What is meant by “autonomy” in the human and social sciences? The concept’s multiple meanings make it susceptible to mishandling. These meanings refer to distinct traditions, from the theme of political autonomy, which can be traced back to autonomia in Ancient Greek cities, to that of individual autonomy, a concern that grew along with the development of the Protestant ethic and humanist thought, as theorized in Kant’s ‘autonomy of reason.’ Autonomy also holds a place of choice in sociological reflections on the conditions of production and circulation of symbolic goods, which will be the focus of this paper: it is used to analyse the relationship between these goods and economic, social, and political conditions, without reducing goods to conditions. However, here too, meanings vary between the three strands of research that have systematically used and discussed it2 since the 1950-1960s: the sociology of professions, Marxist reflection theory, and field theory (1). Despite these differentiated, partly incompatible uses, there are overlap areas and possible bridges to cross between these approaches (2). They suggest proposals for a reasoned synthesis of these uses in sociological analysis of the production and circulation of symbolic goods (3).
1. Uses and Meanings of the Concept of Autonomy Pertaining to the Production and Circulation of Symbolic Goods
In the sociology of professions, the concept of autonomy is closer to the political sense of the word: it refers to the organizational form of professions whose ‘technical’ and organizational autonomy (over the control of access to the profession and practices) is recognized by the state (1.1). In the Marxist tradition, the concept of autonomy is a corrective to the theory that sees in artworks, as in religion and the state, a superstructure reflecting social relations of production (1.2), whereas in field theory, it refers to the (relative) autonomy of universes of production vis-à-vis external (political, religious, or economic) constraints (1.3).
The use of the concept of autonomy in the sociology of professions relates to its political sense of self-determination. This is based on the indigenous definition of professional autonomy—namely, the right and privilege granted by a governmental entity to a class of professionals, and to each of its certified members, to offer services without being subject to control. Beyond the exercise of the profession itself, autonomy pertains to the organization of the professional group, its recruitment, rules of operation, ethics, and internal discipline. For these organized professions,3 autonomy is the result of the state’s delegation of its powers of decision and intervention to representative bodies, particularly regarding the definition of the group’s boundaries—i.e., the inclusion of new members and the exclusion or sanction of those who breach rules of professional conduct.
The Anglo-American sociology of professions made this autonomy one of the distinguishing characteristics of professions, by universalizing the cases of the US and the UK, informed by liberal traditions, where it was written into law (for lawyers and doctors). These privileges were granted to these professions in return for their contribution to the public good. In his pioneering study on medicine, Eliot Freidson (1970: 71-72) explains that only professions have been given the right to control their own work and to declare outside evaluation illegitimate. According to Magali Sarfatti Larson (1977: XII) “this distinctive autonomy is, however, only technical and not absolute. Professions ultimately depend upon the power of the state, and they originally emerge by the grace of powerful protectors.”
Yet the history of professions shows that the exercise of these activities has taken different forms in bureaucratic and authoritarian states: in the French, German, and Habsburg empires, some intellectual services, such as teaching, were considered part of the civil service, which freed them from market demands and clientelism, and ensured that individual interests, especially economic ones, did not prevail over the general interest, and were kept in check ideologically (Siegrist 2004). This affiliation with the civil service gave specialists a monopoly, but placed them under a hierarchy. The situation of university professors in France remains the same today, but a clause exempts them from hierarchical evaluation, which is replaced by peer evaluation—a sign of the autonomy granted to them; in a further sign, the recruitment of new professors works the same way. The system of officiers ministériels [ministerial officers], which applies to notaries and auctioneers in France, combines relative autonomy and state control (Quemin 1997).4
Such a comparative socio-historical perspective (Macdonald 1995) calls for an embedding of the recognition of professional autonomy within the framework of the history of “professional development”5 conceptualized by Andrew Abbott (1988)—a history that includes the competitive struggles between professions over the monopoly of state-recognized areas of expertise (“jurisdiction”), leading to the “division of expert labour.”
The comparative approach (between professions and countries) also raises the question of the circulation of professional models. In France, for instance, the professional organization was imported under the Third Republic (1870-1940), against a background of liberalization (a law authorizing unions was passed in 1884, and associations were made legal in 1901) and secularization, wherein religion was replaced by nationalism as a vector of identity, and by science as a source of truth and knowledge (Sapiro 2006). At the time, doctors acquired a state-sanctioned social authority (especially vis-à-vis the clergy) (Léonard 1987: 5-19). Other professions were structured during that period, including journalism (Delporte 1998) and consulting engineers (Henry 2012).
However, the degree of autonomy granted to professions varies between regimes. The most authoritarian regimes tend to reduce autonomy as much as possible, as illustrated by the cases of the professions of engineer and teacher in Nazi Germany (Jarausch 1990), or doctor under the Vichy regime in France (Muel-Drefyus 1998: 301 sq.).
Eliot Freidson analysed the foundations of professional autonomy in his study on medicine. Rather than conceiving it as an intrinsic trait, he argued that autonomy was the outcome of a process, a conquest allowing professionals to acquire independence from the ideology of the dominant elites, but which is not acquired once and for all, and as such can be lost—since indeed autonomy is granted by the state. He introduces a distinction between technical autonomy, based on esoteric or complex knowledge, and socio-economic autonomy. The former constitutes a criterion for differentiating professions and other occupations: it is recognized in doctors, irrespective of their working conditions (including in the USSR) and largely conditions socio-economic authority (Freidson 1970: 23-24; 45).
Worth noting is that, where it exists, this autonomy was often granted in return for serving the authorities: for instance, the recognition of medicine as a profession libérale in France was closely linked to the recognition of the expertise of hygienists and the ability of doctors to ensure public order and health (Goldstein 1984). Autonomy also involves power, the authority of professionals over the “lay” clientele—as critical studies that built on the work of Freidson (1970) in the 1970s noted (Sarfati Larson 1977). Indeed, the technical autonomy granted by the state allows professionals to exercise a monopoly over knowledge in their jurisdiction and to not be subjected to outside control of content. This technicity is the basis of their authority. It is expressed in a distinct language: “jargon,” which according to Terence Johnson (1970: 56) has the double function of maintaining the professional group’s internal homogeneity and “increasing autonomy from outsiders, both competing specialists and laymen.” As Johnson (1970: 57) points out, it is professionalism that creates charlatanism (in the course of the struggles over the monopoly in an area of expertise), not the other way around.
Creative jobs do not enjoy such professional status. They are rarely taken into consideration by the sociology of professions—when they are, this is a “challenge,” to quote Freidson’s phrase (1986). While they do share some characteristics with organized professions, including the existence of representative bodies, these occupations lack other attributes, such as a specific training programme attested by a diploma in the case of writers and many artists, or a code of ethics in most cases. Nevertheless, the sociology of art was founded partly by addressing artistic activities from the perspective of creators’ careers (Moulin 2009), the division of labour (Becker 1980), and professionalization. Raymonde Moulin (1967, 1983) analysed transformations in the organizations of artistic life in France, from corporative organization in the fourteenth century, to academism in the seventeenth century, and paradoxical professionalization with the emergence of the market, which replaced skill—certified by the academy of fine arts—by vocation, recognized by new intermediaries: gallery managers. Philippe Coulangeon (1999), for his part, showed that in France the (unfinished) professionalization of jazz musicians, and by extension their differentiation from amateurs, has been supported by a range of actors outside the community.
Robert Merton raises a wholly different question in his analysis of scientists—that of the incompatibility between the ethics of science and state ideology in authoritarian regimes (Merton 1996). How can science remain autonomous in such settings? The idea of scientific autonomy requires professionals to be able to escape political diktats. This brings Merton closer to the Marxist problematic—from which he draws the idea of a margin of autonomy of the sphere of knowledge (Merton 1996: 257)—, and to field theory—of which he was one of the sources. After Marx, Karl Mannheim inscribed this autonomy in the position of “unanchored, relatively classless,” “free floating” intellectuals (or “socially unattached intelligentsia,” in Mannheim 2006: 128). This insight inspired the sociology of knowledge, from Merton’s considerations on the scientist’s ethos (1996: 267-277) to Bourdieu’s on the relative autonomy of the scientific field (1975; 2004), but was not picked up by Marxist thinkers. Merton was on the other hand criticized by the advocates of the Strong Programme in the sociology of science, for having dissociated the cognitive and social dimensions of scientific activity (Barnes, Bloor, Henry 1996; Callon & Latour 1991; see also Olivier Martin’s 2005 synthesis). The concept of field precisely aimed to overcome this tension between the cognitive and the social (Bourdieu 2004; see below).
In Marxist thought, the concept of autonomy emerged as a corrective to reflection theory. Under the materialistic postulate, literature contributes, as does religion, to the superstructure, which reflects relations of production. While it appeared reductive from the outset, reflection theory inspired in-depth consideration of the autonomy of artworks from social conditions and on the mediations between them (Goldmann 1970; Macherey 1971; Williams 1977; for a synthesis, see Sayre 2011). Is literature merely a reflection of the social world, as realist literature claims to be, or the expression of an ideology through a “world vision” or “worldview” (Weltanschauung)? For Mannheim (2006), ideology is indeed part of a worldview, a social group’s structure of consciousness or style of thought, divided into two categories: ideology, which legitimates the established order and facilitates its reproduction; and utopia, which refers to representations with a subversive function. Other thinkers have asked whether the “worldview” conveyed by literary and artistic worlds is the manifestation of the dominant class’s collective consciousness or if it echoes the contradictions at work in social relations of production. As the expression of these contradictions, Georg Lukács (1951/1999) contended, Balzac’s work may for instance give us a critical outlook on society regardless of the author’s political positions.
Drawing on research by Lukács, who argued that the relations between literary forms and the social situations in which they emerged are mediated by collective consciousness, sociologist Lucien Goldmann (1955 Engl. ed. 2013) considered that the real subject of the work is not the individual author, but the social group to which he or she belongs (family, profession, nation, and class). The group’s worldview constitutes the mediation between the economic and social infrastructure and the works: “A world vision is [...] the whole complex of ideas, aspirations and feelings which links together the members of a social group (a group which, in most cases, assumes the existence of a social class), and which opposes them to members of other social groups” (Goldmann 2013: 17). The members of the group attain this consciousness “in a more or less coherent manner,” “as a result of a particular social and economic situation, and which then gives rise to the set of activities performed by the real or potential community constituted by this social class” (Goldmann 2013: 18). There is thus a structural homology between Racine’s tragedies or the Pensées of Pascal and the “tragic” world vision of the Jansenists, an expression of the collective consciousness of the noblesse de robe. The first form of Jansenism indeed emerged at the time of the shift from limited to absolute monarchy, among the officers (especially lawyers) who depended on the state and could therefore not commit to active opposition, although they did distance themselves. Their tragic vision is an ideology—or rather a “total vision” encompassing ideology, affect, and behaviour—asserting that it is “impossible to live a valid life in this world” (Goldmann 2013: 1056). Goldmann sees literature, like the arts, philosophy, and to a large extent religious practice, as a “language”—a means for men to communicate with other beings, whose defining feature is that they are dedicated to the expression of specific contents, i.e., “world visions”—a world vision being “the conceptual extrapolation in the most coherent manner possible of the real, emotional, intellectual, and even motor tendencies of the members of a group. It is a coherent pattern of problems and replies which is expressed, on the literary plane, by the creation through words of a concrete universe of beings and things.” (Goldmann 2013 : 314) This is why, according to Goldmann:
“The aesthetic fact consists of two levels of necessary correspondence:
(1) the correspondence between the world vision as an experienced reality and the universe created by the writer; and
(2) the correspondence between this universe and the specifically literary devices—style, images, syntax, etc.—used by the writer to express it.”
In his view, the “coherence” of the extrapolation is what gives literary works value compared with other writings: “All valid literary works have an inner coherence and express a world vision” (Goldmann 2013 : 315). This capacity of extrapolation, which is to him the basis of representativeness, is only found in a few singular individuals.
Frankfurt School theoreticians also tackled the reflection problem. In his Notes to Literature (1958), Adorno (1992) argued that the most hermetic art can express a reaction against the devastation of language by commerce. However, as he emphasizes the ambiguity and polysemy of literary texts, he also shows that they resist ideological reduction.
At the same time, against the idea of a mystification of the masses by the productions of the cultural industries, alienating working-class consciousness, the founders of Cultural Studies considered the relations between culture and society with special attention to the mediation of conditions of production and reception. While, with the notion of “oblique attention” Richard Hoggart (1957/1970: 296) challenged the view that working-class people passively received these products, Williams (1961) studied the effects of the Industrial Revolution on literature and developed a programme for the historical sociology of the institutions of literary life (Williams 1977; 1981). Whereas in Goldmann’s work, the mediation between the economic and social infrastructure and the work was the world vision, in Williams the social conditions of production of the works take on that role. Nonetheless, Williams denied the distinction between literature and other forms of writing.
The concept of “production” spread and replaced the romantic ideology of creation. To Pierre Macherey (1971: 53-54), for instance, a literary work is indeed the “product of work” and the writer is a “worker of his text,” who “does not manufacture the materials with which he is working.” Yet Macherey sees the work as a “broken mirror,” imperfectly reflecting reality. At the same time, Macherey refused a purely materialistic approach to conditions of production, like his mentor, Louis Althusser. The latter indeed called for scientific reflection on “the processes that produce the ‘aesthetic effect’ of an artwork” in the April 1966 issue of the communist intellectual journal La Nouvelle critique (see Matonti 2005: 174). Claude Prévost then looked for a way out of the problem of realism in order to overcome reflection theory, identified with Lukács, and focused on content, to the detriment of form (see Matonti 2005: 172-175). He also recused the concept of “artistic fact,” and called on criticism to refocus on the analysis of language. In the April 1967 issue of La Nouvelle Critique, which was going through a theoretical aggiornamento (Matonti 2005), he published a piece on “Marx and myths,” in which he criticized Roger Garaudy’s interpretation of the introduction to Contribution to the Critique of Political Economy, in which Marx questioned the permanence of aesthetic emotion, and the “charm” of artworks, beyond their conditions of production. According to Garaudy, Marx had introduced a “relative autonomy” of the work there, allowing us to overcome the idea of reflection. Prévost, however, saw in Garaudy’s interpretation a form of idealism. For his part, he emphasized the “demythifying” function of literature and literary language.
In Lire le Capital (1965 Engl. ed. 2015), Marxist philosopher Louis Althusser and his disciples Étienne Balibar, Roger Establet, Pierre Macherey, and Jacques Rancière rejected the notions of causality underlying Marxist analyses – on the one hand, the mechanic causality presupposed by reflection theory, and on the other, the expressive causality entailed by the concept of “world vision,” which meant reading texts as allegorical models of society as a whole (for instance in Lukács’s work). They argued that associated with expressive causality, the very concept of mediation is problematic in that it establishes relations between different levels (superstructure and infrastructure, text and society). They proposed replacing these two types of causality with the idea of a “structural causality,” referring to the presence and immanence of the structure in its very effects, akin to the Darstellung concept in the Marxist theory of value. The concept of ‘ideological State apparatuses’ introduced by Althusser (1970 Engl. ed. 1984) was a means to theorize the idea of a relative autonomy of the superstructure from the infrastructure that determines it, and of the effect it may have in return on the latter, namely “reproduction,” which he illustrates by discussing the role of school.
In The Political Unconscious, US Marxist critic Fredric Jameson (1981) showed that in the literary domain, the two types of causality challenged by Althusser remain locally valid: indeed, did the replacement of the dominant model of the three-volume novel by the cheaper single-volume format following the late nineteenth century crisis of publishing (an external cause) not lead to altering the very form of novelistic narrative? And expressive or allegorical causality works on some level because the great narratives and leading conceptions of history are often inscribed in the texts themselves. According to Jameson, these texts are actually our main pathways to access an otherwise inaccessible historicity, which functions as an absent cause (this is how Althusser interprets Spinoza’s immanentism: effects are inherent in the structure—or in God). It is in that sense that they are sources to access our “political unconscious,” which must be subject to the work of interpretation. According to Jameson, Althusser’s main contribution resides in the relative autonomy he recognizes in the different social spheres, against expressive causality, which tends to reduce them to one another through the concepts of structural homology (for instance, in Goldmann’s work, between class situation, world vision, and artistic forms, or between the novel as a form and the everyday life of the individualistic society resulting from the market economy) or mediation (the institution of family as a mediation between the child’s experience and the class structure, in Sartre’s biography of Flaubert). However, as Jameson notes, the very concept of relative autonomy cannot be used without reflecting on the forms of mediation between different spheres.
Pierre Bourdieu proposed a comprehensive theory of the relative autonomy of spheres of activity with the concept of field. While he borrowed the notion of autonomy from Marxist thought, he gave it a different meaning, by linking it to Weber’s conception of the differentiation of social spaces as a result of the emergence of a body of specialists and to Durkheim’s analysis of the division of labour. Bourdieu was also inspired by the programme laid out by Levin L. Schücking (1931) who, in his Sociology of Literary Taste, addressed the role of authorities specific to the world of literature. It should be noted that none of these authors used the concept of autonomy, which actually seems incompatible with the functionalist conception of organic solidarity that underpins the social division of labour according to Durkheim. However, it can be reconciled with Weber’s analysis of the differentiation of domains endowed with their own references and rules, elaborated upon by Parsons (1939) and Luhmann (1991) in a functionalist systemic approach.
In Weber’s work, the concept of Eigengesetzlichkeit serves to characterize the Wertsphären, value-spheres (religious, economic, political, aesthetic, erotic, intellectual; see the “Intermediate Reflections,” “Zwischenbetrachtung,” first published 1915 and collected in his writings on the sociology of religions; Weber 1984) or Lebensordnungen (life orders); this concept refers to an intrinsic logic, or, according to Roth (1992: 457), an “autonomy” emerging from the process of differentiation of these spheres—whose values may enter in conflict—and of its institutions (see also the entry “value-spheres” in Swedberg & Agevall 2016). Another source for field theory, Ernst Cassirer’s theory of symbolic forms—language, science, religion, myths, arts—also postulates their autonomy and partial incompatibility, which renders them irreducible to logics (he speaks, for instance, of the “conceptual world” -Begriffswelt- of physics as “entirely self-contained”; Cassirer 1953: 26).
Bourdieu’s reflection on autonomy arose from a methodological question: can cultural domains be autonomized as objects of study? To him, this was only possible under the condition of studying the historical process of autonomization of these spaces of activity (whether the literary, artistic, or scientific field, etc.): “Obviously this approach can only be justified in so far as the object to which it is applied, that is the intellectual field (and thus the cultural field) possesses the relative autonomy which authorizes the methodological autonomization operated by the structural method when it treats the intellectual field as a system which is governed by its own laws,” he wrote in his very first article on the subject, “Intellectual Field and Creative Project” (Bourdieu 1969: 89). While the structure of the field may vary “according to the society or the age,” “it remains true that certain fundamental social relationships are established whenever an intellectual society exists which is relatively independent of the political, economic, and religious authorities,” such as the relationships between creators and publics, and with the “granters of legitimacy” (Bourdieu 1969: 109). Bourdieu later dismissed that paper as too interactionist, and adopted a more objectivist topological approach; clearly, however, this early formulation carried the seeds of field theory—with relative autonomy, translating into a ‘refraction’ effect of the structure of the field over external determinations, and specific authorities of legitimation and transmission (“reproduction” was not used yet). The refraction effect is for instance at work in “the relationship which an intellectual has with the social class he comes from or belongs to [...] mediated by the position he occupies in the intellectual field […]. Thus, forces of determinism can only become a specifically intellectual determination by being reinterpreted, according to the specific logic of the intellectual field, in a creative project,” which obliges economic and social events “to undergo a conversion of meaning and value by transforming them into objects of reflection” (Bourdieu 1969: 118-119). While Goldmann argued that the true subject of an artwork was the group to which the creator belongs, for Bourdieu it is the field (Bourdieu 1984: 212; 2004).
As he showed in his 1971 paper on the “market of symbolic goods” (Engl. ed. 1985c), three conditions must be met for a field to be formed: the emergence of a group of specialists of the activity in question; the existence of specific authorities (instances of consecration); the formation of a market. In that paper, Bourdieu analyses how literary and artistic activities became autonomized from clientelism and patronage through the market, but also shows that the market introduced a new constraint, specific to its own mercantile logic. Hence Bourdieu argues that the principle of autonomy lies in the refusal of the market logic, and more broadly of extra-literary (especially ideological and moral) criteria for judging artworks. The primacy of aesthetic criteria over all others also means that only peer judgment is recognized as valid, as opposed to public sanction translated into sales figures. In his Collège de France lecture on Manet, Bourdieu (2017) evidenced the role of the market in the conquest of autonomy from the state, which ensured the monopoly of the Académie des beaux-arts over the training of artists and the access of artworks to public space (by selecting those exhibited at the Paris Salon). The autonomization process must therefore always be related to the constraints at work in a given configuration: indeed, while the market allowed these activities to gain autonomy from the state, the state now ensures them a degree of autonomy from the market (Sapiro, 2003a). The scientific field is one of the most autonomized, as its main audience is composed of peers, but it may also be subject to contextual political, religious, and economic constraints (Bourdieu 2004).
Beyond the study of the process of autonomization vis-a-vis political, economic, and religious powers that determine the conditions of production and reception, autonomy, which is always relative, means that the field has a variable effect of refraction on external—economic or political—constraints, which are nevertheless always present. As we have seen, this refraction effect on social determinisms, which are manifested in the form of dispositions according to habitus theory, operates through the position occupied by creators in the field. Thus, unlike what is often assumed in comments on field theory, the relationship between habitus and field is by no means a mechanical one: dispositions orient the choices made by cultural producers in a structured “space of possibles” with a sense of its own history, which constitutes the field. These choices and socialization in the field lead to a “transubstantiation” of ethical and political dispositions, sometimes making them unrecognizable, especially through the process of the imposition of literary, artistic, or theoretical form, as Bourdieu (1991) shows in the case of Heidegger.
Lastly, Bourdieu (1979) examined the social conditions of aesthetic judgment, defined by Kant as subjective, devoid of concept and disinterested. This view of the purity and autonomy of aesthetic judgment neglects the “impure” conditions of production of judgment—namely, its social conditions and the distinction effect it produces by asserting the superiority of the classes most endowed with cultural capital. Also, the uses of cultural and scientific productions are far from limited to disinterested appropriations, based on a judgment that is autonomous from socio-political and economic concerns.
While these approaches and uses of the concept of autonomy are to a large extent incompatible, attempts at linking them can nevertheless be proposed. Sociological theories, be they the sociology of professions or field theory, take as a starting point the emergence of groups of specialists who are entitled to give informed judgments in certain domains. They concur in showing the social authority that these specialists acquire, and in demonstrating it to be a condition of the autonomization of their activity and the gap this creates between them and laypeople (2.1). Marxist theory addresses cultural productions in their relations with the social world, as a reflection, an expression of a world vision or ideology; Bourdieu introduces the mediation of the field to conceive these relations (2.2).
Assuming the autonomy of specialist judgements to be guaranteed by the state, the sociology of professions has emphasized the internal organization of specialized activities—i.e., control over access to the group, socialization through training, professional ethics, and internal discipline (see for instance Goode 1957; Abbott 1988; Freidson 2001). Field theory, on the other hand, has focused on autonomy vis-a-vis ideological and economic demands, guaranteed by peer judgment of the works (Bourdieu 1996; Sapiro 2003a). These approaches all base the principle of autonomy on an ethic of disinterestedness and responsibility, whose conditions must be ensured.
Field theory has the advantage of allowing us to account for the heterogeneity of professional groups. Largely observed in literature on the sociology of professions (Smith 1958; for a synthesis, see Champy 2012: 95-142), this heterogeneity, in terms of social recruitment, working conditions, or forms of specialization, has not been theorized as such, except with respect to specialization (Abbott 1988: 100-106 and 126; for an analysis of specialization drawing on field theory, see Pinell 2005). Likewise, as Bucher and Strauss pointed out in their 1961 article, by emphasizing cohesion, the structural-functionalist approach neglected internal conflicts within professions. Conflicts were also neither genuinely considered in the approach in terms of “worlds” (Becker 1980), which stresses the division of labour and cooperation. Field theory seeks to identify structural principles of opposition governing a specific activity and competition in a relational way, especially between the most established, who hold a monopoly over the definition of the activity (the “dominant”) and the newcomers who challenge that definition or other groups occupying dominated positions (Bourdieu 1984; 1996).
Here relative autonomy is a methodological principle that calls for relating the internal stakes of struggles of varying openness to these structural oppositions, through which external socio-economic and political stakes are retranslated and “refracted”: the degree of mediation indicates the degree of this autonomy. It is, for instance, lower in authoritarian regimes or crisis situations, where internal cleavages are determined by the relationship to power (Sapiro 1999; Dragomir 2007; Leperlier 2018).
These stakes are also refracted in specific ways in the sub-fields formed by specialties, disciplines and genres (literary, musical, etc.). In the sociology of professions, the concept of “segment” was proposed by Bucher and Strauss (1961) to describe these subdivisions and the tensions they elicit, including methodological cleavages and struggles that may lead to change (for instance in methods or professional ethics). However, the field concept systematizes the relational approach shown to be underlying the dynamic of competition between inter- or intra-segments over the accumulation of specific symbolic capital (or the construction of reputations), while accounting for less organized activities, such as fields of cultural production. Compared with the functionalist and static concept of segment, sub-field also has the advantage of reflecting on the specific and dynamic way in which it echoes the field’s structural oppositions.
These internal divisions are concealed by professional bodies which aim to unify the profession to defend its interests vis-à-vis the outside (Smith 1958), and which tend to operate as “corps” (Bourdieu 1985a). Drawing on Ernst Kantorowicz’s The King’s Two Bodies, Bourdieu harkened back to the reflection of medieval canonists (theologians in Church law) on the concept of corporatio, which also translates as universitas. A “corps” or “corporation” is in that sense an established entity that survives its members, in legal and symbolic terms (Kantorowicz 1957: 200 sq.)—here, the professional group, bound by a shared activity that fosters to varying extents a solidarity of interests revolving around rights or privileges secured or claimed by the group as an entity: the group’s degree of success in identifying the defence of these interests to the good of society as a whole—and by extension in universalizing its own interests—preserves it from or exposes it to accusations of “corporatism,” in the sense of focusing on one’s own interests at the expense of the general interest or universal causes. According to Bourdieu (1999a: 11), “[…] a corps exists when a group of individuals is relatively homogeneous in terms of the dominant principles of differentiation in the social universe under consideration, and united by a solidarity based on shared participation in the same symbolic capital.”
When an authority reaches a high degree of monopoly over a field, to the extent that it can instil an “esprit de corps,” the term “corps” is used instead of “field”: the closing off of recruitment using competitions, a numerus clausus, etc., is a means of controlling access to the field that may lead to its transformation into a corps, as in the case of the professional philosophers under the Third Republic (Fabiani 1988). Organized professions are those that most operate as corps, whereas in the less organized ones, field logics are more visible through internal struggles (Sapiro 2003b). The “field effects” they produce yield forms of resistance against attempts to unify the profession as a corps, as I have documented in my research on writers (Sapiro 2003c; Sapiro & Gobille 2006).
Various sociological approaches have also observed the divide between specialists and laypeople, sometimes considering it as legitimate (from a Weberian or Mertonian perspective) and sometimes criticising it (from a Marxist or Cultural Studies perspective). The concept of disinterestedness has often been used to justify this divide (Parsons 1939), whereas that of responsibility defines the relationship between specialists and laypeople (Johnson 1970: 57). These concepts have circulated between the intellectual professions, who constructed their autonomy by referring to one another and competing with one another (Sapiro 2017; on the ethic of disinterestedness professed by eighteenth-century lawyers, see Karpik 1991: 89-91 and 158).
Beyond the technical and expert dimensions concerning the control of contents7—a point of convergence between the sociology of professions and field theory—as well as the space of possibles underlying these contents, according to field theory—the issue of autonomy raises a triple question: that of the relationship of the activity concerned to external constraints; that of the organizational conditions that ensure it; and that of the ethics of disinterestedness in professional practice.
Points of divergence are observed where the relationship to external constraints is concerned. Reproducing the professional ideology, for a long time, the sociology of professions neglected to address the quid pro quo of statutory autonomy—namely, the service of the state and of the interests of the ruling classes. However, writings by Freidson, Larson, and many others have largely filled this gap by analysing autonomy as a demand by professionals, not as an essence of their activity.
According to Terence Johnson (1970: 41-43), specialization creates a system of interdependence that offers “potentialities for autonomy,” through the “structure of uncertainty” or “indeterminacy” in the relationship between producer and consumer—this uncertainty is reduced by the institutions. Depending on the balance of power at work, uncertainty is reduced to the producer’s or the consumer’s benefit. Johnson’s analytical model accommodates different forms of relationship between producer and consumer, determined either by the former (professionalism), the latter (patronage; corporations with a virtual monopoly over some services) or a third party (the state or entrepreneurs). The producers’ autonomy is high in the first case, weak in the second, and negotiated in the third. These three forms of relationship imply different levels of indeterminacy in the relationship between supply and demand, and by extension different levels of uncertainty, according to Johnson (1970: 41-43).
The organizational conditions ensuring technical autonomy have been extensively studied in the sociology of professions—especially the role of associations (Millerson 1998 ). However, the supposed link between professional bodies and autonomy—even technical—has been challenged by empirical research on authoritarian regimes, such as Jarausch’s (1990) study on engineers and teachers under Nazism, and work on the role of writers’ unions (Garrard & Garrard 1990; Dragomir 2007; I expand on this in Sapiro 2003a). Indeed, even though professionals were able to negotiate or retain a distinct status under these regimes, their technical autonomy was in effect restricted by the ideological censorship imposed on intellectual professions. The concepts of autonomy and heteronomy, as Bourdieu defined them, offer a heuristic framework to approach these more or less ambiguous relationships between professional bodies and the political, religious, or economic authorities in place. This also applies to instances of consecration, which have been neglected in the sociology of professions, and subject to more attention in field theory, but whose role in ensuring continued autonomy can be just as ambiguous, as I have shown in the cases of the Académie française and the Académie Goncourt under the Vichy regime in France; paradoxically, a clandestine organization set up by the Communist Party, the Comité national des écrivains [National committee of writers], stepped up to defend the principles of literary autonomy during these years of crisis (Sapiro 1999). As Alain Viala (1990) noted, the institutions of literary life are crucial links between the field and the political and economic authorities, and their functioning and role must be subject to concrete empirical studies to establish the type of autonomy (or heteronomy) that they ensure.
Regarding ethics, it is methodologically necessary to distinguish between the discourse on autonomy and actual practice, which may contradict discourse. This goes equally for the sociology of professions, which has observed practices wherein the professional’s interests take precedence over the client’s, in breach of the code of professional ethics (Freidson 1970: 360-362), and field theory: for instance, under the German occupation of France, the Nouvelle revue française, a literary magazine whose publication had resumed on the initiative of German ambassador Otto Abetz, ostensibly conveyed a discourse on ensuring the autonomy of art under any circumstances even though at the very same time former contributors were being excluded for reasons entirely unrelated to art, specifically their (Jewish) religious background or their (communist and anti-Nazi) political opinions (Sapiro 1999: 402-408).
Marxist uses of the concept of autonomy emphasize the relative autonomy of artworks vis-à-vis the social conditions of their production. This idea bears similarities with that of technical autonomy, but likewise raises the question of its conditions of possibility.
As previously noted, these artwork-centred Marxist approaches lack an examination of the mediation exercised by their conditions of production and circulation. This is found in another strand of Marxist tradition, running from Gramsci to the founders of Cultural Studies, especially Raymond Williams (1981). Williams does not use the term autonomy in that sense, but his research programme on the institutions of literary life—publishers, magazines, academies, and circles—took these aspects into account (he referred to Bourdieu). Also worth noting is that Goldmann’s thesis on the Jansenist world vision of the noblesse de robe in his analysis of Racine is in no way incompatible with field theory. There was admittedly no genuinely autonomized literary field in the seventeenth century, even though, as Viala (1985) observed, a group of specialists and an instance of consecration, the Académie française, emerged (see also Saint-Jacques & Viala 1994). Yet clientelism still prevailed, and largely conditioned the production of artworks, in the absence of a market of symbolic goods—which would only develop during the following century. On the other hand, Bourdieu faulted Sartre for ignoring the mediation of the field in his masterly biography of Flaubert, The Family Idiot (1971-1972), which relates the writer’s oeuvre to his family history and a class analysis without addressing the conditions of literary production at a time when, as Bourdieu went on to show in The Rules of Art (1992 Engl. ed. 1996), a pole of restricted production developed to defend the principles of the autonomy of the judgment of peers and specialists in the face of the market-oriented approach that prevailed at the pole of large-scale production of the expanding book market.
In light of these possible connections, I propose a distinction between three, sometimes embedded, levels of autonomy—conditions of production, practices, reception and uses.
Let us come back to the three forms of relationship between producer and consumer singled out by Johnson. As previously mentioned, the producers’ degree of autonomy is high when they are in a position to determine their relationships with consumers, as in the model of independent professions that developed in the US before circulating across the world. Professionals enjoy a high degree of technical autonomy, guaranteed by professional bodies (ordres or associations endowed with powers), but this state-sanctioned autonomy depends on the regime’s degree of liberalism (in authoritarian regimes, professional bodies are supervised and kept in check by the authorities) and is never complete (as Freidson  observed, the training that conditions access to the profession in question is not offered by these professional bodies, but by institutions whose diplomas are accredited by the state).
Additionally, the price of this technical autonomy in terms of the service rendered to the political authorities in return also deserves consideration. This raises the question, investigated by Merton and Bourdieu, of autonomy vis-à-vis the dominant ideology and state demand. Organized professions may be conceived as highly centralized fields, around professional bodies that operate with a logic of “corps,” in that they control access to the game and its rules, and have the power to punish the ‘heretics’ that stray from orthodoxy. While this control is in principle aimed at ensuring that the clients’ interests are safeguarded, it is liable to being hijacked for the benefit of the authorities (as in the case of eighteenth-century hygienist physicians), corporations (the lawyers, engineers, and scientists working for them often use their technical autonomy to serve financial interests; on the case of economists working for banks, see Lebaron 2000), the dominant classes, or dominant ethnic or religious groups. Provisions such as the numerus clausus for Jewish doctors and lawyers in interwar Eastern Europe and the ten-year wait for foreign doctors in 1930s France were for instance aimed at restricting competition for professionals belonging to the dominant ethnic or national group. The same applies to the exclusion of women from these professions (Rennes 2007). These practices reveal forms of heteronomy of professionals vis-à-vis various components of the field of power, when they are in a position to regulate relationships with clients. Professional bodies are a site of observation of negotiations with these components, and hence of translations of these forms of heteronomy into professional rules and practices. However, the opposition between autonomy and heteronomy is, alongside that between dominant and dominated (according to the volume of specific capital), a structuring principle of fields and therefore a grid for analysing the practices themselves.
When the clients of services or products are mainly patrons, as in the arts during the French Ancien Régime, or corporations with a virtual monopoly over orders, as in the case of consulting engineers (Henry 1992; 2012), the autonomy of professionals is weak, as their production is largely dependent on demand. In that case, production is also largely informed by reception and the possible uses of symbolic products.
Although Johnson does not discuss them, the case of third-party mediation (by the state or entrepreneurs), wherein the autonomy of producers is negotiated, very much fits the art markets that are characterized by state subsidies on one hand, and the significant role of intermediaries (publishers, literary agents, gallery owners, and film and music production companies) on the other. In authoritarian states, as in communist and fascist regimes, the autonomy of cultural producers is weak, even though they are highly professionalized, and professional bodies act as ideological control bodies. In market economies, their autonomy is weakened by dependence on a logic of profit that drives many cultural entrepreneurs— “double” agents who are more or less expected to reconcile that logic with the specific logics of cultural worlds (Bourdieu 1977)—and submits production to a market-oriented rationale. Empirical studies on recent trends in the publishing field in France and the Anglo-American world have shown that such economic constraints have a growing impact (Bourdieu 1999b; Thompson 2005). Hence the dual structure identified by Bourdieu, with a pole of large-scale production and one of small-scale production, which characterizes all markets of symbolic goods. In response to these economic constraints, the state may play a corrective role by subsidizing part of the production for which demand is lower, but whose value is estimated to be high by the specialists entrusted with evaluation, in a form of recognition of the autonomy of their judgment (Sapiro 2003a).
While they are highly useful to our understanding of the degree and type of autonomy in terms of conditions of production, these three configurations are not in themselves enough to establish them. Indeed, they must be considered within a broader socio-political and economic context according to the type of regime and the ideological control over practices, which brings us to the second level.
While the producers’ discourses on autonomy are useful to establish their demands according to the conditions of production, they need to be confronted with practices.
As soon as the market of symbolic goods emerged in the eighteenth century, some men of letters, such as Karl Philipp Moritz, an influence on Kant, warned cultural producers (particularly artists) of the temptation to flatter the audience to the detriment of the requirements of their art and their professional ethics. At odds with Moses Mendelssohn’s idea that the unity of the arts resides in their effects, meaning the emotions produced by artworks as a result of relations of harmony and symmetry, Moritz also finds it in the formal perfection that underpins ideal beauty, but emphasizes the absence of finality, which Kant theorized by defining the artwork as finality without end (Woodmansee 1994). This is why the huge success of some cultural productions is suspected of resulting from the producer’s compromises, or dishonourable attempts to seduce an undifferentiated audience, lacking familiarity with the aesthetic canons. Except in highly codified forms such as genre films, romances, or crime stories, some intermediaries (publishers, agents, film producers, etc.; on autonomy in cinema, see Duval 2016) intervene upstream of the creative work to make sure that the project will be economically profitable, and subsequently commit to a promotion effort that will be analysed in the next section. The market-oriented rationale may also influence scientific or professional practices, liable to serve economic interests, as previously noted. Many industrial groups have lawyers working for them. Beyond heteronomous practices, the scientific products that may appear to be the most autonomous and abstract, such as mathematized economics, can also be instrumentalized for “impure” purposes, such as justifying austerity policies (Lebaron 2000).
Political and religious authorities have often sought to subdue the sciences as well as the arts and letters to turn them into propaganda instruments; socialist realism as a creative method is an example of this (Robin 1986; Aron, Matonti, Sapiro 2002; on science, see Matonti 2002). However, even when they result in bans and censorship, these constraints do not preclude heterodox practices, in coded form (for instance, “contraband poetry” during the German occupation of France, or the theatrical performances of Polish actors under the communist regime), clandestine form (samizdat), or displaced abroad (on these options, see Sapiro 2003).
Beyond such constraints, Marxist approaches have evidenced the ethical and political dimension of cultural productions using the concepts of “world vision,” “ideology,” and “hegemony.” As previously noted, Althusser and his disciples preferred “ideology,” which conveys the function of reproduction exercised by these representations, over Goldmann’s “world vision.” For his part, Raymond Williams favours the Gramscian term “hegemony” (Williams 2009). Pierre Bourdieu forged the concept of “symbolic power,” which stresses the structuring cognitive functions of symbolic goods in the neo-Kantian tradition, while concurring with Weber and Marx in pointing out their social functions in relations of domination, which are all the more effective as they are unrecognized—“méconnues”—as such (Bourdieu 1977). Indeed, whether they reproduce the dominant ideology or challenge it to support alternative worldviews, cultural productions are never “neutral,” in that they get their significance from their relations with the space of representations and the symbolic forms that exist at a given juncture, both in terms of conditions of production and reception. According to field theory, their degree of autonomy hinges on the refraction effect resulting from a process of imposition of form—mise en forme—inscribed in a historically constituted space of possibles. The reproduction or renewal of these forms is a major stake in worlds where originality is the most important value (Lucien Karpik  calls this “the economics of singularities”).
The question is whether this formal autonomy—which has similarities with technical autonomy in the sociology of professions—is enough to ensure autonomy vis-à-vis the dominant ideology or the social functions given to artworks. Crucially, even the most “autonomous” imposition of form will not prevent a work from conveying ethical and political schemes associated with a given group or class, as Bourdieu (1991) has shown in the case of Heidegger. Jeffrey Mehlmann’s (2003) political reading of French poet Paul Valéry’s “Graveyard by the Sea” finds in the phrase “derelict heads” a reminiscence of the author’s training in craniometry with raciologist Vacher de Lapouge, during which Valéry had dug up and measured 600 skulls in an abandoned cemetery. Without going quite as far as to depict Valéry himself as a racist thinker, Mehlmann shows how the superiority of the Aryan man imbues the worldview conveyed in his work.
The concept of autonomy thus relates to the process of translation of extra-literary issues according to logics specific to the field. The term “monologism” proposed by Bakhtin in contrast to “dialogism” (the heterogeneity of points of view, styles, etc.), and employed by Régine Robin (1986) in her analysis of socialist realism, raises the question of the limits of autonomy and dispositions towards heteronomy of cultural and intellectual producers.
Yet despite frequent assumptions, the politicization of producers of symbolic goods is not in itself in contradiction with autonomy: indeed, the different forms of their engagement are generally related to their activity and position in the field. Their degree of autonomy hinges on their independence from outside demands from social groups or their spokespersons (political organizations) (Sapiro 2009, 2018). Twentieth-century avant-gardes dreamed of revolutionizing the world through purely artistic means (Brun 2014; Gobille 2018).
In addition, even the most formal and abstract works may serve political goals regardless of the author’s intentions. This happened to abstract art, which was instrumentalized by US foreign policy to embody the American conception of liberalism in art in contrast to figurative socialist realism (Guilbaut 2006). However, this aspect rather pertains to the uses and appropriations of artworks, which also partly addresses Marx’s question as to the autonomy of works vis-à-vis their conditions of production, particularly through the canonization process.
Before we move on to this last question, it is worth recalling Williams’s (2009) distinction between residual cultural forms (having emerged in a previous social formation, they linger in the present juncture, but on the fringes of dominant culture), and emergent cultural forms, which include alternative and oppositional forms. It might be worth cross-examining these distinctions (which I contend are more heuristic than the binary ideology/utopia division proposed by Mannheim) and the degrees of autonomy or heteronomy of practices (self-referentiality, inscription in the history of the field, ability to renew forms, autonomy of the project from a pre-formatted demand, etc.). The concept of symbolic form could allow us to link formal aspects with the perception and evaluation schemes conveyed by the works.
Technical autonomy is based on peer evaluation in both the sociology of professions and field theory. While it cannot prevent ideological appropriations and usages of productions and services, it can respond to those with its own criteria for judgment, focusing on ‘technical’ aspects that require specialized knowledge—in science, professional practices, and artistic productions alike.
Following a market-oriented rationale, profitability, measured on the basis of sales figures for artworks and uses for services, introduces heteronomous principles that might be detrimental to the quality of products in that they lead to compromises aimed at accommodating the tastes of lay publics. Here we must effectively distinguish between the three configurations defined by Johnson. In the case of professionalism, the argument of autonomy can be hijacked to the detriment of clients’ interests, who are often helpless in the face of the authority of specialists. The case of heavy dependence on demand from corporations or the state is one of structural heteronomy. The question of autonomy is raised in the third case, in which the relation between producer and consumer is mediated by intermediaries (on that category, see Lizé, Naudier, Roueff 2011; Jeanpierre & Roueff 2014), as the latter have economic interests that they may put in the balance during the upstream process of selection of products and services and during the production process itself.
These intermediaries frequently put forward the allegedly democratic argument of consumers’ preferences versus the supposed elitism of experimental creators and their supporters—a rhetorical trick developed by the cultural industries to justify profit-oriented rather than demand-oriented choices, as exemplified by the aggressive strategies of the largest publishing groups to showcase their books in chain bookstores (by paying for space in window displays or near registers) to outperform the competition (Thompson 2010). Concentration around these large conglomerates in the film, music, and publishing industries thus threatens the autonomous logics of the fields of cultural production by confining them to the sidelines and mitigating the verdict of specific instances of consecration (Bourdieu 1999b; on film, Duval 2017). Research has shown that literary awards, which have themselves been regularly accused of compromising to retain their media sway, have less impact on bestseller lists in the US—a country where the rationalization process in the publishing sector is very advanced—than in France, where these awards are still influential (Verboord 2011). This phenomenon is revelatory of the ways in which some intermediaries pursue strategies to bypass the verdict of the field’s authorities and promote a market-driven approach.
The political interpretations, appropriations, and uses of cultural and scientific productions have been extensively studied. Cultural producers had to fight to conquer their autonomy in the face of censorship and still suffer from restrictions of their freedom of expression in many places (on writers in France since the early nineteenth century, see Sapiro 2011). Regarding reception, while cultural producers may have an interest in leaving things ambiguous to avoid punishment or may reach out to a more diverse audience, appropriations are facilitated when their works are detached from their context of production, through a shift in time or space (Bourdieu 2002). The political uses of theories of justice in France are the ultimate illustration of this. Significantly, Rawls’ introduction to the philosophical canon in France involved a depoliticization process (Hauchecorne 2019).
While the processes of consecration and/or construction of reputations are well known by now, canonization, which gives works the classic status that allows them to survive their conditions of production, has been less studied, despite the pioneering examination by Russian formalist Viktor Shklovsky on the stratification of literary systems into canonical and non-canonical elements, then expanded upon by Even-Zohar (2005) in his polysystem theory. We owe Alain Viala (1993) a reflection on what a classic is. Since the 1980s, the deconstruction of the European canon has inspired studies on the unequal conditions of access to recognition (on French-language writers from sub-Saharan Africa, comprehensive research is found in Ducournau 2017).
Indeed, the claim of autonomy, just like professionalization, has also in many cases served as a lettered justification for excluding producers from the canon, marginalized for reasons of gender and/or social, ethnic, religious or national origin (sometimes this has meant downplaying their contributions; on the exclusion of interwar female French writers from the literary canon, see Milligan 1996). Thus, the autonomization of a philosophical field in the second half of the eighteenth century was characterized by the invention of a new form of philosophy history writing by the Kantians, which dissociated it from salvation doctrines, thereby excluding all the non-Western heritage (Park 2014).
The same form of distinctive elitism sometimes draws on the argument that there must be a break from lay audiences in the name of autonomy, by affirming the disinterestedness of the aesthetic judgment of cultural capital holders, in accordance with Kantian theory, against the utilitarianism or lack of distance of those deprived of that capital (on reading, see Sapiro 2011). Marxist approaches saw in this alleged lack of distance an expression of the mystification of the working classes and the alienation of consciousness, with the exception of the founders of Cultural Studies, who challenged these theses first by making a distinction between mass and popular culture (Williams 1974), and second by disputing the thesis of identification that underlies the idea of alienation (Hoggart 1970). The sociology of taste and uses then reconsidered this dichotomy, favouring a reflection on structural homologies (Lizé & Roueff 2010) and the diversity of the social uses of cultural productions: for instance, Gérard Mauger et Claude Poliak (1998) showed that “aesthetic reading” is only one form of appropriation among a range of other interests: entertainment, education, and salvation.
While the concept of autonomy stems from different, rather incompatible traditions, an informed synthesis of these approaches may be fruitful to explore how spaces of production of symbolic goods operate. It would also be interesting to explore areas where these definitions compete in reality—for instance, professional autonomy versus autonomy in the sense of field theory (see Picaud on this subject). Thus, professional claims and aspirations to autonomy from the state and the market are at odds in the literary field, where professional recognition comes from income, not from quality judgments (Sapiro & Gobille 2006; Sapiro 2016).
More broadly, as I have suggested elsewhere (Sapiro 2006), there is no relationship of necessity between professionalization and autonomy, as the former depends on corps rationales that may be offset to various degrees by field effects. The examples of the engineers and teachers of the Third Reich and intellectual and artistic producers in the Soviet Union, professionalized by the state, who centralized and nationalized their production so that they could better control it and use it for propaganda purposes, shows that we stand to benefit from moving away from professional ideology in order to better study the organization and its practices, including those consisting in excluding or restricting individuals for reasons of gender or geographical, national, social, or ethnic origin. The same applies to market constraints; heteronomy may in fact even favour professionalization, as in the paradigmatic cases of blockbusters and bestsellers.
Furthermore, beyond external constraints, the autonomy of reason and creation can be placed at the service of authorities and serve to justify domination, as has occurred in many professions (doctors, lawyers, etc.) and with ideologues and experts (both civil servants in universities and “independent” producers). And one should distinguish the autonomy of critical reason, which is not exercised in response to an external demand, but in the name of one’s own conscience and values (truth, justice), as a specific form of exercise of intellectual autonomy, liable to make alternative, or even oppositional symbolic forms emerge.
Lastly, both the reflections on professions and fields raise not only the question of competitive struggles over the division of labour and the social recognition of an area of competence (the specialization process), but also that of geographical borders. As it happens, national autonomy does not always mean autonomy from power. Admittedly, like the organization of professions, the autonomization of fields has been linked to the emergence of nation-states and the competition between them, and at the same time, it extends beyond that scale by definition, through the market approach to cultural production, the circulation of knowledge and models for the professions, regulated to varying degrees by states (equivalence), international agreements such as the Berne Convention on copyright, or the ones devised by international bodies, Unesco, and international professional groups (Sapiro 2013). Yet Pascale Casanova (1999) showed that writers asserted their autonomy by turning their backs on their involvement in the construction of national identities, often by importing foreign models. As Bourdieu (1985b) and Paul Aron (2005) demonstrated regarding Belgian literature, the presence of national instances of diffusion and consecration is not enough for an autonomous field to form, as the most autonomous pole in the Belgian field still pursues consecration in Paris.