No Job Name Justice and Solidarity: The Contractarian Case against Global Justice David Heyd The Problem: Bootstrapping Charity begins at home, they say. But of course this does not mean that it ends there. It can and should extend way beyond home. The same might be said of respect for persons and their human rights. However, as I wish to argue in this article, justice begins at home and ends there. Or to put it in more theoretical terms, some fundamental forms of justice, particularly distributive justice, are essentially domestic. If that proposition is correct, some recent attempts to estab- lish principles of global justice are conceptually misguided. Theories of justice in the past have been relatively mute as to the identity and the scope of the groups to which the principles of justice apply. The subject groups were often simply presupposed as given. Take classical Greek principles of justice: they were typically meant to regulate the life of the citizens of the polis, a well-defined group of people, consisting mainly of adult, free, Greek males. Minors, slaves, “barbarians” and women were thus left out, beyond the reach of the benefits (and sometimes the burdens) of “just treatment.” All in all, there was hardly any attempt to theoretically justify the boundaries of the relevant group within which justice held sway. The urge to extrapolate the idea of justice to the international sphere is, as its intergenerational analogue, a modern phenomenon (with the possible exception of Christian theology). Both attempts are associated with historical developments regarding the unprecedented control human beings have gained over the welfare of other people living far away, either in location or in time. Like birth control and environmental policies, which enable us to deeply affect the lives of future generations, mass movements of people, money, and ideas in our age of “global- ization” make us much more responsible than ever in the past for the conditions of people on the other side of the globe. However, such attempts at global and intergenerational extensions of justice face serious theoretical obstacles. In this article, I shall leave aside justice to future generations1 and focus on the international parallel, which has also received much attention, particularly since Kant’s On Perpetual Peace. And I will do so from a contractarian approach, since individualism as a metaphysical assumption and the centrality of the will as a normative basis for political legitimation have become trademarks of liberalism and democracy. JOURNAL of SOCIAL PHILOSOPHY, Vol. 38 No. 1, Spring 2007, 112–130. © 2007 Blackwell Publishing, Inc. The problem is strikingly simple: Who takes part in the contract that serves as the basis for either the legitimation of the state’s authority or the justification of the principles of justice? Locke’s answer is that when any number of Men have, by the consent of every individual, made a Community, they have thereby made that Community one Body, with a Power to Act as one Body, which is only by the will and determination of the majority.2 Hobbes’s answer is similar to Locke’s: A Common-wealth is said to be Instituted, when a Multitude of men do Agree, and Covenant, every one, with every one, that to whatsoever Man, or Assembly of Men, shall be given by the major part, the Right to Present the Person of them all . . .3 Locke and Hobbes show how an indeterminate number of people can, through the very act of a covenant, become a cohesive group, referred to as “community” by Locke and “commonwealth” by Hobbes. The cohesion of the newly created group (as against the diffuse nature of “the multitude”) is expressed by the idea of the “body politick” in Locke and that of the one person representing all others in Hobbes. But both philosophers are completely silent about the process through which the individuals join each other to form that covenant. Is there any basis for the grouping of “any number of people” or “a multitude of men”? In other words, why should an individual (in the state of nature, or in any other imaginary original position) join forces to create a political community with these individuals rather than those? Even if the contract has the power to transform “a multitude of men” into a political community, what makes this multitude convene and make an agreement in the first place? Hobbes and Locke would surely respond that the question misses the purpose of the method of the social contract. Individual human beings have never actually lived outside a social community and hence the question of their choice to assemble in this or that way has never arisen. That is to say, individuals are usually born into human societies rather than choose to form or join them. The whole point of the thought-experiment of the social contract is justificatory, namely, to normatively ground the authority of the state retrospectively, or to offer a “post factum” test for its legitimacy and the justice of its institutions. We may guess that for Hobbes and Locke, England was the framework of the social contract as imagined by Englishmen, as was Geneva or France for Rousseau. As long as political communities were fairly homogeneous and relatively separated from each other, the issue of the scope of membership in the social contract could be left unaccounted for or assumed as given. Thus, despite the search for peace and security, believed by Hobbes to be the prime motivating force for the contract, the distinction between the national and international levels of political relations attests to Hobbes’s view that the contract is ultimately “local” in essence. Only in the “national” context is a contract possible; the international resembles the state Justice and Solidarity 113 of nature. But circumstances have changed since the seventeenth and eighteenth centuries and today’s world is characterized by much more heterogeneity in political communities and by a growing interdependence of countries and states. The issues of the identity and number of the parties to the social contract can no longer be brushed aside.4 But the new awareness of the problem of the scope of the contract leaves us in a logical quandary. If the state’s legitimacy (Hobbes and Locke) and the justification of the principles of justice (Rawls) are decided exclusively by a hypothetical contract and if the scope of the group within which this contract is formed cannot be simply taken for granted as given, it seems that the scope of the group must itself be decided by a contract. But the contours of the group engaged in a contract cannot be fixed by the contract itself. The fairness of inclusion or exclusion from the group of contracting parties cannot be determined by the contract, since it must precede it. This is merely a generalization of the well- known problems of real-life political decisions, like universal suffrage or seces- sion. Who should have taken part in the decision in England in 1918 whether women should get the vote—men only, or men and women? Or, who should vote on the proposal that Quebec secede from the Canadian federation—the Quebecois alone, or the whole Canadian electorate?5 Consider the two radical solutions to the question about the scope of the group of participants in the social contract: everyone or, alternatively, just one. Both are attempts to provide a universal basis to the contract, thus avoiding the pitfalls of controversial contingent restrictions on its range of application. However, if we say that everyone is a party to the contract, who exactly is included? All human beings in the world? Adults only? Are future people to be included in the group of “everyone”? Actual future people, or also possible future people? Should the group include also non-human rational beings as in Kant’s Kingdom of Ends? Clearly, these are substantive philosophical questions relating to the proper reference of “everyone” and they must be settled on a non- contractarian basis. Universal suffrage is a noble ideal, but universality itself is always contextual, relating to persons (or entities) of a certain kind. And this “kind” must be specified prior to the contract. The alternative of taking each (one) person, singly, as the proper scope of the method of legitimation by consent is an interesting idea, which was proposed by Rawls himself. The argument is well known: if we admit into the original position people who are all equally rational, mutually disinterested, free and under a veil of ignorance, the imaginary game of a contract among different people becomes superfluous, since all that is required is an individual’s acceptance of a set of principles of justice for society. There is no need for negotiation between identical individuals and hence there remains no point in a contract. However, the Rawlsian introduction of the veil of ignorance shifts the methodology of justification from the principle of consent to the principle of impartiality, deviating from the tradition of the social contract. It thus obviates the issue of the scope of the relevant group within which the principles of justice (or the legitimacy of sovereign power) apply.6 114 David Heyd I suggest that we return to the original idea of a contract between individuals—more than one but less than everyone. My argument is that even if fairness demands that the principles of justice be agreed upon under a veil of ignorance, we must assume an independent motive for people to enter the contract with those particular people rather than with others, a motive that cannot be hidden by that veil. Note that the existence of such a motive belongs to the conditions or circumstances of justice, not to the justification of the principles of justice. The motive only describes the formal requirement for the applicability of the proce- dure of justifying the normative principles. I therefore suggest that we reverse Locke’s and Hobbes’s view: consent does not create a community; it arises out of a sense of community. In American language, “We the people” cannot be defined by the Constitution, since it is this body of people that creates it. Justice: A Matter of Truth or Consent? Before further examining the problem of global justice, we should emphasize that it arises only within a contractarian point of view. Non-contractarian theories of justice have a long and venerable tradition. The fundamental issue separating the two approaches is whether justice is a matter of truth or a matter of consent; is it a property of the world or a human construction; or, whether it is “given” or a product of the will. The first, realist view of justice takes it as characterizing natural or metaphysical harmony, balance or order. It is associated with concepts such as desert, status, role, or proper functioning, and is accordingly grounded in specific descriptions of human essential nature as well as in social or normative ideals. The second, voluntaristic view of justice arises out of skepticism regarding the metaphysical assumptions about human nature and offers in their place a procedural account of justice as the product of the exercise of individual human will in concert with other individual wills under particular constraints (the ideal- ized conditions of the “state of nature,” or the fairness of the constructed “original position”). Some forms of modern liberalism, being aligned with metaphysical skepti- cism, are naturally biased toward a consent-based analysis of justice. Thus, in Hobbes’s theory, there is no application of the concept of justice in the state of nature. It is only after the sovereign is instituted that by his will justice is created. For Locke, even though moral norms exist in the state of nature, justice, in the sense of the actual distribution of property rights, is the outcome of contractual consent. And of course for Rawls, justice is not given in any metaphysical or realistic way, but is a matter for construction, that is, constitution through human cooperative agreement. Pre-Socratic philosophers treated justice as a matter of natural, cosmic harmony. Plato understood justice in terms of a grand scheme of proper function- ing and an ideal system of a division of labor, reflecting the essential nature of human beings and the realization of their potential. Contract-based liberalism, however, treats justice as a matter of social cooperation, that is to say, a voluntary Justice and Solidarity 115 form of interpersonal enterprise. Justice is no more a given truth or principle, but a product of a negotiated deal. But of course, the characterization of the conditions of such negotiation, the so-called “circumstances of justice,” is crucial for exam- ining the possibility of extending justice to the global sphere. Since Hume, the list of the circumstances of justice has included moderate scarcity, limited sympathy, mutual vulnerability or interdependence, and rough equality. We cannot cooperate when extreme scarcity leads us into a life-and- death struggle; we do not have to cooperate when there is unlimited affluence of resources; we cooperate without having to create principles for cooperation when altruism and love serve as sufficient motives; we have reason to cooperate only when we are not self-sufficient but rather depend on each other, either to avoid harm or to create benefits; and if, as in sharply hierarchical societies, we are not more or less equal, our interpersonal behavior will be based on domi- nation rather than cooperation. These circumstances of justice are determined partly by universal human conditions and partly by historical factors. Accord- ingly, the issue of global justice must be considered in the light of both the universal and historically determined factors that make social cooperation pos- sible and valuable. There seems to be no reason to believe that, at least in principle, the circum- stances of justice apply any less to the world as a whole than to domestic societies. There are indeed differences of degree in scarcity and equality on these two respective levels, and there is some debate on the extent of mutual vulnerability and interdependence on the global sphere. However, my concern is with the parameter of limited sympathy. Limited sympathy has been traditionally inter- preted, from Hume to Rawls, in negative terms, namely, as self-interest, or at least as the mutual disinterest of the parties to the contract-based principles of just cooperation. But social contract theorists have neglected the corresponding posi- tive meaning of the idea of limited sympathy, namely, the degree of actual care for others which I would like to argue constitutes a necessary condition for the operation of justice and should be understood as one of the constitutive circum- stances of justice. The idea is the following. The principles of just distribution of a social cake are analytically connected with the conditions of cooperation which enable the production of the cake (on the assumption that such cakes cannot be made by a single individual). But the cooperation to bake a cake is not simply given, since the crucial question is with whom do we want to do the baking and the eating of the cake. Social contract theories ignore a fundamental motive for cooperation, which goes beyond self-interest, utility, or survival: the wish to cooperate with a par- ticular set of individuals rather than with another. Game theoretic approaches to justice typically take for granted the scope of the group within which “the game of justice” is played, referring to the players in thin (universal) terms as rational beings (or, to be loyal to Rawls, rational and reasonable individuals constrained by the sense of fairness). But there are numerous schemes of cooperation for any given individual, differing from each other in the scope and the identity of the 116 David Heyd other individuals with whom cooperation takes place. Utility or prudence cannot decide this scope or identity since at most it accounts for ad hoc forms of social cooperation (typically in economic enterprises or games) rather than for compre- hensive systems of social justice.7 And once we realize that the option of coop- erating with everybody is indeterminate, since, as we have shown, it can be interpreted in many ways, we have to decide on what basis the group within which the cooperative enterprise occurs is defined. In other words, cooperation is not just a given fact or condition of justice, but something to be chosen. This demonstrates the shortcoming of a Lockean theory of justice. It takes “the cake” as given, namely, natural resources in the world as given by God to all humanity to share. This is also the ground for the Lockean proviso, which takes all human beings as parties who deserve some share in the goods of nature. However, most issues of distributive justice involve the production of the goods (the baking of the cake) rather than just their consumption or use, which leaves open the question, with whom do we want to cooperate in the enterprise of production and related consumption. If God gave the natural world to humanity as a whole for its use, then we all have to share natural resources with each other, but, in the absence of such theological assumption, the fact that we happen to live on this planet does not in itself create a motive or a reason for sharing. Sharing takes place when we feel special care for others, which arises out of either natural ties or cooperative commitments. Before moving on to the positive grounding of the contractarian theory of justice, I would like to address an interesting and challenging objection to the “local” conception of the contract.8 Any restricted group of individuals can engage in a contract which would generate universal principles of distributive justice, that is, principles that apply to all human beings in the world (rather than just to the contractors). Thus, members of the group can agree to distribute the product of their common enterprise with all others, who are not members of the group, in various degrees: from a minimum that would secure international peace, through a more egalitarian sharing, to extreme forms of saintly altruism. In all these cases, the fact that the principles were agreed by a restricted group of people does not imply that their “reach” cannot be universal. Maybe this is exactly the way we expect enlightened liberal states to fix their principles of justice, namely, to apply them globally, beyond their own borders. Such a disparity between the scope of the consenting group and that of the beneficiaries of the principles agreed upon by the group is indeed possible (although, unfortunately, it tends to be of the less altruistic kind). But it does not undermine the theoretical claim made above, which was concerned merely with the way principles of justice are justified rather than with their content (or the identity of their beneficiaries). The point of contract-based principles is that they apply to the contractors only because they agreed to them (under certain idealized conditions). Therefore, they create claims and counter-claims only within the group in which the contract was made. Non-members of this group may be the beneficiaries of the contract, but have no claim or right against the contracting Justice and Solidarity 117 members. An obvious indication for it is that non-members can share the benefits of a contractual agreement of a group of people, but cannot be expected to carry the burdens of such an agreement (to which they were not party). They can enjoy transfers of wealth to them but cannot be taxed. Contract-based principles of distributive justice can therefore apply to non- members but only for some reason lying beyond the idea of pure consent. The contractors may believe, for instance, that non-members deserve some goods, a minimal standard of living, or the satisfaction of their needs. But these are substantive moral principles, which serve as the grounds for the distributive duties. All the contract does in this instance is to help decide the way the non-contractual duties of members toward non-members is to be distributed within the group of members. An obvious example would be duties of care toward animals (which are beneficiaries of these duties without being parties to the contract). A more con- troversial but still powerful example would be a society’s duties to future genera- tions. Even if future people have rights (which I doubt), they can hardly be accounted for in contractarian terms, especially if we think of future people who are just possible and whose existence is subject to our own decisions. Non- members can never be considered free riders by members of the contracting group since they are not bound by any duty of fairness. Solidarity Solidarity is a social force which contributes to the sustenance of the unity of a group of people. Like the solidity of physical bodies, it is what keeps an entity from disintegrating. Both its etymology and historical sources are French, leading back to the revolutionary idea of fraternité.9 It is intimately connected to coop- eration, that is, to intentional common enterprise, calling for a combined and coordinated action by many people. Unlike natural bonding forces of the kind of family love and care, solidarity is mediated by a commitment to an idea or a cause. In that respect, solidarity is “ideological.” Although we say that solidarity is something we feel toward or with others, this is not a raw feeling, but involves cognitive and reflective elements. Solidarity is thus to be distinguished from both family love and the companionship of friends (camaraderie) in being based on the awareness of a common cause, a worthy social goal, a sense of shared past or future destiny. Solidarity, like cooperation, is not simply given but rather chosen or constructed on the basis of certain beliefs. Solidarity has a typically cohesive effect and is often considered necessary for the success of a social struggle. Thus, solidarity is manifest in the struggle of the working class against oppression, the fight of trade unions for the amelioration of working conditions, or the effort of liberation movements of minorities to achieve recognition. Solidarity typically involves the commitment of individuals to remain loyal to the collective cause even when it means ignoring their self-interest and potential personal gain.10 In that respect, solidarity does not come naturally to us but takes a conscious effort aimed at the achievement of an impersonal goal.11 Solidarity has both descriptive 118 David Heyd and normative dimensions: it characterizes the “togetherness” of people and makes demands of loyalty to the group. Since solidarity is created in the course of a struggle for a collective cause, it is necessarily exclusive, presupposing the existence of competing causes. Unlike friendship, tribal community, and family, solidarity is a social bonding that is formed against, or at least in competition with, other groups. Richard Rorty, for example, speaks of “we” or “one of us” as the foundation of solidarity, when “we” is contrasted to “they” on the basis of a historically dependent construction rather than on any essential natural qualities.12 But this does not imply a Schmittian friend/foe model. Identification with a group is exclusionary in some sense, but not necessarily hostile. Chantal Mouffe appropriately describes solidarity as an “agonistic” (rather than “antagonistic”) basis for “we-identity.” Solidarity is a relational concept which presupposes the existence of other groups that are rivals or adversaries.13 Solidarity, I suggest, is a form of bonding, which is partly given, partly created by a group of people, on the basis of a shared past, a present interest, hopes for the future, or engagement in an enterprise directed to the realization of common values. Shame is an interesting indicator of the limited scope of solidar- ity: we can feel ashamed of what our family members do, or of the behavior of our fellow citizens and even government. We do not feel shame for what other human beings as such have done, and if we do, this is only on the assumption that there is a higher order of beings before whom we feel this shame (God, imaginary extra-terrestrial beings).14 The sources of solidarity are often circumstantial and even forced upon people. Thus, sharing a territory with people of a different culture or even “nation- ality” creates a powerful form of economic and psychological interdependence which, in turn, gives rise to a sense of solidarity or may create a normative reason for promoting solidarity. Unlike friendship or tribal identification, national soli- darity is often shown among people who may be in many respects adversaries, like national and ethnic minorities. The state responsibility of justice toward minorities is exclusive, that is, does not include members of that minority group living outside the borders of the state. Majority and minorities in a multicultural society share the same destiny in many respects, a fact that creates a sense of solidarity. Note that the historically based formation of solidarity is fully compatible with the non-historical, hypothetical nature of the Original Position (or the classical “social contracts”): the historically contingent circumstances are the conditions for the operation and scope of the thought experiment of a hypothetical agreement. One important implication of the agonistic character of solidarity is that it cannot be universal. Humanity might be the object of Kantian respect but not of identification. We identify ourselves as human beings, but we do not identify with human beings as such. The idea of a “human community” is as misleading as it is attractive.15 We could imagine indeed circumstances in which our planet is threat- ened by the invasion of a rival, non-human race, leading to the rise of “human solidarity.” But this only serves to confirm the relational analysis of the concept. Solidarity is in its essence partial, in the sense of being opposite to both universal Justice and Solidarity 119 and impartial, that is to say, it is local and biased. This does not in any way detract from the universality of moral duties or human rights.16 It only aims to distinguish between the universal level of morality and the local, solidarity-based level of certain forms of social justice. Rorty is willing to extend the application of the concept of solidarity to the universal level of “all humanity,” arguing that we feel solidarity with any human being vulnerable to suffering and humiliation.17 But this extension, I believe, is artificial, since it does not involve any shared value or cause with which people actually identify, but only a universal revulsion from pain. Furthermore, the extension goes against Rorty’s own analysis of solidarity, since sympathy and compassion with people suffering pain is not a matter of construct- ing a group of “us” on the basis of contingent historically dependent conditions but a universal natural inclination of human beings (which, unlike solidarity, is often expressed also toward animals). Solidarity, accordingly, is an interesting attitude, since it lies halfway between unmediated feelings like love or compassion and a pure rational Kantian recog- nition of the moral standing of fellow human beings who have dignity and rights. One may describe solidarity as sympathy mediated by a belief in a common project. Solidarity combines personal and impersonal components: the objects of solidarity are mostly anonymous in the sense that the subject does not personally know them. But they are also not considered as just human beings abstracted from all their individuating characteristics. We care for people with whom we feel solidarity, but do so on the basis of their belonging to “our” group fighting for a certain goal rather than as just individual human beings. The view I am proposing here is that just distribution is one such social goal and hence it both presupposes and reinforces solidarity.18 This combination of a historical “given” with a self- determined commitment to a cause is constitutive of solidarity. It should be noted that the justification of the particularistic aspect of distributive justice is not grounded in universal terms (such as the general efficiency of dividing the world into smaller units in the attempt to promote overall justice in the world). It is rather that the conditions of the applicability of contract-based justice are particularistic. Durkheim’s work on solidarity can help us see this point. The modern phe- nomenon of the division of labor is (surprisingly) associated by Durkheim with the development of social solidarity. For Durkheim, division of labor is not merely a means for the promotion of individual welfare through increased production, but a moral value in itself. It touches upon the deep foundations of the very existence of social bonding, since it involves mutual complement rather than mere egoistic exchange. It is more similar to Aristotelian friendship than to cooperation in game-theoretic contexts.19 Mechanical solidarity, for Durkheim, is created on the basis of likeness and is primarily expressed by the criminal law of society, its collective passionate reaction to crime, the common response of people sharing a common conscience. Organic solidarity, in contrast, is grounded in the differences between people which call for cooperation and division of labor. It is the recog- nition of the importance of the different functioning of different parts of society, analogous to the harmonious functioning of parts of an organism.20 For our 120 David Heyd purposes, we should emphasize that Durkheim contrasts modern society, based on the functional division of labor, to traditional society, based on kinship or natural tribal affinity. Durkheim is convinced that modern, organic solidarity is no weaker than its traditional predecessor. Furthermore, organic solidarity is expressed in institutions rather than in feelings, primarily in the laws of contract, family, property, and labor. The apparently impersonal structure of social relations creates the mutual commitment of solidarity. Durkheim’s analysis of solidarity and the division of labor can be used to analyze the problem of the social contract raised at the beginning of this article. Even if the terms of the contract itself are described as completely impersonal, self-interested, and impartial, the motive for entering the contract in the first place should be described in more particularistic terms, namely, the Durkheimian drive to complement one another, to cooperate with others in achieving a morally valuable aim. The original position is not only the thought-experiment in which the prin- ciples of just social cooperation are articulated; it is itself a cooperative enterprise, and as such it must have a motive. As in any game, the rules are impersonal and impartial, but the motivation to play the game with particular people cannot be purely impersonal or abstract. Solidarity is such a motivation and operates in a dialectical way: we engage in just cooperation with people about whom we care and, in turn, the cooperation itself reinforces our solidarity with them. This bidirectional operation of solidarity can be demonstrated in the cooperation of the “working class,” or in the mutual commitment of members of a trade union, in Durkheim’s idea of division of labor, or—in my argument—in distributive justice. One should not be misled into thinking that solidarity is a communitarian idea which serves as a basis for a communitarian conception of justice. Durkheim’s concept of solidarity relates to a modern Gesellschaft rather than to a traditional Gemeinschaft.21 It is accordingly expressed in institutional arrangements rather than in compassionate feelings or shared traditional practices. As Chantal Mouffe proposes, it assumes common values but no agreement on their ranking. The ranking is left for political negotiation within the group, but solidarity is the bonding of the group within which politics takes place.22 I may add that in social contract theory unanimity is a condition only in the agreement on the foundational terms of political legitimacy and the principles of justice; once political society is established, the rule of decision making is typically majoritarian and allows for pluralism and competing goals.23 Solidarity does not necessarily generate an egalitarian system of justice, but it does explain the equal weight given to the parties in the initial contract. Solidarity is, therefore, the bond which unites the partners to the contract in their commitment to decision-making procedures in conflict situations typical of heterogeneous societies.24 In other words, we must first decide the scope of the group within which we count heads and whose majority we follow. Majority, like justice, cannot fix the size of the group within which it operates. So it seems that although liberalism in its universal view of moral relations must be suspicious of the idea of solidarity (and indeed this idea has been much Justice and Solidarity 121 more popular in socialism, feminism, and communitarianism),25 it nevertheless presupposes solidarity as the condition for the social contract. This could also serve as an answer to Brian Barry’s query: How can Rawls’s contract take place in a particular society if the contractors are supposed to be mutually disinter- ested?26 Disinterest is one of the conditions in the original position, but it does not and cannot be a condition for entering the original position. The State It is no coincidence that the idea of contract-based legitimation of govern- ment, and hence of political justice, was developed more or less simultaneously with the rise of the modern state. The state is the most extensive social framework in which principles of distributive justice can be meaningfully applied. Smaller- scale social associations have their own distributive norms, but they cover a more restricted scope of goods and burdens to be distributed. Voluntary associations, like clubs and games, define the scope of distributions in a relatively precise way, but cover only little of the main goods and burdens of human life. Natural associations, like families and tribal communities, control much wider systems of distributions and deserts, but the basis of these systems is not voluntary. The social contract is a device for generating principles of justice that are both voluntary and comprehensive in their content. The state is the natural locus for such principles since it enjoys sovereignty, that is, the power to enforce the contract in a wide array of social issues: it distributes the burdens of military service and taxation as well as the benefits of many public and private goods. It is the most effective framework for the division of labor. Critics of the modern state, like Habermas, argue that the state is an imagined community (in Benedict Anderson’s terms), a construction invented in order to create social solidarity where there is no cultural or ethnic basis for a common political identity. Max Pensky says that “ ‘Nation’ consists of an imaginary soli- darity among strangers that cements the otherwise fragile bonds of reasonable cooperation.” Furthermore, he adds, the difficulty of creating solidarity for the achievement of abstract republican ideas makes the nation-state “ ‘double coded’ or indeed Janus faced.”27 On the one hand, the universalism associated with political and moral equality requires the rejection of all particularistic loyalties; on the other, national solidarity defines itself “against” the other. There is, in other words, an inner contradiction between what might be called inward universal egalitarianism and outward particularism. Now, my approach to the relationship between justice and the state considers this tension between the exclusivism and inclusivism of the state as exactly its asset as the framework of distributive justice. For it embodies a form of solidarity that arises out of certain historical factors (like shared territory, common experience, sometimes a dominant language and culture), and it applies egalitarian principles and impartial concern to its own citizens. 122 David Heyd So David Miller is right in claiming that “without a common national identity, there is nothing to hold citizens together, no reason for extending the role just to these people and not to others.”28 But this in no way entails a communitarian view of either the identity or the normative basis of the state. For these are at least partly derived from the very enterprise of instituting agreed-upon principles of justice. There is no circularity in such a characterization of the state, although it is dialectic in nature: it originates in some form of a historical process, but is then reinforced by the sense of the common endeavor to establish and maintain a just society. This is typically the case with the Constitution-based identity of the American nation.29 Even Brian Barry, a passionate advocate of cosmopolitanism, concedes that national identity requires more than a passport and that it is con- stituted by “civic nationality,” namely, the identification with society’s social institutions and the recognition of a common good. The problem with Barry’s view is that it is not clear what he means by common good: is it a universal good (which would make “nationality” vacuous), or is it particular to the national group (which would weaken his argument that political entities have no moral value and are only instrumental in promoting universal values such as human rights)?30 Solidarity-based justice does not exclusively express itself in the framework of the nation-state. But it should be noted that when it crosses national boundaries it is not constituted by a contract. Thus, the solidarity of the working classes in the communist sense is the mutual commitment of people in different countries to the transformation of the social system into a more just one. But justice here means a substantive ideal scheme of power relations and distributions rather than a set of principles generated by an ideal contract. On the other hand, the call for social solidarity of the middle class with the unemployed and the poor within a given society is a hallmark of the welfare state. It is a demand for the special concern of people for their fellow citizens with whom they cooperate politically (in the democratic vote, in defending the country, in producing its wealth, in perpetuating its cultural heritage, and of course in sharing its territory).31 Can Justice Be Globalized? As I indicated at the beginning of this article, theoretical attempts to extend the scope of application of distributive justice across nations arose more or less simultaneously with attempts to extend it across generations. Of the two extrapo- lations, it was believed that the intergenerational (“vertical”) is more intractable than the international because of the crucial difference between possible and actual people. Duties of justice toward future people cannot be fully determined due to the indeterminate number of people who are going to be both subjects and objects of those duties. Not only do we not know the relevant demographic data; the number of future people is to a large extent determined by the economic and demographic policies which stand to be chosen. This is a logical problem for which there is no coherent answer. On the other hand, global justice in the “horizontal” sense concerns actual people whose number and identity are both Justice and Solidarity 123 given and known (not to speak of their needs and preferences which can only be guessed in the intergenerational case). In that respect, at least philosophically, the case for transnational justice is much easier to make.32 However, if my argument about the conceptual connection between justice and solidarity is persuasive, then it seems that the intergenerational application of justice is more natural than its global counterpart. For it is definitely the case that our concern for “our own” descendants (and respect for our predecessors) carries more weight than our concern for contemporary strangers. The former are conceived as partners to the social cooperation of our polity, for whose welfare we strive, and we expect them to contribute to a common enterprise that is transgenerational in nature. It is much more natural to view the state as extending in time than to see it as a global entity comprising all actual (present) human beings. We feel solidarity with our (future) children and grandchildren but not with our remote contemporaries. A single generational “slice” of a community is not self-sufficient in the sense that we cannot invest meaning in our private and collective life without the perspective of both the past and the future of the social project in which we are engaged. Yet, the transgenerational continuous existence of a community makes perfect sense, even independently of other, contempora- neous communities. However, there are a few powerful philosophical voices which press the case for global justice no less than for intergenerational justice. Impartiality means that we should take the geographical index as no less arbitrary than the temporal. Much of the debate revolves around the possibility of applying Rawls’s difference principle globally, and whether Rawls himself is correct in insisting that it can be done only in a limited way. Charles Beitz is one of the first and most fervent advocates of global justice. He argues first against the Hobbesian “realist” view that considers the international realm as lying beyond the grasp of moral judg- ment. Rawls’s difference principle should be applied also globally, ideally on the individual level (i.e., not constrained by intra-national arrangements), or at least on the national level (constraining any deviation from international equality by concern for the worst-off country). Beitz’s cosmopolitan approach grounds the duties of international redistribution not in the duty of assistance but in that of justice.33 For sovereignty, unlike autonomy, cannot serve as a defense against sharing a country’s resources with others. Unlike natural talents, which are not subject to redistribution, being constitutive of our personal identity, natural resources do not constitute collective or national identities, and hence their actual distribution in the world should be considered morally arbitrary and consequently subject to redistribution on the basis of egalitarian justice.34 This appears to be a forceful argument, but we must remember that some important natural resources (even oil) are not like manna, simply there to be used, but have to be searched for and extracted from nature, that is to say, they have value only when a cooperative human activity transforms them into consumable goods. Again, the cake, even when it looks “natural,” has to be baked before being eaten and hence its just distribution is a matter of deserts that accrue to the 124 David Heyd participants in its production. Indeed, this might not apply to some primary goods like health, in which desert or right is not derived from the contribution to its “production.” But even in this case, we view the distribution of health services as part of an overall scheme of social cooperation in which the ability to provide these services is dependent on the productivity in a particular group of people and on their particular order of priorities. Beitz also points out that sovereignty cannot be a morally relevant constraint on global distributions. The boundaries of a political sovereign entity cannot be left to individuals to decide, since—as was argued at the beginning of this article—who will decide the scope of the group of individuals who will make the decision? However, I think Beitz draws the wrong cosmopolitan conclusion from this puzzle of bootstrapping: it cannot mean that all human beings should take part in deciding the principles of just distribution, since that already prejudges the issue of the correct extent of the contract and in any case leaves the range of “all” indeterminate. For this reason, Beitz’s demand to leave the facts about national affiliation under the veil of ignorance when engaging in the thought-experiment of the original position undermines the very possibility or motive for a contractual agreement.35 For when I choose to enter the original position, I want at least to know with whom, with what kind of people I am going to hold justice relations. Rawls, in his later writings, came to the conclusion that distributive justice, particularly the difference principle, applies only on the domestic level, in “bounded societies” which are self-sufficient transgenerational systems of coop- eration into which individuals are born and in which they usually die. Although there are transnational principles of justice, they first apply only to “peoples” (rather than to individuals), and second they are limited in content to raising other societies (or peoples) to the threshold conditions of justice, that is, the ability to maintain just domestic institutions.36 There is, therefore, no duty of justice of equality toward either the worst-off societies or the worst-off individuals in the world. The main reason for this restriction of justice to the domestic sphere is that societies are to a large extent responsible for their own economic conditions, which are the function of free choice regarding demographic planning and other cultural values (such as the rate of savings for future generations, or religious arrangements of property and power distribution). I want to reinforce the Rawlsian opposition to the globalization of justice. Rawls points out that different societies might agree on different principles for domestic justice, for example, non-liberal (but “decent-hierarchical”) societies would not accept the difference principle. But even assuming that (ideally) all societies in the world agreed on the difference principle, this would not mean a global application of that principle. For each society is at best motivated by the idea of advancing the welfare of its worst off members, rather than the world’s. This can be illustrated by an analogy. If two societies adhere to the (same) majority principle, it does not mean that they wish to follow the joint majority of the two societies! Even the other clause of Rawls’s second principle of justice, the equality of opportunity, is social specific; it is artificial to think of giving every- Justice and Solidarity 125 body in the world equal opportunity in the competition for jobs, education, or political participation. But of course, these analogous cases of distributive justice—majoritarianism and equal opportunity—should be distinguished from the commitment to human rights: two countries which sign the Geneva Conven- tion assume a universal responsibility to prevent war crimes in general, not just “their” war crimes. The same applies to natural duties like those of assistance to starving people. So the local character of justice does not derive solely from the culture-dependent nature of its principles, but more fundamentally from the local basis or motive of the cooperation which it is meant to regulate. It is interesting to note that while on the domestic level the contract includes everybody, regardless of his or her comprehensive moral views, on the interna- tional level a contract is made, according to Rawls, only with peoples of a certain kind, namely, “liberal and decent,” and not with “outlaw and burdened” societies. This indicates that any contract must be motivated by some deep motive, such as solidarity, which operates in different ways in the domestic and the international contexts. In the domestic case, it is the motive of self-interested individuals to further their welfare by cooperating on a fair basis with people with whom they feel solidarity. In the international case, it is the motive of promoting justice, peace, and security which leads societies to principles of non-aggression, trade relations, and aid. Indeed, states as such do not feel solidarity with each other, but peoples often do share such feelings (e.g., the United States and Britain in World War II). However, this kind of solidarity leads to alliances and agreements of mutual interest rather than to relations of distributive justice. These agreements, unlike the domestic social contract, are actual rather than hypothetical, and are binding like promises rather than as a matter of fairness. Both the domestic and the global approaches to justice take the individual as the essential unit. Although Rawlsian contract theory is conceived as a mere political construction, the idea of justice as fairness must take the individual as a non-arbitrary, metaphysical foundation in the theory. It is only the scope of the group of contracting individuals, which is left contingent (depending on the kind of solidarity frameworks that happen to develop in history). In other words, individualism is a philosophically necessary assumption of contract-based justice. However, the international approach addresses relations between peoples, states, or societies, which are not metaphysically fixed units, but rather empirically determined entities, and worse, entities whose scope and bounds are themselves partly objects of normative decision. Now, denying that justice applies globally does not mean that there are no moral principles governing international relations. There are natural duties which obligate both individuals and governments, primarily the duty of aid. People who lack the basic necessities, like food and water, immunization or shelter, have a legitimate expectation to be assisted. Unlike the concern for the worst off in society on the domestic level which is based on equality, the concern for starving people in other societies is based on want, on deprivation of universal human means of survival. Human rights and international law are also to be respected 126 David Heyd irrespective of national boundaries and cooperative agreements. So are the duties of compensation for past exploitation or oppression, which are measures of corrective justice. On the level of states, non-aggression and the respect for trade and other agreements are principles that must be respected. But these are all principles of a different sort than those of fair distribution (like the difference principle). Their justification lies in human nature, rationality, or sympathy, rather than in ideal contractual agreement. Indeed, there are also specifically distributive problems that call for global solutions, like the distribution of the burdens associated with fighting global warming, the ozone threat, the dwindling of the fish population in the oceans, the disappearing rain forests, world epidemics, and so on. There are goods which in their nature are global, that is, cannot be linked to any particular sovereign territory. These problems can only be addressed by sovereign states in the world through mutual agreement. But like international trade agreements or mutual disarmament treaties, these do not relate to “the basic structure of society,” but are of an ad hoc nature. They reflect the power relations and interests of the actual parties rather than the idea of fair distribution (although this does not mean that the agreements are free from moral constraints, like the prevention of exploitation or the use of force). Unlike the “basic structure of domestic society,” world order is expected to achieve primarily peace, stability, and mutual advantage, rather than the fair distribution of primary goods. The aim of this article is modest. It does not intend to claim that human action on the global level is free from moral constraints. It does not purport to relieve either individuals or states from duties toward members of other societies. It only tries to show that one of the most powerful methods of justifying normative arrangements, that of the social contract, cannot account for the sphere of global moral relations. If we adhere to a liberal view of justice as a consent-based system of norms created under the strict methodology of a hypothetical contract, then we will have to view international relations and global distributions as deriving their justification from a different source than that of the contract. Justice, particularly in its distributive sense, should be kept as a separate category from natural duties, rational ad hoc agreements, impartial solutions of conflicts, and respect for human rights. This article tries to show that the distinctive feature of contract-based justice is a prior sense or motive which makes the contract possible and defines its scope. This sense is here called solidarity. The answer to Hobbes’s perplexity— how can a multitude of people institute a commonwealth through a covenant?—is quite simple: the covenanting people are not really a multitude to begin with. This article was written during my fellowship at the Institute for Advanced Studies of the Hebrew University of Jerusalem. Justice and Solidarity 127 Notes 1 Alongside numerous philosophers following Rawls, I have discussed that issue in some detail in my Genethics: Moral Issues in the Creation of People (Berkeley: University of California Press, 1992). 2 John Locke, Two Treatises of Government, ed. Peter Laslett (Cambridge: Cambridge University Press, 1988), 331. 3 Thomas Hobbes, Leviathan, ed. Richard Tuck (Cambridge: Cambridge University Press, 1996), 121. 4 For an early discussion in our period of democracy’s inability to determine its scope, see Frederick G. Whelan, “Democratic Theory and the Boundary Problem,” Nomos 25 (1983): 13–47. 5 Cf. Bernard Yack, “Popular Sovereignty and Nationalism,” Political Theory 29 (2001): 529. 6 Sen strongly advocates this shift from the model of consent to that of impartial arbitration. And he is right in his belief that a model of impartial decision can avoid the logical problem of a contract with future people or with potential members of a given polity in deciding the scope of membership. However, I am not sure how the conditions for impartial decision can be speci- fied. The whole point of consent-based legitimation is that it avoids the interminable contro- versies about what constitutes impartiality by letting the parties to the conflict agree on the principles of justice. Impartiality, like neutrality and the “ideal observer,” is hard to define without appealing to substantive values or positions, while the fairness of a process of agree- ment is less controversial. Amartya Sen, “Justice across Borders,” in Global Justice and Tran- snational Politics, eds. Pablo De Greiff and Ciaran Cronin (Cambridge: Cambridge University Press, 2002), 45–46. 7 This seems to go against Russell Hardin’s view in his One for All: The Logic of Group Conflict (Princeton, NJ: Princeton University Press, 1995), particularly chap. 3. I do not deny that self- interest and group identification are often “congruent” and that such identification may cease when costs to the individual become very high (p. 47). However, this does not prove that self-interest can be the exclusive and exhaustive grounds for the formation of a political entity with its own system of justice (even if we include under “self-interest” altruistic motivations such as the interests of others). Hardin’s main concerns with the rationality of identification are contexts of ethnic or national conflicts (where it is natural indeed to approach the issue of belonging from the instru- mental or self-interested point of view); but my focus here is on the context of distributive justice in which the promotion of interest within a fair system of binding rules (rather than pure self-interest) is sought. 8 The objection was raised by one of the anonymous referees of this article. I am grateful in general to his or her detailed and incisive comments. 9 It is interesting to note that most of the major philosophical literature on the subject of solidarity is French (and then German), rather than English and American: Auguste Comte, Emile Durkheim, and Jürgen Habermas, to name a few of the important writers on solidarity. Rawls considers the difference principle as a manifestation of fraternity (A Theory of Justice [Cambridge: Harvard University Press, 1971], 106), but does not analyze its nature. 10 Loyalty is closely related to solidarity in its being “necessary to form a community of reciprocally caring people, committed to counting each other as equal in their common project.” George Fletcher, Loyalty (New York: Oxford University Press, 1993), 20. I follow Fletcher in the idea that “loyalty is a critical element in a theory of justice; for we invariably need some basis for group cohesion, for caring about others, . . . as partners in a common venture” (p. 21). However, soli- darity seems to me to better capture the conditions of justice than loyalty, since it is not a straightforward duty (sometimes even overriding other moral duties), but rather a presupposition of justice (though it has some weak normative force). Solidarity is a less personal relation, and, unlike loyalty, does not apply to marital relations or to the relations between man and God. The opposite of solidarity is not treason or treachery, but simply indifference. It is essentially reciprocal (which is not the case with loyalty): I cannot feel solidarity with a group in which no one else feels solidarity, since the group is (partly) constituted by the sense of solidarity. This also explains why I can be loyal to my (true) self, but cannot be said to feel solidarity with myself. 128 David Heyd 11 Some philosophers distinguish between natural solidarity (of family or community members) and “project-related solidarity” (based on what I refer to as a common cause). See Klaus Peter Rippe, “Distinguishing Solidarity,” Ethical Theory and Moral Practice 1 (1998): 355–73. My approach, however, involves features of both kinds of solidarity. As Joel Feinberg has noted, solidarity consists not only of the overlap of shared specialized interests (like in corporations), but also of a community of interests (like in a family or a nation). See Joel Feinberg, Doing and Deserving (Princeton, NJ: Princeton University Press, 1970), 233–34. It is interesting that Feinberg suggests an argument concerning retributive justice which runs parallel to the one outlined here concerning distributive justice: collective responsibility can be imposed on a group of people only when there is an “antecedent solidarity” between them (ibid., 240–41). 12 Richard Rorty, Contingency, Irony, and Solidarity (Cambridge: Cambridge University Press, 1989), 189–98. 13 Chantal Mouffe, “Politics, Democratic Action, and Solidarity,” Inquiry 38 (1995): 99–108. 14 This very exceptional sense of shame was developed by Karl Jaspers in his moving, post-Holocaust essay on German guilt. Metaphysical guilt is felt by human beings for crimes other human beings have perpetrated. It is felt in the face of God and is close to the sense of shame in belonging to the same species as the perpetrators. Karl Jaspers, The Question of German Guilt, trans. E. B. Ashton (New York: Capricorn Books, 1961). Cf. Feinberg, Doing and Deserving, 236–37, where guilt and shame are presented as indicators of solidarity. 15 Avishai Margalit has pointed out to me the deceptive and manipulative idea of “the family of man” (famously used as the title of a 1950s exhibition of photography). 16 Richard Miller, “Cosmopolitan Respect and Patriotic Concern,” Philosophy and Public Affairs 27 (1998): 202–22. Equal respect applies to all human beings as such; concern applies in more local groups, like peoples. 17 Rorty, Contingency, 192. 18 For a very close view, see Véronique Munoz-Dardé, “Fraternity and Justice,” in Solidarity, ed. Kurt Bayertz (Dordrecht: Kluwer, 1999), 87: “It may appear that the ‘spirit of fraternity and benevo- lence’ must already exist between citizens as a precondition of social justice, and that the difference principle depends upon a conception of the good that makes each member of the political community want to identify with and care for other citizens.” See also Margaret Canovan, Nationhood and Political Theory (Cheltenham: Edward Elgar, 1996), chap. 4, appropriately entitled “Social Justice: Looking after Our People.” 19 Emile Durkheim, The Division of Labor in Society, trans. George Simpson (New York: Free Press, 1964), 63–64. 20 Ibid., 127–29. 21 Cf. Ulrich K. Preuss, “National, Supernational and International Solidarity,” in Solidarity, ed. Kurt Bayertz (Dordrecht: Kluwer, 1999), 283. 22 Mouffe, “Politics,” 103. According to Craig Mataresse, Sartre’s idea of solidarity is based on the individual’s pledge not to betray or defect from society. Being based on choice, it is contrasted to Hegel’s communitarian and organicist view of political society, in which the social bonding is not mediated by a principle or a value. See Craig Mataresse, “Solidarity and Fear: Hegel and Sartre on the Mediations of Reciprocity,” Philosophy Today 45 (2001): 43. 23 I have dealt in detail with this two-tier structure of contract theory in the context of the majority principle. See David Heyd, “Counting Heads or Casting Lots: The Lottery Challenge to Majority Rule,” Iyyun 53 (2004): 3–26. 24 Charles Taylor has argued that an egalitarian democratic regime requires a higher degree of citizen solidarity than a traditional hierarchical society in which the issue of the minority acceptance of the majority decision does not arise. 25 Kurt Bayertz, “Four Uses of Solidarity,” in Solidarity, ed. Bayertz (Dordrecht: Kluwer, 1999), 3–28. This article is an important contribution to the taxonomy and analysis of the concept of solidarity. Justice and Solidarity 129 26 Brian Barry, Theories of Justice (London: Harvester, 1989), 181–82. This resolves the tension pointed out by Barry between Rawls’s commitment to impartiality on the domestic level and his acceptance of mutual advantage as guiding international relations. 27 Max Pensky, “Cosmopolitanism and the Solidarity Problem: Habermas on National and Cultural Identities,” Constellations 7 (2000): 66. 28 David Miller, “Community and Citizenship,” in Communitarianism and Individualism, ed. S. Avineri and A. de-Shalit (Oxford: Oxford University Press, 1992), 94. See also On Nationality (Oxford: Clarendon Press, 1995), 61–62. Andrew Mason takes issue with Miller and claims that a weaker sense of belonging, not necessarily “national,” is a sufficient condition for distributive justice. Andrew Mason, Community, Solidarity, and Belonging (Cambridge: Cambridge University Press, 2000), 134–35. 29 Craig Calhoun, “Constitutional Patriotism and the Public Sphere,” in Global Justice and Transna- tional Politics, eds. Pablo De Greiff and Ciaran Cronin (Cambridge: MIT Press, 2002), 290. Calhoun argues that the public sphere is constructed through discourse, that is, it is a product of “solidarity and choice.” I am told that a major factor in contemporary Swedish national identity is the commitment of its citizens to the Swedish welfare system and their pride in it. This would be a perfect example of the way justice plays a role in the creation of national solidarity. 30 Brian Barry, Culture and Inequality (Cambridge, MA: Harvard University Press, 2001), 80ff. 31 For a very compelling description of the multifarious ways in which citizens are united by “common sympathies,” see John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999), 23 and particularly note 17 in which he quotes extensively J. S. Mill on “Nationality” and its grounds. 32 See Lukas Meyer, “Justice, Intergenerational,” in Stanford Encyclopedia of Philosophy, ed. Edward Zalta. URL: http://plato.stanford.edu/entries/justice-intergenerational. 33 Charles Beitz, Political Theory and International Relations (Princeton, NJ: Princeton University Press, 1979), 128. 34 Thomas Nagel, in a recently published article, also wonders about the moral arbitrariness of local borderlines but ultimately sides with a form of non-universalistic, “associative” model of justice. See Thomas Nagel, “The Problem of Global Justice,” Philosophy and Public Affairs 33 (2005): 113–47. 35 Charles Beitz, “International Liberalism and Distributive Justice: A Survey of Recent Thought,” World Politics 51 (1999): 269–96. 36 Rawls, The Law of Peoples. 130 David Heyd http://plato.stanford.edu/entries/justice-intergenerational