• 1 The author wishes to thank Jérémy Sinigaglia, Adrien Thibault, Julien Duval and Mathieu Pastor for (...)

1Although it was initially dedicated to the “Economic and Social Sciences,” Section VI (roughly the equivalent of a department) of the École pratique des hautes études (EPHE) is known for being one of the institutions where research on literature was conducted1. It was the locus of the emergence of literary theories that developed in the wake of structuralism as well as the cradle of a number of sociological approaches to literature and the arts. The reception of some of these studies has contributed to entrenching the oppositions between the so-called “internal” and “external” approaches to literary works. This article examines how studies on literature fit within the disciplinary regimes of the EPHE’s Section VI, in an effort to understand how they shifted from sociology to literary theory based on structural linguistics and semiology. In the process, they left behind a speciality—the sociology of literature—within a discipline—sociology—to instead follow a trajectory that may qualify as a “‘regression’ into professional purity” (Abbott 2001: 146). Here the concept of “disciplinary regime” reflects a twofold approach to the “disciplinary division of intellectual labour” (Heilbron 2004; Lamy & Saint-Martin 2011). First, it draws attention to the fact that disciplinary divisions are privileged instruments for the regulation and governance of intellectual and scientific output. Second, it emphasizes the temporal nature of these disciplinary divisions, reminding us that disciplines, as “relatively stable academic units” (Convert & Heilbron 2005), routine “frameworks of perception” of scientific activity and “institutional frameworks” for the organization and reproduction of knowledge (Godechot 2011), are not set in stone.

• 2 The collation des grades is the procedure through which the national degrees awarded by some higher (...)

4In the French academic field, Section VI was close to the intellectual field, and overall its members had less temporal power than their colleagues situated at the temporally dominant pole of the academic field (Bourdieu 1984). In literature-related domains, it included few holders of the agrégation (a highly selective competitive examination for teachers and lecturers), had ties with international networks rather than national bodies, relied on teaching staff from somewhat modest backgrounds, and promoted new disciplines rather than the traditional humanities. As it did not fall under the collation des grades procedure2, it initially only awarded the EPHE degree. The introduction of a doctorate in literature in 1958 increased its draw. The fact that theoretically no qualifications were required to attend the seminars and that the degree awarded had little value in the academic market led the institution and its members to stick to distinctive postures in the symbolic struggles that rocked the academic field (Kauppi 1996). The hiring of teachers at the school was not tied to disciplinary affiliations. Olivier Godechot has shown that between 1960 and 1965, scholars were likely to be invited to sit on PhD examination panels based on their familiarity with a common topic, and subsequently with the study of the same “cultural area” between 1966 and 1970, rather than because they belonged to the same discipline (Godechot 2011). Yet, while this disciplinary indistinction was both a deliberate policy and an institutional discourse that reflected the Annales School’s approach, a number of social, organizational, and intellectual dynamics manifested distinctly disciplinary rationales and aligned the institution’s transformations with those at work in the field of higher education and research, which prevents us from looking at Section VI as an institution above the fray, entirely “outside the scope of the ordinary laws of academia” (Revel & Wachtel 1996: 22).

• 3 See for instance Fernand Braudel’s project of creating a “Faculty of economic, social and political (...)

7As soon as these disciplinary distinctions appeared, some courses escaped labelling and were thus potentially geared towards an unspecified audience. Their increase resulted most notably from the combined growth of student attendance and of the number of courses, and their qualitative evolution reveals the shifting nature of the classification of knowledge as “general,” “cross-cutting,” or “specialized.” While there was just one such unlabelled course in Section VI in 1960-1961, there were seven in 1965-1966 (Table 1). They reflected the section’s longstanding concern3 to offer tools and methods applying to all social science in the tradition of the Annales School. “Mathematical methods” quickly prevailed as a tool of choice for formalization; Jacques Bertin’s “Cartography Method” seminar moved from “Methods” to “History and Geography,” and went on to be renamed “Cartography Methods” and then “Semiology of Graphics.” Semantics and semiology made an initial breakthrough in this cluster. Unlabelled courses disappeared in 1966-1967, as all courses were given new disciplinary affiliations (see Table 4). Some courses with no specified disciplinary affiliation reappeared in the early 1970s under the umbrella name of “Social Science Methods and Techniques.”

• 4 Fouchet’s reform terminated preparatory courses and introduced discipline distinctions at an earlie (...)
• 5 The Higher Education Orientation Act or Faure Act reformed the administrative structure of universi (...)
• 6 The so-called “non cumulative” directeurs d’études (DEs), meaning those who mainly taught in Sectio (...)
• 7 EHESS archives, Louis Velay papers, file “Rapport Roncayolo (novembre 1968).”
• 8 Master of Advanced Studies.
• 9 EHESS archives, Louis Velay papers, file “EPRASS-Historique et finalité (1968).”

8The disappearance of courses without a specified affiliation to a discipline in the mid-1960s was one of the consequences of a double trend towards disciplinarization and the reconsideration of disciplinary boundaries and the rising importance attributed to pedagogical matters. These transformations were not specific to the institution; on the contrary, they appear to have been one of the ways in which the demands placed on the academic field (Cardon-Quint 2015; Chevalier & Encrevé 2006; Drouard 1982; Hauchecorne 2016; Soulié 2012) were “refracted” (Bourdieu 1991). The so-called “Fouchet reform4” of 1966 promoted extreme disciplinary specialization; two years later, the “Faure act5” was intended to establish the conditions for pluridisciplinarity (Dormoy-Rajramanan 2017). Section VI responded by establishing a relationship of complementarity and distinction with the CNRS and the universities, with whom it shared many members6. The school gave a largely pedagogical meaning to its discipline labels. It strongly affirmed its vocation for teaching and research and the seminar model, whose titles and modes of classification differed from the disciplinary distinctions in the name of “cutting-edge research7,” in contrast to the lecture-based form of teaching that prevailed in universities. During that period, several pedagogical initiatives reformulated disciplinary questions and were approached in light of what the universities were doing. These included the introduction of preparatory courses for advanced social science research (EPRASS) in 1965 and of the advanced social science studies degree (DEASS), modelled on the advanced studies degree (DEA8) offered by science faculties. The EPRASS was presented as a course that was unavailable at humanities faculties9, in disciplines that were seldom or not at all taught at- universities (such as social anthropology, sociology, and linguistics).

• 10 Lanson established literary history as a discipline at the Nouvelle Sorbonne at the turn of the twe (...)

9Until 1959-1960, there were no courses specifically dedicated to literature in Section VI. Yet literature was not entirely absent, and was never entirely neglected: Lucien Febvre and Roger Bastide, among others, had repeatedly called for social science and literary studies to be combined. In the 1950s, the number of student dissertations explicitly covering literature was fairly small in Section VI. In 1956, Georges Balandier and Claude Lévi-Strauss were the rapporteurs (external examiners) for a dissertation on Black African Theatre and its Sociological Functions; in 1959, Georges Balandier oversaw another on Duala literature and its social function; and in 1960, Jean-Claude Vernant oversaw Poetry and philosophy in ancient Pythagoreanism. Section VI’s first course dedicated to literature—albeit not exclusively—was introduced in 1959-1960: Lucien Goldmann’s seminar “Sociology of Literature and Philosophy” was classified in the “Sociology and Ethnology” cluster. The sociological approach to literary texts, combined with philosophy, was literature’s path towards an official introduction into Section VI—there was no historical patronage, unlike in academic literary studies, which were structured using a chronological model and inspired by philology and the social history of literature developed by Gustave Lanson10, who was initially close to late nineteenth-/early twentieth-century Durkheimian sociology (Compagnon 1983; Sapiro 2004). Throughout his career, Lucien Goldmann devoted part of his teaching to the sociology of literature.

• 11 The discipline grew during that decade: the Revue française de sociologie was created in 1960, and (...)

12The “Sociology and Ethnology” cluster, in which these new specialists of literature were placed, became “Sociology, Ethnology and Social Psychology” in 1965-1966 (Table 3). The most prominent post-war sociologists, who were from or connected to the Centre d’études sociologiques (including Georges Gurvitch, Georges Friedmann, and Alain Touraine), had an important role in it and connected it to the sociological work being conducted outside of Section VI. The rise of sociology in Section VI reflected broader trends within the discipline. While sociology was previously a certificate within the philosophy degree, its institutionalization ensured it lasting student attendance in a relatively autonomous curriculum: the licence (bachelor’s) was created in 1958 and the first theses were defended in the early 1960s11. The growth of sociology was also connected with the expansion of universities in the 1960s: the number of PhDs awarded began increasing by the middle of the decade, and so did the number of CNRS researchers—among whom the number of sociologists rose from 50 to 130 between 1960 and 1975 (Heilbron 2015: 152 sq.). Section VI’s staff also expanded. Sociology gained an increasingly prominent place there, branching out into several theoretical programmes. The rise of structuralism was a fairly minor development for sociology, whereas it allowed other disciplines such as linguistics to renew and structure themselves using new epistemological bases.

• 12 EHESS archives, Louis Velay papers, file “Rapport Roncayolo (novembre 1968)”.

14Several factors go some way towards explaining the new discipline distinctions, starting with the aforementioned transformation of the academic field. Considering the close ties between Section VI and the CNRS, the creation of associate laboratories and research teams in 1966 should also be taken into account, as they favoured and ultimately institutionalized disciplines and groupings of specialists (Picard 1990). The members of Section VI also expressed their “intention to define themselves better in relation to the university system12” as social science expanded in universities and the institution lost its virtual monopoly over training in the related disciplines, which had until then been neglected by faculties of literature. Lastly, the school was forced to depart from the universities’ discipline labels while retaining clearly defined understanding labels of its own that could be converted into resources in the academic field. The distribution of “Sociology and Ethnology” courses in the new subdivisions followed existing thematic and methodological lines (Figure 1). Sociology secured the highest number of courses (twelve), all connected to directions d’études (Directorates of Research). More than half of the sociology DE had been taught for at least ten years in Section VI (see Appendix).

• 13 It would be tempting to see the construction of the SSL subdivision, if not the development of soci (...)

15In quantitative terms, the reassignment of “Sociology and Ethnology” classes into finer disciplinary subdivisions showed that sociology alone did not cover the entire “Sociology and Ethnology” cluster, but rather served as a unifying label masking the scattered nature of the “refuge discipline” at the time (Bourdieu 1980: 19; 1984: 159). The modalities of this reassignment reveal the key role of the teachers’ seniority and grade in the imposition of a speciality, subdivision or discipline. Indeed, not a single subdivision was without a DE. Those containing only one course involved longstanding members of the institution and/or holders of positions of power in the academic field. Lastly, disciplines tended to be increasingly divided up between several labels spearheaded by key figures in various research fields (Lévi-Strauss for anthropology; Friedmann and Touraine, among others, for sociology; Barthes and Greimas for semiology). One of the new labels stood out: “Semantics, Semiology and Linguistics” (SSL), in which none of the most senior members of “Sociology and Ethnology” was involved. Out of the three teachers from that cluster active in the new subdivision, Barthes was the only DE: André Miquel was a maître assistant (assistant lecturer) and Georges Dumézil—a professor at the Collège de France—was a chargé de conférences extérieur (guest lecturer). The SSL subdivision featured solely recent recruits of “Sociology and Ethnology.” Yet its place among the new subdivisions was notable—in terms of staff size, it came fourth behind history, economics, and sociology. It included two of the three specialists of literature at the school13. However, SSL was not the only venue hosting researchers with an interest in literature, since Goldmann remained affiliated with sociology.

• 14 Higher-level competitive examinations for teachers and professors.
• 15 The literary critics Georges Poulet, Jean Starobinksi, and Jean-Pierre Richard, and to a lesser ext (...)

Gaëtan Picon (1915-1976)’s opposition to the theoretical avant-garde was probably one of the reasons why his role was overlooked in the historiography of 1960s intellectual life, although during that time he was situated at the intersection of the literary, artistic, and academic fields. Born into a family of Bordeaux industrialists, a holder of the agrégation¹⁴ in philosophy, an activist in the socialist youth movement and in the Popular Front, Picon, a former student of Gurvitch, was close to André Malraux, who appointed him as director-general of arts and letters in 1959. He ended his political career in 1966—the year he was elected as DE in Section VI with Goldmann’s and Lévi-Strauss’ support. He began teaching at the Paris school of fine arts and pursued a parallel international career. He worked for several publishing outfits (Mercure de France, L’Éphémère, Skira’s collection « Les Sentiers de la création ») and wrote on literature (André Malraux, 1945; Georges Bernanos, 1948; Panorama de la nouvelle littérature française, 1950; L’Usage de la lecture 1960-1963) and art criticism (including articles and books on Ingres, Picasso, Dubuffet, Cézanne, and surrealism). He was close to the literary criticism of the Geneva School15 and to the poets Yves Bonnefoy and André du Bouchet, and approached literature through the perspective of modernity and the subject as creator. He died in 1976 when he was due to become the next director of the Académie de France in Rome (Callu 2011: 519).

• 16 Annuaire 1968-1969: 120.

17The presentation of Gaëtan Picon’s 1966-1967 seminar describes an examination of novelistic and autobiographical narration in the first half of the nineteenth century that places the subject at the centre of literary works. The following year, he considered the “modern notion of art and literature16” based on texts by Baudelaire and lectures by Jean Starobinski and Yves Bonnefoy. In the early 1970s, Gaëtan Picon’s teachings proposed a historical approach to structural questions. He then worked alongside Marc Soriano, a holder of the agrégation in philosophy, a former student of Jean Piaget and the author of a PhD thesis on Charles Perrault’s fairy tales. On examination panels for master’s and doctoral theses, Gaëtan Picon often sat with Pierre Francastel, specialist in the sociology of art, Jean Cassou, who began co-teaching the seminar for that speciality with Francastel in 1965-1966 and went on to teach his own seminar on “Social Conditions and the Spirit of Surrealism” in 1968-1969, and Barthes and Goldmann. Picon evaluated a dissertation on Marc Chagall supervised by Jean Cassou in 1968 alongside Francastel, and one on “writing and painting” also supervised by Cassou, in 1969, with Barthes. The same year, Cassou and Picon evaluated a dissertation on existentialism and Marxism in Paul Nizan’s work, supervised by Goldmann, and Picon and Algirdas Julien Greimas evaluated another on Franz Kafka, supervised by Barthes. Their joint presence on examination panels illustrates several trends, namely: the formation of a network of specialists on literary and artistic themes across the new discipline boundaries; the increasing number of works on literary and artistic themes and the diversification of approaches; as well as the relative scarcity of specialists of literature and art in the institution, leading them to collaborate in spite of their different approaches and discipline affiliations.

• 17 Annuaire 1967-1968: 244-245.
• 18 Ibid.: 245.

The 1965-1966 seminar on the sociology of culture was devoted to the relations between “systems of education and systems of thought” and the production, reception and diffusion of “culture” and artworks. Regarding literature, the presentation mentions a lecture by Jean-Claude Chamboredon on the “social conditions of the creation and diffusion of literary works.” In 1966-1967, the seminar focused on the “critical examination of ethnological and sociological studies pertaining directly or indirectly to the diffusion of culture, and of literature in particular17” for the purpose of “identifying some of the conceptual schemes that are indispensable for the development of a theory of cultural diffusion18.” Lectures addressed various topics (health, teaching, etc.). The uses of the concept of culture in sociology were discussed, based in particular on the writings of Raymond Williams. The diffusion and reception of culture was presented as the main research focus of the group of researchers and of the seminars held in the following years.
Like symbolic goods as a whole, literature had a prominent place in Bourdieu’s work at a very early stage, but it only became a distinct obbject of inquiry at a later point (Martin 2010). In the late 1960s, Bourdieu began relying heavily on literature in his effort to construct and formalize the concept of field (Bourdieu 1966, 1971, 1975; Sapiro 2017). Distinction (1979) examined reading as a cultural practice. In the 1980s, Bourdieu’s body of writings on literature expanded (e.g., Bourdieu 1984, 1985, 1991), leading up to the publication of Rules of Art in 1992. Lastly, Bourdieu’s working groups in Section VI and subsequently at the EHESS and mostly at the ENS Ulm were venues for collective research on literature. Before it gained international exposure, the sociological approach to literature developed by Bourdieu, but also beyond his circle, was primarily implemented, complemented, and discussed by former students, colleagues, and members of these working groups.

• 19 Erwin Panofsky’s Gothic Architecture and Scholasticism and of Richard Hoggart’s The Uses of Literac (...)

20In Section VI, Goldman’s sociology of literature appeared not to find a place either in sociology or in the emerging pole of scholarship on literature and the arts. To understand the reasons for this lack of disciplinary affiliation, we need to situate it within the broader development of the sociology of literature, which indeed was part of a “wider movement, namely that of sociology’s growing concern for the symbolic dimensions of society” (Dirkx 2000: 37). The emergence of this speciality was encouraged by the creation of a Ministry of Culture in 1959. Studies on photography and museums initiated by Pierre Bourdieu (1965, 1966), Jean Duvignaud’s research on theatre, Raymonde Moulin’s work on art markets as well as Pierre Francastel and close colleagues contributed to this movement, which also materialized with the importation of foreign works19. The sociology of literature had long been torn between two disciplines whose respective developments were at odds (Lepenies 1990)—“too ‘sociological’ for literature scholars and two ‘literary’ for sociologists” (Sapiro 2014). The very instability of the “sociology of literature” label, which sometimes became “literary sociology” (a term used in Gustave Lanson’s work, and on occasion by Lucien Goldmann and Jean Starobinski) or the “sociology of the literary fact” (Escarpit), is a clear sign of its weak institutionalization. In light of its relative flexibility regarding disciplinary distinctions, Section VI was a favourable venue for its development. Yet it did not entirely escape the disciplinary rationales at work in the division of intellectual labour. In the 1960s, the sociology of literature detached itself from historical approaches (especially those of Gustave Lanson and Lucien Febvre) in several ways. In 1960, Robert Escarpit (1918-2000), who was in charge of comparative literature at the University of Bordeaux and an advocate of an empirical sociology of the “literary fact,” created the Centre for the Sociology of Literary Facts in Bordeaux. He received funding from UNESCO, whose book development programmes he then promoted. Having authored the first introduction to the sociology of literature in the famous « Que sais-je ? » collection (1958), he played a key role in the speciality’s initial development in French-speaking spaces (Van Nuijs 2007). Lastly, Marxist approaches to literature, which Sartre explored in the early 1970s, were also key to the work of Goldmann and the promoters of an Althusser-influenced brand of Marxism, who extensively discussed the social dimensions of literary works but did not call themselves sociologists.

• 20 Born in Geneva in 1942, Jacques Leenhardt had previously worked with Jean Piaget. He studied philos (...)
• 21 Annuaire 1967-1968: 258.
• 22 Annuaire 1968-1969: 320.

23According to the directories, Goldmann’s seminar was regularly attended by around fifteen auditors. Students who went on to pursue literary careers included the Czech-born researcher and scholar Pierre Václav Zima (1946-), who was also a student of Jean Cassou, and who turned to sociological criticism and the analysis of critical theories of literature. Two other students of Goldmann also played a prominent role in the diffusion of structuralist and post-structuralist literary theories: Julia Kristeva, an academic and psychoanalyst with ties to Barthes and the journal Tel Quel, and Sylvère Lotringer, a promoter of French Theory whose more distant relationships with Barthes and Goldmann turned out to be crucial to his training and career. Starting in 1966-1967, Goldmann’s seminar was complemented by a class led by his assistant Jacques Leenhardt20. Leenhardt hoped to “open up a dialogue with certain so-called a-historical theories of literature21.” He published research on the relations between the sociology of literature and structural semantics, and on the Nouveau Roman (the new novel), airing views that clashed with those of the Tel Quel group (Lecture politique du roman, 1973). His seminar, which excluded philosophy and focused on the sociology of literature, illustrated the Goldmannian sociology of literature’s attempt to carve a space for itself in literary criticism. Leenhardt positioned himself in relation to literary criticism—not sociology. He believed it necessary to “help students find their way in the maze of what is commonly called ‘critical languages’” to “clarify the theoretical and epistemological positions on display in each of these languages as precisely as possible.” He found semantics to be a “major empirical confirmation22” for the sociology of literature.

• 23 One was Michael Löwy, who wrote on literature on several occasions. He also co-authored an introduc (...)
• 24 Interview with Jacques Leenhardt, 5 July 2014.
• 25 Ibid.
• 26 Ibid., and Interview with Annie Goldmann, carried out in collaboration with Quentin Fondu, 2 Novemb (...)

24Leenhardt became a maître de conférences (lecturer) at Section VI in 1974. He succeeded Goldmann, who had died in 1970, as the head of the “Sociology of Literature Group,” while other students of Goldmann went separate ways to pursue their research23. Leenhardt inherited an audience of students and a small structure that included a secretary (Brigitte Navelet), a research engineer (Martine Burgos), and an office in the rue M. le Prince in Paris. Leenhardt was elected DE in 1995, roughly twenty years after becoming a lecturer, which did not help further his career. His outlook on the latter was reminiscent of Goldmann’s—he went back and forth between the assertion of his autonomy and the observation of a sense of isolation, compensated by the transnational circuits of the sociology of literature. Locally, not being a part of the “54” (the MSH on Boulevard Raspail) led to him being “marginalized24” and to some degree deprived of the institution’s “intellectual power25,” and socialization venues—like Goldmann, who would reportedly meet his students outside of Section VI26. Leenhardt says he did not work with the sociologists in Section VI, which further gave him a sense of being “marginal,” again bringing him closer to the figure of Goldmann as an exiled immigrant, seen as lacking the resources needed for a successful integration into Parisian academic spaces—a stance that can also be understood as a way of making a (scientific) virtue out of (social) necessity (Ben-David & Collins 1966, Bourdieu 2001: 152):

• 28 Annuaire 1967-1968: 222.
• 29 Ibid.: 226.
• 30 Ibid.
• 31 Godechot shows that between 1971 and 1975, members of the SSL subdivision were somewhat isolated—ex (...)

27Two main trends stand out in the subdivision’s early years. First, the increasing proportion taken up by literary themes: they were the subject of Barthes’s seminar beginning in 1967-1968, which led to his semiotic analysis of Balzac’s Sarrasine, S/Z (1970); the same year, Greimas examined narrative structures; then, Jean-Claude Gardin studied the application of automated information processing systems to literary texts and Louis Marin analysed relations between images and texts. In 1969-1970, Roland Barthes was replaced by Gérard Genette, who launched a seminar on “Literary Semiotics,” thus strengthening the focus on literature. A maître de conférences at Section VI in 1967, elected DE in 1972, Genette had by then already published the first two volumes of his Figures (1966, 1969) and was about to distance himself from Tel Quel and create the literary theory journal Poétique in 1970, alongside Hélène Cixous and Tzvetan Todorov. New developments in linguistics were not overlooked—being tackled on courses by Eliséo Veron and Sophie Fisher, François Récanati and, most importantly, Oswald Ducrot. There was a regularly reasserted effort to develop general theoretical approaches to linguistic, literary, and artistic objects drawing on semiology, semiotics, and semantics, which were also taught in the EPRASS. In 1966-1967, Rygaloff claimed that the “issues of definition and conception that arise regarding all types of linguistic unit have mainly been addressed in terms of theory, and only incidentally in terms of applications of the theory to Chinese28.” Greimas sought to “specify the status of theories of language, as well as the nature and organization of the epistemological space they cover29,” in the process encompassing a wide variety of possible applications of studies by “specialists, or semioticians looking to specialize30.” A pole of specialists of language was taking shape, appearing relatively autonomous even though it had fewer members than other disciplinary groups31. The SSL subdivision was built around similar theoretical and methodological approaches applied to varied objects, whereas the Goldmannian sociology of literature was trying to assert its place as a speciality within sociology.