Recently several short pieces run in Jacobin have tried to repopularize Analytical Marxism.11. See, for example, Ben Burgis, “If You Want to Understand Marxism, Read G.A. Cohen,” Jacobin, April 10, 2022, and Nick French, “How Analytic Philosophers Made Sense of Capitalism,” Jacobin, January 25, 2023 Analytical Marxism, which gained popularity in latter decades of the twentieth century, is an attempt to reconstruct Marxism along the lines of the predominant intellectual trends in the Anglo-American academy. These accounts accept Analytical Marxism’s view of itself as a nondogmatic approach to Marxism and social theory. But as I will argue, Analytical Marxism turns out to be a dogmatic and reductionist approach to social explanation.
Analytical Marxism is actually founded on two dogmas: the first, its belief in methodological individualism, that social phenomena can be reduced to aggregates of individual behavior; and the second, its faith in rational choice methodologies. When taken to their logical conclusion, these two dogmas substitute rational choice theories for genuinely social theories. This substitution identifies social and historical dynamics with the dictates of reason, leading Analytical Marxists to recreate the Hegelian teleology they ostensibly critique.
Internal debate has led some Analytical Marxists to soften the influence of its founding dogmas. But such attempts mute the critique of “bullshit” Marxism motivating Analytical Marxism’s revision of historical materialism. Additionally, they tend to undermine its purported advantages over competing schools of historical materialism. Consequently, the question as to what is distinct and what is valuable in an Analytical Marxism stripped of its founding dogmas is open to investigation.
Cohen’s Ultimatum: Methodological Individualism or Bullshit?
In his introduction to Karl Marx’s Theory of History, G.A. Cohen proposes “non-bullshit” Marxism as an alternative name for Analytical Marxism.22. G.A. Cohen, Karl Marx’s Theory of History: A Defense (Princeton; Princeton University Press, 2000), xxv. For Cohen, more traditional schools of Marxism often relied on assumptions supported by esoteric but ultimately vacuous theory—in short, by bullshit. While Cohen concedes several forms of Marxism are neither analytical nor bullshit, he also pushed a stark choice: Marxism must either “become analytical, or become bullshit.”33. Ibid., xxvi.
Given the strength of this ultimatum, it’s clear that Cohen believed that analysis was the cure for Marxist bullshit. What then is “analytical” about Analytical Marxism? For Cohen, Analytical Marxism is “analytical” in both the broad sense of using precise argumentation and in the narrow sense of privileging analytical technique. This technique is characterized by “its disposition to explain molar phenomena by reference to the micro-constituents and micro-mechanisms that respectively compose the molar entities and underlie the processes that occur at a grosser level of resolution.”44. Ibid., xxiii. To be sure, the precise argumentation of Analytical Marxism is its most laudatory feature. On the other hand, its commitment to analytic technique—which Cohen characterizes as “absolute” and “un-revisable”55. Ibid., xxiv.—motivates Analytical Marxism’s critical project and its problematic commitment to methodological individualism.66. Cohen makes the connection between analytical technique and methodological individualism explicit at Ibid., xxiii.
Marxism must become analytical because, for Cohen, classical Marxist theory relies on pseudo-causal explanations that encourage bullshit. For example, most Marxists argue that capitalism inevitably produces crises that can facilitate socialist revolution. However, this argument rests on collective “agents”—classes, institutions, and modes of production—as causes for historical transformations. Marxist bullshit relies on such explanations without itself pinpointing the causal power of these agents. Such bullshit is widespread in the use of teleological explanations that rest on reason, the proletariat, or history itself. In a Hegelian vein these supra-individual agents are cited as causes without the accompanying explanation of how these “agents” actually carry out historical change.
For Cohen, these “collective” agents are nothing but aggregates of individual behavior. Put simply, the proletariat does not exist. What does exist is the behavior of working class individuals. On this view, good explanations of social change are those that make sense of the individual behaviors that produce them. As Cohen writes:
to claim that capitalism must break down and give way to socialism is not yet to show how behaviors of individuals lead to that result. And nothing else leads to that result because behaviors of individuals are where the action is, in the final analysis.77. Ibid., xxiv.
Analytical Marxism’s reconstruction of historical materialism is thus meant to go where the action is. It attempts to supply the causal account neglected by “bullshit” Marxism, that is methodologies that unreflectively appeal to supra-individual agents.
The superiority and rational legitimacy of Analytical Marxism is thus founded on methodological individualism, the notion that explaining social phenomena is simply explaining aggregates of individual behavior. If Marxism must “become analytical” or “become bullshit,” it is precisely because Analytical Marxism applies the “most advanced” techniques of economics, science, and philosophy to explain individual behaviors. Thus, the two dogmas of Analytical Marxism entail a reductionist epistemological project. Explanation is reduced to individuals’ behavior.
In Karl Marx’s Theory of History, Cohen attempts to reconstruct key points of Marxist theory on analytic lines. Cohen identified the thesis that technological and economic capacities tended to increase throughout history as the key tenet of orthodox historical materialism. For Cohen, this principle was the basis of both Marx’s purported technological determinism—the belief that social transformations were determined by technological and economic development—and his progressive view of history.
Cohen rooted this developmental view in human rationality, which he saw as the tendency to maximize utility. He held that, rational human beings prefer receiving the most possible benefit from their actions. He then generalized this notion and argued that this preference, combined with the social ability to preserve technical knowhow, meant that humanity’s productive capacities tended to increase over time. Given that rational individuals tend to prefer increases in productive capacities, societies would tend to “select” technologies and forms of economic organization that increased such capacities.
Cohen’s reconstruction of Marx’s theory of history was successful in, Alan Carling writes, “re-orienting the sociological imagination.”88. Alan Carling, “Rational Choice Marxism,” New Left Review 1, no. 160 (Nov/Dec 1986): 28. Instead of resting on sweeping accounts of history, Cohen’s argument reoriented research interests toward the microprocesses of individual behavior. As Carling explains, Cohen’s theory allowed for the reinstatement of the subject in the wake of the critiques of agency launched by the French structuralists of the latter third of the twentieth century.
The minimalist theory of reason allowed theorists to zero in on individual behavior without making robust assumptions about the consciousness of individuals.
Cohen’s configuration of individual subjects as rational utility maximizers avoided the problem of ascribing intentional attitudes to agents. Different positions that ascribe robust motivational structures to historical agents based on considerations of class, nationality, and so forth impute forms of consciousness that often run against the behavior and stated intentions of actual working class actors. Finding ways to get beyond them, as Analytic Marxism promised to do, was understandably attractive. The minimalist theory of reason as value maximization appeared a promising means to that end as it allowed theorists to zero in on individual behavior without making robust assumptions about the consciousness of individuals.
In spite of these gains, Analytical Marxism’s “no bullshit” project failed to escape its own gravity. Within the tradition, Jon Elster has criticized the functionalist character of Cohen’s argument. Functionalist theories explain the causes of historical phenomena in terms of the effects they produce. For Elster, such circular explanations simply stipulate that effects tend to produce their own causes.
Such theories posit a general historical trajectory towards capitalism and the nation-state (in Hegel) or communism (in Marx). For example, the existence of capitalist phenomena is explained by appealing to their necessity for capitalism itself.99. Jon Elster, “Marxism, Functionalism, and Game-Theory: The Case for Methodological Individualism,” Theory and Society 11, no. 4 (1982): 458–59. Thus, capitalism—or any other supposed historical endpoint—becomes the unexplained premise that explains its own genesis. As Elster notes, Cohen’s functionalism omits an account of the causal mechanisms that bring about the historical changes he simply derives from the principle of utility maximization.1010. Jon Elster, “Cohen on Marx’s Theory of History,” Political Studies 28, no. 1 (1980): 127.
This line of argumentation ultimately identifies Cohen’s functionalism with the very methodological holism that he critiques as Marxist bullshit. That is, Cohen’s reliance on rational optimization ends up explaining history in terms of a supra-individual entity. By basing the tendency of economic forces to develop on individuals’ rational utility maximization, Cohen has ultimately posited rationality’s pervasive influence on the world. Moreover, like the Marxist “bullshit” he critiques, Cohen fails to identify the agent that actually causes the changes prescribed by reason. This is a glaring problem.
Put simply, even if it is rational for humans to favor economic changes that increase productivity, historically situated rational agents may not have the capacity, foresight, or inclination to enact them over other desires. Given that methodological individualism is premised on the belief that causation happens at the level of individuals, Cohen’s functionalism, while in principle compatible with an explanation of individual behavior, does not amount to a causal description of the changes it posits.1111. While different rational choice Marxists have more relaxed views on functionalist explanation, Elster’s view is that it is otiose. See Ibid. On its own, rather than showing the microprocesses of individual behavior, Cohen’s theory simply reduces history’s trajectory to the actions of a singular macro individual.
Cohen’s slide back into teleology is symptomatic of the explanatory weakness of general theories of rationality. Cohen’s technological determinism—his contention that the increase of productive forces is the ultimate explanation of historical change—is rooted in his characterization of universal rationality as value maximization. The forces of production tend to increase because human beings, as a whole, tend to maximize value.
However, granting that the principle of value maximization is universal would not establish a causal dependence between that principle and individual behavior. Cohen merely identifies a reason—albeit a pervasive and universal reason—for an agent to perform an action. Thus, if we take Cohen’s characterization of rationality for granted, it can generally be asserted that: humans adopted capitalist productive relations, and their proclivity to maximize value provides a reason to choose more efficient techniques of production. However, the “and” in the previous sentence falls well short of “because.” Explanations of behavior must do more than identify a motive to provide a casual explanation; they must identify the reason that causes the action.
Merely attributing a motive to an individual fails to establish a causal role for that motive because motives invariably coexist with other complementary motives or, alternatively, fail to produce the effects we infer from them. We often have an excess of possible motivations for any action we perform. To take a mundane example, we might apply for a better job in Los Angeles both for the increased financial benefits and because we prefer the climate on the West Coast. In such cases, simply asserting a general tendency in human rationality fails to establish that this motive caused the behavior in question.
Simply having a motive for an action does not provide a causal explanation, as the connection between mental states and actions defies generalizations.
Moreover, the percentage of times that humans perform actions that they have a reason for is vanishingly small. To return to the previous example, although a person may know of numerous better paying jobs, they may never apply for any of them. Thus, the causal role of the general tendency to maximize utility is questionable, given that it fails to produce the effects we infer from it. Simply having a motive for an action does not provide a causal explanation, as the connection between mental states and actions defies lawlike generalizations.
To return to the second example, perhaps arguing that an authentic notion of utility must include the risks and expenses incurred by applying for a new job would undercut its force. However, this line of objections simply concedes my ultimate point: In either case, drawing the connection between an individual’s behavior and a general rational principle requires further subsidiary work in order to sustain a causal explanation. As I argue in the following, identifying the causal role of beliefs, preferences, desires, and other mental states actually occurs at this “subsidiary” level. This explanatory work requires a description of the agent’s practical context.
A common strategy for dealing with the anomaly of the mental is to modify lawlike generalizations about human behavior by adding clauses that stipulate that the generalizations only obtain in most situations. To return to the example of the principle of the development of the productive forces, one could argue that humans tend to opt for more efficient production relations so long as certain conditions are met. One can maintain this generalization by enumerating all these conditions—the technical knowhow to organize more efficient forms of production relations, a balance of political forces propitious to adopting them, the absence of other countervailing exigencies mitigating against economic reorganization, and so on. Or one can simply say that humans tend to maximize utility and, consequently, adopt more efficient relations of production, all things being equal.1212. These clauses are referred to as “ceteris paribus” because they stipulate a lawlike relation between cause and consequent “all things being equal.” The addition of such a proviso, known as a ceteris paribus clause, seemingly upholds the validity of general causal principles in light of their empirical exceptions.
Attention to the function of ceteris paribus clauses reveals the deeper problem of such attempts at explanation. In his analysis of ceteris paribus clauses in practical contexts, Robert Brandom suggests that we should view them as signifying the inherently comprehensive character of practical reasoning, rather than as a deus ex machina that removes it.1313. Robert Brandom, “Actions, Norms, and Practical Reasoning,” Noûs 32, no. S12 (October 1998): 127–39, 132. That is, rather than being determined to a specific behavior by an overriding rational principle, rational agents pursue the action they think is best all things considered. Brandom argues ceteris paribus clauses are best conceived as a shorthand for the total list of conditions that would result in exceptions to our generalizations about rational behavior. We cannot, in principle, complete such a list because the list is by its nature indefinite. As Brandom asserts, we would not even know how to specify what is on the list.1414. Ibid., 133.
Because we do attempt to consider all things, empirical content plays an important function in identifying the intention that motivates an action. Ceteris paribus clauses help mark the consequences that result from our attitudes, all things being equal, because they indicate that practical agents consider the totality of desires and circumstances they face. Moreover, if, as Brandom suggests, the addition of ceteris paribus clauses amounts to listing the conditions which would lead a motive to fail to produce the predicted behavior, the full specification of these clauses amounts to an empirical consideration of the agent’s total practical context.
For example, if a technological determinist argues that humans will transform their mode of production in order to maximize utility in a given situation, the ensuing specification of the situation would entail a number of other conditions. The cost of adopting a more efficient set of production relations must be weighed in relationship to other practical exigencies placed on a set of agents, such as the relative class interests of distinct groups. Even were we to concede an agent wanted to maximize their utility, establishing that this desire caused a change in production relations entails an examination of those exigencies.
Ultimately, the preceding observations lead to two broad conclusions. First, the causal role that a person’s reasons play in relation to their behavior is constitutively connected to the practical context in which they are engaged. Given that the number, weight, and relative effect of these considerations on the motivations of an agent cannot be determined on the a priori level, an explanation of historical transitions must take account of the actual analysis of these conditions.
Second, this practical context is an inherently open system. If, as Brandom suggests, the list of conditions that allow one to draw a strong connection from a motive to a behavior is impossible to specify, it is because humans engage in practical contexts that are open to conditions that can influence their decisions. Since establishing a causal connection between motives and action depends on a description of the empirical context in which an agent acts, the explanatory force of identifying motivations or analyzing reason cannot result from the general principles of theory. A general theory of reason is insufficient to meet Analytical Marxism’s ambition of providing a theory of historical change that accounts for the microprocesses of individual behavior that effectuate those changes.
Cohen’s derivation of the development of the productive forces stems from his desire to root a macro social historical change in the micro level behavior of individuals.
Crucially, the dependence of Cohen’s theory on a general theory of reason is a result of his dogmatic insistence on methodological individualism; Cohen’s derivation of the development of the productive forces stems from his desire to root a macro social historical change in the micro level behavior of individuals. This is motivated by his insight that individuals are causal agents of historical change. “Bullshit” Marxism neglects to provide a causal explanation for the changes it attributes to the social whole. Methodological Individualism is meant to address this explanatory deficiency. As I’ve shown, appealing to a general rational principle fails to meet the ambitions of Analytical Marxism’s first dogma.
Old Wine, New Bottle
The preceding argument brings us to the second dogma of Analytical Marxism: it’s faith in rational choice modeling in general and game theory in particular. The general tendency in attempts to reconcile Cohen’s and Elster’s frameworks has been to hold that game theory and rational choice modeling can supplement functionalism by identifying forms of strategic rationality that explain collective action and class conflict. However, to fulfill this function, game theory must be able to provide an empirically testable theory superior to competing theories of both social explanation and individual behavior. As I argue, the hope that game theory can fulfill this promise is mistaken.
As noted by Elster himself, game theory had failed to generate testable hypotheses and had few applications in nonexperimental settings.1515. Elster, “Marxism, Functionalism, and Game-Theory,” 477. At the time, the empirical purchase of game theoretical modeling was, at best, a promissory note. Presently, the usefulness of game theory in making predictive claims has been limited to closed systems (such as online poker) or in constructing price-setting algorithms modeled on online auctions. That is, the empirical grip of game theory seems to be limited to human interactions nearly identical to the formalized model of a game. Game theoretical models of largescale social phenomena are just as divorced from empirical data as the theoretical models of “bullshit” Marxism.
This failure to capture the day-to-day goings-on stems from game theory’s failure to grasp the very thing it tries to model: rationality itself. Writing in the wake of similar failed attempts to empirically test rational choice models, Donald Davidson interrogated their ability to actually capture the operations of rationality.1616. Donald Davidson, “Hempel on Explaining Action” in Davidson, Essays on Action and Events (Oxford: Oxford University Press, 2001), 219–20. Davidson argued that such theories fail to provide a causal explanation of action. This is because, to expand on our earlier discussion of Brandom’s work, they misidentify what general theories of rationality can actually explain. Davidson argues that they simply cannot give a general causal account of individual behavior.
For Davidson, the dispositions charted by verbs such as “wants” and “prefers” are relative to specific agents. Intentional explanations attempt to identify the motives that cause rational agents’ actions by attributing attitudes, preferences, and beliefs to them. Because such attitudes are invariably tied to the entirety of an agent’s beliefs, distinct beliefs can only be attributed to them within the context of an overarching theory of that worldview.1717. To put the point in more precise terms, the content of our mental states is determined by the history of their acquisition, which is both personal and specific. Cf. Donald Davidson, “Could There Be a Science of Rationality,” in Davidson, Problems of Rationality (Oxford: Oxford University Press, 2004), 121. Consequently, such explanations are necessarily specific and depend on a wealth of empirical evidence about the agent’s beliefs, desires, and intentions, rather than on an axiomatic and general theory of reason itself.1818. Davidson, “Hempel on Explaining Action,” 222.
Davidson gets at an important distinction between intentional explanations and rational choice modeling in general. Rational choice modeling ostensibly explains by virtue of its theoretical axioms. These axioms are foundational postulates about rationality that are used as the basis for the arguments of rational choice theories. The principle of utility maximization serves as such an unargued-for premise in Cohen’s theory.
Game theory, as a variety of rational choice theory, attempts to chart the likely strategic decisions of social actors in different practical contexts. Social actors are conceptualized as players in a game, the constraints of the social context modeled as rules, and the actors’ possible courses of action are ranked in “preference sets” indicating a rational actor’s preferred options relative to those constraints. The power of game theory lies in its ability to model the interdependence between an agent’s decisions, practical constraints, and the actions of other “players in the game.” This type of model captures the simple truth that our rationales for action are based on our expectations about what other people will do. Crucially, the preference sets and rules used by game theorists are just as axiomatic as the principle of utility maximization is for Cohen.
Any disparity between the model’s predictions and reality can be explained either in terms of another axiomatic reordering of preferences or in a further interpretation of the subject’s behavior: Experimenters can either claim the subject’s preferences are other than they thought, or that their behavior indicates the same preferences attributed to them under some alternative interpretation.1919. Ibid., 220. In either case, no data set could possibly falsify the theory. That is, game theory is neither empirically falsifiable, nor testable.
This leads Davidson to conclude that “the main empirical thrust of a decision theory, or of a reason explanation, does not come from its axioms … but from the attributions of desires, preferences, or beliefs.”2020. Ibid., 222. Axiomatic assumptions about the general nature of rationality do none of the explanatory work. As I’ve suggested, the explanation happens on the subsidiary level of an actual analysis of intentions of specific agents. The Analytical Marxists’ neglect of the preceding point leads them to ultimately reproduce the speculative, teleological, and rationalist history they deride in Hegel.
Axiomatic assumptions about the general nature of rationality do none of the explanatory work.
Given that rational choice models such as game theory lack any empirical grip, it is not surprising that they ultimately end with the same reduction of collective agents to macro individuals just as Cohen did. While game theory ostensibly tracks individual behavior, the preference structures used to model individuals are generally applied across collective agents. To quote Elster: “The struggle between capital and labor is a two person game and between members of the capitalist class an n-person game. Often … complicated n-person games can be reduced without too much loss of generality to simpler two person games.”2121. Elster, “Marxism, Functionalism, and Game-Theory,” 465. Given that class formations are collectives, Elster’s use of “person” in the above paragraph is undoubtedly technical. For him, “person” refers to any unit of individuals whose behaviors can be modeled along the lines of identical motives and constraints on their actions.
Elster’s model, by reducing collectives to “persons,” simply applies the theoretical vocabulary he uses to understand an individual person to a collective. Rather than dispensing with collectives and social wholes as causal agents, Elster’s methodological individualism simply stipulates that the rational choice theories used to model individual behavior can be applied wholesale to social phenomena. Such a stipulation led Ellen Meiksins Wood to argue that the preference sets and rational axioms attributed to individuals by game theory are simply social structure that have been smuggled onto the level of individual analysis.2222. Ellen Meiksins Wood, “Rational Choice Marxism: Is the Game Worth the Candle?” New Left Review I– 177 (Sept/Oct 1989): 49. Wood, like many critics of Analytical Marxism, provides a worthwhile critique of the politics implicit in the methodological project. While I agree with its general thrust, I eschew extended commentary on this because it is well trodden ground in the literature on this school, and because I view the connection between methodology and political program to be quite loose.
Elster’s procedure is the consistent application of the two dogmas of Analytical Marxism: the belief that explanations of social phenomena can be reduced to explanations of individual behavior and that individual behavior is best analyzed in terms of rational choice methodologies. He simply stipulates that rational choice theory is social theory.
The problem, of course, is that claiming to be a social theory and being one are different things. At a minimum, a social theory should be able to take into account social evidence. Yet by relying on axiomatic decision theories rather than the ascription of intentions, beliefs, and preferences via intentional explanations, rational choice Marxists sidestep the problem of evidence and simply attribute motives to agents as the causes of their actions. As Elster himself notes, intentional explanations require a wealth of historical explanation.2323. Elster, “Marxism, Functionalism, and Game-Theory,” 461. Game theorists simply punt on the task of amassing such evidence in an obverse manner to the French structuralists. Rather than identifying the individual with a position in the structure, rational choice Marxists simply reduce collective agents to super-sized individuals.
Such a reduction overlooks the problem that collectives and individuals are distinct and operate according to different internal dynamics. For example, the interactions between the individuals who constitute a class are of a different type than the interactions between the various preferences that constitute an individual’s worldview. While Elster certainly meant for game theory to track the interactions between individuals that constitute collective agents, his goal was unreachable given his procedure of reducing collectives of distinct individuals to persons in a game. Simply effacing the distinction between individual and collective obscures the difference between theories used to explain social phenomena and theories used to explain individual behaviors.
This problem derives from a mistaken inference in their interpretation of method- ological individualism. As noted above, the basic insight justifying methodological individualism is that individuals are the causal agents whose aggregate behavior induces social changes. The fact that social changes causally depend on individuals in this sense does not entail that social dynamics are structured in the same way as individuals. Causal dependence does not imply some sort of essential similarity between a whole and a micro-process; put simply, just because a class is composed of individuals does not mean that a class can be conceptualized as an individual writ large. While the debate between reduct-ionist and nonreductionist theories of science is very much alive, the deleterious consequences that flow from the reduction of social theories to theories of individual behavior require that we opt for nonreductionist options.
Let me summarize those consequences. The Analytical Marxist’s substitution of individuals for the supra-individual entities they constitute fails to produce a causal explanation as to how individuals constitute themselves as social agents. This constitution is simply assumed through the ascription of identical preference structures—motives and intentional attitudes—to the constituents of the collective.
Elster is simply modeling collective behavior as if it were individual behavior. Class formation, from this perspective, is simply the realization of reason as a principled optimizing strategy in the same way that historical development is, for Cohen, simply the result of humanity’s rational impulse to maximize utility. Thus, Analytical Marxism ultimately views both the actions of collectives and history’s ultimate trajectory as the realization of reason.
The Hegelian echoes of the previous section are nonaccidental. The critical impetus of Analytical Marxism is a function of the problem opened by the distinction between supra-individual and individual agents. However, as I’ve argued, Analytical Marxism merely short-circuits this distinction by reducing supra-individual collectives to individuals. Consequently they, like Hegel, simply stipulate that historical and social dynamics are reducible to individual rationality.
That is, the first dogma of Analytical Marxism (the view that social phenomena ultimately reduce to aggregates of individual behavior) and its second dogma (that individual behavior is modeled best in terms of rational choice methodologies) ultimately reduce the theoretical vocabulary available for describing social phenomena to rational choice methodologies. Thus, the theoretical reduction of social phenomena to individual behaviors ends with the same teleological view of history as the realization of rational agency characteristic of Hegel.
In other words, it’s the same bullshit. Alan Carling responds to criticisms of their model of historical explanation in his “Defense of Rational Choice Marxism.” He writes:
The possible outcomes one can call France, England and Poland. In France the peasants get the land, and have no incentive towards technological development because they can remain self-sufficiently aloof from market forces. In Poland the lords get the land and dominate the peasants. They impose a second serfdom, with concomitant feudal stagnation. In England alone, the lords get the land, and the peasants get the peasants. This means that in England alone, both the direct producers and the ruling class have no alternative but to become enmeshed in market relations, with the consequences for economic development predicted by the sufficiency claim of Brenner’s Axiom. Once capitalism is established in the English countryside, England can break out of the feudal demographic cycle, and eventually the comparative advantage it enjoys enables capitalism to spread, over-turning feudalism in the process. Moreover, there is bound to be an “English” outcome sometime, somewhere, to the class struggle in the trough of the demographic cycle, given the decentralized character of feudal Europe. Capitalism is inevitable, in short, because there’ll always be an “England.” But this is to say that capitalism comes about when and because it would foster the development of the forces of production. Which is what Cohen says the functional theory of history says. Q.E.D.2424. Alan Carling, “In Defense of Rational Choice: A Reply to Ellen Meiksins Wood,” New Left Review I–184 (Nov-Dec 1990): 104–5.
In place of an explanation, rational choice Marxism leaves us with the same dogmatic and evidence-immured appeal to teleology characteristic of the most orthodox Marxists. The passage from Carling is, admittedly, a full-throated application of what I have termed the two dogmas of Analytical Marxism.
Dissenting voices from within the tradition have put these dogmas into question. Wright’s, Levine’s, and Sober’s essay on methodological individualism have trimmed the explanatory ambitions of rational choice theory and argued for the consideration of competing theoretical frameworks to explain individual behavior.2525. Erik Wright, Andrew Levine, and Elliot Sober, Reconstructing Marxism (New York: Verso, 1992), 107–28. A 2012 essay by Vivek Chibber argues that the cost of shoring up both a robust rational choice theory of history like Carling’s or the more minimalist approach of Wright, Levine, and Sober is their substitution by a class struggle theory of history.2626. Vivek Chibber, “What is Living and What is Dead in the Marxist Theory of History,” Historical Materialism 19, no. 2 (2011): 60–91. I will leave aside an in-depth discussion of Robert Brenner’s membership within the Analytical Marxist school. Suffice it to say, I would be delighted if Analytical Marxism shorn of its dogmas turned out to be nothing more than Brenner.
However, if the critical purchase of Cohen’s ultimatum is lessened, then the signature achievements of Analytical Marxism lack the impetus that made them so attractive. For example, if the grip of methodological individualism is loosened, then does Roemer’s reconstruction of exploitation in terms of rational choice theory—heavily based on hypothetical constructions that ignore existent power relations—have the same weight? The more the two dogmas are relaxed, the less Analytical Marxism has to both distinguish it and recommend it next to competing theories. I suggest that once these dogmas are thrown away, it becomes apparent that the solutions to “bullshit” Marxism are not to be found in the theoretical paths Analytical Marxism has already charted. In that case, attempts to repopularize Analytical Marxism simply pours old bullshit into new bottles. ×
Notes & References
- See, for example, Ben Burgis, “If You Want to Understand Marxism, Read G.A. Cohen,” Jacobin, April 10, 2022, and Nick French, “How Analytic Philosophers Made Sense of Capitalism,” Jacobin, January 25, 2023.
- A. Cohen, Karl Marx’s Theory of History: A Defense (Princeton; Princeton University Press, 2000), xxv.
- Ibid., xxvi.
- Ibid., xxiii.
- Ibid., xxiv.
- Cohen makes the connection between analytical technique and methodological individualism explicit at Ibid., xxiii.
- Ibid., xxiv.
- Alan Carling, “Rational Choice Marxism,” New Left Review 1, no. 160 (Nov/Dec 1986): 28.
- Jon Elster, “Marxism, Functionalism, and Game-Theory: The Case for Methodological Individualism,” Theory and Society 11, no. 4 (1982): 458–59.
- Jon Elster, “Cohen on Marx’s Theory of History,” Political Studies 28, no. 1 (1980): 127.
- While different rational choice Marxists have more relaxed views on functionalist explanation, Elster’s view is that it is otiose. See Ibid.
- These clauses are referred to as “ceteris paribus” because they stipulate a lawlike relation between cause and consequent “all things being equal.”
- Robert Brandom, “Actions, Norms, and Practical Reasoning,” Noûs 32, no. S12 (October 1998): 127–39, 132.
- Ibid., 133.
- Elster, “Marxism, Functionalism, and Game-Theory,” 477.
- Donald Davidson, “Hempel on Explaining Action” in Davidson, Essays on Action and Events (Oxford: Oxford University Press, 2001), 219–20.
- To put the point in more precise terms, the content of our mental states is determined by the history of their acquisition, which is both personal and specific. Cf. Donald Davidson, “Could There Be a Science of Rationality,” in Davidson, Problems of Rationality (Oxford: Oxford University Press, 2004), 121.
- Davidson, “Hempel on Explaining Action,” 222.
- Ibid., 220.
- Ibid., 222.
- Elster, “Marxism, Functionalism, and Game-Theory,” 465.
- Ellen Meiksins Wood, “Rational Choice Marxism: Is the Game Worth the Candle?” New Left Review I– 177 (Sept/Oct 1989): 49. Wood, like many critics of Analytical Marxism, provides a worthwhile critique of the politics implicit in the methodological project. While I agree with its general thrust, I eschew extended commentary on this because it is well trodden ground in the literature on this school, and because I view the connection between methodology and political program to be quite loose.
- Elster, “Marxism, Functionalism, and Game-Theory,” 461.
- Alan Carling, “In Defense of Rational Choice: A Reply to Ellen Meiksins Wood,” New Left Review I–184 (Nov-Dec 1990): 104–5.
- Erik Wright, Andrew Levine, and Elliot Sober, Reconstructing Marxism (New York: Verso, 1992), 107–28.
- Vivek Chibber, “What is Living and What is Dead in the Marxist Theory of History,” Historical Materialism 19, no. 2 (2011): 60–91. I will leave aside an in-depth discussion of Robert Brenner’s membership within the Analytical Marxist school. Suffice it to say, I would be delighted if Analytical Marxism shorn of its dogmas turned out to be nothing more than Brenner.