Department of Philosophy, University College London
G.A. Cohen, universally known as Jerry, died unexpectedly on 5 August 2009. Born in 1941, he had recently retired as Chichele Professor of Social and Political Thought at Oxford University, and had taken up a part time post as Quain Professor of Jurisprudence at University College London. UCL was where he had begun his lecturing career in 1963, before his election, in 1984 at a youthful 43, to his Oxford Chair, which had previously been held by G.D.H. Cole, Isaiah Berlin, John Plamenatz and Charles Taylor. He took up the Chair in 1985, the same year in which he was also elected to the British Academy.
The question of who would be appointed to the Chichele Chair was, somewhat surprisingly, a matter of discussion in the national press. Cohen was relatively unknown and an unlikely candidate, at that time the author of just one book, Karl Marx’s Theory of History: A Defence (1978) and a handful of papers. On his appointment the satirical magazine Private Eye speculated that the committee may have been influenced by Cohen’s reputation as a wit and raconteur, and the need to enliven the quality of dinner table conversation at All Souls. Certainly Cohen had a unique and memorable gift for entertaining those around him – his conversation crackled with jokes, snatches of show tunes, and impressions of great philosophers, real and imagined – but in truth, the committee understood that he also had a rare, perhaps unique, philosophical talent, and their confidence in him was amply rewarded.
Cohen was born into a Jewish Marxist family, and his life and character were woven into his philosophical work in an unusual way, to the point where some of these writings contain extended descriptions of his upbringing and family. For example, Chapter 2 of his 1996 Gifford Lectures If You’re An Egalitarian How Come You’re So Rich? is entitled ‘Politics and Religion in a Montreal Communist Jewish Childhood’ and paints a moving picture of his childhood, his parents, their convictions and their social milieu as factory workers and, in the case of his mother, communist party member and activist.1 To read it is to be transported into another world: the world of a cold-war Canadian child, from an immigrant family, first convinced of the truth of Marxism and the moral superiority of Soviet Communism, but later trying to come to terms with the behaviour of the Soviet Union in the 1950s. Cohen’s upbringing, his family, his Jewishness (as distinct from Judaism) and his need to position his own beliefs in relation to Marx and to Soviet Communism were central to his life and work, both in terms of its content and, often, its presentation.
Equally important to his work was his training in Philosophy, especially at Oxford, where he moved from undergraduate study in McGill, in 1961. There he came under the influence – the ‘benign guidance’2 – of Gilbert Ryle and received a thorough grounding in the techniques of analytical philosophy, with its emphasis on rigour and fine distinctions. It was armed with such techniques that Cohen began his earliest project, resulting in his Isaac Deutcher Memorial Prize-winning book Karl Marx’s Theory of History: A Defence. (KMTH)3 Later he said it was a type of ‘repayment for what I had received. It reflected gratitude to my parents, to the school which had taught me, to the political community in which I was raised’.4 It was an attempt to state and defend Marx’s theory of history in a fashion that met the standards of rigour and clarity of contemporary analytic philosophy, in the face of criticisms from Plamenatz and others that this could not be done. The project proceeded relatively slowly. Cohen first published a handful of papers on Marx-related themes. These include two papers on what might be thought of as social epistemology. One, his first published paper, considers the question of whether one’s social role can determine what one can think and believe; Cohen argues that human freedom requires one to believe as a human being, rather than attributing one’s beliefs to a social role that one plays.5 A second paper asks how a Marxist understanding of the materialist production of ideas affects the question of whether any such ideas can be regarded as true.6 This is clearly a matter of huge importance for a Marxist philosopher, and, no doubt, a question Cohen felt he had to settle before taking any further steps. His response is that while other classes need, falsely, to represent their ideas as universal, in the sense of being in the interest of the great majority, the proletariat have no such need of pretence or deception. For their ideas really are in the interest of the majority.
Both these papers are, in a way, prefatory to his project of defending Marx’s theory of history, in that they are questions that need to be answered in order to carry out the project with confidence. A third paper from the period, however, is much more closely aligned to the book-length project. Published in 1970, it is called ‘On Some Criticisms of Historical Materialism’ and was presented to the Joint Session of the Aristotelian Society and Mind Association at its annual meeting.7 Here Cohen responds to some earlier criticisms of historical materialism by H.B. Acton and John Plamenatz, and Acton then replies to Cohen’s paper.
Although published some years before KMTH, several of the innovative themes of that work are foreshadowed here. First, Cohen praises Acton for applying the standards of rigour of analytical philosophy to Marxism, and suggests that in his own work he will apply even higher standards. For this reason, arguably, this 1970 paper may well be the first appearance of what was later to be called ‘Analytic Marxism’, using the techniques of analytical philosophy and formal economics and social science to defend Marxism, rather then to criticize it. Second, Cohen takes Marx’s 1859 Preface to the Contribution to a Critique of Political Economy as the central source for his reading of Marx’s theory of history. Finally, he presents a sketch of how the device of functional explanation can be used to overcome some difficulties in the formulation of the theory, which was to become one of the central aspects of his later reconstruction. The main topic of the symposium is the question of the relation between the economic base and the legal and political superstructure in historical materialism. The economic base is understood to be the set of relations of production, such as the relations between capitalists and workers, or masters and slaves, within the economy. Thus the base is, broadly speaking, the economic system. Plamenatz had argued that it was impossible to characterize economic relations of production except in terms of legal powers. For example, a proletarian is someone who has the right to sell his or her labour, unlike the serf or slave who has no such right. Yet to use the language of rights is to use a set of concepts belonging to the superstructure, and hence, so it is argued, it is impossible to define the economic structure except in superstuctural terms. If this is so, then, it is argued, it cannot be the case that the economic structure has explanatory priority over the superstructure, as orthodox Marxism dictates.
Cohen does not question Plamenatz’s claim that it is necessary to provide an independent account of the economic structure for it to play the role Marx requires of it. Rather he takes on the challenge of providing such an account – what he calls a ‘rechtsfrei’ interpretation. He argues that the economic base should be understood, strictly speaking, as constituted by powers, rather than rights. The superstructure, as a set of legal rights, exists in order to consolidate the powers belonging to the economic base. This is a direct and explicit appeal to functional explanation. The superstructure exists because it has a function: the function of protecting economic power. The solution is elegant. The base and superstructure can be characterized independently of each other, and while the superstructure has a causal effect on the base, it exists in order to have that effect. Therefore the economic base has explanatory priority even though causal influence goes in the opposite direction.
Although many of the elements were in place by 1970, and other important papers on Marx were published in 1972 and 1974,8 it was not until 1978 that Cohen published KMTH. Part of the reason for delay was his perfectionism in trying to get the details as precisely right as he could. But another explanation was that he was faced with a much more urgent project. In 1973 Robert Nozick published a long article in Philosophy and Public Affairs, which was to become the heart of the libertarian political philosophy elaborated in Anarchy, State and Utopia.9
Cohen reports that Nozick’s ideas were first drawn to his attention by Gerald Dworkin in 1972, and, in an important episode in his life, in 1975 he visited Princeton for a semester, lecturing on Nozick and making important connections with Tom Nagel and Tim Scanlon.10 On encountering Nozick’s arguments Cohen felt a need to divert his focus from his work on Marx , for the time being, to answer Nozick. Nozick, of course, sets out a natural rights based form of libertarianism, defending a minimal state, and condemning any form of redistributive transfer as coercive and unjust. For many left-liberals, Nozick’s was a dazzling defence of an obviously false and heartless view: a view that required attention because of the intellectual strength, wit and elegance of many of the arguments of the book but not because the overall doctrine presented gave them any cause to doubt their own heartfelt convictions. For Cohen, however, the situation was quite different. As he later put it, in a paper revealingly entitled ‘Marxism and contemporary political philosophy, or: why Nozick exercises some Marxists more than he does any egalitarian liberals’, Cohen’s Marxist-inspired critique of capitalism was based on the idea that the relation between capitalist and worker is exploitative, because it involves ‘the theft of another person’s labour time’. Yet in Anarchy, State, and Utopia Nozick argues that redistributive taxation has exactly that character. According to Cohen, Marxists such as himself at that time believe in the principle of self-ownership, that people are the rightful owners of their own powers, but exactly this principle is argued, by Nozick, to yield not communism, but a stark form of capitalist individualism. Refuting this view, then, became another essential ‘ground-clearing’ task in the defence of Marxism, but also very important for its own sake.11
Cohen’s classic paper on Nozick, ‘Robert Nozick and Wilt Chamberlain: How Patterns Preserve Liberty’ was published in 1977.12 (A slightly revised version was published in 1995 in Self-Ownership, Freedom and Equality. Like many of his reprinted papers the later version contains a number of small corrections and amendments.) Nozick vividly argued that any attempt to introduce a ‘pattern’ of distributive justice, such as equality, will require the state to prevent individuals from making voluntary transactions that might disrupt the pattern. Yet if the state were to do this, it would restrict individual liberty, needing to coerce individuals into conformity to the designated distribution, and so those who value liberty should resists any attempt to try to implement a pattern. Cohen makes many points in criticism of Nozick’s argument, but his main response is that Nozick has overlooked the fact that a distribution of property is already a distribution of liberty. One person’s ownership of an item of property entails that other people are not at liberty to use it without the owner’s permission. Therefore it can be the case that a pattern is needed to preserve the liberty of those who would otherwise suffer in an unpatterned distribution. Hence, Cohen argues, patterns preserve liberty. He notes that Nozick attempts to avoid this, by redefining liberty as, essentially, the freedom to do what one has a right to do, and so a non-owner’s inability to use the property of its owner is no longer a detriment to liberty. But if this move is made it then becomes question-begging to try to defend a view of private property in terms of liberty, for any account of liberty already assumes a view of justified property. This critique is arguably the most powerful and influential of those that attempted to engage with Nozick’s argument.
Cohen finally published KMTH in 1979, as well, that year, as publishing a brilliant, critical examination of Marx’s labour theory of value and its relation to the theory of exploitation.13 On the publication of KMTH Cohen established his position as among the world leading interpreters of Marx’s thought. The book is a considerable extension of the earlier paper ‘On Some Criticisms of Historical Materialism’, and sets out a clear account of the core of Marx’s theory of history. According to Cohen the two central theses of historical materialism are the ‘development thesis’ and the ‘primacy thesis’. The development thesis states that society’s productive forces tend to develop throughout history, in the sense that human productivity tends to become more powerful over time. The primacy thesis is a combination of two claims: that the nature of the productive forces explains the economic structure, and that the nature of the economic structure explains the superstructure (the claim we saw explicated and defended in the earlier paper). Put together, this is a form of technological determinism: the ultimate explanatory factor for all other significant facts about society is the nature of technology available. As Marx himself puts it, in The Poverty of Philosophy, ‘the hand-mill gives you society with the feudal lord, the steam mill society with the industrial capitalist’.
This theory is distinctively Marxist in that it divides history into epochs – pre-class society, slavery, feudalism, capitalism, and communism – and understands the transition from epoch to epoch as the result of class struggle and revolutionary change. The claim is that an economic structure, such as capitalism, will persist for as long as it is optimal for the development of the productive forces (an application of functional explanation) but eventually it cannot contain all the growth it has stimulated. At that point the economic structure ‘fetters’ the development of the productive forces, and must give way, to be replaced by a new economic structure that will continue the development of the productive forces.
Although the basic theory can be stated simply, KMTH is a complex book. First, it enters into many disputes regarding the detailed understanding of Marx, attempting to settle contested points of textual interpretation. Second, Cohen builds detailed and ingenious arguments for many of the positions taken. The book, after all, is an interpretation of Marx and a defence of the view. Accordingly the discussion encompasses questions not raised by Marx, such as how to formulate and deploy the central device of functional explanation, or how to argue for the claim that the forces of production tend to develop throughout history.
On publication the book received wide acclaim. At the same time, naturally enough, it received various forms of criticism. Some of this criticism was aimed at the interpretation of Marx. One oft-made charge was that Marx’s theory of history was not, at bottom, one of technological determinism. Some of these critics pointed out that Cohen had down-played Hegel’s influence on Marx, and, accordingly, had not taken seriously dialectical forms of reasoning. However, Cohen’s project of incorporating analytic philosophy into Marxism was designed precisely to overcome what he saw as the damaging obscurantism of Hegelian Marxism, especially that transmitted via the work of Althusser.14 A related, and less doctrinaire, criticism, was that Cohen allowed only a relatively minor role for class struggle. In Cohen’s reading, class struggle is the agent of change from epoch to epoch, rather than the engine of history at all times, as appears to be indicated by Marx’s remark that ‘history is the history of class struggle’. Yet Cohen was convinced that his interpretation of Marx was correct on this point, accepting that class struggle is the ‘immediate driving power of history’ but not its ‘underlying’ driving force.15
Other lines of criticism concerned the theory itself, rather than whether it was a true depiction of Marx’s thought. Jon Elster, for example, strongly criticized the use of functional explanation, arguing that it retained an unacceptable teleology. As Elster observes, suggesting that economic structures rise and fall as they further or impede human productive power seems to assume that history is goal directed, or even that there is some sort of external agency ensuring that progress continues to be made.16 Andrew Levine and Erik Wright take issue with Cohen’s argument that the development of the productive forces can be explained in terms of what they call ‘rational adaptive preferences’. They suggest that Cohen does not take sufficiently into account problems of collective action.17 Joshua Cohen made similar criticisms and adduced evidence that there have been long stretches of history in which the productive forces declined in strength (most notably on the fall of the Roman Empire).18 Richard Miller pointed out that the account of fettering was unclear.19 Did the productive forces have to stop developing, or was it enough that they developed more slowly than they would under some other economic structure? Furthermore, the use of the productive forces and their development are quite different. One could argue that capitalism greatly develops the productive forces but uses them poorly. Is this fettering or not?
Each of these criticisms brought forward important responses and further clarifications of the theory. In response to Elster, Cohen pursued the analogy with the Darwinian use of functional explanation in evolutionary biology, which does not pre-suppose teleology or ‘nature’s purposes’.20 In response to Joshua Cohen, and Levine and Wright, Cohen, together with Will Kymlicka, wrote a detailed rebuttal of their argument,21 and in response to Richard Miller, Cohen broadly accepted the criticism that the theory of fettering was unclear, and wrote a detailed clarification, which was first published in History Labour and Freedom and was ultimately incorporated as an additional chapter in the extended edition of KMTH.
Yet in the face of these criticisms and reformulations Cohen began to see that the theory was not as clear-cut as he had thought. He also had began to develop reservations of his own especially about historical materialism’s apparent neglect of people’s apparent need for self-definition: that is, the need to identify with groups in society that are less than the whole. This in turn leads Marxism to a dismissive and reductionist approach to religion and nationalism. These anxieties are recorded in two other papers that also appeared first in History, Labour and Freedom and then in the expanded edition of KMTH: ‘Reconsidering Historical Materialism’, and ‘Restricted and Inclusive Historical Materialism’. Others might, at this point, have seen the enterprise as a ‘degenerating research programme’. Instead of using the theory to illuminate and explain ever more aspects of empirical reality, it appeared to require increasingly intricate internal development, specification and qualification, to defend it against criticism, thereby reducing its explanatory power. Indeed, Cohen explains that his attitude to historical materialism had changed on completing the book. While writing it he was sure that Marx’s theory of history was correct. After, he said, it was not so much that he believed it to be false, but that he did not know how to tell whether or not it was true.22
During this time Cohen was a founder and very active member of the Non-Bullshit Marxism Group (later called the September Group), which was a remarkable, interdisciplinary group of scholars who first met in 1979 and again 1980, to discuss exploitation. They then met annually, and then biennially, to discuss wider themes. The core membership of the original group, aside from Cohen, were Jon Elster, John Roemer, Hillel Steiner, Philippe van Parijs, Robert van der Veen, Adam Przeworski, Erik Olin Wright, Pranab Bardhan and Robert Brenner, although the membership changed considerably over the years.
The September Group was founded to discuss themes within Marxism, but their allegiance to the themes lasted longer, typically, than their allegiance to Marx. This development was foreshadowed, to some degree, by Cohen’s earlier paper ‘The Labour Theory of Value and the Concept of Exploitation’ in which he had argued that the concept of exploitation does not rest on the labour theory of value. As Cohen continued to work in the 1980s and beyond, one might characterize his writings as working out how to formulate his opposition to capitalism and allegiance to socialism without the underpinnings of Marx’s theory of history. As he put it, ‘In the past, there seemed to be no need to argue for the desirability of an egalitarian socialist society. Now I do little else.’23
This next phase in his career takes up themes that emerged in his criticism of Nozick: the relation between capitalism, socialism and freedom, and the nature and consequences of the thesis of self-ownership. These are the topics of the last few papers reprinted in History, Labour and Freedom, and all of his next collection, Self-Ownership, World Ownership and Equality.
Cohen in the 1980s, was especially exercised by the Marx-inspired question of how to understand the unfreedom of workers under capitalism, given that they exist in a structure which places freedom of choice at its centre, and no worker is forced, so it appears, to work for any particular capitalist. Yet at the same time there seems to be a vital sense in which workers in the capitalist system remain unfree. One important part of the analysis is to provide a definition of being forced to do something in which saying that a person is forced to do something does not mean that it is the only option available to him or her, but that any other options he or she has are not acceptable or reasonable .24
One obvious response to the claim that workers are forced to sell their labour-power to the capitalists, on this definition, is that workers do have an acceptable alternative; they can become petty bourgeois shop owners or self-employed in some way. Here Cohen accepts that this escape route is available to some workers, yet, he argues, although any individual worker is free to leave the proletariat, the proletariat is collectively unfree, for there are nothing like as many escape routes as there are members of the proletariat.25
The notion of the worker’s right to freedom and the thesis of self-ownership are linked through the idea of a person’s right to control their actions and labour. Cohen characterizes self-ownership as the thesis that ‘each person enjoys over herself and her powers, full and exclusive rights of control and use, and therefore owes no service or product to anyone else that she has not contracted to supply’.26 This is, of course, qualified by the condition that rights of self-ownership do not permit one to interfere coercively into the lives of others. In a series of papers Cohen considered the relation between self-ownership and what he refers to as ‘world-ownership’: rights over those parts of the world that are not persons. Essentially Cohen set out to rebut the Nozickian argument that rights to self-ownership entail rights to world-ownership (i.e. individual property rights) that are in principle unrestricted. Nozick had argued that any attempt to redistribute worldly resources in effect conscripts one person, willingly or not, to work for another.
Cohen notes that those who are in favour of redistribution have the option merely to deny self-ownership, and assume that we have non-contractual duties of non-interference. Yet he argues that a stronger defence is to accept, for the purposes of argument, the thesis of self-ownership, and show that the Nozickian conclusion does not follow. In a now-classic discussion of Nozick’s account of justice in initial acquisition, Cohen points out that it is essential to Nozick’s argument that the external world is, initially, unowned, and therefore available for initial acquisition. However, Nozick does not show that the world is not jointly owned by all human beings. If that were the case the conditions for appropriation would be much more strict, and would not yield the type of property rights favoured by libertarians. Hence, at the least, Nozick has not shown that radically unequal distribution can follow from self-ownership. Furthermore, even if the world is not jointly owned, Cohen argues that Nozick’s defence of initial appropriation – roughly, an appropriation is acceptable as long as it makes no one worse off - contains a strong element of paternalism that Nozick would reject in other circumstances.27
As his work on this topic developed Cohen seemed close to endorsing the thesis of self-ownership, especially, as we noted above, because he saw it as very similar to the views that underlie the Marxist opposition to exploitation. Yet he came to believe that self-ownership and an attractive form of egalitarianism were in conflict. In ‘Are Freedom and Equality Compatible?’ he argues that assuming that egalitarianism should be characterized by the thesis that the world is jointly owned by everyone is far too restrictive. It would require everyone else’s consent before anyone could use anything at all. This, Cohen argues, thereby renders self-ownership ‘merely formal’ (a criticism that also applies to libertarianism, for the self-ownership of those without property is also merely formal and they would rely entirely on the co-operation of others for survival.) He continues with the argument that the egalitarian alternative of parceling the world into equal individual shares fares no better, at least from an egalitarian point of view, as it will allow outcomes to be strongly determined by the exercise of differential talent, and fails to guarantee support for those who cannot produce for themselves.28
Eventually, therefore, he found the principle of self-ownership unhelpful, and in a pair of papers published for the first time in Self-Ownership, Freedom and Equality he came to accept the position that many of his liberal egalitarian friends and colleagues had urged upon him for years: the rejection of the principle. While he defends the coherence of the idea of self-ownership, he argues against its adoption. His position is not so much to find a direct argument against the thesis, but rather to demonstrate that the motivations that lead in the direction of self-ownership do not take one all the way. That is to say, one can oppose slavery, advance autonomy and object to treating a person as a means without adopting self-ownership.29
Self-Ownership, Freedom and Equality largely pursues a single theme: the thesis of self-ownership, and it is something of an anti-climax that the work ends on a largely negative note. Yet this should not detract from the point that the essays together add up to the most powerful and influential detailed rebuttal of Nozick’s libertarianism that has been produced, one that is unlikely to be surpassed. The essays, executed with supreme rigour, are full of insight and interest even when their point is to warn against a wrong turning, rather than build a new construction.
Modestly Cohen characterised himself as essentially a reactive philosopher. This, as a more general conception of philosophy, comes out clearly in a remarkable paper ‘How to Do Political Philosophy’, written for use in teaching a graduate class in Oxford, where it is clear that Cohen conceives of philosophy as an activity that takes place against an opponent.30 In the first phase of his career Marx was the clear inspiration, and the opponents were analytic critics of Marx such as Plamenatz, and obscurantist defenders, such as Althusser. The second phase was dominated by the need to answer Nozick’s libertarianism. In the third phase Ronald Dworkin was the focus of his reflections, and in particular Dworkin’s two seminal articles on equality of welfare and equality of resources.31 Cohen found himself very sympathetic to what later came to be called ‘luck egalitarianism’. Dworkin’s achievement, said Cohen, in a much quoted passage, was to perform: ‘for egalitarianism the considerable service of incorporating within it the most powerful idea in the arsenal of the anti-egalitarian right: the idea of choice and responsibility.’32 Indeed, there is a strong residue of Cohen’s earlier reflections on Nozick in this comment.
The leading idea of luck egalitarianism is to make a distinction between those aspects of one’s fate for which one is responsible, and those aspects for which one is not. Dworkin makes a distinction between ‘brute luck’ and ‘option luck’ and on this view the project is to set out principles that allow people to reap the benefits, but also suffer the burdens, of good and bad option luck, but at the same time to insulate people from the effects of good and bad brute luck. There are, at least, two central questions that must be answered in order to settle how this doctrine is to be formulated. One is the question of how exactly to define the ‘cut’ between those factors for which a person is to be held responsible, and those they are not. A second is the ‘currency’ of justice: should equality be defined in terms of welfare, resources, capabilities, or something else again? Dworkin is very clear on the second question: equality of resources is the right currency. His response to the first question – how exactly to draw the cut – was less easy to discern from his writings.
Cohen’s contribution to this debate was initially set out in two papers, ‘On The Currency of Egalitarian Justice’, mentioned above, and ‘Equality of What? On Welfare, Goods and Capabilities’. 33 Cohen broadly accepts Arneson’s characterisation of Dworkin’s theory as one of equality of opportunity for resources.34 However, in opposition to Dworkin and Arneson, Cohen’s preferred position is one of ‘equality of access to advantage’, of which the more important modification is ‘advantage’ instead of ‘resources’ or ‘welfare’. Cohen’s point is that an egalitarian must be sensitive to certain types of resource deficiency, however they impact on welfare, as well as certain types of welfare deficiency, however they impact on resources. Accordingly he defines a new notion – advantage – which straddles resources and welfare (although he does not attempt to specify how the two elements are to be combined).
One primary issue between Cohen and Dworkin comes down to the question of whether low welfare should engage egalitarian concerns. Dworkin admits the immediate appeal of such a view, but marshals a range of considerations to put it into doubt. Perhaps surprisingly, the focal example for deciding between different cases is that of ‘expensive tastes’. If a person cannot enjoy those things others typically take pleasure in – beer and hen’s eggs, say – but to achieve comparable levels of enjoyment, they must consume expensive champagne and plover’s eggs, should they receive a social subsidy so that they can achieve the same level of enjoyment as the rest of the population?
Dworkin’s position is that one should not receive a subsidy for expensive tastes, unless they are a form of compulsion or craving, akin to mental illness. Cohen, by contrast, argues that there is a difference between those people who find themselves with expensive tastes, by bad brute luck, who should be subsidized, and those who deliberately cultivated them, who, in the spirit of luck egalitarianism, should be required to bear the consequences of their freely made choice. Dworkin argues that the key factor for deciding whether or not subsidy is due is whether the person identifies with their tastes. It would be ‘alienating’ to offer people subsidy for aspects of what they regard as their personality. In reply Cohen makes the important distinction between identifying with the taste, and identifying with its cost. One can fully identify with the taste yet regret that it is expensive. If one has not deliberately cultivated it, then, in Cohen’s view, subsidy is due. The debate between Dworkin and Cohen went through several exchanges, both making strong and plausible arguments and neither side prepared to concede ground.35
When luck egalitarianism – in all its versions – came under attack from Elizabeth Anderson36 and others for its apparent inhumanity, such as its ‘abandonment of the irresponsible’ (those who have freely chosen paths with disastrous consequences and would therefore have no claim for help) Cohen took pains to point out that his project was only to define and argue for a theory of equality as an account of distributive justice, and not to argue that any society should adopt an unmodified principle of equality. Rather, he reminded his readers of a point that he had made explicit in his earlier paper. He accepts that concerns other than those of egalitarian justice could turn out to be more important in practice.37 In making this point he develops an early version of a distinction that, as we will see, became important in the last period of his work: the distinction between theories of justice and what he was to call ‘rules of regulation’.
The next phase in Cohen’s work began with three papers that stand with Karl Marx’s Theory of History, and the critique of Nozick, as the high points of his career. These are the Tanner Lectures of 1992, ‘Incentives, Inequality and Community’,38 ‘The Pareto Argument for Inequality’ (1995)39 and ‘Where the Action Is: On the Site of Distributive Justice ‘(1997).40 The last of these papers was also included in Cohen’s superbly readable and engrossing book If You’re an Egalitarian How Come You’re So Rich? and together they also comprise the first three chapters of his final major book, Rescuing Justice and Equality. The essential question is how, as a believer in equality, should one behave in one’s personal economic life? In particular Cohen is concerned to question how it could be consistent both to pursue a high income and to espouse egalitarianism. Much of this work is aimed at the criticism of one particular attempt to defend such a combination to be found in the work of John Rawls, who, in effect, became Cohen’s last philosophical opponent. Rawls’s famous ‘Difference Principle’ states that inequalities in income and wealth are justified when they are to the greatest possible benefit of the worst off.41 An apparently naïve reply to Rawls is to question how inequalities could ever be to the benefit of the worst off. Inequalities can be removed by transferring money from the richer to the poorer, thereby achieving equality by making the worst off better off. The Rawlsian reply is that such a transfer would, of course, be better if it were possible. But the Difference Principle also anticipates situations where such a beneficial transfer is not possible; that is where equalizing would make everyone worse off, at least in the longer term. Broadly this doctrine is thought to be sensitive to the economic argument that everyone can be better off if the highly productive are provided with material incentives to work harder. This, in turn, leads to inequalities that are to the advantage of all.
Cohen, however, pushes the argument to another stage. How can there be circumstances where equality is impossible at a higher level for all? Presumably, only because those who are well off will not contribute as much effort at a lower level of income. This may be an understandable response, if not admirable, for those who do not believe in equality. Yet one of the conditions of Rawls’s account of a ‘well-ordered society’ is that everyone should believe in the Rawlsian principles of justice, and in particular endorse the principle that the worst off should be made as well off as possible. Therefore people in a Rawlsian society should not seek higher wages than others, unless there is some special reason why they cannot (as distinct from will not) be more productive on the same income as others. Therefore, Cohen argues, the Difference Principle justifies much less inequality than it is often thought to do, and that Rawlsian principles of justice must be supplemented by an ‘egalitarian ethos’ to guide choices in every day life.
There are several resources in Rawls to try to combat this line of argument, although Rawls himself never confronted it in detail. One important response is that the Difference Principle is intended to regulate the ‘basic structure’ of society, rather than personal behaviour. This, and several other strategies, are discussed, and rebutted in detail in the first half of Rescuing, Justice and Equality. The second part of the book, while still engaged with Rawlsian theory, changes tack, extending an argument first presented in a paper entitled ‘Facts and Principles’.42 Here the project is to attempt to show that basic principles of justice must be ‘fact free’ in the sense of not depending on any empirical facts. This contrasts with a Rawlsian ‘constructivist’ approach in which facts about human nature and society are taken into account at the most basic level in formulating principles of justice. Here Cohen accuses Rawls and his followers of failing to respect the distinction mentioned above between rules of regulation and (pure) principles of justice.
Although this work has attracted respectful and detailed attention, many readers have been surprised by this turn in Cohen’s work. Although it is the fruit of several years of sustained endeavour, in contrast to most of his other work it is much less clear what the payoff is, as his opponents are not convicted of any substantive error regarding what is to be done, as distinct from conceptual confusion regarding the nature of justice. However, for Cohen conceptual clarity for its own sake was of the supreme importance.
Nevertheless, certainly for the chapters in Part One, the book is already a classic in political philosophy, and it may well be that in time the significance of Part Two will come to be better understood. Furthermore, in presenting his ideas in book length form Cohen came to reflect on a number of items that are foreshadowed in earlier work but explicitly clarified here. For example, there is a short discussion of Cohen’s attitude to moral realism,43 and a more explicit endorsement of pluralism than is found elsewhere44.
Cohen’s final book, Why Not Socialism? was completed before he died but published posthumously.45 The book is very short, and published in small format. It begins with an account of a camping trip and persuasively argues that under such circumstances, the trip would be much more enjoyable for all participants, and more efficient, if the campers adopted certain anti-individualist principles of community and equality that could fairly be described as socialist, rather than capitalist market principles, to govern their interactions. The book continues with the question of why it should be that such socialist principles are not adopted in broader social and economic life. Here Cohen refuses to accept the pessimism about human nature that suggests that natural human selfishness makes socialism impossible. Rather, he points out, we have not (yet?) been able to devise social mechanisms that allow ourselves to organize large-scale economic interaction on the basis of human generosity, in contrast to the capitalist free market, which can turn individual greed and fear to general advantage, although, if course, it has many disadvantages too.
At his death Cohen left a number of works in progress as well as a series of lectures on moral and political philosophy that he had intended to prepare for publication. Much of this work will be published in the next few years. One of the most intriguing as yet unpublished papers is called ‘One Kind of Spirituality’. The importance of Cohen’s Jewish background has already been remarked upon, but many assumed that he had no interest in any issues of religion or spirituality, especially given what he has described as his ‘anti-religious upbringing’. However two of three children, Gideon and Sarah, took a different direction, Gideon adopting Rastafarianism and moving to Ethiopia, and Sarah spending much of her time in an Ashram in the southern Indian state of Kerala. Cohen’s love and respect for his children no doubt encouraged him to take their views seriously. In his Gifford Lectures for 1996 – normally given on a theme in Philosophy of Religion – Cohen stated that he was agnostic, not an atheist. But more surprisingly for many readers, he revealed himself as a long-standing and regular Bible reader of both testaments. In the lectures he showed a respectful and tolerant attitude to religion, and especially Christianity.
It should be clear from the forgoing how important family was to Cohen. His first major philosophical project was seen as a type of repayment to his parents, to whom his first book was dedicated. He married Margaret Pearce in 1965 and they had three children, Gideon, Miriam (who now teaches philosophy in London) and Sarah. The marriage was dissolved but Cohen remained on very good terms with Maggie, and both remarried, Cohen to Michèle Jacottet in 1999. His second marriage was a very happy one, spent in the company of what was now a complex and growing extended family.
Cohen’s contribution to political philosophy has been extensive, defending what many would regard as the most thorough-going and radical egalitarianism to be found among analytical philosophers. However, his own positive view was not developed in the detail of other leading figures such as John Rawls, Ronald Dworkin or John Roemer. Rather, as noted, Cohen considered himself more of a reactive than an individually creative philosopher. In this his skill was unrivalled. His style is often that of an expert demolition worker: finding what might look like a rather banal difficulty, but probing and probing until the edifice collapses. Cohen knew exactly where to locate his criticism, and how to develop it to greatest effect. At first sight the criticisms can look pedantic or fussy, but as the arguments develop something of great power emerges. Those who have the instinct to defend the views he attacks find themselves with a much more difficult task than they first assumed.
Cohen will be remembered for his work, but just as much for his wit and his support for other people. Even in prestigious public lectures he would crack jokes, burst into song, or imitate other philosophers. He would do the same thing in restaurants, drawing waiting staff or diners at other tables into the fun and good-natured mischief. His valedictory lecture at Oxford in 2008 included a series of imitations or parodies of many well-known philosophers, and was said by many members of the audience to be the funniest and most entertaining lecture they had witnessed. Fortunately some video and audio recordings of Cohen survive, most notably a TV programme, No Habitat for a Shmoo, made in 1986, some videos of lectures and impersonations delivered in Madison Wisconsin in 1998, and an imperfect audio recording of the valedictory lecture.46
In 2009, about half a year before his death, a conference at Oxford was held in celebration of his work and career. In remarks at the end of the conference Cohen observed how odd it is that in this country we honour people by attempting to rip their work to pieces. But in these remarks he also made clear how extremely proud he was of his former students--of how confident they had become, and of how much they had become their own people. Cohen was extraordinarily generous with his time, and not only for his own students, and not only on his own topics. His native intelligence – honed by tutorials with Gilbert Ryle – enabled him to grapple with any topic put to him, and fifteen minutes with Cohen would leave anyone understanding both more and less about their own view or argument. All of those who met him, or read his work, will realise what a gap his unexpected death has left. He gave so much, yet he still had so much more. Any attempt to express how much he will be missed by his family, friends, colleagues and even those who never met him, will seem trite or formulaic.47