Human Rights, Legitimacy, and International Law

Human Rights, Legitimacy, and International Law

Abstract

The trajectory of my lecture will follow the thematic sequence foreshadowed in its title. I shall begin with some reflections on the nature, and basis, of human rights considered as moral standards to which a variety of legal and political instruments purport to give expression and force. I will then examine recent attempts by some prominent philosophers to explicate the nature of human rights by reference to the supposedly constitutive connection they bear to political legitimacy. In the second half of the lecture, I turn to the question of the legitimacy of international law itself, especially insofar as that law purports to embody and implement background moral norms of human rights. I offer an account of the legitimacy of international law based on the service conception of legitimate authority and conclude by canvassing the limitations of its legitimacy arising from pluralism, freedom (sovereignty), and exceptionalism.

A general thesis that underlies my lecture is that conceiving of human rights as, fundamentally, natural rights preserves their integrity as a moral notion, gives us a better handle on the notoriously contested issue of their justification, and offers a clearer understanding of their important, but also importantly qualified, role in assessments of political legitimacy, including the legitimacy of international law. The obverse side of that last point is that we should resist the attempts by philosophers in recent years to overburden the idea of human rights, finding in it the key to the resolution of all manner of complex and pressing legal and political questions, such as the internal legitimacy of states and the shape and extent of the sovereignty to which they are entitled in the international domain.

I. Human Rights as Natural Rights

We are all familiar with the powerful human rights movement that sprang up in the latter half of the twentieth century, a movement for which the Universal Declaration of Human Rights is a key manifesto, its provisions having been elaborated, extended and implemented in subsequent declarations, conventions, constitutions and laws. One question worth asking is whether there is a cogent, unitary notion—a notion of “human rights”—that constitutes the primary ethical idea driving this movement, giving it both conceptual coherence and normative force. This is not so much a sociological as an ethical question: since we are looking for an idea that not only has an actual foothold in the development of the human rights movement, but also makes sense of it as a worthwhile enterprise. My contention is that there is such an idea, but that its historical origins significantly pre-date the Universal Declaration of 1948, a fact that would not have surprised many of the Declaration’s chief authors. To this extent, it is misleading to interpret the contemporary human rights movement as embodying a distinctively modernist ethical outlook, one that took root, according to one of the most radical proponents of this now-popular interpretation, only in the late 1970s, during the Carter administration.²

Considered as moral standards, my claim is that human rights are best understood as continuous with what were once known as “natural rights.”³ This continuity is not merely historical, but conceptual: it is not simply that the tradition of natural rights thought is part of the historical lead-up to contemporary human rights discourse, but that the ethical idea at the core of the latter is essentially that of a natural right. For present purposes, it is enough to highlight two key dimensions of this idea. First, human rights are moral rights, possessed by all human beings, simply in virtue of their humanity. In other words, human rights, like natural rights, are universal moral rights. Call this the universality thesis. Second, human rights are to be identified by the use of natural reason, principally ordinary, truth-oriented moral reasoning, as opposed to the artificial reason of some institution, such as law, the conventionally accepted reasons upheld by some culture or tradition, or the deliverances of divine revelation. Moreover, it is important to stress that ordinary moral reasoning, in virtue of being “natural,” need not be saddled with the futile ambition of wringing moral conclusions exclusively from value-free propositions about the natural world. Call this the natural reason thesis.

Some philosophers have challenged the universality thesis, according to which the Universal Declaration is best understood as aspiring to give effect to universal moral rights, on the grounds that there is no plausible way of interpreting all of the items in that instrument as applying to all human beings throughout human history. In what meaningful sense, these critics ask, did Stone Age cavemen have a right to a fair trial, or political participation, or a nationality? Surely, the argument goes, human rights have counterpart obligations, and historical variations in resources and technological and institutional capabilities prevent us from attributing such rights to all people throughout human history.⁴ Or, as Bernard Williams put it, in a related connection: “Of course, one can imagine oneself as Kant at the court of King Arthur, disapproving of its injustices, but exactly what grip does this get on one’s ethical or political thought?”⁵ Now, it seems to me that this line of thought is misguided, largely because it mistakenly assumes that the idea that human rights are possessed simply in virtue of one’s humanity necessarily entails that the self-same schedule of human rights must be meaningfully attributable to all human beings at all times and places. Instead, in talking about human rights, we need not be purporting to operate ahistorically; we can instead refer to the human rights possessed by all human beings within a specified socio-historical context, such as that of modernity. These rights will be possessed simply in virtue of our humanity because their existence does not depend on any particular status or achievement on our part nor on any actual institutional or social recognition. Another way of putting this, as the distinguished historian of natural rights, Brian Tierney, has stressed, is that human or natural rights are not restricted to those that may be possessed and meaningfully exercised even in a pre-political state of nature. This idea is just one strand, but not even the dominant strand, in the natural rights tradition.⁶ In construing human rights as natural rights, the burden of the adjective “natural” is not to signal their possession even in a state of nature, but instead their susceptibility to identification by natural reason, what I called the natural reason thesis, which is the second dimension along which I claimed that human rights are importantly continuous with natural rights.

Now, a radical development in recent political philosophy has been the calling into question of the natural reason thesis by John Rawls and members of his school. According to Rawls, irrespective of whether human rights have a grounding in natural reason, the key question is whether they can be given a grounding in a form of “public” reason whose norms are autonomous with respect to those of the former. Natural reason is standardly truth-oriented and comprehensive in scope, not confined to some specifically political subject-matter. More to the point, reasonable people are prone to disagree about natural reason’s verdict on questions of the good life, morality, human nature and our place in the cosmic order. Therefore, especially when it comes to rights applicable across heterogeneous cultures, it is imperative to justify them in a way that transcends these interminable controversies. Here is a central passage from Rawls’s book, The Law of Peoples, a book whose rather gnomic remarks on human rights constitute the most extended treatment he ever gave to that topic:

[Human rights] do not depend on any particular comprehensive religious doctrine or philosophical doctrine about human nature. The Law of Peoples does not say, for example, that human beings are moral persons and have equal worth in the eyes of God; or that they have certain moral and intellectual powers that entitle them to those rights. To argue in these ways would involve religious or philosophical doctrines that many decent hierarchical peoples might reject as liberal or democratic, or as in some way distinctive of Western political tradition and prejudicial to other cultures. Still, the Law of Peoples does not deny these doctrines.⁷

Admittedly, there is an undoubted allure to the proposal that, in justifying human rights, one should remain steadfastly non-committal regarding the interminably disputed questions about the human good and human nature. Doing so apparently sub-serves the idea that justifications of human rights should be ecumenical, not parochial: graspable and endorseable by reasonable adherents of non-Western cultures. But, as I have argued elsewhere, the constructivist approach to the justification of human rights devised by Rawls collapses back into the very parochialism it seeks to avoid. This is because the content of public reason, for Rawls, simply presupposes ideas taken to be implicit in the liberal democratic tradition. In particular, the idea of free and equal citizens willing to co-operate on the basis of fair terms provided that others are willing to do the same. It is the embrace of these ideas that marks someone out as reasonable, whatever their comprehensive beliefs.⁸

Now, of course, when it comes to human rights, Rawls lowers the bar dramatically: what is necessary is not that such rights be justifiable to people who are “reasonable,” by liberal-democratic lights, but to those who are only “not fully unreasonable.”⁹ But that backhanded compliment—“not fully unreasonable”—testifies to the impotence of the “public reason” approach in the face of the charge of parochialism, since what counts as “fully reasonable” has been defined, in effect by stipulation, by reference to a certain interpretation of the liberal democratic tradition whose authority is simply taken as given. We could, in the face of this failure, seek more elaborate and sophisticated ways of giving the notion of public reason content that is suitably non-parochial, for example, by appealing to a form of global public reason that draws on standards shared across a range of traditions. But the prospects of such a project look dim, because the core problem with the idea of public reason is the attempt to prescind from the idea that judgments and principles of political morality are to be vindicated at the bar of ordinary truth or natural reason, replacing this with a focus on standards of assessment that are actually shared. We do better to hold to the traditional way of responding to the charge of parochialism, at least at the most fundamental level. This involves defending human rights standards, insofar as we are prepared to defend them, as objectively true propositions of morality. How we should properly act on these truths in relation to societies that contravene them is itself a further question, one which itself turns on further truths about topics such as the demands of respect for the self-determination of political communities or the likely geo-political impact of any proposed plan of intervention.

But even those writers on human rights who are not seduced by the questionable blandishments of the Rawlsian “politics-without-truth” methodology into abandoning the natural reason thesis still often labor under a serious disadvantage. This is the fixation of contemporary moral philosophy on a division between deontological and consequentialist doctrines, with the accompanying idea that any grounding of human rights in the moral truth must consist in their incorporation within either a deontological or consequentialist general theory of morality. The paradigmatic consequentialist theory, of course, is utilitarianism. As a consequentialist view, it holds that the rightness of an action is exclusively a matter of the states of affairs that are its consequences. It assesses these states of affairs in terms of the aggregate amount of well-being or interest-fulfillment they contain, and it instructs us to perform those acts which will maximize well-being overall. As critics of utilitarianism have repeatedly observed, however, this theory confronts tremendous obstacles in making sense of the idea of individual rights; hence, the uneasy place occupied by natural or moral rights in the history of utilitarian thought: from outright rejection, in the case of Jeremy Bentham, to a cautious embrace, after being suitably re-interpreted, in the case of John Stuart Mill.

The insuperable difficulty utilitarianism has with individual rights is that they are elements of our moral scheme that break its consequentialist and maximizing template for moral rightness. It is a feature of human rights that they are relatively immune to trade-offs, including trade-offs within the category of human rights itself. For example, one cannot justify violating one innocent person’s right not to be killed simply in order to prevent the violation by someone else of three other innocent people’s right not to be killed. Deontologist critics of consequentialism, such as Thomas Nagel, have diagnosed the problem as stemming from the assumption that the existence of rights depends on the way they serve the interests of their holders.¹⁰ This is because, these critics suppose, utilitarians are correct in assuming that the moral significance of interests is ultimately reducible to the logic of maximizing their aggregate fulfillment. It follows that the special moral significance of rights can only be captured if they are grounded independently of the interests of the people whose rights they are. This grounding, these deontologists insist, will consist in an appeal to their status as moral agents or members of the moral community, a status we can make sense of, and which can generate rights, independently of any appeal to the quality of life of those who possess this status.

Yet deontological attempts to ground human rights in some notion of status, mobilized without any appeal to the elements of a good human life, tend to suffer from a variety of debilitating defects. Insofar as the notion of human status is not left mysterious, it is doubtful that if offers us enough in the way of justificatory materials to ground anything like the full complement of human rights in the Universal Declaration. The alternative, grasped by Nagel, is to characterize the notion of status in terms of a set of human rights. But this fundamentalist move, which makes human rights basic and underived elements of morality, is hard to distinguish from a question-begging appeal to certain parochial value judgments. And, in any case, we are still left with the puzzling fact—surely not readily dismissed as a coincidence—that paradigmatic human rights systematically protect important human interests.

Fortunately, die-hard fundamentalism about human rights is not the only way to preserve them from distortion at the hands of the consequentialist. The beginning of wisdom here is the realization that the consequentialist-deontological dialectic involves a shared, but eminently questionable premise: the assumption that the significance of interests is exhausted by the role they play in an aggregative calculus. To believe that, you have to believe that the injunction to maximize the fulfillment of interests is at least part of the truth about morality. But, as John Finnis, Philippa Foot, David Wiggins, and others have powerfully argued, there is no compelling reason to accede to this view.¹¹ With that obstacle dislodged, it becomes evident that we do not have to choose between status and interests in grounding human rights. On the contrary, I believe that the two notions operate in intimate union in justifying standard human rights. So, on the one hand, there is no reasonable prospect of grounding human rights independently of how they serve the objective interests of their holders: interests such as those in knowledge, friendship, autonomy, enjoyment, accomplishment, play, and so on. But, on the other hand, human beings are not simply the impersonally plottable “locations” at which free-floating interests get realized or frustrated. Instead, each human being enjoys a valuable status in virtue of their membership of a species characterized by a series of distinctive capacities, including capacities for thought, deliberation and action. Since all human beings are, in this sense, equally human, this valuable status is possessed by each in equal measure. The recognition of this status, which is the most promising way of understanding the notion of human dignity, is incompatible with the impersonal maximization of interests that utilitarians place at the centre of our ethical thought.¹²

Of course, I have only gestured towards the kind of pluralist justification of human rights I find most persuasive. It is pluralist at two levels. First, it appeals to both human status and the elements of the human good in generating human rights. Second, it treats the elements of human good, insofar as they have a bearing on the existence of human rights, as themselves irreducibly plural. And I have not here broached the hard question of how one gets from dignity and interests, in any given case, to the warranted conclusion that they underwrite a human right.¹³ But the key point for present purposes is this: keeping in mind the continuity of human rights with the natural rights tradition helps liberate us from the shackles imposed by the consequentialist-deontological matrix of contemporary moral philosophy. For if historians like Tierney and others are to be believed, the idea of natural rights was born as early as the twelfth century, many centuries before consequentialist and deontological modes of thought imposed their respective straight-jackets on moral philosophical inquiry. To those early natural rights thinkers, primarily canon lawyers who sought to harmonize Roman law and canon law, the idea that natural rights had no grounding in the human good would have been just as outlandish as the idea that moral rightness consists in the impersonal maximization of the good across persons. The same holds for the thought that natural or human rights constitute claims of the individual that are essentially characterized by an antagonistic relationship to the common good, the common-place idea of “rights as trumps” that makes most sense against a potentially incoherent, utilitarian conception of the common good. As John Finnis pointed out almost thirty years ago, this thought “trades on an unwarranted assumption that utilitarianism is a moral-political theory sufficiently coherent to yield results which need to be, and can be, trumped by considerations of individual rights.”¹⁴

The more general moral to be drawn is that reflection on human rights highlights the shortcomings of much contemporary moral-philosophical debate, in particular, its impoverished understanding of the way in which a concern with the human good can bear on the existence of recognizably moral demands. A salutary consequence of the characterization of human rights as natural rights is the preservation of a conceptual link with a mode of thinking about morality and the human good that is free of this modernist distortion.¹⁵

II. Legitimacy

I previously mentioned one source of resistance to the idea that human rights are essentially natural rights. This comes from those who want to elaborate a notion of human rights that is a good fit for the contemporary human rights movement, including the provisions of the Universal Declaration of Human Rights. Their objection is that the concept of a human right, interpreted as a natural right, is massively under-inclusive with respect to a list of rights of the sort found in the Universal Declaration. This is because natural rights must be attributable to all human beings throughout human history, whereas this is not plausible of many items in the Universal Declaration. The answer to this objection, I suggested, is to abandon the idea that human rights are necessarily timeless: we can meaningfully talk about rights possessed by all human beings, simply in virtue of their humanity, within a restricted temporal framework, like modernity.

But now a critic, motivated by the same concern with fidelity to contemporary human rights practice, might advance the opposite objection: that the resultant concept of human rights as universal moral rights is over-inclusive. This is because in contemporary conditions universal moral rights likely include rights—such as the right not to be pinched, insulted or personally betrayed—which not only do not historically figure in standard human rights documents, but which would be out of place there. Thoughts such as these have prompted a number of philosophers to reject the natural rights interpretation of the concept of human rights in favor of an explicitly political interpretation. Human rights, on this view, are not simply universal moral rights, not even indexed to some specific historical context, but rights that perform a distinctive political function or complex of functions. This political role marks them out as distinctive within the general class of moral rights. And there is the added bonus on this analysis that, as a sub-set of universal moral rights, human rights represent more minimal standards, and so are more readily insulated from the charge of a wholesale projection globally of Western moral and political standards.

Advocates of the political view of human rights differ as to the political function, or set of functions, that distinguishes human rights. For some, it is that the primary obligations associated with human rights exclusively bind states or coercive political institutions. For others, it is that they set a benchmark that any state must satisfy if its laws are to be legitimate, i.e., if they are to impose moral obligations of obedience on their putative subjects simply as laws. For yet others, human rights are triggers for international intervention or concern in the case of sufficiently grave and extensive violations. And, of course, some proponents of the political view combine a number of functions in their conceptualization of human rights—this is famously the case with Rawls, the most influential advocate of a political view of human rights, who regards them as both benchmarks of legitimacy and triggers of coercive intervention. In what follows, I will concentrate only on the thesis that human rights are rights that have a special role in judgments of political legitimacy, so that an adequate grasp of the concept of a human right must make reference to this role.

But before discussing this thesis, it is worth registering a general objection to all essentially political interpretations of human rights.¹⁶ The objection is that although human rights have significant political implications, some of which may even follow as a matter of conceptual necessity given the nature of human rights and the nature of political institutions such as the state or the system of states, it puts the cart before the horse to conceptualize human rights in terms of any of these political roles. One indication of this is that it makes perfect sense to deploy the language of human rights even if, as an anarchist, one denies the moral acceptability of the state, or, as an advocate of cosmopolitan, one-world government, one rejects the ultimate moral desirability of a system of states. Of course, the concept of human rights may implicate the idea of a state, or a system of states, without those who deploy that concept being committed to endorsing either of those ideas. Nonetheless, it does seem decidedly odd, if not strictly incoherent, to think that the cosmopolitan theorist, who argues for the abolition of the state system precisely on human rights grounds, is deploying a concept that itself can only be understood in terms of the very system he is using that concept to decry.

That, of course, is a very general objection to any attempt to render the concept of a human right parasitic on the concept of a particular kind of institutional structure or geopolitical configuration. But let me proceed now to one broad manifestation of the political view of human rights, that according to which human rights are essentially benchmarks of political legitimacy. By the legitimacy of a political institution, I mean the right of a political institution, such as the state, to rule over its purported subjects. And by ruling, I mean the issuing of directives that purport to be morally binding—that purport to impose obligations of obedience—on those subjects. Moreover, these are obligations that are claimed to exist independently of the content of any particular directive, simply in virtue of the institution’s say-so. It follows from this that a law may be morally binding, because enacted by a body that enjoys legitimacy, while being in some sense unjust in terms of its content or the process whereby it was enacted.

Now, some prominent philosophers have recently advocated different versions of the idea that a constitutive connection exists between human rights and political legitimacy, but they differ both as to the content of the benchmark of legitimacy that human rights establish and as to the benchmark’s modality, for example, whether it is or approximates to a sufficient condition of legitimacy or whether it is only a necessary condition.

For Rawls, in addition to serving as defeasible triggers for coercive intervention by external agents, human rights are standards that must be satisfied if a society’s law is to be legitimate.¹⁷ Political legitimacy is enjoyed only by those social orders that are systems of social co-operation, systems that preclude “command by force, a slave system” and that take the good of all members of society into account in political decision-making.¹⁸ Conformity with human rights is a necessary (but not sufficient) condition for the obtaining of such a system. In a strikingly similar vein, Bernard Williams argues that the violation of human rights, or at least the most basic among them, approximates to a relationship of unmediated coercive power between ruler and ruled, one of might rather than right.¹⁹ For Ronald Dworkin, by contrast, the touchstone of legitimacy fashioned by human rights centers not on the idea of coercion, but on that of contempt. A government is legitimate principally to the extent that it operates on a “good faith” understanding of what the dignity of its subjects requires, and the criterion of good faith is acting on principles that constitute an intelligible, even if ultimately flawed, conception of dignity.²⁰ Human rights are the sub-set of political rights that limn the outer boundaries of an intelligible conception of dignity: a government that violates them exhibits contempt for the dignity of its subjects, as opposed to merely having a mistaken view about what dignity requires. This undermines a government’s good faith, hence also its legitimacy.²¹

Although I cannot undertake a detailed examination of these three views,²² let me indicate broadly why I believe any variant of the general thesis that human rights bear a constitutive relationship to political legitimacy faces a powerful dilemma in meeting the desideratum of fidelity to the post-Universal Declaration culture of human rights. If compliance with human rights it taken to approach anything like a sufficient condition of legitimacy, a problem will arise in capturing a central feature of human rights thought. This is the idea that the self-same set of human rights, including broadly identical normative content in the duties associated with those rights, is attributable to all human beings throughout the globe. The uniformity of content of human rights, so understood, is in tension with the relational character of assessments of legitimacy. By the latter, I refer to the fact that whether or not any particular state is legitimate depends upon the obtaining of a certain kind of relationship between the state in question and its putative subjects. This relational character means that the selfsame law, or laws with identical content, can be legitimate with respect to some putative subjects and not others. We see an illustration of this tension in Dworkin’s theory, in which compliance with human rights operates as something close to a sufficient condition of legitimacy. Deploying his “good faith test” of legitimacy, Dworkin points out that “a health or education policy that would show good faith effort in a poor country would show contempt in a rich one.”²³ This means that the normative content of the human right to health—the substantive level of health care that it makes obligatory—can vary enormously from one country to another. In a rich country, good faith about human dignity may require the provision of IVF treatment or even some forms of cosmetic surgery, whereas inhabitants of poorer countries would only be entitled to the most rudimentary levels of health care, or perhaps would have no positive entitlement to health care services at all. On this sort of view, the idea that the human right to health confers the same entitlement to all human beings has been effectively eviscerated.

One response to this first problem is to accept that human rights are standards whose substantive content is uniform across societies, but to insist that they constitute only necessary conditions of legitimacy. This is Rawls’s approach in The Law of Peoples, but in his hands it has the upshot that only an ultra-minimal set of human rights is vindicated. Rawls’s schedule of human rights makes no room for human rights to freedom of speech, opinion and movement; to non-discrimination on the grounds of sex, race and religion; to education, work and an adequate standard of living. Only Articles 3 to 18 of the Universal Declarations give expression to “human rights proper,” as opposed to “liberal aspirations.”²⁴ There is a similar upshot on Williams’s account. The charge that a practice violates human rights is interpreted by him as “the most serious of political accusations,” one applicable to torture, surveillance, arbitrary arrest and murder.²⁵ Indeed, it may be that Williams takes the connection with legitimacy to pick out only the most important, or “basic,” human rights, rather than all human rights. This is the second horn of the dilemma confronting the benchmark of legitimacy approach: insofar as human rights are taken to be only necessary conditions of legitimacy, it is unlikely that fidelity to the non-minimalist character of the contemporary human rights culture can be secured.

No doubt there will eventually come a point at which an institution’s human rights deficiencies are so egregious as to rob it of legitimacy, whether in general or in some specific domain. This may well be true of violations of the most basic human rights, such as the rights not to be tortured, enslaved or arbitrarily killed. But these rights constitute only a small sub-set of all plausible candidates for human rights status (or, perhaps more accurately, it is only a sub-set of the violations of this sub-set of rights that has legitimacy-depriving effect in all cases). So if, as I suggested at the outset, the rejection of the natural rights conception is motivated by the desideratum of fidelity to the Universal Declaration-centered practice of human rights, including its notable non-minimalism, then moving in a Rawlsian direction is self-defeating.²⁶

The conclusion I draw is that although human rights undoubtedly have significant implications for assessing the legitimacy of political institutions, alongside a series of other moral standards, their very nature is not to be explicated in terms of their role in such assessments. What it is to be a human right is not, in part, to be explicated by any informative relation to the standard of political legitimacy, no more than what it is to be a nuclear weapon is to be understood in terms of the deterrence role they play in geopolitics.²⁷ This still leaves hanging the original problem of the supposed over-inclusiveness of the natural rights model of human rights, assuming we accept the desideratum of constructing a theory that makes sense of contemporary human rights practice. But that problem can be addressed by less drastic measures than jettisoning orthodoxy in favor of a political concept of human rights. These alternative measures include the following two: (1) distinguishing between core and derivative specifications of human rights, so that some of the universal moral rights that do not figure in standard human rights documents are interpreted as deriving from more abstract rights that do so figure (e.g., the right not to be pinched can be derived from a right to physical security); and (2) stressing that human rights documents often differ in the political functions they perform, and it is the political functions of the documents, rather than of human rights themselves, that dictate the inclusion of some universal moral rights and the exclusion of others (hence the absence from standard human rights documents of a right against certain forms of personal betrayal, since its enforcement exceeds the proper bounds of state authority, whereas most of those documents are concerned with human rights responsibilities that states may appropriately discharge).²⁸

III. The Legitimacy of International Law

I turn now to some reflections on the legitimacy of international law, with a particular focus on the legitimacy of that part of international law that characteristically seeks to give expression to the underlying morality of human rights.²⁹ The question of the legitimacy of international law is one that I framed in terms of the right to rule and, hence, of the law’s bindingness; more specifically, in terms of whether there is a content-independent moral obligation to obey the law, an obligation to obey the law just because it is the law.³⁰ Some philosophers, such as Ronald Dworkin and Philip Pettit, would go further, insisting that the question of legitimacy centrally concerns not merely the justification of content-independent obligations to obey, but (also) of the coercive enforcement of the law.³¹ But the more abstract formulation of the problem of legitimacy I have given is, I think, preferable as a specification of the core matter at issue. The obligation to obey is the fundamental claim made by an institution that asserts the right to rule. The question of the legitimacy of particular enforcement mechanisms can then be addressed as a matter of the content of particular bodies of law that claim legitimacy, rather than as integral to the very notion of legitimate authority itself. Proceeding in this way is especially illuminating in relation to international law, a legal system which seldom claims to deploy coercion against its subjects, and so would turn out not to make a systematic claim of legitimacy on the view under consideration, and perhaps in consequence not to enjoy the status of fully-fledged law. So, in addressing international law’s legitimacy, we do not have to follow Dworkin’s drastic step, in recent work, of engaging in counterfactual speculation regarding the bindingness of an international court endowed with compulsory and effective jurisdiction. It is not at all clear what bearing such speculations have on the claims to obedience made by international legal institutions here and now.

Still, there are at least two respects in which international law cannot be neatly subsumed under the traditional understanding of legitimate authority as the right to rule. First, there is no unified legislative body asserting a right to rule by means of international law comparable to those prevalent within domestic legal systems. Second, the two main traditional sources of international law are treaties and custom, and the former, being essentially promise-based in nature, is not as such a matter of obligation grounded in the right to rule. However, multilateral treaties, in particular, often play an important role in the formation of customary international law, so that obligations whose content is originally specified in treaties come to bind non-states parties as well. In light of these points, my primary focus will be on the legitimacy of customary international law, especially as it purports to embody and implement human rights morality.

Law’s claim of legitimacy is the claim to impose a content-independent obligation to obey. Philosophical controversy persists as to the conditions, if any, under which that claim is justified—as to the standard of legitimacy, as I will call it. Consent and, much more recently, democracy, have been advanced as standards for determining the legitimacy of international law. Here, I follow the view of Joseph Raz, himself following in the footsteps of thinkers such as Plato, Aristotle, and Aquinas, according to which the satisfaction of what Raz calls the Normal Justification Condition (NJC) is typically a sufficient condition for legitimate authority, both domestically and internationally. The NJC states the following:

A has legitimate authority over B if the latter would better conform with reasons that apply to him if he intends to be guided by A’s directives than if he does not.³²

So, an authority is legitimate when its putative subjects would likely better conform with the reasons that apply to them by treating the authority’s directives as binding than if they did not. This is aptly dubbed a “service conception” of legitimate authority, but the adjective should not mislead us into supposing that what confers legitimacy on an authority is its role in enabling its subjects to fulfil their subjectively-given preferences or goals. Instead, the reasons in question are ultimately objective: they ultimately concern what the subjects’ goals should in fact be, not what they are.

Raz’s service conception of authority is by now very familiar to students of legal philosophy, but for my purposes it is worth highlighting two of its features. The first is the potential diversity of the reasons relevant to making judgments of legitimacy under the NJC. These include not only reasons of self-interest but, crucially also, moral reasons. One implication of this is that a state with an excellent record of compliance with human rights might, nonetheless, be bound by international human rights law, provided its being so bound sufficiently enhances the likelihood of other states complying with human rights morality. Indeed, it may be so bound even if the upshot of subjection to the international regime is that it has to accept inferior human rights decisions within its own borders as compared with those that would have been reached had it relied exclusively on domestic legal institutions.

The second observation about the NJC is that its application is not all-or-nothing, but potentially fragmentary, and this along two dimensions. First, there is the possibility of domain fragmentation: a legal system’s claims of legitimacy might be vindicated in some areas—the international law relating to use of force, for example—but not others—such as international economic law. Alongside this domain fragmentation, the service conception also makes room for the possibility of subject fragmentation, according to which the relevant area of law or legal institution binds only some, but not all, of the subjects it claims to bind.³³ This follows from the fact that the NJC is essentially relational in character. It asks, for each putative subject of authority at some given time, whether or not they would better conform to the reasons that apply to them by taking the directive as binding. The answer to this question may vary from one subject to another, given cognitive, volitional and other differences among them. I shall come back to subject fragmentation when I discuss the topic exceptionalism.

The possibility of domain and subject fragmentation highlights some ways in which the NJC is a modest standard: it does not set itself to show that, when laws or legal institutions are legitimate, they bind all those they claim to bind regarding all the subject-matters they address. But there is a deeper way in which the NJC is modest, and dwelling on this may help forestall some frequent misconceptions. The NJC is explicitly formulated as the “normal” justification for legitimacy; it sets out, at the most abstract and basic yet informative level, the criterion whose satisfaction typically underwrites a right to rule. But it is not an exhaustive account of legitimate authority. Compatibly with accepting the NJC, one may hold that there are other, supplementary ways of grounding the claim to legitimate authority, some of which may presuppose the satisfaction of the NJC, while others may be stand-alone justifications.

Moreover, there are cases in which the NJC is satisfied without establishing a right to rule, so that it is not even a sufficient condition for legitimacy. This emerges most clearly if we focus on the fact that the right to rule entails a content-independent moral obligation of obedience. Perhaps following my work colleague’s instructions regarding some menial task we habitually perform together would significantly reduce the likelihood of my suffering a short sharp pain in the course of carrying out the task. Let us suppose that my reasons to avoid the pain outweigh the consideration in favor of deciding for myself how to perform the menial task. In such a case, I would better conform to reason by taking my colleague’s instructions as content-independent and exclusionary reasons for action. But even if so, it is a little far-fetched to conclude that I have a moral obligation to obey his instructions, such that failure to do so is a ground for guilt (on my part) or blame (on the part of others). This is because the relevant background reasons in this case, which focus entirely on my personal comfort, are not such as to plausibly generate a moral obligation. So, the application of the NJC cannot be anything like mechanical. It is a general guideline that requires supplementation by case-sensitive judgment as to whether compliance with the directive satisfies the NJC in such a way as to issue in a content-independent and exclusionary reason to obey that is a moral obligation. To this extent, the NJC is more like a template for assessments of legitimacy than an algorithm.³⁴

Can we establish the authority of existing international law, or some component of it, under the NJC? In important recent work, Allen Buchanan has defended the legitimacy of international human rights law primarily on the basis of the cognitive advantages it secures.³⁵ Buchanan contends that, if properly designed, international human rights law institutions can give us vital epistemic assistance regarding both the identification of human rights morality and the most effective ways of enhancing compliance with it. Among the ways they can do so are the following: (a) assessing and utilizing reliable factual information crucial to the justification or specification of human rights norms; (b) achieving a more inclusive representation of interests and viewpoints than is available at the domestic level, thereby mitigating the risk that our judgments about the content of human rights norms and the best way of implementing them are skewed by cultural biases; and (c) providing authoritative specifications of human rights when there is a range of reasonable alternative specifications.³⁶

Buchanan’s novel argument is restricted to the case of “properly designed” institutions, a designation that international human rights institutions very doubtfully satisfy. But, even if they do satisfy it, the crucial consideration here is that the sorts of epistemic advantages described by Buchanan, important thought they are, could not by themselves be a sufficient basis for legitimacy. This is because a person or institution’s epistemic virtues can typically only establish them as an epistemic authority, one whose judgment on a given topic provides others with reasons to believe what they say, even if this is a belief about what they should do. They will not show that they are a practical authority, someone whose say-so generates content-independent moral obligations to obey them.³⁷ Classically, among the further conditions relevant under the Normal Justification Conditions, are considerations of efficacy, i.e., the power to secure general compliance with a putative authority’s directives. So, for example, one fundamental source of the legitimacy of most states is the power to resolve problems of collective action by laying down standards that its subjects have reason to comply with because, among other things, those standards are likely to be obeyed by others in the community. Hence the actual traffic code in the UK has practical authority over inhabitants of that country, whereas the radically different, yet “ideal” code produced by an academic expert on traffic regulation does not.

In one way this is not such a problem for Buchanan’s argument, since not all of the institutional capacities he invokes are purely epistemic in nature. I especially have in mind the vitally important function of “authoritatively specifying the interpretation of a particular human right when there is a plurality of reasonable interpretations.”³⁸If these alternatives are genuinely equally eligible so far as reason is concerned, then what is at work here is an executive function and not just an epistemic one. It is the capacity for an institutional act of volition or choice that picks out one interpretation from an array of “reasonable alternatives.” In other words, when reason leaves a number of interpretations open, reason itself can demand an act of individual or collective will to identify the one that is to serve as a basis for co-ordination. Notice, however, that international law can only reliably generate authoritative co-ordination points if there is a tendency for its putative subjects to conform to it. Just as an “ideal” traffic code is not binding on me if there is no tendency for other drivers to adhere to it, the same is true of other laws intended to perform a co-ordinative function. So, the case Buchanan advances regarding function (c) must be crucially supplemented with an appeal to the “value added” of capacities of international human rights law that are not exclusively epistemic. Primary among these capacities is that of sheer efficacy or de facto authority. The same argument applies a fortiori to the more purely epistemic functions (a) and (b).

So, contrary to the implicit suggestion in Buchanan’s argument, we cannot make significant progress in assessing the legitimacy of international law whilst completely bypassing the vexed question of its efficacy. And a vexed question it remains. International law, unlike standard domestic legal systems, generally lacks the capacity to deploy effective sanctions against non-compliance. And international human rights law, in particular, faces special efficacy-undermining burdens. Compared to many other forms of international law, it derives limited benefits from the logic of reciprocity. The failure of a state to respect human rights law in its treatment of its people does not, of itself, harm other states, nor can the latter meaningfully retaliate by failing to respect their own citizens’ rights. Fortunately, as a philosopher, the efficacy of international law is a question I am neither equipped nor expected to tackle. There is, however, comfort to be derived from recent work by scholars such as Mary Ellen O’Connell and Beth Simmons,³⁹ which seriously calls into question the blanket skepticism about the causal efficacy of international law, especially its human rights components, that has been advanced by some “rational choice” theorists. In arguably the most comprehensive and rigorous empirical study of the consequences of human rights treaty ratification, Beth Simmons has argued that such treaties play a “crucial constraining role” in shaping the behavior of the states that ratify them. The public commitment to be legally bound by human rights norms that is expressed by treaty ratification influences various agents’ expectations regarding how ratifying states will behave. Subsequent shortfalls in state behavior trigger political demands for compliance, mainly from domestic constituencies, but also internationally. Simmons focuses on the role of three domestic mechanisms through which such demands are channeled: elite-initiated agendas, litigation and popular mobilization. Indeed, a notable feature of her analysis is the stress placed on “stakeholder agency”—the role played by those members of society who stand to gain from the recognition of human rights in taking measures to secure the implementation of treaties, as opposed to implementation achieved through the workings of a “white knight” transnational actor, such as the United Nations, a foreign state or a non-governmental organization.⁴⁰

Before leaving the question of efficacy, however, it is worth underscoring one conceptual confusion that underlies some skepticism about the possibility of an international rule of law in the absence of anything comparable to the characteristic features of domestic government at its most effective. Primary among these features are reasonably effective enforcement mechanisms, whether they operate formally, through legal institutions, or informally, through public opinion and the efforts of activists and interest groups. Proponents of radical skepticism treat as a litmus test of international law’s causal efficacy the question whether states conform to it when they have no interest in doing so or perhaps an interest in not doing so. On the basis of various empirical case studies, they contend that in the overwhelming run of cases there is very little evidence for a positive response.⁴¹

For all its vaunted hard-nosed empiricism, radical skepticism often appears to rest on a number of a priori, but deeply problematic, assumptions. To begin with, the notion of a state’s “interest” is usually not well-defined. If it is construed rather narrowly, as referring to objectives such as security, wealth and power, then it is highly implausible that states never obey international law when it is not in their interests, or contrary to their interests, to do so. If, as is commonly the case, the notion of state interest is given a maximally broad interpretation, so that it even encompasses a state’s commitment to moral values—such as global peace, the eradication of extreme poverty, the promotion of democracy, etc.—then the hypothesis that states never act except in their own interest turns out to be trivially true, but is therefore also not amenable to empirical testing.

The failure to register the triviality of the second construal of states’ interests encourages the mistaken assumption that if international law is to be a causally significant factor in influencing state behavior, states must be motivated to obey it independently of any values or interests that might be served by their doing so.⁴² But this fetishistic interpretation of what it would take for international law to be causally efficacious is patently erroneous. Instead, the question is whether a non-negligible number of states, on a non-negligible number of occasions, conform to international law because they take it to bind them. If they believe that international law is binding, it will no doubt be because they also believe that obedience to it serves important interests and values. But this does not mean that its being law is causally redundant. This is because the relevant interests and values are served precisely through treating international law as a source of content-independent obligations. Moreover, a state’s conformity to law may often be over-determined: obedience to law may operate alongside other motivations for engaging in a given pattern of behavior. This contrasts with the tendency of skeptics to discount the causal significance of international law once some independent motivation for the relevant behavior has been identified, as if being directly motivated by interests and values precluded the motivational influence of the perceived legitimacy of law. Of course, all of this is compatible with the existence of significant differences in the comparative efficacy of most domestic legal systems, on the one hand, and international law, on the other. But this is a far cry from licensing the conclusion that international law in general—or its human rights component in particular—is so causally impotent as to be fatally compromised in its legitimacy.

IV. Pluralism, Freedom, Exceptionalism

I conclude with some schematic observations regarding three principled limitations on the legitimacy of international law: value pluralism, freedom and exceptionalism.

Value pluralism is not the relativist claim that the truth of value judgments is brutely relative to some framework or other, with no objective assessment of frameworks themselves being possible. Instead, pluralism is a normative thesis about value, one that claims to be objectively true. It holds that there is a plurality of irreducibly distinct values; that these values can be ordered in different ways in a choice or in the life of an individual or a community; and that sometimes there is no single ordering which is uniquely correct.

If value pluralism of this sort is a compelling doctrine, as I think it is, then it is an important constraint on the operation of the NJC.⁴³ It means that we must attend to the question whether international law embodies an ordering of values that is just one among a number of eligible orderings. This is one way—a more productive way in my view—of giving substance to the repeated complaint that international law is “ethnocentric,” subjecting non-Western societies to a parochial, specifically Western set of ethical-political priorities. Usually, such views are defended, insofar as they are defended as opposed to brutely asserted, by reference to a dogmatic post-modern skepticism about value judgments in general. But a generalized skepticism of this sort is self-defeating, sawing-off the branch from which the post-modern critic launches their critique. A better way to understand the complaint about ethnocentrism is as the claim that there is a diversity of eligible ways of life, and that international law should not, without further justification, impose one of those ways of life, or elements of it, upon societies already committed to different ways of life. Of course, this constraint is a defeasible one, since there may be compelling reasons in particular cases for insisting on a legally-established uniformity of standards. But it does mean that we need to exercise vigilance, embarking on the difficult task of distinguishing between the particular and the universal within the domain of the objectively eligible. For example, when it comes to the human right to political participation, it may be that construing it as “the right … of citizens to choose their representatives through regular, free and fair elections with universal and equal suffrage, open to multiple parties, conducted by secret ballot, monitored by independent electoral authorities, and free of fraud and intimidation,”⁴⁴ should be understood as only one way, among various others, of fleshing out the content of this right. Other no less eligible ways might justifiably dispense with such requirements as equal suffrage (by giving multiple votes to the poorest members of society, or to those belonging to historically marginalized and oppressed groups), secret ballots or even the party system.

One reason the theme of pluralism has salience is that we prize the freedom of communities, and of the individuals who are their members, to choose from an array of eligible options.⁴⁵ Recall that according to the NJC enhanced conformity with reason is the touchstone of legitimacy. Now, among the reasons that apply to us are not only first-order reasons to do, or not to do, certain things, but also second-order reasons. Second-order reasons are reasons for acting on some, but not other, reasons. Among the second-order reasons that apply to ordinary human beings, and also to some collectivities, are reasons of freedom or self-determination. These are reasons to decide certain matters for oneself because of the intrinsic or instrumental value of doing so. It is such reasons that explain why no one has legitimate authority over us regarding whom we should marry or which career to pursue, even if by following their directives we would likely select a more suitable spouse or career than if we exercised our own independent judgment. This follows from the high value that accrues to making and pursuing such personal decisions autonomously, in a way that gives expression to one’s own judgments, tastes, and inclinations.

It is worth noting that affirming the value of communal self-determination does not entail that political communities have an ultimate value comparable to that of the individual human beings who are their members; on the contrary, collective self-determination is valuable insofar as it serves the interests of these individuals. A question also arises whether collective self-determination is only intrinsically valuable in the case of democratic states. In this connection, it seems to me that Rawls rightly considers that some non-democratic states—such as decent hierarchical societies—are capable of realizing this value to an extent that merits the protection of international law. And even if the intrinsic value of collective self-determination is largely confined to democratic states—in a very broad sense of “democratic”—it may be appropriate for international law to extend the protection it justifies, under the aegis of the principle of state sovereignty, to all but a few states. This is on the grounds that the primary responsibility for bringing about the democratic reforms necessary for that value to be adequately realized falls on the members of the society in question and that, in any case, intervention by external agents except in extreme cases is likely to be either counter-productive or to have destabilizing consequences for the global state system.

To the extent that agents have reasons to choose autonomously, the latter too come under the auspices of the NJC. Now, it is a complex question whether states have reasons of freedom to decide for themselves regarding matters within the domain of human rights, such that those reasons limit the legitimacy of international legal institutions, even if obedience to those institutions would have led to enhanced compliance with human rights. I think the verdict here must be a nuanced one, depending on the nature of the human rights issue in question and the kind of authority claimed by international law, e.g., whether it purports only to impose an obligation or also some enforcement mechanism in the event of its breach and, if the latter, what kind of mechanism. However, it is plausible that, when it comes to at least some human rights matters, the idea of a “margin of appreciation” or “subsidiarity” can be defended in part in terms of the importance of communal self-determination, even if the cost of recognizing this discretion is an inferior outcome with respect to conformity with human rights morality on the point at issue.

In addition to these reasons of freedom, which come under the NJC in the ordinary way, there are reasons which international legal institutions have not to interfere with the freedom of states, even though these are not straightforwardly reasons forthe states themselves. So, for example, it might be that although treating international law on a certain topic as authoritative would enhance a state’s conformity with reason, the actual upshot of international law’s claiming such authority would be to trigger a violent backlash against international law generally, one that leads to all sorts of undesirable geo-political consequences. This might be the situation if international law withheld certain incidents of sovereignty from states that failed to meet a non-trivial standard of justice with respect to the treatment of their own members; for example, if international law denied borrowing privileges to states that systematically and persistently flouted certain basic rights against discrimination on the grounds of sex. In other words, we have to recognize a freedom-based consideration that can operate as an exception to, or at least a defeater of, the NJC.⁴⁶

In light of the foregoing analysis, the proliferation of human rights norms within international law acquires a new significance. Even if certain norms of international human rights law accurately reflect human rights morality, they may lack legitimacy because they purport to bind states regarding matters that should be left for them to decide. In other words, even if rights not to be subjected to capital punishment or discrimination on grounds of sexual orientation really are universal moral rights, they may nevertheless in whole or in part fall within a zone wherein different societies should be free to make their own, albeit potentially inferior, decisions without being bound by norms of international human rights law or subject to intervention as a result of their breach. Of course, this understanding of human rights diverges from that endorsed by Joseph Raz, according to which human rights are essentially limitations on state sovereignty, in that their violation in principle justifies intervention by external agents, such as other states or international organizations.⁴⁷ But this view of the relationship between human rights and state sovereignty is problematic on at least two counts. Insofar as it purports to reflect the role of international human rights law, which is Raz’s overriding ambition, it implausibly takes all human rights to disable the protection ordinarily afforded by the principle of state sovereignty and, related to this, it relies on an eccentrically broad understanding of state sovereignty, one according to which sovereignty excludes even formal criticism of a state.⁴⁸

The final limitation on the legitimate authority of international law I wish to consider is that of exceptionalism. This is the idea that, although international law in general or on a certain topic may be binding on the great majority of states, there are some states it does not bind, because of special features those states possess. This sort of exceptionalism represents one strain of American foreign policy: we saw it surface, for example, in some arguments as to why, in launching the war against Iraq, the United States was not morally bound by international law on the use of force. As the sole remaining super-power, with a powerful commitment to democracy and human rights, the argument runs, the United States should not labor under many international legal restrictions that properly bind other, weaker and less benign, states.⁴⁹ However, I hasten to add that this is just one strand of American foreign policy, because there is also another view, eloquently expressed by President Eisenhower: “We cannot subscribe to one law for the weak, another for the strong; one law for those opposing us, another for those allied with us. There can only be one law—or there shall be no peace.”⁵⁰

Now, of course, assertions of exceptionalism are regularly greeted with alarm or derision by other states, as if they could amount to nothing more than the hypocritical evasions of a hegemonic power that wants to benefit from others’ compliance with the burdens of international law whilst selfishly exempting itself from those very same burdens. But although exceptionalism is certainly an idea ripe for abuse by powerful and self-serving states, it would be a serious mistake to dismiss it out of hand as a constraint on the legitimacy of international law. That we need to make room for the possibility of exceptionalism was already foreshadowed in my discussion of what I termed subject fragmentation under the NJC (see section III, above).

The NJC establishes a litmus test of legitimate authority that is essentially relational in character: the question is whether some particular institution enjoys legitimate authority over some particular would-be subject with respect to some particular domain of conduct. And it is perfectly possible that the self-same institution has authority in a given domain of activity over many of its purported subjects, but not all. This is because the exceptional few would not better conform with reason by taking the institution’s directives as binding. This may be because of special qualities, with respect to knowledge, technical capabilities, or moral virtue, possessed by the exceptional few, or perhaps special features of the predicament they confront.

Now, it seems to me highly implausible that any state could carve out for itself a blanket exemption from subjection to international law, one that extends from international telecommunications law to the law on the use of force. More likely is the scenario that some states can properly claim an exemption with respect to someareas of international law. So, for example, it is a serious question to what extent democratic states, with a strong tradition of commitment to human rights, are bound by the international law of human rights. But any case for exceptionalism on their behalf must reckon with at least the following three considerations. The first is that the human rights-related reasons that apply to a state are not only the reasons it has to adhere to human rights morality itself, but also its reasons for promoting such adherence by other states. This means that the reasons liberal democratic states have for promoting human rights beyond their borders may be decisive in subjecting them to the authority of international human rights law. Of course, much here will turn on the answer to the empirical question addressed by political scientists such as Beth Simmons: the question whether the participation of liberal democratic states in an international legal regime of human rights tends to foster compliance with human rights in other (non-liberal democratic) states.⁵¹ The second point is that we are not confronted with a zero sum situation in which either international human rights law or domestic rights jurisprudence is binding but not both. Both bodies of law may enjoy legitimacy with regard to a given state, with the result that some kind of ethical-legal judgment will need to be made in concrete cases in which the laws of the two systems conflict.⁵² The third point to take into account is that, insofar as we are concerned with a liberal democratic state’s own compliance with human rights morality, we have to consider its extra-territorial record in addition to its domestic record. And this may lead to a divided verdict, such that it is bound by international human rights law in its extra-territorial activities, because in relation to them there is no analogue to the safeguards that foster compliance domestically.

One final observation. I have so far only discussed exceptionalism with respect to powerful, liberal-democratic states. But there is no reason to treat exceptionalism as the exclusive preserve of such states, although they are perhaps more likely to claim that status. On the contrary, arguably the most plausible beneficiaries of exceptionalism are severely impoverished, underdeveloped or in some other way disadvantaged states. One reason this may be so is the especially onerous burdens that conformity with international human rights law would entail for such states. One such burden relates in the first instance to knowledge. Consider the following reflections, by the historian Anthony Pagden, on the idea that liberal democracy is a product of “the West”:

This does not mean that the champions of democracy are not also right in claiming that it is also the best obtainable government there is, and with the failure of communism, the most equitable way of distributing power and goods—at least at present. The mistake is to assume that this fact must be simple and obvious, in particular to those who have had no prior experience of modern democracy, and who inevitably equate it with Western imperialism, and Western godlessness, and whose first encounter with it is often at the end of the gun.⁵³

I do not quote this passage in order to give substance to the all-too-familiar suspicion that international law is radically illegitimate because, underneath the facade of its rhetoric about peace, dignity and human rights, it is in reality an instrument employed by Western powers to dominate and exploit non-Western societies. The point I wish to extract is rather more subtle and perhaps even true. It starts from the premise that one can only be subject to an authority if one can reasonably come to know of its legitimacy without undue expenditure of time and effort. This follows from the fact that the Normal Justification Condition is supposed to generate obligations capable of guiding action so as to improve conformity with reason.⁵⁴ The quoted remarks raise the possibility that societies with a very different cultural orientation and history may be so deeply mired in certain errors and misconceptions that they are seriously impaired in their capacity to grasp the legitimacy of much beyond perhaps a minimal core of international human rights law. This is not to deny that these societies are nonetheless bound by human rights morality, although their historically-induced misconceptions may generate some excuse (not a justification) for non-compliance with it. Rather, the idea is that much international human rights law may lack legitimacy with respect to these societies, at least at the present juncture, given their historical experience of international law or of states claiming to act in its name. The distinction is not ad hoc to the extent that there are special obstacles that obtain in the case of the legitimacy of international human rights law that do not apply, or do not apply to the same extent, in the case of human rights morality. Presumably these differences will be found in the fact that international law, unlike morality, is a creation of states, and that Western states, with their long and continuing history of imperialism and often tragically misjudged interventions, have historically played a disproportionately influential role in fashioning and implementing international law.

We should take the possibility of exceptionalism with respect to the binding force of international human rights law seriously. But we need to take that possibility seriously for others, not just for rich, liberal democratic societies.