1The contemporary revival of economic sociology originates in the growing dissatisfaction of numerous researchers with social sciences that are becoming more and more specialized and thereby more and more ignorant of one another: “Contrary to the movement that leads the social sciences to divide intellectual labor and specialize research, economic sociology suggests that it is necessary and useful to connect economic and sociological theories so as to provide better explanations for economic facts than either of these specialties does when they operate in an isolated or, worse, a contradictory manner.” [1] Despite the balanced wording here, it is truly economic theory’s growing tendency toward independence that is first and foremost criticized by contemporary economic sociology. Therefore, although Mark Granovetter is careful to distance himself simultaneously from the oversocialized approaches of sociology and the undersocialized approaches of economics, when he arrives at the definition of his research project, the focus of his concerns is the confrontation with the new economic imperialism, which is oblivious to social structures. For him, what is at issue is combating the “the extreme version of methodological individualism that dominates much of modern economics.” [2] This essential role played by the critique of standard economics in the definition of the “new economic sociology” (NES) is clearly seen in the following quotation, where the three central propositions defining this approach are clearly stated: “My critique (of economic imperialism) builds on three assumptions, each deriving from the classical sociological tradition : 1) the pursuit of economic goals is normally accompanied by that of such non-economic ones as sociability, approval, status and power; 2) economic action (like all action) is socially situated, and cannot be explained by individual motives alone; it is embedded in ongoing networks of personal relations rather than carried out by atomized actors; 3) economic institutions (like all institutions) do not arise automatically in some form made inevitable by external circumstances, but are “socially constructed.” [3] Opposition to economic theory appears even more strongly in the work of Richard Swedberg in “The Paradigms of Neoclassical Theory and Economic Sociology: A Comparison.” [4] Richard Swedberg lists no fewer than seven weaknesses, touching on all the central areas of socioeconomic analysis: concept of actor, arena of action, types of economic action, result of economic action, view of the analyst, concept of time, and scientific method.

2This critical stance led the NES to present itself as a competitor to economic theory on economic theory’s own territory, that is, the understanding of basic economic facts. As Steiner emphasizes, the NES “has for its purview the various domains of the market mechanism on which economic theory pronounces.” [5] This aggressive stance breaks from the earlier situation, that of the period spanning the 1930s to the 1960s, the “Pax Parsonia,” when there was a clear division between the “spheres of influence” of sociology and economics. [6] Herein lies an original aspect of the contemporary circumstances, which is sharply highlighted by Granovetter: [7] although the old economic sociology, because it focused on the institutional conditions of economic life, did not need to propose an alternative description of everyday economic activity, the NES inverts economic imperialism and proposes sociological analyses of fundamental economic topics, such as markets, contracts, currency, trade, and banking. As it does this, one can detect a situation like that of the beginnings of economic sociology. [8]

3With one major exception, however: the question of the unity of the social sciences under the authority of a general sociology, a unity defended in various forms by Durkheim, Pareto, Schumpeter, Veblen, and Weber, has completely disappeared from the contemporary theoretical agenda. [9] As far as I know, Granovetter never raises it. He does not call into question the legitimacy of the disciplinary division between economics and sociology. In addition, he perceives economic sociology as purely coming under the sociological approach. [10] This approach lacks the rigor of the aforementioned “founding fathers,” who, asserting that economic facts are social facts, were logically led to conceptualize the inclusionary nature of “sociology,” whatever the suggested definition and the principles by which this inclusion was realized. In Granovetter’s work, this results in an ambiguous status for economics, the existence of which as an independent discipline seems accepted, albeit on bases that are not entirely specified. This vagueness appears in the following quotation, one of the rare passages, as far as I know, where Granovetter mentions this topic: “… a sophisticated economic sociology will neither throw the valuable corpus of economic reasoning out the window, nor be so seduced by it as to produce a “rational choice” argument that loses touch with the classical sociological tradition; rather, it will seek to understand how modern economics can be integrated with a social constructionist account of economic institutions, and what the division of labor must therefore be between sociology and economics.” [11] While the criticisms levied against the orthodox economic paradigm are precise and radical, the degree of “validity” that this paradigm may bring to the understanding of social facts remains undefined. The hypothesis will be made that this difference in perspective between the contemporary period and that of the “founding fathers” originates in the ever more extensive and triumphant development of individualistic economics, the perceptions and tools of which are progressively imposed on the most varied domains, including those that seemed to be reserved for sociologists alone. As Richard Swedberg writes, “It is clear that to many observers the most important social science discipline in our time is economics, more precisely the kind of economics that is usually referred to as mainstream economics and which has its most powerful stronghold in the United States and its economic profession.” [12]

4However, in spite of this hardly favorable situation, this article will defend the idea that the question of the unity of the social sciences must not be evaded, nor can it be – and this is the point of view even of those who recognize themselves in the undertaking that the New Economic Sociology is pursuing. This idea is developed in the first section of this paper. There we emphasize the inconsistency of the current NES, which develops a radical critique of the individualistic approach in economics on the plane of theoretical principles, but refuses to draw from it the single obvious conclusion – that is, the rejection of this approach and the construction of an alternative approach on new foundations. This inconsistency accompanies the refusal to confront the question of the unity of the social sciences, instead favoring an opportunistic practice that is content with the status quo between economics and sociology, even if it implies an unprincipled coexistence of the disciplines. Understanding why this is the case requires arriving at what constitutes the heart of the problem, that is, the fact that the NES is incapable of understanding individualistic economic thought other than in a negative way because it perceives it as analyzing a virtual world, maintaining only remote connections to the real world. While this perception prevails, it is not possible to lay the foundations for a rigorous dialogue between economics and sociology. At a minimum, this dialogue requires that in terms of principles, the sociological relevance of economic approaches is recognized, that is, that they cease to be perceived as essentially mathematical games that deal with an imaginary universe, in order to appear as describing a social world that is distinct but in accordance with the overall laws of human socialization. This is the subject of this article’s second section. The difficulty is to contradict traditional readings that believe that the specificity of the economic approach consists in the extreme hypotheses of rationality that are retained in order to analyze individual behaviors. These readings propel the NES because they give the economic world the appearance of a universe that is purely fictitious because it is comprised exclusively of experts in mathematics and logic. It is worth noting that this same vision is also to be found in the work of numerous sociologists and economists. In contrast to this perception, this article maintains that what makes the economic approach distinct is to be sought among the institutional hypotheses that serve to characterize the actors’ environment – that is, an extreme objectification/naturalization of the world surrounding them. Here, rationality is not the result of ad hoc, unusual cognitive properties, but the logical consequence of hypotheses made about the extreme objectification/naturalization of the social world. This analytical perspective leads to a consideration of an economic embeddedness that includes not only networks of personal relationships but also the mediation of objects. For those who take this perspective, the world of economic models ceases to be a “sociological UFO.” It is then possible to imagine it as an approximation of the real social world; this introduces the possibility of a comparison with the social logics that are classically analyzed by sociologists. To illustrate this thesis, the third and final section considers the case of the financial markets. In this case, the naturalization carried out by the Walrasian economic approach bears on the relation to the future. The future is regarded as being produced exogenously and knowable by all the actors. The entire conceptual effort of this article aims to show that this situation is only a particular case of a more widespread logic of opinion, a particular case that obtains when all the actors manage to agree on a single representation of the economic future. This perception makes it possible to construct a space that is common to economists and sociologists, which is the very meaning of the unidisciplinary perspective that this article defends.

5Before getting to the heart of the matter, I will mention that this text will make ample use of “conventionalist” works. This should not be taken as the expression of a sectarian will seeking to make one school of thought prevail over another at any cost. I believe that the idea of convention as it is used below easily finds its place in all the currents that are recognized in the NES undertaking, because it is simply a matter of asserting that commodities are social constructs. Therefore, when we speak of a “convention of quality,” we wish to say that goods are not natural facts but the result of a social process. Likewise, “financial convention” runs counter to the idea that that there would exist an objective representation of the future that could not fail to be vital to the agents’ rationality. On the contrary, I believe that the basic scenario that market prices express is a shared belief produced by the game of the market itself. In this sense, it is a convention. I also stress that although this text deals with the question of the unity of the social sciences, it limits itself to contemporary debate and is not placed in the tradition of controversies applying to the chances of one general science of society as can be traced back at least to J. S. Mill’s System of Logic.

1 – The Impasses of the New Economic Sociology

6In a recent work of synthesis explicitly titled “Values, Coordination, and Rationality: The Economy of Conventions or the Time of Reunification in the Economic, Social, and Political Sciences,” the historical conventionalists collectively recall their attachment to the now-abandoned idea of the unity of the social sciences: “The Economics of Conventions aims for an integration that concerns the economic, social and political sciences equally … Rather than a pluri- or multidisciplinary approach that would simply combine the contributions of different disciplines, our perspective seeks to cross the boundaries between economics and sociology in order to uncover their common foundations, and to re-examine them.” [13] I suggest the term “unidisciplinary” to describe such a perspective. Unidisciplinarity defends the idea that economic facts are social facts and consequently share with social facts a model of intelligibility. This adoption of a unidisciplinary stance has led to spirited criticisms. The most charitable critics have seen in it a naive and utopian thesis lacking in practical significance, because in their eyes, the disciplinary division is imposed de facto as the natural framework of the future of the social sciences. The most ill-intentioned critics have seen in unidisciplinarity the hidden expression of a strategy of domination that aims to elevate the “economics of conventions” to the status of the science of sciences! To me, this understands nothing of the true implication of taking this position, which is of an entirely different nature. It is a matter of facing without a red herring the questions that individualistic economics raises for the NES: What is its status? What should be done with its results? What should be said of its models? At present, the NES’s attitude toward these questions is ambiguous.

7On the one hand, in terms of theoretical principles, with respect to orthodox economics the NES develops a fundamentally critical discourse of which three assertions by Granovetter constitute the backbone. It attacks “the reductionist methodological individualism of modern economics.” [14] This criticism applies essentially to the undersocialized nature of the economic approach: the actors neither behave nor make their decisions in an atomized way, independent of any social context. On the contrary, the actions they undertake to fulfill an objective are embedded in concrete systems and continuous with social relations. Therefore, thinking about individual action in a pertinent manner requires that this action be resituated in what constitutes its relevant context, that of the interpersonal relations through which the social world exerts its influence. This is what makes the sociological approach superior: because it alone is able to envision the embeddedness of economic action, it alone is able to produce a fully relevant empirical analysis, that is, by integrating the multiplicity of determinations that shape historic reality. This perspective justifies the aggressive nature of the NES that, as is stressed above, does not hesitate to attack the study of the most important economic phenomena. For all that, although the NES theoretically condemns, in the firmest way possible, the inadequacy of economic formalism, which “overlooks social structures,” the NES does not fight for what would seem to be the logical consequence of this theoretical denunciation, that is, the refusal to consider individualistic economics as a social science. But then, if the NES does not reject individualistic economics, what status should it be given? What dialogue does the former propose to establish with the latter? How can this dialogue be reconciled with the critiques that are made of it with respect to principles? Because these questions are never addressed head-on, but only indirectly and partially, it is not easy to discern the economic sociologists’ exact position on the subject. Leaning on these fragmentary reflections, it is nevertheless possible to attempt the following interpretation: the NES sees in the undersocialized approach of contemporary economics an essentially logical-abstract argument that resorts to mathematics in order to analyze a hypothetical world in which there interact atomized rational individuals who are assumed to have as their sole motivation the maximization of their economic profit. Seen from this perspective, the NES takes up, on its own account, Simiand’s position: “Economics defines possible worlds, that is, logically possible theoretical results.” [15] However, although Steiner, on the basis of this observation, logically concluded that theoretical economics were inadequate and were to be expelled from the social sciences, the NES, as has already been underscored, does not go so far. As Steiner says, the NES takes up Simiand’s position, “but strips it of its critical burden.” [16] What sense can be made of such an apparently contradictory position? If the models of economic theory cannot be interpreted as expressing reality, what use can the economic sociologist make of it? The entire question lies there, and it is clear that the NES does not offer a clear answer. It does not say clearly what this theory, so vigorously criticized for its undersocialized approach, disconnected from reality, can contribute to the knowledge of socioeconomic facts. For this contribution to be realized, conversely, it would need to specify on what common bases a rigorous dialogue may be established between the NES and economic theory. Either economic theory is only a mathematical game and its “seductions must be refused,” or its results are to be taken into consideration, but the reasons why this is to be done and the extent of this consideration must then be justified.

8To illustrate this contradiction, I will take as an example what Granovetter says about his research on the electric power industry in the United States. He seeks to determine why, in the United States, electricity production is handled by private-sector companies and not by a vast public enterprise, like EDF, or by private production that is decentralized to each household and company. The general line of argumentation may be presented thus: the electric power industry in the United States presents a type of configuration in which the technological and economic constraints do not automatically lead to a single organizational form; numerous types of organization are possible. It remains for individual or collective action, channeled through existing personal networks, to determine which option will be chosen. Therefore, even if the economic and political conditions are identical, the final form may vary entirely if the structure of the social networks varies. [17] These reflections reveal a very particular form of articulation between economic and sociological analyses. First, economic analysis determines the possible equilibria given the existing economic and technological constraints. In the case under study, this analysis concludes with the existence of three possible systems of equilibrium: public property, private and decentralized production, or a system of private-sector companies. But the economic analysis cannot go any further. It is not capable of determining which solution will in fact be chosen. Enter the sociologist, who comes to the aid of the insufficient economist. As Granovetter stresses: “This resembles situations in economic dynamics that are characterized by multiple stable equilibrium points. Indeed, I believe that a social constructionist account can help make such dynamic economic models of institutions more sophisticated. These models are frustrating because there is little substantive way to resolve their underdetermination. As in physical cases with multiple equilibria, you can understand which state the system has reached only by looking at its history. But the contingencies involved in this history are typically outside the economic framework, and thus seem ad hoc and unsatisfying to economists; within a sociological framework, however, they can be given a systematic treatment.” [18] This passage is very clear. The respective merits of economics and sociology are distinctly expressed in it. According to Granovetter, one of the essential qualities of sociological analysis is to allow for a historical contextualization of economic models. This contextualization makes it possible to depart from the indeterminacy of the economic model that is content to determine the list of possible equilibria without being able to determine which one will prevail. The principle of this contextualization is found in the sociological clarification of social networks. To specify the network’s form is to provide the means to understand how individual and collective action will be channeled and will lead to the carrying out of one option over another. This method of articulation between sociology and economics is interesting and there is much to say about it. This article will merely emphasize that the method rests implicitly on a perception of what economic theory can imagine, a perception that is completely different from the one supported in terms of the principles of this very NES when it stigmatizes the undersocialization of the economic approach. The contradiction is evident. Obviously, the analysis offered by Granovetter in this case does not consider that economic models are abstract games that are unrelated to reality. The criticism aimed at them is of a different character, that is to say, their inability to make the entirety of this reality intelligible. In the method of alliance between economics and sociology that Granovetter outlines here, the economic approach to the domain of electric power, an approach that is nevertheless undersocialized, founded on the competition between atomized and completely rational homo economicus, who are moved only by the quest for maximum profit, is seen as providing a description that is limited yet relevant to historic reality. However, he does not say why it is like this or why the approach’s undersocialization is not an obstacle.

9My intention in this article is neither to take the side of this specific model of alliance between economics and sociology, nor even to state that this same model of alliance is found in all of Granovetter’s work. Rather, the point is in this specific case to highlight the NES’s lack of theoretical consistency regarding the status of individualistic economics, a question that is central from the perspective of a renewed dialogue between the social sciences: although, theoretically, the issue sees there only an undersocialized discourse that is unrelated to reality, it “practically” insists that one should not “throw the valuable corpus of economic reasoning out the window” (infra), but without specifying what this “value” is like. To me, the roots of this situation can be found in the NES’s refusal to address the question of unidisciplinarity. From the sociologists’ side, this is because the NES remains vague regarding theoretical economics and because it authorizes both the most sectarian and the most opportunistic approaches. From the economists’ side, this is because the NES does not offer any common ground or collaborative work. Consequently, one sees a parallel development of the two disciplines, with each one asserting its universal mission to the detriment of the other, but without true dialogue and without either feeling obliged to take the other’s findings into account. This then leaves the way open to pure relations of force between the disciplines, which today lead to an underlying domination of economic ideas.

10In essence, my entire argument states that it is in the very nature of economic sociology to fall within a unidisciplinary perspective. This appears when one considers the birth of economic sociology (1890-1920), a period during which unidisciplinarity was strongly championed. [19] This also appears today, but, on the contrary, in the contradictions reached due to the NES’s desire to remove this question from its agenda in order to accept the disciplinary status quo and to view economic sociology as a branch of sociology. In the first place, this prudent position may seem realistic because it conforms to the current balance of power in the social sciences. Can sociology, fragmented into multiple schools of thought, demand the unity of the social sciences in its own name in the face of an ever more triumphant economics that invades the most varied domains, such as reciprocity, language, religion, social interactions, crime, or family? What credibility would it have? However, this argument has its limits. As has been emphasized, for want of a declared opinion on the relations that should be established between the economic and sociological approaches, the NES shows itself incapable of providing economists with a true possibility of collaborative work or even dialogue, save on an ad hoc basis, according to persons and situations. What should be feared in such a context is that these local alliances would take the path of least resistance, thus leading the NES to silently and progressively enter the orbit of individualistic economics. In this way, the unidisciplinarity that the NES refuses to imagine would ultimately be implemented by an individualistic economics that does not share its scruples. As Gautié writes, “Becker and many post-Beckerians actually often render homage to sociology for its ‘big and broad questions.’ But this homage is more than ambiguous: it actually remains deeply anchored in the traditional vision according to which sociology, an ancillary discipline, is at best a purveyor of intuitions and/or empirical observations, which only economics is able to formalize in a consistent theoretical framework, a vision that moreover seems internalized by some sociologists.” [20]

2 – The Economics of Conventions and Object-Oriented Embeddedness

11As has been seen, the “economics of conventions” (EC) is distinct from the various approaches that make up the NES, owing to the forcefulness of its unidisciplinary commitment. The incongruity of this position in the field of NES can be compared to another singularity of EC in this same field, the fact that economists and sociologists are represented in equal numbers, to the point that it is not easy to characterize their disciplinary affiliations. In these conditions, it is understood that the question of a common foundation for thought that spans economics and sociology holds a place of strategic importance here because it conditions the very existence of this research. Laurent Thévenot formulated its guiding idea. [21] I would like to show that this provides to economic sociology what, to my mind, it currently most painfully lacks: an aggressive vision of the integration of the social sciences capable of providing economists with the bases for, if not collaborative work, then at least a true dialogue.

12This idea is perhaps most clearly elucidated in Thévenot’s “Rationalité ou normes sociales : une opposition dépassée?” (“Rationality or Social Norms: An Obsolete Opposition?”). As is indicated by the title of this article, Thévenot’s theoretical effort is entirely directed toward the construction of a common framework making it possible to jointly grasp, and thus to reconcile, the sociological approach via norms and the economic approach via rationality. This shared perspective is that of coordination: “The figures of action developed around [the ideas of ‘social norm’ and ‘rationality’] appear to conflict with one another from the outset, with the one tracing the contours of collective or social behaviors and the other defining an individual decision. In order to encompass them both in a single perspective and to find an agenda common to all the social sciences (including economics), each of these two ideas must be resituated on the same plane: the study of the integration of individual acts into an order, an equilibrium, a coordination – the terminology to vary according to the framework for this integration. The two concepts from which we began thus find a place within two sharply contrasting frameworks: that of an order governing social behaviors, and that of an equilibrium resulting from rational choices.” [22] In the second place, by utilizing the model of justifiable action, Thévenot shows that the norm-order and rationality-equilibrium pairs can be analyzed as different specifications of a single system of adjustment of actions characterized by the search for a shared judgment and thus presupposing a shared form of evaluation. [23] The concepts of norm and rationality correspond to two ways of envisioning the manner in which the actor grasps the situation. These concepts must be situated in a wider range of methods of judgment. [24] The strength of this approach is to give competitive coordination the robustness of sociology. In this perspective, Walrasian logic ceases to be a “sociological UFO” produced by an abstract imagination disconnected from the real world, instead appearing as a specific case of a more general social logic, that of justifiable action. I will now examine in detail the originality of this unidisciplinary alliance that conventionalist thought offers to economists and sociologists.

13As is stated above, this approach differs fundamentally from that followed by the NES in that EC does not judge the Walrasian model as a purely logical-abstract construct devoid of connections with reality. On the contrary, this approach recognizes in it a specific model of coordinated actions that it can name precisely, that of justifiable action, which is seen to be involved in other forms of judgment that all lean on the definition of a common good surpassing people’s particularities and enabling their agreement. [25] In so doing, EC escapes the impasses to which the NES is led due to its principled rejection of orthodox economics. By its very nature, the radicality of this critical position makes it impossible to establish any theoretical dialogue between the two disciplines. It permits only compromises that are cobbled together on an ad hoc basis. EC’s perspective on Walrasian economics is entirely different. To it, the Walrasian world is not at all hypothetical. It corresponds to a structure of interactions that the sociologist recognizes and can study. This difference in evaluation is based on a divergent analysis with regard to the question of embeddedness, which is central to economic sociology.

14I believe that the NES tends too much to identify social bonds with personal bonds. From this follows a predisposition to think of discontinuous and impersonal transactions as being, at worst, “outside of the social” and, consequently, great difficulty perceiving the reality of market atomization. [26] However, the separation of individuals in the marketplace is not just a Walrasian fantasy. If the commercial universe can be satisfied with weak, discontinuous, and fragmentary interpersonal connections, this is because of a social structure that privileges relations to objects. In addition, herein lies an utterly fundamental fact for anyone seeking to understand market-oriented individualism, as Louis Dumont stresses: “In the individualistic ideological configuration, the relation of humanity to things (to nature, to objects) is valorized in contrast to relations among men.” [27] This role of objects is a dimension strongly emphasized by “economies of scale,” for example, when they state that market coordination rests on two pillars: an identification of commercial goods, the exchange of which defines the course of action, and a shared evaluation of these objects via prices that enable the adjustment of various actions. [28] These objects, the importance of which is highlighted by the conventionalist perspective, are socially constructed objects, having previously been formed in such a way as to allow them to be engaged in operations of justification and evaluation, according to the specificities of the scale under consideration. “Our approach has consisted in conceiving the question…of coordination in connection with the notion of the object. The burden carried by the understandings of individuals in their speculations, judgments, and justifications can thus be eased and partly transferred to the objects. This theoretical approach assumes that we can recognize several forms of objectivity connected to several forms of coordination of actions. We can easily recognize in market coordination the role of commodities that make actions comparable by orienting them toward the appropriation of goods that are detached from persons.” [29] Here, the social operation of aligning certain objects to certain scales is envisioned through the concept of “investment of form.” As François Eymard-Duvernay stresses, the investment of form emphasizes the operations that must be undertaken for goods to acquire generality (objectivity) by establishing relations of equivalence. [30] From this standpoint we can see a social embeddedness of individual action as resting less on a network of interpersonal relations than on a network of objects previously shaped in such a way as to be adapted to operations of market evaluation. To simplify as brutally as possible what separates me from the NES, then, the latter undervalues the role of objects in its definition of social embedding in order to focus solely on interpersonal relations, making it difficult to understand the atomization and impersonality of market relations clearly. Note that this assessment does not mean that we have to reject the economic sociology of networks. Certainly this assessment offers an indispensable clarification to a number of important questions. What is more important here is that the idea of embeddedness ought to be extended to the consideration of “objects.”

15This emphasis on the question of objectification leads to a new understanding of what makes individualistic economics distinct, not only in relation to economic sociologists, as is stressed above, but also in relation to economists. This is not surprising inasmuch as one may notice a strong convergence of perspectives in the way that each of these two groups understands economic theory, even if they diverge when it comes to evaluating the capacity of this discourse to understand historical reality. In particular, neither of the two groups seems to perceive the crucial role played by the hypotheses of naturalization and objectification in economics’ modeling of market relations. Therefore, when they think about the increasing independence of their discipline within the social sciences, economists usually see this as the effect of the two traits taken to be specific to their discipline, that is, mathematization and the hypothesis of rationality, as being specific to their discipline. [31] To me, this analysis remains superficial. Neither the use of mathematics nor the emphasis on individual rationality is a trait that is truly specific really peculiar to economics. One may simply note that the use of mathematics and the hypothesis of rationality is more widespread and systematic in economics than in most other disciplines. However, the reason for this must be sought in the manner in which economists formalize the environment in which they place their actors. Consider, for example, the model of the consumer. The use of maximization may seem troubling there when compared to what we observe on a daily basis in the area of consumption. One would thus be tempted to see it as overestimating the individual’s cognitive abilities. But this is not at all the case. In judging this behavior, it must not be forgotten that the consumer, as envisioned by the economist, is situated in a highly distinctive world, (1) where all goods are assumed to be of an absolutely known consistent quality, (2) where the list of all goods is made public, and (3) where unit prices are represented in such a way that the consumer may instantaneously compare the prices, qualities, and uses of all the baskets of possible goods. Focusing on the consumer’s ability to determine the optimal basket in order to see in it the effect of an improbable mastery of mathematics is to understand nothing of the situation. The question of actual calculation is entirely secondary here. The reader who is concerned with realism needs only to imagine an automatic computer device connected to an online purchasing program or placed on a shopping cart. What is essential in this model is the fact that the commercial sphere is imagined by the economist as a completely transparent space as a result of an extreme objectification of the relation to commercial goods. If, in such a universe, rationality is reduced to calculation, this is not tied to the hypothesis of an expanded rationality, but to the fact that the world was previously constructed so that it would be this way. More generally, if economic actors do not cease to pursue maximization, this is because the social world in which they have been immersed is such that no ambiguity or indeterminacy regarding either qualities or quantities remains there. In other words, the systematic use of mathematics and rationality finds its basis in the frenzied objectification that orthodox economics imposes on the social world until all its elements take on a form suited to operations of commercial calculation.

16The strength of this analysis appears clearly if one considers what remains the central theoretical reference for neoclassical economics: the Walrasian general equilibrium. The typical critique consists of seeing in it a completely desocialized world that brings atomized, rational, calculating individuals face-to-face with one another. Thus the NES emphasizes the radically undersocialized dimension of this world in which individuals react only to price signals, without ever coming into contact with one another, nor even meeting or becoming acquainted with one another. In this respect, Granovetter quotes Hirschman, who writes that the markets, gathering “large numbers of price-taking anonymous buyers and sellers supplied with perfect information … function without any prolonged human or social contact between the parties. Under perfect competition there is no room for bargaining, negotiation, remonstration or mutual adjustment and the various operators that contract together need not enter into recurrent or continuing relationships as a result of which they would get to know each other well.” [32] The reading of perfect competition offered here is entirely different. It highlights the fact that the economic universe as the general equilibrium formalizes it is, on the one hand, a transparent world in the sense that all the facts that an actor needs in order to make decisions are unambiguously defined and readily available, and on the other hand, a shared world in the sense that all the actors share an interpretation of this environment. In such a framework, there are three central hypotheses that formalize the relation to goods, the future, and price: the hypothesis of the nomenclature of goods, the hypothesis of the nomenclature of states of the world, and the auctioneer. [33] Let us turn to the least well-known hypothesis, that of the nomenclature of the states of the world. It posits that the uncertainty of the future may be represented as a list of m exogenous events or states of the world that is supposed to exhaustively describe all the events that could occur in the future. In addition, this list is assumed to be common knowledge for all economic actors. The technical hypothesis of common knowledge is the way in which the objectivity of this collective representation is modeled so that no ambiguity exists about the fact that this description is universally known. Walrasian formalism completely ignores the question of how economics would acquire such a universally accepted representation of the future. More precisely, this representation is imagined as being “naturally” established due to the fact that it would be the exact description of the various occurrences that could take place in the future. From this perspective, the treatment of the relation to the future is exactly like that of quality. We find here the same process of objectifying naturalization, leading to the hypotheses of common knowledge and consequently constructing a transparent world that is immediately intelligible to the actors.

17Consequently, in the context of general equilibrium, individuals seem no longer to need to concern themselves with one another. In order to figure out what to do, an individual does not need to know what others are thinking or doing. She needs only a representation of the prices of all goods, a good being defined as a particular quality in a particular state of the world. However, this presentation is deceptive because it overlooks the fact that in order to be this way, all members of society would previously have had to come to a general agreement on the definition of the qualities, the representation of the future, and the mechanism for setting prices. Similarly, when a driver automatically and unthinkingly passes through a green light, it may seem that she is not concerning herself with others’ actions. However, this is a superficial analysis, for if she acts in this way, it is because she assumes that all the other drivers are respecting the rules of the road and will stop at a red light. If everyone were to stop respecting the rules of the road, the driver would slow down at the green light in order to speculate about the behavior of the drivers who encounter the red light. In these two situations, the driver is concerned with others’ actions, but this concern may be expressed in two distinct ways depending on whether or not objectified rules prevail. In the Walrasian general equilibrium, although it may seem that individuals are indifferent to one another, it is because the presentation of this equilibrium does not call to the reader’s attention the importance of the preliminary social labor required for all members of society to come to share a definition of qualities and a representation of the future. It is thus because this shared understanding is presented as natural, uncomplicated, and self-evident, although even the Walrasian accord draws its most vital resources from it. [34] Consequently, the fact that, in short, competitive regulation is based on the postulated existence of common references (goods, states of nature, prices) is not clearly perceived by the economists themselves, as they are used to thinking of them as natural realities that are directly perceptible and do not require any special social construction. Economic space is modeled as a transparent space of immediate understanding that reduces cognition to information and calculation. The idea that economic agents, because of these hypotheses, find themselves de facto immersed in a highly structured shared space of belonging and evaluation is made to disappear, with many economists even going so far as to think that this analysis would provide a spontaneous modeling!

18This critical perspective leads to a research program aiming to deepen our understanding of market coordination by developing a thought that restores the dimension of social construction to all these “objects,” which are mistakenly seen by economics as exogenous and natural facts. Therefore, it is possible to summarize this project with a word like “denaturalization.” This perspective belongs to economics as much as to sociology; only the labels change. In the terms of economic sociology, this program takes its place in the already-rich body of work seeking to explain “the social construction of the markets.” For the economist, it is a question of returning to more tentative hypotheses like those with which the theoreticians of information asymmetries began when they analyzed what happens when the quality of goods is no longer homogeneous and commonly known. However, in substance, this is a shared project, the success of which requires the skills of sociologists and economists alike. By way of conclusion, a succinct illustration regarding the financial evaluation of actions is in order.

3 – Financial Convention: An Example of the Objectification of a Shared Belief

19From an economic perspective, the value of a stock lies in the flow of future dividends to which holding it gives a right. In this regard, the term “intrinsic value” or “fundamental value” is used. Its valuation is not easy because it strictly depends on the ability to predict today what dividends will be paid in the future. As was already shown with the hypothesis of the nomenclature of states of the world, the dominant approach in economics holds the hypothesis that the future is, at an instant t, objectively given in a probabilistic form. Under these conditions, it is possible to define an optimal valuation, one that best uses all the available information and is thus independent of investors’ idiosyncratic opinions. Herein lies a vital premise that forms the idea of informational efficiency. An informationally efficient market is one in which at any moment the price conveys to investors the correct valuation of the fundamental value. This is what Fama says when he writes, “in an efficient market at any point in time the actual price of a security will be a good estimate of its intrinsic value.” [35] Therefore, one may say that the hypothesis of informational efficiency understands finance as an accurate “reflection” of the real economy. [36] Within this theoretical framework, the fundamental value objectively preexists the financial markets, the central role of which is to provide the most accurate and precise valuation of that value.

20In this presentation can easily be seen a new illustration of economists’ penchant for imagining the actors’ environment as a natural fact. However, this analysis stumbles on a simple fact: the extreme precariousness, to say the least, of the objective bases of our knowledge of the future. Keynes, stressing this point, understood that this idea logically leads to a radical reconsideration of the results of classical economics. The uncertainty attached to the future, as he sees it, escapes all scientific knowledge, even in a probabilistic form. As he writes, “About these matters there is no scientific basis on which to form any calculable probability whatever. We simply do not know.” [37] The future must be conceived as indeterminate. It strictly depends on individual decisions that will be made now and later, including financial decisions. Therefore, the hypothesis of the ex ante objectivity of fundamental value must be abandoned. It does not preexist financial transactions.

21A simple way to model this idea is to recognize that valuations are irremediably subjective. They correspond to the subjective bets made by investors according to their own ideas of what the future will be. The hypothesis of an objective future imposing itself on individual valuations and leading to their necessary convergence toward the right valuation does not fit reality. This case was seen at the time of the Internet bubble, for example. When an economist criticized an investor for his excessive valuations, stressing that they completely contradicted everything that had been observed up to that point, the exuberant investor could only retort that owing to the development of new information technology, the economy had entered a new era with no connection to what had been experienced previously; it was an era governed by new laws. This is an indisputable argument. However, as soon as it is permitted to push away the lessons of the past on the grounds that the world is not static, that new things emerge from it, and that consequently, the future does not look like the past, it becomes possible to dismiss all objections. On the terms of this argument, nearly any valuation can be justified. There follows an intransigent subjectivity of valuation that fits perfectly with what is seen in the markets. It must come down to this. It must be admitted that individual valuation is nothing more than an opinion. It may be more or less well-informed, but it remains an opinion in the sense that there exists no procedure that allows any one of them to be confirmed as superior to any other ex ante. However, economists have the greatest difficulty accepting such a point of view. They see it as the end of any serious economic theory. The concept of financial convention exists in order to demonstrate that it is nothing more. Understanding it nevertheless imposes on economic discourse a radical broadening of its concepts so as to imagine economics in a way other than the formation of natural constraints. Economics must evolve toward an intelligibility of economic relations that accords full importance to opinions, to collective representations, and to the social forces that can shape them – what may be called a “cognitive turn.”

22It is difficult to overstate the importance of this question. The opinion/objective value opposition is at the very heart of how economics has always perceived its distinctive place within the social sciences. This is clearly seen as early as the reactions inspired by Durkheim’s 1908 lecture to the Société d’Économie Politique (Political Economy Association). Asked what was “the place of political economy in the social sciences,” Durkheim replied: “The difficulty… is that the facts with which political economy deals and those which are the object of the other social sciences seem at first sight to be very different in nature. Ethics and law, which are the subject matter of determinate social sciences, are essentially questions of opinion…. The speaker nevertheless believes that economic facts can be considered from a different viewpoint. To a degree that he does not seek to determine, they too are a matter of opinion.” [38] Such is the basis of the unidisciplinary convictions of Durkheim and his disciples. The response of the economists was immediate and scathing. Edmond Villey professed to have “felt somewhat shocked.” He formulated his objection thus: “[Opinion] does not determine value, which is determined by rigorous laws of nature… it is always the law of supply and demand, completely independent of opinion, which determines prices as it determines all values.” [39] Paul Leroy-Beaulieu went further and concluded: “[Opinion] is doubtless a powerful factor whose influence is to modify certain forms of the economy, but what it will never be able to transform are the great economic laws which are immutable.” [40] Contemporary economists have hardly evolved on these matters. The idea of convention that I propose aims precisely to undo this opposition by thinking of objectivity as a social construction rather than as a fact of nature.

23The case of the stock market provides the perfect opportunity. In the framework of fundamentalist thought, collective or individual representations do not play a role because only the fundamental value counts. However, in the real markets, the logic is entirely different. What is important first and foremost in order to make stock market profits is to predict not the fundamental value, but the opinion of the market itself. What counts for the investor is the evolution in prices as the majority opinion determines it. From this follows a structure of interactions quite distant from the fundamentalist model in that it posits as a norm not an objective reality that is external to the market, i.e., fundamental value, but an endogenous variable, i.e., the opinion of the market. This is described as a “self-referential structure.” [41] Contrary to the fundamentalist model, predictions are not oriented to the real economy, but to the predictions of other participants. This rule of behavior leads to a disjuncture between finance and the real economy inasmuch as what is important on a market is not the real content of a piece of information in relation to the fundamental facts, but rather the way in which collective opinion is supposed to interpret it. There results a singular rationality that is fundamentally mimetic in that it seeks to imitate the market in order to precede it in its evolutions, as erratic as they may be. If I believe that tomorrow the market price will rise, my interest dictates that I should buy shares even if I think that, in relation to the fundamentals, this rise is an aberration. Imitation is a perfectly rational behavior at the individual level even if it leads collectively to very damaging situations for the economy, as is the case when excessive increases appear. This line of thought emphasizes the phenomena of influence and diffusion of beliefs. This mimetic logic of interactions leads to the polarization of individual opinions around a single interpretation that then becomes the legitimate reference, that which is contained in the prices. The notion of convention can be deduced from this. To develop this theory of the intelligibility of finance, we need both the tools sociologists traditionally use when studying the “force of ideas” (sociological convergence) and the tools economists use when studying rational interactions between actors in self-referential situations (mimetic convergence).

24In conclusion, in the traditional approach, the representation of the future is imagined as a natural fact and the market’s goal is simply to list the various events that could occur. In the conventionalist approach, the role of the market is to bring out a shared reference that allows the economy to be projected into the future. The convention of the “New Economy,” which dominated the financial markets in the late 1990s, is a particularly significant example. This convention contains an idea of the future development of the economy that is structured around new technologies and, more specifically, electronic commerce. Its impact was fundamental in that, on the faith of this idea, many individuals thought it would be profitable to invest in these new areas. There followed a very strong shift in investment and, consequently, significant growth. Here can be seen the crucial economic role played by collective beliefs, light years away from fundamentalist notions. Consequently, elucidating the reasons that led to the appearance of this particular convention requires an approach that closely combines sociology and economics. [42] It is simply not possible to stick exclusively to economic factors. It is also useful to invoke cultural factors such as the strong American penchant for belief in technical progress, or socioeconomic factors such as the aging of the Western population, which leads them to make long-term investments. The fact that taking only economic facts into account is insufficient is particularly visible in the case of this other, still current, convention of requiring 15 percent return on equity (ROE). This requirement is not at all economically justified because of the underemployment to which it automatically leads, since few enterprises are so profitable. Therefore, to understand the foundations of this convention, we need to develop a general analysis of contemporary financialized control allowing us to conceptualize the shareholders’ power from an economic as well as from a social and political point of view. Finally, it should be stressed, to respond to one frequent criticism, financial convention does not describe a financial world in which everything is possible. In that sense it is an unfounded criticism. In fact, if ex ante it is not possible to determine objectively what the true value is, ex post it is possible to judge whether or not the conventional predictions turned out to be accurate or not. In order for a convention to endure, it is important for the observed facts to accord with the predicted facts. This has been seen in the case of the Internet bubble, which burst when the financial and technical efficacy of the new-economy companies turned out to be much less auspicious than predicted.

Conclusion

25The currently prevailing situation in the field of social sciences fully demonstrates the failure of the traditional interdisciplinary model. Despite the apparent support it enjoys, even from the most prestigious public research institutions, this model has not been able to block the growing trend toward the autonomization of theoretical economics. The exacerbation of this trend has been so obvious in recent years that it is difficult to overstate its intensity. In my opinion, the most emblematic and troubling manifestation of it can be found in the total loss of shared references. A few decades ago, it was still possible to think that historians, sociologists, and economists shared some fundamental authors, such as Durkheim, Marx, Mauss, and Weber, to name only the oldest and most paradigmatic. However, today the movement toward specialization is leading economists to completely abandon these authors and, more broadly, references from the generalist social sciences. Like those in the so-called hard sciences, professional economists now read only the most recent articles dealing with their specialized areas of research. This tendency is also completely seen and accepted by the crushing majority of economists, who see it as the very proof of the distinctive status of their discipline and a justification of its claim to a Popperian scientism. For those who, like me, do not adhere to this analysis, this autonomization is highly damaging for economic thought, among other things. The goal of unidisciplinarity is precisely to break with this trend by seeking a shared model enabling the establishment of a dialogue between sociologists and economists, i.e., to set the bounds of a rigorous discussion. The “new economic sociology” must be blamed for having abandoned this goal. The most damaging consequence of this abandonment is that the new economic sociology has nothing to offer the economists. This can be seen when the NES, in order to define itself, systematically emphasizes its differences from the economic approach, in the vein of Granovetter’s three assertions or the table of comparisons between “the paradigms of neoclassical theory and economic sociology” suggested by Swedberg, without ever defining what they have in common. [43] Conversely, economic sociology, conceptualized in a unidisciplinary perspective, is no longer understood as the branch of sociology that studies economic facts. Its ambition is much wider. Its purpose is the dialogue between economics and sociology and, consequently, the reassertion of the fact that economics belongs to the field of the historical sciences. [44]

26More specifically, my guiding idea has consisted of emphasizing the vital role played by the question of the naturalization of social contexts in the autonomization of economics, when ordinarily the hypothesis of rationality is emphasized. My use of the notion of conventions is deduced from this. It is a matter of deconstructing the “natural” objects of which economic modeling makes such extensive use (goods, representations of the future, currency, prices) in order to conceptualize their social production. Thus understood, the idea of convention defines a very broad research program of which Boltanski and Thévenot’s “economies of worth” constitute only one possible application. [45] Other analyses would also find a natural place within this overall vision – for example, financial studies that dwell on the explication of calculation techniques or the analysis of symbolic power. [46]