Microsoft Word - Gilabert--Justice and Beneficence (CRISPP).doc   1   “Justice and Beneficence” [Penultimate version; forthcoming in a special issue of Critical Review of International Social and Political Philosophy] Pablo Gilabert Philosophy Department, Concordia University, Montreal, Canada What is a duty of justice? And how is it different from a duty of beneficence? We need a clear account of the contrast. Unfortunately, there is no consensus in the philosophical literature as to how to characterize it. Different articulations of it have been provided, but it is hard to identify a common core that is invariant across them. In this paper, I propose an account of how to understand duties of justice, explain how it contrasts with several proposals as to how to distinguish justice and beneficence, respond to some objections and suggest further elaborations of it. The conceptual exploration pursued in this paper has practical stakes. A central aim is to propose and defend a capacious concept of justice that makes a direct discussion of important demands of justice (domestic and global) possible. Duties of justice can be positive besides negative, they can be imperfect as well as perfect, they can range over personal besides institutional contexts, they can include multiple associative reasons such us non-domination, non- exploitation and reciprocity, and they can even go beyond existing national, political, and economic associative frameworks to embrace strictly universal humanist concerns. We should reject ideological abridgments of the concept of justice that render these possibilities, and the important human interests and claims they may foster, invisible. 1. Introduction It is common to assume that there is an important distinction between justice and beneficence (or charity, humanitarianism—I treat them as equivalent here). This distinction between duties to others seems important, for example, when we think about our distributive responsibilities. Some believe that certain duties to help the needy or the relatively poor are duties of justice, whereas others claim that they are only duties of beneficence. Some believe that some of those duties are duties of justice in domestic, but not in global contexts. But what is a duty of justice? And how is it different from a duty of beneficence? We need a clear account of the contrast. Unfortunately, there is no consensus in the philosophical literature as to how to characterize it. Different articulations of it have been provided, but it is hard to identify a common core that is invariant across them. In this paper, I propose an account of how to understand duties of justice (section 2), explain how it contrasts with several proposals as to how to distinguish justice and beneficence   2   (section 3), and respond to some objections and suggest further elaborations of it (section 4). My focus will largely be on distributive justice (i.e. on the just apportioning of access to material goods and conditions affecting subsistence or well-being, including for example housing, health care, education, alimentation, opportunities for productive and meaningful work, income and wealth, instruments of production, and natural resources). The conceptual exploration pursued in this paper has practical stakes. A central aim is to propose and defend a capacious concept of justice that makes a direct discussion of important demands of justice (domestic and global) possible. Duties of justice can be positive besides negative, they can be imperfect as well as perfect, they can range over personal besides institutional contexts, they can include multiple associative reasons such as non-domination, non- exploitation and reciprocity, and they can even go beyond existing national, political, and economic associative frameworks to embrace strictly universal humanist concerns. We should reject ideological abridgments of the concept of justice that render these possibilities, and the important human interests and claims they may foster, invisible. 2. A proposal for how to understand duties of justice Let me start by presenting my view about how to understand duties of justice. I suggest that we conceive of them as follows: Duties of justice are duties to preserve or promote people’s access to important conditions or goods to which they are entitled and whose fulfillment is prima facie enforceable. This enforcement is all things considered justifiable if it is necessary for or strongly contributes to securing the required preservation or promotion and can be feasibly introduced without imposing unreasonable costs.   3   There are different ways to develop this concept into a more specific conception of justice.1 That development would involve providing substantive accounts answering questions such as the following: What conditions or goods are important enough to give rise to rights and duties of justice?2 What duties of preservation or promotion are reasonable to accept? What makes a proposed form of enforcement necessary for or strongly contributory to the fulfillment of a right? What considerations bear on the appraisal of the feasibility and reasonable costs of the enforcement of a duty? What areas of social life are amenable to demands of justice (i.e. what is justice’s site)? What are the criteria for identifying duty-bearers and right-holders (i.e. what is justice’s scope)? Different conceptions of justice can differ in their answers to these questions while sharing the concept as characterized above. They can agree on the role of justice in identifying prima facie enforceable duties to secure people’s rights even if they disagree about what makes duties fit for that role. I will explain these points further in section 3. But before proceeding let me say something about the status of my discussion. The account of the concept of duties of justice just provided is neither a stipulation nor an elucidation, but something in between these that we can call a deliberative interpretive proposal. A deliberative interpretive proposal shares with elucidation the interpretive interest in relating the account of a concept to the practices in which that concept may be used. But it differs from elucidation (and gets closer to stipulation) in that its core aim is deliberative rather than descriptive: it does not merely report what some agents in fact think, but makes a suggestion as to what they are to think. The aim is to propose an understanding of justice that relates to certain practices in which elements of the concept may be already in use, or where the concept may be relevant, while offering an articulation of the concept that may improve on the current concept or its understanding and thus help ameliorate the practice itself.   4   In proposing and defending the deliberative interpretive proposal mentioned, it is helpful to proceed by considering some reasonable desiderata, or adequacy conditions, for an account of duties of justice. What do we want to achieve by identifying some duties as duties of justice? What theoretical and practical work would a distinction between such duties and duties of beneficence do for us? The following are some of the desiderata I think worth recognizing (I do not claim that this list is exhaustive): (a) Capturing the stringency of justice. Justice involves demands that are very strong and that normally override competing considerations. An account of justice should pick out this feature. (b) Enabling rather than suppressing substantive debates. We should not propose an understanding of the concept of justice that is so narrow that it renders important substantive debates about duties of justice irrelevant. We want to engage intelligently in those debates and to reach a considered view about what position to support. Eliminating options by definitional fiat impoverishes our practical reasoning. (c) Determinateness. Our understanding of justice should be tolerably determinate. Ideally, we should identify necessary and sufficient conditions for the application of the concept of justice. Failing that, we should at least identify typical features holding for an important range of cases.3 (d) Capturing the critical role of justice. Conceptions of justice have a practical role in criticizing social practices and institutions. We should be suspicious of accounts of justice that do not help us make sense of challenges to the status quo. (e) Illuminating practices. Our account of the concept of justice should help us clarify the stakes of important practical questions.   5   3. How the proposed account relates to several proposals as to how to distinguish justice from beneficence The following are the main contrasts that have been proposed to identify the specificity of duties of justice: I. Due. Duties of justice, unlike duties of beneficence, are concerned with giving people their due. II. Negative / Positive. Duties of justice are negative, whereas duties of beneficence are positive (the former demand that we avoid depriving others of access to certain important conditions or goods, while the latter demand that we provide them with such an access or protect them when they already have it). III. Perfect / Imperfect. Duties of justice are perfect, whereas duties of beneficence are imperfect (the former are, and the latter are not, such that it is always clear who owes what to whom in what circumstances). IV. Enforceability. Duties of justice are justifiably enforceable, whereas duties of beneficence are not. V. Rights. Duties of justice have correlative rights, whereas duties of beneficence do not. VI. Institutional / Personal. Duties of justice are institutional, whereas duties of beneficence are interpersonal. VII. Associative / Non-associative. Duties of justice are always associative (they hold only amongst those who share a certain associative framework—such as a system of economic cooperation, a state, etc.), whereas duties of beneficence (sometimes) are not.4 The understanding of justice I proposed in section 2 combines some elements of contrasts I, IV, and V (although, as it will become clear, it does not collapse into any of them). On this understanding, duties of justice range over what is due to people. Just distribution is a response to   6   people’s rights to what is distributed. And distribution warrants enforcement. Such enforcement is seen as only prima facie warranted, however. Enforcement imposes constraints on the freedom of agents and involves burdens for them, and such costs must themselves be justifiable as feasible and reasonable. Hence, a duty could be a duty of justice even if a specific implementation of it is not justifiably enforceable, all things considered, in certain circumstances (either because in the circumstances such an enforcement is not feasible or because it imposes costs that are unreasonable given other, stronger, conflicting demands of justice). But since a pro tanto ground for action persists, agents may have to find alternative (feasible, reasonable) ways to honor it, or change the circumstances so that some form of honoring the ground becomes practicable. Consider, for example, the duty to help eradicate severe poverty. Access to basic nutrition, housing, education, and medical care constitute urgent interests that give rise to rights. The fulfillment of the correlative negative duty not to deprive others of access to basic necessities is clearly worthy of enforcement. So arguably is a positive duty to help others gain and maintain such an access. Now, the latter duty is clearly feasible at reasonable cost in most contemporary societies. It could be argued that this is not the case in extremely poor societies where corruption and lack of economic dynamism are rampant. If this is true, agents in poor societies still have a duty to improve their political institutions and bolster their economic development so that more poverty can be eliminated. And the members of wealthier societies have duties to facilitate, and of course also not to block, such improvement. They may have to stop using international institutions (such as the International Monetary Fund or the World Trade Organization) as tools for dominating and exploiting poorer societies, and help turn them into tools for expanding the opportunities of the poor to escape penury. I will defend the proposed view of duties of justice by explaining how it relates to the seven contrasts mentioned. In doing this, I will draw on the five desiderata listed in section 2. I will   7   argue that my proposed account satisfies the desiderata, that it does not neatly coincide with any of the traditional contrasts, and that none of the contrasts fulfills all the desiderata. The text in this section partly draws on, but systematically revises and expands, the discussion on the contrast between justice and humanitarianism presented in section 1.6 of my book From Global Poverty to Global Equality.5 (i) Capturing the stringency of justice A common intuition is that duties of justice are particularly stringent moral demands. Many agree with the view that, as Adam Swift puts it, “[p]olitical philosophy … is a very specific subset of moral philosophy, and one where the stakes are particularly high. It’s not just about what people ought to do, it’s about what people are morally permitted, and sometimes morally required, to make each other do”.6 The account I suggested captures this by saying that duties of justice are prima facie enforceable. However, it might be objected that my account does not fully capture the stringency of justice because it qualifies enforceability by reference to “reasonable costs” to duty bearers. This might seem to render duties of justice less stringent than their contrast with beneficence intuitively involves. On a construal of contrast IV (Enforceability), we would be problematically blurring the distinction between justice and beneficence if we allow duties of justice to turn out to be not enforceable given computation of costs. It would be better to deem such duties demands of beneficence, which are typically less stringent, and not necessarily enforceable. To illustrate this objection, consider two scenarios. Both include duties whose fulfillment involves severe costs to the duty-bearer (such as their inability to pay for housing). The content of the duties in the two cases differs as follows. In the first, Federico pays back a loan to his creditor. In the second, Gillian helps several destitute people by giving them a considerable amount of money. According to the objection I am considering, it seems intuitively correct to   8   subject the second duty, but not the first, to a “reasonable costs” qualification, and this difference is one of the things we want to mark by calling the first a duty of justice and the second a duty of beneficence. The duty of beneficence to help the needy can be limited if its fulfillment involves the severe impoverishment of the duty-bearer, but this caveat does not apply to the duty of justice to pay back a loan. The objection that taking demands of justice to be sensitive to costs would deprive them of their characteristic stringency (as opposed to the lesser stringency of beneficence) is mistaken. First, it seems reasonable that demands of justice be sensitive to costs. This is because justice should attend to the claims of everyone, including duty-bearers besides right-holders.7 Thus, we can reasonably reject the judgment that Federico should sell his home and liquidate all his assets and run to risk it all in a casino tonight even if doing this offers the only opportunity for him to raise enough money to make the payments that are due tomorrow. There may be extenuating circumstances such that someone’s debt payments may be justifiably lowered, or delayed, or cancelled. For example, they may cater for a sufficientarian norm of justice that no one should fall into destitution. Second, even if Gillian is permitted to refrain from giving so much money, she may still have some stringent and demanding duties toward the destitute. Consider the view that it is not wrong to tax people like Gillian to help the destitute. We think that the level of taxation should be determined not only by considering the potential beneficiaries of policies funded by taxation, but also by attending to the situation of those who would be taxed. This point, like the previous one, suggests that duties of justice and sensitivity to costs can legitimately be coupled. Third, we can see some duties of justice as weightier than other duties without having to say that the latter are not duties of justice. There can be demands that outweigh, or even constrain, others, while being all demands of justice. It is of course often difficult to determine the precise   9   relations among different demands of justice, but we can recognize the general point that a plurality of them warranting prima facie enforcement exists. We can then envisage appropriate articulations in particular exercises. A clear and familiar example is Rawls’s “lexical ordering” of his two principles of justice. The first principle protecting equal civil and political liberties constrains the second principle promoting economic justice, and the part of the second principle concerning fair equality of opportunity constrains the other part (the “Difference Principle”) demanding the maximization of income for the worst-off.8 There is a common view according to which duties of justice normally have logical priority over duties of beneficence because the former, unlike the latter, are constitutive of an account of what is rightfully owned.9 Beneficence focuses on what people should do with what is theirs, assuming an account of what belongs to whom. Now, the account proposed here agrees with this view. Some positive duties to help others can be the counterpart of their positive entitlements. We can then justify the claim that rich people like Warren Buffett should be taxed to fund public education for children from destitute families by saying that the destitute children are entitled to the part of the rich’s pre-tax income that could be used to support them. (ii) Enabling rather than suppressing substantive debates A common temptation that should be resisted is to define disputed terms in ways that make some substantive claims couched in such terms impossible. This maneuver or shortcut does not solve substantive problems. The positions rendered definitionally impossible will simply reassert themselves through challenges to the definitions proposed, or by disputing the weight in our overall reasoning of the claims using the terms as defined. Consider contrast II (Negative / Positive). On this contrast, no thing is a duty of justice unless it is a negative duty or a derivative of it. Thus, there is a negative duty of justice not to deprive others of the means of survival. There is no positive duty of justice to provide them with such   10   means (unless such provision is a form of compensation for previous deprivations, or the counterpart of a promise or a contract whose violation would flout a negative duty). Now, this view of justice as based on negative duties should be seen as a substantive conception of justice, not as a claim about the concept of justice. If the latter were asserted, this would beg the question against those who say that people like Warren Buffett have positive duties of justice to contribute to the subsistence of the destitute. If those who think that there are underived positive duties of justice were told that they are thinking about duties of beneficence rather than justice, they will contest the definition of the former as unduly narrow. Alternatively, they will say that duties of justice do not have the primacy often claimed for them (so that an alleged negative duty of justice not to tax Warren Buffett to help the destitute would be outweighed by a positive duty of beneficence to make him help them, even if this involves taxing Buffett). Hence it is better to keep the definition of justice broad enough to make room for substantive debate. Thus, a libertarian view of all fundamental duties of justice as negative should be seen as a substantive theory of justice, not simply as a definition of the concept of justice.10 A defense of it through a definitional maneuver would be a Pyrrhic victory. Does what I have just said warrant the rejection of contrast II? A defender of this contrast could reply that we should accept it because of its fit with the first desideratum concerning the stringency of justice. This is because a negative duty regarding a certain object is normally weightier than a positive duty regarding the same object. Stealing food from the hungry is worse than refraining from giving them food. But this rebuttal is unsatisfactory. First, it does not ground a general claim of supremacy of negative over positive duties (which seems necessary for contrast II to hold). A negative duty regarding a certain object may have less moral weight than another positive duty regarding another object when the two conflict in practice. My duty not to deprive you of access to an object you own without your consent may have less weight than my   11   duty to save someone’s life. This is so, for example, when I can only take a mortally wounded person to the hospital by taking your car. Second, some negative duties may be weightier than positive duties with the same object without this entailing that the former are, and the latter are not, duties of justice. Even if they have unequal weight, both duties may be sufficiently weighty to be duties of justice. Thus, the duty to help secure educational opportunities for destitute children may have less weight than the duty to refrain from prohibiting destitute children from attending school, but it may be weighty enough to justify taxing Warren Buffett to fund its implementation. The account I have proposed is broad enough to enable the substantive discussion that contrast II represses. Similar considerations apply regarding contrasts IV and VI. Consider contrast IV (Enforceability). The problem with this contrast is that justifiable enforceability is neither sufficient nor necessary for something to be a duty of justice. Some contributions could be justifiably enforced without being duties of justice. For example, Allen Buchanan has argued that some forms of enforcement can be seen as securing the provision of certain collective goods that do not respond to correlative rights. Buchanan’s discussion of this thesis may be problematic, as the collective goods he mentions—clean air, energy conservation, national defense—may in fact be the objects of rights.11 But we can imagine cases in which rights are not involved. An example might be the promotion of certain cultural practices of the kind Rawls discusses when he identifies the tasks of a hypothetical “exchange branch” of the state in a well-ordered society.12 An instance of this is public funding for opera houses. Another example might be some forms of renewed assistance to people who repeatedly and blamably squander the resources they receive.13 Justifiable coercion is also not a necessary condition for justice. Sometimes coercive enforcement of certain demands may be all things considered unreasonable. An example of debate over this issue concerns the introduction of “speech codes” that impose penalties on agents performing   12   speech acts that carry clearly racist or sexist content. Some argue that although people have a duty of justice not to speak to others in racist or sexist ways, it would be overall a bad idea to have the state patrol linguistic practice. This position might be mistaken, but it is not conceptually incoherent. What explains the wide support for contrast IV? It seems to me that at least part of this support depends on the important intuition that demands of justice are strong enough for their enforcement to warrant serious consideration because of the involvement of rights claims. The reference to rights establishes a strong, stringent ground for proposals of enforcement that is lacking in cases of beneficence. But my proposed account captures this point while avoiding the pitfalls of IV by seeing enforceability as being only prima facie and conditional upon considerations of feasibility and reasonable costs. Proposals of enforcement can be rejected when they involve unreasonable costs, or when they would not really help fulfill a relevant duty. The point just made also affects some versions of contrast VI (Institutional / Personal) that see justice as only ranging over coercive institutions. As G. A. Cohen has argued, some duties of justice that constitute the ethos or moral culture of a just society may inform choices by individuals that are beyond the purview of coercive legal structures.14 A society without a racist culture would in one respect be more just than a society with a racist culture even if the coercive institutions of both societies were the same. An institutionalist about justice could reply that the importance of an ethos for justice can be accounted for while insisting that only institutions populate the primary focus of demands of justice. This is because institutions have significant cultural effects and can thus help shape a society’s ethos.15 This true and important point is not however sufficient to justify an exclusive primary focus on institutions. One reason for this is that the relation between a society’s institutions and its culture is one of reciprocal rather than unidirectional influence. A certain   13   cultural environment may be necessary for certain institutions to arise in the first place. This point is crucial when we address the (relatively underexplored) issue of the accessibility of just institutions besides the (often explored) issue of their stability. The latter can to a large extent be accounted for by considering how just institutions shape political culture by socializing people in certain ways. But the former cannot readily be accounted for in this way. An important part of the struggle for justice in unjust societies is the generation of a sufficiently strong political will supporting reform, and this often involves an ideological battle in civil society through which political actors succeed in shaping practical attitudes of large sectors of society before new, just institutions formalizing the spirit of those attitudes are introduced. A striking example of this is the long cultural struggle by feminists before legal frameworks implementing nondiscrimination were established.16 Furthermore, it is worth noting that some demands of justice are directly rather than indirectly focused on people’s informal attitudes and choices in their interactions with each other. Daily, informal interactions in which persons address each other with respect are part of what a just society consists in. Even if institutions were enough to shape such attitudes (something I find unlikely to be true17), their role would be partly instrumental, and thus not wholly constitutive of the primary focus of justice, as holders of contrast VI seem to assume. (iii) Determinateness We want a concept of justice that is tolerably determinate. A difficulty with contrast I (Due) in this respect is that it does not fully capture the fact that duties of justice are different from other moral demands. The characterization of justice as giving people their due is too broad. On this account, most moral duties could be seen as a duty of justice, as most actions that are moral duties (rather than, say, something that would be praiseworthy but not wrong not to do), could be construed as delivering what is due, or owed, to someone. Consider the duties to fulfill the following promises: to call you on Wednesday to go to the cinema and to pay back the money   14   you loaned me at the agreed upon time. Both duties involve my giving you what is due to you. But the first, unlike the second, does not strike us as a good candidate for a duty of justice. As I said in section 2, my proposal partly absorbs contrast V (Rights) because it sees duties of justice as responding to correlative rights. However, my proposal does not simply collapse into this contrast. One reason is that, as we saw, it also refers to prima facie enforceability. Doing this makes the concept of justice more determinate. Not every right warrants demands of justice. Some moral rights (such as your right that your friend show up at the cinema at the agreed upon time) may not be even prima facie justifiably enforceable. On the other hand, we should not be too strict about determinateness because this may lead to overlooking important moral issues. Consider for example a version of contrast III (and a view of contrast V that is tied to it), according to which someone’s right to an object is only linked to duties of justice on the part of someone else if those duties are perfect (if they are duties to do certain clearly specified things to preserve or promote access to that object by the right-holder). This seems to me to be an unduly narrow account of rights that impoverishes our thinking about them. It overlooks the possibility that we can identify rights with different levels of precision, so that some rights may correlate to both perfect and imperfect duties.18 Take for example a right to health care. Governments have perfect duties not to arbitrarily deprive any resident of access to existing public heath care programs. They may not, for example, discriminate on the basis of political opinion, race, ethnicity, or gender. But surely governments also have a duty to create some needed health care programs that do not yet exist. This duty is imperfect because it is indeterminate as to what programs of the many that are possible should be chosen (when not all can be feasibly and readily introduced at reasonable cost). It can be made more determinate at the level of policy. But its more general and imperfect (and thus somewhat indeterminate) form is itself significant. It is one of the reasonable counterparts of a constitution that includes social   15   rights, and citizens and courts could press governments to acknowledge it and take steps to fulfill it in some of the many ways possible.19 Thus, the distinction between perfect and imperfect duties does not provide a model for the distinction between justice and beneficence. As we saw, there can be duties of justice that are imperfect. As an additional and important example, consider the duty to contribute to institutional change in an unjust society. This is one of the “natural duties of justice” for individuals mentioned by Rawls in A Theory of Justice.20 It is clearly imperfect because it does not by itself specify what, and how much, any person should do to fight injustice.21 Furthermore, there may be duties of beneficence that are perfect. An example could be a duty to help certain people escape circumstances of severe deprivation when they can only do so with our help, they are the only ones who need our help, we can help at reasonable cost, and the deprivation they face is the result of their repeatedly squandering resources previously received (perhaps in fulfillment of duties of sufficientarian or egalitarian justice). (iv) Capturing the critical role of justice A fourth desideratum in accounting for duties of justice is that we avoid automatic deference to the status quo. A critical and dynamic attitude toward our current circumstances is among the crucial things we expect the standpoint of justice to help us achieve. An important case in this respect is the ability to support dynamic duties to expand current feasible sets of political action so that certain just distributions that are infeasible (or have very low feasibility) in the present become feasible (or more feasible) in the future.22 This helps notice a further problem with contrast III (Perfect / Imperfect). For example, it would be a mistake to say that there is not a duty of justice to help eradicate global poverty because it is currently unclear who should do what for whom, or because we currently do not have international institutions scheduling and enforcing specific forms of contribution securing the eradication of poverty for all. Currently imperfect   16   duties can, and often should, be made perfect through the generation of new forms of collective action and institutional structures articulating specific provisions.23 It is not unreasonable to see the introduction of such articulation as a demand of justice. But its ground may be the importance of the pre-existing imperfect duty of justice to eradicate poverty that such articulation would help specify and implement. It would have been a worrisome capitulation to the status quo to deny the existence of these demands before the construction of welfare states in domestic contexts. A similar mistake at the global level could and arguably should be avoided. The identification of perfect duties and the existence of institutional mechanisms are extremely important. They can foster efficiency by coordinating the action of many agents, they can help identify a fair allocation of responsibilities given agents’ diverse needs, abilities, and associative relations, and they can cement motivation to comply for those who are not sufficiently morally motivated and for those who need assurance that others will also do their share. Thus, we do need to focus extensively on institutional structures when we develop a conception of justice. (Of course, their coercive nature besides their far-reaching influence also makes their justifiability an important concern.24) But this does not entail that there could be no duties of justice that are not already institutionally specified. This point obviously affects some versions of contrast VI (Institutional / Personal) in addition to contrast III (Perfect / Imperfect). We have already identified several reasons for rejecting the reduction of duties of justice to institutionally articulated duties. Another reason that relates to the desiderata of avoiding question-begging views and the critical role of justice is the following. It is not a conceptual impossibility to think of a just society that has no robust, or coercive, institutional structure. An example might be an anarchist society securing liberty and equality for all through arrangements that are thoroughly voluntary.25 Although such a society is hardly achievable, its characterization as just is not a conceptual mistake. Furthermore, a picture of it may work as an ideal with the   17   help of which other, more feasible social structures can be evaluated and compared. Some may turn out to be preferable if they approximate the ideal more closely than others. (v) Illuminating practices A further desideratum for an understanding of justice is that it should help us make sense of problems in current political practices. For example, an understanding of distributive justice as ranging over enforceable rights is in tune with the modern emergence of the view that duties to eradicate severe poverty and reduce inequality are among the key responsibilities of states.26 This desideratum should also apply to the current debate on global justice, in which we can identify two important shifts. They involve the emergence of discussions over whether responsibilities of justice can rely on associative relations different from co-membership of a state and on humanist claims concerning important interests of human beings regardless of associative ties. Both shifts involve a departure from a narrow construal of justice which, as a variant of contrast VI (Institutional / Personal), holds that two persons can have duties of distributive justice toward each other only if they are both subject to state institutions that act coercively and in their name.27 The first shift involves the view that two persons may share duties of distributive justice if they are intertwined in associative frameworks that are less comprehensive than a state but are still morally consequential.28 A typical example concerns relations of deep economic interdependence. Their existence may give rise to duties to avoid exploitation of members of poorer societies (for example through unfair bargaining mechanisms in international institutions such as the World Trade Organization, or through humiliating labor conditions in sweatshops run by multinational corporations in relatively poor countries). The previous shift is compatible with contrast VII (Associative / Non-associative), but demands a wider view of the kinds of associative facts that give rise to concerns of distributive justice. The second shift is more radical, and challenges it. According to this contrast distributive   18   justice is confined to duties arising among co-members of already existing associations (economic, political, national, etc.). On this contrast, a humanist view of justice according to which some duties of justice correlate to rights that protect interests that human beings as such have is a category mistake. But such a humanist view is an important strand in the contemporary theory and practice of global justice. Universal sufficientarian claims against severe poverty are a central part of human rights discourse. And some philosophers are starting to argue that some egalitarian demands regarding some goods (such as health care, education, and work opportunities) can also take non-associativist forms. If we recognize that all human persons are ultimate units of equal moral concern and respect for everyone, then we might entertain the thought that we have duties to help others access important advantages even if due to the natural lottery they were born outside of our associative networks.29 This view has practical significance for debates about the strength and content of global duties of aid, and about the extent to which borders should be open to immigrants. The desideratum of illuminating current political practice would be flouted if we work with a concept of justice that does not help us grasp these shifts as expressing valid moves in the debate about what justice demands. The desiderata concerning the critical and dynamic role of considerations of justice and concerning the need to avoid begging important substantive questions should also make us suspicious of an understanding of justice based on contrast VII. Eliminating humanist conceptions through a definitional maneuver would be philosophically unsatisfactory. It would also uncritically defer to the status quo, which includes on the part of many the belief that they do not have duties of justice tracking rights of human beings who are not already co-members in some parochial association. To reach a considered view about what is the just way to approach the ongoing process of globalization, we need to be able to raise and   19   assess diverse associativist and humanist proposals. And to enable that discussion, we need a capacious concept of justice like the one I proposed. 4. Objections and elaborations I conclude by considering some possible objections to, and elaborations of, the account of duties of justice presented above. (1) A first worry is that the reference to “important conditions or goods” in the characterization of duties and rights of justice may not be necessary. Imagine a contract between Juan and Peter in which Peter commits to giving Juan a massage for a fee. Juan is entitled to Peter’s massage, and perhaps Peter can be said to have a duty of justice to honor the contract. But the good involved is not important. In response, we can make two points. First, we can consider specific rights and duties as cases of wider principles. In the example discussed, we can consider the importance of making and honoring contracts. Having access to successful practices of this kind is clearly an important good. Alternatively, it could be that some instances of a practice are relatively unimportant (as in the example above) while others are quite important (as in, say, transfers of real estate or full- time job contracts). We can then simply distinguish subsets of contract rights and duties, and say that only those ranging over sufficiently important goods give rise to rights and duties of justice. (The standard of “sufficient importance”, of course, is a matter of substantive conception, not of concept.) (2) Another worry targets the component of “prima facie enforceability” in the account of duties of justice proposed. The general difficulty concerns the explanatory significance of this component when contrasted with the other component referring to rights. It seems that in the order of explanation, something is a duty of justice because it protects or promotes the fulfillment   20   of an important right, and a duty is enforceable because it is a duty of justice. It does not seem true to say that something is a duty of justice because it is (inter alia) enforceable. In response, I agree that referring to rights is explanatorily more important than referring to prima facie enforceability (this is why I said, in section 2, that enforcement is justifiable when— inter alia—it is necessary or strongly contributory to the fulfillment of rights). But I think that both are explanatory significant. Prima facie enforceability is a dependent, but relevant component. This is because intuitively duties of justice involve, prima facie, justifiable limitations on people’s liberty to act. What the worry just mentioned forces me to acknowledge is that there is an explanatory structure within the account of justice such that the importance of the relevant rights determines (together with other considerations such as feasibility and reasonable costs) the justifiability of the enforcement of the duties to protect or promote their fulfillment. (3) The previous point helps me to address another worry, and to entertain a reformulation of the account proposed. The worry is that the account of justice proposed may force us to exclude imperfect duties from the domain of justice because they may be unenforceable even prima facie. If Gillian has an imperfect duty to help the destitute, she has discretion as to whom to help and how to do it. Given this discretion, we have no clear target for enforcement.30 A first response is that enforcement may still be possible when the duty-bearer has discretion. One way to do this is to demand that the agent show periodically, at defined times, that they have chosen some of the possible ways to fulfill their imperfect duty or face penalties. The agent still operates with a disjunctive set of possible ways to help (for example, they can help A or B or C…). But they can be asked to show that they have chosen at least one of the items in the disjunction. There is also the option, considered above, of “perfecting” the imperfect duty through institutional specification by for example demanding that Gillian pay taxes, which are in turn   21   used to fund specific policies of aid for certain individuals. But important as it is, this solution is partial, as there will still be some additional duties that remain imperfect. The choice by officials of relevant policies to assist the destitute may itself involve some discretion. This point has been overlooked by proponents of the strategy of “perfecting duties.” Here the imperfection is eliminated at one level but at the cost of generating it at another. Thus, this strategy cannot yield the elimination of imperfect duties. This circumstance is not, however, a source of despair if we reject contrast III, as I recommend in this paper. There may be other cases in which certain duties are unenforceable because they are simply not the kind of thing that can be enforced and are, thus, not even prima facie candidates for enforcement. This might hold in the case of certain psychological and cultural traits and dispositions that are not directly under the voluntary control of agents (such as a spirited concern for the plight of the worse off, and a readiness to participate in democratic politics when there is no legal articulation of the relevant duties). But these phenomena can still be shaped, indirectly and over the long-term, by agents. We can increase the presence of such traits and dispositions through education, material incentives, and other mechanisms. And we can be under prima facie enforceable duties to support the use of those mechanisms. For example, we may have duties to publicly fund educational opportunities, democratize access to the mass media, and much else that affects people’s knowledge and attitudes. This is another case of application of the idea of dynamic duties (see 3.iv above). A more radical response to the current worry would be to revise the account of duties of justice by relativizing the enforceability clause even further. In addition to seeing it as explanatorily dependent on the clause referring to important rights, we can see prima facie enforceability as typically, but not always or necessarily a feature of duties of justice. This could be a case in which strict analysis in terms of sufficient and necessary conditions breaks down,   22   and we rest content with identifying a set of typical features of a concept. On this weaker approach to conceptual inquiry, the characterizing features need not all hold in every case. Another possibility is to see enforceability as marking a species within the genus of justice (as pertaining, for example, to something that could be called “political” or “institutional justice”). (4) A further worry is that the acceptance of imperfect duties of justice is incompatible with the view that duties of justice correlate with rights. Some could argue that only duties that are clearly directed to specifiable claimants can correlate with rights. If Gillian can fulfill her duty to help some within a set of three persons, none of those persons has a claim against Gillian that they specifically be helped by her. This challenge can be answered in many ways. Some were already mentioned (when referring to mechanisms for “perfecting” duties). But the issue of directedness can also be tackled by noticing that Gillian’s imperfect duty is in principle directed to all of the three potential beneficiaries. Each has a claim that they be considered as a potential beneficiary. None would be wronged by Gillian if another potential beneficiary is chosen in their stead (as Gillian is permitted to choose), but they would be wronged if they were not so considered, and, of course, if none of them were helped. If the relevant duty is primarily based on the importance to the destitute of avoiding their penury (rather than, say, on the sense of elevation that Gillian would get from helping) then the duty already has a direction toward right-holders. Notice that this directedness should orient not only immediate responses in the present, but also strategies of action for the future. Besides having the normal duty to help some of the destitute now, Gillian, together with others, may have a dynamic duty to foster social arrangements under which all (or more) of the destitute get the help they need. Another possible response is to adopt a narrow view of rights as bounded up with perfect duties but to relax their characterization. We could do what was suggested above when referring   23   to enforceability, and see claimability (narrowly conceived) as a typical, but not always necessary, feature of justice. 31 Still another response is to drop acceptance of imperfect duties as duties of justice. This would simplify the discussion in (3) and (4) because there would be no need to contest or revise the reference to any of the components of the account of duties of justice proposed. However, my current state of reflective equilibrium says that we should avoid the dramatic surgery involved in this suggestion. It would require, implausibly, that we omit from the set of duties of justice some that seem to be among their paradigmatic cases (such as the duty to contribute to the reform of unjust societies and to support just ones). (5) Another potential worry is that my account of justice is too beneficiary-centered. This impression might arise because rights are said to be based on the importance of certain conditions or goods to right-holders. This seems to omit consideration of the perspective of duty-bearers. But the interest of a putative right-holder in certain conditions or goods is not sufficient for grounding a right to it; we also need to show that the agents with the putative correlative duties of protection or promotion in fact ought to acknowledge such duties. I agree that the perspective of duty-bearers is important in determining whether there is a right. Such a perspective is already taken into account when we see the enforceability of the implementation of demands of justice as depending on their feasibility and reasonable cost (which partly concern the duty-bearers’ ability to do what is demanded without shouldering unacceptable burdens). I now suggest that such concerns of feasibility and reasonable cost can also be seen to apply to the determination of the existence of the duties themselves. We can bring the perspective of duty-bearers further into focus when developing a conception of rights. In my own view, we can proceed on the basis of the following schema: If A (a right- holder) has a right to O (a certain object involving an important condition or good) against B (a   24   duty-bearer), then there are feasible and reasonable demands on B that they respect or promote, in some significant ways to be specified, A’s access to O. The specification of what B owes to A regarding O tracks the moral importance of A’s interest in O, the feasible ways for B to respect or promote A’s access to O, and the subset of such feasible forms of respect and promotion that do not involve morally unacceptable burdens on B or others (given the importance of their own interests) and on A (given the importance of other interests of A besides that concerning access to O). The specification and justification of A’s right is coterminous with the specification and justification of B’s (and other duty-bearers’) duties. No right is justified only by considering what would benefit A. Even if access to O would greatly benefit A, A does not have a right to O against B if B cannot feasibly affect A’s access to O, or could only do it in advantageous ways to A by incurring unreasonable costs. As we identify B’s duties to A, we can also factor in additional moral considerations such as responsibility and fairness. Sometimes the fact that A’s lack of access to O has resulted from blamable voluntary choices by A may diminish B’s obligations; sometimes B’s obligation may be more stringent if A’s lack of access to O is the result of what B has blamably done; and sometimes the computation of the extent of sacrifice B can reasonably incur in supporting A has to include what other agents C can and should do. The kind of holistic assessment just mentioned would also help capture the multiple considerations that can go into the identification and justification of rights and duties of justice. These may depend on associative concerns (such as those regarding justification of coercion, reciprocity in economic cooperation, equity in relations of production and political decision- making so that exploitation and domination are avoided). My rejection of a definitional reduction of the concept of justice to associative responsibilities does not deny that justice includes them. Its point is to prevent a definitional suppression of other responsibilities of justice that are not associative (as arguably at least some human rights based ones are).   25   Rights of justice, as I acknowledged, cannot simply be articulated by reference to the demand side (the interests of right-holders). We also need to illuminate the perspective of duty-bearers. But we do not need to assume as a conceptual matter the substantive view that the demand side can never, as it were, outstrip the general negative duties and special associative responsibilities of suppliers. A substantive view of justice as responding positively (within the limits of what is reasonably feasible) to the important interests of outsiders is a lively option that must be taken into account and engaged in normative argument. There are views of justice as reduced to duties not to take goods away from others, or as only recognizing frameworks of cooperation in which every participant “breaks even” (never giving more than they receive), or as deepening mutual service (when it goes beyond strict equivalence) only when it happens among people already wrapped in the flag of a particular nation, state, or common economic venture. But these views are just some candidates among others. They have to face the alternatives rather than outsource the concerns those alternatives articulate to the realm of beneficence. What would be lost if such concerns are indeed outsourced to the realm of beneficence? What would be lost, I think, is the opportunity to see how those concerns may affect the content, existence, and weight of the negative and associative duties that we already accept. Expanding the circle of justice forces us to seek more perspicuous, integrated, and practically lucid pictures of what we owe to each other.32 We should take everyone’s important interests to heart before deciding whether we can permissibly see some of them as giving rise only to weaker or supplementary calls on our charity. (6) How significant is the discussion about the distinction between justice and beneficence? Why should we care about what does and what does not belong to justice?33 Perhaps we could do all the theoretical work we need to do if we drop reference to justice and beneficence and simply talk about the different reasonable moral demands we can make on each other: some negative, some   26   positive; some grounded in rights, some not; some institutionally articulated, some interpersonal; some associative, some general; etcetera. We could simply identify these different demands, their mutual constraints, and their relative weight in different domains of application. What is crucial, it might be said, is that we have a clear picture of the different duties that capture what we owe to each other, their content, relation, and weight. Whether we call them duties of justice or beneficence is immaterial. On reflection, I think that we should keep referring to justice (construed as I suggest) as different from beneficence for four reasons. A first reason is pragmatic. The invocation of the distinction between justice and beneficence is pervasive in moral and political theory and practice. A call for dropping the terms “justice” and “beneficence” is not likely to succeed. Furthermore, it is common to say that demands of justice are more important than, or have priority over, demands of beneficence. If we want to make sure that certain important demands (such as those that are positive, and go beyond state-based associative reasons) are not downplayed or made to wait, it is wise to construe them as part of justice. It could be replied that if the distinction between justice and beneficence, with its in-built hierarchy, is bound to remain, it is also likely to fail to include the capacious view of justice that I recommend. Is not that view arbitrary in any case? The answer (and this is my second point) is that the construal I offered is not arbitrary. As I said at the outset, it is not a mere elucidation, and so it is not meant to simply report current usage. But no view that merely reports current usage will be satisfactory because there is no thoroughly agreed upon usage anyway. So we need a constructive proposal. The proposal I offer, on the other hand, is not a mere stipulation. It builds upon ideas of rights and enforceability, which are in fact quite central to the actual practice of moral and political debate (even if it construes them in a specific way).   27   Third, my proposal about how to construe justice meets the five reasonable desiderata. It helps us identify stringent demands, enables us to engage in important substantive debates, is tolerably determinate, helps us fulfill the critical role of considerations of justice, and makes it possible to clarify the stakes in important contemporary debates about justice (especially about global justice). A concept of justice with those virtues is worth using. Finally, although capacious, my account of justice does not swallow all putative demands of beneficence.34 Some of those demands are in fact absorbed by justice (as is the case with easy rescues attending to urgent needs35), but others remain beyond justice (as is the case with some ways of benefitting others that cannot plausibly be construed as responding to prima facie entitlements). My account enables the grasp of the wealth of reasons of justice, but it also honors the widespread intuition that there are some ways of making the life of others better that are not demanded by justice. We do need a separate conceptual domain for beneficence. (7) Another worry about my account of duties of justice is that it fails to capture the relational nature of justice and the active dimension of persons. The account seems to be non-relational but instead focused on end-states and it seems to see those states as befalling persons in a passive way. The account, it might be said, is fixated on the issue of what some person A gets. But it should, instead, focus on how another person B treats A, on what kinds of relations exist between A and B, and on what A and B have done to each other. Access to goods is an important, but subordinate dimension. Justice should range over persons as related and active agents, not over persons as isolated and passive recipients. Otherwise, we cannot capture the difference between B’s duties toward A if the disadvantage experienced by A is the result of a natural disaster and when it is the result of relations of economic or political exploitation or domination imposed by B. B’s action to help improve A’s condition is, in the first case, a matter of charity, whereas in the second it is a matter of justice proper.36 Besides illuminating the involvement of B, we must   28   also consider the agency of A and ponder whether A deserves B’s support, and what kind of support is fitting if A is to function as an independent person. This worry captures two important intuitions, but does not debunk my broad account of justice. The intuitions are that it matters for justice how circumstances of advantage and disadvantage result from social relations, and that it also matters that people see themselves and each other in those relations as active agents with power and responsibility. But those intuitions can be captured without surrendering the approach proposed. If B helps A to overcome the penury resulting from a natural disaster, B could be adopting a certain relational stance toward A, one in which the penury of another human being may trigger an obligation of support. We should construe the idea of a relational stance capaciously. It may include associative concerns about non-exploitation and non-domination, and general negative duties to avoid undue harm. But it may also involve a general commitment not to let fellow human beings down when they are stuck in penury. Some forms of moral solidarity might be duties of justice (while others, of course, might be duties of beneficence: for example, we may have a duty of justice to help provide clothes to the destitute, but we might have only a duty of beneficence to give them their favorite clothes37). In such cases, when we consider others’ states of disadvantage we do not only ask “Have I done that to them?” or “Was it my fault?” We also ask “May I allow that to go on?” or “May I let that happen to them?” Injustice may be done to others through indifference. Several issues of agency are of course important. Resources for support do not normally materialize without being the result of productive efforts by some agents (thus the reference to fairness to duty-bearers mentioned in 4(5) above). And those helped are not to be seen as mere passive receptacles of beneficial stuff, but as agents who could be participants in their own amelioration38 (and sometimes morally responsible for bringing about their own disadvantage). But we can, and on some substantive views of justice we should, also pay attention to the   29   significance of need and mutual dependency. A radical ideal of independence is arguably infeasible and undesirable. It is part of what persons are that they need the support of others to avoid the bad and the wrong and to pursue the good and the right. We can, and arguably we should, relate to each other so as to respond to that vulnerability with solidarity. Persons are not merely passive recipients, but they are passive to an important extent. They should not have to feel humiliated if they are helped, and they should not have to feel guilt when they help others. We can capture these points by acknowledging (multiply qualified) general rights and duties of justice of mutual support. (8) I conclude these reflections by considering briefly one of the hardest issues in the discussion about the idea of justice: How do we identify the difference between a concept and a conception? Is this difference sharp or porous? A view I find worth entertaining is that this difference is in fact not tight. We can see the concept of justice as marking a territory of debate that is valuable to identify and explore. But the boundaries of that territory depend on our substantive interest in areas of it, and in our concern for engaging each other in its examination even if we disagree about the content of the important areas. On this view, our conceptual commitments are partly driven by our substantive normative beliefs. The former are meant to help us articulate, voice, and assess the latter. But why include in a concept the possibility of a conception one considers false? Because (inter alia) one has a quite general, substantive commitment to engage other human beings in argument about what we owe to each other. To do this, we need conceptual tools that allow us to work (and also quarrel) together. I realize that this picture of the debate about concept and conceptions of justice may assume a humanist cosmopolitan commitment to the framing importance of equal respect and concern for all human persons. But if one cannot fully disengage concept from conception at some point, this seems a point worth resting at.   30   Acknowledgments For helpful comments or conversations I thank referees, Arash Abizadeh, Elizabeth Ashford, Charles Beitz, Rainer Forst, Francisco Garcia Gibson, Mariano Garreta-Leclercq, Waheed Hussain, Carl Knight, Rahul Kumar, Barbara Herman, Louis-Philippe Hodgson, Alistair Macleod, Julio Montero, Mark Nelson, Henry Richardson, Andrea Sangiovanni, Peter Singer, Laura Valentini, Alex Voorhoeve, and participants in colloquia at the Centro de Investigaciones Filosoficas (Buenos Aires), Princeton University Center for Human Values, Queen’s University, University of Montreal, and Glasgow University. My research was supported by a grant from the Social Sciences and Humanities Research Council of Canada. Note on contributor A native of Argentina, Pablo Gilabert is an Associate Professor in the Department of Philosophy at Concordia University (Montreal, Canada). He has been an HLA Hart Visiting Fellow at the University of Oxford, a DAAD Fellow at the University of Frankfurt, a Visiting Fellow at the Australian National University, and a Laurance S. Rockefeller Visiting Faculty Fellow in the Center for Human Values at Princeton University. His papers appeared in journals such as The Journal of Political Philosophy, Political Theory, The Philosophical Quarterly, Philosophy and Phenomenological Research, Philosophical Studies, Political Studies and Ethical Theory and Moral Practice. He is the author of From Global Poverty to Global Equality. A Philosophical Exploration (Oxford: Oxford University Press, 2012).   31   References Ashford, Elizabeth, 2007. The duties imposed by the human right to basic necessities. In: T. Pogge, ed., Freedom from poverty as a human right. Oxford: Oxford University Press, 183-218. Beitz, Charles, 2009. The idea of human rights. Oxford: Oxford University Press. Buchanan, Allen, 1987. Justice and charity. Ethics, 97, 558-575. Buchanan, Allen, 2004. Justice, legitimacy, and self-determination. Oxford: Oxford University Press. Caney, Simon, 2011. Humanity, associations, and global justice. Monist, 94, 506-534. Cohen, Gerald, 2008. Rescuing justice and equality. Cambridge, MA: Harvard University Press. Cohen, Joshua, 2002. Taking people as they are? Philosophy and Public Affairs, 30, 363-386. Cohen, Joshua and Sabel, Charles, 2006. Extra republicam nulla justitia? Philosophy and public affairs, 34, 147-175. Fleischacker, Samuel, 2004. A short history of distributive justice. Cambridge, MA: Harvard University Press. Forst, Rainer, 2012. The right to justification. New York: Columbia University Press. Gauri, Varun, and Brinks, Daniel, 2012. Human rights as demands for communicative action. Journal of political philosophy, 29, 407-431. Gilabert, Pablo, 2009. The feasibility of basic socioeconomic human rights. A conceptual exploration. Philosophical quarterly, 59, 559-581. Gilabert, Pablo, 2010. Kant and the claims of the poor. Philosophy and phenomenological research, 81, 382-418. Gilabert, Pablo, 2011a. Humanist and political perspectives on human rights. Political theory, 39, 439-467. Gilabert, Pablo, 2011b. Feasibility and socialism. Journal of political philosophy, 19, 52-63.   32   Gilabert, Pablo, 2012. From global poverty to global equality. Oxford: Oxford University Press. Gilabert, Pablo, 2013. The capability approach and the debate between humanist and political perspectives on human rights. A critical survey. Human rights review, 14, 299-325. Gilabert, Pablo, 2015. The socialist principle ‘from each according to their abilities, to each according to their needs’. Journal of social philosophy, 46, 197-225. Gramsci, Antonio, 2000. The Antonio Gramsci reader, ed. D. Forgacs. New York: NYU Press. Kagan, Shelly, 1998. Normative ethics. Boulder, CO: Westview Press. Hart, HLA, 1994. The concept of law, 2nd. ed. Oxford: Oxford University Press. Miller, David, 1999. Principles of social justice. Cambridge, MA: Harvard University Press. Miller, David, 2007. National responsibility and global justice. Oxford: Oxford University Press. Miller, Richard, 2010. Globalizing justice. Oxford: Oxford University Press. Nagel, Thomas, 2005. The problem of global justice. Philosophy and public affairs, 33, 113-147. Narveson, Jan, 1988. The libertarian idea. Philadelphia, PA: Temple University Press. Nozick, Robert, 1974. Anarchy, state and utopia. New York: Basic Books Nozick, Robert, 1981. Philosophical explanations. Cambridge, MA: Harvard University Press. Nussbaum, Martha, 2000. Duties of justice, duties of material aid: Cicero’s problematic legacy. Journal of political philosophy, 8, 176-206. O’Neill, Onora, 1996. Toward justice and virtue. Cambridge: Cambridge University Press. Pogge, Thomas, 2008. World poverty and human rights, 2nd. ed. Cambridge: Polity. Rawls, John, 1999a. A theory of justice, rev. ed. Cambridge, MA: Harvard University Press. Rawls, John, 1999b. The law of peoples. Cambridge, MA: Harvard University Press. Sangiovanni, Andrea, 2007. Global justice, reciprocity, and the state. Philosophy and public affairs, 35, 2-39.   33   Sen, Amartya, 2004. Elements of a theory of human rights. Philosophy and public affairs, 32, 315-356. Sen, Amartya, 2009. The idea of justice. Cambridge, MA: Harvard University Press. Shue, Henry, 1988. Mediating duties. Ethics, 98, 687-704. Singer, Peter, 1972. Famine, affluence, and morality. Philosophy and public affairs, 1, 229-243. Swift, Adam, 2006. Political philosophy, 2nd. ed. Cambridge: Polity. Tan, Kok-Chor, 2004. Justice without borders. Cambridge: Cambridge University Press. Van Parijs, Philippe, 2007. International distributive justice. In: R. Goodin, P. Pettit, T. Pogge eds., A companion to contemporary political philosophy, 2nd .ed. Oxford: Blackwell, 638-652. Waldron, Jeremy, 1993. Liberal rights. Cambridge: Cambridge University Press. Wiggins, David, 2006. Ethics. Cambridge, MA: Harvard University Press. Wittgenstein, Ludwig, 2009. Philosophical investigations, 4rd. ed. Oxford: Wiley-Blackwell.                                                                                                                 1 For the distinction between concept and conception see Rawls (1999a, p. 5) and Hart (1994, pp. 160, 246). 2 The phrases “conditions and goods,” “entitlements,” “rights,” and “duties,” are broad enough to allow for whatever a conception takes to be the distributively relevant (e.g. they allow comparative considerations about the standing of persons in relation to others). 3 Some such features (even if they do not all apply in every case) may mark the “family resemblance” amongst duties of justice. On “family resemblance” see Wittgenstein (2009, sects. 65-7). The use of this approach might make a proposal less open to challenge via counterexamples. The latter can be seen as identifying borderline cases that lie beyond the core of the concept. For deployments of this strategy see section 4. 4 For the point that justice concerns giving each person their due, see Wiggins (2006, Lecture 10) and Cohen (2008, p. 7). For elucidation and critical discussion of contrasts II through V, see Buchanan (1987). For a presentation of contrasts IV and V see Miller (2007, pp. 248-9). See also Miller (1999, p.76): “One test of the distinction between justice and humanity is whether those in need are regarded as having enforceable claims to the resources that will meet their needs, and correspondingly whether potential donors are regarded as being under enforceable obligations to provide those resources.” An example of contrast VI appears in Tan (2004, p.21). Tan actually focuses on the related question of the distinction between “justice” and “ethics,” claiming (on the basis of an interpretation of Rawls’s theory of justice) that the former has an “institutional focus” while the latter has an “interactional focus.” Tan distinguishes his view from Pogge’s view, according to which an “institutional approach not only has an institutional focus, but it also has an institutional basis or justification,” and he does not “share the justificatory claim in Pogge’s approach that we have duties of justice only in so far as we are causally via our institutions responsible for injustices” (Ibid.). For Pogge’s view, see his (2008). Rawls (1999b, pp. 195ff.) contrasts distributive justice and humanitarian assistance. Contrast VII is embraced by the majority of philosophers in the current debate on global justice (although the normatively relevant forms of association are characterized in different ways—see notes 27-28 below). For challenges to it see, however, Caney (2011) and Gilabert (2012, chs. 5-6). My discussion focuses on contemporay debates. For historical explorations, see Fleischacker (2004); and Nussbaum ((2000). 5 Gilabert (2012, pp. 11-8). This paper differs from that earlier text in the following ways: an account of deliberative interpretive proposals is advanced, the characterization of justice is changed to include reference to access to goods   34                                                                                                                                                                                                                                                                                                                                                                           and conditions rather than merely to advantages, the list of desiderata is explicitly stated and is expanded to include determinateness and illumination of practices, the contrast Associative / Non-associative is added, the arguments defending the proposed characterization of justice are rephrased to make them sharper and additional arguments are advanced. The final section of this paper takes the discussion further by addressing possible challenges and by providing further elaborations of the proposed characterization of justice. Finally, this paper does not endorse a deflationary conclusion about the importance of the distinction between justice and beneficence. 6 Swift (2006, p. 6). 7 Nozick (1981, pp. 498-504). As Francisco Garcia Gibson pointed out to me, considerations about costs to agents may affect duties of justice in different ways (e.g. by impacting their content, weight, or enforceability). 8 Rawls (1999, pp. 266-7). 9 See, e.g., Van Parijs (2007, p. 641). 10 For the libertarian view, see Nozick (1974) and Narveson (1988). 11 Buchanan (1987, pp. 562-9). 12 Rawls (1999a, pp. 249-51, 291-2). 13 Miller (2007, p. 249). 14 Cohen (2008, ch. 3). See also Sen (2009, pp. x-xi, 10, 67-9, 75-86). 15 See Cohen (2002). See also Rawls, (1999a, sect. 41). 16 Another instructive example is Gramsci’s view that socialist transformations in Western democratic societies are only likely to succeed if socialist activists manage to make their values and demands hegemonic within the sphere of civil society before they find realization through state action. See Gramsci’s discussion on the phenonemon of hegemony within civil society in Gramsci (2000, pp. 195, 205-6, 211-2, 249, 306-7, 333-4, 345). 17 A similar skepticism might partly explain why it is a good thing that the Convention for the Elimination of all Forms of Discrimination Against Women ranges over “the political, economic, social, cultural, civil or any other field” (Article 1). Part of the difficulty of this discussion turns on how to handle the distinction between what contitutes and what causes justice. See Cohen (2008, pp. 377-81). 18 I agree with Waldron (1993, pp. 25, 212) that rights generate “waves of duties” of many kinds. 19 On the role of courts in getting the political process to acknowledge and specify rigths see Gauri & Brinks (2012). 20 Rawls (1999a, pp. 99, 293-4). 21 The natural duty of justice has another part that concerns the obligation to support existing just schemes. A critic might say that this part connects only to perfect obligations. But this would be a mistake. One aspect of the duty to support just schemes, the obligation to obey just laws, is a perfect duty. But the members of a just scheme have a wider, imperfect obligation to actively engage in activities that help keep the just arrangements alive, such as voting (when it is not legally mandatory), participating in political parties, public debates, etc. 22 For elucidation and exploration of the category of “dynamic duties,” see Gilabert (2009; 2011b). 23 Shue (1988); Gilabert (2010). 24 Furthermore, institutional structures sometimes create certain roles, and thus give rise to specific interests and responsibilities (Leif Wenar, “Justice and Charity. Rawls, Roles, and Rights”—Talk delivered in the wokshop “Justice and Beneficence.” Princeton University, 9 November 2012). 25 These arrangements would include certain conventions (including those of language). These are of course, in one sense of the term, “institutional facts.” But I am here focusing on more formalized institutions whose rules are shaped and applied in a way that involves top-down command, without requiring the direct consent and control on the part of (at least some of) their subjects when they are applied. 26 For a historical study of the emergence of this idea, from Babeuf to Rawls (and their surrounding political contexts), see Fleischacker (2004, ch. 3). See also Miller (1999, ch. 1). Miller suggests that socialist challenges to capitalism played an important role in the emergence of the contemporary idea of “social justice” (p. 3). On socialist justice, see Gilabert (2015). 27 Nagel (2005). 28 Cohen & Sabel (2006); Sangiovanni (2007); Miller (2010). 29 Just as we can recognize agent-neutral besides agent-relative duties in the case of human rights, we can acknowledge both kinds of duties in the case of some egalitarian claims. See Buchanan (2004); Caney (2011); Gilabert (2012). For a different, “practical” characterization of human rights that does not rely on a humanist approach see Beitz (2009). Gilabert (2011a; 2013) argue that the insights in such a view can, and should, be combined with the humanist approach.   35                                                                                                                                                                                                                                                                                                                                                                           30 The problem of the imperfection of duties in fact involves many possibilities depending on what aspect of a duty we focus on. The difficulties may be to identify the duties’ corresponding (a) agents, (b) acts or omissions, (c) beneficiaries, or (d) circumstances of application. 31 For other discussions on claimability see the specific proposal by O’Neill (1996), and the criticisms to it by Sen, (2004) and Ashford (2007). 32 This does not involve obliterating the distinctions between different kinds of duties of justice (negative and positive, general and associative, and the different varieties of each). They can and should all be articulated. The key point is that their mutual relations become an explicit and necessary target of discussion. 33 Cohen (2008, pp. 289-90 n.14) reports Joseph Raz’s puzzlement on this score. Some authors claim that the reference to justice does not add anything fundamental to a full account of moral duties, and that it is in fact wholly derivative. See Kagan (1998, pp. 176-7) and Buchanan (1987, pp. 574-5). 34 It also does not obliterate (and in fact invites the grasp of) different reasons of justice. See note 32 (and surrounding text). 35 Thus, Peter Singer construes the positive duty to help reduce global poverty as a matter of beneficence (on the basis of the famous analogy with the duty to rescue a child drowning in a shallow pond). See Singer (1972). But we can construe this positive duty as a demand of justice. We could then preempt the common reaction (for example by officials of governments of wealthy nations) that attending to this duty must wait until other, more stringent demands of justice have been taken care of first. 36 Forst (2012, p. 4). 37 I owe this example to Peter Singer (who does not, however, support my account of the distinction between justice and beneficence). 38 I think that this should include recognizing a right to participate in the political regulation of practices of production and distribution. For this important point see Forst (2012,p. 4).