In 1989, the UN General Assembly declared the 1990s the "decade of international law." In retrospect, the decade seems to have earned its name. We have witnessed remarkable legal accomplishments, in areas ranging from the law of the sea to trade and the environment. Legal action against human rights offenders has also made great strides, with criminal trials of defendants from Bosnia and Rwanda, and a recently-completed convention on the establishment of an international criminal court.

These efforts at legal construction have been easily matched, however, by persistent violations of human rights norms, not just in Bosnia and Rwanda, but in Iraq, Iran, the Sudan, the Congo, and too many other places around the world. For the countless victims of state violence and terror, this "decade of international law" has been a dismal failure.

The recent dramatic efforts by Spanish courts to extradite Chilean General Augusto Pinochet from England provide a good opportunity to think about why international human rights law always seems to have this dual appearance of progress and defeat.

The Conflict over Pinochet

In 1970, Salvador Allende was elected President of Chile. In September 1973, General Pinochet led a military coup against the elected government, in the course of which hundreds of people– including President Allende himself– died. From 1973 to 1990, Pinochet headed a repressive Chilean regime, which pursued a dirty war against those it deemed threats to the conservative political order. While exact numbers are not available, at least 3,000 people were murdered under his rule, including many non-Chileans. Indeed, it now seems that his government worked in coordination with other repressive military regimes in South America-especially the Argentine government-in an international extension of their dirty war.

Pinochet has never been subject to any legal process for his conduct. Instead, Chile worked out a peaceful transition to a democratic regime. As part of this negotiated transition, Pinochet agreed to a gradual withdrawal from governmental leadership. Still, at 83 he is a permanent senator, and as such enjoys immunity from Chile's legal process.

Though Chile chose not to prosecute Pinochet, a Spanish judge has taken up the case. The judge, who began by investigating the actions of the Pinochet regime with respect to Spanish citizens caught in Chile, has since expanded his inquiry, and seeks now to prosecute Pinochet for his actions toward Chileans and non-Chileans alike. As a first step, he has requested Pinochet's extradition from England, where he was recuperating from a surgical procedure.

Pinochet, it is alleged, engaged in a campaign of grave violations of human rights. Such violations are grounds for "universal jurisdiction"-a term of art in international law, which means that Pinochet may be prosecuted by any national regime that chooses to extend its jurisdictional reach to cover him. International law has long recognized universal jurisdiction over pirates and slave traders. Increasingly, the human rights violator too has been seen as hostis humani generis: an enemy to all mankind. So a US court concluded when it allowed a tort action for violation of human rights to go forward against a Paraguayan official. The International Criminal Tribunal for the former Yugoslavia similarly concluded that grave breaches of international humanitarian law are crimes that "are universal in nature." Moreover, several specific treatiesããincluding the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment, to which England is a party-require states to punish or extradite persons guilty of certain human rights offenses.

Over vigorous objection from the present Chilean government, a panel of English law lords decided that Pinochet may not claim head-of-state immunity for human rights violations, and thus may be extradited to countries that seek to prosecute him. The panel's decision has since been vacated on the grounds that one member of the panel failed to reveal his links to Amnesty International. And at this writing (January 8, 1999), there has been no final decision on the Spanish extradition request.

Whatever the ultimate results of the Spanish and British legal maneuverings, the events thus far provoke two quite different responses, sometimes within the same person. The first is that we are finally witnessing the maturation of an international regime of human rights. Pinochet, it is said, violated legal standards that have had a formal life at least since Nuremberg. But their real enforcement was stymied for decades by Cold War political conflicts and the absence of suitable international legal institutions. With the end of the Cold War, we now see the emergence of multiple institutions that are translating formal human-rights law into a living rule. Not only are there the ad hoc criminal courts for offenders from Bosnia and Rwanda, but-as with the arrest of alleged violators in Bosnia-we see the slow expansion of international peacekeeping forces into a human-rights police function. More generally, the UN Security Council is newly concerned with human- rights issues from Haiti to Iraq. That a Spanish judge would pursue the "internationalization" of domestic jurisdiction is yet another institutional development demonstrating an increasingly effective commitment to what have long been unenforced legal norms.

This line of argument about progress in international law is part of a broader contemporary story-about globalization, the retreat of state sovereignty, and the emergence of a cosmopolitan moral code. Emerging international human- rights conventions and institutions express moral standards that transcend national borders. Responding to serious human rights abuses is a moral obligation that falls on each of us as members of the single human community, and a legal obligation of state governments, even when they have little if any direct connection to the violations. Human rights are now everyone's concern. Tyrants can no longer hide behind state lines. The writ of justice knows no jurisdictional borders.

A second, less celebratory response, focuses on Chile as a self-governing political community. On this view, Pinochet and his regime are a problem for Chile to work out. A country must make its own history. Chileans need an honest reckoning with their own past, not moral instruction from Spain. More fundamentally, their real problem is not simply Pinochet, nor can it be addressed by holding him, or any other single person, accountable. They need to come to terms with a regime that was supported by many Chileans who remain members of the community. Partly because of such continuing membership, overcoming civil wars and authoritarian regimes are among the hardest problems faced by any society. Focusing on the individual is the approach of criminal law. But there is often a tension between our perceptions of individual criminal accountability and the judgments of collective political responsibility that are essential to reconstructing a political society.

When the Spanish prosecute Pinochet, they emphasize individual accountability: their narrative of the 1970s is about an evil man who gained control of the state and must be held responsible for his own conduct and that of subordinates. But competing narratives present Pinochet's actions as representative of a particular historical constellation of forces in Chile. Instead of seeing the criminal behavior of an individual, a society may look back and acknowledge that the dreadful deeds were things that at least in some sense it did to itself. Prosecution of the individual can work to displace responsibility from the community to the individual. Indeed, just this moral displacement was part of the Allied effort at the Nuremberg trials. Communities, of course, need scapegoats at times, and better to choose a morally evil scapegoat than an innocent one. But even this decision is ordinarily left in the hands of the relevant community.

From this community-centered perspective, reconciliation is as much a problem as retribution. Accountability must go forward in ways that do not fracture the community, leading to new cycles of civil conflict. This is not merely a practical imperative, reflecting the often tenuous circumstances of successor regimes-although that is part of it-but also a moral imperative, just as compelling as the punishment of individuals. A community must take responsibility for its past as it builds its future. No single prescription can set forth how this is to be done. For just this reason, outsiders have a moral obligation to respect a community's own choices, and the Chilean government is not wrong when it views the Spanish-and British-actions as offensive to community self-determination.

How should we respond to these two different perceptions of the Spanish prosecutorial effort? In the first, the international community stands for certain universal standards of individual behavior, and we are better off now that a number of legal institutions are becoming available to enforce these standards. In the second, Chilean political society should be respected as an independent moral agent, and should not be made to conform in its recovery from civil conflict to a single standard of law and punishment. We want to respect both individuals and communities, but the Pinochet case seems to force a choice on us.

There is no easy escape from this confrontation of normative perspectives. To be sure, we can claim that the international community has, in fact, already worked out a response to this dilemma. Human rights activists are likely to make this case, urging that the Spanish judge is only putting into effect legal norms established decades ago. The choice between these two patterns of response was already made, they will tell us, with the adoption of the Convention on the Prevention and Punishment of Genocide and the International Covenant on Civil and Political Rights, both of which have been ratified by Chile and Spain. Communitarians may not like it, but the Spanish judge is operating under the law, and that should be the end of the matter.

Such a retreat into the formalism of legal conventions, however, is unconvincing. We now have a decade of experience with transitions from authoritarian to democratic regimes, and trials under international human rights standards have not figured prominently in these transitions. Diverse methods of community reconciliation have been deployed in place of trials, with many ways of acknowledging violations of moral and legal norms, including "truth commissions" writing reports or conducting hearings, civil actions, opening secret files, prohibitions on holding public office, reparations, and efforts at social redistribution. Argentina did experiment with domestic criminal trials, but the results were not promising: attempted military coups, truncated procedures, regime change, and ultimately amnesty.

Whatever we think of these transitional arrangements, we cannot condemn them as violations of law. International law is not simply a matter of formal rules, written in treaties; it is also a system of mutual behavior among states. And no state has asserted that successor regimes that fail to hold trials are in violation of international legal obligations: no sanctions have been imposed for such alleged failures; no resolutions of condemnation have been passed; no one has intervened to correct this alleged abuse of a remedial requirement. We find widespread tolerance, indeed encouragement, of diverse community responses to an authoritarian past, even when those responses include grants of amnesty. At least that's what we found until the Pinochet case.

The other prosecutions we have seen, those arising out of Bosnia and Rwanda, differ precisely in that they have not been interventions in a transitional process achieved within the domestic political community. Instead they have been imposed as part of an international peacekeeping program. Their legal ground has not been the human rights norms themselves but the Security Council's authority to intervene in situations of violent conflict.

The debate over judicial prosecution of human rights offenders has always gotten more attention than it deserves. There is undeniable moral satisfaction in seeing such prosecutions. But this moral satisfaction is easily confused with fulfillment of a legal obligation. Worse than that, prosecutions come to symbolize a legal order, as if success in prosecution is the equivalent of achieving an international legal regime of human rights. The failures of human rights law are, however, far deeper than failures to prosecute. Bringing Pinochet before a Spanish court, or even establishing an international criminal court, will hardly begin to redress these failures, which go to the very core of the idea of law.

Law and Self-Government

A legal regime cannot be said to exist in the absence of a pattern of compliance with the rules, and thus an expectation of such compliance. A rule that has the formal appearance of law and has been enacted through an official procedure but fails to govern behavior is not law. It may describe a community aspiration; it may set forth an ideal state of future law. But it is not law unless we can see how it works as a normative standard for some community.

International law is strewn with false hopes. In 1928, the countries of the world formally declared a legal prohibition on war in the Kellogg-Briand Pact. Was it law? Was World War II against the law? In the immediate aftermath of that war, the new United Nations produced the Genocide Convention. Was this law? Are we to understand the last 50 years of history in places like Cambodia, Indonesia, Iraq, Liberia, Ethiopia, Guatemala, and Rwanda as legal problems? In 1966, the General Assembly adopted the Covenant on Civil and Political Rights. It entered into force in 1976. If that was law, how many of the countries of the world have lived in a chronic state of illegality since then?

We cannot usefully describe a political order as in a chronic state of illegality. Such a regime, we should say, is one without law. Neither can we usefully describe as law rules that fail to organize any community's behavior.

Ordinarily, we can read the legal order off the on-going behavior we observe in a community. If we see a tax code, enacted through a formal rule-making procedure, but no payment of taxes, we should be wary of describing this as legal regime of taxation. When we hear that in Russia no one pays taxes, we confront just this situation. To say that there is law but no enforcement only produces confusion. It gives too much normative weight to rules that may have no substantive presence. A functioning tax regime, for example, might need to look very different from the formal tax code. Or consider whether blue laws are really law in American communities. A rule with no bearing on behavior is law only in name, not in fact. If a prosecutor suddenly decides to prosecute someone under norms that no one follows or expects others to follow, that action itself seems to violate the rule of law.

The failure to take account of this required connection to a community's behavior has led to some of the more extreme claims of international lawyers. For example, the International Court of Justice was recently asked to declare the use or threat to use nuclear weapons a violation of international law. It came very close to doing so, despite the fact that the nuclear weapons regime has been at the foundation of the international order in the entire post-war period. The court was willing to contemplate the possibility that the international community has organized itself in a fundamentally illegal manner. What could such a legal judgment mean? Can we imagine the international community reorganizing itself to meet such a legal judgment? It would be as if the Supreme Court declared property an unconstitutional violation of the legal norm of equal protection.

But if the enactment of rules through a formal procedure does not suffice to produce the rule of law, neither does effective enforcement of norms. Regimes in which compliance is brought about only by threat are authoritarian or dictatorial-hardly the virtues of the rule of law. The old regimes of Eastern Europe did not lack for compliance through enforcement, but we are not likely to describe them as regimes operating under the rule of law. We expect a rule of law to assert a normative pull on citizens. If that pull comes from nothing other than the threat of punishment, then the legal rule makes no normative claim in and of itself. Law becomes simply another way to control behavior through the application of force. Of course, the use of force is not necessarily a bad thing: it depends on the ends for which it is deployed.

We need a different set of categories-different from origins, or compliant behavior-to describe the normative pull of the rule of law. Here, some traditional terms can help: a legal regime expresses the "sovereign will" of the community. Conceptions of law that are disconnected from the idea of sovereignty all fail to see the crucial connection of the rule of law to self-rule. In a democratic regime in particular, sovereignty and law are tightly linked: the sovereign people govern through the rule of law; and by following the law, citizens participate in popular sovereignty and achieve self-government.

We expect citizens to use laws as standards for regulating their own conduct as well as standards for evaluating the behavior of others. But the quality of the commitment to the legal rules matters, too. We begin to understand the special normative character of law if we start from the observation that legal regimes demand sacrifice from citizens, and citizens accept this demand as legitimate. If the demand is rejected as "illegitimate," then we are likely to say that law does not rule in that community, but that the system of enforced norms is imposed from without or imposed by a minority on the majority. Legal legitimacy points to an identification of the individual with the political community. The quick collapse of the regimes of eastern Europe demonstrated just this absence of a willingness to sacrifice on the part of citizens. What purported to be law in those countries did not represent the sovereign will; it was therefore rejected as illegitimate-as a rule of some people over other people in the guise of the law.

The unique quality of a legal regime is revealed in this dimension of sacrifice. Unlike the moral order, the legal order expects and demands that it will be defended by citizens as an ultimate value. The rule of law is not an incidental characteristic of the modern state. Rather, it defines the state as a democratic project in which individual members so identify with the community that in following the law they act freely: as Rousseau put it, they act on laws that they give to themselves. I am not concerned here with the conditions under which this proposition linking the rule of law to autonomy and compliance can be held to be true or morally defensible. Rather, I am concerned with the experience of law as a living, normative regime. We don't have a regime of law until legal norms are observed in the community's behavior, and observed because they are seen by citizens as an expression of their own identity-which identity is itself understood as worth defending.

Law is so important in the modern period not simply because it promises order, but because it expresses an understanding of the political order as a community of equals that freely creates itself on the basis of deliberation and choice. This was the idea behind the two great projects of state creation in the modern period: first revolution and then decolonization. For both, the first task of the new state was to write a constitution in order to define itself as a self-formed community. This explains why the decolonization process could be so remarkably indifferent to ethnic diversity within inherited colonial borders. That diversity made no difference from the perspective of the modern project of writing the nation-state into existence through the construction of a legal regime. Decolonization represented a commitment to the same idea of the state that earlier made possible the idea of a "nation of immigrants." In both, citizenship under law-not ethnicity or origin-is the primary category of political identity.

This idea of the place of law's rule in the modern nation-state offers a means of partially bridging the gap between the competing conceptions of universal moral values and community autonomy. If the state under law is a communal project of self-creation, then the discourse of state-creation is necessarily a normative discourse that will appeal to current ideas of justice. We create law in order to achieve a just political order. If we are to see ourselves in the legal order, that order must hold forth a normatively attractive vision. Legal discourse is, therefore, unavoidably a moral discourse in its foundations. This does not mean there is an exact overlay of moral and legal discourse, but it does mean that law's creation and maintenance necessarily borrow a normative vision from the larger discourse of morality.

We expect modern constitutions to include guarantees of human rights because this is the dominant moral discourse today. We are mistaken, however, if we believe it is the content of the values, rather than law's constitutive character within the community that supports the compelling claim made upon individual citizens. It is not the case that liberal democratic regimes everywhere can claim our sacrificial support; expressing correct moral values does not allow a state to draft non-citizens. We may contribute sympathetically to the causes with which we agree, but we do not think the state should compel sacrificial acts for other political communities, regardless of our agreement with their values.

International Law?

These last observations about law and justice allow us to make some sense of our conflicting intuitions in the Pinochet matter. Of course, the values affirmed in the human rights conventions of modern international law are congruent with the values we expect to see affirmed in the modern nation-state: rights to life, liberty, and due process; individual accountability for wrongs done; and tolerance for difference. These values are the currency of our present moral discourse, both within and outside the borders of the state. But the affirmation of these values does not amount to a rule of law at the international level because of the absence at that level of all of the essential characteristics of the rule of law in the modern nation-state.

First, we cannot speak of international law as expressing the will of a sovereign of which we are each a part. International law results from delegations of power from multiple state sovereigns. There is as yet no world community with a sovereign will, and no conception of citizenship as a part of this single, world sovereign. There is, of course, an apparatus of international institutions. But, once more, rules do not themselves create an international community under law. While the production of law may contribute to the development of a single political community, it cannot do all the work. The human rights law that we have is not part of a richer discourse or set of practices that characterize a single international community. Instead, the formal statement of law may be all the community we have.

Second, we lack the idea of a legitimate demand for individual sacrifice in support of international law. We cannot now conceive of the United Nations instituting a draft or compelling the use of state forces; we are most uneasy with the idea of using state force to defend international norms in situations of limited relevance to national interests. Just this failure to act leads to perpetual charges of hypocrisy when action is taken: against Iraq but not Israel, against Rwanda but not South Africa. International peace keepers are deployed at times, but only under exceptional circumstances, which rarely include situations in which the parties do not want the peace kept.

Third, the construction of international legal norms remains a project of elites, not an expression of individual and community self-formation. Human rights law, though it purports to apply to individuals, is still made by states. Despite the increasing presence of non-governmental organizations, for most citizens most of the time the relationship to international law is mediated through the political community that is constitutive of political identity: the nation-state.

Of course, all of these characteristics are under substantial strain, particularly in Europe. There, we see an emerging concept and practice of "European citizenship"; we can imagine an international draft; and there is an increasingly direct relationship between the individual and a transnational community. We do not know whether Europe is emerging as a successor nation-state to the traditional states of Europe, or if instead Europe is a first step toward an international order. But we do know that nothing even remotely approaching the transnational community that is Europe exists anywhere else. Indeed, elsewhere, the strains on the modern nation-state come not from an emerging internationalism, but from a resurgence of nationalism that fractures rather than aggregates states.

International human rights law, then, has not been part of a legal regime: it is a system of formal rules of undoubtedly praiseworthy moral quality, but lacking the connection to a practice of self-government that I have associated with the rule of law. What would it take for international human rights law to become part of a legal order? At a minimum, we would need to see the international community become much more serious about these norms as standards for behavior applicable to all members of that community. In view of this requirement, the Pinochet prosecution comes too late. An international community that allows massive violations of human rights as long as a regime is in power, but then threatens to prosecute offenders once they are out of power-and even then only if they happen accidentally to fall within a state's jurisdictional grasp-is not a regime of law's rule. The question is not, what does the law demand of Pinochet now? But what did law demand of all those states that maintained relations with Chile during his regime? A regime in which no one is willing to make sacrifices in defense of the formal norms is not a legal regime.

If international human rights law is to be law, it must be so consistently, and above all when it counts. It must be a rule of behavior for governments-even powerful governments-toward their own citizens and toward other governments that violate the norms. Human rights law confronts its moment of truth when the violations are threatened. If the world will not protect victims in Chile, Bosnia, or Rwanda, then it is wish, not fact, to describe human rights norms as law. The promise of international human rights is a promise to the potential victims that a regime of law governs the behavior of the state toward its own citizens. It is little solace to the victims to learn that violators may be punished in the future. Imagine a domestic law of murder that made no effort to stop murderers but promised to prosecute them after they committed the offense.

In Bosnia, human rights law failed completely-whatever the successes of the criminal prosecutions before the International Criminal Tribunal. The same is true in the Pinochet case. No regime of international human rights law protected the victims in Chile. There was no such law, because no state was willing to sacrifice to give it life.

Why is the international community so reluctant to enforce human rights law as law? Partly because, as I have said, citizenship remains bound to the state. However morally compelling, international human rights claims do not appear with the force of law in any politically organized community. But this fundamental role of states itself reflects a commitment by the international community to the nation-state as the politically relevant community, even when this organizing commitment carries large moral costs. The idea of domestic jurisdiction is itself protected in the UN Charter. The defense of human rights continues to appear as "intervention" rather than as law enforcement within a single world community. And this "anti-intervention" impulse draws its normative force precisely from the conception of the political community as a self-forming, historical presence. Despite universal human rights norms, we continue to be drawn to the idea that political communities have a right to their own villains, their own civil wars, and their own recoveries from these conflicts. The history of separate political societies, not of mankind, continues to dominate our political sentiments.

Law, Morality, and Pinochet

How then should we understand the Spanish effort to prosecute Pinochet for the violation of international human rights norms? We should understand this effort as a use of legal means to pursue moral ends. The means are justified by the morality of the end, not by any fact of law.

Consider a harsh analogy. Suppose that the Spanish took up as their own mission the pursuit of justice against violators of international human rights. Without stretching too much (given what we know about Spanish activities against Basque separatists), we can imagine the Spanish government forming a "hit squad" whose mission was to go after grave human rights violators around the world. This decision would have a disquieting effect on tyrants everywhere, who would have to contend with the thought that they would never be safe from moral retribution for their deeds.

The Spanish would clearly not be acting under the color of law in this pursuit, but their pursuit of punishment might be a morally good thing. At least, I would be willing to defend their actions on a variety of moral grounds, from retribution to utilitarianism. We might wonder why this was any business of the Spanish, yet countries do concern themselves with humanitarian enterprises in other countries. The Spanish human rights hit squad might accomplish more good than any foreign aid the government could offer. If we cannot imagine a state pursuing international justice in this way, then imagine a group of committed human rights activists adopting these ends. I do not think the moral issues would be any different if we substitute private actors for the state.

We might worry about the lack of legal process, but we should be clear about the nature of our worries. Does anyone really believe that there would be a problem identifying wrongdoers-perhaps Pinochet has a twin brother? Some may want to argue that Pinochet lacked the requisite personal knowledge to hold him accountable. I think that is pretty far fetched with respect to the likes of Pol Pot, Karadzic, or Pinochet. Are we worried that Pinochet needs a public forum in which to make his defense, perhaps arguing that there was a communist threat to which he had to respond? He has had 25 years to present this argument. It has been well aired, but not convincingly made. There is something reassuring about a legal process, but law does not absorb all of our moral concerns, and its reassurances may contribute little to those concerns.

This is just how I understand the Spanish extradition request. If the Spaniards want to devote their resources to hunting down and punishing the Pinochets of the world, and to use legal resources in pursuit of this end, I am not going to object. I see their efforts as a kind of guerrilla warfare against human-rights violators: one uses the tools at hand, including legal tools, to pursue morally compelling ends. The morally defensible end justifies the means here because Pinochet himself can raise no countervailing value against using these means. He is treated harshly, but not unfairly. He is getting no more than he deserves, whatever the law has to say about this.

But what if Chile, not Pinochet, makes the counter argument, claiming the right of community self-determination to deal with its own villains? That is indeed a compelling value. But that is all it is. It is not a superior claim; it is an equal claim. It is not a ground for condemning the Spanish action, but only an equally valid ground for alternative action. When faced with this sort of moral equivalence, we are all on our own. It is wrong to think that every serious issue has only one possible moral resolution. In this situation, one might say the British are lucky: they cannot go wrong from a moral point of view. They can respect the claim for autonomous self-formation by Chile or they can respect the universal norms defended by the Spanish.

To be sure, these competing positions each involve familiar dangers. We should worry about the moral judgments of human rights hit squads-even if they take the form of the Spanish government. Will they always identify the worst violators? Will their judgment be colored by politics or personal interests instead of human rights norms? These are serious worries that might raise a practical bar to recognition of any such self-appointed group. Yet, on the other side, how do we know that Chilean objections reflect the self- determination of the Chilean community as opposed to the perversion of that community by lingering military threats? Are we witnessing a community writing its own narrative, or the continued control of that community by the human rights offenders themselves? Arguably, the threat of foreign prosecution of Pinochet might liberate factions of the Chilean domestic community that have thus far been immobilized by fear or threat. These, too, are good arguments that should make us cautious about moralizing in this situation.

Yet caution is not an excuse for inaction. Something must be done, and we can do no better than to act on our own best judgment under the circumstances. But make no mistake: the situation requires judgment. Law cannot serve as a crutch. The Spanish may be engaging in highly praiseworthy action, but they cannot rightly claim that they are enforcing the rule of law, or that international law requires such action, or that what is at stake is law and not morality. The issues in the Pinochet case remain moral, not legal.