Analysis Terminated? Towards a Post-Analytical Marxism
Or: What Can Bullshit Marxists Learn from G. A. Cohen?
I never encountered Jerry Cohen, the man. I only knew G. A. Cohen, the author of important and influential books and essays in analytical Marxism. Jerry Cohen, so I’m told, was playful, funny, kind, and generous. The G. A. Cohen I knew was dead serious – if also capable of wit – harsh in his judgments, and quite intimidating. He was also incredibly sharp and really knew his way around Marx. I read his work – everything relevant to the study of Marx, that is – when I was in Amsterdam working on my dissertation on Marx’s concepts of labor. I found some of it helpful for my project, much of it ever so slightly disagreeable, and all of it rather less enticing than the Althusser I was reading at the same time. Nonetheless, even in my youth I recognized that Karl Marx’s Theory of History was the most formidable exposition of that other kind of Marxism, the attempt to make sense of Marx in the terms and by the conventions of Anglo-American academic philosophy and social science. I took Cohen much more seriously as a Marxologist than I did Jon Elster, for instance (indeed, footnote references to Elster in my dissertation included a parenthetical “sic!” after the title of his book, Making Sense of Marx). It was quite apparent to me that Cohen was seeking to clarify what Marx wrote precisely because he was convinced that, in rough outline at least, Marx was right: right about history, right about society, right about capitalism and the need to overcome it. His was not my kind of Marxism, but it seemed to me an intellectually honest, respectable, and challenging kind of Marxism, nonetheless.
I don’t think Cohen would have had the same judgment of me. In the introduction to the 2000 edition of Marx’s Theory of History, Cohen recollected that “before others taught me to call what we were doing ‘analytical Marxism,’ it was my own practice to call it ‘non-bullshit Marxism’” (KMTH, xxv). He admits that the term is “aggressive,” since “when you call what you do non-bullshit Marxism, you seem to imply that all other Marxism is bullshit.” He seems for a moment to undercut this aggressiveness by conceding that “there exists Marxism which is neither analytical nor bullshit,” but this concession has a sting in its tail, for he concludes by naming this non-analytical, non-bullshit Marxism “pre-analytical Marxism,” and declaring that whenever “pre-analytical Marxism encounters analytical Marxism, then it must either become analytical or become bullshit” (KMTH, xxv-xxvi). Since my Marxism encountered Cohen’s analytical Marxism in 2003, and did not, after that encounter, become analytical, then it seems that I have been, for the last six years or so, a bullshit Marxist. Hence, the alternate title for this talk. Since I nonetheless find much to respect and value in Cohen’s work on Marx, I want to press his definition of and commitment to analysis, and to see whether or not it makes sense to proclaim myself – not to mention numerous others who are similarly situated vis-à-vis Marxist theory – a post-analytical Marxist.
To that end, I want to do what Cohen claimed in 2000 he and his fellow analytical Marxists never did, put analysis in question. The first task will be to get clear on just what analysis is, on Cohen’s account. It has both a broad and a narrow sense, and I will proceed to challenge each sense, beginning with the narrow one, anti-holism. I will argue that, in the narrow sense defined by Cohen, analytical Marxism is not actually analytical. That is, it is committed to a certain sort of holism. Then I’ll move on to discuss the broader sense of analysis, which is opposed to “dialectical reasoning,” something Cohen does not actually think exists. In other words, Cohen takes analysis to be identical to reasoning as such. I think there are good reasons for resisting this view, and that Marxists in particular ought to be wary of it. In order to show why this is so, I will enter into the realm of Cohen’s Marx interpretation. I think that Cohen makes a number of observations about Marx and Marx’s project that can actually be read as motivations for a post-analytical Marxism.
On Cohen’s telling, “The fateful operation that created analytical Marxism was the rejection of the claim that Marxism possesses valuable intellectual methods of its own” (KMTH, xvii). This rejection can be seen to take two forms, one of which is necessary and sufficient to constitute analytical Marxism, the other of which is a not entirely necessary augmentation of the first, and therefore constitutes something like a hyper-analytical Marxism. One can be an analytical Marxist without being hyper-analytical. The necessary form of the rejection is the rejection of “so-called ‘dialectical’ thinking,” and I will turn to this later. The more exacting sense in which “much” work in analytical Marxism rejects any special Marxist methodology is the sense in which analytical thinking “is opposed to what might be called ‘holistic’ thinking” (KMTH, xvii). My contention in this section is that Cohen’s work does not meet this more exacting sense of (hyper-)analyticity. However, contra Elster, this is not because of Cohen’s reliance upon functional explanation, for I shall further argue that Elster’s rejection of functional explanation is at least as “holistic” as is Cohen’s acceptance of such explanation. Indeed, I think anti-holism simply fails as a criterion of (hyper-)analyticity.
To begin, Cohen defines the narrow sense in which analytical Marxism is analytical as “its disposition to explain molar phenomena by reference to the micro-constituents and micro-mechanisms that respectively compose the entities and underlie the processes which occur at a grosser level of resolution” (KMTH, xxiii). I want to call attention to the three metaphors intermingling in this statement. Of course, there is 1) the metaphor of parts and whole (molar, constituents, compose) that naturally attends a discussion of holism, but there are also, alongside this, 2) a metaphor of comparative size (micro-, grosser), and 3) a metaphor of structural levels (underlie). I think the metaphor of structural levels goes along with most any effort at explanation, and that it meshes unproblematically with either of the other two metaphors. That which underlies has an explanatory primacy with regard to whatever it underlies. However, I think the inclusion of metaphor 2, the metaphor of comparative size, suggests more than it illuminates. It introduces the notion that explanatory primacy belongs with the smallest constituent parts and/or mechanisms. I think there are good reasons to resist this suggestion, and that purging this metaphor from the canon of analytical Marxism reveals that whatever anti-holism there is in its program, it is only an anti-one-particular-sort-of-holism.
Taken seriously, the metaphor of comparative size would suggest a program of thoroughgoing physical reductionism, but of an extremely old-fashioned, Democritean kind. That is, it suggests that the being, characteristics, and activity of everything big and complex is to be explained by the being, characteristics, and activity of basic, “atomic” constituents. To explain something, on this model, just is to show how it is an epiphenomenon of the atoms. Regardless of how widespread or defensible such a model might be in the sciences more generally, it seems out of place here, for two reasons. First, it suggests that only a reduction of social science to physical science would satisfy the (hyper-)analytical Marxists, but nothing aside from the metaphor of comparative size indicates that they held this view. Secondly – and this is the crux of the matter for me – it suggests an essential and direct proportion between smallness and simplicity. This suggestion illicitly supports Elster’s contention that his methodological individualism is an advance over Cohen’s functional explanations because, among other things, although the individual agent remains to some extent a black box, it is a smaller black box than the economic structure. This notion of a smaller black box expresses in a condensed form everything I find worthy of suspicion in the metaphor of comparative size, for it assumes what it is supposed to support, that a decrease in the size of the explananda is a decrease in the amount of explaining left to do. In the social scientific context, the claim that social structures and processes are, ultimately, to be explained in terms of the beliefs, desires, and actions of individual persons cannot rest on the observation that persons are smaller than societies.
Alongside this misleading metaphor of comparative size, Cohen also offers up the real reason for preferring that explanations of social phenomena be couched in terms of individual agents. Comparing current social science to thermodynamics before statistical mechanics – that is, to a descriptive science that as yet lacks a mechanism for explaining the regularities it observes and predicts – he avers that “to claim that capitalism must break down and give way to socialism is not yet to show how the behaviors of individuals lead to this result. And nothing else leads to that result, since behaviors of individuals are always where the action is, in the final analysis” (KMTH, xxiv; my emphasis). I take Cohen’s pun seriously. It suggests that individual agents are the loci of social scientific explanation because they are agents, not because they are individuals. Society is not a thing composed of lots of smaller things; it is a structured complex of material and social relations, a complex that is enacted by the agents it relates. (Hyper-)analytical Marxism is invested in methodological individualism, I contend, because it is committed to the explanatory priority of agency over relations, the notion that how we relate to one another is determined by what we do.
I think there is quite a bit of – admittedly somewhat circumstantial – evidence that Cohen maintained the primacy of agency over relations. I’ll only point to his endorsement of Marx’s claim that “material relations are only the necessary forms in which [human beings’] material and individual activity is realized.” Note that, since relations are the necessary forms in which individual activity is realized, the explanatory primacy of agency over relations does not imply any commitment to the existence of agency outside or otherwise independent of relations. Just as, according to Cohen, every base needs a superstructure, agency needs relations, and gets the relations it needs because it needs them.
Now, whatever one might say for or against this thesis that agency has explanatory primacy over relations, it is not, I think, a consequence or expression of anti-holism, that is, of any sort of thoroughgoing program of reductionism. It does coincide with a denial that society is an agent, a denial which implies that society is not a whole of a certain sort. But this goes hand in hand with identifying an irreducible whole in the individual human agent. Individual behavior is where the action is; action is not to be found at either a grosser or a more microscopic level of resolution.
On this reading of Cohen’s work, it seems to me that bullshit Marxists actually have a claim to be more (hyper-)analytical than the (hyper-)analytical Marxists, since post-Althusserian continental Marxism has been just as opposed to any holism of the individual subject as it has been to any holism of society or of class. In short, if (hyper-)analytical Marxists seek to explain every social process and structure by reference to the beliefs, desires, and actions of individual human agents, continental Marxists seek to explain the constitution of individual human agents, together with their beliefs and desires, out of the material relations in which they always already find themselves. Both seem to me to be valid and important avenues of research. It’s not clear to me that analyticity or anti-holism is the most appropriate criterion for differentiating them.
By way of turning to the broader or more basic sense in which analytical Marxism is supposed to be analytical, I want to advance the notion that, in fact, bullshit Marxism has been, for at least fifty years, busy thinking through certain problems that Cohen’s re-articulation of Marx indicates but does not really explore. There are two problems that particularly interest me: the problem of theory and practice indicated by Cohen’s concept of fettering, and the problem of ontology indicated by Cohen’s concept of power. In both of these cases, Cohen’s theoretical return to the historical materialism of the Second and Third Internationals is helpful primarily insofar as it clarifies why continental Marxism moved away from that paradigm in the first place. That is, I want to read Cohen as motivating theoretically the move to bullshit Marxism that was originally motivated practically by the rise of Stalinism.
In Cohen’s construal of historical materialism, 1) the forces of production underlie and explain 2) the relations of production – i.e., the economic structure – which, in turn, underlie and explain 3) legal and political institutions (the superstructure). The forces of production are natural or material powers, the relations of production are social powers, and the legal and political institutions are juridical rights and duties. One of Cohen’s most important contributions to Marx scholarship is his rigorous attention to the difference between natural and social powers, on the one hand, and to the difference between social powers and juridical rights and duties, on the other.
According to Cohen’s historical materialism, the underlying trajectory of human history is the expansion of the natural power of humanity. That is, despite all local and temporary setbacks, there is a tendency for the forces of production to grow (in Cohen’s parlance, the forces develop). However, since the productive forces produce use-values, of which there is no common measure, the development of the productive forces is measured by “the amount of the day which remains after the laboring time necessary to maintain the producers has been subtracted” (KMTH, 61). The core of historical materialism, then, is this historical tendency of surplus material power to grow.
Historical materialism is scientific because it is explanatory, because it maintains that the historical development of the productive forces explains the form taken by the dominant relations of production, that is, by the economic structure of society. Economic structures that allow or encourage this development in natural power win out over those economic structures that hamper this development. Hence, any post-capitalist economic structure must be better able to produce surplus material power than is capitalism. As Foucault said to Chomsky, the proletariat will overthrow the bourgeoisie because it wants to and can, not because it is just. (I’ll return to this.) Hence, also, social revolutions can be premature, just in the case where the forces of production are not developed enough to support the relations of production introduced by the revolutionaries. In other words, social or economic power – i.e., the dominant economic structure – is functional for natural power, and takes the form that is compatible with the current level of development of the latter.
Except, that is, when it is not and does not. For it is also true that, as Cohen puts it, “the production relations are capable of fettering, that is, restricting the use and development of the productive forces” (KMTH, 41). During any period of time when fettering is occurring, “[d]ysfunctional relations persist” (KMTH, 161). Cohen generally ignores this dysfunctional state of affairs – indeed he announces that he will generally ignore it (KMTH, 161) – but it seems to pose a significant difficulty for his construal of historical materialism, at least in that theory’s practical comportment. (That the theory is supposed by him to be practical is evident in his claim that “scientific socialism” is “the study of the nature of, and the route to, socialism, using the most advanced resources of social science” [KMTH, xxvii].) For, during periods of fettering, the relations of production are not explained by their functionality for the forces of production on hand, since they are dysfunctional given those forces. Hence, during periods of fettering – and it must be noted that Cohen thought that at least the US was in such a period (KMTH, 297) – the perseverance of the economic structure, if it is to be explained at all, must be explained by something other than the material powers, as these are defined by Cohen. The material powers at hand are incompatible with the economic structure at hand, and call for a new economic structure, and yet the economic structure persists.
This state of affairs – and, again, Cohen thought it to obtain presently, at least in the US – seems to throw us back on the other explanation Marx offers for historical change, the class struggle. Cohen claims that this cannot be explanatory in the final instance because reference to the class struggle does not tell us “why the strong were strong and the weak weak” (KMTH, 148). This is certainly fair. Nonetheless, once you admit that fettering occurs, you must also admit that while fettering is occurring – and hence during revolutionary periods – the status of the class struggle is irreducible to the level of development of the forces of production, and so there must be other forces, other powers, that enter into the equation. Nothing but contesting the current relations of production will tell whether those social powers will fall. In short, the very concept of fettering indicates that taking the epochal view of the determination of economic structure by forces of production is inappropriate for the political and social struggle to transform that economic structure. Waiting for the forces of production to burst their fetters is like waiting for Godot.
I don’t think I’m saying anything that is contrary to Cohen’s own declarations. Nonetheless, I do think that Cohen’s historical materialism cries out for what it does not itself supply, and that his thumbing his nose at appeals to “determination in the last instance” and other such “opaque ideas” is no help. The functional explanation of social power by natural power breaks down right where those who are concerned not to interpret the world but to change it begin their work of thinking and acting. I think this has far reaching consequences, to which I will return after a closer consideration of Cohen’s understanding of power.
Cohen rather famously attempts to purge all legalistic language from the definitions of the various relations of production, and hence from the definition of the economic structure, replacing it with “a distinct language of effective powers” (KMTH, 216). The point of so doing is to articulate “a method of conceiving the economic structure which excludes from it the legal, moral, and political relations of men” (KMTH, 235). This conception is in the service of a “general explanatory thesis […] that given property relations have the character they do because of the production relations property relations with that character support” (KMTH, 226). If production relations explain property relations (and other legal, moral, and political categories), then production relations cannot include or be defined in terms of property relations (or other legal, moral, or political categories).
Instead, production relations are defined in terms of power, of being “able to φ, where ‘able’ is non-normative” (KMTH, 220). Cohen seems to be at pains, however, to avoid any sort of ontology or physics of these social powers. Hence, he claims that his “definition of production relations does not stipulate how the powers they enfold are obtained or sustained” (KMTH, 223), and that “your power is what you are able to do, regardless of what makes you able to do it” (KMTH, 234n1). And yet I cannot see why he should be so reticent, since it would seem that some account of the sources and concrete forms of social power would go a long way towards both a) clarifying the distinction between natural and social powers that undergirds the distinction between forces and relations of production, and b) pointing the way to some explanation for the persistence of dysfunctional relations of production.
This is the path taken by continental Marxists when they turned to analyses of the culture industry, of hegemony, of the reproduction of labor-power in ideological state apparatuses, of technologies of discipline and security, of the organization of labor in Taylorism, Fordism, and post-Fordism, and so forth. Many of these investigations also encompassed returns to early modern philosophy – to Machiavelli, Hobbes, and Spinoza, especially – in order to flesh out an ontology of modern social power. The point was to explain the sorts of phenomena that Cohen merely points to: that, in modern societies, “rights screen powers” (KMTH, 225); that “might [i.e., power] frequently needs right in order to operate or even to be constituted” (KMTH, 231); that, hence, social powers “are sustained by the law, morality, and the state” (KMTH, 235); that “[c]apitalist society propagates and reinforces ignorance of power” (KMTH, 244); etc. It seems quite obvious to me that, while the historical materialism defended by Cohen brings these phenomena into view, nothing in Cohen’s definitions of natural or social powers helps us explain them.
Here, as in the case of fettering, Cohen exhibits a frustrating tendency to fall back on something like common sense at just the point where his explanatory project ceases to explain. It’s just obvious that legal systems and codes of morality help stabilize relations that would otherwise rest solely on force. But, to borrow a line of analysis Cohen uses to discuss fetishism, laws and norms don’t stabilize social powers autonomously. Stabilizing power does not inhere in laws and norms, but is delegated to them by other forces and powers. What is missing in Cohen, then, is any effort to naturalize law, morality, and the state, in the way in which Marx tried to naturalize the commodity and capital. That is, what is missing is the dual attempt to a) trace the power of law, morality, and the state back to natural powers, and b) trace the origin of the separation of these powers from their natural base back to the operations of that base. If the state’s power is really rooted in coercion, to make up an example, why does coercion take the form of the state, rather than being directly and transparently expressed as coercion? Cohen quotes Marx’s Grundrisse: “The bourgeois economists have a vague notion that it is better to carry on production under the modern police than it was, e.g., under club-law. They forget that club-law is also law, and that the right of the stronger continues to exist in other forms even under their ‘government of law” (Grundrisse, 88; KMTH, 225). What I am protesting in Cohen is his inattention to the problem of explaining the origin and operation of these other forms of the right of the stronger, and his corresponding blindness to what is actually going on in the continental Marxists he ignores and or dismisses.
c. the ‘necessity’ of analysis
It has been noted that both Althusser’s and Cohen’s work – the two landmarks of post-war Marxist philosophy – coincided with the greatest period of crisis in the Marxist labor movement and revolutionary program (excepting only the collapse of the Soviet system, if this latter can be separated from the crisis of the ‘60s and ‘70s). And yet, as Grahame Lock notes:
Cohen's book shows no signs of having been written in this context of crisis.This appearance seems quite essential to Cohen’s work, and seems to me to be of a piece with what he refers to as the broad sense in which analytical Marxism is analytical. Cohen was committed to a set of values (“socialist values”), about which we are bound to hear more today, and he was committed to analysis as the form of reason as such. Marxism, as a movement, as a tradition, as a set of problems and texts, as a set of critical theses about capitalism – this was secondary. But if analytical Marxism is necessarily committed to analysis, if “it is always Marxism, not analysis, that is in question” (KMTH , xxiv), then it seems that one cannot be committed to Marxism without being a bullshit Marxist. Cohen compares this to turning Marxism into a religion, but doesn’t say what differentiates a commitment to a tradition or movement from a commitment to a set of values, or why the former is religious, while the latter is not. (Given his late appreciation of Christianity, perhaps there is no difference.)
Rather, it was written as, and has been received as, a more or less purely
philosophical treatise, in a quite traditional sense of this phrase: as a work
to be read and assessed in abstraction from any consideration of its possible
roots in or impact on the political and ideological situation. (“Louis Althusser
and G. A. Cohen: A Confrontation,” Economy and Society 17:4 (1988), 501)
Perhaps it all comes back to the title of Cohen’s book. Cohen set out to clarify and defend a theory of history, but nowhere does he clarify or defend his decision to read Marx primarily as a theorist of history. Cohen seems to suppose that the socialist movement finds its origin and justification in a theory of history, rather than the other way around. In actual history, the theory of history Cohen defends seems to have served primarily as a motivating myth, a means by which workers could transcend the opportunistic and purely economic struggle for better wages, safer workplaces, etc., and become political subjects with a world to win. (Lars Lih’s work on Lenin’s relation to German social democracy is illuminating on this point.) In other words, the theory of history was a weapon in the sense Marx meant when he wrote that “theory also becomes a material force as soon as it has gripped the masses.” This notion that the struggle is primary, and that theory is something that emerges from and finds its end in the struggle, seems alien to Cohen’s work. And that is why I don’t lose a lot of sleep over the fact that I’m a bullshit Marxist. Nonetheless, precisely because Cohen recapitulates a certain history in the realm of theory, I would only second the assessment of Althusser's friend and student, Dominique Lecourt, who, in reviewing Cohen’s book, wrote that it is “one of those books on Marx, which are so rare, whose grandeur consists in their forcing the reader to rethink the whole structure of Marxism, from the philosophical foundations up, even if he goes on to draw conclusions diametrically opposed to those which the author proposes.”