1 Talent, creation, creativity, genius—these scarce phenomena have traditionally represented enigmas for the social sciences. The lexicon of giftedness, inspiration and singularity attached to the arts has long held sociological analysis of this field of activity at bay, both sociology defined as the science of collective forces and sociology that embraces the language of self interest. Pierre-Michel Menger has constructed a scientific approach that works to disqualify this commonplace of the discipline, his wager being that the study of art worlds will help renew the social sciences (and economics) and perhaps offer new insights—less accessible in other fields—into societies and individuals. The decisive stages of his undertaking have now been brought together in Le travail créateur (Menger 2009), thirteen chapters initially published as articles from 1989 to 2004, some reworked for the occasion; the book also contains a new essay of over 130 pages, in the equivalent of a small book.

2 Le travail créateur, while partially opening up the author’s workshop to his readers, also partakes of the comprehensive summary and the manifesto. Gradually, through successive movements, the author unveils a general analytic model of artistic creation and the social worlds of which it is composed, thereby providing the substratum of some of the research studies, syntheses and general introductions to his work which he has published in recent years (Menger 2002, 2005b). In critical mode, the book points up the limitations of existing economic and sociological explanations of the creative process and artistic labor market operation. As positive sociological study, it offers a set of concepts that prove useful for analyzing these phenomena, together with a fully developed alternative model for explaining them. Moreover, the work attests to the many different theoretical problems that interest Menger once they have been partially disengaged from the specific populations (contemporary composers, musicians, actors, senior audiovisual technicians, intermittents du spectacle [freelance and contract performing arts and media industry workers], etc.) and topics he has been studying for the last thirty years (Menger 1979, 1983, 1989, 1997, 2005a). In this respect Le travail créateur is also a book of sociological theory, in the form of critical commentary of several social science thinkers and currents of thought. For example, a chapter on incompletion in art and the case of Rodin leads to a discussion on the limits of ontology and esthetics of artworks. Another chapter is primarily concerned with theories of action, while a third offers an oblique reading of Durkheim in connection with the art question. One feels one is reading a collection of exercises that might be designated analytic sociology [2] because they so often make use of thought experiments, as in analytic philosophy, from which Menger draws several of his arguments.

3 The collection does make substantial use of empirical material from sociological and economic studies of the art professions in several countries, also drawing on time-use surveys and surveys of cultural demand and consumption in Western populations, social history of music and art, together with the author’s own studies of French realities, namely the situation of freelance and contract performing arts and media industry workers in the 1990s and 2000s. The most original feature of this heterogeneous collection is that from the outset it combines sociology of art with socio-economics of labor and labor markets—thus following in the path forged in France by Raymonde Moulin—in turn combining these with sociology and economics of professions and of culture. Le travail créateur will therefore be of interest not only to sociologists of creative work and the arts but all researchers in these various social science and economics sub-specialties. And as attested by its subject index, the book’s problematization level allows for asking and answering more general theoretical economics and sociology questions. It therefore deserves a wide readership.

Artistic creation as uncertain action

4 Menger takes off from the observation that artistic creation worlds are unique—or at least that they differ from the social worlds and labor markets usually studied in sociology of work and economic sociology—and that the motivations of the individuals who populate these worlds are necessarily paradoxical if considered in relation to the rationality generally attributed to actors by the social sciences. How, for example, are we to explain the strong and growing attraction to artistic careers, careers regularly punctuated with competitive ordeals, careers to which many seek entry but few are ever admitted? For the author, these specificities are best explained by what he calls the uncertainty “principle,” as this is what structures action and transactions in art worlds. Artistic creation differs from other action regimes in that it must always and everywhere cope with uncertainty. Indeed, this is what makes it attractive, for such activity is not at all likely to be repetitive or dull. But uncertainty is also the wellspring of all the specific constraints of creative work: art work is risky activity very likely to subject those who invest in it to a precarious material existence. Without the central idea that creative work is structured by uncertainty, we cannot understand the paradoxical social and labor logic governing it, of which Menger apprehends both the individual and collective dimensions: activity levels—the production process—on one hand, value levels—labor market exchange—on the other.

5 “Rationality and uncertainty in the artist life” (Chapter 4), a seminal article first published in 1989, focuses on the theme and problem of uncertainty in connection with art analysis. Menger reviews the difficulties of studying creative activity and how sociologists and economists have sought to resolve them. How are we to understand the fact that some individuals choose activities that pay less than those to which their educational degrees and training would give them access in other sectors? After explaining why the answers put forward by determinist sociology and neoclassical economic theory cannot account for the logic behind that choice, Menger highlights that becoming an artist involves something like a preference for risk-taking. Actors justify their choice of the artist life in two ways: it “offers the psychic benefits of non-routine or relatively non-routine work”; it provides them with experience and the opportunity to undergo competitive trials that regularly evaluate—better than any metrics could do—their chances of succeeding as creators. In this framework, “the artist career can be understood as the interaction between self-development as assessed through comparison of the individual’s successive states as she acquires capital and information about herself and the gradual revelation of individual value via the interindividual contests” that are a regular companent of life in art worlds (Menger 2009: 139 [unless otherwise indicated, all references are to this work]). Here Menger mobilizes theories of “job matching” and the idea that workers modify their expectations and sense of their own aptitudes over the course of working in different jobs. “The strong earnings variance [among creators] is therefore explained on one hand by the high proportion of inexperienced art workers accepting low income in exchange for the information they can acquire (which leads many to abandon quite quickly the idea of an artistic career, or at least any plan to make a living from artistic activities alone) and on the other by the success of a small minority, who manage to turn their accumulated experience to advantage.” (p. 213).

6 The other side of the preference for risk-taking that Menger imputes to persons planning a life in art or creative work is pluriactivity, a feature consistently noted in all sociography of art sectors. He likens how an individual divides his self-investment himself between artistic and other income-producing activities to a “business portfolio” as analyzed by Kenneth Arrow. Once again this is a case of individuals calculating how to cover the risks they take, but Menger offers a more dynamic analysis than most economists, who tend to naturalize differences between risk-seeking and risk-averse individuals. In passing the author notes that the less professionally organized a given artistic activity is, the higher the proportion of pluriactive artists in that activity.

Time as creator

7 Since the defining feature of the artist life is action “whose horizon is uncertainty,” the opening chapter of the book, a reworked version of an article published in the Revue Française de Sociologie, examines in general terms how the social sciences account for action uncertainty. Menger’s purpose here is to assess the theoretical contributions to “action analysis” (p. 33) made by sociology and economics, the two disciplines he makes the most use of. From the outset he rejects the partition between the two fields, likewise rejecting the tired opposition between individualism and holism, focusing instead on two other ways of doing social science: determinism, defined as a scientific discourse in which “causal analysis places the agent under the control of forces whose properties derive from the actor’s past and trajectory” (p. 36) and interactionism (granted a wider meaning here than it has in the eponymous sociological tradition), which emphasizes “actor intentionality” and attributes “a crucial role to [actors’] representations” of “their resources and action objectives” (p. 45). For Menger, both these approaches prove limited when it comes to analyzing action. While determinist positions manage to account for differences in actor behavior, this is only at the cost of collapsing time and ignoring the play of uncertainty. They substitute “logical time” for “historical time.” The interactionist approach, on the other hand, apprehends the contingency of situations and social relations but cannot specify either the causes or the aggregated effects of differences in individual behaviors.

8 Menger has thus sought an analytical path that steers clear of these two pitfalls. He suggests integrating time as an irreducible factor of potential novelty while preserving a fundamental social science objective; namely, to explain rather than merely observe interindividual differences and the structuring of social worlds. This leads to a “notion of action as process” (p. 94) informed by analytical philosophy (Vincent Descombes’ in particular) and the thinking of Gilles Gaston-Granger and Paul Ricœur. Menger invites us to conceive of historical time as “stochastic time,” in which events “are distributed according to the probabilities that lead actors to formulate conjectures about their environment and the behavior of other actors in it,” and in which every situation has its own “coefficient of uncertainty.” Under these restrictive conditions, the rationality attributable to an actor can be defined as a coherent, consistently applied manner of handling uncertainty (pp. 96-97).

A theory of art world social stratification

9 Once the general effects of the “uncertainty principle” on individual and collective behavior in art worlds have been weighed, individual differences within those worlds need to be explained. This question is the work’s other center of gravity and may be read as a theory of social stratification in artistic labor markets. Here Menger calls upon several studies showing that experience acquired “on the job” in these activities is more relevant than initial training for understanding the success obtained by candidates for admission to art worlds. As specified, he suggests thinking of that experience as the individual’s “demand for information”—a demand with a cost; this then justifies, writes Menger in passing, the lower average income at equal resource and skill levels found in the cultural sector (compared to other sectors of activity). But it is also important to account more specifically for differences in choice of activity within the overall creative sector.

10 Individual’s initial economic resource level surely plays a role in that choice, but only alongside of and as a complement to differences in innate and acquired aptitudes—Menger devotes many pages to this last idea. The “hypothesis of differences in aptitudes” he explains, “becomes particularly significant if we assume that the various types of learning—schooling, occupational training, ongoing education, on-the-job experience—are interdependent” (p. 116). He demonstrates that in art worlds the “human capital” an actor accumulates can never be known in advance and will only be revealed gradually, as he or she goes from experience to experience. This is why occupational choices in artistic sectors cannot be modeled simply in accordance with standard economic theory as human capital investment choices.

11 Menger’s primary innovation here is to introduce a dynamic mechanism: “selective job-matching” between employers and employees. For artists, he explains, the “trajectory made up of formal learning followed by on-the-job learning… is strongly determined by work done with experienced partners, as this is what offers an individual in the process of becoming professional the best chances of developing his abilities, enabling him as it does to participate in demanding projects that will put him in contact with partners themselves chosen for their potential. Art worlds combine labile organizational architectures (networks, individual projects, vertical disintegration) with a team structure, the team being composed of professionals of equal quality or reputation—that is, selectively matched individuals” (pp. 354-355). This means that the status of the persons an artist or aspiring artist works with in the framework of a given project acts as a “signal” of both her value in the particular area of activity and her aptitude level. It also means that artistic labor markets can be analyzed as highly stratified worlds in which each stratum brings together protagonists who have obtained fairly equal degrees of recognition, a feature making them likely to form collaborative networks.

12 Differences in artist success levels, then, are not like “lottery” results as Adam Smith may have seemed to suggest. But neither are they determined solely by differences in initial or accumulated “assets,” as both the Bourdieu-inspired sociological vulgate and classic human capital theories would have it. Interindividual differences in “talent” are, for Menger, initially unknown to art labor market protagonists; they are literally revealed and amplified by the cumulative selective job-matching dynamic. [3] The logic behind this mechanism may be compared to what Merton identified as the “Matthew effect.” Not only does the series of matches between artists and other art world actors produce a sum of information on the various protagonists but it acts as a multiplier of initial differences between individuals working in the same activity. This is how virtually non-existent initial differences get amplified in the course of a career, generating highly divergent success trajectories.

13 Menger is concerned here to refute the Marxian regulatory idea that creativity is equally distributed and that a social world could exist from which comparative evaluation and competition have entirely disappeared. Using counterfactual reasoning, he shows that the notion of zero aptitude difference—aptitude being defined here not as an essential property of individuals but as social recognition that a given individual is more competent than another to perform a given task—is anthropologically impossible; if it were possible, there would never have been any division of labor in the history of human societies. Moreover, Menger understands interindividual comparison (auditions, scholarship and grant allocation, contests, awards, festivals, etc.) as a socially necessary art-world mechanism because it works to reduce the uncertainty that transactions in those worlds necessarily imply. On this point he remarks that the number of contests, competitive trials, established prize-winner and bestseller lists increased considerably over the twentieth century in both the arts and sciences.

Rejection of constructionist analyses

14 Chapter 7 is entirely devoted to refuting a purely constructionist reading (in Hacking’s sense) of the “selective job-matching” mechanism and the workings thereof. The demonstration here is based on an analysis of sociological and historical explanations of Beethoven’s “genius.” Though Menger is right to claim that in the arts, differences in career choices and would-be artists’ success levels depend on the information about actors that is accumulated and exchanged in the series of “creative projects” they work on together, is it not also true that actors quite different from the creators themselves—i.e., critics, mediators, gatekeepers, audiences, etc.—are likely to be responsible for “producing” the value of artists and artworks? Menger maintains that this argument would only shift the initial sociological problem from how to explain differences among artists to how to explain differences among intermediaries and gatekeepers. As he sees it, social constructionists reject the idea that there could be differences in the “intrinsic strength of individual talent” (p. 424) (granted that talent cannot be known in advance), whereas in his model individual aptitude cannot be reduced to other factors. Much of Le travail créateur is devoted to demonstrating this last point, which the author takes up in several ways throughout the work.

15 Menger recapitulates his propositions by asking that three factors alone be taken into account to explain “talent and inequality in the arts” (p. 359):

16

• “differences… in individuals’ quality,” understood exclusively in terms of aptitudes that win social recognition, and despite the fact that these differences “are not fully observable” (p. 360);
• action and evaluation by peers, other professionals, audiences; that is, “the selective attention focused on individuals and works within a given professional community or audience… [attention] that can give rise to rational contagion, … a disproportionate advantage” due to a cumulative advantage dynamic (p. 361);
• the particular dynamic of “selective job-matching,” which offers “matched individuals higher returns on their respective aptitudes than what they would get through chance matchings,” as well as “increased experience” and “much higher earnings than what they would obtain by way of a mere additive function” (pp. 362-363).

17 These three factors are not only necessary but sufficient, Menger argues. For while, as mentioned, he duly takes into account differences in protagonists’ initial resources, types of work, opportunities, and ability to actually seize them, none of these other differences, as he sees it, constitutes a factor that could itself explain art world careers and success dispersion. In some of his analyses (the case of Beethoven, for example) he also takes into account variations in the material, legal, and political environment that artistic activity is being done in. But once again, none of these dimensions, as he sees it, is irreducible.

Jobs in art: the archetype of flexibility, the reign of “project-based organization”

18 As an extension of his thinking on stratification in creative activities and its social origins, several chapters of Le travail créateur are devoted to how artistic labor markets are organized. Chapter 11 looks at the employment arrangements of theater, film and television actors. In those activity sectors, the number of creators rises faster than the volume of paid hours. Turnover is considerable. Menger’s survey also shows that up against the economic uncertainty in which works in these performing arts get produced and the competitive need to find new esthetic formulas, relations between employers and artists tend, at least in theater, to become stabilized around a privileged relationship between actor and director. Moreover, acting contracts are becoming shorter and actors are signing more of them. Menger differentiates the network types that enable performers to land jobs in the three subsectors studied and prevent these labor markets from becoming entirely atomized. The role of personal, non-professional ties is greater in film and television than in theater, where “recurrent professional ties” are more important than “weak” ones when it comes to acquiring information on available jobs. Nonetheless, in all these subsectors, “networks are the mechanisms that structure interindividual relationships, and they operate in place of what would otherwise be a world of sporadic, ever-changing relations—a world without memory—or a world in which activities would be organized within a formally self-contained business or professional group” (p. 527). That hyperflexible artistic labor markets are structured this way also enables us to understand spatial concentrations of artistic activities (Chapter 12).

19 These developments belong to what in sociology is called the “project-based organization” characteristic of performing arts, film and audiovisual work. Discontinuous employment—frequent alternation between periods of activity and unemployment—is part and parcel of this organization mode, itself due to the downsizing of big corporations and vertical disintegration in these sectors. According to Menger, the assembling and disbanding of production teams by artistic project offers both economic advantages (lower fixed costs) and esthetic advantages (a plurality of creative contributions). France accelerated the transition to this way of organizing employment relations in 1964 by setting up the world’s first unemployment insurance program for “freelance and contract media industry and performing arts workers” (the system was further developed in 1969). This policy has been a source of controversy in France for over a quarter of a century. In the early 2000s, Menger obtained access to data from UNEDIC (Union Nationale Interprofessionnelle pour l’Emploi dans l’Industrie et le Commerce; National inter-occupational union for employment in industry and commerce) and the Caisse des Congés Spectacles [association managing paid leave for performing arts professionals who have not been employed continuously by the same employer for the twelve months preceding their application for this benefit]. This enabled him to study contemporary forms of employment and unemployment in the performing art sectors in France; the substance of his analyses is reexpounded in Chapter 10. When first published, these analyses provoked much criticism, particularly from sociology and economics researchers implicated in the struggle of this set of workers against recent changes to their unemployment insurance program (Corsani and Lazzarato 2008). Menger was interested in the system’s growing financial imbalance, which he intended to prove was structural. [4] “The employment flexibility secured by this unemployment insurance setup has spread, and increased the risk it was designed to cover rather than reducing it” (p. 502). Who is responsible for this? No one or everyone, depending on point of view. According to Menger, the contemporary way of organizing the artistic labor market—facilitated but in no way caused in France by the country’s unemployment insurance program for part-time and “intermittent” performing arts workers—is responsible for poor distribution of information about and among workers in this market.

20 In an employment system characterized by “extreme flexibility sheltered by unemployment insurance” and by considerable inequalities in activity and income levels, selective job-matching no longer works to signal reliable information to either applicants for art world jobs or their employers, Menger argues. “Employment situations are so variable and fluctuating that intermittent employees are quite incapable of identifying the causes of interindividual inequality in the allotment of work and earnings.” (p. 497). It is therefore in the interest of employers to become associated in a relatively enduring way with a few unchanging partners (in the 1990s in France, 60% of the volume of paid freelance work in these sectors was accounted for by work done by individual freelancers for a single employer) and to use the unemployment insurance as the adjustment variable in risk strategies. To overcome the program’s predictable chronic deficit, due to these strategies, Menger called for setting up a reward and punishment system in which “[a given] employer’s contribution to the unemployment insurance pool (i.e., contribution on wages paid out) would vary by what that employer has spent to cover the unemployment periods of the particular freelance and contract workers he/she/it has employed” (p. 509).

21 Despite the overall coherence of these results, there can be no doubt that Le travail créateur will meet with strong resistance from French sociologists—and not only because of its demystifying approach to art worlds. Much of that resistance will be due to profound disagreement with the author’s epistemological orientations and a critical view of the philosophical anthropology underlying his sociology, particularly the sophisticated—some will say unrealistic—type of rationality he attributes to actors. That competence surely deserves a more precise definition than the one sporadically offered. To conceive action in a situation of radical uncertainty of the sort operative in all art worlds, Menger turns to the philosopher Gilles Gaston-Granger and defines minimum-level individual rationality as a certain “consistency in the way one deals with uncertainty” (p. 97). As he sees it, following Mead, taking action implies a particular kind of reflexivity wherein the subject internalizes the way others see him. In the creative process this reflexivity takes the form of self-comparison; that is, between what one has already accomplished and what one might possibly yet accomplish, as the philosopher Hintikka understands it (p. 466). If we combine all these features, we can only conclude that Menger has accepted a highly demanding version of actor rationality wherein we are all Bayesian statisticians. I do not wish to engage directly in a meta-theoretical debate that would require me to object to his very conception of the actor, thereby running the risk of being confined to that debate without being able to discuss the main results of his proposed sociology of artistic creation—to do so would not do justice to Menger’s labor and what he has constructed. My comments shall therefore focus on two points in his overall argument: his analysis of relationships generated through selective job-matching and how those relationships work; his definitions of two concepts central to his construction: uncertainty and talent. My remarks will be seen to converge on the point that Menger’s model imputes much of the differences in artistic careers to a mechanism— “selective job-matching”—whose social logic he does not sufficiently clarify.

22 Specifically, the author leaves in shadow the activity of artistic labor market intermediaries, brokers and gatekeepers and the role they play in creating inequalities among artists and artistic careers. More fundamentally, he does not grant enough weight to the constant struggles among the different art world actors to impose their criteria for judging and ranking artists and artworks and thus to control the process through which talent hierarchies are established. Sociological study of art world stratification and of the inequalities in those worlds—only part of Menger’s ambitious program but a crucial one in Le travail créateur—has to specify whether or not art world employment conventions and conventions for evaluating creative work are efficient or arbitrary and what their possible imperfections might be due to. In this connection, can a perspective such as Menger’s do without the notion of asymmetry in social relations or the idea that social power relations—at least some of them—are a necessary feature of inequality-producing mechanisms? [5] It is this question that I now put to Menger’s sociology of art and inequalities.

Matching, social capital and transaction costs

23 As we have seen, the selective matching mechanism can be summed up by the idea that on artistic labor markets, protagonists with the same status in their respective activities, the same “reputation level,” the same degree of “artistic and social influence” (p. 423) are likely to work together. According to Menger, “sequential matching games” and successive “competitive tournaments” that produce rankings are what reveal differences in talents and skills—differences unknown to the actors themselves beforehand—and so account for the gradual crystallization of hierarchies of artists, activity volumes and incomes. “The successful career dynamic,” he adds, “amounts to upward mobility within a stratified world of interacquaintance networks and recurrent collaboration” (p. 355). [6] This dynamic was confirmed by the American sociologist Robert R. Faulkner ( [1983] 2005) in his analysis of employment and activity among Hollywood musicians. Faulkner showed the existence of “recurrent coalitions” of producers and composers, coalitions correlated with members’ productivity levels; that is, number of films they had worked on.

24 This outcome naturally grants a determining role to social capital as a factor of artistic success. Amplification of initial differences in talent and how “robust” artist actions prove in situations of uncertainty depend on each partner’s social capital (Padgett and Ansel 1993). [7] Here it seems to me we would need to inquire into differentiated investment returns not only on different volumes but also on various types of social capital (e.g., the ratio of strong to weak ties, number of structural holes in a given individual’s network, etc.). This resource and its effects tend to be subsumed in Menger’s model under the concept of “selective matching.” And though the author convincingly criticizes the almost magic concept of “homology” between fields, dispositions and positions in Bourdieu’s sociology for its limited explanatory power, dubbing it the “black box” of structural reasoning in the social sciences (p. 44), the question arises whether his “selective matching” notion is not doing the same duty and being granted comparable powers in his analytic construction. Menger seems to think this idea will explain the virtually natural coming together of art world actors whose relative social positions are “homologous” in that they are at a comparable level in their respective professional areas.

25 The fact is that employer-employee matching modes vary in time, space and within the different regions of social space. The question arises, for example, whether the increased transaction costs linked to demographic growth of creative worlds affect the social shape and dynamic of selective job-matching and how efficiently and fully the matching mechanism fulfills its function of revealing and amplifying differences in talent. For example, is it the same thing for writers to be selected by editorial committees, publishing editors, or external literary agents of the sort that brought about the decline of independent American publishing houses by giving authors incentives to turn to big corporate publishers (Thompson 2010)? Another example: Do the current short intermediation chains between artists and audiences on the internet have the power to modify the mechanisms by which artists’ reputations and careers are constructed? And what happens to the mechanisms that fuel circulation of information on skills along with talent assessment and talent revelation when artistic labor market intermediaries become prescribers, as in the audiovisual industry? In sum, there is no reason to assume that the various types of job-matching will always work to reveal talent and bring together individuals or groups with similar reputation and skills levels. For Mark Granovetter (1974), for example, not all the relations that help a person get hired can be acquired through strategic calculation; the effects of some of those relations are unpredictable because due to preexisting ties developed outside the work world.

A disembodied sociology?

26 All these questions are reason to examine more attentively than Menger does the sociography of artistic labor market and art market intermediaries, brokers and gatekeepers, and exactly how the intermediation function gets socially organized as a “system of professions” and a territory in which several tasks are shared by a multitude of protagonists (Abbott 1988). [8] Such an undertaking requires entering into the “production mode” of selective job-matching and the arrangements among intermediaries who play a role in the social selection of creators on the various artistic labor markets. Several economic sociology studies have explored markets this way, including the artistic job market. In French-language social science, for example, Christian Bessy and François Eymard-Duvernay (1997) sought to differentiate between types of intermediation in hiring as a function of worker skills, on the one hand, quality of work assessment modes on the other. In the case of modern and contemporary visual art, Moulin (with P. Costa [1992] 1997) showed that the intermediation system of selecting artists changed over time: whereas what used to produce artist and artwork value was an alliance between critics and gallery owners, the more relevant pair is now the gallery owner and the exhibition organizer or curator. Clearly, then, the weight of the economic dimension of artworks when it comes to assessing artists’ “talent” has changed. Just as Menger points out in passing how the unemployment insurance system for freelance and contract performing arts and media industry workers distorts information exchange between employers and artists on individuals’ respective skills and strong points, so it is reasonable to hypothesize that other organizations implicated in matching artists with available jobs produce “imperfections” in the ideal circulation of “signals” about artists.

27 Sociology of intermediation professions together with sociology of the networks underlying market transactions would provide two powerful tools here for penetrating the mystery of job-matching, differentiating the types of social logic and structures it involves, analyzing what goes wrong in connection with the adjustment function Menger attributes to them. One benefit of supplementing his study this way would be a more precise characterization of the variables making up what he calls the “status” (pp. 329-331) of artistic labor market intermediaries—agents, impresarios, gallery owners, managers, administrators, etc., the privileged partners of professional career artists—and the way that “status” itself operates as a “signal” (Podolny 2005).

Artistic creation up against multiple evaluations

28 The second benefit of taking into account the many ways that job-matching is actually organized is to fully recognize the multiplicity of evaluation criteria that artists are subjected to and the consequences thereof. Menger is fully aware of this diversity, but as far as he is concerned it always seems to get absorbed into a hierarchy of quality-determining criteria which, if not fair or just, is at least always justified. First, he understands comparison, particularly numbered rankings, to be the appropriate judgment (and justice) mechanism in art worlds. But is it the only legitimate one? On what Lucien Karpik calls “singularity markets”—of which artistic labor is one—a multitude of devices exist for bringing together supply and demand: the “cultural complex” of critics and enlightened art or music lovers, for example, or the fact that a certain work force is concentrated in a certain area, meaning that an employer there can recruit virtually anyone, regardless of skills sought (Karpik 2007). [9] Sociological study of recruitment agencies has shown that the higher their prices, the more new, privy information on job candidates they are able to produce, whereas less expensive agencies offer their clients sets of candidates drawn up on the basis of coarser criteria. In this context, a fall in hiring costs in a given sector or company may bring about hiring discrimination by limiting available information, or information deemed relevant, on candidates (Bessy and Eymard-Duvernay, 1997: xviii-xix).

29 One weakness of Menger’s analysis is that the element of arbitrariness in job-matching forms and formats and the way artists (and other art world protagonists) are perceived and ranked gets reduced to a problem of market disequilibrium or imperfect information on skills. For their part economic sociology and the “economics of conventions” have highlighted the constraining and possibly discriminatory nature of evaluation conventions, conventions consubstantial with all market operation and crucial in dissipating the uncertainties that are part and parcel of market transactions. On labor markets, explains Eymard-Duvernay, there are several “conventions about skills,” several “grammars of information on how well workers work, the particular aim of these grammars being to provide a basis for evaluations of work quality and productivity” (ibid. p. 19). These grammars organize information that is useful for action and make that information available, but they also provide rules on what can (and therefore what cannot) legitimately be done. For this reason they may lead to various sorts of social boycotts.

30 Le travail créateur also uses the sociological concept of “convention,” but in the more general sense that Howard S. Becker gave to it of a set of habits, techniques and organizational solutions that make it possible to save on the costs of acting and coordinating protagonists in art worlds. It is these conventions that temper and above all channel the love of risk and novelty that Menger rightly attributes to great artists. “Without conventions,” he writes, “without interaction rules, without more or less stable procedures for dividing up tasks and mutually adjusting expectations and exchanged meanings, there could be no cooperation between the people called upon to work together to produce, diffuse, consume, evaluate and preserve artworks” (pp. 204-205). From this perspective, highly similar to Becker’s, there can be no doubt that how artists and employers get matched partakes of the set of conventions particular to any given art world. However, Menger is hesitant to point out that conventions are constraints—though he is occasionally inclined to do so (see, for example, p. 329). This reluctance prevents him from inquiring more deeply into the existence of errors in matching and the exclusion caused by the cumulative advantage mechanisms he brings to light. Citing, as Menger does, Merton’s explanation of social hierarchy construction by means of the “Matthew effect,” Harriet Zuckerman (1977), at the end of her study of the scientific elite of American Nobel Prize winners, emphasized the injustices caused by those mechanisms.

Two levels of uncertainty

31 It is not only artist quality and artwork success that are radically uncertain. There is also an even more radical, second-order uncertainty, namely as to what devices or rules should be used to reduce uncertainty about artists’ qualities; i.e., uncertainty about the criteria used to judge their talent. Up against this second-order uncertainty, symbolic goods market intermediaries struggle to prescribe their own categories of perception and criteria for assessing competence. [10] This is why modes for matching artists with employers and for judging artists’ respective quality levels are initially fairly independent of each other and may well be in conflict before possible ranking. Moreover, contemporary science and research is also a theater for these battles between quality assessment modes. Do sociologists have to give up accounting for these battles and the socially grounded reasons for which they are lost or won?

32 Menger has anticipated some of these questions. There is little to be gained, he warns between the lines, by raising the level of the question put to the social sciences by the existence of uncertainty in social transactions; in other words, by inquiring into competition among “judges” and market intermediaries rather than competition among creators. Doing so seems a sure-fire way of getting caught once again in the social constructionism trap and its denial of objective differences in individuals’ talent or skills. Attentive readers of Le travail créateur will see how wrong it is to view Menger first as a fierce theoretical enemy of Bourdieu and Bourdieu’s sociology. Menger is indeed critical of what he sees as his former master’s determinism. But the most consistent enemy in this book is social constructionism. The author’s first purpose is to lay low the idea that social structuring is contingent. He has also been criticized for having “naturalized” ever-socially-constructed differences. Nothing could be farther from the truth. But he does mean to put a stop to the infinite regression to which, as he sees it, constructionist reasoning leads. [11] To do so he postulates two art world universals: uncertainty, understood as an integral part of action, and talent scarcity (not, as some have said, talent as natural and innate).

Difficulties in defining uncertainty and talent

33 This creates two problems. The first concerns the concept of uncertainty. In Menger’s theoretical architecture, uncertainty is an ontological category that allows for characterizing artistic action, as in the perspective opened by Dewey (1934), though this author is not at the core of Menger’s model. But uncertainty is also a consistent feature of the ordinary social experience of artistic labor market protagonists. The danger here is to collapse the second meaning of the uncertainty notion onto the first, thus giving in to a kind of ontological paralogism. [12] This would preclude sociological inquiry into the possible existence of differences in how uncertainty works in the various art worlds and whether these differences are perfectly homologous with observed differences in talent. We can at least hypothesize that subjective ways of experiencing and reacting to uncertainty or envisaging one’s own future are unequally distributed, and that this distribution is independent of the creative aptitudes hierarchy precisely because that hierarchy is never immediately revealed to art world actors.

34 The second problem concerns the relation that can be constructed in Menger’s model between talent and value, be it economic, reputational or esthetic. The specific risk in the model is that the argument will end up being based on a circular relation in which value proves talent and talent value, just as an impresario’s status or reputation is thought to confirm the talent or reputation of the actor s/he works for. To escape this there is really no other solution than to get down to examining the real making of matches: how artists are selected and ranked by other artistic labor market actors.

Two approaches to inequality

35 These criticisms are certainly not meant to suggest that Menger’s idea that artistic labor market networks produce their own “endogenous hierarchies” should be scrapped because those hierarchies can be assumed unfair or exclusionary. For my part, I would suggest that the right way to analyze the origin of inequality in the arts is to account for that possible injustice in some other way than individual resources (with the understanding that talent is a variable that aggregates other variables) and social mechanisms whose cognitive (natural) component (comparison, ranking, anticipatory calculations, etc.) is emphasized to the detriment of its relational (social) dimension. And to do so, for example, by inquiring, as Ronald S. Burt did, into intra-collaborative network position differences, information asymmetries, the power relations those asymmetries reveal and their effects on artist activity and careers.

36 We have gotten into the habit of distinguishing between two major social science model types: equilibrium models and power models. It can be said that each of them analyzes what the other has failed to conceive. Power models generally have trouble distinguishing the arbitrary from the necessary. Equilibrium models such as Menger’s have difficulty distinguishing between a multitude of possible equilibria. But even in the latter type of model, power can have a place. Andrew Abbott (1988: 134-135), for example, called “professional power” a force “retarding competitive equilibrium among professions in a system of interprofessional competitions in order to appropriate and preserve tasks for itself.” In the sociology of artistic careers and activity that Menger has developed, is power also a factor of temporary disequilibrium? Or is inequality in the arts founded on mechanisms that are not only unfair but illegitimate? While Menger is right to say that by its very principle social constructionism leads to a regression in sociological explanation, that regression is not infinite in survey practice or real social worlds. In the worlds he has studied, it can be brought to a halt by the observation of power relations among individuals and groups, relations that force inadequate or controversial categorizations and ranking criteria (good artist/bad artist) on other artistic labor market protagonists. This is why sociologists should also analyze relations among intermediaries, brokers and gatekeepers on these markets and those between intermediaries and artists. What are art-world intermediaries’ interests? What kind of cooperation and competition relations do they entertain? Are those relations institutionalized, and are they imposed on artists? Have they changed over time? Have they always had the same degree of influence in determining talent? All these questions call for answers in analysis of creative work. Answering them is crucial if we are to understand inequality in the arts.

37 At this point the question arises as to the scope to be attributed to Menger’s accumulated results. He has told us of his intention to test some of his hypotheses for the worlds of science and academia. He also raises the question in Le travail créateur as to whether his analyses apply beyond art worlds. Positing the existence of aptitude differences to explain social hierarchies might also be relevant for sports, politics and even business, he suggests (p. 294). We may even ask whether the current proliferation of comparative tournaments and qualifying contests, itself competitive, is a characteristic of contemporary cultural markets alone. The “project-based” organization that provides a social grounding for the “selective job-matching” dynamic is now in place in several social sectors and companies. Lastly, doesn’t the “uncertainty principle” that Menger shows to be of such crucial importance for artists also structure an entire set of markets, not to mention several innovative activities at the microsociological level? If so, does this mean Menger’s model of social hierarchy production should be extended to all so-called immaterial activities or to the most innovative sectors of contemporary capitalism and even beyond? This is what his Portrait de l’artiste en travailleur seemed to suggest, with its demonstration of how today’s artist is “the future worker incarnate” (2002).

38 A major contribution of Menger’s studies in the last two decades is to have invalidated irenic representations of art worlds as loci not of activity freed from all constraint (as in Romantic representations of creative work) but as outposts of the transformation of the capitalist production mode into more horizontal relations involving creative and cognitive cooperation. Indeed, Menger is continually pointing out the extremely inegalitarian character of social worlds of creation compared to other activity spheres. And though it is true that art worlds are today in the avant-garde of capitalism, Menger has made it clear that this is first and foremost in terms of job flexibility and wide differences in symbolic and economic earnings. The organizing and structuring of artistic labor and cultural capitalism markets thus represent excellent focus areas for sociological study of contemporary capitalism.

39 In the study of art worlds specifically, the scientific challenge now is to better determine the origin of observed inequalities in the careers and success levels of individuals with equivalent, similar or unknown social attributes and resources. On this point, models in which hierarchy is an emerging function of the social interdependence characteristic of creative activity, i.e., models in line with the one Menger has constructed, will surely differ in the future by how they situate and define power and the arbitrariness of artistic labor market cooperation conventions. As I have indicated, this debate should also focus on the historical dynamic of contemporary organization of cultural capitalism, a dimension that is touched upon but not sociologically analyzed in Le travail créateur. In Menger’s thinking, the social features of art worlds can be deduced from a limited number of principles and laws. But do we really know whether these principles and laws have operated in the same way in all periods? In other words, is inequality in art worlds produced according to the same laws at different periods in history? And should we then differentiate—and if so how?—between the current age of social relations between artists and their employers and earlier ones? How can sociologists account for shifts from one social mode of selecting artists to another over the course of history?

40 These questions and comments in no way diminish the sense of fulfillment one has upon closing this long, highly cultured book. There can be no doubt that Menger’s demonstrations will be cited and debated in the future. Moreover, several of the internal tensions I have noted and the questions that subsist are due above all to the fact that Le travail créateur is not a general theoretical work or a single, closed argument but a collection of reworked articles. Menger’s creative and intellectual project is extremely coherent. And he occupies a unique position in French sociology, with his one-of-a-kind scientific style, deliberately open-ended, unfinished—consistent in this with his image of the creative process itself, that of a somewhat fragile submersible setting out for a blue, uncertain sea, ready to take reiterated plunges into deep, occasionally dangerous waters but also to change course if necessary.