1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

The wars in Iraq and Afghanistan have put front and center the problem of dealing with non-uniformed combatants. They have also made central deep questions of the legitimacy of resorting to martial violence, and responsibility for picking up the pieces thereafter. I argue here that the special problem of non-uniformed combatants and the general problem of justifying war are profoundly linked. War, I shall argue, is but one form of a more general species: collective violence. Collective violence poses a particular set of challenges to the application of moral principles. In what follows, I identify a conflict between two themes in our response to collective violence. I call these themes of inculpation and exculpation. I illustrate these themes with three stories derived from actual events.

Crime Story1

Smith and Daniels approach Taylor. Daniels tells Taylor that Jax Liquor would be a good target for a robbery. All they need is a car and getaway driver. If Taylor will sit outside the liquor store till Smith and Daniels come out, he’ll get a third of the haul. Taylor is in.

Smith and Daniels come into the store, Smith waving his gun, both shouting, demanding the money. But the situation fails to unfold as planned. Linda West, the owner's wife, is working in the back of the store when she sees the men. Fearing that Smith is about to shoot, she grabs the gun in her pocket, and shoots and kills Smith, then wounds Daniels as he runs to escape in Taylor's car.

Taylor and Daniels are later arrested. Taylor, although he did nothing more in preparation for the robbery than sit in his car, is charged with both robbery and murder, for Smith's death. Though Ms. West actually shot Smith, the death is treated as causally flowing from Smith's own and Daniels’ frightening gun-waving during the robbery, and as a species of murder because it manifests an extreme indifference to human life. By the logic of accomplice liability, according to which any member of a criminal group is liable for any reasonably foreseeable acts done in furtherance of the group's common design by any other member, Taylor is also responsible for Smith's death. The result is that Taylor may be convicted of a murder he did not commit, or even cause.

War Story

Imperioland has invaded its small but oil-rich neighbor, Petrostan, in order to seize its oil wells. Sergeant Blue, of Imperioland's volunteer Army, is aware that world opinion holds Imperioland's invasion to be a flagrant violation of international law, but he follows the judgment of his political leaders. Blue, however, intends to fight the war in full compliance with the international law of combat, known as jus in bello or, more currently, as International Humanitarian Law (or IHL). IHL is independent of the legality of the conflict itself (the rules governing which are known as jus ad bellum). Among its principal requirements are that soldiers proportion the violence they deploy to military necessity, discriminate between combatants and non-combatants (a category including civilians and wounded and surrendered soldiers), and respect the life and well-being of anyone not currently a threat, including surrendered or injured enemy combatants.

Blue's squad is ordered to capture an engineering building at one of the refineries. Blue enters the building. He shoots and kills the Petrostan soldiers on guard. His mission appears successful. And because Blue killed only combatants, it was unquestionably consistent with IHL.

Suddenly a company of Petrostan's soldiers arrive and capture Blue. He is sent to a detention camp, called before a military tribunal, and charged under Petrostan's domestic criminal law with murder, for the intentional killing of the sentry, not in defense of himself or others. He is sentenced to death.

Before the sentence is carried out, a member of Petrostan's foreign ministry arrives. Petrostan (like Imperioland) is a Geneva Convention signatory, and the minister is waving a copy of the Third Geneva Convention, which deals with combatants taken prisoner (often abbreviated as GPW, for Geneva–Prisoners of War). According to the GPW, Blue, as a regular, uniformed soldier, must be treated as a “privileged combatant” and can only be held as a Prisoner of War. This means he cannot be punished for his killing (assuming it did not breach the laws of combat). He may be held in captivity only until the cessation of hostilities.2 Though Blue kills without justification, as a soldier he is impunible.

Rebel Story

The tide turns in the invasion, and Imperioland's troops begin to rout Petrostan's army. Remaining members of the army doff their uniforms, move to the back country, and become a partisan resistance. They are joined in their efforts by Petrostan citizens, and foreigners from the region who infiltrate the border and join the resistance.

Gray is a foreigner who wants to join the partisans. She too crosses the border, affiliates with a partisan unit, receives weapons training, and is sent out to fight.3 The partisans’ resistance is classic guerrilla strategy: they hide among the population, and seek low-intensity engagements. To paraphrase Raymond Aron, they believe they will win so long as they do not lose their ability to inflict losses, and that Imperioland will lose so long as it does not wipe them out.4 Their goal is to protect and restore the political institutions of Petrostan, as well as to defend a religious and cultural tradition they reasonably see as under threat by the occupation. The partisans strike only at military targets, and are as scrupulous as Sgt. Blue about observing the international law of combat. But, unlike Blue, they do not wear uniforms or otherwise reveal their identities as combatants, because it would be certain death or capture. Only when they draw their weapons in battle do they reveal themselves as combatants.

Gray is preparing for an assault when her house is swarmed by Imperioland soldiers. She is armed but not uniformed. Imperioland has ratified the GPW, which accords POW status only to combatants who wear uniforms or otherwise bear “a fixed distinctive sign recognizable at a distance.” But it has not ratified the additional Protocol I to the Conventions, which broadens combatant status to non-uniformed, “liberation”-seeking members of the armed forces of a party to the conflict, and who bear their arms openly while engaging in or preparing for military operations.5

Like Blue, Gray is brought before a tribunal and charged with conspiracy to commit murder and sabotage. Her claim that she is a combatant entitled to POW status is dismissed. While she is spared from the death penalty Imperioland's regulations permit, she is sentenced to indefinite confinement at an Imperioland prison. (The Imperioland army fears, reasonably enough, that Gray when released will rejoin the fight.)

Two Themes in the Key of Collective Violence

  1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

These stories reflect the differential treatment of collective violence in law and ethics. I put this in terms of two conflicting themes. The first is the theme of complicity, and every jurisdiction in the world plays a variant of it. Ordinarily moral responsibility and criminal liability attach to an agent only on condition that the agent has performed a wrongful act, perhaps producing a wrongful result. This is a principle of individual culpability, and requirements (in Anglo-American terms) of the existence of a culpably done criminal act and proximate causation of a result undergird and limit the attribution of wrongs to individuals. Complicity doctrine, however, attaches liability through a different route. Even though individuals on their own might have done nothing wrong, they can be held responsible for someone else's wrongful act, if they are members of a group whose other members do wrong in furtherance of a joint criminal plan. To put the point more strongly yet, so long as any member of a group with a criminal project does a foreseeable wrong, each member of that group bears responsibility for the wrong.

Take Taylor, from Crime Story: driving a car to, and sitting in front of, a liquor store one hopes to rob is not itself wrongful. Those acts, on their own, would probably not support a conviction of attempted robbery in most jurisdictions, as they fall short of a “substantial step” towards the crime's completion.6 Taylor's liability rests not on what he actually does, but on a combination of what he intends to do—participate in an armed robbery—and what he might expect his fellow participants to do— instigate a shooting. His complicity in the group robbery renders him liable for another's killing. This I call the theme of collective inculpation.

A contrasting theme, of collective exculpation, runs through the law of war. The function of the law regulating the conduct of war (IHL) is to demarcate a zone of impunible violence: killing, maiming, and property destruction. The boundaries of this zone are set chiefly by the rules of proportionality and discrimination mentioned above; but the central presupposition of the zone is the collective, political character of the violence: these acts are only impunible when committed by a member of the armed forces of a state or insurgent party to the conflict (provided they are otherwise in compliance with IHL).

Sergeant Blue kills by his own hand and without justification, and so would be guilty of murder if he were simply trying to rob the refinery. But because he is a member of Imperioland's army, no liability attaches to him personally. Even if Blue fires the only shot in the war, he bears no liability for the killing. Moreover, the injustice of his army's war is irrelevant. Blue's permission to kill depends on the fact that he is part of a certain sort of group collectively intent on violence. This ought to be shocking but it is all too familiar: participants in normalized mass killing, territorial occupation, and political transformation enjoy permission to do together what would be infamous crimes if done separately.

Non-uniformed fighters like Gray mix both themes. Is rebel Gray more like Taylor or more like Blue? Should she be inculpated or exculpated? Gray's cause, Petrostan independence, is presumptively just, unlike Blue’s. But, Gray, unlike Blue, may be criminally liable, and executed or detained indefinitely. Her legal status depends on a two-step analysis: first, Gray's acts are removed from the context of a collective partisan resistance and she is treated as an individual with criminal intent. Next her collective status is reasserted in the complicity or conspiracy charge. Like Taylor, she is liable for rebel-caused deaths whether or not she fires a shot.

My subject is the contrast between the themes of collective inculpation and collective exculpation, and the tension that arises when the two themes encounter each other in the treatment of irregular, usually non-uniformed combatants.7 These are individuals engaged in the ordinary business of war who, if they were part of conventional military units, would enjoy impunity so long as they proportion their violence to military necessity, and discriminate between civilians and combatants. The case of non-uniformed, irregular fighters is of course an especially current practical challenge for the law of war. It also brings into the open the question why certain forms of collective action privilege violence, while others serve as the basis for punishing it.

The European partisans of World War II fighting Nazi occupation are exemplars of this category, including the storied Maquis of France. Others include the anti-colonial rebels of the developing world. The anti-colonialist movements were a major motivation for the 1977 Protocols amending the Geneva Convention provisions; Article 44 of the First Protocol (hereafter PI 44) specifically deals with question of irregular combatancy. PI 44 permits violence by insurgents and partisans who conceal their status generally but engage openly in combat. Protocol I was widely ratified, thus binding its signatories, who do not include the United States but do include most other major powers.8 Modern examples of irregular fighters, to whom the application of PI 44 is controversial, include the Taliban and Al Qaeda fighters in Afghanistan, the Fedayeen and Baathist resistants in Iraq, the posses of Afghan and Somali warlords, and some of the Colombian anti-government rebels, in whose disputes U.S. forces are entangled. More pointedly, so may be U.S. Special Forces soldiers and CIA field operatives, who typically serve out of uniform and without clear insignias of their national affiliation. (Recall the photos during the Afghanistan war of U.S. Special Forces riding their horses in the company of the Northern Alliance.9)

The category of irregular combatants is not new but its instantiations have increased (perhaps because of greater U.S. military adventurism). As has been widely discussed, this is a consequence of three principal “developments” in modern violence.10 First, state military conflict today rarely occurs in the form of major battles between armies, but increasingly through the tactics of “asymmetrical” warfare, including guerrilla raids, hiding among either one's own or one's enemies’ populations, infiltration of enemy lines, sabotage, and joint operations with collaborating civilians. Second, recent conflicts are increasingly transnational in character, where the transnational element includes collaborations between intelligence units of one nation and military units of another, or involves foreign volunteers linked by ideological or religious affiliations. Again, this is not new—witness the Spanish Civil War—but it is resurgent with militant Islam. Relatedly, some recent conflicts have been neither internal to a state nor transnational, in that they have taken place in political conditions where no state exists because power is too fragmented. Somalia is a prime example.

The third development is the renascent phenomenon of war through mercenary proxies, which predated the modern era of war, subsided during the consolidation of state power, emerged again during decolonization, and then subsided once more. It is now again on the rise through the distinctly post-modern phenomenon of the “corporate warriors,” who provide outsourced logistical and “tactical” (read lethal) support to everyone from the U.S. Army to the UN to Sierra Leone to the petroleum industry.11 Modern combatants look increasingly unlike the army regulars around whom the Geneva conventions were drafted.

The results of these developments are troubling. It is, at the least, conceptually anomalous that greater numbers of combatants in modern war fall outside the regime crafted to control war's violence. It poses a practical problem, in that if combatants lack impunity for engaging in violence bounded by the norms of proportionality and discrimination, they have no incentive to observe these bounds. And it is a legal problem, in that we lack criteria to assess the legitimacy of the treatment of the large number of irregulars captured on the battlefield and held indefinitely by occupying powers. As ever more warfare involves stipulatively unprivileged combatants, the normative systems controlling war become more and more strained. If lawlessness is a problem, an even deeper problem is normlessness.

Designing Norms for New Wars

  1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

What norms should we adopt? What difference should uniforms make? I look at and reject some traditional answers to the problem, including answers generated by pre-modern conceptions of sovereignty, and by straightforward consequentialist reasoning. Instead I turn to a modification of a tradition inaugurated by Rousseau, who conceived political authority as resting in a special relationship among individuals. When individuals’ wills are linked together in politics, this affects the normative valence of what they do individually as part of that politics, even to the point of rendering impunible what would otherwise be criminal. The salient Western form of these political relationships is democracy, which I understand here as involving some form of majoritarian decision making, coupled with a universal franchise. But by “political” I shall mean any forms of social action oriented around state or institutional formation, where power may in some sense be seen to rest at the level of individual voluntary commitment to the shared project. (Thus, I mean to contrast “political” relationships with authoritarian, fear-motivated hierarchical relationships.) A consequence of my conception of political authority is that permission to engage in collective violence turns on combatants’ attitudes and relations to one another, not any external sign of their obedience, including wearing a uniform. Put directly, citizen–soldiers enjoy combat privileges because they enjoy the political status of citizens, not because they wear the uniform of a soldier.

In actual policy terms, this article defends a regime like that of the First Protocol (PI 44), which permits combat by non-uniformed combatants fighting for “liberation” or “self-determination,” a paradigmatic political category of collective violence. I depart from that regime in one important respect, however. PI 44, as a matter of positive law, is fully consistent with the separation of jus ad bellum from jus in bello. My argument opens conceptual space for denying the privilege to some otherwise lawful combatants waging clearly unjust wars, a position considered and rejected by the drafters of the First Protocol. For a number of reasons, both practical and conceptual, this logical space may be closed for all likely cases. But principled reflection demands that we understand the deep links between responsibility for war and the privileges of warfare, rather than simply assert their separateness. The very idea of an ethical regime of war generates paradoxes, which I now consider.12

The Paradoxes of War

  1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

The first paradox is substantive: even if a state is illegally engaged in war (in violation of the UN Charter now, or of just war principles in an earlier day), its forces enjoy a right to wound and kill enemy combatants subject to IHL's norms of proportionality and discrimination.13 This is puzzling: domestically, no one could defend a murder on the grounds that he had shown special delicacy, à la Hannibal Lecter, in the manner of his killing. Means are normatively inert. Yet it is a commonplace that the rules of IHL are independent of the justice of the war itself.14 This commonplace obscures a deep puzzle: how can there be permissibly violent means of pursuing impermissible ends? The very premise of the normative independence of IHL brings into question the nature of its justification. This is the paradox of permitting the impermissible.

A variant of this paradox has frequently provoked puzzlement among newcomers to the law of war: how can there be any significant distinctions within the field of killing? If a war is unjust, then any killings done in its prosecution are unjust, even if they are permissible. It is therefore hard to see how a normative regime can determine that some of these unjust killings (for instance, killings not using dum-dum bullets, or killings by uniformed combatants) are categorically better than others, such that they are permitted and the others banned. Even in a just war, killing is a terrible thing, permitted out of necessity rather than utility. Once necessity is in play, one might think, distinctions among necessary killings seem somehow beside the point. In domestic criminal law while we sometimes grade punishment in relation to the manner of killing, reserving the most severe sanctions for the most heinous forms of killing, we do not distinguish among the varieties of justified killings. But international law promulgates precisely such distinctions.

An instrumental answer comes forth immediately: restrictions on methods and targets of killing in war reduce the suffering of combatants, risks to non-combatants, and the costs to states, and hence are justified by their good consequences. The permission to kill within the bounds of these restrictions is the bribe paid to combatants to induce their compliance with them. I mention this justification now to acknowledge it, but, for reasons I elaborate below, I do not believe it accounts for IHL's normative authority, and I think it particularly fails to justify a central feature of it, the categorical quality of its rules.

There is a related historical point. Many of the customary rules of IHL come from the chivalric tradition, particularly rules regarding the treatment of those hors de combat. The rules thus have their ground in a conception of warrior virtue; and again an instrumental account seems inadequate to the underlying ethical view on which they draw.15 This point is hardly decisive, since a revisionary account of our intuitions might in fact provide the best justification of the norms these intuitions support (as, for example, Mill argued of utilitarianism). But it is a prima facie objection that an instrumental justification seems to “argue back” to a conclusion more certain than the path of argument itself.16 However one assesses the force of these considerations, then, we need a framework of principles within which those considerations can be deployed.

Without these broader puzzles in sight, the question of whether to grant battle privileges to the irregular combatant appears easier than it is: just a matter of estimating the marginal costs and benefits of additional suffering that a change in the rule would impose. We need a deeper solution.

Conceptual Sources of the Combat Privilege

  1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

We can identify three sources for the conceptual foundation for the privilege of uniformed combatants. The first source is the early modern conception of sovereignty itself, where the concept of the state was wholly identified with its ruler. This notion, theorized most radically by Jean Bodin's 1576 Six Books of the Commonwealth [Six livres de la République], was as much a logical and metaphysical claim as a prescription for political unity. According to Bodin, the very idea of political authority requires a distinction between the agent who exercises authority and the subject who receives it.17 The idea of an agent who was at the same time a subject, or, alternatively, a subject who was bound by laws he himself imposed, was for Bodin a logical impossibility.18 With a firm distinction in place between the state, embodied in its ruler, and its subjects, the moral qualities of the state cannot flow logically to its inhabitants. Just as the fact that the sovereign might incur a debt does not mean that a given peasant in his realm is also liable for that debt, so the fact that the sovereign was at war with another state would not mean that his subjects were at war with the other state. War could not be, in moral terms, a relation between the soldiers actually doing the fighting. They are merely the technology for resolving the interstate dispute.

The moral and metaphysical separation of state from subjects thus opens up a logical space for a distinct code of ethics for soldiers, an ethics independent of the legitimacy of their sovereigns’ dispute. The war is not about them, it is about their sovereign. Within the field of combat, there is room for codes of chivalry, especially with regard to the norms of respecting surrender and discriminating between civilians and soldiers. The permission to kill within these limits, under this theory, is not a deep justification of killing, in the sense that it does not justify the killing itself. Rather, the permission reflects the limited moral status of the soldier qua soldier, who was not expected to justify his role in the war before God or his conscience, but only his conduct in the war. Responsibility for the war itself belonged solely to the sovereign.

A further norm restricting the privilege to the uniformed makes sense in the context of this conception of sovereignty, although the regular wearing of uniforms post-dates Bodin considerably. While uniforms were hardly unknown before the modern period, they did not feature prominently (at least in Europe) as the garb of national militias until the seventeenth century, when Oliver Cromwell dressed his citizen army uniformly; and the trend came to a head with the elaborate uniforms of Frederick the Great.19 The systematic uniforming of armies in fact tracks the post-Westphalian establishment of a system of internally ordered, sovereign states. Disciplining the army and disciplining the nation-state go hand in hand.20 A norm that war should be between uniformed combatants simply mirrors the claim that war is a relation between states, not citizens. Because the basic relation of sovereign to subject is an external relation, on this conception—a matter of the power of the sovereign to compel obedience21—it follows that the relation of privileged combatant to sovereign would also be established through external mark. The uniform is, in effect, the stamp of ownership the sovereign puts on his army, and this stamp renders the external quality of what they do, namely killing others, attributable to the sovereign rather than to themselves.

The inadequacies of an account of the privilege grounded in Bodin-esque sovereignty need not be belabored: the separation between state and citizens it depends upon is not sustainable under conditions of mass, bottom-up politics. But a second and more resonant conceptual source of the privilege emerges from the rival conception of sovereignty that superseded Bodin's in modern, post-Enlightenment thought. This is the conception we take from Rousseau. Rousseau famously argued in the Social Contract that not only can a subject, collective or individual, give itself law, but that giving oneself law is a necessary condition of political freedom and legitimate authority. It follows from this, Rousseau thought, that a people is sovereign when and only when their individual agency, in the form of their wills, is linked in the structure he calls the “general will.” A people whose wills are so linked are committed to acting together in the interests of all, on the basis of a distribution of rights and responsibilities that guarantee their equal freedom. When this is so, a people produces

a moral and collective body made up of as many members as the assembly has voices, and which receives by this same act its unity, its common self [moi commun], its life and its will. The public person thus formed by the union of all the others formerly assumed the name City and now assumes that of Republic or of body politic, which its members call State when it is passive, Sovereign when active, Power when comparing it to similar bodies.22

The sovereign, on this conception, is dependent upon but not reducible to the individual citizens taken together. This is because the sovereign is a relation among wills, not a set of persons. The individual citizens retain their personal wills, notwithstanding their voluntary commitment of their rights to their collective sovereignty. Indeed, this retention of their personal wills is what explains the self-evident strains of committing oneself to even a just polity: the temptations to free-ride for personal benefit do not disappear merely because one acknowledges the force of the public interest. Thus sovereignty reflects an aspect of the citizens of a state, their public face in a sense. Their relations as members of the sovereign—or, better, as participants in the collective achievement of sovereignty—to themselves as private individuals is what enables Rousseau's response to Bodin as to how a sovereign can bind itself.23

So war, conceived as a relation between peoples linked constitutively as sovereigns, can still be distinguished from a relation between individuals per se. What would seem to follow from Rousseau's account is that in war, soldiers relate to one another as citizens rather than as individuals. Thus, an ethics of international relations, not an ethics of interpersonal relations, constrains their conduct.

Interestingly, this is not what Rousseau says. What he says instead is: “War is not then a relationship between one man and another, but a relationship between one State and another, in which individuals are enemies only by accident, not as men, nor even as citizens, but as soldiers; not as members of the fatherland, but as its defenders.”24 On its face, this is puzzling: why should men in war encounter each other only as soldiers and not as citizens? As with much of Rousseau's writing, answering this demands recognizing an imprecision forced by context. Rousseau's concern in the sentences above is to limit the power of victors by defining the scope of the relation of enmity. His specific task is to deny the traditional victor's right to enslave the vanquished. His argument must therefore be that, if war is between states, and if states consist of citizens (appropriately bound), and if soldiers confront each other as citizens (as well as soldiers), then in prosecuting a war against another state it is not sufficient simply to disarm its solders; one must further kill or enslave its citizens. To deny this line of reasoning, Rousseau must show that on the battlefield norms appropriate to the circumscribed role of the soldier, not the more expansive role of citizen, determine the range of permissible acts.

Rousseau has two arguments for doing so. The first argument is at work in his claim that

[t]he foreigner, whether he be a king, a private individual, or a people, who robs, kills, or detains subjects without declaring war on their prince, is not an enemy, he is a brigand. . . . Since the aim of war is the destruction of the enemy State, one has the right to kill its defenders as long as they bear arms; but as soon as they lay down their arms and surrender, they cease to be enemies or the enemy's instruments, and become simply men once more, and one no longer has a right over their life.25

A declaration of war is a special kind of collective act, reflecting the will of one sovereign to engage in hostilities with another. The collective aspect of a citizen's agency in the domestic sphere lies in his participation in forming a general will, constituting sovereignty. But on the battlefield, the collective aspect of his agency consists simply in fighting as part of a unit, that is, as a soldier. In the external relations of state to state in war only the potential for belligerency is significant to the citizen's normative identity. Once a citizen-soldier is disarmed, that external aspect of the citizen's identity is destroyed, he can no longer properly be considered an enemy of his victor. He is simply an individual, and there is no ground for the victor to claim any right to kill or enslave a private individual.

The second argument amplifies the first: “It is sometimes possible to kill the State without killing a single one of its members: and war confers no right that is not necessary.”26 Sovereigns formed by interdependent citizen wills are “killed” when the relation among those wills is broken; and that relation can be broken by isolating an individual citizen-soldier, not just by killing him. Even in authoritarian states, where sovereignty is vested in an individual prince, killing the soldier does not kill the state. More generally, so long as sovereignty is understood as an abstract property of an individual or individuals, killing disarmed soldiers is neither a necessary nor a sufficient means for vanquishing a state. This argument too, then, rests on the special nature of political organizations.

The logic of this position supports both the permissibility of killing and its subordination to a non-partisan system of rules. While the citizen-soldier is at war with other citizens, he bears no personal relation of enmity to his foes. The general will in which he participates creates in him only an obligation of military service.27 Since he has an obligation to fight, and since ought implies can, it must be permissible for him to fight. It also follows that since he engages in battle as a soldier, the chivalric ethics appropriate to the soldier's role are appropriate. Thus, Rousseau's account would seem to deliver an account of the normative autonomy of the battlefield, one derived from the collective aspect of war. That autonomy is a consequence of the fact that wars are relations between collectives, fought through individuals.28

Moreover, one can see how Rousseau's argument for the limited right of the victor, grounded in sovereignty as the product of the general will, can support (though not entail) a uniformed condition for the permission. What motivates his argument is the isolation of the citizen's identity in the context of battle, and (as with Bodin) the construction of that identity in external, functional terms. A citizen in uniform has permitted his identity to be reduced to the aspect of soldierhood. His relation to the state is not, as it was with Bodin, a mere tool of the sovereign's will; but it is still limited to the functional role of “defender” obliged by the terms of the social contract to fight for the state. By contrast, the irregular, non-uniformed combatant can be taken as asserting an individual rather than a collective identity: he presents himself as an individual force vector, not a part of an armed host. Since reducing their battlefield identity from citizen to soldier is why vanquished soldiers retain rights to life (thus, impunity for normal acts of war), it makes sense to condition that right on individuals’ formal acceptance of that identity: by donning a uniform.

I now want to argue that while Rousseau's account suggests a path forward, it will not justify the normative autonomy of the battlefield, much less the restriction of the privilege to the uniformed. If what really links citizens to the state is an internal relation of their wills, garbing in uniform is ultimately window-dressing. That a group of soldiers wears uniforms might be external evidence of internal collective organization within a larger political community, and requirements of providing such evidence have clear instrumental value. But the evidence of the tie is not itself constitutive of such organization or ties; a squad of undisciplined mercenaries might be uniformly clothed.

Rousseauean sovereignty poses a major problem for the independence of jus in bello from jus ad bellum. The problem arises because the conceptual isolation of the identity of soldier from that of citizen cannot be maintained. After all, under the victor's sword there is but one person, whose normative identity has different aspects. A father does not cease to be a father when he becomes a soldier; it is simply that his fatherhood is not relevant on the battlefield. But an individual's identity as a citizen does seem relevant on the battlefield, as well as his identity as a soldier. Insofar as he has partly authorized a war, why not hold him responsible for that choice? If the collective decision to wage war is unjust, then as a citizen he is responsible for that injustice.

It may be true, that, as an individual, he is obliged to fight in the service of the collective waging of war. But all that follows is that he should not be punished as an individual for his belligerency, assuming it meets with the norms of proper combat.29 (Even this point does not hold if fighting is voluntary.) It does not follow that he may not be punished as a member of a collective, that is, he and his fellow soldiers may be held collectively responsible for the war they wage. 30 Think of a criminal sentence passed on a business entity: if the sentence is just, then the costs of that sentence are legitimately borne by the business's members: its partners, for example, or shareholders, or employees. Though they are not being punished as individuals, they are punishable as members of the corporate entity.

Sources of the Privilege: the Instrumentalist Strategy

  1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

Thus a Rousseauean argument fails to account for a blanket privilege to kill in war, a privilege independent of the justice of the war itself. One might well respond, So much the worse for the privilege of collective, unjustified violence. But in the service of trying to make sense of current norms, we should pursue the matter further. Indeed, a third and now dominant strategy remains for defending the privilege: the consequentialist strategy I mentioned above, which plays a central role in the International Committee of the Red Cross's understanding of the case for IHL. This strategy effectively links a political realist premise—wars happen—with a normative premise demanding the minimization of human suffering in their wake. Since wars will happen whether or not combatants have special international legal status, the question is, What incentives can limit the suffering they impose? Impunity for certain forms of violence, coupled with the special treatment for the captured, is stipulated as necessary to induce restraint in combatants.31 The further restriction of the privilege to the uniformed is then justified by its role in promoting the distinguishability of combatants from non-combatants.

Like most consequentialist arguments, the force of this is difficult to assess. The privilege must be defended not only at the margin but also categorically. For example, if a war might be shortened through relaxing efforts at discrimination, as Allied forces claimed in World War II when they initiated strategic bombing campaigns, a purely instrumental rationale must permit this. Presumably shorter wars cause less net death and suffering than a prolonged and discriminating war, and must be permitted. But the strategic bombing campaigns are now widely regarded as a grotesque moral mistake, whatever their strategic value.32 For those who consider them a moral mistake, moreover, the mistake clearly does not consist in an aggregate miscalculation, for instance that relaxing area bombing restrictions will increase suffering in other conflicts. That might be true, but the real mistake lies in the tolerance of the wholesale slaughter of civilians per se. The instrumental account suffers the problems of any two-level form of consequentialism: it is unable to offer categorical support for the categorical rules it defends.

There is an analogous point as well, familiar in ethics: if the rules of IHL are justified instrumentally, then that fact must be kept from combatants. For a combatant who knows that IHL is justified on the basis of wholesale calculations of humanitarian advantage will always have reason to ask himself in a given instance whether playing by the rules makes sense, or whether it is a case of what J.J.C. Smart has famously called “rule-worship.”33 What we want to inculcate instead is a combatant's thought that the rules of IHL, and the system of values that sustain them, command categorically. Since soldiers, being human, are reflective creatures, this means that we must provide a non-instrumental argument for those rules. So we must anyway exit the path of instrumental justification.

Furthermore, the empirical assumptions underlying the argument are open to question. First, the realist premise assumes that the amount of combat is fixed independently. But it is hardly clear that the amount is fixed; and indeed it might well be thought that the amount of combat is increased when all participants are guaranteed impunity, especially those fighting criminal wars. It is now widely thought that individual prosecutions for war crimes are necessary or at least useful in reducing the number of war crimes that might occur. Individual prosecutions for unlawful belligerency could also, by the same reasoning, tend to deter individual participation in that belligerency. This is especially true in states with volunteer armies; but even for conscript armies, the prospect of post-capture prosecution might well dampen the ardor of the soldier. A similar argument can be deployed against the familiar claim that the absolute privilege rule reduces suffering by making surrender more attractive. That may be true once the war has begun, but if fewer wars might be initiated in the first place under a privilege restricted to just wars, then killings might be yet further minimized. Without a way to assess the realist claim of the inelasticity of violence, the consequentialist arguments are indeterminate.

Granted, there will be profound disagreements about what constitutes an unjust war, whether in relation to positive international law, or in a broader justificatory argument, as with NATO's Kosovo 1998 intervention.34 Resolving those disagreements would be necessary to justify punishing cases of unlawful belligerency. But those disagreements already need to be resolved for the post-Nuremburg, routine practice of prosecuting political and military leaders for wrongful aggression. Convictions of captive line soldiers could simply adhere, as a form of accomplice liability, to the leadership convictions. As well, a prospective soldier's uncertainty about the permissibility of engaging in combat could be a good thing, insofar as it might dampen efforts in dubious wars (and, more generally, might hinder recruiting and deploying combatants). Cases of clear justification, for example territorial self-defense, would present no problem, as the permission would be clear.35 Third, it is unclear whether combatant privileges really do function as incentives to comply with IHL. A soldier in combat cannot know in advance whether in fact he will receive the treatment he is due under IHL, and a little knowledge of history should make him dubious. (The Allied Forces’ and Germany's treatment of their POWs appear to be historical exceptions.) A rational combatant conditioning his conduct only on the proposed benefit of POW status would have to discount that benefit greatly. On the other hand, a credible threat of greater marginal prosecution for violations of IHL, on top of a prosecution for belligerency itself, would seem more than sufficient to motivate compliance.

The consequentialist argument for a uniform requirement is even weaker. A rule demanding no visible distinctions between combatants and non-combatants might result in much higher civilian casualties than a rule requiring that combatants bear a “distinctive mark, visible at a distance.” But two further claims are also plausible. First, by the “in for a penny, in for a pound” rationale, non-privileged irregular combatants have little interest in refraining from indiscriminate violence; their incentive is just the marginal difference in punishment for war crimes over the punishment for belligerency itself, and both may be death.36 Thus the gain in the ability of the uniformed side to discriminate comes precisely at the cost of a reduced interest on the non-uniformed side of discriminating themselves. Second, and conversely, if it makes sense to provide uniformed combatants killing privileges in order to induce IHL compliance, then it must make sense to offer the same incentive to non-uniformed combatants. The only question is whether costs outweigh benefits, and this cannot be settled from the armchair.

A consequentialist can offer a stronger response: apart from the instrumental value of any particular rule, the existence of some determinate scheme of rules makes a profound welfare contribution. A regime of absolute combat privileges for the uniformed improves decision making in the fog of battle, makes for clearer policy choices at the state level, and provides for stability in international cooperation and treaty formation. Indeed, the even partial regularization of war is one of law's great achievements.37 Nonetheless, the claim is overstated. First, at the level of fact, the world we live in is, as I said above, increasingly characterized by asymmetrical and non-conventional warfare. Distinguishing innocent civilians from perfidious enemies is already a central, and extremely debilitating, part of modern warfare, at least for occupying armies intent on minimizing the killing of the innocent. No system of rules can really dispel the fog of war, and it seems an exaggeration to think that granting POW status to non-uniformed soldiers otherwise innocent of war crimes will do much to thicken that fog.

Second, a moderate form of rules is clearly available that provide for some, but less, discriminatory effect than a uniform. This is precisely the theory behind PI 44's requirement that, when exigencies exist, combatants need only distinguish themselves during combat by carrying their arms openly.38 Third, and most broadly, whatever humanitarian benefits flow from restricting combatancy generally have to be set off against the real costs of discouraging irregular resistance. In historical retrospect (and in many national narratives), some fights against alien occupation or for national self-determination are worth fighting par excellence.39 These are fights that can only be waged plausibly by guerrilla techniques. Giving an asymmetric advantage to a uniformed occupier, whatever the justice of its occupation, means resistance struggles will be rarer or harder than, by hypothesis, they ought to be.

So simply in its own terms, the consequentialist argument for the limited privilege is too indeterminate to serve. The costs and benefits of privileging combatancy are speculative and necessarily involve the kind of gross estimates of long-term consequences that invite contamination by wishful thinking. But this merely confirms a deeper point: if there is an objection to prosecuting combatants for IHL-consistent killings, that objection comes from the domain of right (or fairness), not cost-benefit calculation.

Another Tactic: Combatancy as Complicity

  1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

I began this argument by emphasizing the puzzling distinction between the themes of collective inculpation and collective exculpation, between Crime Story and War Story. Why, in the context of war, should doing violence together make right what in the domestic context it makes wrong? But the discord of these two themes might also be taken as an invitation to harmonize them. In fact, as I argue now, the same logic of collective action that underwrites complicity law also underwrites the law of war. With some help from Rousseau—at least some help from what he should have said, rather than what he did say—we now have the materials to explain and justify a limited form of the privilege of combat.

Take the ethics and law of complicity first, as well as its partner, conspiracy. Complicity functions not as an independent crime in its own right, but as a distinctive form of moral and legal responsibility that links agents to outcomes by way of their participation in a collective effort, and largely independently of their individual causal contributions.40 Recall Taylor, in Crime Story: if he genuinely has thrown his lot in with the armed robbery, then he bears responsibility for Smith's killing, and punishing him for that killing is just, even if we do not regard him as Smith's literal killer. Or consider the British case of DPP for Northern Ireland v. Maxwell.41 James Maxwell, a standing member of the Ulster Volunteer Force (UVF), was asked by a fellow member of the UVF to help on a “job.” In Maxwell's case, this meant driving his own car to guide a following car to an inn. Maxwell drove past the inn, but knew that the tailing car stopped. In fact the tailing car had left a pipe bomb at the inn, a bomb that, fortunately, the son of the inn's owner was able to defuse.

Although Maxwell did not know the specifics of the terrorist “job,” and though he neither touched nor saw the bomb himself, he was nonetheless convicted of planting of an illegal bomb, on the grounds that Maxwell knew some form of terrorist action was afoot, and that he had played a significant role in guiding the bombers. Maxwell was criminally liable for the foreseeable acts of the group in which he participated, for when we act together, we individually bear responsibility for what we together bring about, within the scope of our common venture.

The logic of complicity is the logic of collective action more generally, and that logic pervades our social, ethical, and legal existence. It explains and justifies, I believe, much of the pride we take in our collective accomplishments, even when our own contributions lie at the insignificant margin. It explains the special importance we attach to the signal act of collective freedom, voting, an act whose individual causal significance is far outweighed by its costs.42 In addition, it explains and justifies much of the shame and guilt we feel when the groups in which we live do wrong, even when we have been dissenting voices within. In all these cases, we begin with a group act and then derive and distribute the individual responsibilities thereof. Individual pride makes sense because of our participation in a collective accomplishment; the decision to vote makes sense because the collective selection of political authority is a necessary condition of freedom; our shame makes sense because the wrongs we do together are consequences of the collective systems and institutions to which we contribute.

Our individual responsibility for these collective acts is point one. Point two is that individual responsibility is not the same thing as collective responsibility. When I take pride in, say, my orchestra's brilliant performance, I do not regard myself as individually responsible for that brilliant performance. When I feel shame for my nation's prosecution of an unjust war, I do not regard myself as personally responsible for that war. Recognition of my responsibility involves recognition that that responsibility is a relation in social space, one that links me in normative terms both horizontally to the other members of the group, and vertically, to those whom my group affects (or to the outcomes it produces). My response to, and responsibility for, what we together do is essentially mediated by membership in the group and grounded in my individual participation therein.

In the present case, the logic of collective action both enables and disables an account of the combatants’ privilege. As Rousseau saw, under modern conditions of politics war is also something we do together, a normative relation we bear as a group to another group.43 As an individual, I share in responsibility for the decision to go to war. But my responsibility as an individual is not identical with the responsibility of the group. My individual responsibility is, rather, a duty to serve if called and if the war is not clearly criminal, and to protest if it is (and perhaps to refuse service as well). The fact that my nation is at war, not me, does not absolve me of responsibility towards my enemy, but it does create a normatively distinct relation between us, one structured through a set of rules specific to our interrelationship as individual members of warring nations in confrontation with one another. This is the logical space in which jus in bello can claim independence from jus ad bellum.

Specifically, the logic of collective action can make appropriate a limited scope for an essentially political permission to do violence, because when I do violence, I do it as a member of one group towards another. The privilege to kill as part of a collective is not a moral permission attaching to the individual soldier. A soldier who kills as part of an unjust war morally wrongs those he kills, and bears a share of responsibility for their deaths. But it does not follow that an enemy state can legitimately punish him, even if it can kill him in battle. Rousseau was right: the victorious state encounters the individual only accidentally: its essential normative relations are with the soldier's state, not with him. As Rousseau says, enemy soldiers confront each other as defenders.

Or as attackers, and there's the rub. The argument I have just given seems to me the best case for making jus in bello independent of jus ad bellum. But, as with Rousseau's argument, which it parallels, it requires an over-strong distinction between individual and collective responsibility. For it is plausible, particularly on a retributive theory, to say that the soldier who kills while prosecuting an unjust cause is fit for punishment.44 After all, it is as an individual that he participates in the unjust war. A collective response to the enemy state does not preclude an individual response to the enemy soldier.

Perhaps this is the proper conclusion: collective decisions to go to war confer no individual immunity from punishment (for those participating in the collective).45 When the injustice of the war is clear, so is the justice of prosecuting the aggressors in that war. There would be, of course, profound questions about the appropriate degrees of punishment given the range of pressures placed on individuals to fight, and difficult issues of post-punishment reintegration. But at the level of principle, there is not only conceptual room, but conceptual pressure towards linking jus ad bellum to jus in bello. Nonetheless, there is another aspect of the collective nature of war that tells against drawing too tight a link between the individual and the state for which he fights. Wars, like many of history's uglier monuments, come to look very different in retrospect than they do in prospect. Many belligerent acts, like many violent revolutions, are easily condemned at the time but become praiseworthy in retrospect. This is because history happens in messy ways, and it involves a kind of normative mistake to apply ex post the same criteria that one applies ex ante.

To take some recent, albeit controversial, examples: Israel's 1981 preemptive destruction of Iraq's Osirak reactors seemed an outrageous violation of limits of aggression at the time, and now like a prudent and regionally responsible intervention. NATO's Kosovo intervention, intensely debated at the time, now seems one of the alliance's finest moments. And if the war in Iraq, which seems thus far morally and practically disastrous, nonetheless leads directly to a peaceful and democratic Middle East, then doubtless my retrospective judgment will surely shift.

This does not just concern the difficulty of establishing uncontroverted criteria for assessing the justice of war. It is, rather, a point about the vulnerability of judgments of a war's justice to an analogue of what Bernard Williams called moral luck, and what we might call political luck.46 Williams’ example was painter Paul Gauguin, who (in Williams’ version) went to Tahiti to paint and in so doing abandoned his wife and children to poverty in Paris. According to Williams, if Gauguin's paintings had been aesthetic failures, then his trip would have been a moral failure. Since they were (at least stipulatively) aesthetic successes, however, his trip cannot be condemned in moral terms.

Williams’ argument may not fully convince, for we might well come to a more complicated judgment: “Gauguin may be a louse, but he painted some beautiful pictures.” In fact Williams is fairly non-specific about the normative consequences of the aesthetic triumph; he does not claim that the trip becomes morally justified, but only that condemning it is beside the point. In political and historical contexts, Williams’ claim is amplified. Retrospectively we care less about the properties of actions and more about the possibilities and constraints inhering in the outcomes they produce. The immoralities of acts are swept with the economists’ broom into the dustbin of sunk costs. Because politics is fundamentally about the question of what we together should do, its perspective is anchored in the now and moves forward, aggregating over collective interests and values. Conversely, its outcomes can only be assessed in retrospect, and that in the longer term.

Criminal judgment also applies in retrospect, but the gap in time between act and judgment will usually be too short to accommodate vicissitudes in judgments of some wars’ justice. The normative autonomy of the battlefield, at least for the great range of conflicts in which judgment might reasonably be thought to vary in time, might then be thought to reflect the gap between the immediately post-war assessment of individual battlefield conduct and the longer-term assessment of the war's justification. A war's justification might emerge post bellum, in the epistemological sense that while in advance the warrant for military action might have been deeply controversial—perhaps because facts on the ground were in dispute, as in a developing genocide—facts available after the war might render that initial judgment much less controversial. The Kosovo intervention might be an example of this phenomenon; and so, contrarily, might be the Iraq war. But this justification may only emerge long after the battles, and after prosecutions would have begun. More radically, the judgment whether war was warranted, made by victors or third-party tribunals in a position to permit prosecutions, might end up turning on the costs of the war or the success of the post-war peace, with “good” but unjustified wars grounding immunity, and costly but perhaps justified wars grounding prosecution. Given the likely vicissitudes of these essentially political and post hoc judgments, it would be unfair to punish line soldiers except in the cases of the most grossly unjust wars, such as extraterritorial genocide.47 By contrast, the norms of proportionality and discrimination can be easily deployed in judging individual conduct, so the regulatory force of IHL is preserved.

In any event, the question whether to expand combatancy's privilege to the non-uniformed can be resolved quickly and independently of resolving the precise scope of that privilege for the uniformed. On either of the accounts I have offered for the general combat privilege, the privilege is grounded in the relation of individual combatants to a collective decision to go to war. That relation is a matter of individual commitments to the collective: their mutual orientation around each other as fellow agents in a collective project. If an essentially intentional relation among individuals grounds the privilege, then the privilege ought logically to be extended to any who together constitute a collective at war, whether or not they are uniformed. Instrumental considerations of the sort canvassed above might tip the decision one way or another; but if those considerations are as indecisive as I argued, then there is no reason not to extend the combat privilege and a good reason to do so. Thus, something like the moderate regime of PI 44, requiring only open carriage of weapons in deployment and combat, can be justified as a matter of principle and defended as a matter of practice.

This conclusion may seem a bit quick, for it cannot be that any group of individuals, merely because they act as a group, can earn for themselves the privilege of combat. This would, obviously, be to erase the line between criminal law and the law of war, in favor of the latter. We do surely want a way to distinguish between Crime Story's gang and Rebel Story's partisans. What was implicit above needs to become explicit: only political groups engaged in violence in support of political goals, in the sense of aiming at creating (or restoring) a new collective ordering, can rightly claim the privilege.

Whether a group's violent acts count as political, or as merely criminal, turn principally on three factors: the existence (or not) of an internal ordering, the character of its aims, and the degree of success on the ground. The existence of internal order is necessary, because it is a legitimate condition of extending combat privileges to a group that it be itself capable of regulating its own conduct by the laws of war.48 Groups on the verge of internal anarchy would thus fail to meet this condition. As to the second factor, the character of its aims, the substantive criteria of recognition created by the First Protocol—that groups be engaged in projects of national liberation or self-determination—mark an understandable starting point, albeit a contentious one.49 For liberation and self-determination are political aims, and are prima facie the sort of causes that can justify violence if anything can; but so might also be disputes over regional autonomy or the flow of resources to particular regions, as with the Zapatistas, or struggles for religious or cultural autonomy, as with the Kurds. This criterion effectively excludes groups like the Colombian narco-trafficking groups that have sought and attained powers of territorial governance (and popular acquiescence), but only for the sake of securing their coca supply, not for the sake of political aims.

The third criterion, degree of success, is more problematic. The point of such a criterion is to recognize the practical need of state authorities to suppress disturbances to the peace that lack the legitimating force of broad popular support and thus to deny the privilege of war to groups whose violence, however symbolically justified, can do nothing but create civil unrest. There is no sharp way to define such a criterion. In principle, popular support (however gauged) or territorial control might be the right guides. Such measures of success indicate that a group may be able to bargain effectively to achieve some of its goals, even if it cannot force concessions of all of them. In practice, only those groups that actually have popular support or territorial control will have the leverage necessary to force recognition, and such recognition might only come with time.50

More abstractly, the success criterion recognizes that engaging in politics is not just a matter of positing wishes, but of creating a real, mutual, social ordering. Politics, and political violence, must be anchored in real possibilities of social formation and transformation.51 This criterion makes the moral permission to fight dependent on non-moral factors, so that rebels fighting clearly just but hopeless causes are subject to punishment. This is a hard position, for it denies the privilege to groups who might have had popular support but for the success of state terror. But it seems right. Occupations may be real usurpations of self-government, but if and when they bring civil order, any group opposing that order bears a large normative burden in justifying its resort to violence. The game may be worth the candle, but only if it is a winnable game, at a tolerable human cost. A group engaged in violence but whose aims are part of no actual or reasonably possible system of social ordering engages not in politics but rather in a deadly solipsistic fantasy.

It is a feature of this account, indeed a virtue, that whether a group of irregulars engaged in combat count as political, and are thus entitled to combat privileges, may change over time. In fact, the status of those captured may turn from criminal to POW as their colleagues find success in the fields and in the towns. (Such transformations of status happen anyway as a matter of negotiations between states and increasingly powerful insurgencies.) In any event, the importance of POW status for groups on the margin of criminality may be oversold. As a matter of practice, states will deny them that status until the groups are sufficiently powerful to demand it, whether or not the groups would be entitled to POW status as a matter of law. Since even lawful combatants may be held until the cessation of hostilities, which in civil or quasi-civil conflicts may be indefinite, and since they may also be interrogated exhaustively (but not punished for refusal to answer), the state loses little security by acknowledging combatant status. Moreover, since violations of the law of war can be punished among lawful and unlawful combatants alike, granting POW status hardly precludes prosecution for terrorist acts. In short, although expanding combat privileges to irregulars brings risks and disputed judgments, it may actually be less disruptive than the resisters fear as well as more in consonance with the best case to be made for the categorical character of IHL norms.


  1. Top of page
  2. Introduction
  3. Two Themes in the Key of Collective Violence
  4. Designing Norms for New Wars
  5. The Paradoxes of War
  6. Conceptual Sources of the Combat Privilege
  7. Sources of the Privilege: the Instrumentalist Strategy
  8. Another Tactic: Combatancy as Complicity
  9. Conclusion

I offer here a brief survey of the concrete implications of this view. Where on the traditional view, War Story is the easy case and Rebel Story is the hard case, on my view the situation reverses. Rebel Gray, although not a Petrostan army member, is a member of the group seeking Petrostan's liberation, a political goal. She has linked her will with theirs and so inhabits a common normative space, in pursuit of a paradigmatic political goal. That she wears no uniform is irrelevant to the collective aspect of her individual action; and it is the collective aspect that underwrites her privilege. Assuming she has obeyed the laws of war, she ought to be impunible.

Sergeant Blue's case is harder, because the question of his combat privilege now depends on whether the injustice of Imperioland's invasion is so great as to fall outside the scope of reasonable disagreement or reasonable retrospective re-assessment of Imperioland's case for war. On the bare facts I stipulated, this is unclear. An invasion to acquire another nation's resources looks clearly illegal, but the question becomes more complex for resources located near hastily drawn or colonially imposed borders, or when legitimate international disputes exist about access to those resources. So long as some of these factors are relevant to assessing Imperioland's case, it seems appropriate to defend a privilege for Blue as well. He may have acted badly, in moral terms, insofar as he took part in collective violence on grounds he knew or had reason to know were morally dubious, and the deaths he caused should sit uneasily on his conscience. The question of whether it is legitimate for Petrostan (or an international body) to punish him, however, is far more difficult.

Closer to home, my view entails that Taliban fighters and the foreign volunteer “Afghan Auxiliaries,” whether or not they were garbed “distinctively,” ought to have received lawful combatant status, assuming they displayed their weapons openly in conflict and respected IHL norms. The Taliban regime may have been unjust, but self-defense of even a wicked regime sits squarely within the scope of privilege for uniformed soldiers, and we have dispensed with the reasons for discriminating against the un-uniformed. So too, I think, combatant privileges could belong to members of Iraq's Baathist insurgency, provided again that they obey the rules of war. None of this turns on approving the regime they aim to install or defend, and it is fully consistent with punishing all humanitarian excesses. This is only to recognize that the claim to engage in the form of collective politics known as war belongs prima facie to all, and can logically be withdrawn from each, uniformed or not.

Clearly these cases pose difficult questions of policy, and controversy will inevitably remain for any set of legal rules that might apply to them. Seeing the law of war through the lens of the criminal law of complicity reveals an underlying logic of collective action that can make sense of both bodies of law. Further, seeing that logic in the special collective context of politics can help us understand a deeper rationale for the core of the law of war, the combatant privilege. More importantly, understanding war in terms of collective action forces us to reckon with the real individual responsibilities that come with participation in collective violence: relations of value that go beyond dulce et decorum est pro patria mori.52

  • I am grateful to audiences at the Humboldt University, Boalt Hall Law School, Columbia Law School, Yale Law School, and Stanford Law School, for their reactions to this article (or to its ancestors). I am especially grateful to Richard Buxbaum, Jules Coleman, Lori Damrosch, Michael Doyle, Barbara Fried, Andrew Guzman, Thomas Grey, Jeff McMahan, Neyson Mahboud, Gerald Neuman, Peter Strauss, Jeremy Waldron, and John Yoo for further discussion and correspondence: and to Shahla Maghzi and Eric Shen for superlative research. I am also grateful to the Editors of Philosophy & Public Affairs for their marvelous comments.

  • 1

    Based loosely on Taylor v. Superior Court, 3 Cal. 3d 578, 477 P.2d 131 (1970).

  • 2

    1949 Geneva Convention relative to the Treatment of Prisoners of War, Art. 118. (Future citations to the Geneva Conventions will be of the form GPW 118.)

  • 3

    I draw upon the “facts” offered by the government in the Yasser Hamdi case. See “Declaration of Michael H. Mobbs,” Special Advisor to the Under Secretary of Defense for Policy, filed in Hamdi v. Rumsfeld, No. 2:02CV439 (E.D. Va). The Supreme Court has since ruled that U.S. citizens taken on foreign battlefields are constitutionally entitled to a legal forum in which they can contest the facts governing their legal status. Hamdi v. Rumsfeld, No. 03-6696, 542 U.S._ (2004) accessible at <>. If honored by the government (unlikely at this writing), this ruling would bring U.S. practice with respect to its own citizens back into conformity with Article 5 of GPW, which requires adjudication of all dubious cases by a “competent tribunal.” U.S. practice for non-U.S. battlefield captures does not yet conform to even the weak combatancy status hearing requirements of GPW 5.

  • 4

    Raymond Aron, On War (New York: W. W. Norton, 1968).

  • 5

    GPW 4(A)(2)(b); 1977 Protocol Additional to the Geneva Conventions of 12 August 1949, and relating to the Protection of Victims of International Armed Conflicts (PI), Art. 44 (3). Under PI 1(4), only persons involved in interstate conflicts or “conflicts in which peoples are fighting against colonial domination and alien occupation and against racist regimes in the exercise of their right of self-determination” have access to the relaxed standard of combatancy of PI 44(3). In addition, GPW 4(A)(6) extends battlefield privileges to citizens who, as a whole, rise up as a foreign invader arrives. This “levée en masse” clause is almost never triggered, and would not be triggered by partisans resisting an occupation. Non-privileged combatants may be killed on the battlefield, as well as be prosecuted after conquest for their belligerency.

  • 6

    A “substantial step” is the Model Penal Code's rule for attempt liability, Sec. 5.01(1)(c).

  • 7

    The United States calls these “unlawful combatants.” See, e.g., Ex parte Quirin, 317 U.S. 1, 31 (1942). One of the pernicious features of the term “unlawful combatants” is that it effectively conflates crimes like killing civilians with not wearing a uniform in combat. Another is that it concludes, rather than leaving open, the question whether they enjoy any privilege to kill.

  • 8

    Currently, 146 states have ratified Protocol I, though many have made reservations to Art. 44. Given the number of ratifiers, some argue that PI 44 (like common articles 3 of the Geneva Conventions, which lays down general limits to violence in all conflicts) now has force as a universal, customary (rather than treaty-based) norm. See Antonio Cassesse, “The Geneva Protocols of 1977 on the Humanitarian Law of Armed Conflict and Customary International Law,”Pacific Basin Law Journal 3 (1985): 55118, at pp. 72–73. But U.S. non-compliance with that regime undermines the argument for customary force.

  • 9

    Special Forces soldiers may have dressed distinctively, however, in the garb of the militias with whom they were affiliated. If so, and given a reasonably generous interpretation of the requirement of GPW 4(a)(2), which requires that combatants wear “a fixed distinctive sign,” then they would be lawful combatants. Ironically, however, this reading is denied by the United States officially, as part of their ground for not treating captured Taliban as POWs. See W. Hays Parks, “Special Forces Wearing of Non-Standard Uniforms,”Chicago Journal of International Law 4 (2003): 493547, at pp. 496–98. It is unclear whether CIA personnel who took part in the hostilities wore any distinctive garb; since presumably they were already unlikely to gain POW status if captured, they had no reason to.

  • 10

    These developments are nicely surveyed by Herfried Münkler, “The Wars of the 21st Century,”International Review of the Red Cross 85 (2003): 722.

  • 11

    Civilian contractors, for example, routinely operate surveillance aircraft, provide direct logistical support for weapons systems, operate combat-zone radar equipment, and fly armed drug interdiction efforts in collaboration with the U.S. military. Many of these roles seem close enough to the criterion of “direct participation” in the hostilities to render them combatants under GPW 4. For a survey of this phenomenon, see Peter Singer, Corporate Warriors (Ithaca, N.Y.: Cornell University Press, 2003). As Singer points out, the new mercenaries can contribute to social peace (as they did in Sierra Leone, at least until their contract expired) as well as to escalate conflicts between weak states that would not otherwise be able to engage in sophisticated levels of violence.

  • 12

    An early, important philosophical discussion of this point is by Thomas Nagel, “War and Massacre,” in his Mortal Questions (New York: Cambridge University Press, 1979), pp. 53–74. I am also indebted to Jeff McMahan, “The Ethics of Killing in Wars,”Ethics 114 (2004): 693733, which is especially perspicuous in pointing out the puzzles of IHL's normative authority.

  • 13

    These norms further proscribe certain disproportionate or indiscriminate killing means, for example poisonous gas. See, e.g., Hague Convention of 1925, Protocol for the Prohibition of the Use in War of Asphyxiating, Poisonous or Other Gases, and of Bacteriological Methods of Warfare; Declaration of 1899 Declaration on the Use of Bullets Which Expand or Flatten Easily in the Human Body.

  • 14

    See Michael Walzer: “It is perfectly possible for a just war to be fought unjustly and for an unjust war to be fought in strict accordance with the rules.”Just and Unjust Wars, 3d ed. (New York: Basic Books, 2000), p. 21. This formulation, of course, leaves it ambiguous whether any strong substantive normative value attaches to the formal criterion of “playing by the rules.” See also Gabor Rona, “Interesting Times for International Humanitarian Law: Challenges from the ‘War on Terror’,”The Fletcher Forum of World Affairs 27 (2003): 5574, at pp. 67–68: “The very essence of jus ad bellum is the distinction between just and unjust cause—between entitlement and prohibition to wage war. Jus in bello, on the other hand, rightfully recognizes no such distinction. While one party may be a sinner and the other a saint under jus ad bellum, the jus in bello must and does bind the aggressor and the aggressed equally.”

  • 15

    See Theodore Meron, Bloody Constraint: War and Chivalry in Shakespeare (New York: Oxford University Press, 1998), as well as Allan Rosas, The Legal Status of Prisoners of War (Helsinki: Suomalainen Tiedeakatemia, 1976), pp. 44–84.

  • 16

    The phrase is Samuel Scheffler’s, “Individual Responsibility in a Global Age,”Social Philosophy and Policy 12 (1995): 21936, at p. 222.

  • 17

    See Jean Bodin, Bodin: On Sovereignty, ed. Julian Franklin (New York: Cambridge University Press, 1992), Bk. I, Ch. 8, “On Sovereignty.”

  • 18

    “For although one can receive law from someone else, it is as impossible by nature to give one's self a law as it is to command one's self to do something that depends on one's will.” Bodin, Bk. I, Ch. 8, [360–61], p. 12.

  • 19

    See Geoffrey Parker, The Military Revolution: Military Innovation and the Rise of the West 1500–1800 (Cambridge: Cambridge University Press, 1988), pp. 71–72; “Uniforms,” in The Oxford Companion to Military History, ed. Richard Holmes (New York: Oxford University Press, 2001), pp. 931–35; Toni Pfanner, “Military Uniforms and the Law of War,”International Review of the Red Cross 86 (2004): 93130, at pp. 95–99.

  • 20

    See Münkler, “Wars of the 21st Century,” pp. 14–16.

  • 21

    “From all this it is clear that the principal mark of sovereign majesty and absolute power is the right to impose laws generally on all subjects regardless of their consent.” Bodin, Bk. I, Ch. 7.

  • 22

    Jean Jacques Rousseau, “The Social Contract,” Bk. I, Ch. 7, par. 10, in The Social Contract and Other Writings, ed. Victor Gourevitch (New York: Cambridge University Press, 1997), pp. 50–51.

  • 23

    Rousseau, Social Contract, Bk. I, Ch. 7, par. 1 (p. 51).

  • 24

    Rousseau, Social Contract, Bk. I, Ch. 4, par. 9 (pp. 46–47).

  • 25

    Rousseau, Social Contract, Bk. I, Ch. 4, par. 10 (p. 47).

  • 26

    Rousseau, Social Contract, Bk. I, Ch. 4, par. 10 (p. 47).

  • 27

    What I have called in other work a “participatory obligation” to do one's part in a collective project to which one is committed. See my my “The Collective Work of Citizenship,”Legal Theory 8 (2002): 47194.

  • 28

    Compare George Fletcher, Romantics at War (Princeton, N.J.: Princeton University Press, 2002). Fletcher argues that Rousseau's conception of war is essentially Romantic, a form of self-expression by an organically united people. While Fletcher is right that Rousseau's thought featured prominently in later Romantic conceptions of peoples and their self-expression (as, for example, in J. G. Herder), his reading ignores the Enlightened and contractarian aspect of Rousseau's own conception of sovereignty, as well as Rousseau's view of the contingent nature of a politically united people.

  • 29

    Hence conscripts in authoritarian regimes might be seen as doubly impunible: as individuals they may have a claim of duress, and they lack any collective citizen responsibility. This does, however, assume that duress is true in fact (i.e., that conscripts will be killed if they do not kill) and that it excuses even killing, which is generally not true in Anglo-American law. See McMahan, “Unjust Wars,” p. 700.

  • 30

    McMahan also rejects traditional jus in bello principles and argues that individual soldiers in unjust wars bear moral responsibility for what they do. Although, as I elaborate below, I agree that individual soldiers may bear moral responsibility, one still needs to argue for the legitimacy of punishing combatants for their (morally) impermissible killings. Impunity, in other words, is not the same as justification or moral permissibility.

  • 31

    See, e.g., Gabor Rona, a Legal Advisor to the ICRC, in “Interesting Times,” p. 57: “[H]umanitarian law is a compromise. In return for these protections, humanitarian law elevates the essence of war—killing and detaining people without trial—into a right, if only for persons designated as ‘privileged combatants’, such as soldiers in an army.”

  • 32

    They may also have been strategic miscalculations, at least in Europe, where there is little evidence that they made a difference to already quickly declining German power. See, e.g., Michael Sherry, The Rise of American Air Power: The Creation of Armageddon (New Haven, Conn.: Yale University Press, 1987), p. 260.

  • 33

    J.J.C. Smart, “An Outline of a Theory of Utilitarianism,” in Smart and Bernard Williams, Utilitarianism: For and Against (New York: Cambridge University Press, 1973).

  • 34

    For doubts about its lawfulness, though not about its prudence, see Louis Henkin, “Kosovo and the Law of ‘Humanitarian Intervention’,”American Journal of International Law 4 (1999), accessible at <>.

  • 35

    Problems would arise for humanitarian military intervention by national, as opposed to UN, forces. This is a difficult problem, as arguably only national actors have the will to intervene early enough to make a difference. I think it sufficient here to say that the benefits of expanded permission to engage in such missions have to be offset by the costs that that permission will underwrite clearly unwarranted interventions.

  • 36

    This assumes that they do not care independently about civilian casualties, as indeed they may not if fighting on enemy territory.

  • 37

    I am indebted to Jeremy Waldron for discussion of this point.

  • 38

    I would guess that real challenges for discrimination come in two settings. The first is urban combat, where discrimination is already difficult even between distinctively marked troops. Second is the long-distance aerial strike, where small arms might not be visible to a target spotter. But since combatants, uniformed or not, might well be camouflaged in buildings or vehicles that require no distinctive marking, the problem of discrimination does not seem to me appreciably greater with PI 44 than without. (This may underestimate the difficulties faced by U.S. troops facing guerilla warfare in Vietnam, though France, with similar guerrilla experience, ratified PI without reservation to Art. 44.)

  • 39

    Arguably, protection should also be extended to foreign volunteer groups aiding partisans in struggles for self-determination, or against extermination, so long as the national partisans themselves qualify for protection. (Foreign members of national groups, like Gray, are protected under the current PI criteria.) While there is a cost to this position—it increases instability by making foreign intervention more attractive—it has the corresponding benefit of increasing the likelihood of success in just struggles. The logic is thus parallel to the national group case, and ought to be governed by the same deeper principle, namely that these volunteer groups have come to joins their wills in the collective struggle as well. (I thank an Editor of Philosophy & Public Affairs for this suggestion.)

  • 40

    I explain and defend this claim extensively in Complicity: Ethics and Law for a Collective Age (New York: Cambridge University Press, 2000).

  • 41

    [1978] 1 W.L.R. 1350.

  • 42

    I develop this argument in “Collective Work of Citizenship.”

  • 43

    This is one of many reasons why “terrorism” cannot be the opponent of a war. If the protean abstraction of global terrorism is the opponent, then anywhere terrorists act is a scene of “battlefield” combat, governed only by the laws of war. This means states might target and kill virtually anyone suspected of terrorism, subject only to constraints of reasonable discrimination and proportionality. For discussion, see Rona, “Interesting Times.” Questions about the proper legal analysis governing the conflict with terrorist groups such as Al Qaeda are very complicated, and beyond the scope of my argument, which concerns unproblematic deployments of the idea of armed conflict and battle. For discussion, see John C. Yoo and James C. Ho, “The Status of Terrorists,”Virginia Journal of International Law 47 (2003): 20728.

  • 44

    It is even more plausible on a deterrence theory, of course: a state has every legitimate interest in deterring attacks on its soldiers.

  • 45

    It is a clear implication of my view that non-combatant citizens are also in principle exposed to punishment for the belligerency. However, for the much-discussed consequentialist and slippery-slope reasons of trying to avoid total war, I regard the impermissiblity of attacking non-combatants as much easier to defend than the permissibility of intercombatant killing.

  • 46

    Bernard Williams, “Moral Luck,” in Moral Luck (New York: Cambridge University Press, 1981), pp. 20–39.

  • 47

    For a rich discussion of post-war justice, see Gary J. Bass, “Jus Post Bellum,”Philosophy & Public Affairs 32 (2004): 384412.

  • 48

    As GPW 4(A)(2)(d) requires.

  • 49

    This raises the question of the fairness of prosecuting national leaders for waging unjust wars. They too, after all, may be prosecuted long before opinions are clear on the justification for their legitimacy. Nonetheless, a distinction between leadership and line prosecutions is acceptable. Prosecutions of national leaders are likely to be such rare events, involve so few persons, and to be so constrained by the exigencies of international politics that the risk of unfairness is surely lower than that courted by routine prosecutions of enemy combatants. It also seems appropriate to hold national leaders to higher standards of compliance with standards of just conduct than soldiers, whose views about the permissibility of their nation's conduct are likely to be more permeated by jingoistic false consciousness than their leaders’.

  • 50

    I thank an Editor at Philosophy & Public Affairs for pressing me on this point.

  • 51

    Here I extend into politics H.L.A. Hart's argument that the necessary and sufficient conditions of a legal system's existence are the acknowledgment among officials of the normative force of a system's rules, and a popular practice of obedience to those rules. The Concept of Law, 2d ed. (New York: Oxford University Press, 1994), p. 113.

  • 52

    There is, then, an asymmetry for soldiers of egregious regimes: those fighting voluntarily to extend their nations’ sway would not be privileged, even though they retain the privilege in defending their states. But since the privilege of killing comes with the correlative privilege of their enemies to kill them, this asymmetry is not such a benefit.