Version classiqueVersion mobile

Difference, Competition and Disproportion. The Sociology of Creative Work

Pierre-Michel Menger

Difference, Competition and Disproportion. The Sociology of Creative Work

Inaugural Lecture delivered on Thursday 9 January 2014

Pierre-Michel Menger
Traduction de Liz Libbrecht

Texte intégral

1Mr Administrator,
Dear colleagues, dear friends,
Ladies and Gentlemen,

2While this lecture is addressed to an audience whose attention I will endeavour to capture, my words also reach out to a host of “significant others”, as sociologists call them. There are the elders, there are those who have their place in the pages of a personal and professional biography, and there are all those with whom a daily conversation develops, in the solitude of reading and mental ruminations. Research work is constantly establishing its bearings, its genealogies and its debts, as well as its proximities and distances. However for all the easily traceable knowledge and hypotheses accounted for by citations, countless fruitful intuitions arise in our work, through a confused alchemy of conversations, reading and encounters that are more difficult to credit.

3Roger Guesnerie and Pierre Rosanvallon proposed to create this Chair, and Pierre Rosanvallon put forward my candidacy. I would like to tell them just how appreciative I am. Mentioning our discussions on the figures of equality or on the convergences between sociology and economics cannot convey my indebtedness to them. I am deeply grateful to Raymonde Moulin, who taught me to be a researcher, and much more. I would also like to pay tribute to two institutions: the Centre national pour la recherche scientifique (CNRS), where I worked for most of my career before joining this institution, and the École des hautes études en sciences sociales (EHESS), to which I remain loyal.

4This Chair is devoted to the sociology of creative work. Its disciplinary identity places it in the illustrious footsteps not only of sociologists who have taught here, but also of two who should have done so. I have in mind Émile Durkheim, whose failure to be elected at the Collège de France negatively marked those times, and Maurice Halbwachs who, shortly after being elected, was arrested and deported in 1944, never to return.

5While the titles of three Chairs seemingly indicated that they were devoted to the field of work, they were actually history Chairs. A careful exploration can easily show that work has been studied in many ways, in many disciplines taught at the Collège de France. Later on, I will mention some rather unexpected ones.

6My teaching will approach the creative dimension of work, one which points not only to a common theme – the study of creative work, especially in the arts and the sciences –, but also to the additional terms attached to the adjective, and that are required to define work as socially creative: value, identity, collective protection, status, etc.

7These terms, along with others, can easily be spread across a continuum between two contrasting but ordinary functional definitions of work: its instrumental and its expressive values. According to the former, work is the engagement of individual energy in more or less difficult conditions of physical effort and mental strain. The latter defines work as self-realization through the act of production.

8The distinction between these two values was shaped by a long philosophical tradition and was taken up by the social critique of the mutilation of work through its specialization and commodification. It only became analytically operational with the multidimensional characterization of all types of work.

  • 1 Jon Elster, Making Sense of Karl Marx, Cambridge, Cambridge University Press, 1985.

9The critical ambition underlying the argument according to which complex work is but a multiple of simple labour is well known. Based on this premise, Marx somewhat ambiguously proposed that work be homogenized, and that the economy of value, the remuneration of work and the situation of exploitation in work relationships be made to correspond with one another. As Jon Elster has shown, this correspondence falters when the full heterogeneity of work is acknowledged.1 The sociology that I wish to develop here consists of a multidimensional exploration of work.

10I will start this lecture by situating my research in relation to the evolution of the sociology of work. I will then explain how I drew on the study of artistic and scientific professions to understand why the most rewarding work, that which has expressive value, also has the most uncertain outcome, and why it is remunerated highly unequally. After examining some of the innermost mechanisms of inventive work, I will present the twofold understanding that is required in the analysis of work if we wish to identify the conditions under which it is creative.


11The research agenda on work has evolved with the transformations of our societies and their economies. The sociology of industrial labour was central to the earliest sociology of work. The employee category, which later became prominent, raised another issue: that of the wide range of occupations and employment conditions in an economy that had come to be dominated by services. From the 1970s onwards, rising levels of qualification in the workforce and increasing numbers of highly skilled jobs generated growing interest not only in the work of managers and the intellectual professions, but also in professions with high levels of expertise, a legal monopoly of practice and market positions that were often at the interface between the private and public spheres – primarily doctors, lawyers and architects.

12Studies of the growth of female employment have been instrumental in bringing to light the full diversity of visible and invisible discrimination mechanisms at work on the labour market. While women’s qualifications have progressed faster than men’s, pay inequalities subsist even after controlling for the effects of the various individual characteristics, disparity in returns on initial training in different occupations, over-exposure to part-time work, and self-selection as far as participation in the labour market and the choice of professions are concerned. One of the most eloquent studies on discrimination against women looked at the hiring process in classical music orchestras: with the introduction of blind auditions, the proportion of women in orchestras increased enormously.2

  • 3 Alvin Gouldner, The Coming Crisis of Western Sociology, New York/London, Basic Books, 1970.

13In 1970, Alvin Gouldner argued that sociology was becoming the assessor of the failures and achievements of the welfare state.3 The tasks that it could undertake in France were ambitious: analysing and evaluating the extensions of social welfare, the institutions and regulations of labour markets, and the expansion and performance of the education system. Ten years later, however, unemployment was starting to take root, and contractual forms of employment proliferated. The opposition on the labour market between insiders, who were protected, and outsiders, who were exposed, became stronger and entrenched, to the detriment of recent entrants. This also hampered professional mobility. After criticizing the disintegration of work, sociology became critical of the disintegration of employment.

14Professions and employment systems in the arts, which I have studied extensively, have been part of the sociological research agenda for about thirty years. In the 1950s, the Institut national de la statistique et des études économiques (INSEE) had grouped artists, the clergy, the army, the police and sports professions into a single category, along with a few other unclassifiable professions like mediums and radiesthesists. In the new classification of occupations, which was revised under the influence of the most advanced sociological theories and was implemented in 1982, the INSEE grouped artists with managers and higher intellectual professions, alongside researchers and teachers. However the first census using this taxonomy revealed singular traits among artists: their economic position was similar to that of intermediate occupations, and they presented widely disparate incomes with a high rate of underemployment and pluriactivity, unparalleled in the higher professions category.

15As all international comparisons have shown, inequality scores in artistic and scientific professions are some of the most elevated amongst higher professions. The distribution of volumes of activity and of monetary gains in the arts and the distribution of reputation and consecration in scientific professions are highly asymmetrical: 80% of the monetary or symbolic remuneration in circulation is concentrated with 20% of the individuals in these professions.

  • 4 Raymonde Moulin, Jean-Claude Passeron et al., Les Artistes. Essai de morphologie sociale, Paris, La (...)

16Moreover, in the arts, the proportion of people either not making any profit or making a loss – i.e. after expenses incurred to practice their activity – in a given year is greater than in any other higher profession. It thus comes as no surprise that at least eight possible definitions of artistic professionalism exist. Raymonde Moulin and Jean-Claude Passeron were able to untangle this imbroglio by studying the visual artist population.4 The work on conditions of employment and activity in the performing arts that I carried out with my EHESS and CNRS team directly benefited from their research.

  • 5 Christian Baudelot, Michel Gollac et al., Travailler pour être heureux? Le bonheur au travail en Fr (...)

17The particularities of these professional worlds can raise the question as to whether creative work is the right side of work, the flipside, or a possible ideal thereof. Answers vary, depending on the criteria. Creative and research jobs are highly appealing: they score very well on scales of occupational prestige, for example in the measurement of work satisfaction, as Christian Baudelot and Michel Gollac’s study on happiness at work has shown.5 But they present an equally high risk of failure, underemployment or hybrid practice, in combination with less interesting side jobs.

18While social and personal gratifications are therefore important, they come with high probabilities of economic failure, at least in the practice of the most highly valued part of the activity! Does the choice of such jobs represent an exceptional case of preference for risk?

19This question can be answered by arguing that the remuneration of work is twofold. By this I mean its monetary component (income) and its non-monetary component (the personal, psychological, and social gratifications that are associated with the practice of stimulating, rather than routine work.)

20The analysis of the double cost of work is rooted in history: from Aristotle to Hannah Arendt and beyond, the value of work has been assessed both according to its degrees of direct or indirect utility and to its certain or uncertain chances of accomplishment. The double semantics of work expresses this well, opposing labour to work, and burden to creation. This argument enabled Durkheim to see art as a fortunate embodiment of the civilizing value of activities that were neither assigned to instrumental purposes nor treated according to the usual laws of market economics, even though Durkheim was wary of artists’ individualistic intemperance and of the contagious hedonism that their imagination could trigger in society. Max Weber stressed the rationality of what he called artists’ “anti-economic behaviour” which, according to him, was the only thing that could ensure the charismatic exercise of activity and the deployment of inventiveness.

21Yet artistic and scientific work is steeped in evaluation. Judgement, comparison and the differential attribution of value underpin the allocation of jobs and resources and the organization of careers. Competition is omnipresent.

  • 6 Pierre Bourdieu, Science of Science and Reflexivity, Cambridge, Polity Press, 2004.

22In his Collège de France lecture, which he devoted to science, Pierre Bourdieu strongly emphasized the control scientists have over their activity, through the mechanisms of internal competition.6 A comparison with the arts should reveal considerable differences since the artistic field is also affected by the power of entrepreneurs and intermediaries, along with their multiple ways of influencing the public’s preferences, to the point of exacerbating individual and collective sensitivity to differences of quality and originality. And the objectification of value does not imply the certification and cumulativeness of results, as it does the case in the sciences.

23However, both in the arts and the sciences, competition is a driver of professionalization. Personal reputation is founded on the verifiable sum of achieved and credited productions, and on what they promise for the future. Reputation contains the information necessary to the functioning of what are called professional labour markets, the characteristics of which are well known: competition based on expertise; the transferability of skills from one organization to another; and the functional role of peer recognition, which reflects the influence of the professional community, beyond the walls of each organization.

24The sciences and the arts have shaped not only the tools and the culture of evaluation, but also the critique of metrologies and their instrumentalization. When an activity presents low risk, innovation is simply incremental: it can easily be evaluated. But the greatest value is attributed to originality and breakthrough innovations. Their emergence and the instant valuation of their products are however far more uncertain. Trial and error and the rejection of the tyranny of the short term are efficient principles for estimating their impact. A rationality of inattention exists with regard to evaluations.

25The fact that, in creative activities, each person’s work is embedded in a network and a chain of collective projects makes it more complicated to evaluate. This collective production needs to be broken down into individualized contributions, into name credits, and into lists of co-authors of publications and works. To attest to the share of work attributable to each person and to its relative importance is to give the activity its fair share of recognition. A significant component of the non-monetary compensation of complex work lies in the named identification of individual contributions to the collective result. This identification is also a way of informing efficiently professional communities on each contributor’s qualities and responsibilities. Thus, through the attribution of personal reputation, career continuity can be distinguished from the discontinuity of work sequences.

26The individualization of activity and of its identification is not the opposite but the very substance of a certain form of collective organization of creative work. The coefficient of individualization of work and of its rewards varies, as does the organization of work. A writer neither works nor innovates like a composer or film-maker, no more than does a sociologist’s work directly compare to that of a biologist or a CERN physicist.

27In many scientific disciplines, the list of article co-authors is lengthening. Jobs are diversifying, specializations are more and more finely divided, and the volume of work collaborations is growing. Without trying to force a comparison, we witness a similar trend in film credits which are far more detailed than before, particularly though not exclusively because budgets and teams are larger. The expansion of professional communities goes hand in hand with the segmentation of activities, and with that of research domains. Incidentally, the meshing of collaborative links and the frequency of disciplinary hybridizations make it impossible to consider this segmentation as the simple juxtaposition of clusters of specialization.

28With the expansion of professional worlds in the arts and the sciences, the demand for evaluation is constantly increasing. Whether regarding people, teams or organizations, the number and variety of comparisons, rankings, records of achievements, bibliometrics and all kinds of metrologies has grown faster than the variety of products or services offered.

29Has this resulted in success being less scarce, in reputations being more volatile, and in a greater quantity of social esteem to go round? The answer provided by research on these subjects is unambiguous: inequalities of material and symbolic remuneration have in no way decreased.

30Of course, activities with uncertain success do not all speculate on talent with the same intensity. Expected returns are highest in the entertainment industry and that of the most popular sports, as these represent the largest markets. The winners of competitions reap the greatest rewards and flaunt them. Nevertheless, the least industrialized creative activities also have their own accolade systems. The value of some is revealed more gradually, but also more sustainably, which constitutes a good index of the degree of control exerted by a professional community over its competition system and over the guarantees provided to legitimize its evaluations.

31Whether pertaining to individual achievements or to the overall result, evaluations are designed to identify quality differences and to highlight and reward them. The cost of these differences is correlated to the size of the markets and organizations that sustain and remunerate them. All evaluations are measurements. In infinitely differentiated worlds of production, the spread of a culture of measuring qualities and their commensurability seems to be a tool of perfect competition. Such is the case if it reduces the strategic exploitation of information asymmetries and corrects the influence of those reputational hierarchies that have become more lasting than legitimate – and their associated incomes. But a culture of reasoning and action founded on comparison and benchmarking weakens measurement’s emancipatory power if it overlooks its arbitrariness.

32The following question arises: do evaluations identify quality differences? Or do they create them? Or do they amplify them disproportionately?

  • 7 Ian Hacking, The Social Construction of What?, Cambridge (Mass.), Harvard University Press, 1999.

33The constructionist critique suspects that they create them, arguing that causality is reversed: far from being markers of incompressible differences, categories of talent and excellence are merely products of market competition systems, or of systems mimicking the market, and they grant incommensurate value to qualities selected arbitrarily. The only differences that should matter are horizontal differences of originality, which are multidimensional. This suspicion, at the apex of the scale of constructionist radicalness imagined by Ian Hacking, implies the need simply to abolish all forms of singularizing differentiation and to dismiss most of the gaps in competency, expertise and value.7

34This turns into a devastating critique if used to suggest that by hiking up work and career individualization, creative activities constitute extreme examples of competitive worlds in which the professional game is a zero-sum game, where the winner takes virtually all. In this case, the asymmetrical distribution of material and symbolic remuneration is nothing but the product of a social rarefication mechanism.

  • 8 Jean-Paul Sartre, Critique of Dialectical Reason, vol. 1: Theory of Practical Ensembles, London, Ve (...)

35Jean-Paul Sartre defined scarcity as “the negative unity of the multiplicity of men”.8 Whereas the fundamental human relation is reciprocity, i.e. the recognition of the other as freedom and a project, scarcity, Sartre argued, transforms this relation into a force field, since “there is not enough for everybody”. In a society reduced to a zero-sum game, everyone sees the other as a predator. By inference, work can become expressive only in a world where resources of all sorts are sufficiently abundant for competition no longer to come into play.

36Thus radicalized, the constructionist critique separates and opposes what creative occupations associate in a singular way: the significance of autonomy at work, the importance of collaborative relations, these professions’ appeal, their openly competitive nature and their prestige, i.e. the high quantity of social esteem they attract.

37The problem must be rephrased if it is to be addressed. We can agree that creative imagination and inventiveness are inexhaustible resources, perhaps our planet’s only inexhaustible resources. Using them intensively can actually make them even more abundant. So are they underused because of the system of inter-individual competition and the scarcity of the resources allocated to these activities? Or are they not rather articulated through relations and transactions that do not add up to a zero-sum game, and that greatly complicate the simple demarcation between competition, reciprocity and cooperation?

38One of the challenges facing the social sciences is to explain, using a single explanatory model, the social prestige of these activities, the divergent trajectories of professional achievement and the huge differences in earnings, whether they be monetary or not. From the reverse perspective, some of the most important studies of the causes of professional inequality and the mechanisms of its amplification have their roots in sociological and economic analyses of the arts and the sciences.

39There is another, complementary challenge: how can one explain that such inequalities are better tolerated in the arts and the sciences than in other professional worlds and, in some cases, are even powerfully consecrated as legitimate forms of identification and remuneration of quality? Is it because of the intrinsic uncertainty of the professional game? Or is it because there are two remuneration currencies? Or yet is it because of the hardly questionable difficulty of objectifying value? These are serious issues, which I will now begin to explore.

40In the study of earnings, a factorial reasoning based on human capital generally accounts for only a third of the income variance. And with regard to the professions I am concerned with, these explanatory models are much harder to adjust, particularly because of the differences within these professions: discrepancies in gains or renown can reach considerable levels, even between two professionals with the same training and amount of professional experience.

41In order to explain the choice of such professions, should we automatically refer to non-monetary gratification as a perfectly equalizing complement, which makes up for the financial shortfall? If so, then non-monetary income or psychological gratification from creative work should be distributed exactly inversely to monetary income. This would ultimately justify all forms of precariousness. Romantic writers invented the concept of “the loser wins”, according to which psychological gratification is provisionally a substitute for income, and is even the only possible guarantee of remote future success, since selflessness was thought to protect creativity from corruption. But the practical issue of the immediate remuneration of work was thus sidelined from the outset.

42One way of pursuing the factorial reasoning is to introduce an individual heterogeneity variable relating to latent or underlying determinants of individual value. This is a well-known problem in studies of returns on education, and consists in identifying endogeneity biases. Not only do the length and orientation of studies depend on individuals’ resources, but the quality of the successive learning sequences also acts cumulatively, both to shape the preferences for a type and a level of studies, and to increase or reduce the returns on educational investment.

43Individual heterogeneity involves various determining characteristics of social origin, as well as markers of individuation and context. In the case of differences in the remuneration of work, the heterogeneity argument can be formulated as follows: individuals initially have capacities which they develop unequally, based on the opportunities for activation provided to them by their early socialization, by the multiple forms of support for individual and collective investment in education, and by the formative power of the whole everyday extracurricular environment. Their preferences differ, irrespective of how they developed, which can explain, for example, varying sensitivity to the non-monetary dimension of remuneration for work.

44The visible and measurable result is a range of knowledge and behavioural dispositions, which become more effective if they are well combined professionally. But the substance of individuation, in other words everything that stems from the multiple factors of differentiation, is neither entirely detected nor activated.

45This is where the argument of ability or aptitude comes into play: the range of skills held by an individual is the product, among other possible ones, of the sum of qualities that the individual could develop. This is indeed what the semantics of competences relates to in the work evaluation toolkit. Reacting to labour deskilling theses, sociologists of work have stressed the tacit competences that are applied even in the most basic forms of work, the real value of which is not recognized.

46All in all, the individual is greater and more diverse than the contents of the knowledge that he or she has acquired and has had certified, and that he or she is paid to apply in his or her work. Disproportionate differences in professional trajectories can develop from the interaction of two sets of factors yet distributed symmetrically: individual qualities, and chances to develop those qualities.

47Attempts both to identify abilities and to develop them are constant: in higher professions, this is evidenced in the fevered search for what is called talent. The word is used to aggregate an unstable variety of traits and qualities into a mysterious, evocatively-named grandeur. The vocabulary of talent, which was introduced and disseminated in the business world over the last fifteen years, denotes a mix of knowledge and aptitudes identifiable only through comparisons between individuals, in order to extract all the “potential” they hold.

48The study of creative work provides useful tools to decipher what individual heterogeneity means. Let me start by saying that there are two ways of missing the point. The zero-sum game reasoning I mentioned earlier stresses the exclusively destructive nature of competition. It therefore also calls into question all forms of domination, which are actually challenged by competition itself. The only solution is to call for a world in which there would be no more gains to expect from individual differences than from inter-individual interactions. When Marx saw the artist’s creative activity as a positive ideal of work and of its possible transformation, it was based on the premise that work would no longer be a matter of market and competition. But that was at the cost of an untenable anthropological requirement: proscribing not only all specialization of work, but also all comparison and even all inter-individual communication likely to reveal gains from exchange and specialization.

49The second way of missing the point is the essentialist tautology: the proof of individual differences is right in front of us, in the exceptional value of some people’s work, which we can appreciate and admire. The argument does not go much further other than adding that all obstacles to the expression of talents should be removed, so as to free their potential.

50I will advocate another approach, which consists in introducing a high variability coefficient in inventive activity. The aim here is not to emphasize the randomness of a genetic or social lottery, which endows some individuals with fortunate dispositions to being inventive and innovative, but rather to break down work into sequences with very different properties. While this approach is centred exclusively on the resources and behaviour of the individual alone, and its value is above all heuristic, the sociologist can use it if he or she exploits the argument to highlight the collective dimension of the process leading to invention. I will first examine the argument, and then detail the sociological framework needed to enrich it.

51Before the proliferation of widely diverse research on creativity over the last forty years, the cumulative effect of which is uncertain, many essays were written from the late nineteenth century to provide a psychological characterization of creative imagination and of scientific inventive work. The authors included, notably, Helmholtz, Ribot, Souriau, Paulhan, Poincaré and Wallas, and a little later on Hadamard, Simon, and Campbell.

  • 9 Donald Campbell, “Blind Variation and Selective Retention in Creative Thought as in Other Knowledge (...)

52This research generally described the succession of conscious, infraconscious and subconscious phases of creative work. A host of ideas and hypotheses is developed and then subjected to multiple tests, trial and error exercises or encounters through unlikely comparisons and associations. The result is closely examined to identify what it is promising to keep and exploit and, far more often, what must be discarded. To paraphrase Donald Campbell, creative activities require the intensive accumulation of material, a mechanism generating variations in associations, combinations and recombinations of ideas, a selection process, and a mechanism to preserve and reproduce the variations selected.9

53One can call “luck” the variability needed for this psychological activity, and “intuition” the capacity to identify fruitful solutions. This is a way of pointing out, perhaps too conveniently, the unpredictable and emergent nature of the result. But this description resonates with researchers and creators’ accounts, even if it results from introspections that are difficult to verify. François Jacob spoke of “day science” and “night science”: infraconscious associative activity and intuition are the nocturnal side, while organized and rational work makes up the diurnal side.

54What matters is the relationship and tensions between the two forces, or between the two sides. Were it just a matter of accumulating material, the act of creation or invention would amount to the imitative recombination of existing solutions. But were creative work too closely associated with randomness, it would be out of the reach of any will or committed effort whatsoever: everything would happen by lottery, as naïve understandings of genius or talent maintain.

55The point to remember, in order to extrapolate the analysis to another level, is that a considerable disproportion exists between the quantity of what must be made available – accumulated knowledge, exploration time invested, trial and error, reviews, repetitions, rejections, starting again – and the selectiveness of the ultimate results of the process.

  • 10 Paul Valéry, “Variété”, Œuvres, Paris, Gallimard, coll. “Bibliothèque de la Pléiade”, 1957, vol. 1, (...)
  • 11 Roland Barthes, Writing Degree Zero, London, Jonathan Cape, 1967.

56Over-abundant trial and error to carry out explorations with an uncertain outcome: a truly harsh economy and morality of endurance. Paul Valéry expressed the negative accounting of creative work as follows: “Rigorous work, in literature, manifests itself and operates through refusal. One can say that it is measured by the number of refusals, … the number of options rejected”10; and, I should add, at the risk of excessive and sterilizing self-criticism. Roland Barthes saw the insistence or even conceit of writers who from the nineteenth century sought to depict themselves as artisans or home-based workers, as a justification of the value of work, which was to legitimize the cost of an activity rooted from then on in the systematic search for originality and novelty.11 This goes to show just how important the work-uncertainty twosome is. It activates a somewhat indirect rationality: in an activity with little guarantee of its chances of success, success will be obtained essentially as a by-product, as Jon Elster has argued.

57Far from confining us to the description of a series of states, psychological and cognitive research on creators’ work encourages us to enrich our understanding of action.

  • 12 Gilles-Gaston Granger, Essai d’une philosophie du style, Paris, Armand Colin, 1968.

58Gilles-Gaston Granger developed a stylistics for exploring scientific production as work and in many respects for bringing it closer to artistic creation. In this kind of work, individuation is the somewhat stereoscopic outcome of multiple competing and simultaneously possible structuring.12 This takes us to the dynamics of activities that, because of their degree of inventiveness, hold multiple possible pathways, not only when the work is first undertaken, but while that work is actually underway. In short, the argument that can be found in Maurice Merleau-Ponty or in John Dewey’s pragmatic philosophy is that the tension needs to be maintained between the unpredictability, the uncertainty, the game and the surprise associated with variability during work, as well as routines developed through previous learning.

59Realizing that the course of their work is not entirely decipherable, creators may want to retain multiple traces to try to understand its process, to review its result, or to draw from the wealth of unexploited possibilities. I will examine this in one of my lectures, so as to answer a seemingly trivial question: how does one go about finishing a work or not finishing it?

60In order to show what sociology can draw from such studies, a theoretical framework is needed to analyse work as a system of action, and to configure the environment of the action using the same terms. The creative work scene, which until now has focused overly on the individual, will thus rapidly be populated.

61Social reality can be understood as a system of relations. Each entity bears the imprint of its relations with its environment. While a considerable share of this environment is simply the persistence or reproduction of earlier states, part of it does vary, based on constant recombinations of the relations between all the significant units of the field of action. The situations in which the action unfolds are formed as changing doses of stability and variability.

  • 13 John Levi Martin, Social Structures, Princeton, Princeton University Press, 2009.

62If the action occurs in a system of relations, interaction constitutes its cornerstone. The rational substance of individual action is then defined by the perspective adopted by each actor on the behaviour of others and on the conclusions that can be drawn to orient their own behaviour and decisions in joint action. Relations in interaction can be symmetrical (an alliance), asymmetrical (reciprocity, which is not guaranteed to consolidate), or anti-symmetrical (domination).13

63The actor’s environment is defined in similar terms: when this environment deviates from its previous state, the action is no longer automatically and routinely adjusted to it. If the environment varies, it can be defined as a sum of possible situations, the materialization of which depends on events that can occur without my intervention, and events that occur only because of my initiative. In professional activity, taking risks or on the contrary choosing simple cases to resolve, or yet learning the various facets of a job and determining which ones to focus on and which ones to delegate therefore amounts to experiencing the contingency and variability of work situations and of organizational routines.

64Work can thus be interpreted in two ways. Jobs can be distinguished through degrees of complexity and responsibility, and differences of remuneration. Stratification mechanisms bring into play levels of competency, extents of control, and means for monopolizing certain fields of expertise. But this interpretation of work is not sufficient.

65A complementary understanding holds that the environment with which an individual interacts can be broken down into varyingly recurrent and predictable relations. And the activity itself can be broken down into a set of tasks and related skills. These tasks and their combinations are drawn on differently, depending on the environment in which they are performed.

66Qualifying these “thick realities” constituted by professions and organizations in probabilistic terms constitutes an efficient heuristic principle. Let us examine the part of the distribution of professions where work variability is at its highest.

67Mastering a task varies most certainly with its complexity, but also with the probability of exercising it in a range of diverse situations. The frequency at which a complex task is practised and the formative potential thereof are determined by the previous rate of success of the individual or the team performing it, and by the quality of the work partners. A cumulative learning and advantage mechanism can be triggered, amplifying expertise and skill disparities.

68All in all, the deployment of individual abilities depends on a variable combination of factors: what skills do the tasks to carry out draw on and develop? Who are my work partners? How does the person for whom the service is intended react? What are competitors doing? Etc. This variability makes it impossible to adopt a simple vision of the distribution of abilities within a professional group. While the personal accumulation of professional experience does matter, the opportunities for accumulating this experience and the profiles of those who do so create differences.

69This process-based and relational approach leads to a characterization of jobs and occupations as sets of tasks, routines, unknowns, solutions to be invented on the spot, and risks and trials of uncertainty to manage individually and collectively. It is grounded in the horizontal dimension of inter-individual relations, with its elements of cooperation, conflict and reciprocal learning, to enrich the analysis of the structure of social relations.

70If I ask what my chances of getting a job are, I invoke the predictive power of training and qualifications, along with the job categorization that they underpin. If I now ask “how does one go about doing one’s work?”, I adopt a complementary perspective which opens the field of comparison.

71Chicago School sociologists have shown what makes work commensurable, beyond its simple calibration based on income, employment status and the functional specialization of an activity. What is therefore common to all the occupations they studied: doormen, scrap merchants, undertakers, taxi drivers, rabbis, primary school teachers, jazz musicians, hospital psychiatrists, osteopaths, executive officers, pharmacists, secretaries, lawyers and plumbers?

72In our economies, based primarily on services and their large interactional component, is it not salutary to ask why the asymmetry of information, knowledge and credibility on which the doctor-patient or lawyer-client relationship is founded is not so different from the relationship characterizing the services of a plumber or religious services? By suspending hierarchies of credibility and prestige attached to different occupations, we are able to access a whole dynamic of the social processes whose analysis completes the vertical approach that studies inequalities in skills, abilities, and earnings.

73This exploration allows us for example to distinguish between all the routine and non-routine elements of work. We can grade the various tasks constituting an activity according to their cognitive, emotional and physical requirements, their gratifying, tiresome, off-putting or difficult nature, and the variety and flexibility of the routines they involve.

74This analytical breakdown of work makes it easier to investigate occupations’ exposure to the effects of technological innovations and to the globalization of labour markets. While this is not a new perspective, it has gained importance on the agenda of sociological and economic research on work. Certain tasks are directly complementary to technological trends. Others, whilst carried out by workers with average or high qualifications, are comprised of a rather inflexible set of routines; they can therefore be automated, and the occupations, of which they were the core, disappear when absorbed by technological tools. Other tasks are subcontracted or relocated, depending on local or international competition around labour costs. Finally, some tasks, at various levels of qualification, are not exposed to technological substitution or price competition: this applies to some personal services, or small-scale activities, which have long interested sociologists, particularly from the perspective of the social critique of fragmented work. The study of the current polarization of jobs and earnings reveals a collapse at the centre of distribution: the notion of middle class, a statistical aggregate, is becoming an area where the qualities of work are subjected to centrifugal forces as they come into contact with cascades of technological and organizational innovation.

75The sociology of professions, as revived by sociologists such as Everett Hughes, Howard Becker, Arthur Stinchcombe and Andrew Abbott, has made another contribution: the analysis of the constant reshaping of occupations through the aggregation and disaggregation of tasks. A profession evolves if it manages either to delegate routine tasks to a less desirable profession or, conversely, if it achieves recognition of the competences required to perform demanding tasks, or yet if it manages to compete with higher professions by harnessing tasks previously monopolized by them. Ultimately, there is no absolute and permanent link of necessity between a profession and the tasks it claims as its own. Take for instance a gynaecologist, a midwife, a nurse and a nursing auxiliary: a cascaded delegation of tasks in caring for patients is organized between them, with a changeable geometry.

76The analytical characterization of work also explains why the learning content of work varies considerably. Work is not only qualified, but qualifying. The argument of variability and risk is articulated here to define a balance. On the one hand, practising a desirable occupation often affords the possibility to choose cases and problems to deal with, so as to maximize the learning potential of work by reducing routine. On the other hand, a sufficient intensity of practice needs to be maintained, and therefore a virtuosity borne out of routine needs to be sustained, so as to reduce risks of error. Routine also has protective value.

77Thus emerges one of the great lessons of the sociology of work, which is all too often forgotten since the whole apparatus of work and worker evaluation has imposed its objectives and sparked controversy. Rather than simply identifying factors of success and performance in work, it is also necessary to understand what we stand to learn from the risk of failure, experimentation and the unknown, which gives rise to uncertain initiatives and rectifiable decisions. This affords insight into how individual work is embedded in a set of collective systems intended to manage errors and performance fluctuations, and in some cases to convert leeway into resources for action rather than weaknesses that are harmful to the organization. In work situations, everyone must be able not only to fulfil their responsibilities but also to save face by taking on some non-transferable risks and by legitimately delegating others.

78Inherent to the world of work being steeped in relations of exchange and reciprocity is the fact that each individual is placed in a cascade of constant interpolations. The discrepancy between my urgency as a client and the routine processing of my request by the professional is that which, in the reverse situation, someone else would experience if I had to answer their request for my competencies and attention. Mutual understanding requires the multiplication of perspectives, without the relations of interdependence doing away with differences of position. What is trivial or insignificant for someone else may be decisive or vital for me. The need for meaning in work cannot be understood outside of these interpolations.

79Thus, there are two complementary understandings of the world of work. The one is vertical, situating individuals in a cumulation of differences and inequalities – of expertise, authority, credibility, social power and remuneration. In its most separative form, this interpretation opposes labour and work [ouvrage]: work with instrumental value and work with expressive value.

80In a relational or horizontal understanding of work, human dignity stems from the interchangeability of perspectives. Knowing how to put oneself in someone else’s shoes is an infraconscious mechanism, essentially steeped in behavioural and cognitive routines, which surfaces only when a situation becomes problematical. This relational substance is what cements relations of reciprocity. Articulating the dimension of reciprocity here brings me closer to Pierre Rosanvallon’s work on equality.

81Are these two vertical and horizontal interpretations antinomic? I believe they are complementary, as they allow us to break work down so as to identify the conditions facilitating coordinated action, without falling into the irenic vision of a world devoid of social gravitation.

82As mentioned above, describing the process of creation and discovery by dividing it into diurnal and nocturnal phases or layers can easily be linked to the analytical breakdown of work that I have just presented. Let us look at these phases. Acquiring, recording and exploiting available knowledge are normal activities and necessary routines of research work. However the more enigmatic activity readily called productive imagination work refers to the share of variability and unpredictability in an unprogrammable process. Finally, the control and exploitation phase where the solutions chosen are stabilized falls within competences whose efficiency increases with the success rate of the imaginative work and with the quality of the work partners who themselves act as screens. The experience acquired by performing this work increases reflexive attention on these phases, without ever removing awareness of the uncertainty of the process.

83The organization of activity in the arts and the sciences affords us insight into how the overabundance of ideas, intuitions, qualities, ambitions and productions offered is a matter of professional demographics, and why the identification of fruitful solutions generates disproportionate differences of success and reputation. But it also teaches us why interaction with others is a crucial resource that fosters alliances and reciprocity, and not only a strategic field of competition.

  • 14 Andrew Abbott, Chaos of Disciplines, Chicago, The University of Chicago Press, 2001.

84My approach to work therefore consists in showing that the same mechanisms are at play at the different levels of analysis: it amounts to a fractal sociology if, as Andrew Abbott did, I wish to depict my endeavour to reduce complexity.14

85Let us examine the relationship between the demographics of professions and the distribution of individual abilities sought therein. Some occupations present a normal or Gaussian distribution of competences and individual performances: competent individuals know how to perform the tasks required. Yet, irrespective of how appreciated it is, an outstanding individual performance will not have a commensurate effect on the organization’s success. It is essentially disastrous performance that can disrupt the mechanics of functional interdependence of activities: for an airline pilot, landing the aeroplane on time or causing a chaotic landing are two performances with highly asymmetrical consequences, both for the passengers and for the airline.

86There is a second category of activities in which the situation is very different: the value of some workers’ performance has a multiplicative effect on the overall result or, in other words, their contribution to the success of the project or organization is more than proportionate to the difference between their abilities and those of their work partners.

87Artistic creation and scientific research activities fall within this second category. Their professional demographics is essentially imbalanced, for the abilities which give an individual distinctive value cannot be identified in simple and certain terms, beyond his or her particular set of skills. These occupations’ appeal stems from their less routine content and their capacity to singularize individual contributions to the overall result.

88But what do we know about the individual abilities so sought after in these jobs? If these abilities were easy to define and to observe, there would be no uncertainty regarding success. They would have been directly identified and developed in the initial training, and the evaluation of outputs would follow absolute and univocal criteria. How can these qualities be identified so as to develop them? This is done through relative comparisons that cause reputations to emerge. What is called “talent” can be defined as the degree of quality attributed to an individual through these comparisons without absolute external markers. The difficulty of defining talent comes from the fact that it is a purely differential quality.

89A host of individual, cognitive, emotional, behavioural and social resources can of course be identified, the implementation of which is correlated to chances of success. What matters however is essentially the combination of these resources. Yet the formula to dose them in optimal combinations is too variable to be measured with certainty.

  • 15 Robert Merton, “The Matthew Effect in Science. The reward and communication systems of science are (...)

90Initially, individuals have poor knowledge of their abilities and must gradually learn to process the information supplied to them through comparisons of their work with that of others. What exactly are these evaluation procedures worth? They are all the more imperfect given that the promising dosage of individual characteristics is difficult to define and to gauge. Robert Merton assumed that chance could very well be responsible for considerable career differences.15 But the succession of work sequences reduces the importance of unknown factors and can act as a fair principle: a career is more than an instantaneous success.

91As a result, in situations of uncertainty regarding the value of individuals, of their outputs and of their potential, it is rational to resort to an excess supply – an excess of work and knowledge produced – and to bring into play professionalization mechanisms based on multiple projects. Some individuals rapidly stand out, though perhaps not lastingly, while others persevere, and yet others are distributed across the variety of related occupations or reorient themselves when they have taken stock of their mismatch.

92Have I returned to the destructive competition and social zero-sum game argument where a few take all? Actually, creative and inventive work acts more subtly. Novelty is contingent on the limitless variety of projects; it feeds on the density of interactions within professional communities.

93This sheds light on reciprocity as a common foundation crucial to the growth of inventiveness: the constant articulation of ties of collaboration, mentoring, and gift reciprocation. We must also remember that individuals’ behavioural plasticity activates a variety of roles in changing environments: without it, the variability of work experiences would be too limited, or impossible to control. Role diversity balances risk and security in one’s activity, as I have shown in my research on artistic careers.

94In the sciences, every researcher is a matrix of links within and outside their organization; every team enters into relations of collaboration of varying intensity and recurrence with a network of partners. An original science of scientific work in teams and in networks is developing. It shows the extent to which interactions, exchanges and collaboration are subtle combinations of cooperation, competition and transmission, and how reciprocity is embedded in this network of balances.

95It also demonstrates that these relationships of exchange and collaboration are structured by a mechanism of assortative matching or preferential attachment: individuals or teams co-opt one another according to what their identifiable abilities have to offer to collaboration. This mechanism amplifies the chances of accumulating new skills and knowledge, when collaborations are successful. The chances of inventiveness increase or decline, depending on the potential of this environment. This is a social ecology of work, with its own density and plasticity.

96The assortative matching argument does not ipso facto lead to a rigid stratification of professional worlds. It can be optimal for an individual to get involved in a project with less experienced or less specialized partners, not only because the project seems more novel, more risky and more formative than predictable collaborations, but also because a team’ diversity provides more than its homogeneity does.

97Without this dynamic, one could make sense neither of the fact that careers are coupled with innovation, nor of how diversity can permeate a stratified system. The ecology of inventiveness is not one of oligopolistic concentration. But without analysing the multiplicative nature of the production function in creative activities, it would not be possible to understand co-opting mechanisms. Innovation is – and must be able to remain – capillary, whether in the arts, the sciences or entrepreneurship. Yet the ranking industry overlooks this.

98Finally, for individuals, professional communities and organizations to be able to engage in the risky game of creative invention, they must have effective means of managing job and career uncertainties. Without these means, there would be no collective benefits to an economy of differentiation and risk. Redistribution mechanisms, the production of statutory guarantees and collective protection, public funding and the search for tolerable balances between control and autonomy within organizations are all common means of reinforcing solidarity and interdependence at the very heart of the competitive force fields of the world of work.

99Yet the analytical breakdown of work reveals less visible means of support. Creative and inventive work would not be socially sustainable if the occupations were not structured in such a way as to absorb part of the risks of failure. A researcher’s or a creator’s activity itself contains a diversification formula that allows for the coexistence of the two profiles I mentioned above: functions with additive value and functions with multiplicative value.

100This is what the association of teaching and research in academic jobs represents. Whereas teaching constitutes the foundation of unquestionable social utility, research enjoys higher social value but less certainty of success. These two roles are more firmly linked if the complementarity between the quality of the teaching staff and that of its student audience, particularly at PhD level, operates. Ultimately, the idea is to embed the share of risk in the activity whose outcome is secure. Similar multi-activity formulae are frequent in artistic careers.

101Teaching means exploiting, recombining and transmitting acquired knowledge. Researching and inventing means exploring through experimentation and variation. In terms of occupations and work teams, this presents a possible transposition of the separation between the various phases and tasks of creative work. In the lectures that I will devote to transformations of the coupling between teaching and research, I will study this duality in the acts of work. The asymmetrical value attributed to them has increased role tensions and conflicts, since the generalization and globalization of competition between organizations designed to improve their research performances, to attract the best scientists and the most promising students, and to obtain competitive project-based funding.


102During this lecture, I mentioned the names of colleagues from various disciplines who teach within this institution or did so in the past. Reading their work is part of this scientific companionship that stimulates productive imagination. I will pursue this dialogue in my teaching and engage in it with the colleagues who, here and elsewhere, are curious to know more about academic and creative work.

  • 16 Harriet Zuckerman, Scientific Elite: Nobel Laureates in the United States, New York, Free Press, 19 (...)

103For those who are familiar with the recruitment and co-opting mechanisms in professions with uncertain success, entering an institution as prestigious as the Collège de France can be experienced both as a challenge and as an honour. In a famous apologue, Harriet Zuckerman imagined what the creation of a 41st seat at the Académie française would look like.16 He or she who would come to hold it would certainly have abilities comparable to those of the colleagues who would have co-opted him or her. But the latter are and remain 40 in number. Through this counterfactual reasoning, we can understand that the social scarcity of positions creates statutory disparities, whereas the underlying distribution of individual abilities most likely has a far less separating effect. This is something to remember when doing one’s job here, so as to maintain the appropriate humility and the sense of challenge.


1 Jon Elster, Making Sense of Karl Marx, Cambridge, Cambridge University Press, 1985.

2 Claudia Goldin and Cecilia Rouse, “Orchestrating Impartiality: the impact of ‘blind’ auditions on female musicians”, American Economic Review, vol. 90, no. 4, 2000, pp. 715-741, DOI: 10.1257/aer.90.4.715.

3 Alvin Gouldner, The Coming Crisis of Western Sociology, New York/London, Basic Books, 1970.

4 Raymonde Moulin, Jean-Claude Passeron et al., Les Artistes. Essai de morphologie sociale, Paris, La Documentation française, 1985.

5 Christian Baudelot, Michel Gollac et al., Travailler pour être heureux? Le bonheur au travail en France, Paris, Fayard, 2003.

6 Pierre Bourdieu, Science of Science and Reflexivity, Cambridge, Polity Press, 2004.

7 Ian Hacking, The Social Construction of What?, Cambridge (Mass.), Harvard University Press, 1999.

8 Jean-Paul Sartre, Critique of Dialectical Reason, vol. 1: Theory of Practical Ensembles, London, Verso, 2004, pp. 127-128.

9 Donald Campbell, “Blind Variation and Selective Retention in Creative Thought as in Other Knowledge Processes”, Psychological Review, vol. 67, no. 6, 1960, pp. 380-400, DOI: 10.1037/h0040373.

10 Paul Valéry, “Variété”, Œuvres, Paris, Gallimard, coll. “Bibliothèque de la Pléiade”, 1957, vol. 1, p. 641.

11 Roland Barthes, Writing Degree Zero, London, Jonathan Cape, 1967.

12 Gilles-Gaston Granger, Essai d’une philosophie du style, Paris, Armand Colin, 1968.

13 John Levi Martin, Social Structures, Princeton, Princeton University Press, 2009.

14 Andrew Abbott, Chaos of Disciplines, Chicago, The University of Chicago Press, 2001.

15 Robert Merton, “The Matthew Effect in Science. The reward and communication systems of science are considered”, Science, vol. 159, no. 3810, 1968, pp. 56-63, DOI: 10.1126/science.159.3810.56.

16 Harriet Zuckerman, Scientific Elite: Nobel Laureates in the United States, New York, Free Press, 1977.

Le texte et les autres éléments (illustrations, fichiers annexes importés) sont sous Licence OpenEdition Books, sauf mention contraire.


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search