Social Philosophy & Policy Vol. 6 Issue 2, ISSN 0265-0525 IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN ?* BY DAVID MILLER INTRODUCTION This paper stands at the confluence of two streams in contemporary political thought. One stream is composed of those critics of liberal political philosophy who are often described collectively as 'communitarians'.1 What unites these critics (we shall later want to investigate how deep their collegiality goes) is a belief that contemporary liberalism rests on an impoverished and inadequate view of the human subject. Liberal political thought - as manifested, for instance, in the writings of John Rawls, Robert Nozick, and Ronald Dw'orkin - claims centrally to do justice to individuality: to specify the conditions under which distinct individuals, each with his own view about how life should be lived, can pursue these visions to the best of their ability. But, the critics claim, liberalism is blind to the social origins of individuality itself. A person comes by his identity through participating in social practices and through his affiliation to collectivities like family and nation. An adequate political philosophy must attend to the conditions under which people can develop the capacity for autonomy that liberals value. This, however, means abandoning familiar preoccupations of liberal thought especially the centrality it gives to individual rights - and looking instead at how social relationships of the desired kind can be created and preserved. It means, in short, looking at communities - their nature and preconditions.2 The other stream comprises various attempts to recast the principles of socialism with the aim, broadly speaking, of bringing it more closely into line * I should like to thank the participants in the conference on "Capitalism and Socialism" organized by the Social Philosophy and Policy Center for helpful discussion of an earlier draft of this paper, and Jerry Cohen, Andrew Williams, and Lesley Jacobs for sending valuable written comments. 1 See, for instance, Amy Gutmann, "Communitarian Critics of Liberalism," Philosophy and Public Affairs, vol. 14 (1985), pp. 308-22. 2 This sketch of communitarianism is deliberately ambiguous in one aspect. We may read the communitarian critics as basing their argument on a core liberal ideal - personal autonomy - but as proposing a more adequate account than mainstream liberalism of the conditions under which autonomy can be realized. Alternatively, we can read them as departing in a more fundamental way from liberal assumptions, substituting a different conception of the self, a different conception of freedom, and so forth. This ambiguity runs deep in communitarian writing (I return to the point later, particularly in relation to Charles Taylor): does communitarianism come to fulfill liberalism or to destroy it ? 52 DAVID MILLER with the aspirations of the majority of people (including the majority of workers) in the advanced societies. This means not only discarding outdated policy proposals, such as extensive schemes of nationalization, but at a more fundamental level looking critically at traditional socialist ideals (for instance, the belief that it is intrinsically better for people to enjoy goods and services in common than to enjoy them privately as individuals). We can identify the central ideal of the new socialism as equality of effective choice: people should have the rights, opportunities and resources that enable them to choose effectively how they are to live their lives. Socialism is not the enemy of freedom, but its best friend; whereas libertarians and liberals claim that their proposals provide people with the greatest equal liberty, only socialist policies can make that liberty effective.3 How are these two streams of thought related ? In one, we find people attacking liberalism in the name of community: in the other, we find socialists trying to divest themselves of traditional commitments, including communitarian commitments, and to outflank liberals in their devotion to individual freedom. Should the communitarian critique give the new socialists any pause for thought ? Are the ideas it advances in any way integral to socialism itself? Or should it be regarded as an essentially conservative response to liberal institutions with which a modernizing socialist should have nothing to do ? To answer these questions, we need to look more closely and critically at the often obscure views of the 'communitarians'. First, though, I shall offer a schematic interpretation of the socialist tradition which is intended to bring the questions above into sharper focus. I. Two STRANDS IN THE SOCIALIST CRITIQUE OF CAPITALISM It hardly needs saying that 'socialism', like other terms designating ideologies, resists straightforward definition. It is impossible to provide a set of necessary and sufficient conditions for a political outlook to be socialist. Wittgenstein's strictures about family-resemblance terms like 'game' apply with their full force here. We can, however, say uncontroversially that socialism arose as a radical reaction to nineteenth-century capitalism, and it is certainly a necessary condition for an outlook to be socialist that it advocate a major transformation of that order. For the purposes of the present discussion I want to isolate and contrast two strands in the socialist critique of capitalism, which I think on any reckoning would count as fairly central though this is not to say that every socialist has embraced them both.4 3 For statements of this view by two promii.ent members of the British Labour Party see Bryan Gould, Socialism and Freedom (London: Macmillan, 1985); Roy Hattersley, Choose Freedom: The Future for Democratic Socialism (London: Michael Joseph, 1987). 4 Nor do I want to say that the socialist critique is exhausted by the two elements I identify. It has other strands too: for instance one charge often made by socialists is that capitalism is a highly IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 53 The first element in the socialist critique focuses on the distributive inadequacies of capitalism. Capitalism, it is alleged, distributes resources, freedom and power in a way that is grossly unfair and/or prevents a large section of the population from receiving decent quantities of these benefits. Socialist institutions would allocate such benefits in a far more egalitarian fashion, in order to conform with socialist criteria of justice (which, in the extreme view, would prescribe perfect equality - in less extreme views, those criteria would find a place for limited inequalities based on desert or merit). The most obvious target of this critique is the distribution of wealth and income in capitalist society. As I interpret it here, however, the critique also extends to issues such as the distribution of labor-time (workers sweat while capitalists stand idle), the distribution of power in economic enterprises (capitalists command while workers obey), and the distribution of power in society more generally (capitalists control the state, benefit from the legal system, etc.). If we construe 'resources' very broadly to include benefits such as these, we can summarize the socialist charge as one of maldistribution of resources: the indictment of capitalism is not that it generates the wrong resources, but that it allocates them in a way that is unfair and inhumane.5 The second element in the socialist critique, in contrast, focuses on the quality of life in capitalist society, including the quality of the resources it generates. Included here, we find a number of different charges: for instance, capitalism involves production for profit rather than production for use, and therefore fails to provide people with the goods and services that they really need; it stifles creativity and robs work of its aesthetic content; it promotes the consumption of privately-purchased commodities, rather than enjoyment of goods and services in common; it fosters competitive relationships between people rather than relations of cooperation and fraternity; it renders inefficient system, making poor use of the welfare-generating resources available to it. There is a good discussion of efficiency arguments in Allen Buchanan, Ethics, Efficiency and the Market (Oxford: Clarendon Press, 1985), ch. 2. 5 Marxists often claim that their critique of capitalism does not involve a charge of maldistribution. Their meaning, I think, is that they are not centrally concerned with the allocation of income; in particular, they want to dismiss the suggestion that capitalism can be made acceptably fair by income redistribution schemes. What this suggests is either that they see a fairly rigid connection between distribution in the narrow sense (income distribution) and the structural inequalities of capitalism (e.g., the power structure of enterprises) or that they see distribution in the former sense as a comparatively trivial matter. In my wider sense, however, the Marxist critique, in this aspect, would properly count as a distributive critique. Marx himself recognized that his critique of capitalism could be expressed in distributive terms. For the evidence, see G.A. Cohen, "Freedom, Justice and Capitalism," New Left Review, vol. 126 (March-April 1981), pp. 13-14 n. 7. 54 DAVID MILLER people's interactions instrumental and formal, rather than encouraging the spontaneous meeting of human hearts (and so forth). These claims, made singly or in combination, add up to the thesis that capitalism does not and cannot provide the good life for man, and that what must be brought about is not a mere reshuffling of resources, however radical, but a qualitative change in human relationships and motivations. These two components of the socialist critique are not, of course, mutually exclusive, and can indeed be employed in tandem to good effect. In Marx, for instance, the first element is represented by the idea of exploitation - the claim that, under capitalism, the surplus value created by the labor of workers is systematically expropriated by capitalists - and the second is represented by the idea of alienation - the claim that work under the conditions of capitalism fails to realize man's "species-being" (i.e., his nature as a creative and communally-oriented being). Nevertheless, it is in most cases possible to disentangle the two elements, and doing so may throw some light on the revisionary socialist project mentioned at the beginning of the paper. For that project can be interpreted as one of pursuing the distributive critique to the exclusion of what I have identified as the "quality of life" critique. Socialism, on this view, is entirely a matter of the fair distribution of resources, taking "resources" in the broad sense as above. It is not concerned with what people may do with the resources they are allocated, with what motivates them, or how they are related to one another - except insofar as these matters have repercussions for the allocation of resources itself. It is not difficult to understand the pressures that push contemporary socialists in this direction. First of all, the distributive critique can be reconciled with major features of modern industrial societies far more readily than the "quality of life" critique. Consider two of these features: the market economy and the legal system. There is primafacie no incompatibility between the distributive critique and economic markets as such; that is, it seems a feasible project to reallocate resources in such a way that markets produce outcomes that are acceptable on grounds of distributive justice. How this could be done would depend on the criterion of distributive justice employed, but I am thinking generally of schemes such as equalization of capital holdings, the conversion of enterprises into worker cooperatives leasing capital, progressive income taxes, and so forth. To oppose this, one would have to hold that there was an inexorable connection between markets and capitalism such that, if markets are allowed to flourish, standard capitalist patterns of ownership must inevitably re-emerge; or, on the other hand, if these arrangements are outlawed, markets will be unable to work effectively. (Although this view is sometimes expressed - both by Marxists and by libertarians - the argument for it remains obscure to me.) By contrast, most versions of the "quality of life" critique - including the IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 55 6 Marxian theory of alienation - entail the condemnation of market relationships as a distortion of genuine human relationships. No matter how radically resources are redistributed, activity in the market must be governed by norms of instrumental rationality, people must behave non-tuistically (that is, each must aim to maximize his holdings, regardless of the welfare of his partners in exchange), and so forth. The "quality of life" critique seems therefore inevitably to point beyond markets towards some other method of coordinating economic behavior. By the same token, however, it lays itself immediately open to a charge of utopianism. If we want a feasible form of socialism, it seems that we have to accept a major role for markets, and to that extent we must abandon the "quality of life" critique.7 A similar point can be made with respect to the legal system. A modern legal system can be regarded as a system of uniform general rules enforced by formal procedures, which confer rights on individuals. The distributive critique includes nothing at odds with this idea of legality itself. Socialists characteristically allege that, in capitalist societies, a particular set of rights favorable to capitalist interests is embodied in the law, and moreover that enforcement procedures favor those already well-endowed with resources. A socialist system would rectify these defects by, for instance, recognizing enforceable rights to welfare and allocating resources in such a way that access to the legal system was effectively equalized. In contrast, the "quality of life" critique contains elements hostile to the very idea of legality: rights are dismissed as "the prized possessions of alienated persons,"8 and the formality of the law is contrasted with arrangements whereby people could deal with one another as complete human beings, each responding to the full particularity of the other. The problem, once again, is to see how an alternative to the legal system can be made to seem feasible in a modern industrial society.9 Considerations of realism, then, are one major pressure inducing contemporary socialists to abandon the "quality of life" critique in favor of the distributive critique of capitalism. A second pressure is loss of faith in the assumptions needed to back up the former critique. The "quality of life" critique requires us to judge some modes of human life as better than others, 6 I have examined this theory critically in "Marx, Communism and Markets," Political Theory, vol. 15 (1987), pp. 182-204. 7 Evidence for believing that a feasible form of socialism must allow a major role to markets is usefully presented in Alec Nove, The Economics of Feasible Socialism (London: Allen and Unwin, 1983). 8 Ruth Anna Putman, "Rights of Persons and the Liberal Tradition," ed. Ted Honderich, Social Ends and Political Means (London: Routledge and Kegan Paul, 1976), p. 102. 9 I have argued this point briefly in Anarchism (London: Dent, 1984), ch. 12. For a fuller defense of the idea of legality against left-wing criticism, see Tom D. Campbell, The Left and Rights (London: Routledge and Kegan Paul, 1983), esp. ch. 3. 56 DAVID MILLER regardless of the preferences that people actually display. If we are going to condemn competition, say, or the kinds of goods produced for consumption in market economies, we must be able to deploy some theory of human good which allows us to make the necessary discrimination. But the contemporary intellectual climate is very hostile to any such theory. The high-minded assumptions about the nature of the good life that socialists made a century ago - as indeed did many liberals, most notoriously John Stuart Mill - now strike us as elitist and somewhat pious. We are far more self-critical in the matter of elevating our own preferred mode of life to the status of universal truth. And while most socialist intellectuals privately persist in their taste for improving literature, healthy hikes in the country, and political discussion as forms of recreation, they are far less keen to have those predilections held up as the image of socialism itself. The current preference is, if anything, for 'designer socialism,' that is a view of socialism that warmly allies itself to current fashions in clothes, music and life-style generally. It is clearly impossible to celebrate modes of consumption thrown up by present-day capitalism while at the same time holding on to a view of socialism that embodies a strong "quality of life" critique of that very system. These are two of the pressures, therefore, that incline contemporary socialists to put forward a slimmed-down version of socialism, defined more or less entirely in distributive terms. Socialism, then, is exclusively a matter of allocating resources (broadly conceived) in the appropriate manner. As examples of this tendency in political philosophy, I would cite the following: Hillel Steiner's proposals for a laissez-faire economy grounded in equal entitlements to natural resources; 10 Ronald Dworkin's conception of equality of resources, including an insurance scheme to compensate for inequalities in personal endowments; 11 John Roemer's argument for equality of productive assets as the best way of capturing the point of the Marxian theory of exploitation;12 Robert Van der Veen and Philippe Van Parijs's advocacy of a "Capitalist road to Communism" whereby returns to labor and capital are taxed at progressively higher rates to provide each person with an unconditional grant to satisfy their needs.13 Although these 10 See Hillel Steiner, "The Natural Right to the Means of Production," Philosophical Quarterly, vol. 27(1977), pp. 41-49; "Slavery, Socialism and Private Property," eds. J. Roland Pennock and John W. Chapman, Norms XXII: Property (New York: New York University Press, 1970); "Liberty and Equality," Political Studies, vol. 29 (1981), pp. 555-69. 1 ' Ronald Dworkin, "Equality of Resources," Philosophy and Public Affairs, vol. 10 (1981), pp. 283345. 12 John Roemer, "Equality of Talent," Economics and Philosophy, vol. 1 (1985), pp. 151-87; "Should Marxists Be Interested in Exploitation?", Working Paper No. 221, Department of Economics, University of California, Davis. 13 Robert J. Van der Veen and Philippe Van Parijs, "A Capitalist Road to Communism," Theory and Society, vol. 15 (1986), pp. 635-55. IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 57 proposals differ in important respects, they share the aim of radically altering the distributive outcome of conventional capitalism, but without requiring any corresponding change in the quality of human relationships that prevail under that system. I should make it clear that I am broadly in sympathy with this tendency towards a justice-based socialism; in particular, I have been arguing for some time that markets are both an economically essential and an ethically acceptable component in a viable form of socialism.14 Nevertheless, I do not believe that socialists can discard everything in the "quality of life" critique of capitalism; they especially need to hold on to some form of community as an essential part of their vision. Which form of community is the major issue addressed in this paper. But first let me present the minimalist case for communitarianism of some kind as an ineliminable part of the socialist project.15 In presenting this case, I make two assumptions which I hope are not controversial. The first is that we want our version of socialism to be democratically suported: whatever view is taken about the transition to socialism, socialist institutions should command the willing assent of at least the majority of the population once those institutions are installed. The second is that the rules of the system (the rules governing economic transactions, etc.) should, for the most part, be complied with voluntarily: the level of coercive enforcement should be no higher than, say, that prevailing under present-day capitalism, and preferably it should be a good deal lower. Clearly, if such assumptions are to hold good, the socialist arrangements we 14 See David Miller, "Socialism and the Market," Political Theory, vol. 5 (1977), pp. 4 7 3 - 9 0 ; "Jerusalem Not Yet Built: A Reply to Lessnoffon Capitalism, Socialism and Democracy," Political Studies, vol. 28 (1980), pp. 584-89; "Marx, Communism and Markets"; David Miller and Saul Estrin, "Market Socialism: A Policy For Socialists," ed. I. Forbes, Market Socialism: Whose Choice?, Fabian pamphlet No. 516 (reprinted as "A Case for Market Socialism," Dissent, Summer 1987, pp. 359-67). These ideas are developed more systematically in a forthcoming book, Market, State and Community (Oxford: Clarendon Press, 1989). 15 1 call this the minimalist case because it hinges the argument for community entirely on elements drawn from the distributive critique; it makes no appeal to the inherent value of community. Now an argument of this kind might appear inherently paradoxical, at least insofar as it is addressed to the public at large. For either the addressees already see themselves as belonging to a community, or they do not. If they do, then it is redundant to offer them a justifying argument that appeals to extrinsic distributive considerations; if they do not, then a sense of community cannot be conjured out of thin air because it would be helpful from a distributive point of view were it to exist. Thus the argument that follows might seem to have an unavoidably esoteric character. We may, however, take the addressees of the argument to be people who both see themselves as members of a community and espouse principles of distributive justice, but who as yet see no necessary relationship between these commitments. Community membership is felt to be inherently valuable, but irrelevant from a distributive point of view. The purpose of my argument is to enhance the value of community by connecting the two commitments. This has a practical point insofar as we are now in a position to make political decisions that will influence the nature of our community in the future - strengthening or weakening people's allegiance in the long run. 58 DAVID MILLER have in mind must be legitimated, in the sense that most people must hold a conception of justice that corresponds to the one that these arrangements embody. For instance, insofar as the arrangements we envisage rest on a conception of equality, the people subject to them must be, or become, egalitarians. Now, on one view of the matter, this question of legitimation poses no real problem. If we think that good, rational arguments can be given for the view of justice that we favor, then we may believe that most people - given sufficient time, perhaps - will come to share this view, and so the appropriate conception of justice can be developed apart from, and prior to, the arrangements of socialism itself. In particular, the view of justice people hold doesn't depend on the kind of relationships they have with those around them, so the order of decision goes as follows: first, a consensus emerges on justice itself; second, this consensus is embodied in institutions which allocate resources to people; third, people use the resources they have been allocated to pursue their personal goals, perhaps including the goal of developing relationships of particular kinds. This, very crudely, is how John Rawls sees the matter, 16 and his idea has been influential among many, including some socialists who interpret the substance of justice in a more egalitarian way than he does himself. I hold this view to be badly in error. Against it, I want to argue that our ideas of distributive justice are powerfully affected by our perception of the relationships generally prevailing in the set of people within which the distribution is going to occur. Perhaps this is best illustrated by starting with an extreme case. Suppose we conceived of a 'society' made up of individuals who had no social relations with one another, each living an entirely independent and self-contained life - a set of Robinson Crusoes, each on his own island. What would justice mean to the inhabitants of such a 'society' ? They would endorse something like Nozick's view that justice means noninterference with the rights that each has acquired by his own legitimate efforts (where "legitimate" is in turn spelled out in terms of noninterference). Charles Taylor has put this point well: ... there is a mode of justice which holds between quite independent human beings, not bound together by any society or collaborative 16 This is not the place to discuss the finer details of Rawls's theory. He expresses some concern about what he calls the "strains of commitment" - the possibility that people might no longer be able to accept the principles of justice they have endorsed in the abstract when faced with their concrete results - but he sees this as a problem about justice and material interests: can people who do badly out of the application of a theory of justice be expected to continue embracing it ? (See John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971), esp. section 29.) He does not raise the question whether the practical acceptability of a theory of justice might not depend on the quality of social relationships in general. IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 59 arrangement. If two nomadic tribes meet in the desert, very old and long-standing intuitions about justice tell us that it is wrong (unjust) for one to steal the flocks of the other. The principle here is very simple: we have a right to what we have.17 Conversely, any view that is more redistributive than Nozick's - any view holding that people can make claims on one another that go beyond simple non-interference - must presuppose a background set of social relationships against which claims of this sort would appear legitimate.18 This, indeed, is the nub of Michael Sandel's criticism of Rawls.19 Rawls advances a distributive principle - the difference principle - which gives people a claim on what others have produced by exercising their talents and skills, but says nothing about the communitarian relationships which, in practice, would be needed to underpin this principle. What applies to Rawls applies a fortiori to principles of distributive justice that are more strictly egalitarian. These principles may deprive some asset-holders of large quantities of the holdings they would have enjoyed under the "Nozick Constitution" which we are using as a benchmark. We can only expect them to consent to institutions that enforce the preferred distribution if they regard themselves as bound to the beneficiaries by strong ties of community: the stronger the ties, the more egalitarian the distribution can be. 20 It is not an adequate answer to this line of thought to say that a distributive practice can, of itself, create the necessary ties. No doubt there is a process of reinforcement such that implementing a practice of distributive justice appropriate to a particular community will tend to buttress the sense of community that already exists. But if, starting from the Crusoe 'society,' an external agency were to impose an egalitarian redistribution of assets but do 17 Charles Taylor, "The Nature and Scope of Distributive Justice," Philosophy and the Human Sciences: Philosophical Papers II (Cambridge: Cambridge University Press, 1985), p. 289. 1s Let me stress that I am concerned here about the conditions under which a socialist system of distribution could be legitimate, in the sense of being congruent with widely-held and spontaneouslyformed notions of justice. I am not directly concerned with the transition to socialism, i.e., with the circumstances under which those who are the chief beneficiaries of capitalism would be willing to renounce the privileges they already enjoy. Although democratic socialists will want both the transition and the ensuing arrangements to have broad popular support, it would be unrealistic to set the standard of consent as high for the former as for the latter. 19 See Michael Sandel, Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982), ch. 2. 20 Obviously, this is not a claim about logic but a claim about social psychology. Although social psychologists cannot create genuine communities in the laboratory, their simulations provide some support for the claim. In particular, people give less weight to merit and more weight to equality in distribution when they expect to interact with their partners over a period of time. See E. Gary Shapiro, "Effect of Expectations of Future Interaction on Reward Allocation in Dyads: Equity and Equality," Journal of Personality and Social Psychology, vol. 31 (1975), pp. 8 7 3 - 8 0 ; Melvin J. Lerner, "The Justice Motive: 'Equity' and 'Parity' among Children," Journal of Personality and Social Psychology, vol. 29 (1974), pp. 539-50. 60 DAVID MILLER nothing else to change relationships between the Crusoes, I can see no reason to expect that they will begin to think of themselves as forming a community or to regard the redistribution as legitimate. The kind of ties we are looking for are not external and mechanical, but involve each person seeing his life as part and parcel of the life of the wider group, so that the question of how well his own life is going depends in some measure on how the community as a whole is faring. This brings in issues of common good, historical identity, and so forth which reach far beyond the scope of distributive justice. Rawls's notion that adherence to a shared conception of distributive justice could itself form a sufficient basis for community is quite implausible.21 I have not yet said anything about the idea of community that socialist proposals require, or about whether such an idea is feasible given the circumstances of an advanced industrial society. My argument so far is simply that a form of socialism which defines itself primarily in terms of distributive justice must still consider questions about the quality of social relationships if it wants to be something more than a nice intellectual construct. If socialism is to be politically feasible - if socialist arrangements once installed are to command the willing consent of the population - social relations generally must support the preferred conception of distributive justice. Even those who regard community as having no independent political value must rely on it in practice to underpin their distributive concerns. So at this point I turn to see whether there is anything in the recent communitarian literature that might be helpful to the socialist case. II. T H E AMBIVALENCE OF CONTEMPORARY COMMUNITARIANISM I shall focus on the work of Alasdair Maclntyre, Charles Taylor, and Michael Sandel. Each of these writers would endorse the general argument I have just advanced: namely, that ideas of justice cannot be separated from a broader understanding of the community within which distributive practices exist.22 It does not follow, of course, that the conceptions of justice and community they advance are socialist conceptions: indeed, in at least two out of the three cases, the evidence is rather to the contrary.23 So our approach to their work must be a critical one. We can try to get clearer about socialist ideas of community, in part by seeing where they need to diverge from the ideas of Maclntyre, Taylor, and Sandel. 21 Rawls, A Theory of Justice, section 79. See Alasdair Maclntyre, After Virtue (London: Duckworth, 1981), ch. 17; Taylor, "The Nature and Scope of Distributive Justice"; Sandel, Liberalism, ch. 4. 23 Only Taylor seems in any way sympathetic to socialist views, and even he is mainly concerned to present socialism as trapped in the same modernist predicament as other outlooks: see Charles Taylor, "Socialism and Weltanschauung," eds. Leszek Kolakowski and Stuart Hampshire, The Socialist Idea (London: Quartet, 1977). 22 IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 61 Maclntyre's understanding of community initially derives from his narrative view of the self. A person can only make sense of his actions, he argues, by placing them within a narrative structure - a self-told story which runs through the person's life from birth to death. The narrative I construct for myself, however, intersects with other people's narratives - I am a character in their stories and they figure in mine. This set of narratives, in turn, makes references to the wider communities within which the individuals in question bear social identities - say, as the occupants of kinship roles or as members of institutions. In particular, Maclntyre argues, moral activity itself depends upon an understanding of moral value which can only be provided within a particular community. Maclntyre's claim, to sum up, is that people can only make sense of their lives by seeing themselves as members of large communities, which above all provide the preconditions of narrative unity. As to the scope of community, Maclntyre remains agnostic: he talks of "the family, the neighborhood, the city, and the tribe," and so forth.24 What he does make clear, in contrast, is that the modern state is not an appropriate location for community. It cannot embody community, since in Maclntyre's view that would require a moral consensus at the political level which simply does not and cannot exist. Moreover, the mode of operation of the modern state tends to destroy such communal ties as still exist: bureaucratic procedures create individuals who are abstracted from any social identity, and whose residual sense of self is that of a pure chooser of ends. In his remarks on the Jacobins, Maclntyre does not exclude the possibility of an understanding of moral virtue that would give political participation a central place, but he argues that such an understanding is impossible to sustain in modern conditions. "The true lesson of the Jacobin Clubs and their downfall is that you cannot hope to reinvent morality on the scale of a whole nation when the very idiom of the morality that you seek to re-invent is alien in one way to the vast mass of ordinary people and in another to the intellectual elite." 25 So although patriotism - a special regard to the interests of my national community remains a virtue for Maclntyre, it is no longer an idea which should inform my relations to the government of the day.26 I should assess the state purely in instrumental terms: in terms of how effectively it keeps the peace between different communities, protects rights, and so forth. It would be wrong to describe this outlook as politically conservative since, 24 Maclntyre, After Virtue, p. 205. ibid., p. 221. 26 See Hid., pp. 236-37 and Alasdair Maclntyre, "Is Patriotism a Virtue ?", Lindley Lecture, University of Kansas, 1984. 15 62 DAVID MILLER as Maclntyre notes, it involves a complete rejection of "modern systematic politics, whether liberal, conservative, radical or socialist";27 it can fairly be described, however, as morally conservative, in the sense that it defines moral virtue in terms of the traditions of such defacto communities as are salient for each particular person. Although Maclntyre, in his remarks on tradition, argues that living traditions always involve critical argument about the best way of carrying the tradition on, there is no wider forum within which the merits of different traditions might be debated - and, of course, Maclntyre rejects the idea that there are transcendent standards of justice which might be used to adjudicate between them. 28 From a socialist perspective, therefore, Maclntyre's view of community must appear dangerously tradition-bound, not so much in the sense that it starts with defacto communities as in the sense that it has no resources for getting beyond the notions of virtue embodied in each community except insofar as the community itself engages in critical reflection. I shall later connect this deficiency to Maclntyre's other view that political arrangements are irrelevant (in modern conditions) to communal life. If we turn now to Taylor, we find once again that claims about the nature of personhood are advanced to underpin a (fairly unspecific) commitment to community. There is, however, this contrast with Maclntyre: whereas Maclntyre bases his account of the self on a pre-modern understanding of morality, Taylor assumes from the outset "the modern identity" - that is, a view of human nature which breaks decisively with the idea that human fulfillment can be understood as alignment with some given, cosmic order (as in older Christian views). Taylor accepts the modern idea that human selfrealization involves choice as well as discovery; what he rejects are individualist accounts of that process of self-realization. As far as I can discern, there are two major strands to Taylor's argument; I am not clear as to how these strands are supposed to be related. The first strand starts from a familiar liberal ideal, the ideal of personal autonomy. Taylor's claim is that individualist liberals fail to understand the preconditions for autonomy. They see it as unproblematically given and needing only protection against external constraints, whereas in fact it requires a certain kind of cultural background. People can only make authentic choices about their own lives against the background of a civilization in which, for example, moral questions are debated in public, certain aesthetic experiences are available, and so forth. Community makes 11 Maclntyre, After Virtue, p. 237. For further reflection on the difficulties this entails, see my "Virtues and Practices," Analyse und Kritik, vol. 6 (1984), pp. 49-60. 28 IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 63 its appearance here in the guise of a common culture, participation in which is a necessary condition of liberal aspirations to autonomy.29 The other strand in Taylor's argument moves further away from liberal premises. Taylor points to the importance, in the modern consciousness, of an attitude which he calls 'expressivism.'30 This is the ambition to see the world around us as an expression of our authentic nature, an idea that was particularly prominent in the thought of the German Romantics and their followers. Now this attitude can take either private or public form, depending on whether the 'nature' to be realized is the essence of a particular individual or a nature common to the members of some collectivity. Taylor, however, regards private versions of expressivism as somehow deficient: he points to the nuclear family as the current embodiment of the Romantic ideal of a life according to nature, but claims that the family lacks the moral resources to contain narrow self-absorption. . . . if the business of life is finding my authentic fulfilment as an individual, and my associations should be relativized to this end, there seems no reason why this relativization should in principle stay at the boundary of the family. If my development, or even my discovery of myself, should be incompatible with a long-standing association, then this will come to be felt as a prison rather than as a locus of identity.31 Hence, expressivist aspirations can only be fulfilled if we can discover some common identity that might be expressed in a public world. Where might such an identity be found ? Taylor finds the answer in language, an institution that is necessarily public, and at the same time embodies a distinctive way of experiencing the world. The community here becomes the speech community. In talking to one another, we convey a shared view of the world: and this view is constantly open to modification as we change our language in order to express ourselves more adequately. Taylor summarizes: . . . the expressive conception gives a view of language as a range of activities in which we express/realize a certain way of being in the world. And this way of being has many facets. It is not just the reflective awareness by which we recognize things as - , and describe our surroundings: but also that by which we come to have the 29 See Charles Taylor, "Atomism," Philosophy and the Human Sciences: Philosophical Papers II (Cambridge: Cambridge University Press, 1985). 30 See Charles Taylor, / / ^ / ( C a m b r i d g e : Cambridge University Press, 1975), esp. chs. 1 and 2 0 ; "Legitimation Crisis?", in Philosophy and the Human Sciences: Philosophical Papers II (Cambridge: Cambridge University Press, 1985). 31 Taylor, "Legitimation Crisis?", p. 283. 64 DAVID MILLER properly human emotions, and constitute our human relations, including those of the language community within which language grows.32 So we find two "communitarian" trains of thought in Taylor, one beginning with the liberal ideal of autonomy and ending with a view of community as common culture, the other beginning with expressivism and ending with a view of community as language-sharing. Rather than investigate the relationship between these conceptions, I want to point to three traits which they have in common. First, on either view, the scope of the relevant community is exceedingly difficult to determine. How does one attempt tofixthe boundaries either of a cultural or a linguistic community ? On Taylor's first argument, there seems no reason to restrict the scope of community at all, since, presumably, the greater the cultural variety to which I am exposed, the more chance I have to develop my capacity for autonomous choice. The second argument does seem to imply a more particularistic view of community, but then we run into the familiar difficulties of individuating languages (is American-English the same language as English-English?). Taylor's communitarianism is unavoidably amorphous, and this immediately limits its power as a weapon in political philosophy. Second, Taylor's view has the consequence of restricting the extent to which our communal relations are open to rational reflection. The ties that hold us together must, to some degree, remain opaque to critical investigation. Taylor makes this explicit when he describes languages as A pattern of activity . . . which can only be deployed against a background which we can never fully dominate; and yet a background that we are never fully dominated by, because we are constantly reshaping it. Reshaping it without dominating it, or being able to oversee it, means that we never fully know what we are doing to it; we develop language without knowing fully what we are making it into. 33 Now it may turn out that communal relationships are indeed unavoidably opaque in the way that this view of language suggests; this is a matter requiring further investigation. But, pritnafacie, it undermines the belief held by some socialists, most notably Marx, that what we should be aiming for is a society in which human beings consciously and collectively control their destiny, and their relationships become entirely transparent. Community, on 32 Charles Taylor, "Language and Human Nature," in Human Agency and Language: Philosophical Papers 1 (Cambridge: Cambridge University Press, 1985), p. 234. 33 ibid., p. 208. IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 65 Taylor's view, is something that we are immersed in, but whose nature cannot be grasped fully and consciously, let alone shaped completely according to our will. Anxieties about the potentially conservative character of Taylor's conception may be heightened when we notice that community has, for him, no necessary political dimension. It is true that he attempts to build a political argument onto his claim about the role of common culture in providing the conditions for autonomy. The logic of this argument must be that cultural forms and institutions are public goods, and it is unlikely that self-interested individuals will provide them voluntarily. That this is an empirical claim is made clear by Taylor's reply to the anarchist who thinks that these goods will be created spontaneously. "There is nothing in principle which excludes anarchism in the reflection that we owe our identity as free men to our . civilization."34 Political institutions, then, appear simply as instruments for protecting the elements of community by, for instance, subsidizing the arts. They are not part of the framework of community itself. Maclntyre sees politics as irrelevant to community; Taylor sees it playing only an instrumental role. Sandel is, in contrast, more firmly set in the republican tradition. When he speaks of community, he appears to envisage a set of people engaged in, among other things, political deliberation. Corresponding to this is a stronger claim about the importance of community to personal identity. Maclntyre and Taylor both argue, in their different ways, that people's identities can't be satisfactorily defined without communal relationships in the background. Sandel's claim is that identity must, in part, be constitutedby communal attachments. When people discover who they are (discover, not choose), part of what they discover is that they are members of this or that collectivity - a membership which they cannot relinquish without becoming different people in one important sense. What is the status of Sandel's claim here? He presents it not as a description of present-day reality but as a presupposition of liberal theories of justice, especially the Rawlsian theory. His core argument is that the difference principle, which treats people's talents and abilities as common assets, can only be acceptable if we adopt the constitutive conception of community outlined above. He is far less sanguine about the practical feasibility of such a conception. In his brief remarks on the evolution of American politics, he traces a process whereby, starting with local political communities (which, by implication, were genuine communities), politics was progressively "nationalized" in response to economic pressures; but the 34 Taylor, "Atomism," p. 207. Taylor later adds to his argument the claim that political deliberation forms an essential part of freedom, but this has the appearance of an afterthought. See ibid., p. 208. 66 DAVID MILLER attempt to foster a new sense of community at this level was unsuccessful. "Except for extraordinary moments, such as war, the nation proved too vast a scale across which to cultivate the shared self-understandings necessary to community in the formative, or constitutive sense."35 Hence what we are left with is a liberal politics of rights that lacks a coherent communal underpinning. What lies behind this conclusion ? Sandel must be assuming that a sense of community strong enough to foster constitutive attachments can only exist where people have face-to-face relationships (as in the traditional town meeting) or perhaps have strong cultural affiliations. Clearly, a large modern nation cannot expect to meet either of these conditions. So the upshot of Sandel's argument must be uncongenial to the socialist. Although the link between community and politics is forged, community is seen as a phenomenon of localities, and this cannot satisfy the socialist, who needs it to underpin distributive justice across whole societies.36 By strengthening the conditions for community, Sandel at the same time precludes it from playing the kind of political role that socialists want it to play. None of the three 'communitarians' we have considered advances a conception of community that seems well suited to socialist purposes. What lessons might we draw from this fact ? For all three authors, the very idea of community is problematic in the modern world. Maclntyre sees us as clinging to fragments of community inherited from the pre-Renaissance period. Taylor sees communitarians as fighting a rear-guard battle against what he calls "Enlightenment naturalism" - a view of man as an agent who regards the surrounding world merely as an instrument to the optimal fulfillment of his freely-chosen desires. Sandel, as we have just noted, is pessimistic about constitutive community in the face of the scale of modern politics. Now one lesson we might draw is to take these authors' findings as confirmation of a certain negative view of socialism, a view which sees the socialist project as anachronistic from the very moment of its conception. On this view, socialism became a popular ideology precisely in response to the breakup of traditional communities under the impact of the industrial revolution. It became popular because it promised to restore the coherent ; \ i 35 Michael Sandel, "The Procedural Republic and the Unencumbered Self," Political Theory, vol. 12 (1984), pp. 81-96. 36 If communal relationships foster a sense of justice that is relatively egalitarian, why shouldn't a society made up of small local communities develop a society-wide scheme of redistribution? Unfortunately, there is no reason to expect the scope of a practice of distributive justice to extend beyond the community that supports it. There is ample historical evidence of small communities (tribes, guilds, cooperatives) practicing quite radical forms of egalitarian redistribution internally, but dealing with outsiders on very different terms. Socialists must look for community at the level at which effective policies can be made for whole societies, which, in practice, means at the level of the nation-state. j \ . j ] j j j IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 67 moral life found in the disappearing communities, while at the same time providing all the material (and other) benefits of industrialization. But these two promises could never be fulfilled together. In industrial societies the appeal to community is always nostalgic and backward-looking, whatever its proponents may think. This is not the lesson I want to draw, though I feel the force of the argument in the last paragraph which, as noted, reflection on the ideas of Maclntyre, Taylor, and Sandel tends to support. We see, once again, the attractions of a purely distributive view of socialism, which aims to discard community as an essential element in the socialist framework. On this view, particular communities may flourish under a socialist distributive regime, but this is, so to speak, an optional extra, not something that the regime itself requires. I have explained already why this attractive view cannot, in the end, be maintained. Socialists must take up the quest for community, but they should do so in a chastened spirit, in full realization of the obstacles that lie in their path. In the final part of the paper, I sketch in a socialist view of community which I hope meets these strictures. III. A SOCIALIST CONCEPTION OF COMMUNITY A socialist who wants to avoid the charge that he is merely nostalgic for pre-industrial forms of life cannot appeal to thick-textured, face-to-face community as the building block of his system. Where such communities still exist - as, for instance, in certain mining villages or other places where a fixed pattern of working life has persisted over decades - there is no reason to disparage them; equally, it would be wrong to make them integral to the socialist project, or in particular to suppose that the whole of a society could come to take on the character of these local communities. The tendency of an industrial economy is always to erode community in this strong sense, and, whether or not one thinks that economic policies should be designed to protect particular communities against such erosion, there is nothing here firm enough to support a socialist project. It would be equally wrong to conclude that socialist community must comprise nothing less than humankind in general. Some socialists do seem to take it as their aim to extend fraternal sentiments to embrace all other human beings, and in the course of so doing to sweep away all local and particular attachments as relics of an unsavoury past. This aim is often thought to embody a certain idea of rationality. There is no good reason to treat our neighbors or our compatriots any differently from equally needy people elsewhere in the world, so any sense that we owe them special loyalties must vanish under critical scrutiny. This view neglects the fact, evident enough in the writings of the communitarians we have considered, that communities just are particularis- 68 DAVID MILLER tic. In seeing myself as a member of a community, I see myself as participating in a particular way of life marked off from other communities by its distinctive characteristics. Notions such as 'loyalty' and 'allegiance' make no sense unless there is an identifiable something towards which these attitudes are directed. Moreover, to say that such attitudes are necessarily irrational is to adopt a contestable view of rationality, one that presents it as a property of the deliberations of a detached subject reasoning entirely from universal principles. Socialists need not and should not take up the view of ethical rationality implied here. They are better served by a form of ethical particularism that allows existing commitments and loyalties a fundamental place in ethical reasoning - which does not entail that every commitment must be accepted uncritically.37 They should, in short, prefer the Hegelian idea of ethical life to the Kantian idea of morality as an account of practical reasoning vis-a-vis other members of the community. There is a further implication to be drawn here. A realistic form of socialism must start out from actually existing communities. It cannot hope to invent the communities that it might be thought desirable to have on abstract grounds. Communal relationships, for reasons already given, are inevitably fragile in the circumstances of a modern industrial society; they must be husbanded with care. Now if we take this point together with the earlier point that community is needed to legitimize practices of distributive justice, our conclusion must be that the relevant communities are nations. On the one hand, people do, in general, identify themselves with national communities in a way that they rarely do with wider constituencies. On the other hand, community at this level is normally broad-based enough to provide the conditions for an effective practice of distributive justice. The nation as a form of community must have a privileged position in socialist thought, at least in any future we can envisage.38 And this, of course, runs directly counter to a well-entrenched tendency in socialist thought, which regards nationality less as a resource than as a problem to be overcome. For many socialists, the future has seemed to lie either with local community or 37 1 discuss ethical particularism more fully in "The Ethical Significance of Nationality," Ethics, forthcoming. See also Andrew Oldenquist, "Loyalties,">Br«a/ of Philosophy, vol. 74 (1982), pp. 17393; John Cottingham, "Partiality, Favoritism and Morality," Philosophical Quarterly, vol. 36 (1986), pp. 3 5 7 - 7 3 ; Philip Pettit, "Social Holism and Moral Theory," Proceedings of the Aristotelian Society, vol. 86 (1985-86), pp. 173-97. 38 There is no need, I think, to commit ourselves on the question of whether it is ultimately preferable to have a world order made up of distinct national communities or a global community. On one side stands the value of diversity; on the other, the problems of international distributive justice. The point is that the most extensive communal identities that people currently have are national identities, and there is no sign that this about to change. Insofar as there is any movement, it appears to be in the direction of smaller, more intense forms of nationality rather than towards internationalism. IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 69 with global community, or with some combination of these, but in any case not with existing nationalities. Why is there such a resistance in socialist thought to the idea of national community? There is the belief in moral universalism which I have already discussed (national boundaries are morally irrelevant). There is also, of course, the horrific experience of "national socialism" in its German incarnation. But there is a further point. If one starts out by thinking of a community as involving a set of face-to-face relationships in which each person has full and direct knowledge of the qualities of the other members, then the idea of a nation as a community must indeed seem peculiar. Nations are, in Benedict Anderson's phrase, "imagined communities."39 They exist only because of beliefs each of us have about our compatriots, beliefs not acquired by direct experience but culturally transmitted. As Anderson points out, nations can't exist without mass media (originally the printed word) to disseminate an understanding of national identity. But with this comes the possibility of distortion. The picture of national life that becomes embedded in the culture may not accurately reflect what is actually the case, both now and historically. The extreme version of this is, of course, Orwell's 1984, where the telescreens project a version of events that bears no relation at all to reality. But even in societies that are much more open than Orwell's Oceania, we find that national identities contain a greater or lesser admixture of myth. And this apparently contravenes an idea of rationality that the socialist must find attractive - not the abstract idea of reason that I have already rejected, but simply the idea that, to be rational, I must regard all of my beliefs as potentially subject to critical scrutiny. It seems that if I adopt this policy with respect to nationality, I am bound to end up by rejecting a good part of my existing national identity. I have elsewhere tried to assess how far national identities can survive the critical rejection of certain of their components.40 Here I want to develop a different point. There may be built-in limits to the process of critical scrutiny itself. We may simply not be able to formulate everything that goes to make up our nationality in a way that makes rational scrutiny possible. Here we see the significance of Taylor's conception of language as "a pattern of activity... which can only be deployed against a background which we can never fully dominate"; it is a conception which can, I think, be extended more widely to apply to many of the cultural phenomena which constitute nationality. Consider, for instance, the importance of symbols of various kinds - flags, emblems, festivities - in national life. When we respond to these symbols, as 39 Benedict Anderson, Imagined Communities: Reflections on the Origin and Spread of Nationalism (London: Verso, 1983). 40 See "The Ethical Significance of Nationality." 70 DAVID MILLER most of us do, we cannot spell out in propositional form precisely what it is we are responding to, although we may, of course, be able to say something about their significance. What passes through our minds as we stand before the Cenotaph to commemorate our war dead ? Is it pride in the heroism of our soldiers, or horror at the carnage of war ? Or perhaps both of these at once ? If we can't be clear about what the ceremony means, how can we say whether it is rationally acceptable or not ? Yet occasions such as these are an important component of national identity. When assessing Taylor's view that communal relationships are unavoidably opaque, I pointed to its potentially conservative implications. Unquestionably, nationality is open to the same charge. If we define our community in terms of a spontaneously evolving national identity, we shall remain, to a large extent, the prisoners of our past. To avoid this implication, we must appeal to politics. The political forum must be the sphere in which we reshape our common identity through reasoned argument. A socialist view of community must therefore give a central place to citizenship. Not only should we be related as bearers of a common national identity, we should also be related as citizens, as co-determiners of our collective future. Let me expand a little on this. It is important to see how nationality and citizenship are related in the view that I am outlining. Nationality is the identity we have in common, an identity in large measure inherited from the past, and not fully open to rational scrutiny. Citizenship is a political status which allows each of us to participate in reshaping that identity. For instance, we scrutinize our institutions and practices to see whether the meanings they convey (so far as these are determinate) are meanings we still want to endorse (to take a relatively trivial case, we may decide to abolish one public holiday and institute another); we decide which cultural activities are worthy of public support; more generally, our legislation may involve an attempt to influence future understanding of the meaning of membership in this society (consider the case of race-relations legislation). But this exercise never occurs in a vacuum. We take part in political debate already endowed with the shared understandings that come with a common nationality. Since critical reshaping starts with these understandings, we cannot get beyond them entirely, or certainly not beyond them all at once. Taylor's metaphor, borrowed from Neurath, of sailors rebuilding their ship at sea, seems to describe the case pretty well. What must politics be like if it is to fulfill the function we are assigning to it? Socialists will certainly want to insist that citizenship must be a role available to everyone (or else the reshaping of communal identity won't be democratic), so we must envisage arrangements that permit everyone to be politically active. More significant for the present argument, people must engage in politics as citizens, that is, as members of a collectivity committed , I { ', ] ] IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 71 to advancing its common good. They cannot enter it in their capacity as private persons, each with a particular interest - say, an economic interest to advance. This is a formidable requirement, as Rousseau understood in his pessimistic chapter on the silencing of the general will.41 Socialism must draw on an ample stock of republican virtue. It must also become adept at what Michael Walzer has called the art of separation.42 It must find ways of demarcating a person's role as citizen from other roles that he might perform, for instance, as enthusiast for a particular cause, or as spokesman for a sectional interest. A modern society will inevitably embrace a whole gamut of forms of private life (unless they are artificially suppressed), and these are bound to generate demands on public policy. There is always a danger that the force of these demands may obliterate citizenship. The sailors who are rebuilding their ship on the open sea may need to refurbish their tools at the same time. This may indicate the distance between the form of communitarianism developed here, and the simpler view which sees community as a general, undifferentiated characteristic of relationships in socialist society. Community, on the present view, has a restricted character. It describes one respect in which members of a society may be related, but it does not exclude the possibility that they may also be related in other ways - say, as competitors in the marketplace. It also has an artificial character, at least to some extent.43 I have argued that national identities will remain partially opaque to rational scrutiny; I have also argued that we may need devices - symbolic devices and so forth - to protect citizenship from invasion by private interests. This takes us very far from a view of community as the expression of natural sentiments, and of social relations as transparent to the participants. It is not clear to me, however, that socialists must disdain artifice, if that turns out to be the best way of achieving their essential goals. I draw comfort here from Jon Elster's recent discussion of the problem of constitutional choice, which takes up the idea that a constitution is a device for collective self-binding - a way of protecting ourselves from making certain kinds of decisions in the future and applies it to the transition to socialism.44 Socialists should discard the 41 "In the end, when the state, on the brink of ruin, can maintain itself only in an empty and illusory form, when the social bond is broken in every heart, when the meanest interest impudently flaunts the sacred name of the public good, then the general will is silenced: everyone, animated by secret motives, ceases to speak as a citizen any more than as if the state had never existed; and the people enacts in the guise of laws iniquitous decrees which have private interests as their only end." Jean-Jacques Rousseau, The Social Contract (Harmondsworth: Penguin, 1968), p. 150. 42 Michael Walzer, "Liberalism and the Art of Separation," Political Theory, vol. 12 (1984), pp. 315-30. 43 I use "artifice" here in its neutral, Humean sense. 44 Jon Elster, Three Lectures on Constitutional Choke, (mimeo) Oslo, 1981. 72 DAVID MILLER naive view that everything they want can be achieved by following majority opinion at each moment in time. CONCLUSION This paper began with the claim, made recently by several socialists, that socialism centers on the value of freedom - of ensuring that each person enjoys equal effective freedom. In pursuing the implications of that claim, we may seem to have turned a somersault, for there is a well-established view that sees the politically organized community as essentially totalitarian in its upshot. If we are to concern ourselves with our collective identity, and use politics as a means of remodeling that identity, what place is left for the freedoms that liberals characteristically cherish: artistic freedom, religious freedom, privacy, and so forth? How can socialism with a communitarian face possibly claim to be freedom-maximizing? The argument for that claim, to recall, runs as follows. Freedom depends on the distribution of resources. To equalize effective freedom, we need a system of distributive justice. But such a system can't be legitimized unless people see themselves as tied together communally. Politics enters the picture to prevent communal ties becoming merely traditional, to honor socialist demands for rationality. Such a view does not entail a rejection of liberal forms of freedom, or of practices such as the creation and enforcement of rights which protect those freedoms. Communitarianism, in general, may differ from liberalism more radically in its basic premises than in the practical policies it recommends ; 45 this applies a fortiori to the streamlined version defended here. That is not to say that a socialist set of rights would have precisely the same content as the standard liberal set. Socialism requires the introduction of new rights, especially in the field of economics. Equally, it may require the abrogation of some liberal rights, particularly in areas in which forms of private culture threaten to have a destructive impact on the public culture which sustains a common identity.46 So, for instance, a socialist society may wish to impose some limits on educational freedom, seeing the school as an important source of the political understandings that future citizens will bring to their public life. This may have implications both for the structure of the education system (should private schools be permitted ?) and for the content of what is taught in certain fields (should teachers be permitted to transmit any version of the history and politics of their own country ?). Where two sets of rights * 5 For an attempt at conciliation, see Gutmann, "Communitarian Critics of Liberalism." See also the brief discussion in Michael Sandel, "Morality and the Liberal Ideal," The New Republic, May 7, 1984, pp. 15-17. 46 I have grasped this nettle in "Socialism and Toleration," ed. Susan Mendus, Justifying Toleration (Cambridge: Cambridge University Press, 1988). IN WHAT SENSE MUST SOCIALISM BE COMMUNITARIAN? 73 intersect but neither includes the other, there is the familiar problem of deciding which is the more extensive - in the present context, the problem of aggregating liberties. Although I have not tried to show here that the solution to this problem must favor the socialist case, I can see no reason why it should not. In that sense, there is nothing incoherent in beginning one's intellectual trajectory as a socialist from a commitment to freedom and ending up at the circumscribed form of communitarianism I have been delineating in this paper. Politics, Oxford University