asked to write for the Que sais-je? collection, and he cut structuralism up into its disciplinary pieces while reminding readers that neither the notion of structure nor its use in such diverse fields as mathematics, physics, biology, linguistics, and sociology was new. These observations made the range of conceptual advances manifest, and amply demonstrated the scientificity of the project, but only, according to Piaget, so long as other methods were not excluded and the human and historical aspects of things were included. In this respect, Piaget opted for a genetic structuralism close to Lucien Goldmann's positions. His own work on child psychology was one of the possible illustrations of reconciling history and structure. This Que sais-je? was intended as a university tool but quickly became the structuralist handbook, so much so that many today still identify structuralism with Piaget, whereas he was one of its critics.

In 1967, Payot published Saussure's Course in General Linguistics8 in a critical edition prepared by Tullio de Mauro and translated by Louis-jean Calvet. The rather arid, hefty tome was ripped from the bookshelves, and not only by linguists. The return to Saussure, the rumor of the discovery of the philosophers' stone in the realm of the social sciences, ensured an extraordinary audience for Andre Martinet in 1967-68 at the Sorbonne. "In the Descartes amphitheater, there were all sorts of people. It was because there was something new, which was appealing. At the time I had Michele Cotta [a political TV commentator-Trans.], filmmakers ... "9

But the important editorial enterprise in 1968 took place at the Olympus of structuralist ideas, the Editions du Seuil, the project of one philosopher-editor in particular: Francois Wahl. The project dated back to 1966, the wonder year for structuralism, when Wahl--editor for Lacan's Ecrits, for Barthes, and, a bit later, for Derrida-who was concerned with editorial coherence and very interested in what was going on in the social sciences (he even regretted "not having had the opportunity to publish Claude Levi-Strauss"},"? decided to undertake a collective work to answer the question, "What is structuralism?" in philosophy as well as in the social sciences borne along by the phenomenon of modernization. Wahl therefore invited articles by Oswald Ducrot on linguistics, Dan Sperber on anthropology, Tzvetan Todorov on poetics, Moustafa Safouan on psychoanalysis, and wrote the chapter on philosophy himself. The book was so successful that it was reprinted in small discipline-specific paperbacks in the Points-Seuil collection beginning in 1973.11

The structuralist designation was not just a media effect or a simple fantasy, as some have claimed, but had its place at the heart of the production of the phenomenon. In his general introduction, Wahl clearly translated the vision. "Under the term 'structuralism' we find the sciences of the sign, and sign systems."12The phenomenon cast a wide net and made it possible to set its sights quite high since, in the view of Prancois Wahl, it was the model of models, and made it possible to attain the status of a science. "Whatever the case, structuralism is a serious thing, as we have seen. It gives anything having to do with signs a right to science."13 The volume clearly expressed the euphoria of the moment, the way in which semiology bathed everything in a scientific wash. Today, Francois Wahl recognizes that there was something of an "epistemological naivete whose measure we have progressively taken____We were somewhat blinded since we thought we were

in the process of discovering the key." 14

Beneath and beyond Structuralism

Francois Wahl's contribution in philosophy allowed him to discern a structuralist underside in Foucault and a beyond with Lacan and Der-rida, while the middle ground corresponded to Althussero-Lacanian positions. His choice made a nod in some measure to the day's fashion since he made no mention of Martial Gueroult or of Victor Gold-schmidt, although today he considers their reading of Plato's Dialogues and of Descartes to "represent a historical culmination, and I am absolutely sure of that."!' But since they were working only in philosophy, closed to the social sciences and unknown to the public, their contributions to philosophical structuralism went undiscussed. At the same time, Wahl gave considerable attention to Foucault, by investigating his notion of episteme. Although Wahl saw in Foucault's episteme the trace of structuralist concerns, he nonetheless considered Foucault as belonging more to philosophical nominalism. For him, Foucault did not make the announced break with phenomenology but remained a prisoner of it. When he sought the being of the sign, defined through its specific properties as an essence, or when he tried to grasp it at its originary presence, he was faithful to Merleau-Ponty's thinking. "To look at things as a phenomenologist, which is to say in the undersides of structuralism, the being of language defined by structuralism is a contradictory project that can only assign the status of that which remains to being."16 Wahl did of course recognize that Foucault sought an organization of perception, but he was at the edge of vision, a nominalist with an impossible project that attempted to reconcile two incompatible models: the one phenomenological, the other structuralist. "We are beneath the sign, beneath discourse, beneath structure."17 Foucault only led his reader to the edge of the Rubicon without crossing it, in order to go fishing. "Does an episteme of structuralism exist? And how is it that on this topic The Order of Things does not take a position?"18 Wahl was referring to the break required to let this episteme exist, but saw it in Althusser's work and in his explicitly scientific project. "We cannot consider structuralism without thinking about science." 19

Later, Foucault firmly rejected the structuralist label. But at the time, he was a full member of the movement. In The Order ofThings he even went so far as to present himself as the philosopher of these important breaks in the episternic anchor. So he hardly appreciated

having his philosophical project set at a distance from the structuralist project. "He was extremely angry at this, and I can even tell you that he was momentarily furious. "20 Wahl, in his presentation of philosophical structuralism, gave priority to Lacanian-Althusserian constructions, particularly those of Alain Badiou and jacques-Alain Miller.21

In 1968, he brought together jacques Derrida and jacques Lacan, the two warring brothers who politely hated each other, in a beyond of structuralism. He called one jacques and the other jacquot in order not to mix them up. On the one hand, Lacan made it possible to reconsider the subject, which until then had been eliminated, without, however, claiming a plenitude for it. This subjected subject could not reaffirm itself as a secure basis for thought, and was forever out of sync with respect to itself. Blocked doubly in its return as its own master, the subject was subordinated to language and to its structuration within the structures of the signifier. "The letter precedes the subject ... the letter comes before meaning."22 Derrida's beyond of structuralism had to do with the way he went beyond philosophical discourse by his Other, his argument against the notion of limits and of origins, using his notion of trace. For Wahl, structuralism was defined by this break: "Structuralism begins when the system of signs sends us elsewhere. "23

Single and Plural Structuralisms

The few preparatory meetings before the publication of this collective work produced no real common theme, since the points of view often differed. Dan Sperber, back from Los Angeles where he had taken Noam Chornsky's courses and had been put in touch with Francois Wahl by his friend Pierre Smith to translate Chomsky's work, was assigned the chapter on structuralism in anthropology. During these meetings, "there were not really any discussions other than my insistence that Ducrot talk about generative grammar, which he had not been planning to do. "24 Indeed, Oswald Ducrot, who was responsible for the linguistics section, presented the general outlines of Chomsky's thinking next to the other currents of structural linguistics, without claiming any hegemony for his discipline or arguing for its place as a pilot science. That the work began with the piece on linguistics was Francois Wahl's choice, and clearly reflected the role linguistics played in the development of the structuralist paradigm. "I recall having said to Wahl that I saw no reason to start with linguistics. For him, it was obvious, and any other choice would have seemed scandalous to him."25

Dan Sperber had a choice morsel in having to describe Levi-Strauss's work. Starting from his Chomskyan positions, Sperber criticized Levi-Strauss and read his work in such a way as to give priority to everything having to do with structures of the human mind and the deep structures that recalled Chomsky's model of competence. He criticized Levi-Strauss for not having gone far enough and for having maintained a contradictory tension between his ethnological ambitions of inventorying cultural variations, and his anthropological ambitions of determining the specific learning capacities of the human race that orient these variations. "Personally, with generative grammar, I had reservations about structural linguistics from the beginning, and when I was asked to do the chapter on structuralism in anthropology I did not conceive of it as a manifesto for structuralism, but as a chapter intended to be partly critical. "26

The psychoanalytic chapter, which Moustafa Safouan was asked to write, took a Lacanian line. Safouan, an Egyptian philosopher who had translated Freud into Arabic, had converted to psychoanalysis thanks to Lacan, and worked under his supervision for more than ten years. He addressed a number of themes that Lacan himself had addressed during his seminars at Sainte-Anne Hospital between 1958 and 1963, although his approach to the unconscious was less genetic and historical and more spatial and structural than was typical. "By saying that the unconscious is a place, we do little more than confirm the fact that Freud presented his doctrine on this subject as 'topical,' a metaphor, certainly, but which means that beyond everything that constitutes our relationship to the world, an Other Place exists. "27 The structure discovered by psychoanalysis was not located in any hidden meaning to be revealed to itself in its presence, but is to be found there where the subject did not know, in a rupture "that only the Law maintains against (and from) the temptation that pushes man to rediscover-in vain-his initial enclosure."28

Each of these four contributions represented a particular and institutionally sanctioned discipline. In addition to these, Tzvetan To-dorov, who had introduced Russian formalism in France, wrote a long piece on poetics that was both new and connected to structuralism's roots. He wanted to demonstrate how the structuralist method could profoundly renew literature. Poetics was defined as an approach both abstract and internal to literature, and that sought to restore the general, underlying laws of each work. Like Gerard Genette, Tzvetan Todorov did not present poetics as the exclusive activity of interpretation and hermeneutics but as its necessary complement: "Between poetics and interpretation, the relationship is one of complementarity par excellence,"29 although only poetics participated in the semiotic project, for it alone was anchored by the sign. Poetics differed from a specifically linguistic analysis, where, according to Todorov, the process of signification had two major limitations. Linguistic analysis ignored the playful aspect of language, the problems of connotation and metaphorization, for example, and "barely went beyond the sentence as the basic linguistic unit. "30 Todorov was not only criticizing linguistics, but early structuralism as well, preferring plurality and polyvalence.

Both Bulgarians, Julia Kristeva and Todorov, were influenced by Mikhail Bakhtin. "It was Bahktin who first formulated a true theory of intertextual polyvalence."31 An entire mode of operative analysis resulted based on dialogics, an initially literary idea making it possible to link up once again with Roman Jakobson's initial thrust, when he declared in I9I9: "The object of literary science is not literature but literality, which is to say that which makes a given work a work of literature." 32

Even if each of the authors of this collective work was moved by considerations belonging to his or her particular discipline, the possible bridges between them were clear, making it possible to articulate a general understanding around the structural paradigm. A manifestly theoretical ambition drove the whole project and whetted the appetite of the moment for the structuralist key. At the same time, the work clearly translated the situation of general semiology, which had reached a turning point after having undergone diverse assaults to make it more receptive and ensure its imminent collapse. But the intellectual public was unaware of these internal contradictions, which appeared as so many sources of encouragement for this new and productive mode of thought. When Francois Wahl overheard a philosophy professor from a suburban high school express amazement that his twelfth-grade student was reading Freud, and heard the student answer that it was in order to understand Lacan, "At that point, I said to myself, I have won!"33 Without being aware of it, this student was working toward the return to Freud and encouraging both Lacan and his editor. How could anyone resist the collective euphoria under such circumstances?

The Four Musketeers

In 1967 and 1968, Levi-Strauss was working on the publication of his monumental Mythologiques,34 He was still the uncontested master and true force behind this effervescence, even if he carefully remained aloof from any of the extensions of his method. Levi-Strauss refused to assume any kind of paternity that might become burdensome or dangerous, but he was present in the many media echoes. He even increased the number of press interviews in order to present his work, but it was in order to remain within the strict limits of his structural anthropology, one way of keeping to the margins of a speculative structuralism in full bloom. Le Nouvel Observateur played a major role in reaching a wide and cultured public. On January 25, 1967, it devoted three pages to Levi-Strauss. He defined structuralism, implicitly refusing certain uses of the paradigm: "Structuralism is not a philosophical doctrine, it's a method. It took a sampling of social facts from experience and transported them to the laboratory. There, it tried to represent them in the form of models, always considering not the terms but the relationships between the terms."35

By carefully limiting the phenomenon to a method, Levi-Strauss firmly maintained an approach that he judged to be purely scientific, and that differentiated his from certain speculative and ideological uses, for he intended to confederate anthropology with the natural and physical sciences and was about to win his institutional wager without any specific university cursus. He was so successful that "we are obliged to discourage the students from this vocation."36 Despite his complaint about inadequate funding, the number of chairs in anthropology went from five to thirty in twenty years (including the EPHE), and ethnology had gained ground in the university since ethnology was being taught in five universities outside of Paris: Lyons, Strasbourg, Grenoble, Bordeaux, and Aix-en-Provence.

Philosophically, Michel Foucault held the high ground in 1967-68, following the heralded publication of The Order of Things in 1966. Sartre, and then two Sartreans, Michel Amiot and Sylvie Le Bon, violently attacked Foucault in two very critical articles in Les Temps modernes in 1967. Still, he could count on the considerable support of

Georges Canguilhem, a man rather unused to throwing himself into the ring but who enjoyed the greatest admiration and prestige among philosophers. Canguilhem took up his pen to defend Foucault in the review Critique, in which he attacked, not without a certain degree of humor, a nascent league for the defense of human rights that seemed to be mobilizing to block Foucault's thesis, behind the slogan, "Humanists of all stripes, unite!"3? He emphasized the major contributions of Foucault's work, thanks to his notions of episteme and archaeology, which avoided the errors of anachronism, so often encountered in the history of sciences. Canguilhem paid homage to this other history that used original texts of the period and whose related events "affect concepts rather than men."38He aligned Foucault with Jean Cavailles, who effected a similar shift from the point of view of consciousness to that of concepts, and saw in him the great contemporary philosopher who might realize this philosophy of the concept to which Jean Cavailles had appealed.

In 1967, Foucault was one of the four musketeers in Maurice Henry's sketch, which shows Levi-Strauss, Barthes, Lacan, and Foucault dressed up as Indians and chatting, with Foucault happily squatting among his structuralist peers, an enthusiastic participant in the community of thought wherein the press placed him." This explained his angry reaction when Francois Wahl's book confined him to a period predating structuralism, whereas at the time he defined himself quite explicitly as a structuralist. In an interview with a Tunisian newspaper in 1967, he distinguished two forms of structuralism: that of a fruitful method that could be used in different, specific fields of knowledge, and another structuralism, "which would be an activity through which nonspecialist theoreticians tried to define possible current relationships between this or that element of our culture, this or that science, this practical realm and that theoretical realm, and so on. In other words, this was a sort of generalized structuralism no longer limited to specific scientific realms. "40 This second structuralism, clearly the one in which Foucault saw himself, let him maintain his specificity with respect to the rising social sciences because it alone could confirm or invalidate their scientific conclusions, thanks to its critical distance from the individual fields of study.

Another musketeer at a turning point in his own work and with respect to literature, Roland Barthes was moving closer to the notions of subjectivity and of historical dynamism. This did not prevent him, in 1968, from proclaiming his complete agreement with the basic principles of the structuralist approach. Moreover, the text he wrote announcing the "death of the author," which was the literary equivalent of Foucault's "death of man" in philosophy, made a considerable impact. An author would be nothing more than a recent notion born at the end of the Middle Ages thanks to capitalist ideology, which dignified the person of the author. But this mythical figure was on the verge of dissolving, for when "the author is entering into his own death, writing begins. "41

Surrealism had begun to jolt the myth of the author, but linguistics would finish it off by furnishing "a precious analytical instrument for destroying the Author, by showing that the entire utterance is an empty process. "42 In its place came the scriptor, a sort of being outside of time and space, set within the infinity of the signifier's unfolding, making any attempt at deciphering a text hopeless: "To assign an Author to a text is to impose a stopping point, granting it a final signified, closing off writing. "43 Barthes joyfully celebrated the birth of the reader on the ashes of the still smoldering body of the Author.

The other front on which Barthes reiterated the structuralist orthodoxy was in its relationship to history. Although he had assimilated the notion of intertextuality, which allowed him to dynamize structure, he continued to refuse to fall back into historicism. Both of his 1968 articles on "The Reality Effect" and "Writing the Event" implied at once a rapprochement with the idea of transformation, of dynamism, and a reiterated rejection of history.t' He suggested a complicity between literary positivism and the reign of "objective" history in their common concern to authenticate a "reality." He saw the historian's discourse as based on a myth, an illusion called "the referential illusion," transforming "reality" as a denotational signified to a con-notational signified.P If one of the tasks of modernity was to dismantle the sign, and if this was happening with respect to realism in literature, here there was an error, regressive because it operated "in the name of referential fullness. "46

The fourth musketeer at the structuralist banquet in Maurice Henry's sketch was jacques Lacan. Lacan was surprised to find himself in such company: "I assigned myself to the so-called structuralist bucket,"47 but in order to start a review in 1968 based on the structuralist principle of the death of the Author. Lacan even invoked the Bourbaki mathematicians in order to justify the principle of publish-ing anonymous articles in this new review, Scilicet. And yet, the anonymity of scientific writing stopped with the Name of the Father, Lacan. "Our name, the name of Lacan, cannot be hidden in the program. "48 Lacan alone could sign his articles in the review, and those who did not participate in it "cannot be recognized as having been one of my students. "49 The sanction was clear for any potential foot draggers and the project was neatly tied up: maximal visibility for the words of the Master and anonymity for the others. The masses had to pay through the nose for the theorization of the death of the Author, the disappearance of the signature in the name of a scientific superego incarnated by Lacan, who was none other than Lacan's Other.

In 1967, a more serious undertaking was published by PUE Jean Laplanche and jean-Bertrand Pontalis coauthored The Language of Psychoanalysiss? defining all of the notions of psychoanalysis. A precious tool indexing Freud's work with citations, as well as Lacan's notions, this work also concretized this return to Freud that Lacan had managed.

The Seventh Art

Triumphant structuralism even included a new field in its vast empire: the seventh art, cinema. In 1968, Christian Metz published his Essays on Signification in the Cinema.i) which heralded an entirely new area of semiology. Metz had already written an article in the programmatic issue of Communications in 1966.52 The book assembled his texts written between 1964 and 1968, and extended the applications of linguistic concepts to film criticism. "In a word, I wanted to go to the limits of the metaphor of a 'cinematographic language' and try to see what it encompassed."53

Ever since adolescence, Christian Metz had been an impassioned cinephile, but for a long time with no particular outlet except organizing programs in film clubs. Metz had studied linguistics, and "the idea of a semiology of cinema came to me by the connection between these two sources."54 Given this, he went from cinephilia to a new approach to cinema, to which he applied the conceptual grid that he refined with his "grand syntagmatic." "The object of my intellectual passion was the linguistic machine itself." 55

Metz's first semiological essay in 1964 began by reacting against the cinematographic criticism that ignored linguistic renewal and remained impervious to semiology while multiplying the invocations to

a specific cinematographic language. "I began with Saussure's notion

of language____It seemed to me that the cinema could be compared to

language and not to speech."56Dealing almost exclusively with fiction films, Metz believed that he had found a model that could be applied to all of cinematographic language. His "grand syntagmatic" divided films into autonomous segments based on grand syntactic types (in I966 there were six, and in 1968, they numbered eight). The autonomous shot (a single shot equivalent to a sequence); the parallel syntagm (parallel montage); the accolade syntagm (undated evocations); the descriptive syntagm (simultaneity); the alternating syntagm; the scene properly speaking (coincidence of the unique consecution of the signifier: what happens on-screen; and the unique consecution of the signified: the temporality of the fiction); the sequence by episodes (discontinuity become the principle of construction); and the ordinary sequence (the ordered arrangement of dispersed ellipses). These eight sequential types "are responsible for expressing different sorts of spatiotemporal relationships."57 The validity of this code includes, in fact, classic cinema, which goes from the thirties to the new wave of the fifties.

This extremely formalized cinematographic language drew its linguistic inspiration essentially from Hjelmslev whose notion of expression defined the basic unit of filmic "language," according to Metz, whereas codification belonged to a purely formal, logical, and relational approach: "The way Hjelmslev understood it (equals the form of the content plus the form of the expression), a code is a commut-able field of meaningful differences. So there can be many codes in a single language."58

On the eve of May 1968, structuralist France was bubbling. A new theory was shooting up at every moment between the Parisian paving stones, and the world was being remade from a topic and for want of a utopia. Structuralist energies seemed to represent the great fracture of modernity until another-this time historical-fracture came to jolt its convictions.


Structuralism and/or Marxism

A confrontation did occur in I967-68 between structuralism and Marxism, the two important totalizing and universal philosophies. Marxism's decline seemed to feed structuralism's success, but in exchange, could not the Marxism of the late sixties find its second wind through structuralism? Could there be a reconciliation between these two approaches, or, on the contrary, were they incommensurable?

Marxists could no longer duck the issue. Althusser's work and its impact prevented them from doing so, and the spectacular interest in structuralism made the theoretical debate with structuralist positions unavoidable. Lucien Sebag had already begun this debate prior to I968, by publishing Marxism and Structuralism (I964) at Payot. Like Althusser, he wanted to reconcile Marxism and contemporary rationality with the progress that had been made in the social sciences.

Lucien Sebag: An Attempt at Reconciliation Lucien Sebag was trained as a philosopher and a researcher at the CNRS. Like his friends Alfred Adler, Pierre Clastres, and Michel Cartry, he had turned to anthropology and thus to fieldwork. A student of Levi-Strauss, he left France in I96I to spend nine months among the Euyaki Indians in Paraguay and the Ayoreo Indians in Bolivia. He was caught among the modernist pulls of the moment. A structuralist, he, like his teacher, considered that structure was purely methodological and not a speculative conceit. Interested in psychoanalysis, he had begun an analysis with Lacan, who had a privileged relationship with this young philosopher who seemed able to lay the foundations for some new bridges to make his ideas known. A semiol-ogist, Sebag was a student of Greimas, with whom he planned to work on structural semantics in order to make it receptive to the unconscious. A Marxist and member of the French Communist Party (PCF), he had been increasingly critical of the party since 1956. The rigor of the social sciences offered a good counterpoint to the vulgate diffused by the party leadership. He was critical of the Marxist eco-nomism of the time, which saw economic life as a reality in itself with a direct causal role in social relations.

Sebag appreciated the fact that Marxism had replaced the ambient idealism with a concern for studying objective reality, and more particularly, economic reality. However, using the linguistic edge of the structuralist theses, he criticized Marxism for having somewhat fetishized its privileged object and for having underestimated the underlying, immanent principles that organized economic reality, particularly those making it possible to transcend the differences between societies, this "creation of language that defines the very being of culture."! Sebag defended the humanist positions that led him to view structuralism as anthropology and to be skeptical of certain speculative extensions. "Man produces everything that is human, and this tautology prevents us from making structuralism into an extra-anthropological theory about the origin of meaning."2 Many were hopeful that Lucien Sebag the theoretician could modernize Marxism transformed by its relationship with all forms of structuralism. But the book proclaiming the union between Marxism and structuralism also aspired to set the seal on another union, between its author and the woman to whom the book was dedicated: Judith, Lacan's daughter. The denouement was as brutal as it was intolerable: Lucien Sebag committed suicide by shooting himself in the face in January 1965. If Lacan admitted his distress to his close friends, Sebag's editor at Payot, Gerard Mendel, considered that the analyst had failed his patient: "For Sebag, it was tragic, for Lacan mixed everything up: private and public life, the couch, and he accepted anyone as a patient, even serious depressives."3 Nicolas Ruwet, a friend of Lucien Sebag who, until that point, had been interested in Lacan's ideas, turned away from the man who could not save his friend from his ultimate despair,"

The PCF Opens the Discussion

The plan to have the Marxist and structuralist paradigms face off against each other was rather quickly accepted by the PCF leadership. Without adopting Althusser's positions at the March 1966 session in Argenteuil, the central committee nonetheless emphasized the importance of the excitement and work in the social sciences. "Given the many new questions, we can no longer let our tools of expression grow old. Philosophical debates today are taking place not only on the terrain of principles, but also that of specific fields of knowledge (economy, psychology, sociology, ethnology, and linguistics)."5 Thanks to the CERM (Center for Marxist Studies and Research) and to its two journals, the monthly La Nouvelle Critique and its weekly cultural journal, Les Lettres françaises, there was a new policy of receptivity to debate designed to ensure that intellectuals remained within the party and to stop the hemorrhage among members that had begun in 1956.

Communist intellectuals were responsible for two colloquiums to address literary theoretical problems held in April 1968 and April 1970 at Cluny. Designed to seal the union "between literature and professors,"6 and to give rise to a structuralist Marxism, they were organized by La Nouvelle Critique, the CERM, the Vaugirard Group for Interdisciplinary Study and Research, and Tel Quel.

Tel Quel embodied the avant-garde, which a number of Communist intellectuals were discovering. "This colloquium at Cluny was extraordinary: Kristeva was the diva, and others were on bended knee before her. It was even pathetic intellectually to seethe relationship."? Next to julia Kristeva, who addressed the structural analysis of texts, Philippe Sollers gave a talk on the topic "The semantic levels of a modern text" in which he cited the material anchor of the text in the author's body, not the body that can be simply described "anatomo-physically," but the fragmented body, "a body of multiple signifiers."8 Implicitly addressing Althusser's trilogy of three generalizations, Philippe Sollers discerned three levels of textual approach-deep, intermediary, and superficial. These formed a transformational matrix with three functions: translinguistic, gnoseological, and political. Jean-Louis Baudry spoke about the structuration of writing, and Marcelin Pleynet about structure and signification in the work of Jorge Luis Borges.9

The Tel Quel group was clearly the theoretical organizer of this collective thinking, and two months after the colloquium, Philippe Sollers, in the euphoria of the avant-garde role he could play with respect to the party "of the working class," created a Group of Theoretical Studies that identified its objective as putting together an overall Marxist-structuralist theory. The group included Barthes, Derrida, Klossowski, and many others, and met weekly on the rue de Rennes. "Lacan made a brief appearance."lO

This excitement coursing through the social sciences drew a number of intellectuals into the PCP. One notable example, Catherine Clement, a member of the Lacanian organization the École Freudienne de Paris, joined in the autumn of 1968. At La Nouvelle Critique, she was responsible for a number of encounters around the theme "psychoanalysis and politics."

Structuralism and the Test of Rationalism In early 1968, on the initiative of another Marxist journal, Raison presente, directed by Victor Leduc and under the auspices of the Rationalist Union, daylong lectures organized around the theme "Structures and Men" took place at the Sorbonne and drew quite large crowds. Shortly thereafter, the proceedings were published as Structuralism and Marxism.n In the eyes of the organizers, structuralism was an ideology turned against Marxism and humanism, but the debates also drew detractors as well as the partisans of this new mode of thought. 12

Henri Lefebvre warned against inappropriate applications of the linguistic model and Andre Martinet answered that there was no single model but, on the contrary, many linguistic models. Francois Bresson defended generativism and its applicability to activities other than natural languages. Victor Leduc presented the problem the colloquium organizers raised, of knowing whether or not this was simply a Parisian fashion or a new type of rationality.

The major issue revolved around the respective positions accorded structure and human initiative. "Based on a certain theory of structure, which would apply to all levels of reality, is there still a place for the historical initiative of human beings?"13 Francois Chatelet became the devil's advocate for structuralism, even if he rejected the use of the term and only considered the epithet justifiable. "What characterizes structuralism, I believe, is much more a common state of mind."14 Above all, he saw in the phenomenon a possible emancipation of the social sciences, which could establish themselves in their scientificity if they could stop fetishizing the notion of the subject, which had been dominant since the classical age. Structuralism was above all characterized by a refusal: "the rejection of humanism,"15 and the effort to rid itself of ideology in order to liberate theory. This radical split assumed the elimination of man: "In order to address the social sciences with this objectivity, everything depended on radically eliminating the concept of man. "16 The social sciences had to assert their positivity using the disappearance of the subject, just like the physical sciences, which defined themselves by rejecting the illusions of perception.

Olivier Revault d'Allonnes, the philosopher expert in aesthetics and a close friend of Francois Chatelet, was less enthusiastic about structuralism. He, of course, considered the notion of structure to be fundamental for the social sciences in the Durkheimian perspective of his teacher, Charles Lalo, professor of aesthetics at the Sorbonne in I943-44, and grandnephew of the composer whose course he had taken on structural analysis of aesthetic consciousness. "Charles Lalo showed us that ostensibly purely affective, obscure, and spontaneous reactions on the part of the subject perceiving the work of art were in fact in constant and structured relationships with the whole of the psychic life of the society."l? Because of his work on aesthetics, he reacted quite early on against the general use of pathos at the time and gave priority to structuralism before its time. And yet this orientation was not supposed to lead either to static structures or to structures without human beings. Taking the example of musical structures, he demonstrated that any musical system included areas of disequilibrium with which composers worked, reworking them until the system slid irreversibly toward a new structure. The possible paths of freedom were to be found in the search for the limits of structure. "What thrilled me in Bach was Debussy.... What thrilled me in Debussy was Schoenberg, and in Schoenberg, Xenakis."18 Knowing the structures was necessary in order for human abilities to work at transforming them. This was the price of creation. In the absence of such an effort, creation would sign its death warrant and be reduced to static structures.

Jean-Pierre Vernant was equally lukewarm about structuralism, even if he had adapted Levi-Strauss's model to ancient Greece. But he opportunely reminded Francois Chatelet of his first work, The Birth of History.t? in which he had demonstrated the complementarity between the demos, the collectivity's determination of its own political future, and the birth of a historical consciousness thanks to this realization that human beings can be active agents in history. Vernant calmly and lucidly announced, "I am not worried about man because, when he is thrown out the front door, he returns by the back door. We need only examine the contemporary evolution of linguistics to see this. "20 Vernant's other question concerned the status of history in structuralism, which he thought better suited the ethnologist and which ran the risk of reducing events ta irrational contingencies, as, for example, when Levi-Strauss took the "Greek miracle" into account as a purely fortuitous phenomenon that could just as well have happened elsewhere.

Historians in general were less fascinated with structure, even those who worked on the structures underlying the fabric of events. They insisted on the necessary dialectic between structure and dynamic to make history, as Ernest Labrousse had defined it, the science of change: "A science of movement, history is also the consciousness of movement. "21 In the same spirit, Albert Soboul defined the historian's task as understanding the interplay of the forces of change endogenous to the structure. He studied contradictions, whereas the structuralist insisted rather on the systems of complementarity at work in the reproduction of structure, "so that even the soul of history is lost. "22

However, regarding ancient Sparta, Pierre Vidal-Naquet demonstrated how fruitful the structuralist approach could be. Taking pairs of opposites, provided that they be set in a changing context, could shed light on ancient societies. "In Levi-Strauss's language I would say that the foot soldier is on the side of culture and of the cooked and that the crypt is on the side of nature, and the raw. "23 Madeleine Reberieux credited structuralism with having allowed historians to escape their Eurocentrism and with having thus transformed the way history was taught in high schools, which henceforth included the study of either a Muslim or a Far Eastern civilization. Reberieux praised this change, but she resisted a discontinuous vision of history.

Words against Things

Marxism thus seemed able to adjust to a bit of structuralism, but Michel Foucault's work, which incarnated the speculative dimension of the phenomenon, was harder to swallow. He would be the object of severe criticisms by Marxists, with, however, some nuances. For

Jacques Milhau, Foucault's excommunication was total. Had he not committed the crime of relegating Marx to the nineteenth century? "Michel Foucault's antihistorical prejudices can only be subtended by a neo-Nietzschean ideology that, whether he realizes it or not, serves all too well the designs of a class whose major concern is to mask the objective paths of the future."24 Jeannette Colombel saw in Foucault's work a trompe l'oeil choice between the desert and madness, "lucidity or despair, the lucidity of laughter. Made in USA."25 However, in presenting the general arguments of Foucault's demonstration, she also insisted on its richness and value. Two longer studies went beyond book reviews and raised some methodological questions.

In 1967, Raison presente published Olivier Revault d'Allonnes's article "Michel Foucault: Words against Things," which was reprinted in Structuralism and Marxism in 1970. In it, Revault d'Allonnes denounced Foucault's attack on the historical approach, the expression of managerial technocratism, the excessive taste for words that allowed things to be repressed, the priority of instants, and a resolutely relativist conception and discontinuity of approach.

What surprised me the most and practically stupefied me in The Order of Things was that Foucault, whom I had known as a militant, claimed that the subject no longer existed, that it was writing in air.... He gives us remarkable but static snapshots; he takes care not to tarry over that which, within these epistemic spaces, already calls them into question.sf

The other basic criticism came from the historian Pierre Vilar and was published in June 1967 in La Nouvelle Critique.27 For Vilar, by choosing discursive formations as his sole object of analysis, Foucault implicitly marginalized the historical reality that contradicted the conclusions he drew. There again the subordination of things to words was the issue, which led Foucault too hastily to conclude that there was no political economy in the sixteenth century. Pierre Vilar countered that the elements of a macroeconomics of national accounts were already in place in Spain in the Golden Age, which was then discovering the importance of the notion of production. The contador of Burgos, Luis Orty (1557), even attacked laziness by concrete political decisions, which contradicted Foucault's epistemic construction of political economy not being born prior to the nineteenth century.

Marxist intellectuals were nevertheless not overwhelmingly hostile to Foucault's theses, which were well received in Les Lettres [rancaises in particular. Pierre Daix was becoming an enthusiastic structuralist, which led to the publication of Structuralism and Cultural Reuolution.w In Daix's journal, Raymond Bellour interviewed Michel Foucault a second time, on June 15, 1967, which gave Foucault the opportunity to respond to a number of criticisms.

Foucault had not looked for absolute breaks or radical discontinuities between epistemes. On the contrary: "I showed the very form of the transition from one state to the other. "29 However, he did defend the autonomy of discourses, the existence of a formal organization of utterances to be restored, something historians had neglected until that point. He defined a horizon that could not be reduced to formalism, but sought to put this discursive level into relationship with underlying social and political relationships and practices. "This is the relationship that has always haunted me."30 Responding to the criticisms of antihistoricism, Bellour recalled the last chapter of The Order of Things where Foucault privileged history. The author confirmed, "I wanted to do historical work by showing the simultaneous functioning of these discourses and the transformations that describe their visible changes,"31 without giving excessive importance to a history that would be the language of languages or the philosophy of philosophies. And, in response to the war cries in the name of history provoked by The Order of Things, Foucault invoked the effective work of professional historians who recognized that his work was fully historical, such as the Annales historians, and he cited the new adventure represented by "the books of Braudel, Furet, de Richet, and Le Roy Ladurie. "32

Structuralism and Marxism

The important theoretical monthly of the PCF was also mobilized in this confrontation. The October 1967 issue of La Pensee was devoted to the theme "Structuralism and Marxism." The philosopher Lucien Seve, a more official voice, wrote an article to present the theoretical position of the PCP. For Seve, the structural method was an outdated epistemology rooted in the early twentieth century when evolutionism was in crisis and before dialectical thinking took hold in France. The method implied an epistemology of the model, an ontology of the structure as an unconscious infrastructure, a theoretical antihuman-ism, the rejection of the conception of history as human progress, substituting in their stead human diversity. Indeed, the method was quite old, theoretically rooted in Saussure (1906-II), the German historical-cultural school of ethnology (Grabner and Bernhard Anker-mann, 1905), Gestalttbeorie (1880-1900), and Husserl's phenomenology (Logical Research, 1900).

Therefore, according to Seve, one could not be satisfied with a division between the structural method (considered scientific) and a structuralist ideology (to be rejected as unacceptable). Those who made this division in order to reconcile dialectic and structure were in error. His target was less Althusser, whose theses had been condemned by the PCF leadership, than Maurice Godelier. "The goal of M. Gode-lier's research ... : a structural science of diachrony. "33 The price to be paid for such a conciliation was the elimination by Godelier of the class struggle as a motor force, inherent in the structure of dialectical transformation. For Godelier, "structure is internal, but the motor force of the development is external."34 According to Seve, Godelier missed the point of the very nature of dialectical thinking, which is to describe the logic of development by adopting the structural method. There could be no theoretical construction synthesizing the structural method and the dialectic, as he saw it. Although he recognized that the structural method clearly offered something on certain levels ("A Marxist can recognize the validity of the structural method next to the dialectical method");" he opened the narrow path of a union considered to be a struggle.

But Seve could not deny the fruitfulness of this new paradigm, whose eminent representatives also contributed to this issue of La Pensee. Marcel Cohen gave a historical synopsis of the use of the notion of structure in linguistics in the Continental school as well as in the United States. Jean Dubois wrote a veritable apology for structuralism in linguistics, showing that it had made it possible to liberate linguistics from the most destructive aspects of the prior methodology, "an outrageous psychologism and mentalism,"36 and to establish it as a science. Jean Dubois acknowledged that two problems---ereativity and history-had arisen because the implications of the subject had been minimized, because what was said was more important than enunciation, and he thought that the Chomskyan model of competence and performance "indirectly facilitated this reintroduction of the subject,"3? which he considered necessary. Jean Deschamps presented the structuralist theses in psychoanalysis-Lacan's notions-explaining the respective roles of metonymy and metaphor, which allowed for "a coherent theory of the status of the unconscious."38 But Deschamps was critical of an approach that eliminated the dimension of lived experience by relegating it to the role of an insignificant epiphenomenon, and therefore losing even the Freudian notion of repression as a dynamic phenomenon by separating conscious and unconscious as two incompatible languages. Other articles led to a critical dialogue with Levi-Strauss's ideas. The issue showed how seriously the PCF took the structural challenge to Marxism, and intended to respond to it.

The Structuralist Answer to the Crisis of Marxism In 1967 and 1968, La Nouvelle Critique and Les Lettres françaises took advantage of the relative eccentricity of their position with regard to the party leadership to give broader coverage to the structuralist event. In March 1968, a debate moderated by Christine Buci-Glucksmann, Louis Guilbert, and Jean Dubois at La Nouvelle Critique raised the question of whether this was "a second linguistic revolution." For Jean Dubois, Chomsky "seemed to reintroduce movement into the dead structure, a dynamic rather than a static approach."39

Antoine Casanova brought to La Nouvelle Critique the new methods of the Annates school. Looking at the relationship between history and the social sciences gave many historians the opportunity to write in the journal and led to another collective work, Today, His-tory.40 The Annates school clearly appeared to take a middle road with respect to structuralism. A historical dialectic remained possible, although the primary goal was the search for structures. Thanks to the Annates school, some reconciliation and combination between structure and movement was possible.

Above all, structuralism won over Les Lettres françaises. Pierre Daix and Raymond Bellour wrote articles to familiarize readers with the progress made in different social sciences. Benveniste, however little drawn to media pronouncements, agreed to an interview with Pierre Daix on July 24, 1968. He was surprised by the excitement and interest aroused by a doctrine that was both poorly and tardily understood since it was already forty years old in linguistics, where "things had already developed beyond structuralism. "41 Pierre Daix had, however, become the most resolute defender of structuralism: when

Mikel Dufrenne published ForMan,42 in which structuralism was made the guilty party, he went to the battlements to defend it.

He attacked the elimination of man to the benefit of the system. He linked structuralism and technocratism, and saw in this kind of thinking a return of nineteenth-century scientism. For Foucault, wrote Dufrenne, "man is only the concept of man, a fading figure in a temporal system of concepts."43 Pierre Daix replied that this decentering was nothing more than a demystification for structuralists. Dufrenne brought together all the elements of structuralism that similarly intended to dissolve man. "Between Heidegger's ontology, Levi-Strauss's structuralism, Lacan's psychoanalysis, and Althusser's Marxism there is clearly a certain common theme that, in a word, concerns the mar-ginalization of lived experience and the dissolution of man. "44 Dufrenne argued for a humanism that, in Pierre Daix's view, resembled what scholars claimed for God in the nineteenth century; structuralism, to the contrary, needed to "substitute the understanding of man's condition for his privileges, and to understand his condition in all senses of this word."45

While the official Marxist current, that of the peF, tried to consolidate resistance to structuralism, the fissures were increasing between those who, like the Althusserians, had chosen to adopt the structuralist perspective in hopes of renewing Marxism, and those who adopted structuralism as a way of leaving Marxism behind. Thanks to this confrontation, a number of commonalities in the two approaches became clearly apparent, and would link their destinies: initially a triumphant destiny, in 1967-68, but which would quickly go into a decline that affected structuralism as well as Marxism.


Media Success: A Criticism-fed Flame

As structuralism was being theoretically fissured, its media triumph was being consecrated. The media image was one of a convivial gathering of traditionally clad gentlemen. These years, 1967-68, witnessed a veritable "structuralist contagion,"} despite the fact that the structuralist feast was finished. But "had it ever happened? The diners deny ever having been there."?

The two important weeklies of the period, L'Express and Le Nouvel Observateur, gave the phenomenon a lot of press, although UExpress was the more critical of the two. jean-Francois Kahn humorously described the careful conquest of structuralism which had already found its credo with The Elementary Structures of Kinship, its king in Levi-Strauss, its completely horrible language and linguistic alphabet, its best-seller (The Order of Things): "Structuralism is the apex of the imperialism of knowledge."3

In La Quinzaine litteraire, Francois Chatelet wrote about a pseudo-school and an artificial unity established by unscrupulous adversaries. He nonetheless wrote a long article, illustrated by Maurice Henry's famous drawing, in response to the question, "Where is structuralism at?"4 Chatelet considered the different elements of the movement known as structuralist and concluded that no homogeneous, doctrinal corpus could easily be discerned. "We can barely speak about a method."! And yet he saw a common feature in the rejection of empiricism. For to remedy the crisis of ideologies, all of these authors sought not to replace the great, dead Subject of history (the proletariat) with small facts belonging to empirical sociology, but to define scientific methods of investigation in order to know "what we can, in fact, accept as a fact."> After having denied any structuralist unity, Chatelet acknowledged that despite its differences, structuralism did exist, since he saluted a "French" thought that was in the process of rediscovering, in disparate fashion, "the rigor of a theoretical vocation."?

Le Nouvel Observateur became a particularly effective trampoline for the structuralist adventure. Levi-Strauss answered Guy Dumur's questions, and when the ORTF (Office of French Radio and Television) broadcast a show produced by Michel Treguer on ethnology on January 21, 1968, Le Nouvel Observateur printed Levi-Strauss's remarks as well as his definition of structuralism. Benveniste also agreed to an interview with Guy Dumur at the end of 1968 and voiced his optimism about the evolution of all the social sciences. He perceived portents of a grand anthropology, in the sense of a general science of man, taking shape." When Foucault reviewed Erwin Panofsky's work? in Le Nouvel Observateur, the journal's editorial team presented the article this way: "This language and these methods seduced the structuralist Michel Foucault.I't?

Le Magazine litteraire presented a long article by Michel Le Bris in 1968 entitled "Masterpiece: Saussure, the Father of Structural-ism,"l1 illustrating the major tenets of Saussure's thinking by a series of photographs of the four musketeers of structuralism, labeled "Saussure's heirs."

Television was only slightly less involved in the celebration, but when Gerard Chouchan and Michel Treguer brought together Francois jacob, Roman jakobson, Claude Levi-Strauss, and Philippe L'Heritier in front of the camera to debate "Living and Speaking" on February 19, 1968, it was a real event.

Structuralism: A "Religion of Technocrats"? The invasion reaching from research laboratories to editorial rooms seemed to reduce structuralism to a single form of expression. This provoked a certain number of reservations if not to say exasperation, a mixture of theoretical rejection of and irritation with a discourse that, once it became dominant, went from theorization to a certain intellectual terrorism, disdaining any adversarial arguments as simply imbecilic.

Among those who expressed a discordant note in the concert of praises was jean-Francois Revel, the philosopher become chronicler at L'Express, and responsible for culture in the ghost cabinet of Francois Mitterrand. Revel had published a polemical essay in 1957, Why Philosophersit? and had already radically criticized Levi-Strauss's work. He attacked his formalism, a system that was too abstract and that progressively slid from sociological considerations into an ethnological discourse, suggesting, beyond the description of behavior, the existence "of a mental and sentimental system that is not to be found there."13 In 1967, when he reviewed From Honey to Ashes, the second volume of Levi-Strauss's Mythologiques, Revel called Levi-Strauss a Platonic thinker in the realm of sociology. The key to Levi-Strauss's method was the assumption that what is hidden is real, whereas what is commonly called reality is the illusion from which we must divest ourselves. Opposed to functionalism, Levi-Strauss "formalizes, geo-metrizes, algebrizes."14

Somewhat later, jean-Francois Revel reviewed the work of the Marxist philosopher Henri Lefebvre, who attacked structuralist ideology as the expression of the arrival of technocratic power.l- He did not share the Hegelian-Marxist assumptions of Henri Lefebvre, but nonetheless accepted the analogy between structural thinking and the society being prepared by technocracy, and entitled his article "The Religion of Technocrats."16 Passive consumer society and the dialogue-less communication of modernity concentrated the power of social laws in the hands of a machine that escaped human control and that had no other goal than to reproduce itself. "Politics is no longer a battle but an observation. "17 Thus, structuralism would be the continuation at a theoretical level of this technocratic society, a real white-collar opium. Similarly, with structuralism, the individual escapes the meaning of his own acts since he is already spoken before being. Linguistics operates as the basis of all science by suppressing language's referentiality.

Later, Revel deplored "the death of general culture." 18 He saluted the linguistic takeoff at the beginning of the century thanks to Saussure, a veritable "Galileo of this metamorphosis,"19 but regretted that the emancipation of the social sciences had diluted the notion of general culture a little bit more every time and that in order to become sciences, the humanities had to stop being human. Going against the current with respect to those who saw structuralism as the decisive shift toward scientificity, Revel saw rather the natural tendency of any philosophical doctrine, its ability to infiltrate and influence everything during any given period of time with a certain language that quickly became "an Esperanto into which all disciplines were translated."20 Claude Roy did not attack the four musketeers in his article in Le Nouvel Observateur. He did, however, criticize the use that was made of their thinking, and the application of the "structuralist source or logic to strange mixtures."2l He attacked the false Levi-Straussians and denatured heirs of Althusser who were making a very curious use of structuralism in the Latin Quarter, and particularly in the Cahiers marxistes-leninistes. What they had drawn from the structural lesson was simply the fact that only the relationship between terms counted, whereas the terms themselves did not. Limiting structuralism to this postulate opened the door to total confusion, making it possible, among other things, to present the Moscow Trials as terms to be contrasted to one another without defining any of them. "Alice in Wonderland always asked for the definition of the terms being used. This concern is not the one most of the world shares today. The pseudostructural delirium in literary criticism and political theory clearly demonstrates this. "22

Another critical voice raised in 1968, even if he recognized that the structural method was valid in certain limited cases, was that of Raymond Bourdon. He adopted Karl Popper's theories on falsifiabil-ity as an indispensable criterion of scientificity, and listed the different uses of the notion of structure, judging their validity by their ability to be verified. Boudon argued that there could be no general structuralist method, but only particular methodologies, which could be applied to specific disciplines. He contrasted those for whom structuralism was a simple operational method (Levi-Strauss, Chomsky) with those, like Barthes, for whom structuralism was a simple fluid. He insisted on the "polysemic character of the same notion,"23 which prevented any claim of a unique doctrine. Boudon found the notion of structure to be particularly obscure, although it could be used to construct verifiable hypothetico-deductive systems, as was the case of C. Spearman's factorial theory, and jakobson's phonology. Indirectly, Jakobson could deduce the order of complexity of phonemes, but "not the necessary coincidence between this order and the order of appearance of phonemes in a child, for example."24 The structural method applied to no specific object, but to different objects to which a more or less experimental and verifiable method could be applied. Boudon's angle of attack targeted every quest for an essence behind the structure, for the revelation of the hidden side of the visible world. But his criticism, which sought to impose some limits on the application of structuralism, went unheard in the euphoria celebrating the vague ambitions attributed to the promoters of structuralist thinking.