The agency-structure antinomy as social form: Towards a historical critique of sociological reason

Abstract

This article argues that the agency-structure antinomy, far from constituting an ontological feature of social reality or a mere product of conceptual confusion, is a historically specific social form produced by and through capitalist relations of objectification. Through an immanent critique of the post-Parsonian proposals of Pierre Bourdieu and Margaret Archer, it is argued that successive attempts to transcend this antinomy have reproduced it at a new level of abstraction, due to shared neo-Kantian presuppositions. Drawing on Marx’s theory of commodity fetishism as developed by the Neue Marx-Lektüre and WertKritik traditions, the article argues that an adequate sociological account of the antinomy requires understanding it not as the basis of explanation, but as what must itself be theoretically constituted and historically explained.

Introduction

Historically, Sociology has purported to answer the question of “how and why is society possible?” (Turner, 1977). To answer this question, Sociology has traditionally navigated within the ontological background of the agency-structure antinomy. Social phenomena can be related to the actions of given individuals, or they can be related to the social structures which influence the behavior of these agents. Theory can explain how agents act on their daily lives, the values that they hold and what they think about society. Theory can likewise explain how structures influence these actions, values and thoughts. However, traditionally, the field has struggled to explain both within the same framework (Clarke, 1991: 238). This impasse, in part, was responsible for the emergence of the post-Parsonian generation of theorists that shared a common purpose of transcending the traditional antinomies of Sociological Theory (Clarke, 1991; Mouzelis, 1998; Pleasants, 1999), a sort of “third-way” of theorizing (Alexander, 1987: 376). More than 50 years after the first interventions by Bourdieu and 30 years after Archer’s Realist Social Theory, it seems as though they have succeeded in convincing the broader Sociological community that this is a worthwhile endeavor. If in Archer, the main question of Sociology is now to analyze the relationship between agents and structures at any given moment in time through practical research, Bourdieu proposed a series of conceptual and epistemological considerations of great use for the refinement of Sociological research, specifically in order to clarify this relationship between agents and structures. The hegemonical, and borderline common-sense position seems to be that good Sociological research is defined by its ability to analytically explain the relationship between agents and structures.

This “revolution,” however, hasn’t seemed to have produced a clarification of these concepts. A common complaint in the existing literature is the normalization of diffuse and ambiguous conceptualizations of these concepts, often in contradictory fashion (Crossley, 2022; Emirbayer & Mische, 1998; Hays, 1994; King, 1999). Nevertheless, there seems to be a consensus that it is effectively through this transcendence that the progress of Sociology should be sought after. One of the questions this article attempts to answer is if there might be a problem with the presuppositions of the contemporary discussion around the agency-structure antinomy. A common absence from this discussion, for example, is the possibility for the historical apprehension of the poles of this antinomy, and more precisely its correlation with specific historical social formations. There is an implicit presupposition in the manner in which this discussion has developed, namely that agents and structures are the fundamental components of social reality, as if they were the basic variables when one analyzes a given social formation. The evident result of this revolution that has been noted by several authors, is how Bourdieu and Archer end up reproducing what they were attempting to transcend, but it is important to understand why and how they have done this. This article argues, alongside Rose (2009), that this happens necessarily due to the shared presuppositions of neo-Kantianism that Sociology has incorporated into its theoretical cannon. Specifically, this assumed proposition that society presents itself to agents as an objective and external structure, and that the role of Sociological Theory is to explain the linkage between these two poles. What this article proposes is that Sociology should cease to treat agents and structures as the pre-given starting points of explanation and begin to ask how and why they appear as separate at all, that is, to constitute them theoretically rather than presuppose them.

This argument proceeds in four steps. The first establishes the neo-Kantian origins of the antinomy in Weber and Durkheim. The second performs an immanent critique of Archer and Bourdieu as the most sophisticated post-Parsonian attempts to transcend it. ¹ The third proposes Marx’s theory of commodity fetishism, as developed by the Neue Marx-Lektüre and WertKritik traditions, as the conceptual resource for understanding the antinomy as a historically specific social form. The fourth draws out the implications for Sociological Theory.

Weber, Durkheim, and the neo-Kantians

Within Sociological common sense, it is customary to contrast the works of Weber and Durkheim as divergent intellectual projects. Weber finds the object of his Sociology in the realm of actions, intentions, and values of individuals while Durkheim takes the unintelligible, the fait sociaux, as his main concern. While such distinctions in their respective works are unambiguous and cannot be ignored, insufficient attention has been given to their similarities beyond a shared ideal of Sociology. The aim of this section is to explore one such point of convergence that is directly relevant to the broader argument within this article, namely, the indebtedness of both authors to their neo-Kantian roots.

In Weber’s case, this connection was studied early on, given his well-known proximity to Rickert – one of the leading intellectual figures of a particular strand of neo-Kantianism – as well as to Windelband (Brand, 1979; Giddens, 1991; Parsons, 1949; Rose, 2009; Zagirnyak, 2023). In Durkheim’s case, while the connection to Kant is obvious and widely referenced (Garlitz, 2020; Giddens, 1991; Turner, 2017), it is not uncommon to place greater emphasis on the positivist inheritance in his Sociology (Alexander and Smith, 2005; Parsons, 1949), or even to identify the “late” Durkheim with Hegelian ideas (Gangas, 2007; Garlitz, 2020). Without developing an extensive discussion of Kant’s work, two considerations are sufficient for present purposes. Firstly, human rationality depends on certain categories and intuitions that do not originate in experience but are required to understand it. Second, these transcendental categories are universal because they are simultaneously the conditions of possibility of experience and the conditions of possibility of the objects of experience, thus only through them can human consciousness access reality objectively (Kant, 2009: B197). The central problem this bequeathed to Kant’s successors was the inability to conclusively determine the origin of these universal forms of consciousness (Arato, 1974; Rose, 2009: 4), or the problems raised by the tautology at the heart of the Kantian system whereby the conditions of possibility of experience are simultaneously the conditions of possibility of its objects.

Windelband and Rickert, through the crucial mediation of Hermann Lotze (Rose, 2009: 6–10), addressed this problem by separating the question of objective validity from the question of human consciousness altogether. Where in Kant the conditions of possibility of objective knowledge were contained exclusively in the forms of human consciousness, in these strands of neo-Kantianism the origin of objective validity must necessarily be external to consciousness, pertaining to culture, to values, to a sphere independent of consciousness itself, since the question being addressed is now about the nature of objective validity as such rather than its possibility (Rose, 2009: 10–14). It is in this reasoning that neo-Kantianism found its utility for Sociology: the answer to the dilemma about the origin of the universal forms of consciousness would fall upon the shoulders of Sociology.

Durkheim’s originality lies in the claim that the origin of the transcendental categories is social. In The Elementary Forms of Religious Life (Durkheim, 1961), he advances a fundamentally epistemological argument – a socio-empirical proposal for the validity of these categories – rather than a Sociology of knowledge (Rawls, 1996: 438, 462). Moving almost entirely within Kantian categories, Durkheim translates the universality of the forms of human consciousness as their social generality (Fuller, 2013: 250): these categories are valid and necessary because they are endowed with the moral authority of society – validity becomes the sociological foundation of morality (Rose, 2009: 15). The epistemological problem this creates is immediate: if Society is the pre-condition of knowledge, how can it be taken as an object of knowledge? How can something be simultaneously the pre-condition of experience and its object? Durkheim was aware of this, comparing it explicitly to the metaphysical argument for the existence of God (Durkheim, 1964, cited in Rose, 2009: 16), yet could not resolve it. Weber, despite his apparent divergence from Durkheim, performs the same foundational move. In his preface to the Objectivity of Social Science and Social Policy essay, where he makes his deference to Rickert, Windelband, and Simmel on the topic clear, (Weber, 1949: 50) Weber poses culture as a value-concept and the a priori of human action: “empirical reality becomes culture for us insofar as we relate it to value ideas” (Weber, 1949: 76). The validity of social science rests on its relationship to values, a relationship that can never be conclusively established since the validity of values can only be a matter of faith (Weber, 1949: 55). This text also served as an implicit critique of Rickert’s constitutive role for value-spheres since Weber wanted to claim a constitutive logic for Society as a transcendental reality rather than for value as a Sollen (Brand, 1979: 10; Rose, 2009: 21). Nevertheless, the antinomy persists, since ideal types are themselves value-laden and the quest for validity can only be an unending task that is itself a value.

An identical problem unites Weber and Durkheim: from the moment one accepts the existence of a Sociological a priori, the relation between the pre-condition and the conditioned becomes necessarily ambiguous (Rose, 2009: 15). In treating Society as a transcendental reality, both are committed to presupposing a transcendental subjectivity and a universalist, naturalist vision of human ontology – reducing the separation between society and consciousness to something natural to the human condition, and therefore ahistorical. Ultimately, Weber and Durkheim are confronted with the reality that Society necessarily appears as a reality distinct from individuals. Some of the main theoretical disputes throughout the decades have stemmed from their proposed use of transcendental inquiries to solve this issue. What is normally attributed to Parsons as his originality is already implicit in both: individuals are confronted with objective and objectified economic and cultural structures, and their normative action is conditioned by these structures (Clarke, 1991: 227). Parsons, it can be argued, simply takes open argumentative tensions within Weber and Durkheim’s theories and transforms them into Sociological cannon. In the post-Parsonian generation, as this article argues, the crux of this question is not resolved but rather ignored.

The post-Parsonian reproduction: Archer and Bourdieu

The criticisms levelled at Parsons were primarily directed at its over-structural determination, its supposed conservative foundations, and its inability to adequately account for social change and conflict. What is less frequently noted, however, is that the main theoretical responses to Parsons – whether the anti-Parsonian alternatives of the mid-20th century ² or the post-Parsonian synthesis projects of authors like Giddens, Bourdieu, Archer, Habermas – tended to share the same underlying diagnosis: that the central problem was the theoretical imbalance between the objectivist and subjectivist moments of social analysis. Bourdieu (1989) argued for the transcendence of this divide through his constructivist structuralism, while Archer (1995) appealed to the transcendence of an illegitimate conflation of ontologically distinct strata of social reality. Mouzelis, while explicitly criticizing Bourdieu, Giddens, and Elias for attempting to transcend the vocabulary that the agency-structure antinomy presupposes while retaining its fundamental logic, nevertheless accepts the criticism of Parsons’ over-structuralist emphasis (Mouzelis, 1998). Alexander argued that the main opposers of Parsons’ system merely retorted Parsons’ one-sidedness with their own (Alexander, 1987: 376).

What these diagnoses share, despite their considerable and substantial differences, is the implicit acceptance of the agency-structure antinomy as the central starting point of any theory. The main objective of a coherent and useful Sociological Theory is to account for the relationship between agents and structures, whether it may be through conceptual refinement, ontological statements, or methodological clarity. They all fail to question this separation between agents and structures in itself, they all equally fail to justify its theoretical existence. The result is that the poles of this antinomy are postulated as a given: it becomes methodology. This article, through the immanent critique of Bourdieu and Archer’s proposals, proposes a different point of departure. Rather than asking how agents and structures are related, it asks why is that they appear as separate entities at all, and under what specific historical conditions has this separation become the apparently invariant starting point of Sociological Theory. This article thus argues that the failure to bridge both poles of this antinomy is tied precisely to the fact that it is accepted as this fundamental presupposition. These theories, ultimately, presuppose precisely what they are trying to constitute since the separation of agents and structures is the basis of the theory, a statement that Rickert, for example, recognized as a difficulty in his own conceptualization (Rickert cited in Beiser, 2009: 24).

Critical realism: The retreat to ontology

Archer’s (1995) Realist Social Theory presents the most theoretically developed version of the Critical Realist Approach to the agency-structure dilemma, explicitly positioning itself as an attempt to transcend Sociology’s traditional ontological divides. Margaret Archer begins her seminal work arguing that “the problem of the relationship between individual and society was the central sociological problem from the beginning” (Archer, 1995: 1). From the outset, the main motivation behind her work has been to bridge this gap, that is not only interpreted as a gap between theorists and ideological perspectives on society, but also a part of the nature of society. After all, the problem of agency and structure “will always retain this centrality because it derives from what society intrinsically is” (Archer, 1995: 1). There are several aspects of Archer’s proposal towards this linkage but, arguably, the most central tenant of this theory is what Archer refers to as the “vexatious fact of society,” namely, what she sees as the “fact” that human beings are “simultaneously free and constrained” and that they “have some awareness of it.” Ultimately, Archer is speaking in purely ontological terms, deriving this fact from the nature of social reality and “human nature’s reflexivity” (Archer, 1995: 2). It is thus quite coherent that her project of Sociological Critical Realism has ontological stratification and emergentist ontology as its main components.

• “Properties and powers of some strata are anterior to those of others precisely because the latter emerge from the former over time.”

• “Once emergence has taken place the powers and properties (. . .) have relative autonomy from one another.”

• These properties in turn “exert independent causal influences in their own right” and these may be “non-observable” powers (Archer, 1995: 14).

There is an evident particular importance of time in Archer’s proposal. It is time that ultimately proves structural causality in the present (Archer, 1995: 77–81). It is worth remembering that what is at stake here is an ontological postulate of the notion of both agency and structure being qualitatively distinct “aspects of society” (Archer, 1995: 11). It is here that we see Archer’s main concern with retaining a valid concept of causality as the core of Sociological theory and practice in its full splendor. The only way that Archer sees fit to justify these ontological postulates is by claiming that because structure “pre-dates the action(s) leading to its reproduction or transformation,” and that because said reproduction or transformation “post-dates the action sequences which gave rise to it,” the only practical solution for the practitioner of sociological research is to simply accept this inevitability. It is almost as if it were impossible to theoretically unite these qualitatively distinct aspects of social life other than treating it as a matter of the passing time. It is telling, however, that Archer sees analytical dualism not as a methodology of historicity but rather as a “methodology based upon the historicity of emergence” (Archer, 1995: 66). From the valid concern that Sociology must develop a scheme of thought that allows for a simultaneous explanation of both agents and structures, Archer and Critical Realism seem content with separating them analytically and unite them in research. The relationship between agents and structures ceases to be a theoretical problem, becoming a merely practical problem on a case-by-case basis. Albeit very useful for researchers – since it validates them – this is hardly useful for a theoretical linkage between agents and structures. Critical Realism does not question the antinomy but rather affirms it positively as ontology. Archer speaks of an inconsistent linkage between ontology and methodology in the traditional debates, as if the prevailing antinomy was but an error of reason. Faced with the endemic circularity involved in attempting to understand Modern social forms, Archer interprets it as something ahistorical, as a problem of method, and fails to see that the objectivity of structures and agents is only possible in a specific form of society, it is not tied to a specific social structure but rather to a structure of the social. Archer’s notion of the sense of self as an a priori to sociality and human beings as the “ultimate fons et origio of (emergent) social life” (Archer, 2000: 7, 18) betrays more that it may seem: It is not that Archer has a problem with postulating a transcendental basis for society, or with the “regress” to a discussion on the “ultimate constituens” as something useless to social theorizing (Archer, 1995: 50). Rather, these criticisms of other theorists reveal at each paragraph an ideological notion of the transcendental self, as the basis and origin of experience. ³ At its core, CR merely transforms Kant’s transcendental subject into the empirical subjects, or in other words, the conditioned is presented as the unconditioned (Adorno, 2005: 248), the historically produced is presented as “ontologically vexatious” (Archer, 1995: 167), socio-historical reflexivity into “human nature’s reflexivity” (Archer, 1995: 2). The justification for the “ontological” distinction between structures and agents shows itself to come down to this concern with preserving a mythological notion of agents as the ultimate constituens. It presupposes exactly what it aims to constitute (Adorno, 2005: 248–259). However, the role of, if it is to go beyond mere historical contingency, must be to account for this constitution of things as they are. Archer’s “transcendental” account of society turns out to be an inversion between the constitutive and the empirical question: the empirical, capitalist society that presupposes an objectification of social relations and the Modern form of subjectivity are turned into the constitutive.

Bourdieu and the habitus of a sociologist

Bourdieu, it would seem, plays a role in Sociology akin to that of Kant in Philosophy. He inaugurates a “Critical Period,” which attempts to synthesize – and transcend – two opposing views when it comes to the possibility of knowledge, in this case, knowledge of “the social.” From the outset, Bourdieu had as his main motive to transcend the opposition between an objectivist view of the social – where Sociology could only develop true knowledge by a “break with primary representations” – and a subjectivist one, where this knowledge “is in continuity with common sense knowledge, since it is nothing but a ‘construct of constructs’” (Bourdieu, 1989: 15), that seeks to “make explicit the truth of primary experience of the social world” (Bourdieu, 1995: 3). It is important to understand how Bourdieu viewed this opposition, a view that seemed to shift with time, at least in the realm of adjectives. On one occasion it is seen as one among several oppositions that “artificially divide social science” (Bourdieu, 1990: 25), sometime as the “the ritual either/or choice between objectivism and subjectivism” (Bourdieu, 1995: 4), as “ultimately grounded in social oppositions (low/high, dominant/dominated, and so on)” (Bourdieu, 1988: 778) and “inscribed in social reality” but also as “false antinomies” (Bourdieu, 1990: 179).

To transcend this antinomy, Bourdieu stresses that any knowledge of the social world must develop a logic of practice as practice, or in other words, a “break with objectivist abstraction (. . .) of constructing their [practices] generative principle by situating itself within the very moment of their accomplishment.” (Bourdieu, 1995: 3), and a break with subjectivism that “does not reflect on itself and excludes the question of the conditions of its own possibility.” Bourdieu is essentially concerned with a “third-order knowledge” (Bourdieu, 1995: 4) that retains the objectivist moment but draws back on itself, by questioning the conditions of possibility of this mode of knowledge, that is “prepared to inquire into the mode of production and functioning of the practical mastery which makes possible both an objectively intelligible practice and also an objectively enchanted experience of that practice” (Bourdieu, 1995). In the simplest of terms, Bourdieu (1995) argues for the necessity to “pass from the opus operatum to the modus operandum” (p. 73). A more thorough explanation of this logic of practice as practice, or at least its necessity, is postulated as the necessity to escape the “realism of the structure,” that converts “objective relations (. . .) into totalities already constituted outside of individual history” (Bourdieu, 1995). It is through this logical succession that Bourdieu postulates the habitus as a system of “durable, transposable dispositions, structured structures predisposed to function as structuring structures.” An emphasis is explicitly placed on the habitus as a set of “principles of the generations and structuring of practices (. . .) collectively orchestrated without being the product of the orchestrating action of a conductor” (Bourdieu, 1995: 72), hence why it would be so misplaced to accuse Bourdieu of structural determinism.

Perhaps, due to some conceptual radicalism, Bourdieu may in fact seem a determinist: on some occasions habitus becomes a deus ex machina of social practices, habitus as “history turned into nature, that is, denied as such” (Bourdieu, 1995: 78), as a representation of a “particular state of this (social) structure” (Bourdieu, 1995: 78), habitus as “automatic and impersonal” (Bourdieu, 1995: 80). Simultaneously, Bourdieu emphasizes (at times in the same sentence) the practical conscious action that the habitus presupposes, that social agents are “virtuoso(s)” that intimately know their social contexts and how to operate within them (Bourdieu, 1995: 79), that “corrections and adjustments the agents themselves consciously carry out presuppose their mastery of a common code” (Bourdieu, 1995: 81). Now, and at risk of being ironic, is this a contradiction, a lapse in argumentation, an unbridgeable antinomy? Or is it perhaps the habitus of a sociologist, of which “he is not the producer” (Bourdieu, 1995: 79)? There seems to be a latent ontological tension within this argumentation.

For Bourdieu (1995), “[I]t is just as true and just as untrue to say that collective actions produce the event or that they are its product” (p. 82). The Sociologist is always confronted with a pre-made reality that is constituted through the “dialectical relationship” between the habitus as a logic of practice, that encapsulates the possibilities and limits of practice, and the “objective event (. . .) calling for or demanding a determinate response” that can only exert its “conditional stimulations” on “those who are disposed to constitute it as such” (Bourdieu, 1995: 83). Social practices are never “totally co-ordinated” but still products of the causality of objective structures (Bourdieu, 1995). Bourdieu leads the antinomies of Sociology to their limits, by attempting to conceptually transcend them: in the break with subjectivism, he gives primacy to validity, the pre-made coercive social structure that creates – but doesn’t dictate – the conditions of possibility of experience. In the break with objectivism, he gives primacy to value, the everyday experience of doxa as he terms it (Bourdieu, 1995: 164), the intimate knowledge agents possess of their social context. By simultaneously describing the habitus as a structural determination and an agentic competence, Bourdieu attempts to treat this contradiction as a feature of social reality rather than a conceptual problem. And in the last instance, Bourdieu is obliged to preserve this antinomy in his theoretical considerations. Bourdieu is entirely aware of sociological antinomies but in the last instance, true to his intellectual project, he can only see these as a product of scholastic fallacies, of ideological and/or conceptual confusion: sociologists mustn’t forget that objective structures are themselves products of historical practices that are constantly reproduced by historical practices “whose productive principle is itself the product of the structures which it consequently tends to reproduce” (Bourdieu, 1995: 83).

The same tension can be seen in the concept of field. Bourdieu argues that there is a “two-way relationship between the habitus and the field, where the field, as a structured space, tends to structure the habitus, while the habitus tends to structure the perception of the field” (Bourdieu, 1988: 784). Essentially, as was developed in a different article by Bourdieu, the relationship between habitus and field intends to provide a form of mediation to the confrontation, in every historical moment, between objectified history, “the history which has accumulated over the passage of time in this, machines, buildings, monuments, books, theories, customs, laws, etc.,” and embodied history. Habitus is thus seen as a “historical acquisition which makes it possible to appropriate the legacy of history” (Bourdieu, 2015: 305). This is Bourdieu’s way of preventing a merely mechanistic interaction between agents and structures, postulating a relationship of “ontological complicity” between habitus and field. It is the dilemma that Archer seems to discard as such, as a “vexatious fact,” that Bourdieu treats as a problem of history: “this native relationship to a familiar world is a possessing which implies the possession of the owner by his belongings” (Bourdieu, 2015: 306). The field is thus seen as the constellation of “objective relations between positions,” social positions that are somewhat congealed into objectified and objective relations of power or “the structure of the distribution of the species of power (or capital) whose possession commands access to the specific profits that are at stake in the field” (Bourdieu and Wacquant, 2007: 97). The concept of field, however, merely displaces the traditional antinomy into a distinct level of abstraction: it is treated as an objective structure that confronts and influences agents but the problem of the nature of this relationship is simply encapsulated by the concept of field, rather than explained by it.

In the relationship between both concepts this ambiguity reveals itself as a repetition of the problem with the traditional solutions towards the agency-structure dilemma, since these concepts presuppose what they are trying to constitute. In Bourdieu, the presupposition that the separation between agents and structures is a non-problem in itself, it is a transhistorical reality. The fact that there exist objective structures, both material and dispositions, that are virtually or de facto separated from social agents is still the main presupposition in Bourdieu’s system. The ambiguity in Bourdieu’s argumentation can be understood in this light, in the herculean effort to re-unite what is separated by Sociological reason. The existence of this separation can be asserted but taking it as an ontological postulate is to situate Sociology as what it has always been, no matter how critical: an ideology of the actual. What if there was an alternative manner of thinking the agency-structure dilemma? Could Sociology leave the plane of historical contingency and produce historical substantive arguments, which would help understand the social forms of Modernity? Could Sociology cease to think in universal categories of “the social,” to cease to think of “social structures” and think of structures of the social?

Marx and the historical specificity of objectification

If the post-Parsonian generation has been unable to escape the antinomy, it is in part because the alternative that Marx’s work contains has remained largely inaccessible to Sociology. Writing on the “several Marxs” within Sociology, Smart (2013: 40) proposes a triadic understanding of the sociological interpretations of Marx. Firstly, those who consider Marx’s work to be merely ideological and hence incompatible with the sociological project. Secondly, those who attempt to extract concepts from Marx to thus operationalize them in research and in general theories, inevitably decontextualizing them. And finally, those sociologists who see in Marx the potential for a critical or liberational Sociology. (Smart, 2013: 2–3) In Smart’s view, the second tendency is the most common within Sociology. The most crucial distinction deployed by Marx, and the one that represents the most crucial import of Marx to Sociology, is his distinction between exchange-value and value, something that seems to be completely absent from the relevant sociological literature. By value, Marx means neither the price of a commodity nor the magnitude of its exchange-value, nor its utility, but the specifically social form in which labor comes to be objectified under capitalism: abstract, socially general labor as the principle through which otherwise private and isolated labors are rendered commensurable. Exchange-value is the necessary phenomenal form in which this social substance appears. Alexander (1982: 171–182) having been arguably the first sociologist to see the significance of Marx’s notion of the fetishism of commodities as an objective rather than subjective reality, still fails to make this distinction and it is this that informs his reading of the commodity-form as an explanation for instrumental reason, and the fetishism of commodities as an explanation for the alienation of the individual worker. Giddens (1991) equally fails to make this distinction and manages to represent both the second and third tendency identified by Smart, by over-emphasizing Marx’s philosophy of history and the notion of class society and reading the former onto Capital. Turner (2013: 162) proceeds in the same manner, claiming that Marx sees the value of commodities as inhering in the labor necessary to produce them, a notion that Marx explicitly mocks. Elder-Vass (2022: 29) criticizes Marx for not demonstrating why it is “valid to abstract from the qualitative differences of different kinds of labour,” a question which can only be posed in the absence of this distinction. While all these authors may seem to differ radically in their interpretation of Marx, they wholly fit into the tendency identified by Backhaus (1980) of proceeding as if Marx’s object of study in Capital was the measure of value, and not its substance, that is, as if Marx were offering a quantitative labor theory of value in the tradition of Ricardo, rather than a qualitative critique of the social form in which labor appears under capitalism. This misreading has a precise consequence: it renders invisible precisely the aspect of Marx’s work that is most relevant to the agency-structure debate.

Value and labor in Marx

The beginning of Marx’s (1904) critique is present in the unpublished Preface to his Contribution to the Critique of Political Economy, that later appears in the Grundrisse (Marx, 1981). In it, Marx criticizes Classical Political Economy for falling into the error of taking the simplest categories of their object of study for granted, namely Labor and Money. Labor seems to be the “simplest and most ancient relation in which human beings – in whatever form of society, play the role of producers.” However, a closer look reveals that this abstraction – labor as such, without further qualification – is “as modern a category as are the relations which create this simple abstraction.” This indifference towards particular and concrete forms of labor – tailoring, agriculture, etc. – is already in itself a manifestation of a specific society, where “no single one (concrete forms of labor) is any longer predominant,” and this is a fundamentally modern reality, whereby “labor (. . .) has become the means of creating wealth in general” and labor “achieves practical truth as an abstraction” (Marx, 1904: 297–299). It is in this logical sequence that Marx introduces the question of the logical and historical sequence of the categories of capitalism. Marx posits that the relationship between the logical and the historical is troubled if one fails to consider that this specific form of society, capitalism, “by no means begins only at the point where one can speak of it as such” Marx, 1904: 302). Capital, being the “all-dominating economic power of bourgeois society” and being historically posterior to the existence of money and labor, couldn’t possibly be the starting point of a historical sequence. Rather, “their order of sequence is rather determined by the relation they bear to one another in bourgeois society, and which is the exact opposite of what seems to be their natural order or the order of their historical development.” (Marx, 1904: 304). Hence, capital is always presupposed throughout the exposition of the commodity-form and the deduction of money, as it must “form the starting point as well as the end” (Marx, 1904: 303).

Some confusion may have arisen since Marx postulates that bourgeois society and the categories that express its relations, being a society where they reach their higher development, “thus supplies the key” to understand older modes of production but “this is to be taken only with a grain of salt.” Marx intends to argue that if one is to proceed in this manner, to understand tribute and tithe by understanding ground rent, one cannot “smudge over all historical differences and see bourgeois relations in all forms of society” (Marx, 1904: 301). This duality persists throughout Marx’s remaining writings and there is ample philological evidence to sustain two contradictory interpretations. Thus, it can simultaneously be argued in citing Marx, that the categorial exposition in Capital is merely logical or that it is the exposition of an historical process whereby the commodity-form evolves from the simple form of value into money and capital. This sort of tension runs throughout the whole of Marx’s Critique of Political Economy, a tension between a very Modern philosophy of history whereby history is an evolution towards the most abstract principles of human life, and between a sort of negative philosophy of history that attempts to apprehend the categories of capitalism as fundamentally historical, to deduce a categorial criticism of capitalism as a social formation that appears to be natural, to correspond to human nature (Kurz, 2014: 31–62).

Capital

Marx considered the first chapter of Capital, namely the section analyzing the commodity-form, the most difficult aspect of his book and the fundamental key to understanding the remainder of it (Marx, 1986: 90–91). In several passages, Marx alludes to the unanswered question on the qualitative aspect of labor under capitalism. A commodity contains within itself a tension between use-value and exchange-value that both Smith and Ricardo allude to but can never consistently resolve. As a use-value, a coat is simply an object of utility. As exchange-value, however, it appears as a quantitative relation, as something that is mutually commensurable with other commodities regardless of their qualitative differences. It is this question, of what makes this commensurability possible in the first place that Classical Political Economy never poses (Marx, 1986: 173). Ricardo, whom Marx regarded as the most rigorous representative of the classical tradition, confined his investigations exclusively to the quantitative determination of value, never asking why labor is expressed in value, why value must take the form of exchange-value, or why the products of human labor must take the form of objects (Marx, 1971: 129–131).

If one commodity can be exchanged for several different commodities, their use-value cannot be the common denominator – there must be something else at play (Marx, 1986: 125–131). In the simple equation 1 pair of shoes = 1 coat, only one commodity has an active role. The pair of shoes needs to have its value expressed, while the coat serves only as the material in which that value appears – it counts as nothing but value, its use-value is entirely irrelevant to the equation (Marx, 1986: 140). What Marx accused CPE of overlooking was the fact that this equation presupposes that these commodities can only be comparable in quantitative terms when reduced to a common unit. Crucially, it is not theory that performs this reduction – the equation does it. The abstraction happens within the process itself (Marx, 1981: 230). The magnitude of value is irrelevant here – it is not the object of study in Capital. In all these equations the fundamental inversion is still at play: use-value becomes the form of appearance of its opposite, value (Marx, 1986: 147). The expression of the value of one commodity always presupposes its subsumption to another, expressing its value-existence as something wholly different from its substance and properties (Marx, 1986: 149).

Looking at the labor contained in these commodities, the coat as it appears in the expression of the value of the shoes is nothing but the realization of human labor in the abstract. The concrete labor of the tailor is irrelevant as such – it becomes the form of manifestation of its opposite, abstract human labor (Marx, 1986: 150). Once we look at the residue of the products of labor there is nothing left of them in each case but the same phantom-like objectivity – they are merely congealed quantities of homogeneous human labor (Marx, 1986: 128). The substantive argument is that in capitalist society human labor is necessarily objectified into a congealed mass of labor in the abstract, and it is this objectification that lies at the core of the mystery of the value of commodities. The result and cause of this process of abstraction is capital: a society whereby the labor of individuals must attain an objectified form in order to be social. The coat, however, although it counts as a congealed mass of undifferentiated labor, is nonetheless the product of the labor of private and isolated producers, while simultaneously being represented as labor in its directly social form (Marx, 1986: 151). This is the distinction between value and exchange-value in its full development. When the value-form becomes generalized – 1 pair of shoes = 1 coat = 1 kilo of potatoes – a universal equivalent arises that takes on the material form of abstract labor as a social determinant of commodity production. The logic of money is to be deduced from the analysis of the value-form rather than from its historical origins. Money bears the stamp of a society where all labor is general and abstract, where it is the basis of the social.

The fetishism of commodities names the process whereby this inversion becomes the constitutive principle of social life. The relationships between producers, within which the social characteristics of their labors are manifested, take on the form of a social relation between the products of labor (Marx, 1986: 164–165). This is not a theory of illusion or false consciousness – it is a theory of how social relations are constituted in capitalism. Modern society is not unique in producing relations among humans that transcend them, but it is unique in the sense that what transcends them is the product of their own labor (Marx, 1986: 175). This fundamental inversion takes on the form of an automatic subject: both money and the commodity function only as different modes of existence of value itself, constantly changing from one form into the other without becoming lost in this movement (Marx, 1986: 255). Most importantly, the movement through which this process has been mediated vanishes in its own result, leaving no trace behind (Marx, 1986: 187). Capital is thus both the condition of possibility and the object of Modern society, the autonomous pre-condition and object of experience. This socially structuring character of the process of abstraction is what Sohn-Rethel (1978: 20) sought to foreground in his reading of Marx’s theory of value. ⁴ Value, as abstract labor, as labor without further qualification, thus turns itself into a very true social a priori that erects an apparently natural social structure. Rather than being a product of ideological illusion, Modern social structures are structured as such through this principle of the relationship between individuality and universality or in other words, it is not that Modern social relationships appear to be objects but rather that they appear in the form of objects. (Clarke, 1991: 253)

Marx and agency-structure

It is at this point that Marx’s analysis touches most directly on the agency-structure dilemma. In commodity-production, the labor of private individuals is social only to the extent that it is produced as a congealed mass of human labor in the abstract – Modern society thus appears as the subsumption of humanity to labor. The sociability of producers presupposes capital. The universality of commodity production, its sociability, can only be sustained as a false universality, as an external structure to the individuals it concerns, which nevertheless depends on those individuals carrying out their interests and volitions as private agents. ⁵ In their everyday life, capitalists and workers merely confront each other as bearers of capital, something which is posited in the economic relation itself (Marx, 1981: 230). The separation between agents and structures is therefore not an ontological given of social life but a product of this specific form of social mediation. What Sociology has taken as its invariant starting point – that individuals confront an objective social world that transcends them – is precisely what Marx’s critique constitutes theoretically rather than presupposes. This is not a solution to the agent-structure dilemma but a reorientation: rather than asking how agents and structures are related, Marx’s mode of inquiry asks why the products of collective human activity must appear as objective forces confronting the individuals who produced them. Agents and structures are not the furniture of social reality – they are historically produced forms that require theoretical constitution, and it is in this sense that the critique of political economy offers Sociology not an alternative set of concepts to operationalize but a fundamentally different form of inquiry.

Structure of the social, not social structures

The primary argument this article wants to put forth can now be stated summarily. Sociologists have historically argued over which pole of the antinomy should be prioritized and, more recently, how the two poles should be theoretically linked, when they should have been arguing over the historically situated social determination of the antinomy itself. The fundamental issue is that the tautology between the conditions and objects of experience has been made the implicit foundation of a discipline that sees itself as scientific. An antinomy cannot be the unmediated point of departure of a social science, particularly because it bears in itself the mark of a specific structure of the social: affirming the antinomy means affirming the fundamental mark of Modern society, while avoiding its socially produced necessity means only that one will reinvent it at a different level of abstraction. What the alternatives of Bourdieu and Archer share, despite their considerable differences, is the involuntary concealment of the historicity of the existence of agents and structures – this existence is turned into an invariant of human history, in the former explicitly and in the latter only implicitly. In Bourdieu, the attempt to transcend the antinomy remains constantly in the realm of epistemology, treated as an error of reason. In Critical Realism, the valid recognition that the antinomy is inscribed in social reality itself leads to its naturalization as ontology. The path that Bourdieu traced could never leave the realm of epistemology and the path that CR traced could never leave the realm of ontology. The criticisms presented of both are not meant to suggest that their theories should be discarded – they have undoubtedly represented a genuine scientific advance in the attempt to transcend an endemic split. What this article suggests is that sociologists should go with them and beyond them in the urgently necessary path they attempted to steer Sociology into.

Modern society does appear to be structured in such a way that one finds agents and structures separated from one another. There is nothing fundamentally wrong with describing society as the result of an intersubjective creation that presents itself to actual agents as an objectified system of norms and material coercions. The problem reveals itself once one considers the logical development of treating this circularity as the basis of a theory. To speak of social structures distinct from the agents that compose, alter and reproduce them already bears in itself the mark of a society whose totality presupposes the existence of individual agents whose nature as agents presupposes that they are individual and socially individualized. The abstractedness behind these concepts betrays their own historicity: to speak of structures without further qualifications (Marx, 1981) presupposes a stratified and differentiated society, presupposes that human socialization acquires an objectified form – a very specific structure of the social that is as Modern as this abstraction.

Behind the usage of this antinomy as the basis of sociological theory there lays an implicit philosophy of history: that of taking the actual as the pre-condition of itself. An historical abstraction from past societies passes for an empirical observation. Agents and structures are themselves what needs to be theoretically explained and constituted. What we have seen in the alternatives of Bourdieu and Archer is precisely the leap over this crucial step, this separation between agents and structures remains the presupposition, and therefore circularity is not only endemic but unavoidable. In the new ontological cage, the social relation of disunity between agents and structures is theoretically united and thus suppressed (Rose, 2009: 55–58), the social relation that creates the rift is either accepted as nature or conceptually mediated, and therefore not understood as such. What disappears is the very real separation between agents and structures that is socially produced and reproduced. Modern society becomes unknowable as such, converted into a specific and contingent manifestation of nature.

What this article proposes is not a new philosophy of history but rather to make explicit the current philosophy of history that Sociology is anchored in. Sociology is trapped in an eternal return of the neo-Kantian logic of conflating the pre-condition and the object of experience. Marx’s critique was introduced not to argue that it is an extensive account of Modern Society – it is not – but to argue that it represents a fundamentally different form of inquiry: one that attempts to see the logical in the historical and vice versa, not in the classical Hegelian sense that sees a unity between them, but as a negative approach that develops the contradictions between them. An approach that attempts to discern the contradictions between how a phenomenon appears in thought, how it functions, and how it has historically developed.

It is through this reconstitution of Marx’s fetishism of commodities, this article argues, that Sociology has access to a path of reconsideration of its most fundamental categories. Ultimately, it is a theory that allows for a rejection of the separation between agents and structures as a given, that seeks to go beyond the apparent objectification of social relations. It is an epistemological consideration that allows for an apprehension of the objectification of social means of reproduction and socialization as historical product, rather than as ontology. This opens the door for several methodological and epistemological considerations, namely one that allows for an apprehension of structures of the social rather than social structures. That allows for an apprehension of degrees of objectification and degrees of agency and, fundamentally, that allows for more historical rigor.