An Introduction to Karl Marx 
by Jon Elster.
Cambridge, 220 pp., £5.95, October 1986, 0 521 32922 1
Show More
Making sense of Marx 
by Jon Elster.
Cambridge, 556 pp., £32.50, May 1985, 0 521 22896 4
Show More
Analytical Marxism 
edited by John Roemer.
Cambridge, 321 pp., £27.50, March 1986, 0 521 30025 8
Show More
Show More

The relationship between philosophy and Marxism has always been an awkward one. ‘Philosophy stands to the study of the real world in the same relationship as masturbation stands to real sexual love,’ said Marx himself. Was this a dismissal of all forms of philosophy, or only of the overblown Idealism of Hegel? Would he have been equally dismissive of pragmatism or empiricism; would Pierce or Mill have received the same short shrift? Marx was unwilling to waste time on such questions. The philosophical and methodological remarks scattered through his major works are scrappy, undeveloped and not entirely consistent; they take a poor second place to what he conceived of as an empirical inquiry into the logic of capitalist society and the sociology and politics of its supersession, and they leave wide open the question of what positive role he saw for philosophy.

Engels was more ready both to assert that Marxism had abolished philosophy, which from now on could only be a ‘resumé of positive science’, and alternatively (or inconsistently) that it possessed a philosophy of its own. This was the materialist dialectic which revealed some general laws permeating both the natural and the social world, such as the famous ‘law of the transformation of quantity into quality’. Lenin, Kautsky and a host of successors went further. Either they felt that Marxism’s status as science had to be philosophically underpinned, or they felt the need to supplement the brutal determinism of ‘scientific socialism’ with a philosophical demonstration that the socialist future was not merely inevitable but desirable.

Until very recently, neither activity was couched in an idiom which appealed to orthodox analytical philosophers. The ‘philosophical’ Marxism of Lukacs and Gramsci rested on revisions of Hegel and Croce; it is debatable how Marxist the Frankfurt School have ever been, but their debts to Hegel and dislike of empiricism leave no room for argument. The French contempt for Anglo-Saxon empiricism meant that philosophical Marxism in France divided its allegiances between Existentialism and Structuralism. The Anglo-Saxons were not so much contemptuous as puzzled: whatever it was that was going on in Paris didn’t look much like philosophy. The cautious ‘ordinary language’ philosophers of the Fifties and early Sixties found Sartre ‘interesting’ in the same way they might have found an avant-garde novelist interesting; the few who thought him interesting as a philosopher and moralist were then hard put to it to see how he supposed himself to be a Marxist. The rebarbative and authoritarian prose in which Althusser announced that Marx had been a structuralist and anti-humanist aroused in most Anglo-American philosophers a puzzled anxiety at best.

It must be remembered that this mutual incomprehension was only true of the ‘professional’ philosophical reaction; it reveals as much about the diversity of post-war philosophy as about the gulf between Marxism and analytical philosophy. Modern linguists, literary theorists, some dissident social scientists, were much more open to these European influences; that they made so little headway elsewhere merely proved what an ‘establishment’ enterprise academic philosophy was. The New Left Review from its earliest days attempted to break down the barbarian indifference with which British universities regarded the European Left. But even in the pages of the NLR it was long taken for granted that Marxists and analytical philosophers could hardly be on speaking terms. If Marxism was a science, it was a science of a new kind whose insights were unamenable to the empiricist obsession with verification and falsification. Moreover, it was an article of Marxist faith that Marxism adjusted both its findings and its methods to ‘the real movement of things’. Marxism was inherently progressive, both intellectually and politically. Analytical philosophy, whose model of intellectual virtue was physical science, could make no sense of the dialectic, and had nothing to say about a discipline which focused on the novel and the fluid. Its stance was intrinsically unprogressive, concerned only to rationalise existing understandings of the world.

Karl Popper’s violent assault on Hegel and Marx was one more contribution to mutual animosity. Popper’s philosophy of science is in fact Kantian, not empiricist or positivist: but this has always mattered much more to him than to his critics. It was sufficient for the opponents of analytical philosophy that The Open Society and its Enemies and The Poverty of Historicism were all of a piece with his ‘falsificationist’ philosophy of science and that both savaged Marxism as intellectually, morally and politically disastrous. Popper argued that absurd scientific pretensions led inexorably to the closed society and the totalitarian state. Marx and Lenin were the heirs of Plato’s philosopher-kings, demanding the right to build utopia on the wreckage of bourgeois liberalism. Marxists who dismissed Popper’s attacks as far-fetched or misconceived concluded that Marx could not expect a fair hearing from Anglo-American philosophy.

‘Analytical Marxism’ answers that charge. Its essential purpose is to show that much – or most, or some – of what is central to Marxism stands up to critical inspection. This isn’t simply a matter of translating an esoteric knowledge into empiricist terms; its adherents have no time for Parisian high fashion or Frankfurtian obscurity. What is worth learning from Marx can be stated clearly, and defended in the plainest prose. Lest that give the impression that Jerry Cohen, Jon Elster and John Roemer see themselves as Marxists first and philosophers second, it must at once be said that their work is remarkable for the unflinching way in which they throw out whatever won’t pass a dispassionate scrutiny. Elster’s Making sense of Marx has been attacked by its Marxist critics as an exercise in showing that Marx mostly doesn’t make sense. Cohen’s Karl Marx’s Theory of History is subtitled ‘A Defence’,1 but even when it appeared nearly ten years ago, he was committed only to the bare plausibility of Marx’s claim that social change is to be explained by its function or purpose and, more specifically, that it occurs ‘in order to’ promote the development of the forces of production. Since then, he has steadily given ground to his critics – among the most tenacious of whom has been Jon Elster. Cohen has largely stood by his interpretative claim that Marx explained social change in functional terms, but he now agrees with Elster that such explanations are indefensible. If he is a Marxist at all, he isn’t by his own standards an orthodox one.

Similarly, Roemer’s scrutiny of Marx’s theory of exploitation has taken him far beyond Marx’s own account of the subject; his essay ‘Should Marxists be interested in exploitation?’ (reprinted in Analytical Marxism) draws on J.S. Mill and John Rawls as much as on Marx himself. The result is to downgrade Marx’s obsession with exploitation. In Marx’s own work, exploitation is a key concept, explanatory and evaluative. All societies where private property exists are built on forced labour; in all class-divided societies the surplus product is ausgepumpt from the producers to the exploiters. Marx created the theory of surplus value to explain how it was that forced labour existed under capitalism, and how intangible value, rather than corn or cattle, was ausgepumpt from the industrial worker. Roemer denies that exploitation is a ‘fundamental’ term of explanation or appraisal; if what we mean by ‘exploitation’ is ‘forced labour’, then a denial of freedom is involved and freedom is the fundamental notion; if we mean ‘unrequited labour’, then it is a denial of justice, and justice is the fundamental notion. It is only the superstitious Marxist who thinks that ‘exploitation’ is a fundamental term of social analysis and moral appraisal. We deplore exploitation because we approve of freedom and justice. Marx was inclined to dismiss talk of justice as a ‘bourgeois ambush’ and ‘moralising filth’, so it is easy to see that analytical Marxists are far from engaging in hagiological exercises.

Because Marx’s work was in intention a seamless whole in which historical, economic and political phenomena were to be explained as interrelated aspects of the development of the social organism, it is difficult to split off particular topics to illustrate the distinctive approach of the analytical Marxists. Indeed, many interpreters of Marx would think that the very attempt distorts Marx’s achievement – one more point on which they disagree with the analytical Marxists, who think that much of Marx is intelligible and right, much else empirically false or worse.

Still, one might begin from the scientific aspirations of Marx and Engels themselves, and ask two questions: in what does the ‘scientific’ standing of historical materialism consist, and how, if at all, does ‘scientific’ socialism come to supersede ‘utopian’ or ‘moralising’ socialism? It is plausible to suppose that Marx, adopting Hegel’s ‘holistic’ view of social phenomena, and sharing Hegel’s view that history was essentially the history of freedom, turned from philosophy to science by substituting the driving force of productive development for ‘the cunning of reason’ and the struggle of Geist to self-knowledge and self-realisation. Cohen, defending a ‘technological determinist’ interpretation of Marx, takes this line, referring to what he describes as ‘the identity of structure across diversity of content’ which links Hegel and Marx. For Cohen, the superiority of Marx over Hegel lies in the fact that Hegel was only a philosopher of history. In Cohen’s eyes, Hegel’s explanatory categories – ‘culture’ and ‘consciousness’ – are duds; this is not the case with Marx’s. ‘The concepts of productive power and economic structure,’ says Cohen, ‘(unlike those of consciousness and culture) do not serve only to express a vision. They also assert their candidacy as the leading concepts in a theory of history, a theory to the extent that history admits of theoretical treatment, which is neither entirely nor not at all.’

Cohen’s interpretation of Marx’s theory argued that Marx explained social change functionally – institutions change so as to facilitate an increase in the productive forces available. To argue this, Cohen had to rescue functional explanation from two kinds of criticism. One was that of Marx himself, for Marx was adamant that he had abandoned the teleological perspective of Hegel. The other was that of orthodox analytical philosophy of science, which has long held that functional explanation is suspect. ‘Cows have tails so as to swish off flies’ is, according to the critics, not an explanation at all. Nor is ‘capitalist property relations developed so as to release productive forces hampered by feudalism.’ It is, however, hard to deny that Marx constantly talked in functional terms. For instance, after it emerged that the Ten-Hour Act and the Factory Acts had not damaged profits but had probably increased them, he explained their passage in terms of the benefits they had produced for capital. Elster, indeed, treats this approach as the core of Marx’s method and the main defect of his social theory.

Squaring Marx with Marx is not very problematic; the teleology he was against was Hegelian and transhistorical – in effect, it made the whole of history the plaything of some larger scheme or plan. Since Marx was impressed by the organic quality of social life, he was not in the same way sceptical of functional explanation in detail. Cohen’s larger problem was to defuse the old complaint that functional explanation requires ‘backward causation’ – the good the tail is going to do the cow brings tails into existence, the benefits capitalist property relations are going to confer on the forces of production bring capitalism into existence. His account of what he called ‘consequence explanations’ was ingenious, but never quite persuasive. Conseqence explanations take the form ‘if A will have consequence X, then A will occur,’ or ‘the fact that A will bring about X brings about A.’ This neatly picks up the force of ‘cows have tails because they do them good’ or ‘capitalist production relations exist because they promote the development of productive forces’: it is the fact that tails help cows which explains the existence of tail-wearing cows; it is the fact that the ability to hire free labour at will promotes the development of technology, skill and organisation which explains the existence of the ability to hire free labour at will.

But ‘consequence explanation’ raises the question: how does the fact that A will achieve X bring about A? We can connect the good that tails do and the presence of tails on cows by appealing to natural selection: tail-wearing cows outperform non-tail-wearing cows, so they survive and the rest do not. What about social selection? This is the question that every sceptic has directed against sociological functionalism. Is the mechanism which explains why capitalism succeeded feudalism in Britain a raw natural-selection mechanism? It is not implausible to suggest that ‘modern’ economic life has spread over the globe because of the competitive advantage of well-armed and well-organised societies, but in the case of endogenous change there seems little to be said for such a suggestion. Cohen relies more heavily on the thought that human beings are rational; they choose the best means to achieve their goals. The search for efficiency puts pressure on the productive forces, and so institutions will ‘take’ or not according to their ability to promote such forces.

The trouble with this is twofold: as exegesis, its view of social evolution is more Spencer than Marx, and its basis in individual rationality is explicitly denied by Marx; and as a theory of historical explanation it is dubious. The premise of individual rationality is largely false (for reasons which Marx himself spelled out) and even if it were not, it would yield indeterminate results. Marx was insistent that ‘rationality’ in this means-end sense was forced on the members of capitalist societies, that it was not an essential element in human nature and was not something to rely on as the foundation of historical change. The sense in which it is Marx’s theory of history that is being defended is therefore ambiguous at best. Marx was surely right to eschew an appeal to the doings of rational man, even if this causes difficulties for him. If Greek society was permeated by a taste for the pleasures of politics – as Marx followed Hegel in supposing – rather than the desire to make profit, it is hard to see how the materialist conception of history can be sustained; no doubt some pressure would be put on productive forces by the desire to acquire the resources for political success, but hardly very much, and it might well be a precondition of political success that one took a staunchly ‘un-economic’ attitude to affairs. Hence Engels’s retreat to the thought that economic factors explain social change only ‘in the last instance’ and Marx’s equally feeble claim that where economic factors do not play the main role in social change, it is economic factors which explain why they do not.

While Cohen is vulnerable to the accusation that his reconstructed Marxism isn’t Marx, Elster avoids the accusation that he is defending something other than Marx by simply refusing to defend Marx at all. Elster dismisses functional explanation and argues that we should explain as much as possible by appeal to individualist rational-choice considerations of the kind that all economists employ. The great virtue of Marx so far as Elster is concerned is that he raises the right questions; he does not supply the right answers, and characteristically sets about finding them in quite the wrong way. The right questions are questions about ‘counter-finality’ of the kind Elster tackled in Logic and Society and Ulysses and the Sirens and questions about ‘distorted’ or non-autonomous beliefs and desires of the kind Elster tackled in Sour Grapes.

The right way to tackle the first kind of question is by asking how it is that individually rational people, each pursuing his own good, each well informed about the benefits available, and about everyone else’s behaviour, contrive between them to produce results which leave them all worse-off. Booms and slumps are the obvious examples in a Marxian framework; the ‘anarchy of production’ against which Marx railed at such length is a central feature of capitalism – though in Elster’s universe there is none of Marx’s obsession with production rather than distribution, and Elster’s target is all forms of marketplace anarchy.

Similarly, Marx appealed in a decidedly rough and ready way to processes whereby workers were given the sort of ‘false consciousness’ which would sustain the power and wealth of the owners of capital. Workers believed that ‘a man may do as he chooses with his own’ and thus that the capitalist’s right to hire and fire was an expression of the same freedom as their own right to leave one employer and work for another; or they believed that the capitalist’s profits were a just reward for the risks the capitalist bore, and that their employers were entitled to profits in just the way they were entitled to their wages. Marx more or less took it for granted that such beliefs were imposed on the workers in order that their employers should sleep more easily. Elster rightly insists, against this, that the fact that our beliefs suit somebody else doesn’t explain how we came by them: a proper theory of ideology needs to spell out the causal processes by which we come to hold beliefs which are false, or destructive of our own interests. We can’t just assume that if our beliefs suit somebody else, they must have got us to hold them.

Elster’s contention that Marx was wrong wholesale, as it were – that he was wrong to explain social change holistically and functionally – and that ‘methodological individualism’ of a sort familiar in economics or the theory of games is inescapable, conflates two issues which in his more fastidious moments he rightly separates out. The first is that any claim about why an event or series of events occurs raises the question how, and the only acceptable answer to that question is a causal explanation of a step-by-step kind. Whether it is linking the good that tails do for cows to the existence of tail-wearing cows, or linking the spread of new forms of property to the good they do the forces of production, we have to fill in the causal steps by which the process happens. This is what one might call particularism rather than individualism, however. The second claim is that when we fill in such steps in a sociological or historical explanation we should employ the premises of ‘rational choice’ theories of behaviour. But this claim does not follow from the first, and is a much more suspect claim, not least because of the tendency of rational-choice theories to linger uneasily in the no man’s land between simple falsehood and complex vacuity. The priest celebrating mass follows a series of steps which count as such a celebration; we can explain what he does and why he does it by appealing to these steps and his beliefs about them. Such everyday explanations are individualist causal explanations in the sense that we explain what people do in terms of their desires and beliefs. Only in the most vacuous fashion can we reduce all such explanations to accounts of the rational pursuit of self-interest. Not all ‘individualist’ explanations are ‘rational choice’ explanations. Elster himself says as much.

Elster defends his enthusiasm for rational-choice explanation by observing that with the work of economists and games theorists these are the best-developed form of explanation we have. But that defence is much too like the drunk’s defence of looking for his keys under the street light – he knew the keys weren’t there, but at least there was light; rational-choice theories are very illuminating about behaviour that presents few puzzles, and unilluminating about behaviour that presents more. In any event, even Elster concedes that Marx’s lack of concern with individual reactions would have been warranted if Marx’s belief that social conditions overwhelmingly determined those reactions had been right. It is therefore not Marx’s holism which is intellectually indefensible so much as Marx’s determinism. I think that this is absolutely the right view of the matter: but it sets up an uneasy tension in Making sense of Marx between the root-and-branch condemnation of Marxian methodology to which Elster first commits himself and the much more nuanced and piecemeal criticism of particular claims which follows. It also raises a question to which no very satisfactory answer is forthcoming: if what is most distinctively Marxian is methodologically and factually flawed, and Marx ‘makes sense’ only when translated into the terms of Ulysses and the Sirens and Sour Grapes, why bother with Marx at all?

I take it that the answer is what comes at the end of Making sense of Marx and An Introduction to Karl Marx. The suggestion is that it is Marx’s values which make him worth reading. This is an awkward claim, as I have already suggested: Marx was exceedingly hostile to ‘moralising socialism’, and exceedingly rude to those who demanded justice, in the form of ‘a fair day’s wages for a fair day’s work’, or who talked of anything resembling workers’ ‘rights’. He seems to have had a real loathing of all appeals to morality, a genuine, visceral hatred of what he thought of as the sentimentality of those who thought the workers would get anywhere by asking their employers to be nice to them. As Sorel and others pointed out, Marx’s deterministic picture of social conflict was almost Homeric; workers and owners were doomed by historical fate to fight one another. All the theorist of the proletarian cause could do was to spell out the nature of the struggle and encourage his own side to fight bravely but not brutally. Discussions of justice eternelle and the rest were wholly beside the point.

Oddly, Elster, who is well-read, ingenious and endlessly energetic, does not pursue the question of what Marx could have been up to in simultaneously denouncing capitalism as inhuman, oppressive, brutal and exploitative and denouncing morality as a tissue of illusions.2 Nor is it quite clear why he says he admires Marx’s values, when he is decidedly sceptical of the most distinctive of them. Marx, at any rate as a young man, saw work as the essence of man, and thought the human species destined to create a world which would perfectly reflect all human talents. So central was this ‘creativist’ image that Marx even thought of relatively passive experiences such as listening to music as exercises in what one might call ‘co-creation’. What socialism would realise would be a kind of mutual love, expressed in the creation of things made for our delight, not for the market. Elster’s doubts about this are partly logical doubts about the intelligibility of a society in which altruism totally replaces self-interest, partly doubts about the obsession with creativity – for, after all, some of us must more or less passively provide an audience for others’ creative efforts if these are to succeed. He is more enthusiastic about what he conceives to be Marx’s desire for a just society, one based on the justice of ‘to each according to his needs’. On Elster’s view, the central value of socialism is the substitution of egalitarian, ‘needs-based’ justice for the ‘property rights-based’ justice of capitalism.

John Roemer’s successive explorations of Marx’s theory of exploitation culminate in a similar result.3 Roemer’s work is interesting not least for the technically imaginative way in which he shows that on the most plausible view of what Marx thought exploitation is – receiving less than the labour-value of the commodities we trade – there is no necessary connection between being exploited and being propertyless, and there is at least a theoretical possibility that workers can exploit capitalists. It is this that explains why it is that, as we saw before, exploitation ceases to be a central term, either of analysis or of moral condemnation. The conclusion is damaging to anyone who thinks that it is important to preserve the unity of explanation and evaluation which Marx seemed to aim at: but it does not matter to anyone who thinks that capitalism is simply unjust. Profit isn’t the appropriation of the proceeds of forced labour, but we may still object to the narrowed opportunities of workers who have no resources other than their labour power, and to the gap in income between those who do the actual work and those whose ownership of resources allows them to cream off non-work incomes. What replaces Marx’s denunciation of exploitation is a demand for egalitarian justice.

It is unfair to Cohen, Elster and Roemer to give so bald an account of their work. Their patience, fairmindedness and tenacity are beyond praise, and they have done much of what they set out to do: they have extracted the academically reputable elements of Marxism and presented them lucidly and coherently; at the same time they have made issues close to Marx’s heart topics of respectable debate.

Nonetheless, it seems to me that analytical Marxism is a dead end. It is hard to avoid the feeling that throughout Cohen’s work an essentially sentimental and old-fashioned communism of a pretty simple kind is vainly fighting for its life against an intelligence which erodes one Marxist claim after another. Precisely because he is so fair-minded and scrupulous, he produces an account of Marx from which all the fire and passion of Marx himself have been evacuated, without ever confronting the nasty political question of how what’s left is ever to form the basis of a popular socialism. Elster induces the same unease. He, too, puts his socialist cards on the table: he thinks socialism is economically possible and morally desirable, and that colours the questions he thinks worth asking. But at about the point where the reader concludes that Marx’s legacy is a major obstacle to achieving democratic socialism, Elster backs away. Equally, Roemer’s egalitarianism is attractive enough in its own right, but it is not Marx, nor Marxist, nor self-evidently able to stand up against the considerations which drive John Rawls’s Theory of Justice away from egalitarianism – roughly, that rational people would rather be absolutely better-off and relatively worse-off than vice versa.

Viewed as history of ideas – which is not how it asks to be viewed – analytical Marxism doesn’t illuminate the historical Marx. Viewed as a contribution to social science, it suggests awkwardly that we should somehow combine the analytical techniques of neoclassical economics with moral enthusiasms which are to be left free-standing. This may be the best we can do; I often think that it is. But it is a very far cry from Marx, who tried to produce an intellectual system that left no room for moral debate just because he was so conscious that theories of justice and morality were two a penny. This is why this reconstruction is evidently doomed to run into the political sands; if Marxism is reduced to a demand for more freedom and more justice, Marxists are doomed to become participants in the endless and repetitive debates over Rawls, Nozick and Dworkin which constitute the staple diet of academic political theory.

Viewed as philosophy of science, analytical Marxism is unimaginative and conservative4; if Marx had been right, we should have accommodated our ideas about scientific respectability to his successes – just as empiricist philosophers once accommodated their ideas to Newton’s achievements and stopped objecting to action at a distance. Marx’s sloppinesses and vaguenesses would have been glossed over, just as we take no notice of Darwin’s innumerable obscurities. Successful science quite properly just tramples on philosophers who stand in its way. For all these reasons, therefore, the project of showing that Marx can be made to talk in ways that Oxford philosophers of thirty years ago would have thought sufficiently sanitary seems a mistake. It satisfies neither the desire to get inside the mind of the historical Marx nor the need for an intellectual apparatus adequate to the complexities and absurdities of the late 20th century. The first is, of course, exceedingly difficult; the second may well be impossible – until hindsight provides the answers.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.


Vol. 9 No. 20 · 12 November 1987

The price of the paperback edition of Jon Elster’s An Introduction to Karl Marx, reviewed in the issue of 17 September, is £5.95 (not £10.95).

Editors, ‘London Review’

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN

Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences