The concept of humanitarian intervention has evolved as a subset of the laws governing the use of force and has very quickly come to occupy an institutional position alongside self-defense and Security Council authorization as a legal and legitimate reason for war. It is both widely accepted and yet still highly controversial. This article considers whether humanitarian intervention is legal under international law. This is a common question but one that produces an uncertain answer: humanitarian intervention appears to contradict the United Nations Charter, but developments in state practice since 1945 might have made it legal under certain circumstances. Those who argue for its legality cite state practice and international norms to support the view that the prohibition on war is no longer what it appears to be in the Charter. The debate suggests that humanitarian intervention is either legal or illegal depending on one's understanding of how international law is constructed, changed, and represented. Since these questions cannot be answered definitively, the uncertainty remains fundamental, and the legality of humanitarian intervention is essentially indeterminate. No amount of debate over the law or recent cases will resolve its status; it is both legal and illegal at the same time.
This article examines the implications of this finding for the idea of the rule of law in world politics. It suggests that the traditional emphasis that scholars have put on compliance with international law is misplaced; that is, the power of international law in this case comes not from its ability to differentiate rule breakers from rule followers, but rather from its ability to shape the terrain for political contestation in international relations. To the extent that state practice alters the meaning of international law, the distinction between compliance and noncompliance is unsustainable. Disputes over compliance and noncompliance are proxies for disagreements over the substantive behaviors in question, and they cannot be resolved by reference to the rules themselves. As I argue, international law should be seen as a resource that is used by states, rather than as a fixed standard against which we can assess behavior.
The first section of this article reviews the main elements in the legal regime on the use of force. These begin with the UN Charter, and especially Articles 2(4) and 51, but also include other treaties, such as the Genocide Convention, and other organizations, such as the African Union and NATO. Customary and treaty laws on self-defense are relevant as well. Together, these pieces help define the legal conditions under which states can use force against others. They constitute the current legal environment in which war is conducted. The second section considers how humanitarian intervention fits into this environment. It examines the evidence that humanitarian intervention is illegal and then the arguments for its legality. The former view rests on the plain language of Article 2(4) and the UN Charter as a whole, while the latter position considers the behavior of states and finds that their actions have modified the black-letter law of the Charter. These two competing views cannot be reconciled, and so the third section argues that this indeterminacy is inherent in the idea of the rule of law for world politics.
It is not my goal here to argue for or against humanitarian intervention; I do not conclude that humanitarian intervention is wrong or unwise or illegitimate, or the opposites of these.Footnote 1 Rather, my aim is to show that the practice of humanitarian intervention exists in a space between legality and illegality, one where each instance of the practice can be plausibly seen as either compliance or noncompliance with international law. This article begins by trying to clarify what we know about the existing laws on humanitarian intervention. In a well-ordered world, knowing the state of the law should make it possible to assess the compliance or noncompliance of governments in particular cases. However, I conclude that the legality of humanitarian intervention as a category is indeterminate, and as a result the idea of compliance in particular cases is close to meaningless. Despite this, I find that the rules on humanitarian intervention are indeed consequential, but not as a yardstick for measuring compliance. What, then, is the power of law if an act can be simultaneously a violation and a compliance? The contribution that international law makes to international politics does not come at the moment where states make a choice between compliance and noncompliance. Instead, it comes in providing the resources with which states interpret, justify, and understand their behavior and the behavior of others. This is both a constraint and an opportunity for states.
The Law on the Use of Force
International law is centrally concerned with regulating war between states, and well-developed bodies of law exist on state conduct in war and the decision to use force. Both have long histories in European public international law. They originate in Christian doctrines of natural law, merge with European great power accommodations in the nineteenth century, and progress through the codification movement in the twentieth century.Footnote 2 The fundamental piece of law on the legality of the recourse to war by states is the UN Charter. It makes two contributions that are central to today's legal regime on war: it outlaws the use of force on the part of individual states, and it empowers the Security Council to make all decisions on collective measures that involve military force. Article 2(4) establishes the first element by requiring that states not use or threaten force against other states: “All members shall refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any state, or in any other manner inconsistent with the Purposes of the United Nations.” This is a general prohibition, set in the section of the Charter that defines the common and primary obligations of UN membership and of the organization itself, and it is often cited as the primary contribution of the UN system to international order.Footnote 3 It goes along with Article 2(3), which insists that UN members settle their interstate disputes by “peaceful means.”
Article 2(4) takes away from states the legal right to use force, and Articles 24, 39, 42, and others then deliver this power to the Security Council. These sections of the Charter establish that the Council has the “primary responsibility for the maintenance of international peace and security” (Article 24) and that it can take what measures it deems necessary in that pursuit, including military action against states or other threats (Article 42). The goal of the framers of the Charter was to centralize the enforcement of international order in the hands of the great powers at the time, and to pacify the relations among other states by depriving them of independent legal channels to war. This was motivated by the understanding that the lesson of the two world wars was that state aggression must be forestalled with a forceful and collective response. Thus, intervention that is authorized by the UN Security Council is unambiguously legal, as long as it conforms to the Council's authority over “threats to international peace and security” (Article 39).Footnote 4
In this legal environment, the principal legal justification for war by states is self-defense. States have long claimed that military force used in response to an attack by another constitutes a distinct category in law and in practice, and as a result the canon of international law generally recognizes such a right. The customary understanding of self-defense goes back as far as the field of public international law, which is to say that it was recognized by Grotius and others in seventeenth-century Europe as existing already.Footnote 5 The concept is defined as a military response to an armed attack where the response is both necessary and proportionate to the attack. In the history of the concept, it is these ideas of necessity and proportionality that generate controversy; the concept itself is not contested. Each application of the concept in practice has a productive effect that further elaborates its meaning, sometimes making it clearer and sometimes making it more complicated.Footnote 6 For instance, Israel's claim of acting in self-defense in its attack on Iraq in 1981 was widely rejected, including by the Security Council,Footnote 7 but it incidentally may have helped define the outer bounds of “necessity.” The Caroline affair, which arose from skirmishes between the United States and Britain in 1838, provides a case in which a state's justification for its behavior has become constitutive of the categories of lawful and unlawful uses of force. The British eventually apologized for their incursion into U.S. territory, and the Americans conceded that the idea of anticipatory self-defense might exist within the concept of self-defense, but the most lasting effect of the incident was the language that it generated to judge claims of preemptive war: that the threat must be “instant, overwhelming, and leaving no choice of means, and no moment for deliberation.”Footnote 8 Michael Byers has rightly argued that prior to laws banning war, self-defense acted as a political justification rather than a legal exemption since, without laws to demarcate between legal and illegal wars, the justification of self-defense is politically useful and not legally necessary.Footnote 9
The UN Charter as originally proposed by the United States in 1944 did not contain any reference to self-defense as a complement to the ban on war. It relied on the fact that self-defense was a widely accepted custom according to international law. What became Article 51 of the Charter was added during the 1945 San Francisco conference that founded the United Nations, and was inserted at the initiative of Latin American states, concerned that the ban on war might be interpreted to mean that they could not ask the United States to come to their aid while the Council deliberated on its response to an attack on them. The great powers did not oppose adding it, believing that it did not change the underlying customary law. The language of Article 51 reflects the peculiar status of a customary principle in a codified environment: “Nothing in the present Charter shall impair the inherent right of individual or collective self-defense if an armed attack occurs.” This makes it explicit that a right to self-defense in the event of an attack exists prior to and alongside the Charter, and that Article 2(4) is immaterial to that right.Footnote 10
The international legal regime on the use of force is therefore constituted at the intersection of Articles 2(4), 39, and 51 of the UN Charter: the use of force by states against other states is prohibited by Article 2(4); the collective use of force is allowed, and is controlled entirely by the UN Security Council by Article 39, among others; and self-defense in response to an attack is defined by Article 51 as legally distinct from what is prohibited by Article 2(4).
This is the legal environment into which humanitarian intervention was presented as a justification for the use of force. It is an environment in which there are clear black-letter law prohibitions on the use or threat of force in interstate relations, and the development of humanitarianism has therefore taken place in and around that prohibition.
Is Humanitarian Intervention Legal?
In the face of these laws, can humanitarian intervention ever be legal? Recent events, from Rwanda to the Balkans to Libya and onward, have forced to the surface the tensions between humanitarianism and sovereignty, and the resulting debates have produced a set of positions on either side that are clearly identifiable.Footnote 11 Disagreements about deep points of international law, including how law changes in response to practice, how treaties are interpreted, and the meaning of compliance and noncompliance in particular cases, overlay a remarkable consensus that humanitarian intervention is an important tool for states and international organizations whether it is legal or not. The disagreements over how international law works, alongside a consensus in favor of the practice regardless of its legality, suggests that humanitarian intervention is likely to exacerbate the ambiguities inherent in the idea of the rule of law for sovereign states.
The Case for Illegality
The case for the illegality of humanitarian intervention rests on the plain language of the UN Charter. Article 2(4) outlaws the use of force by states and gives no suggestion that the motive behind the action matters at all. Nothing in the Charter opens the possibility that the use of force for humanitarian purposes should be understood any differently than other uses of force. Indeed, as Nikolas Sturchler reminds us, the article outlaws both the use of force and the threat of its use, ensuring that the domain of illegality is much broader than merely cross-border military attacks.Footnote 12 The prohibition is very widely drawn, and this was no accident: the overarching purpose of the San Francisco conference was to ban war and to build an architecture in the Security Council to enforce that ban and deal with violations.Footnote 13 Bringing this idea to the present day, Ian Brownlie has said that “whilst there have been obvious changes in the political configuration of the world . . . these changes have not had any particular effects on the law.” He suggests that humanitarian intervention was understood as legally defensible prior to the Charter, but became illegal in 1945:
By the end of the nineteenth century the majority of publicists admitted that a right of humanitarian intervention (l'intervention d'humanité) existed. A state which had abused its sovereignty by brutal and excessively cruel treatment of those within its power, whether nationals or not, was regarded as having made itself liable to action by any state which was prepared to intervene. . . . [By 1963] few experts believed that humanitarian intervention had survived the legal regime created by the United Nations Charter.Footnote 14
The Swedish government summarized the conventional understanding in a response to the Israeli Entebbe incident in 1976: “The Charter does not authorize any exception to this rule except for the right of self-defense and enforcement measures undertaken by the Council under Chapter VII of the Charter. This is no coincidence or oversight. Any formal exceptions permitting the use of force or of military intervention in order to achieve certain aims, however laudable, would be bound to be abused, especially by the big and strong, and to pose a threat, especially to the small and weak.”Footnote 15
The prohibition on war may be narrowed somewhat by the fact that Article 2(4) outlaws force only “against the territorial integrity or political independence” of a state. The substance of this clause has never been made clear in law or in practice, and its original intent appears to have been to expand rather than contract the scope of the ban on war.Footnote 16 It was added to the draft Charter at San Francisco by a group of small and medium-sized states that wanted to be satisfied that their independence was well protected. The argument could by made that humanitarian intervention does not involve an attack on “the territorial integrity or political independence” of its target state, and as such falls outside Article 2(4). This was essentially the argument presented by Britain to the International Court of Justice in the Corfu Channel case (1946–1948), claiming that its uninvited minesweeping in Albanian waters did not rise to this standard of intervention.Footnote 17 The argument failed in that case, and it has little basis in the text of the Charter or its traveaux preparatoires.Footnote 18
The Charter does include references to human rights and “fundamental freedoms” for individuals, which might be read as endorsing a kind of humanitarianism. However, these do not attach to any legal commitments by the signatories, and so do not create a possibility for armed intervention in their pursuit. The famous passages of the Charter that refer to the “faith in fundamental human rights, in the dignity and worth of the human person” (Preamble) and to the “universal respect for, and observance of, human rights and fundamental freedoms” (Article 55(c)) all arise in a nonbinding context: they are goals that the UN “shall promote” or that its members are “determined” to “reaffirm.” They do not create legal obligations or commitments, and they do not modify the general prohibition on the use of force. Had it been proposed in 1945 that these goals could trump the ban on war, the idea would undoubtedly have been soundly defeated by a large majority of the delegations, including all five of the Security Council's permanent-members-to-be.
A number of other international treaties subsequent to the Charter may also be relevant to this question. The Genocide Convention (1948), for instance, is sometimes understood to encourage or permit intervention against genocidal regimes, based on its Article I, which states: “The Contracting Parties confirm that genocide, whether committed in time of peace or in time of war, is a crime under international law which they undertake to prevent and punish.” Here the controversy becomes whether this undertaking to prevent and punish is an authorization to use force across state boundaries or whether it refers only to the more limited set of measures described in the rest of the convention, such as prosecuting, punishing, or extraditing suspects found in one's territory. While the language of the convention can be interpreted more expansively, the limited view is the most defensible, not least because the expansive interpretation requires that a right to intervene be seen as implicit in the text. In the absence of an explicit recognition of such an important right, the more conservative reading of the law is probably appropriate.
The treaties that establish the Organization of American States (OAS) and the African Union (AU) also make possible the use of coercive collective action against their own member states, and so are sometimes read as legal pathways toward humanitarian intervention. The African Union's Constitutive Act creates a “right of the Union to intervene in a Member State pursuant to a decision of the [AU's] Assembly in respect of grave circumstances, namely: war crimes, genocide, and crimes against humanity” (Article 4(h)). This is a collective right of the Union, not an individual right of member states, and in that way it resembles the interventionary authority of the UN Security Council relative to UN member states. The AU's authority has not yet been enacted, but it does seem to establish a legal basis for humanitarian intervention among its member states. The OAS Charter does not go as far: it expressly forbids interference across borders (Article 19) while reaffirming that its members have abandoned aggressive war (Article 3(g)); but the organization is also committed to sustaining democratic governance in its members, and it has described democratic governance as inseparable from the respect for human rights.Footnote 19 There may be some possibility to combine these three elements into a right to intervene in defense of human rights or democracy, but this is doubtful. More likely, the OAS has the authority to pass judgment on the domestic governance and human rights of its members, but not to invade; its enforcement capacity is limited to suspending a misbehaving member from the organization.
For any of these treaties to modify the UN Charter's prohibition on intervention, it must overcome the further problem posed by Article 103. This clause governs conflicts between the Charter and other treaties, and it answers decisively in favor of the Charter: “In the event of a conflict between the obligations of the Members of the United Nations under the present Charter and their obligations under any other international agreement, their obligations under the present Charter shall prevail.” Thus, the Charter arrogates to itself the status of constitutional law in interstate relations and nullifies contradictory laws.Footnote 20 If Article 103 is read to apply to future treaties and not just those in existence when the Charter was signed, then it appears legally impossible for a later treaty on (for instance) human rights to ease the ban on military interventions.
To a legal formalist, it is therefore clear that existing treaty law from the Charter to the present day makes no room for a legal category of humanitarian intervention. The case for the illegality of humanitarian intervention rests on the plain language of existing treaties and emphasizes the clarity of the UN Charter, as well as its near constitutional status in international politics and its universal adoption. Together, these lead to the conclusion that the purpose behind the use of force (other than self-defense) is irrelevant in law, and the effort to respond to humanitarian emergencies in states that refuse to cooperate ends up confronting the same prohibition on interstate war that was meant to stop aggression. As Byers notes, “The UN Charter provides a clear answer to these questions: in the absence of an attack, the Security Council alone can act.”Footnote 21 One might follow this tradition and yet still argue in favor of a specific act of humanitarian intervention. In so doing, however, one must confront the fact that the act is illegal. The now classic example of this is the post hoc explanation that the NATO intervention in Kosovo was “illegal but legitimate.”Footnote 22 Brownlie considers this to grant “a waiver of the illegality” of the act, and he opposes the claim that it provides any evidence of a change in the law itself. Thomas Franck agrees on its illegality but maintains that international justice is better served by sometimes breaking the law rather than respecting it, and that Kosovo/NATO is one such case.Footnote 23 This is a provocative position since it suggests that the idea of the rule of law is not as absolute as is usually maintained; other values might be more important than rule following.
Three Cases for Legality
To stop at the black-letter law on the use of force requires that we ignore developments in the language and practice of intervention. This may be a mistake since these changes may arguably have created a category of lawful war that encompasses humanitarianism. This is especially compelling with respect to state practice and conceptual development after 1990 or so. The case for legality rests on claims about changes in the law as a result of state behavior in the Charter era, and so is useful for examining the dynamic relationship between international law and state practice.
Two forces in international politics have repeatedly pressed humanitarian intervention onto the international legal agenda despite Article 2(4): the extension of the ideology of cosmopolitanism and human rights, which provides a moral imperative to respond to outrages against people regardless of their citizenship; and the strategic manipulation by states who see in humanitarianism a useful instrument to justify their military interventions. The second may well presume the first, since the language of humanitarianism would not be a useful tool if it did not have political resonance with deeply held beliefs about justice and obligation. But the second is also an independent force, reflecting the incentives that many actors see in adopting the language of humanitarian rescue. This contributes to its development and persistence. Both forces keep the language of humanitarianism alive in the legal discourse of states and activists, and therefore may be propelling developments in the law, though they are very different drivers for the concept and they apply, reinforce, and change it in different ways.
The case for legality can be made using three distinct arguments. All three make their case by joining together an interpretation of recent state practice with a theory of international law, but they draw on different interpretations of practice and lead to distinct implications. The first suggests that the ban on war in Article 2(4) has lost its legal force by being repeatedly violated by states in practice. There is therefore no operative international law left in that article. The second suggests that the normative environment of world politics has changed such that the rule of nonintervention has receded in the face of the progress of a norm of humanitarianism. These normative changes, it is claimed, have driven consequent changes in the formal laws and made lawful what was formerly unlawful. Finally, it is sometimes argued that the two concepts of sovereignty and humanitarian intervention are in fact complementary rather than contradictory, in the sense that sovereignty is conditional on a government respecting the obligation to protect its own people. This view argues that humanitarian intervention is lawful because the legal protections for sovereign states cease to exist if the state is engaged in the worst kinds of abuses of its citizens.
1. Desuetude and Article 2(4). The idea that Article 2(4) has lost its power due to repeated violation rests on an empirical claim about the frequency of violation and a separate conceptual claim about the legal effects of those violations. The two claims are independent of each other and each involves its own controversies. The empirical record regarding compliance with Article 2(4) was summarized by Thomas Franck in the following terms: states have “violated it, ignored it, run roughshod over it, and explained it away . . . [they] have succumbed to the temptation to settle a score, to end a dispute or to pursue their national interest through the use of force,” precisely in contradiction to the rule.Footnote 24 Franck, writing in 1970, used the twenty-five-year history of post-Charter uses of force as evidence that the war practice of states had not changed much from their pre-Charter practice, and concluded: “The prohibition against the use of force in relations between states has been eroded beyond recognition.”Footnote 25 Michael Glennon picked up the argument in 2001. Finding nothing in more recent history to temper Franck's pessimistic claim, he too concluded that “the upshot is that the Charter's use-of-force regime has all but collapsed. This includes, most prominently, the restraints of the general rule banning use of force among states, set out in Article 2(4).”Footnote 26
The conceptual claim is that the legal force of Article 2(4) has been erased by this history of rule violation. It is commonly said in scholarship on international law and international politics that rules lose their force if they are frequently violated. Glennon's claim adds a formal legal element to this idea: he says “international ‘rules’ concerning the use of force are no longer regarded as obligatory by states.”Footnote 27 That is, they have lost the quality that formerly gave them their legally binding character. As a formal legal process, the idea that law fails as law if it is routinely bypassed is common in domestic and international legal systems, and is known as desuetude. This is the concept that allows some outmoded laws to remain on the books despite relevant and major changes in sensibilities. In such instances, courts often refuse to enforce laws that they judge to have become irrelevant and unusable.Footnote 28 In international law, the concept is endorsed in the Vienna Convention on the Law of Treaties (1969) as one reason why a treaty might lose its force, and has appeared from time to time in opinions of the International Court of Justice, including in the Nuclear Tests case and the Aegean Sea Continental Shelf case.
If this applies in the case of Article 2(4), then the use of force by states is no longer regulated by the Charter and it is conceptually impossible for a state to be in violation of the rules. Glennon uses this line of reasoning to conclude that the United States was unconstrained in its military interventions in Afghanistan, Iraq, and elsewhere after 2001. The logic is equally applicable to uses of force for humanitarian purposes.Footnote 29 If state practice has caused the legal forms of war to become unlimited, then humanitarian intervention is no longer illegal for the same reason that all other intervention is no longer illegal: the laws have ceased to regulate it. Indeed, it follows by implication that aggression itself is also once again legal, and we have returned to the pre-1945 state of affairs, though no state has yet used this argument to justify its use of force.
2. Humanitarian intervention as norms-into-law. A very different mode of argument also maintains that the progressive development of international law now accepts humanitarian intervention as legal. This position affirms the legal force of Article 2(4), which the previous view denies, but argues that its scope has shifted as a consequence of recent practice. Like the claim on desuetude, this argument rests on an interpretation of practice joined with a theory of international law, but the practice in question comes from the statements and justifications made by states and others arguing for the legality of humanitarian intervention. Glennon and Franck reach their conclusion by arguing that these statements are something like self-serving cheap talk, but this competing view takes them seriously as evidence that states desire that humanitarian intervention be legalized.
The key element in this argument is the claim that the law has changed through the twin mechanisms of the power of norms and the power of state practice. These powers work together to force a reinterpretation of Article 2(4) by asking that it be understood in light of “emerging normative ideas.”Footnote 30 The Charter is thereby made subordinate to the normative and political environment in which it rests. It is not enough in this view to point out that a norm of humanitarian intervention exists; it must also be the case that the law has changed as a result.
Thomas Weiss and Ramesh Thakur, among others, make this case by invoking recent innovations in state practice.Footnote 31 These include instances in which the idea of humanitarian intervention was used by states to justify their use of force, statements by governments and others, and a reading of legal theory that shows historical strands of the concept. They point out that in 1998, Kofi Annan made the first of many claims to the effect that “state frontiers . . . should no longer be seen as watertight protection for war criminals and mass murderers.”Footnote 32 This was institutionalized further through the concept of the Responsibility to Protect (RtoP), and reinforced by Annan's successor, Ban Ki-moon, on many occasions. The World Summit in 2005 included an affirmation by all states in the General Assembly of their “willingness to take timely and decisive collective action” for humanitarian purposes (though only with Security Council approval).Footnote 33 The Security Council explicitly endorsed the concept in Resolution 1674 in 2006, and has applied it with varying degrees of ambiguity in relation to Darfur (Resolution 1706), Somalia (1814), Libya (1973), and elsewhere.Footnote 34 The key cases of state-led humanitarian intervention include Kosovo in 1999, where NATO used humanitarian rescue to justify its bombing of Yugoslavia, and the no-fly zones of Iraq in the 1990s, designed to protect certain civilian populations from the Iraqi government. These are among the cases most often cited as evidence for the norms-into-law argument about the legality of humanitarian intervention. Each is contestable, of course, but they may form a pattern of official practice that modifies the legal regime on the use of force.
In this view, the available evidence that states are disregarding their obligations under Article 2(4) must be understood differently than Michael Glennon would have it. For Glennon (and Franck) these are violations of Article 2(4) and they suggest that states are ungoverned by international law in their use of force. For the norms-into-law approach, they are instead evidence that states have reconstituted their legal obligations around a new legal principle. According to this approach, such an intervention may not count as a “violation” at all. It is instead constructive noncompliance, which signals that humanitarianism is becoming legal even while Article 2(4) remains in place. Seeing international law as fluid in this way turns it into “social practice” rather than a set of fixed and external standards against which conduct can be measured.Footnote 35
Empirically, the activist case for legalized humanitarian intervention encounters two limits. First, much of the formal support for RtoP, including the 2005 World Summit, is for the relatively easy case of intervention approved by the Security Council. This minimizes the legal innovation as well as the practical scope of the concept, because it essentially repeats what is already accepted about the Council's legal powers. More novel is the argument that humanitarianism can be legal without Council consent. Second, as Paul Williams and Alex Bellamy point out, once we include state practice as relevant to considering the law, we must also recognize the many cases where states that support RtoP in principle failed to carry it out.Footnote 36 It is unclear how to interpret these failures, since the norms-into-law case argues only the permissive case, that humanitarianism can be legal, not that it is consistently applied. And yet the practice of failing to intervene is presumably as politically powerful as the practice of intervention, and must somehow be accommodated into the argument.
3. Contingent sovereignty. The third path to legality suggests that state sovereignty is contingent on a government providing a basic level of human rights protection to its people. This is often bundled together in a broader narrative about the responsibility to protect, but it is conceptually and legally distinct. The idea of contingent sovereignty suggests that statehood itself is legally dependent on acceptable government behavior, such that failure of a government to meet certain minimum standards nullifies its claim to noninterference. This may or may not involve a “responsibility” for outside states to intervene; but is does mean that the government in question has lost the protection entailed by sovereign statehood such that any invasion no longer involves transgressing a legal or physical boundary of nonintervention. This was neatly expressed early in the recent revolution against Muammar Qaddafi when the Italian government declared that because the Libyan state “no longer exists” its treaties with Italy ceased to have any legal content.Footnote 37 The idea of contingent sovereignty is in the end an argument about the role of international laws on sovereignty, rather than on the law on the use of force; it describes the moment at which the protections of sovereignty vanish from within. Once this happens, intervention does not count as a use of force against another state.
Allen Buchanan makes a related argument with respect to international recognition of secessionist movements, suggesting that the international community should assess the human rights performance of claimants to “national self-determination,” so that morally defensible behavior toward one's citizens will become a necessary condition to statehood.Footnote 38 This proposal would link the institutional legal framework of sovereignty to the practice of respecting human rights and make the former conditional upon the latter. Many writers believe such a link has already been made, or has always been implicit, in the law of state sovereignty. Fernando Tesón, for example, has said that “to the extent that state sovereignty is a value, it is an instrumental not an intrinsic value. Sovereignty serves valuable human ends; and those who grossly assault them should not be allowed to shield themselves behind the sovereignty principle.”Footnote 39 As early as 1992, in “An Agenda for Peace,” the UN secretary-general said “the time of absolute and exclusive sovereignty, however, has passed; its theory was never matched by reality.”Footnote 40 A decade later, Gareth Evans and Mohamed Sahnoun wrote of RtoP that “even the strongest supporters of state sovereignty will admit today that no state holds unlimited power to do what it wants to its own people. It is now commonly acknowledged that sovereignty implies a dual responsibility. . . . In international human rights covenants, in UN practice, and in state practice itself, sovereignty is now understood as embracing this dual responsibility.”Footnote 41 Where humanitarian intervention is necessary, they suggest, sovereignty properly understood no longer exists.
All three of these views conclude that under some circumstances humanitarian intervention can be legal. The first two cases suggest that the laws on the use of force have changed since 1945 due to the behavior of states. The third suggests that the legal institution of sovereignty encompasses the possibility of legal intervention because sovereignty itself disappears at some extreme of government misbehavior. They all reconcile the tensions in the law in a way that accommodates the innovation of intervention with the traditional structures of international law and international politics: Security Council approval is not necessary; the rule of law is preserved; humanitarian and legal impulses point in the same direction.
Between Legality and Illegality
The difference between these two sets of views rests on how one understands the relation of international law to state behavior. The argument for illegality is constructed by reading the UN Charter and, after finding the answer to the question there in black-letter law, it reaches its conclusion based on the priority of state consent and treaty law over other legal resources (such as practice, custom, and so on). It suggests that state practices that contradict a piece of treaty law should be coded as “noncompliance,” and that the political, moral, or transformational motives behind it are irrelevant. The law stands, independent of state behavior, even accepting that rule violation will sometimes occur. The cases for legality turn this around. They rest on the view that sustained patterns of state behavior in opposition to the rules have creative effects in international law, such that if there exists a sustained pattern of humanitarian intervention, then legalization may be taking place. This presumes that international law is a product of interstate interactions in settings beyond formal treaty negotiations. Rhetoric, recent behavior, and the apparent intentions of states are all resources for interpreting how states understand their obligations, and thus for learning what the rules are. Evidence of noncompliance with Article 2(4) is in this view a sign that states are in the process of changing the rules. Sufficient movement in this direction can come to constitute a new understanding of the law, equally binding as the ex ante law. This is the standard account of customary international law; and the intuition behind much of the writing on humanitarianism is that it also applies to treaty law in the Charter. This view puts practice conceptually ahead of the text of the treaty, and it puts the agency of states ahead of the constraining quality of external rules. It obviously carries its own dilemma: if legality is a function of practice rather than treaties, then in what ways are treaties constraining or even relevant?
These are deep differences, and they cannot be bridged by reference to the laws themselves. Rather, they require answering the eternal questions in the philosophy of international law about sources, foundations, and interpretation—and this is unlikely to happen. This situation suggests we should admit that international law contains both positions at once and that humanitarian intervention can be made to seem legal or illegal depending on one's needs. This allows us to see that both traditions have honorable pedigrees in international law. However, it also requires that we rethink what we thought we knew about the role and power of international law.
Since the contents of international law can be read to support either position, it is natural that a government will take the view that what it thinks should be done in any particular case is acceptable under international law—and will point to what it opposes as a violation of the law. Deciding in favor of one of these conclusions requires selecting the subset of evidence that supports one's favored view. This freedom to choose among interpretations of the law gives rise to a sea of self-serving claims, and to unending academic debate. Each effort to champion one side produces a predictable response from the other, and the debate has less to do with the merits of one legal interpretation than with the political needs of those making the arguments.
This is the strategic manipulation of international law, which is often taken as the opposite of the rule of law and as therefore negative. This is how Brian Tamanaha sees it in his critique of the “political” uses of law in the American context, and it is also how Michael Glennon sees it in the international context.Footnote 42 However, the humanitarian intervention case helps to show that this is wrong. Tamanaha argues against seeing the law as an instrument with which to pursue political goals. He suggests that this instrumental attitude toward the law is a threat to society because it leads to “a Hobbesian conflict of all against all carried on within and through the legal order . . . [where] law will thus generate disputes as much as resolve them. Even when one side prevails, victory will mark only a momentary respite before the battle is resumed.” The preferable alternative, he says, is to return to the model of “a few centuries ago” when “law was understood to possess a necessary content and integrity that was, in some ways, given or predetermined.”Footnote 43 This noninstrumentalist attitude to law reflects the compliance model in international law, where treaties and state consent provide the integrity of predetermined rules, which stay constant despite the interpretive manipulations and rule-breaking acts of self-serving states.
What Tamanaha identifies as a problem for domestic law is in fact the normal condition of international law. The strategic manipulation of law is inherent in international law, at least with respect to the laws on the use of force and others (such as torture) where there is no international judicial body to settle disputes over the meaning of compliance. International law on self-defense, preemption, torture, and humanitarian intervention does indeed generate, as much as resolve, disputes. States invoke international law in a variety of settings to explain and justify their behavior, and to criticize and embarrass their opponents, and so the instrumental use of law is inseparable from the law itself. The political use of international law is not an aberration or a misuse of the law; it is the normal and inevitable result of striving for rules-based international politics.
Of course, it is widely accepted that international legal obligations must be interpreted in part by reference to state practice. My argument is not that this is novel, nor that it is a problem that needs to be solved. Rather, I seek to show that the ease with which we use practice to understand the content of international law works against the equally common presumption that compliance with international law has a consistent meaning.
State practice has a productive effect on the content of the law. This is evident in the progress of the concept of humanitarian intervention, which has arisen as a legal category out of the practice of states invoking the rules on the use of force in certain ways. The content of these rules is in part a function of how they have been used in the past, especially but not only the recent past. As states began to claim that Article 2(4) should not be understood as banning wars with humanitarian motives, the certainty over the meaning of the Charter eroded. Breaking international law is intrinsically linked to making international law, and both are subsets of the broad category of “using” international law. This is true for both treaties and custom, and not solely in the traditional hierarchy by which treaty trumps custom—the humanitarian intervention case shows that many states and scholars are willing to have state practice trump treaty law.
Ultimately the attempt to organize international law scholarship around the question of compliance is misplaced, at least for a significant subset of international rules. There is a growing literature that seeks to link political science and law around the empirical measurement of compliance, asking what qualities in a piece of law contribute to higher rates of compliance, or what kinds of domestic institutions correlate with higher rates of compliance, or how compliance relates to the decision to join a treaty.Footnote 44 These studies adopt an empirically oriented compliance model to studying international law, in which the causes and effects of compliance and noncompliance are the focus. Compliance is also sometimes used to try to distinguish constructivist from rationalist hypotheses about state behavior. For instance, Judith Kelley asks whether states comply with their obligations to the International Criminal Court out of an interest in material gain or out of a normative commitment, seeing the former as rationalist and the latter as normative or constructivist.Footnote 45 In all this work, it is the reasons for and effects of compliance and noncompliance that are under investigation.
The “compliance model” requires that we be able to differentiate between behavior that is compliant and behavior that is not, and in the case of humanitarian intervention this is clearly not possible. Interpretive challenges here mean that the definition of compliance is itself contested, and disputes over the meaning of the law are best understood as proxies for fights over the underlying substance of the case in question. Many areas of international law have this quality, where the parties insist on their own claims to compliance and provide legal resources that support them. Where we cannot differentiate compliance from noncompliance, the law's effect on behavior must be measured in some currency other than the rate of rule following. Robert Howse and Ruti Teitel have recently made this point, and suggest that the contribution of international law is in changing the terms and shape of the interstate bargaining that takes place in and around the rules.Footnote 46 For them, the law is a resource for states rather than a standard that distinguishes between lawful and unlawful behavior. This is supported by the humanitarian case presented here.
Despite all these ambiguities in international law, the idea of the rule of law remains powerful in international politics. States remain convinced that they should comply with rather than violate the law, and in the humanitarianism debates all sides generally represent themselves as being compliant with the rules.Footnote 47 The idea that international relations does or should take place in a rule-governed context is widely shared. This is a fundamental premise of international law both as a practice and as a scholarly field. The power of state consent is that it marks an explicit moment at which states take on commitments, and compliance is expected thereafter. While there are disagreements about whether a legal obligation provides an independent reason for compliance distinct from any other underlying reasons for action (that is, for reputation, for instrumental gain, or other logics),Footnote 48 there is very little dissent from the idea that states should comply with these obligations. Few commentators suggest that states can or should ignore their legal obligations. Those who argue against a particular obligation almost always do so in the language of compliance with some other obligation.
Conclusion
Wars begun in the pursuit of humanitarian rescue are now seen as different from wars fought for other purposes. They are now legally, politically, and conceptually separate from wars of conquest and wars of national security, even as the category of humanitarian intervention remains fiercely contested in practice. Contemporary international law can be read as either allowing or forbidding international humanitarian intervention, and the legal uncertainty around humanitarian intervention is fundamental and irresolvable. Contradictory and plausible interpretations about the legality of any act of intervention exist simultaneously, and neither can be eliminated. This does not mean that the law is unimportant; there are evident costs and benefits to states in being seen as following the rules. It means instead that law and law following should be seen as resources in the hands of states and others, deployed to influence the political context of their actions.
The debate over the legality of humanitarian intervention raises deep questions about international law. Can the statements of leaders modify the obligations contained in treaties? If states contradict established international law, does this change the law or is it a simple case of noncompliance (or can it be both)? Does the practice of humanitarian intervention (if it exists) sustain the legality of humanitarian intervention? There is no consensus over the legality of intervention, in part because there is no consensus over the sources of international law more generally. The intervention problem is inseparable from questions that have been at the heart of international law for centuries, and that we cannot expect to be answered in order to reconcile the different views on humanitarian intervention. The legality of humanitarianism is therefore contingent on one's theory of how law works and changes. The law may well be incoherent, and it may be unable to distinguish between compliance and noncompliance, but it remains politically powerful and therefore important. The challenge for scholars is to explain how it is that the commitment to the rule of law coexists with this fundamental ambiguity.
