It was hard to read Ellen Wood’s article ‘Rational Choice Marxism: Is the Game Worth the Candle?’ without mixed feelings.footnote1 The general thrust of her critique is undoubtedly correct: in the hands of Jon Elster, John Roemer, Adam Przeworski et al., the attempt to reinterpret historical materialism along methodological-individualist lines has deprived the theory of much of its specificity and substance. She is also right to set Rational Choice Marxism (rcm) alongside post-structuralism as the two main intellectual tendencies which, in the past decade or so, have provided the reaction against Marxism with a ‘left’ guise. Wood sought, however, not merely to demolish rcm, but to do so in part by demonstrating the existence of another, better version of historical materialism. And here the difficulties begin. For while I share most of her criticisms of rcm (indeed, I’ve made quite a few of them myselffootnote2), her own account of what is distinctive to, and worth defending in, Marxism seems to me seriously inadequate.

This account emerges most clearly where Wood discusses putative candidates for a rcm theory of history (pp. 59–75). She regards it as a tacit acknowledgement of the inadequacy of rcm theories of exploitation and class such as that constructed by Roemer that they should require supplementation by some separate account of the sources of historical change. Two such accounts are considered by Roemer in his book Free to Lose. One, G.A. Cohen’s restatement of orthodox historical materialism, is indeed compatible with Roemer’s static models; but the reason why this is so, namely that the development of the productive forces provides an ‘exogenous cause’ of social change, is indicative of the sense in which Cohen’s is not a proper theory of history, since it invokes to explain social transformations, not the properties internal to the mode of production in question, but rather a ‘transhistorical rationality’ which leads human beings in conditions of scarcity to improve their methods of labour (pp. 69–71). Wood looks with much more favour on the other candidate, provided by the work of Robert Brenner, but argues both that his account of the transition from feudalism to capitalism is inconsistent with the idea of any ‘historical necessity for less productive “economic structures” to be followed by more productive ones’, and that it involves a theory of history whose ‘focus’ is ‘on the specificity of every mode of production, its endogenous logic of process, its own “laws of motion”, its characteristic crises—to use Brenner’s formula, its own rules of reproduction’, in both respects sitting ill with rcm’s tendency to rely on explanations derived from transhistorical features of human societies (pp. 68, 70).

This is by no means the first time that Wood has used Brenner’s work to distinguish her alternative reading of historical materialism from Cohen’s. Indeed, at one point she adopted for this reading the label given to Brenner’s work by one of his Marxist critics, Guy Bois, namely ‘political Marxism’. Bois elaborates: ‘It amounts to a voluntarist vision of history in which the class struggle is divorced from all objective contingencies, and, in the first place, from such laws of development as may be peculiar to a specific mode of production.’footnote3 Wood rejects the charge of voluntarism, but takes Marx himself to say that ‘capitalism is unique in its drive to revolutionize the productive forces, while other modes of production have tended to conserve existing forces’ (p. 70 n. 47). The ‘explanatory force’ of the development of the productive forces is subject to ‘severe limits’; to understand social change we must look instead ‘in the direction of class struggle as the operative principle of historical movement’.footnote4 Thus the main sense in which historical explanation draws on features intrinsic to particular social systems seems to be that it identifies the specific form of surplus-extraction, thereby providing the context of the class struggles which provide the motor of change; as, for example, Brenner does when he argues that the breakthrough to agrarian capitalism in England depended on the specific outcome there of the Europe-wide struggles between lord and peasant at the end of the Middle Ages.footnote5

This is a version of Marxism that it is hard not to have great reservations about. In part, these reservations stem from difficulties specific to Brenner’s account of the rise of agrarian capitalism. His writing has undoubtedly provided a valuable corrective to those accounts of the transition to capitalism which, from Pirenne and Sweezy to Braudel and Wallerstein, have accorded prime importance to the expansion of the world market.footnote6 Brenner is, moreover, right to stress the crucial role played by the emergence in England of a distinctively capitalist agriculture, especially in making possible that country’s establishment of first military and then industrial primacy over its rivals—particularly France—after 1689.footnote7 Nevertheless, Brenner’s exclusive focus on agrarian capitalism has encouraged, perhaps contrary to his own intentions, some wildly one-sided readings of the process of capitalist development. Probably the most lamentable example is George Comninel’s Rethinking the French Revolution (based, incidentally, on a dissertation supervised by Wood), which argues that, since there was no equivalent in pre-revolutionary rural France of the capital–wage-labour relations increasingly prevalent in the contemporary English countryside, ‘there simply were no capitalist relations—no appropriation of surplus-value, as opposed to commercial profittaking—that can be attributed to the [French] bourgeoisie.’footnote8 What such arguments leave out of account is the extent to which early modern merchant capitalism, though still rooted in feudal social relations, provided a framework for the emergence of what Lenin called ‘transitional forms’ through which capital began to acquire control over production.footnote9 One such form was what Robin Blackburn calls the ‘systemic slavery’ of the British and French West Indies, and later Cuba, Brazil and the American South: the large-scale exploitation of slave labour, producing for the world market either mass-consumption goods (sugar) or industrial inputs (cotton).footnote10 ‘Proto-industrialization’—the spread of rural industry, usually producing textiles, often on the basis of the putting-out system—represented another form in which labour was partially subsumed under capital, and arguably a more decisive one, since the abolition of slavery led often to a fragmentation of productive units, while the limitations of the putting-out system tended to drive capitalists to centralize the labour process in the factory.footnote11 The development of agrarian capitalism, on which Brenner and his followers concentrate, was part of a much broader process through which bourgeois social relations progressively undermined the old feudal order.

But it is not simply doubts about the historical claims advanced by Brenner (or, perhaps better, by those influenced by him) which give one pause when confronted with Wood’s employment of his work to construct ‘political Marxism’. Historical materialism explains social transformations as the outcome of two mechanisms: first, the structural contradictions that arise between the development of the productive forces and the prevailing production relations; and secondly, and only in the context of the socio-economic crises generated by these contradictions, the class struggle. Capital does not only elucidate the conditions and forms of the extraction of surplus-value within the production process; it also locates capitalism’s chronic liability to recurrent economic crises in the tendency of the rate of profit to fall—the form of the contradiction between the forces and relations of production specific to that mode of production. Some of the greatest recent triumphs of Marxist historiography have been to delineate more precisely the nature of this contradiction in pre-capitalist modes. As Perry Anderson points out, G.E.M. de Ste Croix’s explanation of the decline of classical antiquity is an instance of the kind of ‘systemic contradiction’ that occurs ‘when the forces and relations of production enter into decisive contradiction with each other’.footnote12 Similarly, there is little doubt that, despite their disagreements, Brenner and Bois have greatly advanced our understanding of the form taken by the similar contradiction responsible for the late-medieval crisis of feudalism.footnote13

The trouble is that Wood is plainly hostile to giving any explanatory weight to structural contradictions between the forces and relations of production. ‘The proposition that history is propelled forward by the inevitable contradictions between forces and relations of production, contradictions that emerge as developing productive forces come against the “fetters” imposed by production relations’ is, she says, ‘vacuous’.footnote14 Wood also suggests that Marx’s attachment to this proposition represented ‘an undeveloped phase of Marx’s work, still uncritically bound to classical bourgeois thought’ (p. 69), a claim developed at great length by Comninel, who argues, implausibly, that the development of the productive forces is a central theme only of The German Ideology, which must therefore be consigned to the flames as a piece of ‘liberal materialist ideology’, and plays no part in Capital.footnote15 But, once structural contradictions between the forces and relations of production have been excised from historical materialism, it is not clear that what is left amounts to a theory of social transformation in any real sense. Class struggle alone cannot account for the transition from one mode of production to another. Open or concealed conflict between exploiter and exploited is an endemic feature of class societies. But it assumes a greater intensity in periods of what Gramsci called ‘organic crisis’, where the very viability of the prevailing social system is placed in question.footnote16 Marxism can only provide the theory of history it purports to offer if it can explain the emergence of such crises. To do so in terms of the class struggle itself, as some contemporary versions of Marxist economic theory (for example, the ‘capital logic’ school and regulation theory) tend, is not merely to commit a vicious circularity, in which intensifying class struggle explains intensifying class struggle; it is also to reduce historical materialism to a voluntarist social theory, where the motor of change is the clash of hostile class wills. Andrew Levine argues that versions of Marxism ‘that do not theorize transitions, that fail to postulate a direction of change between epochal structures’, represent ‘not a materialist theory of history’ but ‘a materialist sociology’.footnote17 Wood’s ‘political Marxism’ is little more than a sociology of domination. It is good that, unlike other such sociologies, chiefly of Weberian provenance, Wood’s attaches primacy to class exploitation, but it is very far from being enough.

The conflict between the forces and relations of production can only serve as a mechanism of social change if the productive forces tend to develop and thereby become incompatible with existing relations. It is one of the great merits of Cohen’s Karl Marx’s Theory of History to have so forcefully redirected attention to this simple fact. Wood, when seeking to evade its implications, resorts to Brenner’s argument that the ‘rules of reproduction’ in pre-capitalist societies, in particular the fact that both producers and exploiters have direct, non-market access to the means of subsistence, rules out the intensive development of the productive forces, which becomes possible only when economic agents’ dependence on commodity production forces them to compete and therefore to innovate.footnote18 But even if we readily grant that capitalism is incomparably more dynamic a mode of production than its predecessors, how far are we to take Brenner’s argument? He surely isn’t saying that there was no development of the productive forces under feudalism (the main pre-capitalist mode with which he concerns himself). Apart from being plainly false, such a claim conjures up a vista of endless stagnation unlikely to issue in any new social form. Brenner’s argument is better taken as setting limits to the development of the productive forces in pre-capitalist societies, and therefore requires supplementation with an account of how such societies nevertheless permit a degree of technological progress. The most obvious candidate for such an account, Cohen’s Primacy Thesis, unfortunately won’t do, for well-known reasons—its postulation of a general human interest in the development of the productive forces, its reliance on functional explanations, and its requirement that social revolutions are inevitable.footnote19 But one can imagine some elements of a less vulnerable account. One is what Erik Olin Wright calls the ‘weak impulse’ for the productive forces to develop arising from, inter alia, the fact that ‘under conditions in which increases in labour productivity have the consequence of reducing the toil of the direct producers, direct producers will in general have interests in developing the forces of production.’footnote20 Another is an analysis of the mechanisms which permit specific pre-capitalist modes of production to achieve productive progress over their predecessors. One weakness of Brenner’s discussion of pre-capitalist societies is his failure to differentiate between them, so that slave and feudal modes of production are treated as representing the same level of development, which, once again, ill accords with the historical record.footnote21