•III•
•7•
As far as I can tell, one of the most useful effects of the first edition of Basic Rights was to help to throw open a number of issues about the duties that need to be performed if rights are in fact to be secure. Or perhaps I should say “help to throw into confusion.” In any case, during the subsequent years I along with others, some of whom I will mention below, have continued to struggle with issues about both the specification and the assignment of right-grounded duties, two intertwined complexes of problems concerning the “division of moral labor.”1 Here I would like to rework once again some of the persistent questions about rights and their duties, indicating a little about how my picture of them has evolved so far.
John Locke had taken for granted that the right to accumulate private property was limited by a universal right to subsistence.2 A balanced picture of rights of this kind, including guarantees for both liberties and subsistence, had been reaffirmed by various U.S. political leaders, most powerfully perhaps by Franklin Delano Roosevelt.3 In the first year of the Carter Administration (1977), Secretary of State Cyrus Vance announced a balanced view as the basis for the enhanced attention that, it was promised, would be given to human rights in U.S. foreign policy.4 The secretary of state’s affirmation of rights nevertheless quickly came under philosophical attack, including a politically influential critique in 1978 by Charles Frankel, maintaining that “the list proposed by Secretary Vance, and particularly the economic and social rights of which it speaks, must be characterized as somewhat puzzling.”5 My discussion of duties in chapter 2 was initially intended primarily to assist, by portraying a little more realistically how rights function in practice, with the refutation of some bad arguments against balanced accounts of rights like those embraced by Locke, Roosevelt, and Carter/Vance, and of course embodied in international law. Several bad arguments depended upon, among other things, a highly overabstracted and thereby oversimplified picture of rights as falling into two neat types: “negative rights” and “positive rights.” (To this day many theorists portray this crude polarity as a distinction crucial to the architecture of all rights.) The supposed “negative” and “positive” rights were being forced into this simplistic abstract dichotomy on the basis, not of the nature of the substance of the right (what the right was a right to), but of the supposed nature of the duties required to fulfill the right, which were in turn being given only half-hearted attention.
Each of a number of rights has as its substance one of the varieties of liberty, for example. If everything must simply be divided into negative and positive, some liberties are clearly “positive”; for example, many political liberties involve active participation in self-government, which is a highly active, and in that sense positive, undertaking. So even if one did concentrate on the substance of liberty rights themselves, and one was determined to label everything either positive or negative, some rights would turn out to be positive because some liberties are positive. But other liberties, looked at simply in themselves, would indeed turn out to be “negative”; some types of liberty consist mainly in being left alone by all other persons, not in, say, participating actively in one’s own self-government. So one could have reasonably said that a right to a liberty of noninterference was a negative right in the sense that what is protected when this right is protected is negative: the protected liberty consists at its core in not being interfered with. This obviously leaves out all that must be done to protect, and otherwise provide for, even such a negative liberty, which is by no means mainly negative, but to take implementation seriously is precisely to turn to duties.
Yet an appeal to the core substance of rights was not the kind of argument being made two decades ago. When Charles Frankel, for example, attacked Cyrus Vance’s balanced list of rights for including the economic and social rights that are entrenched in the international law of human rights, Frankel did not argue that the protected content of all economic and social rights is more like positive rights to political liberties than it is like negative rights to noninterference. Instead, he emphasized the onerousness and alleged impracticality of the duties allegedly distinctively involved in fulfilling economic and social rights—the supposed “dangerously utopian overtones” of seriously attempting to honor them.6
The picture underlying such criticisms seemed, then, fundamentally to be that rights are negative if the duties necessary to fulfill them are (exclusively) negative, and rights are positive if the duties necessary to fulfill them are (at least partly) positive. It was then announced that the performance of the positive duties would be too much reasonably to ask of people, if not literally impossible for them to bear, and that, accordingly, the positive so-called rights to which these onerous duties were correlative would have to be disqualified from being genuine rights, although they could be kept as the goals of aspirations. Most critics of the balanced view simply helped themselves as well to the crucial, and false, assumptions that all economic and social rights are positive and no civil and political rights are positive, on the basis of one or two quick examples.
Against this I suggested, in chapter 2, that while some duties are at the negative end of the spectrum and others are at the positive (and many are in between), no right can, if one looks at social reality, be secured by the fulfillment of only one duty, or only one kind of duty. If one looks concretely at specific rights and the particular arrangements that it takes to defend or fulfill them, it always turns out in concrete cases to involve a mixed bag of actions and omissions. Two points became evident. First, in general, while one can line up particular duties correlative to various rights on a spectrum running from onerous actions to costless omissions, what one cannot find in practice is a right that is fully honored, or merely even adequately protected, only by negative duties or only by positive duties. It is impossible, therefore, meaningfully and exhaustively to split all rights into two kinds based upon the nature of their implementing duties, because the duties are always a mixture of positive and negative ones.
Second, specifically concerning the economic and social rights that are the intended target of the dismissive dichotomy, it is simply not the case that all, or most, civil and political rights can be fulfilled entirely or mostly by negative duties, while all, or most, economic and social rights must be fulfilled entirely or mostly by positive duties. This second point is true for a myriad of reasons, not least that civil and political rights, on the one hand, and economic and social rights, on the other, often cannot be fulfilled separately from each other.
The main point so far is simply that duties are complicated, far beyond anything reducible to, or helpfully summarized as, negative/positive. Duties do not come in only two kinds, but in several, and the provisions for the secure enjoyment of any one thing that people have a right to will involve duties of several kinds. Jeremy Waldron has recently captured this in the apt image of “successive waves of duty”: “Each right is best thought of not as correlative to one particular duty (which might then be classified as a duty of omission or as a positive duty of action or assistance), but as generating successive waves of duty, some of them duties of omission, some of them duties of commission, some of them too complicated to fit easily under either heading.”7 We may, for example, first tell others not to interfere with what a person has a right to. When we find that some are interfering anyway, we establish police to stop them. Then when the police turn out to be corrupt, we reeducate the police and go after the drug dealers who corrupted them. Going after the drug dealers may be possible only after defeating the incumbent politicians, which may depend upon an uncorrupt judiciary. And so forth, until the person actually enjoys what Thomas W. Pogge nicely calls “secure access” to the substance of her right.8 The only way to be safe from interference in one’s private life may be first to become politically active, as many a ghetto resident has reluctantly discovered. That requires usable political rights, the implementation of which depends upon the performance by a wide range of people of another wide range of duties. The need for such second, third, and further waves of duties is readily foreseeable. Adequate arrangements to secure any right will include advance provisions for Waldron’s predictably necessary successive waves.
The intricate interweaving of rights of diverse kinds also suggests the image of a fabric, especially one of those rough fabrics with threads of different sizes and strengths. Some of the threads are stronger and more crucial than others—the basic rights—but even the strong threads support each other and support and are supported by the weaker threads. A seeker of privacy will be more secure in a society in which rights of political participation are well protected, because the addicts will keep breaking in as long as the politicians ignore the drug trade. The privacy seeker will also do better, other things being equal, if there is a free press and due process.
One pitfall for philosophers is already evident here: the danger of drifting into unhelpfully saying merely that no right can be safely enjoyed unless numberless people perform innumerable duties. One strategy I employed in the first edition to guard against this danger was the strategy of attempting to identify some of the basic rights. While I never intended to suggest, as some people have misunderstood me, that the three basic rights I discussed at greater length—physical security, subsistence, and liberty (taken to include both social participation and physical movement)—were the only basic rights, I would still defend the attempt to identify strategically critical rights.9 However, I do not want to pursue that here but instead to stick with issues about duties.
Unfortunately for us theorists, while it would indeed be unhelpful to say merely that no right can be safely enjoyed unless numberless people perform innumerable duties, this is in fact the case. Indeed, notoriously, most people whom we ever actually encounter must perform their negative duties if we are to enjoy even a reasonable level of physical safety. Anyone who has ever blundered into a situation in which it was entirely unclear whether large numbers of the strangers present had peaceable intentions can testify to the level of insecurity, psychological and physical, that arises from actually not being able to count upon negative duties generally being performed, as those of us in fortunate situations normally can. Theorists, however, must find more to say than that it is all complicated and that a lot of people, including total strangers, bear a lot of duties toward each other.
In chapter 2 I suggested my “very simple tripartite typology of duties” (p. 52), shortly adding a few subcategories (p. 60). Evidently, chapter 2 in general and the tripartite typology in particular have been by far the most discussed portions of the book. A number of elaborations seem now to be called for. First, the practical focus of this book is, as indicated in the subtitle, the foreign policy of the United States. The two most prominent ways that consideration of human rights is built into the legislation governing U.S. foreign policy are both sections of the Foreign Assistance Act: Section 116, usually referred to as the “Harkin Amendment,” after its sponsor Senator Tom Harkin of Iowa; and Section 502B. Both Section 116, which governs economic assistance including food aid, and Section 502B, which governs “security assistance”—that is, military weapons and training—take the form of requiring the cessation of foreign aid to regimes whose violations of human rights are especially egregious. Both these sections were inserted into foreign policy legislation in attempts by Congress to rein in the policy, regularly practiced by Henry Kissinger and others, of subsidizing any regime with an anti-Communist foreign policy no matter how outrageous its domestic policies of torture, arbitrary execution, imprisonment of dissidents, and so on.10 This meant that the portion of U.S. foreign policy on which consideration of human rights was getting what little bite it ever got was the relatively small portion concerned with foreign aid.
In order for me to focus at the time on the relevant portion of foreign policy, then, I needed to focus on foreign assistance. Actually the focus, for the sake of direct relevance, needed to be on foreign aid for two independent reasons. The less important reason was that, since the action required by the two relevant amendments was the cessation of aid, it was useful to be aware of any countervailing reasons for continuing aid that might flow from some duty to aid. Far more important was the fact that the United States was an active accomplice in many of the worst violations of human rights on the anti-Communist side of the Cold War because we were lavishing assistance upon governments hated by their own people in attempts to buy these governments’ support of U.S. foreign policy goals. In some notorious cases, like Iran and Guatemala, the Central Intelligence Agency had literally put the governments into power through “covert” action, and in many others, like South Korea and the Philippines, U.S. weapons and training were pouring in to dictators who had helped themselves to power. Much of the outrage in Congress was because congressional hearings in the 1970s, prior to the Carter presidency, had revealed the fact that, for example, arbitrary executions and torture were being carried out by military officers not only armed but trained in special schools by the United States.11
Of the three general categories of duties I had suggested, then, it was the third, duties to aid, that I thought I needed to concentrate on (p. 113) in chapters 5 and 6, in spite of the fact that duties to aid become relevant only after failures to perform the first two general kinds of duty. Thus the figure (p. 115), as indicated by the sub-title of this section of the chapter (p. 114), is specifically about duties to aid people whose rights have already been violated, which, I believe, makes the position embodied in it less extreme than it may otherwise seem.12
From a more purely theoretical point of view, my concentration of attention on duties to aid, which are, in Waldron’s image, one of the last waves of duties—what you must do when you have not done anything else you should have done—was unfortunate.13 Devoting so much attention to duties to aid was unfortunate above all because it obscured the importance of the second general kind of duty, the duty to protect, and most specifically what I labeled duty II-2, the duty to protect people from violation of their rights by the design of better institutions (p. 60). While I fairly prominently mentioned a duty to create effective institutions (for example, pp. 17, 59–60, and 198) and analyzed some ways that rights deprivations are essential to, or inherent in, institutionalized economic strategies (pp. 46–51), I gave the creation of better institutions designed to ensure that rights are respected in the first place less attention than I now believe it deserves.
Worse, I conceived of duty II-2 much too negatively and much too narrowly. Duty II-2 is said to be “to protect from deprivation . . . by designing institutions that avoid the creation of strong incentives to violate duty I [which is to avoid depriving people of what they have a right to]” (p. 60). Certainly it is a good thing to produce institutions that avoid incentives to violate rights; this is why the CIA, for example, ought to be totally transformed or abolished. The CIA systematically generates temptations for its employees to violate fundamental human rights to which they routinely succumb. Merely eliminating evil (and phenomenally wasteful, corrupt, and incompetent) institutions like the CIA is, however, far too negative and narrow. Accordingly, some scholar/activists have retained my suggested three underlying categories (avoid depriving, protect from deprivation, and aid the deprived) and given broader and more positive content to the second duty to protect.14 Others have proposed a typology containing a more constructive third category, “to respect, to protect, and to fulfill,” where to fulfill rights clearly involves more than aiding those whose rights one has already violated and includes the creation of more effective institutions to see that rights are honored in the first instance.15
Now, almost everyone involved in these discussions realizes that typologies are not the point. Typologies are at best abstract instruments for temporarily fending off the complexities of concrete reality that threaten to overwhelm our circuits. Be they dichotomous or trichotomous, typologies are ladders to be climbed and left behind, not monuments to be caressed or polished. Thus, there is no ultimately significant question of the form, how many kinds of duties are involved in honoring rights? Three? Four? A dozen? Waldron is closer to the mark in saying “successive waves of duty.” How many waves? Lots—more sometimes than others.
The “very simple tripartite typology of duties,” then, was not supposed to become a new frozen abstraction to occupy the same rigid conceptual space previously held by “negative rights” and “positive rights.” The critical point was: do not let any theorist tell you that the concrete reality of rights enforcement is so simple that all the implementation of any right can usefully be summed up as either positive or negative. The constructive point was: look at what it actually takes to enable people to be secure against the standard, predictable threats to their rights—focus on the duties required to implement the right.
This form of analysis means treating the securing of rights as ends and arriving at adequate arrangements through means/end, or strategic, reasoning.16 James W. Nickel has put it this way: “Where human rights violations are deep and systematic, rights advocates must devise strategies for political change that are not in the scopes of human rights. Here respect for and implementation of human rights becomes a goal, and something like consequentialist reasoning must be used to pursue this goal.”17 I have quoted a negative point and a positive point from Nickel. The negative point is that “strategies . . . are not in the scopes of human rights.” He had earlier noted that “rights are to some freedom or benefit. We can say that the scope of a right specifies what it is to.”18 The point is, then, that knowledge of what a right is to is insufficient to tell us how to guarantee it, and I take this to entail that conceptual analysis alone (of the scope, substance, or content of a right) provides inadequate information for grounding judgments about implementation. It is necessary, but not sufficient, to understand the conception of the right and what the right is to. Knowing how to protect the right against violation, or to restore the right after violation, depends as well on historical and empirical understanding of the relevant social, economic, political, legal, and psychological factors. As long as theorists remain narrow specialists, adequate analysis of how to institutionalize a right requires interdisciplinary collaboration. It is certainly nothing that ethical theory alone can settle.19
Nickel’s positive point is put: “something like consequentialist reasoning must be used to pursue this goal.” The emphasis ought to go on “something like,” or this phrasing can be misleading. Better, I think, simply to say that means/end, instrumental, or strategic reasoning is needed.20 One’s goal is for everyone to enjoy everything they have a right to enjoy. One’s next task is to figure out how this can be arranged and how to allocate the tasks involved in the establishment and operation of the arrangements. One must, in short, help to design social institutions that protect and fulfill rights.
Thomas Pogge suggests what he calls an institutional understanding of human rights, which he contrasts with what he labels the interactional understanding.21 On the interactional understanding, which he notes found its canonical expression in Hohfeld’s ritualistically invoked diagram, each human “right entails certain directly corresponding duties.”22 On Pogge’s recommended institutional
understanding, too, human rights (conceptually) entail moral duties—but these are not corresponding duties in any simple way: The human right not to be subjected to cruel or degrading treatment gives me a duty to help ensure that those living in my society need not endure such treatment. Depending on context, this duty may . . . generate obligations to advocate and support programs to improve literacy and unemployment benefits when such programs are necessary to secure the object of this human right for a class of my compatriots (domestic servants).23
Earlier in the article Pogge has explained the connection between literacy and unemployment benefits and protection against cruel and degrading treatment:
In some of these societies, inhuman or degrading treatment of domestic servants by their employers is perfectly legal. In others, certain legal prohibitions are in place but ineffective: Most of the servants, often illiterate, are ignorant of their legal rights, convictions for mistreatment are extremely difficult if not impossible to obtain, punishments are negligible. Moreover, servants are also often forced to endure illegal conduct on account of economic necessity: They do not dare file complaints against their employers for fear of being fired. This fear is both justified and substantial, because they have only minimal financial reserves, there is a general oversupply of servants and/or they have reason to believe that their present employer would refuse to issue them the positive reference requisite to find new employment.24
Pogge has a number of separable theses, not all of them relevant here. One, which might be called the indirection thesis, is that duties are not “directly corresponding” to rights and “are not corresponding duties in any simple way”: the route to the fulfillment of a right is often not the direct route through a supposed corresponding, Hohfeldian duty. The simple, direct picture Pogge is rejecting is the one, I believe, that leads people to think that the right not to be subject to cruel and degrading treatment is a purely negative right because its fulfillment is completely constituted by the noninterference prescribed by a purely negative duty.25 If employers are molesting their maids, one simply says, in effect: “keep your hands off your maid”—or, in the language of duty, “you have a duty not to deprive your maid of her physical security” (especially since, if she lives-in, she is basically trapped, having no other home to flee to if she wants to earn a living). If employers would keep their hands off their live-in maids, there would naturally be no problem.
The real problem, Pogge correctly sees, is that many employers do not in fact leave their maids alone, and this will in fact not be changed merely by the proclamation of duties of noninterference. Nor will the police in fact be checking the homes of the well-to-do to see whether their owners are respecting the rights of their live-in servants and to protect any servants who need protection. In reality, live-in servants will be secure only if they are, in a word, empowered—made agents in their own rescue. The servants must be taught their rights, taught how to demand them, and provided with some third alternative to submission inside the employer’s house or starvation outside in the streets. Learning about the rights may require first being taught to read and write, and having a viable third option may require being provided with unemployment insurance, as Pogge is rightly indicating. In reality the route to physical security against inhuman or degrading treatment often runs indirectly through primary education and minimal economic security.
James Nickel emphasizes the same web of interdependent rights and duties, and consequent importance of the agency of victims, in the case of Third World hunger:
Most people who experience hunger and malnutrition are functional, are getting water and a little food, and are capable of doing things to find food such as moving or seeking work. If we think of hungry and malnourished people as agents, albeit agents with limited capacities and options, we will avoid assuming that self-help is impossible and that only donated foreign food or money can address the problem. Further, viewing hungry people as agents is a more respectful stance that provides a barrier to the paternalistic attitude that it is mainly rich people from the First World who are competent to address problems of hunger in poorer countries. The purpose of reframing the problem of world hunger is not to get people in rich countries off the hook, but to have a better idea of which hook they should be on.26
The “hook” that the rich-country folk need to be on may be the urgency of ceasing military assistance to a regime that denies its own people avenues of protest while supporting the rich country’s foreign policy goals, that is, a nonintervention hook that allows domestic agency to work, not an international intervention hook. While there certainly is a right to food, the more urgent rights for many hungry people are the right to peaceful assembly and the right to vote. Figuring out what is urgent in a particular context depends upon the means/end or strategic reasoning that Nickel had earlier recommended in Making Sense of Human Rights.
I said earlier simply that duties are complicated. I have been emphasizing one particular respect in which that is true. One cannot stay at the purely conceptual level, reasoning simply that if there is a right to have x, there must be a duty for others to provide x. There may instead be a duty to stay out of people’s way while they take x for themselves, or a duty to teach them to read so they can figure out how to make or grow x, or a duty to let them form a political party so that they can effectively demand that the government stop exporting x (instead of having the CIA arm their police so that they can suppress all dissent). Sometimes there is a duty to provide x to them, conditionally or unconditionally. Positively put, which duties there are depends upon means/end reasoning in which secure enjoyment of the substance of the right in question is the end. Means/end reasoning must be not simply what philosophers like to call “empirical,” but strategic, involving judgments about policies and institutions. Often the means to the fulfillment of one right will include the fulfillment of other rights, because rights may be of great instrumental value in the fulfillment of other rights regardless of whether they are of intrinsic value as well.
Moreover, one certainly cannot definitively allocate duties among potential bearers—determine who ought to do what—until means/end reasoning has established what needs to be done, or at the very least, what are the most promising ways to accomplish what needs to be done if the rights in question are to be enjoyed. In other words, the allocation of duties among bearers depends upon the specification of which duties need to be performed, and hence allocated. Unfortunately, which duties can reasonably be specified also depends upon which allocations of duties among bearers are generally reasonable and, most specifically, fair. The depth of the difficulty here emerges from one aspect of Pogge’s contrast, as I understand it, between the institutional and the interactional perspectives. This difficulty has two elements.
First, one can reason, as it were, from either end of the problem. On the one hand, one can take an institutional perspective on the honoring of rights by asking: what institutions would need to be functioning effectively in order for people generally to have secure access to what they have rights to? This is the strategic reasoning that Nickel and I—but not, I gather, Pogge, who is concentrating on institutions that are already in place—are advocating: taking adequate provision for rights as the goal to be reached, one asks which institutional means are needed in order to get there from here. The specification of adequate institutions is then one basis for an assignment of duties to individuals: the assignment of duties to individuals must be an assignment—more than one different assignment may satisfy this test—that enables an adequate set of institutions—more than one set may do the job—to function. The institutions specified are one test for allocations of duties among individuals.
On the other hand, one can ask: which allocations of right-grounded duties would be fair to individual duty-bearers?27 I intend to invoke this perspective when I say that we should not forget that for the duty-bearers too this is the only life they will live, by which I mean that, however terrible the prospects from the point of view of potential right-bearers if certain rights are not acknowledged and implemented, the point of view of the bearers of the duties implicated by those rights also should be taken into account before one decides whether those rights should, all things considered, be acknowledged. What the latter point of view may reasonably be naturally would have to be spelled out much more fully than I am doing here, but the central thought is that some test of fairness to individual duty-bearers must be satisfied by any allocation of duties, however effectively that allocation would serve the purpose of honoring otherwise reasonable-seeming rights. From this other point of view, then, fair assignments of duties to individuals are one basis for the specification of institutions. The duties permitted by fairness are one test for acceptable institutions.
The reasoning from these two opposite directions is not a contradiction. It is perfectly possible for a necessary condition of the specification of acceptable institutions to be that at least one assignment of the duties the institutions require for their successful operation be fair to individual duty-bearers and a necessary condition of an acceptable assignment of duties to be that they enable the successful operation of at least one set of the institutions specified. In short, one wants institutions that function effectively to honor rights while imposing only duties that make fair demands of those who bear them.
The second element of the difficulty here, however, is that while these reciprocal requirements are not contradictory, there unfortunately is also no guarantee that they can both be satisfied simultaneously in fact. It is like digging a tunnel under a river by having two teams work simultaneously, one from each side, planning to meet in the middle—the nightmare is that they will not arrive at the same place. In practice, it may be that in the situation we actually face the only institutions adequate to secure even the most minimal set of rights for everyone could function only by demanding far more than is fair of many duty-bearers. Given a precise understanding of a fair allocation of duties, this would be an empirical question about the workings of the various institutions we could conceive. Institutional adequacy, as the duties are specified, and individual fairness, as the duties are assigned, are each a test that cannot in principle be surrendered.28 We simply must find out whether both can be satisfied together in practice, being as imaginative as we can.29
AN INSTITUTIONAL TURN FOR CONCEPTIONS OF HUMAN RIGHTS
Several important considerations come together, then, to recommend an institutional turn in thinking about human rights. First, in order to take the implementation of rights seriously at all, one must think strategically. Taking the honoring of the rights as the goal, one needs to look for genuinely effective means, rather than quickly declaring the rights to be utopian or impractical. Second, any effective means to a goal as ambitious as the honoring of fundamental rights requires a division of moral labor. Neither can rights ordinarily be protected one at a time, nor can any one right be honored by small numbers of individuals acting alone. Waves of duties may need to be performed and—to switch images—webs of duty-bearers may need to become involved at various stages. Where the tasks to be performed are predictable, they often should be allocated in advance. Whether the division of moral labor is formal or informal, it must presumably be fair. Third, the ineliminable need for varieties of duty-bearers must not obscure the fact that, when it is possible, the most effective arrangement is to empower the victim. The best arrangement is often one that allows victims of rights violations to become the agents of their own salvation, but this often depends upon institutions that support empowerment.
Complementary considerations about the third point need to be kept in balance. My primary contention has been that taking rights seriously means taking duties seriously. This is of course a thesis about what people other than the right-bearer ought to do on behalf of the right-bearer. This is a vital consideration because in the worst cases of rights violations the right-bearer may have been turned into an at least temporarily helpless victim, who is doomed unless others act in accord with their duties. However, this should in no way obscure, and is not at all in conflict with, the contention that a major part of the means, as well as the goal, of providing for people’s rights is bringing it about that they are in a position to look out for themselves as much as possible. The best institutions for implementing rights will do both.
Theories of rights have been alternately criticized for being statist and for being asocial. Some theories fail in the one direction, and others, in the other, but neither failure is necessary if the complementary considerations just above are held together. An account of rights may be statist, or overly bureaucratic, or even paternalistic, if it allocates more of the work than necessary to people other than the right-bearers. The very first wave of action normally should be responsibility or initiative on the part of the right-bearer himself or herself. Arrangements for fulfilling rights should never be arrangements to do for people what they can equally well do for themselves. Right-grounded duties are, at the earliest, a second wave of action on behalf of people who cannot act effectively for themselves. However, right-violating institutions often leave people precisely unable to act for themselves, either by depriving them of effective channels for action or by actively blocking their action.
On the other hand, a theory of rights may be asocial, or atomistic, in its account if it is simply about individuals demanding things for themselves. This is a picture that omits every wave of action except the first, in contrast with the statist picture that omits the first. Such theories result when rights are conceived as concerned with little more than maximum negative liberty and unlimited property accumulation, each construed as mainly involving keeping other people at bay.30
So, what should the institutions that implement rights be like? Philosophers alone cannot say. Nor can they leave it entirely to others to say.31 Philosophers, insofar as they reflect only about concepts and about the basic information about how the world works that is available to any reasonably well-informed person, can reason, as I did in the first edition, that while everyone everywhere might have enjoyed all their rights if everyone else had simply refrained from violating anyone’s rights, this is in fact not going to happen. In fact, many individuals, firms, and governments are precisely in the business of violating rights for fun and for profit. Anyone who reads a serious newspaper knows this much. So philosophers can be quite certain that beyond a duty not to deprive people of whatever they have rights to, there must be some kind of duty to protect the victims against the violators.32 And given the way the world has been going in recent centuries, it is fairly obvious that we do not have the protection of the victims very well organized yet, which means that at least some of us sometimes have a duty to conceive, nurture, and support better institutions.33
The Thai government does nothing while Burmese kidnappers deliver Burmese peasant girls to Bangkok brothels to serve as prostitutes until they contract AIDS, whereupon the Thais dump the infected women back into Burma.34 The U.S. government announces, and aggressively presses, new restrictions on peacekeeping operations, largely motivated by the completely unnecessary deaths of eighteen courageous young Americans in Mogadishu in 1993, just before and during the slow spread through Rwanda in 1994 of gangs of young boys hacking approximately 500,000 of their compatriots to death with machetes.35 African nations take turns holding famines and civil wars. Wealthy nations spew greenhouse gases into the global atmosphere as if the entire stratosphere were their own national territory exclusively for waste disposal.36 Meanwhile, our pesticides poison even the songbirds (who, I admit, have no rights that it should be otherwise, although that is hardly the whole story). One could rave on and on, but surely it is clear that while the most advantaged carry human technology and culture to spectacular new heights, the most victimized live and die in beastly conditions. We are not managing the fabulous resources of this planet at all well. We ought to build better institutions.
The question is: how do we bring about the transition? Philosophers and rights theorists working alone cannot design better institutions. Some aspects of institutional design are best understood by, for example, lawyers or economists or historians. Yet philosophers ought not simply to flee the scene when things start getting practical. For practical arrangements involve allocations of burdens (and rights and privileges), and such allocations always raise questions of fairness, which philosophers should be inclined to raise and able to help to answer. Philosophers cannot help, however, merely by formulating some general principles of fairness, or general guidelines for fair procedures, and leaving them behind on the table as they hand the institutional design over to the experts, for at least two reasons.
First, good philosophers should be better than most other specialists at sensing where questions about fairness in the assignment of duties arise. Sooner or later those who by then had been treated unfairly perhaps would, if they were not intimidated by an entrenched allocation of power, raise their own questions about the fairness of their treatment. They would probably, however, by then be at a disadvantage, precisely because of the unfairness of their treatment, that would compound the injustice and multiply the difficulties in undoing it. In general, prevention is much better than cure in cases of unfairness. A philosopher might be able to foresee the unfairness before a victim had to feel it. The specific unfairness of a concrete set of arrangements can be foreseen, of course, only by someone who thoroughly understands, and appreciates the workings of, those arrangements. Philosophers and rights theorists who opt out of the details are no help.
Second, philosophical work does not in any case end with the formulation of general principles or general procedures—interpretation and specification are philosophical tasks too. The specification of concrete embodiments of principles does indeed involve extra-philosophical information and skills in addition to philosophical ones, as I have just been emphasizing, but this only makes the philosophical work more challenging, not less essential.37 Imaginative construction of alternatives, careful analysis of what is and is not presupposed and implied—the sensitive attention to what is fair as well as what is efficient or customary, which was the previous point—these and the other philosophical talents are at least as valuable in the making of concrete judgments as in more abstract pursuits. The work is not pure philosophy—it is practical philosophy—but philosophy nonetheless.
Institutional design must combine judgments about what it is fair to expect people to do, what it is efficient to ask people to do, and what it is possible to motivate people to do.38 All this depends upon subtle judgments about sense of duty, sense of fairness, sense of identity, sense of solidarity, self-interest, incentives, and coercion—especially hard choices about which aspects of individuals and societies can be changed while which others remain fixed. For example, one cannot specify and assign duties as if individual people generally embody a moral ideal of a person. Ideal persons would not deprive each other of what they had rights to enjoy, so a simple clarification of the duties not to deprive would be sufficient for such well-motivated people. In that ideal world there would be no point in proceeding to thought about duties to protect against deprivation by the ill-motivated, duties to aid the deprived, and whatever other general categories of duty have point in this nonideal world. In this world we deal, or fail to deal, with “leaders” of ethnic groups who orchestrate the genocide of the rival groups and with “entrepreneurs” who trick trusting peasant girls into brothels where they contract fatal diseases. Such behavior is totally unacceptable and should have been crushed by force long ago; no subtle judgments required. But in the vast middle range between the barbaric and the ideal, and between the way we are now and the best we might actually become, intellectually difficult moral and psychological judgments need to be made about how to arrange to get the necessary transition started, while bearing in mind as the historic transitions are planned that the lives individuals are living now are the only lives they will get.
To be more specific, consider the question that I regularly raise, but never answer, about the limits of right-based duties.39 I have said above that I think one must think strategically about the assignment of duties, with the full enjoyment of (at least) basic rights as the goal. The duties should, then, be assigned in a manner that will make adequate provision for the rights. One could assign the duties necessary to the enjoyment of a right in such a way that, provided everyone did their duty, the right would be secure. That, however, would be the assignment appropriate to a world that was ideal in at least one respect, namely, that everyone did their duty. Indeed, as we already noticed, if everyone could be counted on to honor their duty not to deprive anyone of the substance of their right, we would not need to assign any other duties at all. Since violations of the duty not to deprive people of their rights are deeply characteristic of the world we live in, we must, in order to be in the least practical, move to the assignment of some later waves of duty: default or back-up duties, like duties to protect, duties to aid, and whatever others are needed.
Yet essentially the same problem (of relatively how ideal or real to make our assumptions about how people will generally behave) reasserts itself at each stage. To be minimally practical, we must specify at least duties to protect and some duties to aid (aid those for whom protection in turn fails). How, then, do we assign each of these kinds of default duties among potential bearers: on the assumption that each bearer will perform the duty assigned, or on the assumption that some will fail to perform it, so that a further assignment to a back-up bearer must be made? On the latter choice, we would specify back-up bearers for default duties, for example, who should aid the victims if the people who originally should have aided the victims do not do it, after the people who should have protected the threatened victims against the original right-violators failed to provide adequate protection? This may sound artificially intricate, but it is perfectly ordinary. Who should now pay the medical expenses of the Burmese women already infected with AIDS in the Bangkok brothels openly tolerated by the Thai authorities and patronized by Japanese business executives on organized sex tours? Who should now pay for the orphanage care for the Rwandan children whose parents were already slaughtered in the 1994 genocide that no major military power was willing to bear the expense and danger of stopping?
Rwanda is a classic case of the greater efficiency, not to mention humanity, of performing earlier waves of duty.40 The amount of money spent in July 1994 by the U.S. military dealing with the famine and epidemic in the camps of the refugees in Zaire produced by the civil war in Rwanda in response to the genocide would have paid several times over for a military force that very likely could in April 1994 have prevented the genocide—and many of the horrors that continue to unwind, to this day. The question remains: since the young executioners were allowed by the rest of humanity to hack slowly through Rwanda for weeks until half a million lay dead, who ought to aid innocent survivors now that the duties to protect have been definitively ignored? And what if we/they, whoever we/they are, do not fulfill the duties to aid either? Then what, for the orphaned Rwandan children?
If one simply proceeded mechanically to assign backups to the backups, and defaults for the defaults, one would presumably finally zero in on whichever dedicated people were sufficiently committed to human rights that they would do whatever duties remained to be done until whatever remained possible to honor rights was accomplished, irrespective of who the original failed duty-bearers had been. Such specifications and assignments of duties would be at the other extreme from the specifications and assignments that would be reasonable if everyone carried out their initial assignments—these are the assignments for the world in which many fail to carry out their original assignments. We have swung from the assignment for an ideal world to the assignment from Hell.
Manifestly, such a mechanical devolution of duties has gross faults. First, expecting some individuals endlessly to be willing to step into the breaches left by the failures of others to do their prior duties is wildly unfair. These lives would simply be consumed by (default) duties—this is precisely to ignore that for duty-bearers too, as much as for victims of rights violations, this is the only life they will live. Second, expecting some endlessly to pick up whatever others have dropped would ignore normal human motivation, creating a powerful perverse incentive. People surely could reasonably reject any arrangement that effectively counted on the radical inequality embodied in a moral two-class system consisting of one group of people who were allowed to dodge their right-based duties and another group who had to shoulder whatever the first group dodged. And reasonably or not, people with a normally healthy sense of self-interest would refuse to lead default lives, determined by the moral failures of others. Third, for right-bearers, earlier is of course always better. Every devolution to a later duty leaves violated rights, and violated people, in its wake. Prevention is always better than cure, even where there is something resembling a cure. Aid to previously unprotected victims, or to their now orphaned children, is scarcely a cure.
The solution seems to me to be to draw the line, insofar as possible, at the first level of default: at duties to protect, broadly interpreted to include the design and maintenance of institutions that make it as easy as possible for people to honor their duties not to deprive others of the enjoyment of their rights or, put differently, as hard as possible to commit, and to get away with, deprivations. On the one hand, it would be utopian in the worst sense to attempt to hold the line at an even earlier level—all the way back at duties not to deprive. The world contains, and will contain for the foreseeable future, evil people who will violate the most basic human rights. Large numbers of people will ignore their duties not to deprive others of their rights. We can do everything humanly possible to instill respect for rights and to punish violators, but it would be truly unrealistic to expect that attempts at violation will cease. On the other hand, it does not seem unrealistic to dig in hard at the next level, the level of designing institutions that protect potential victims against, first, those who would violate their rights and, second, those who would neglect their duties to protect them against the violators, rather than falling back entirely beyond protecting potential victims to aiding actual victims. This requires that many decent people not only fulfill their own negative duties not to violate rights but also take on positive duties to block violations against which others had the primary duties to provide protection, that is, also take on default duties to protect. This involves genuine double duty: the negative duties not to deprive that fall upon everyone and some positive duties to provide protection that should have been provided by others.41 Yet it need not involve endless duties.
CONDITIONAL SOVEREIGNTY
These positive duties to protect are duties to conceive and nurture institutions that, taking people as they are, and as they can next be brought to be, make at least the basic rights of all reasonably secure. We have only begun to think seriously about institutional forms for the possibilities here, and my own suggestions are completely tentative. It seems to me, however, that one plausible step would be building a general, global consensus that state sovereignty is conditional upon the protection of at least basic rights and that the international community not only may but ought to step in when the failures of states to protect rights become egregious, as happened in the early 1990s in, for example, both the disintegrating Yugoslavia and the imploding Rwanda. When a state utterly or egregiously fails to protect the rights of the people residing within its jurisdiction, the rest of humanity must have capacities to do more than sit idly by until the slaughter is finished or merely assist the victims after they are violated without resisting the violators.42 It is not permissible to substitute fulfillment of duties to aid that came into play only because victims went unprotected, for fulfillment of back-up duties to protect.
I am still talking about a back-up arrangement for the failure of so-called national governments—few sovereign states consist of precisely one nation, patriotic myths to the contrary notwithstanding—rather than a replacement for the current state system. This is a practical judgment, and an especially shaky one. Others have long ago made the cleaner and more radical suggestion that we work to replace the modern state system with a world government of some sort. Quite apart from the dangers inherent in such a centralization of power, I simply see not the slightest prospect of its happening. Although it is hard to imagine that anyone creating a world from scratch would have made one dominated by any institution remotely resembling the modern state, this is what we nevertheless have.43 If the state cannot be eliminated, as one certainly might wish, the question becomes whether it can be civilized. Is there some way that the modern state can become more of an instrument of good than most states have so far been, or, at least, pressured into protecting the rights of the human beings who live under its sway?
The basic idea is that states should have to behave with minimal decency if they want any respect.44 Sovereignty should be conditional upon performance, and performance should be judged by international norms: conditional sovereignty, judged by minimal international standards, including the provision of protection for basic rights. Rather than global institutions, which may be dangerous and are in any case most unlikely, we would pursue minimal global standards for national institutions. Different states can have different institutions in whatever respects, and to whatever degree, they like, short of failure to protect basic rights. It is unacceptable, however, for whichever bunch of thugs has a monopoly of force within a particular territory simply to allow people to be murdered, raped, or herded around like animals, as has been happening in recent years in some parts of both Europe and Africa, including Bosnia and Rwanda.
Anyone can see that the current practice of leaving people to the mercy of their own government is barbaric. The challenge is to come up with a cure that is not worse than the disease, as critics of intervention regularly note. This is why I emphasized at the beginning of this Afterword that the original human-rights impulse in the U.S. Congress in the 1970s, which predated the Carter campaign in 1976 and of course the Carter presidency, was an anti-interventionary impulse. The congressional leaders, like Donald Fraser of Minnesota, were first of all repelled by U.S. complicity in violations of human rights by regimes we had been supporting, most of all in South Vietnam but also in Chile, Iran, South Korea, and elsewhere. The initial congressional demand, relentlessly resisted by Henry Kissinger in the executive, was simply that the United States stop providing economic and military assistance to gross and systematic violators of human rights (in the name of building alliances against communism). The general principle was merely that where a government was violating the basic rights of the people living under it, no other government, like ours, should strengthen the hand of the violators. The U.S. government was not being asked to take the side of the victims—only to stop taking the side of the victimizers. When President George Bush in 1991 actually took the side of the Kurdish victims of the Iraqi state, by establishing a no-fly zone to protect them, he was taking a more positive action than most people would have dreamed of requesting in the 1970s—and was of course severely compromising Iraqi sovereignty.
One way of putting our question is: is U.S. protection for the rights of the Kurds against the depredations of the state that had power over them a paradigm for the future? Is this the model that should be generalized? Should we simply wish that President Bush had done essentially the same for the Bosnian Muslims against the Bosnian Serbs and Serbia in April 1992, and that President Clinton had done the same for Tutsis and moderate Hutus in Rwanda in April 1994? Not exactly, although I believe that action ought to have been taken in each case to prevent what predictably followed.
I believe that the Bush decision to protect the Kurds was commendable, and that the Clinton decision to write off the Tutsis was despicable, all the more contemptible for having been carefully considered and coldly calculated. Each of these U.S. presidents, however, was probably attempting to do mainly what he took, correctly or incorrectly, to be in the U.S. national interest, on some understanding of it. That is one underlying problem. We do not have adequate institutions if other states intervene or not to protect people against genocidal assaults by their own state depending upon whether it is believed to be in the national interest of the potential intervenor. It appears especially likely that in a world with a single superpower, the superpower will act irresponsibly.
It is, as I have been emphasizing, far from a purely philosophical question which alternative arrangement to embrace. One might, for example, work toward some kind of monitoring system for violations of rights around the world—intergovernmental or nongovernmental—that might alert some agency to act at a sufficiently early stage that action other than military intervention would be effective. One might hope that some nonnational agency—either a United Nations agency or a less than global but still multilateral agency—could be created to conduct any interventions that are called for. Or one might urge that national actors base their decisions on considerations broader than exclusively national ones. Or all the above. In large part this depends upon other, deeper trends, some of which may not be amenable to conscious human design.
The various measures that may be taken can serve either of two different intermediate goals on the way to the ultimate goal of individuals’ enjoyment of their rights. First, some measures need to provide backup in the most literal sense: where the state with the primary duty to protect rights fails—for lack of will or of capacity—to fulfill its duty, some other agent at least sometimes must step in and provide the missing protection. Where the failure is from want of capacity, not want of will, the state may invite the external agent in. Invited or uninvited, the external agent is engaged in intervention, which carries with it myriad difficulties irrespective of the worthiness of its goal. While I certainly believe that military intervention is fully justified by a systematic pattern of gross violations of basic rights of the magnitude of, for example, the genocidal slaughter in Rwanda in 1994, external military intervention is always the last resort because of its inherent tendencies to be self-defeating.
Better, then, would be measures designed, not to take over from the state with primary responsibility the protective duties that it so far failed to perform, but to stimulate the will or provide the capacity for the original state to carry out its own duties. Now, states usually pay remarkably small prices for utter failures to carry out their primary duties. Where those residing on the state’s territory somehow happened to enjoy basic rights to political participation, while other basic rights were violated, they might be able to replace a corrupt or otherwise ineffective government with one competent to protect their other basic rights. But, as I emphasized in chapters 1 and 3, the basic rights tend to go together: it is almost unimaginable that during genocide (failure to protect security rights) or famine (failure to protect subsistence rights), rights to political participation will be protected. People may nevertheless in rare cases be able to engage in political participation—not the same as enjoying the right to political participation, of course—by overthrowing the government by force, but the question is whether the rest of us should allow their enjoyment of their basic right routinely to depend entirely upon their own might.45
Instead of supplying arms to revolutionaries or intervening militarily on their side—the first option discussed above—we should be able to design less violent means of raising the price for governments that fail to protect the rights of the people over whom they claim jurisdiction. Thomas Pogge has made some imaginative suggestions about a different but somewhat parallel problem: how the international community could reduce the benefits to those who overthrow democratic governments, for example, “proposing, as a principle of international law, that a people need not repay loans incurred by a government that ruled them in violation of constitutionally recognized democratic procedures.”46 Just as this suggestion is intended to make unconstitutional seizure of power a little less appealing, we could attempt to make failure to protect basic rights less costless, short of launching military interventions. I do not have a well-conceived institutional proposal to offer. The United Nations is currently considering, for instance, a permanent International Criminal Court that might be able to try politicians implicated in major violations of rights. In any case, we clearly need more middle ways between sitting idly by while horrors are perpetrated, under which I include toothless U.N. Security Council resolutions, and full-scale military interventions. Decisive action short of military intervention, for at least some cases, should not be beyond human powers of imagination and institutional innovation.
What I am calling default duties to protect come, then, in two kinds. Less desirable is literally stepping in to provide protection where those with the primary duty to protect fail to provide it. Often those who will have failed are a government, and stepping in involves military intervention. Hard thought needs to be given to the best agents for such interventions. Far more desirable are institutions that pressure those with the primary duties to protect to perform their duties, especially institutions that impose severe costs upon governments that fail to perform this first duty of government: the protection of basic rights.
This all presupposes logically that the division of moral labor for the protection of basic rights does not respect state boundaries. Does it presuppose it chronologically as well? That is, does a consensus about implementing institutions—for example, arrangements for decisive action short of military intervention—have to wait upon a prior consensus on principle? This is a position held by many and perhaps most forcefully presented by Michael Walzer. It is premature, Walzer has argued, to try to discuss global institutions until there is, if ever there actually will be, a global community whose shared values could be embodied in the institutions: “The only plausible alternative to the political community is humanity itself, the society of nations, the entire globe. But were we to take the globe as our setting, we would have to imagine what does not yet exist: a community that included all men and women everywhere. We would have to invent a set of common meanings for these people, avoiding if we could the stipulation of our own values.”47 In short, “there cannot be a just society until there is a society.”48
The warning against the danger of simply stipulating our own view as the universal one cannot be repeated too often, especially for Americans with our tendency to overmoralize international affairs. However, the stipulation of our own view and the passive wait for a globally shared view to emerge on its own are far from the only alternatives. Indeed, the initial apparent necessity of “there cannot be a just society until there is a society” rests upon a false dichotomy: an artificially constricted picture in which there can be no chicken until there is an egg. In fact, international society and just international society can be built at the same time through the same activities.49 A large part of what makes a collection of people a society is, precisely as Walzer maintains, shared understandings about matters like rights and justice. Rather than waiting for a society somehow to emerge on its own before asking its members to think about what would make it a just society, one can attempt to build a society through agreement in theory or practice on just arrangements. And nothing prevents attempts to seek agreement among those who initially disagree, not that the attempts to reach agreement are guaranteed to succeed.
I think it is clear that this world’s institutions for the protection of rights are grossly inadequate. Conceiving of institutions that could function more effectively while making only reasonable demands upon those who would make them function is challenging and controversial. It is too difficult to be accomplished by isolated theorists, who individually do not know enough and have not experienced enough—this kind of institutional design is a social task. For the foreseeable future, much of the protection of individual rights will, like it or not, depend upon national governments, but national governments seem to me to be among our most inadequate institutions. Institutions that cannot perform deserve little respect, and they certainly are not entitled as a matter of principle to sovereignty.
What can we do to make the protectors protect? And how long do we stand by after it is clear that they never will? In the most extreme situations, like famine in Somalia in 1992 and genocide in Rwanda in 1994, outsiders must step in and do with military force what the national state has failed to do: protect the utterly vulnerable against relentless forces of death. Military force is blunt and negative—it can protect some (by killing others, including innocent bystanders), but it cannot plant or build. Everyone, and most of all professional military people, can see that military force is the last resort. We need, as I noted above, decisive action short of military intervention. For that we need better international institutions. We will not know what our duties are, nor will we know what rights people can expect to enjoy, until we have constructed these institutions. Meanwhile, our common humanity requires at the very least, I would suggest, that we participate in the conversations about the institutions that might protect basic rights.