Oxford Moral Philosophy Seminar, 3/11/08
Let me explain where the following discussion fits into a larger project of mine. The project is about the moral foundations of just war theory. Given the ubiquitous nature of military violence and the changing character of warfare, this issue is of practical as well as theoretical import. Just war theory has recently been in intellectual ferment, especially regarding the justice of going to war, jus ad bellum. Factors such as the unfreezing of the international political order at the end of the cold war and economic and cultural globalization have led to uncertainty, dislocation, a weakening of state sovereignty, and an explosion of military violence, often intra-state violence. Add to this a growing awareness of the importance of international human rights, and the stage was set for recent concerns about military intervention, humanitarian intervention and, more problematic, preventive intervention. In response to this, part of my project is to see if there is a need to account for such factors by significantly recasting jus ad bellum.
But there also been ferment on the question of just conduct in war, jus in bello, due to factors such as reaction to the civilian toll in World War II and the rise in unconventional forms of violence, such as guerrilla warfare and non-state terrorism. It is jus in bello I address today. Specifically, I consider some issues raised by recent theorists who have challenged our traditional understanding of this aspect of just war theory.1 I would like to offer special thanks to Seth Lazar, who provided an excellent set of comments on an earlier draft of this paper.
Alberto Gonzales, then Bush’s Attorney General, famously advised the President that “the war on terrorism is a new kind of war, a new paradigm [that] renders obsolete Geneva's strict limitation on questioning of enemy prisoners and renders some of its provisions quaint.”2 In using the terms “obsolete” and “quaint,” Gonzales presumably meant that some of the rules of the Geneva Conventions had been rendered out-of-date by the “new kind of war” and so should be abandoned in favour of other rules (or perhaps none at all). This suggests the idea that the rules, understood as subject to change in this way, are mere conventions.
By “convention” I mean an effective agreement among members of a group to conduct their behaviour according to certain norms. An agreement is effective when it is generally adhered to and is regarded by members of the group with what Hart calls a critical attitude. Some conventions are formally agreed upon, like the Geneva Conventions. The phrase, “the Geneva Conventions are conventions” is not a tautology, but contains an equivocation. The first use of the term “conventions” refers to a certain formal agreements, while the second adds the claim that those agreements are effective.
By “mere convention” I mean a convention not also backed directly by a deontological moral rule of some sort. Adhering to a mere convention may or may not have moral value, but, if it does have moral value, the value is consequentialist or pragmatic, and so usually depends on general adherence to the norm. The convention in domestic society against killing innocent people is clearly not a mere convention, and there is moral value in adhering to it even if others are not. The convention requiring driving on the left side of the road is a mere convention because, though adhering to it does have moral value, this value is dependent on others doing so. In the absence of general adherence, my adherence would have little or no moral value. The moral value of adhering to a mere convention, if any, is extrinsic to the individual act of adherence; it is dependent of the actions of others. But the moral value of adhering to a convention that is backed by a moral rule is intrinsic to the individual act of adherence. I will refer to a convention directly backed by a moral rule, in contrast to a mere convention, as a “moral convention.”
One important reason that conventions come into being or become extinct is that reciprocal adherence may be mutually beneficial. Thus, for example, when motorists adhere to the driving convention, there is mutual benefit in the avoidance of accidents. The prudential motivation to adhere is usually based on an expectation of reciprocity. When adherence to a convention begins to fall off, this undercuts the expectation of reciprocity, and the convention may be on its way to extinction. But there is another dynamic sometimes at work.
When there are only a few players, I may adhere to a convention without the immediate expectation of reciprocity in an effort to create or recreate reciprocity by example, showing others that general adherence would be beneficial to them. Indeed, it is in such terms that then Secretary of State Colin Powell reacted to the Gonzales memo, arguing that the abandoning of the Geneva Conventions by the United States would “undermine the protections of the law of war for our troops, both in this specific conflict and in general.”3 If the United States violates the Geneva Convention by treating prisoners in the “war on terrorism” in harsh ways, it can expect that its own combatants who become prisoners will be treated likewise. The conventions prohibiting torture were worth maintaining, Powell asserted, in the hopes that this will generate reciprocal behaviour by our foes.
The expectation of mutual benefit serves not only as a (partial) explanation for a convention, but also as a justification for it. The justification is, first of all, prudential—because of mutual benefit, the convention works for me. But the justification may also be moral, in a consequentialist sense, because when everyone benefits individually, it may well be that the total utility is increased. Should I balk at adhering, someone is apt to respond, “What if everyone did that?” and this could be taken as either a prudential or a moral appeal. But in the case of a moral convention, another justification is available, because adherence is morally valuable in itself.
The principle of discrimination of jus in bello (henceforth, simply “discrimination”) requires that combatants discriminate in their fighting, directing attacks only at opposing combatants and not at civilians. This principle is a convention, but is it a mere convention or a moral convention? I raise this question to join a debate over the nature or status of jus in bello rules. In its recent incarnation, this debate began over three decades ago when George Mavrodes argued that the “immunity thesis” (as he calls discrimination) lacks a “morally relevant basis.” Traditional defenders of the immunity thesis cite as its moral basis the concepts of moral guilt and innocence, the idea of moral fault. Combatants are guilty in the sense that they are morally responsible for the harm they pose or threaten, while noncombatants are innocent because they do not pose or threaten harm. But fault, Mavrodes argues, cannot be the moral basis of the immunity thesis because it does not track the distinction between combatants and noncombatants. Some noncombatants are enthusiastic supports of the war, hence share moral responsibility for the harm it does, while some combatants do not fight voluntarily and so are not morally responsible for the harm.4 So discrimination is not morally justified in itself, hence is a mere convention.
Mavrodes claims instead that the immunity thesis has moral justification as a convention, in that it may be understood as a norm (among other possible norms, such as that of combat of champions) adopted to limit the destructiveness of warfare. There is consequentialist moral advantage if there is general adherence to the convention. Mavrodes refers to our obligation to adhere to the convention as “convention dependent.” (By contrast, our obligation to adhere to a moral convention is convention independent.) (The distinction between convention-dependent and convention-independent moral obligations is analogous to the categories of malum prohibitum and malum in se in domestic law.) It is important to appreciate that, in general, a convention-dependent moral obligation exists only when the convention promotes utility.
An important difference is that a convention-dependent moral obligation holds only so long as the convention that generates it holds. In contrast, if a moral convention fails, the obligation to adhere remains, as it is supported by the moral rule. In the case of a mere convention, the lapsing of the obligation to adhere as the convention heads toward extinction is related to the failure of the expectation of reciprocity that this process involves. To the extent that a failure of reciprocity is a sign of the decline of general adherence, this failure may free me morally from the constraints the convention imposed. Because the moral advantage of individual adherence depends on there being general adherence, in the absence of general adherence, the convention loses its moral point. In short, when there are only a small number of players, the failure of another party to adhere to the convention may be sufficient to free me from my convention-dependent obligations under it.
In a response to Mavrodes, Robert Fullinwider has argued that there is a moral basis for discrimination in the principle of self-defense, according to which, “only those are liable to be killed who pose the immediate and direct jeopardy.”5 In war, combatants pose such jeopardy, while noncombatants do not. So discrimination is, after all, a moral convention, being directly supported by the moral rule allowing self-defense. The principle of self-defense establishes a different form of liability to force than that of moral fault discussed by Mavrodes. As many commentators have pointed out, the innocence of noncombatants in Fullinwider’s sense is based etymologically on the Latin nocens, doing harm; the innocent are those not doing harm. This is a different notion of innocence than the absence of fault. The principle of self-defense permits the use of force without a finding of fault. I will refer to the moral liability of a combatant to attack based on self-defense as defensive liability. In contrast, liability based on moral fault I will refer to as fault liability. These two forms of liability often overlap, but not always. Sometimes those who have fault liability do not also have defensive liable, such as the guilty noncombatant, or those who have defensive liability do not also have fault liability, such as the misinformed conscript or, more generally, the innocent attacker. (I will assume here without argument that defensive liability is distinct from fault liability in that one is entitled to defend against an innocent attacker.)
Michael Walzer, our third figure from the 1970s, sees discrimination as having direct moral relevance, or (in my terms) as a moral convention. Concerning the part of discrimination that permits attacks on combatants, Walzer seems, like Fullinwider, to appeal to defensive liability in that combatants are permitted to defend themselves from attacks by the other side. Considerations of fault cannot justify attacks on combatants because combatants “fight without freedom,”6 and so presumably without fault. Concerning the part of discrimination that prohibits attacks on noncombatants, however, Walzer seems to base part of his appeal on fault liability, specifically, as articulated through a notion of human rights. All persons have the human right not to be attacked, and noncombatants have done nothing to lose that right; they are not at fault because they have not made a choice to fight. Civilians “have done nothing, and are doing nothing, that entails the loss of their rights.”7
There are two distinctions important in this debate, (1) that between combatants and noncombatants, which I have been discussing, and (2) that between combatants fighting a just war (just combatants) and those fighting an unjust war (unjust combatants). Walzer argues that distinction (1) has direct moral relevance, but that distinction (2) does not. Unjust combatants are allowed to attack just combatants, as well as the converse, so in this sense the distinction between them makes no moral difference. The lack of direct moral relevance of the second distinction for Walzer is implied by two theses he holds. The first is the “moral equality of combatants” according to which combatants “face one another as moral equals” because (when war is a crime) “it is not their crime.”8 The second is the “independence thesis,” which is that “considerations of jus ad bellum and jus in bello are logically independent.” Were combatants not moral equals, the two aspects of just war theory would not be independent because one could not apply jus in bello rules without appealing to jus ad bellum determinations. These two theses depend on rejecting the moral import of the second distinction, but what would happen if that distinction were morally relevant?
Walzer’s position in defending the independence thesis is in one respect an anomaly within the historical sweep of the just war tradition. Until the modern era, just war theorists largely ignored jus in bello; to the extent it was considered, it was regarded as secondary and derivative from jus ad bellum. In contrast, in the modern era, with the emphasis on the prerogatives of sovereignty, the idea of “regular war” came to the fore, for example, in the claim that raison d’etat was a sufficient justification for war, making jus ad bellum irrelevant and leading to the flourishing of jus in bello. The twentieth century saw a return to interest in jus ad bellum, culminating in the “post-war settlement” of the UN Charter outlawing aggression. In Walzer’s theory we see an effort, through the independence thesis, to continue a vigorous jus in bello along side the resurgent jus ad bellum. But is this position sustainable, or is there some logic recognized by the historical alteration in ascendency between the two aspects of just war theory? This is the issue I seek to get at through the question whether discrimination is a mere convention or a moral convention.
A set of arguments have been advanced over the past fifteen years by several just war theorists, most notably Jeff McMahan, that argue against the independence thesis and in favor of the view that, in my terms, discrimination is a mere convention. These arguments were explored in the recent volume edited by Henry Shue and David Rodin, Just Warriors. I see the arguments as, to some extent, a reassertion of the historical pattern that only one of the two aspects of just war theory can be in ascendency at a time, that the falsity of the independence thesis shows that one will always be secondary or derivative of the other. But even apart from this grand historical claim, it is worth considering whether the arguments against the independence thesis are tenable.
McMahan rejects the independence thesis and the moral equality of combatants.9 He argues that the second distinction, that between just and unjust combatants, not the first, is morally central. Fault liability is the relevant criterion, and the relevance of the second distinction is that just combatants are not at fault (so long as they fight justly) because their cause is just. Because just combatants are not at fault, there is no moral basis for attacking them. Unjust combatants are at fault, and hence liable to be attacked by just combatants, so long as they have chosen to fight a war they know (or should know) is unjust. If unjust combatants have not freely chosen to fight an unjust war, due either to duress or ignorance, then they are excused from fault liability, though their lack of fault does not justify them in fighting, as the just combatants are. So, there are three sorts of combatants to be concerned with: (1) just combatants, (2) voluntarily unjust combatants, and (3) nonvoluntarily unjust combatants. Those in (2) have fault liability, and those in (1) are justified in attacking them. Being excused, those in (2) have no fault liability. Those in neither (2) nor (3) are permitted to attack those in (1). McMahan argues that this is just like the domestic case where individual A attacks individual B: A is not permitted to continue his attack on the grounds that he is defending himself against B’s defensive actions.
In a sense, Walzer and McMahan make a different appeal to the “domestic analogy” of self-defense. Walzer applies the analogue of individual self-defense at the jus ad bellum level, but not the jus in bello level, while McMahan applies it at both levels. For McMahan, as only the defender state is allowed to engage in war, so only the combatants of that state are allowed to fight. In contrast, Walzer, speaking of the jus in bello level, observes: “The domestic analogy is of little help here. War as an activity (the conduct rather than the initiating of the fighting) has no equivalent is settled civil society . . . The moral equality of the battlefield distinguishes combat from domestic crime.”10 It should be noted that Walzer’s approach to jus in bello is a greater departure from domestic morality than the notion of defensive liability may suggest. As McMahan observes, self-defense is not a principle that can be appealed to by a domestic attacker who suddenly finds herself under threat by the defensive actions of her victim. So the everyday ideas of self-defense and defensive liability do not get us to the independence thesis nor to the moral equality of combatants; these ideas permit a defender to attack an aggressor, whether or not at fault, but not an aggressor to counter-attack against a defender. The notion of defensive liability we need to represent Walzer’s position is stronger than the everyday one, in that it permits aggressors to counterattack against defenders, making combatants morally equal. Defensive liability cannot be completely domesticated. I will refer to this stronger sense of defensive liability as defensive liability-2.
Walzer asserts that “jus in bello represents an adaptation of morality to the circumstances of combat, to the heat of battle.”11 Interestingly, McMahan would assent to this statement, though understand it differently. McMahan contrasts the conventions of war with what he refers to as the “deep morality” of war, which is the set of the moral rules that apply to individual interactions, such as his principle of fault liability. Although discrimination “is false as a criterion of moral liability to attack in war, it ought nevertheless to be upheld as a convention to which all combatants are bound.”12 The reason is that the “rules of war have to accommodate our epistemic limitations and to be formulated with a regard for the ways in which their announcement is likely to affect people’s behavior.”13 In practice, applying deep morality directly to military conduct is both difficult and dangerous, given likely consequences. Approaching the moral issues of fighting pragmatically, McMahan believes, will lead us to adopt a set of conventions, such as discrimination, that are at odds with the moral rules that apply to war. For McMahan, discrimination is a mere convention, adopted because of its consequentialist value, and yielding only convention-dependent moral obligations.
So the dispute between Walzer and McMahan represents a version of the question whether discrimination is a mere convention or a moral convention. Walzer views it as a moral convention, while for McMahan it is a mere convention, distinct from and at odds with the “deep morality” of the moral rules that apply to war. There are, however, for McMahan, consequentialist reasons to ignore the direct moral rules and apply instead the conventions such as discrimination. As he notes: “The laws of war, in effect, are conventions in Mavrodes’ sense that are justified by their utility. Only by permitting what is immoral can they best fulfill their morally sanctioned purpose.”14 So both Walzer and McMahan are committed to the application in practice of discrimination. This has led Walzer to the criticism that McMahan’s proposal represents a distinction without a difference. McMahan argues in response that individual combatants should be guided by their consciences in cases where deep morality diverges from convention and the rules of deep morality can be applied. There are some cases in practice where they would diverge, for example, when one is fighting an army of child soldiers.15 In any case, I would argue that the theoretical difference between the positions is sufficient reason to see the distinction as making a difference.
There are a series of criticisms of McMahan’s position I will now rehearse. Some are stronger than others, but cumulatively, I believe, they raise serious doubt about the tenability of his position.
(1) The first criticism raises the question whether fighting in a defensive war is morally permissible for McMahan. I assume that unless he can show that fighting a defensive war is consistent with the deep morality of war, this would be an important problem with his view. The difficulty is that his criterion of fault liability not only does not permit attacking just combatants, but also may not permit attacking unjust combatants. Many unjust combatants are without fault not because they are fighting a just war, but because they are not voluntarily fighting an unjust war, due either to their (reasonable) belief that the war is just or to their fighting under duress. Moreover, there is no practical way that an attacker can separate those with fault from those without. As a result, it is not clear that any action against unjust combatants would pass a proportionality test (or an application of the doctrine of double effect). To avoid this problem, McMahan makes two claims about unjust combatants. First, they are generally at least partly at fault, partly morally responsible, for the unjust harm they threaten; none of them is equivalent to an implacable pursuer or an innocent bystander.16 Second, because of this partial responsibility, an unjust combatant may be attacked by a just combatant on the reasonable presumption of the unjust combatant’s responsibility.17 But the first claim, on which the second depends, is not very plausible. Many of those who fight are young and uneducated, so attributing any significant degree of partial responsibility to them seems a stretch.
(2) The second criticism concerns a further problem with McMahan’s appeal to fault liability. Fault liability is the basis for punishment. In the context of war, it is usually seen as applying postbellum in war crimes trials, where it presupposes a judicial determination of fault. In terms of the domestic analogy, contexts of self-defense operate differently—given the exigency of the attack, there is no opportunity to determine fault, and defensive liability may justify defensive force in the absence of fault. (In war, defensive liability is easy to determine—it applies to those who are wearing a uniform and bearing arms.). So, the problem for McMahan is that fault liability, conceived as liability to punishment, seems out of place in questions of jus in bello liability to attack. But McMahan doesn’t view defensive force as punishment. He distinguishes between two forms of fault liability, which he calls simply liability (I’ll call this liability-M) and desert. Desert justifies punishment, while liability-M justifies force only when necessary to avoid the harm that the person who is liable is threatening.18 With liability-M, McMahan introduces a form of liability that is more suited than desert to be the basis of jus in bello liability to attack. But this does not avoid the problem, as a determination of fault is still necessary. In addition, the use of liability-M in determining jus in bello liability to attack has a strongly counterintuitive implication, though one that McMahan embraces. In removing the condition of proximate cause that characterizes defensive liability, liability-M implies that culpable noncombatants and prisoners of war are liable to attack.19 Prisoners of war who voluntarily participated in fighting an unjust war and civilians who were strong advocates of an unjust war may be liable to attack by just combatants should it be militarily advantageous to attack them.
(3) The third line of criticism considers some problematic implications of McMahan’s view that the moral obligations of the rules of war are convention-dependent. McMahan sees the conventions, and the reciprocity they involve, as mutually beneficial. Due to their mutual benefit, “all have a moral reason to recognize and abide by” them. But if reciprocity fails on one side, “it may cease to be rational or morally required for the other side to persist in its adherence to them.”20 I referred to this earlier—convention-dependent obligations hold only so long as the convention exists. If the convention is abandoned, and reciprocity fails, the obligation goes with it. In contrast, moral rules represent obligations that do not depend on their being adhered to by others or their having the status of a convention. Discrimination, however, seems to be the sort of thing that combatants are obligated to observe even if the other side does not, implying that the obligation is convention independent.
For example, consider Walzer’s discussion of reprisals. The usual justification for reprisals is “they did it first,” but, Walzer argues, “we must condemn all reprisals against innocent people.”21 This means failure of reciprocity resulting from one side’s violation of discrimination cannot justify the other side’s violation of discrimination, making the obligation to respect this principle convention-independent. McMahan might offer the counter-argument that, should discrimination be abandoned as a convention, combatants would still be bound by the deep morality of war, which would, for the most part, imply that noncombatants should not be attacked. But two points may be made in response. First, the phrase “for the most part” in the previous sentence is crucial. As we have seen, McMahan admits that the deep morality of war permits some attacks against noncombatants; but discrimination does not. Second, McMahan’s counter-argument may not work in the case of other rules of war, like the rule against harming prisoners of war. Defensive liability explains this rule well—POWs are not a threat. But if this rule were a mere convention, as McMahan suggests,22 then, should it fail, for example, by the other side’s adopting a policy of killing our combatants it has captured, then we would be no longer morally obligated not to kill their combatants, should there be a military advantage in doing so. But this seems wrong. The same argument might be made using the example with which this paper began, torture.
A related problem with the idea that the rules of war are convention dependent is that they may lack the plausibility that moral rules need to have to achieve general adherence. Walzer observes: “No limit is accepted simply because it is thought that it will be useful. The war convention must first be morally plausible to large numbers of men and women; it must correspond to our sense of what is right.”23 While the conventions of war for McMahan have the deep morality of war as a background, they depart from deep morality for pragmatic or consequentialist reasons, and, as a result, they require combatants to do what is contrary to the deep morality of war, as McMahan acknowledges.24 So the conventional rules may lack plausibility. McMahan’s response is to deny that moral plausibility is required for the general acceptance of a conventional rule of war, citing as an example the prohibition on the use of chemical weapons.25 This prohibition, he claims, is generally accepted, so presumably morally plausible, despite its lacking direct moral justification. So a mere convention, lacking moral plausibility, can be generally accepted. But this one example is a thin reed on which to base his denial, and, in any case, it seems not to be a counterexample at all. Contrary to his claim, the negative attitude regarding the use of chemical weapons (and other WMDs) is generally based on a view that these weapons are inherently immoral. As there is a “nuclear taboo,” a rule against the use of nuclear weapons founded on an attitude of moral revulsion, so it seems there is a chemical taboo, similarly founded. Indeed, despite the fact that chemical weapons may do less damage than conventional explosives, there is an argument, based in part the close connection between chemical weapons and civilian damage, that they are inherently immoral.26
(4) My final criticism concerns McMahan’s efforts to bolster his view about discrimination’s being a mere convention by appealing to a comparison between these conventional rules of war and domestic law. He observes that “domestic law cannot simply restate the principles of individual morality, because the declaration and enforcement of laws have effects that must be taken into account in the formulation of the law,” and he argues that this is the case with the law of war.27 Some domestic laws conflict with morality because there is a need to balance pragmatic or consequentialist considerations with deontological moral rules in the formulation of individual laws in contexts in which extraneous factors influence people’s responses to the rules. For example, selling adulterated milk is a strict liability offense, and so departs from a moral perspective, which would assign penalties only where there is fault. But basing liability on fault in the milk case would not be very effective because merchants could too easily choose to sell the bad milk and dissemble about their knowledge of the fact. The conventions of war and the morality of war diverge in a similar way, McMahan argues. As fault is ignored in the law against selling bad milk, so it may be ignored in discrimination convention. Given the gap between morality and domestic law, we should not be surprised that there is a gap between the morality of war and the law of war as well.
But there is a crucial disanalogy that vitiates McMahan’s argument. Most of domestic criminal law accords nicely with morality, for example, in the mens rea requirement for fault; strict liability is strictly limited to a small part of the law. But, on McMahan’s view, the discrimination convention completely cuts off the law of war from considerations of fault. Combatants are strictly liable and noncombatants are strictly nonliable, across the board. Domestic law is a piecemeal departure from morality in its concern with fault, while, on McMahan’s view, discrimination puts the law of war in wholesale departure in this regard. Some years ago, Lady Barbara Wootton proposed that all criminal statutes be matters of absolute or strict liability because that is the best way to reduce the number of socially damaging actions. But most people react to such a policy with moral horror. On McMahan’s view, however, the law of war is completely separated from moral concerns of fault, and is equivalent to Wootton’s proposed policy. Domestic law is mostly malum in se, but the law is war, given the principal role of that discrimination plays in it, is, on McMahan’s view, primarily malum prohibitum. Thus, the analogy between the law of war and domestic law does not well support his claim that the law of war is a matter of mere convention separate from the deep morality of war.
Related to the issue of the role of fault in domestic law and the law of war, is the contrast between warring and policing. If warring and policing are morally different enterprises at a fundamental level, this would leave room for discrimination to be a moral convention, distinct from the strictures of fault liability, rather than a mere convention. If warring were policing, fault would be morally basic in war. McMahan doubts that warring and policing are separate matters: “I think that just war is police action of a sort while unjust war is criminal action of a sort.”28 It may be that war should be police action, that is, that it should no longer be war, but this is not the case in our world lacking a global executive authority.
Consider the contrast between McMahan and Walzer on the comparison of the moral relation between a police guard and a bank robber and the moral relation between a just combatant and an unjust combatant. The robber is not permitted to fire at the guard, even if the guard fires first, because the guard is defending the bank’s money against an unjust attack. McMahan says that, at the level of deep morality, the two combatants have the same sort of moral relation—the unjust combatant is not permitted to fire on the just combatant. In contrast, Walzer says that the unjust combatant does not do wrong in firing on the just combatant, for “in our judgments of the fighting, we abstract from all consideration of the justice of the cause.” So, the nature of the moral constraints on the bank robber is not a proper guide to the nature of the moral constraints on the unjust combatant: “there are rules of war, though there are no rules of bank robbery (or of rape or murder).”29 Policing and warring are different kinds of activities from a moral point of view, and different basic moral rules may apply. Despite the fact that the conventions of war are at odds with the morality that applies to domestic interactions, especially regarding the role of fault, this is not a sufficient reason to treat discrimination as a mere convention. Fault may simply play a different role in the morality of war than it plays in the morality of domestic relations.
I want to conclude in a moment with some speculations on where these arguments leave our understanding of the morality of war, but let me first return briefly to the comments of Alberto Gonzales that the Geneva Conventions are anachronistic. I suggested that they could be so only if they were mere conventions. The conclusion of my subsequent argument is that there is good reason to regard those rules, at least discrimination, as more than mere conventions, but as moral rules justified in themselves. Jus in bello holds its own as a moral theory; it has its own integrity. This helps us respond to Gonzales. He is wrong to abandon the Geneva Conventions as a matter of basic morality, not simply wrong, as Colin Powell would have it, as a result of misunderstanding the importance of maintaining reciprocity regarding the conventions.
The traditional rules of war, at least discrimination, proscribe actions that are malum in se, not malum prohibitum. As the example of Gonzales is meant to suggest, there are strong pragmatic reasons for treating discrimination as a basic moral rule. Allowing one side to abandon discrimination, if the other side does so, opens up war to escalating civilian destruction. To take the example peculiar to Gonzales, we cannot justify torture on the grounds that the enemy practices it. The proscription of torture is not a mere convention, but, like the prohibition against attacking noncombatants, is a basic moral rule. Further risks ensue from abandoning the independence thesis, which McMahan also recommends. Doing this opens the door to the likes of General Sherman, who can argue, because of the dependence of in bello judgments on ad bellum judgments, that the rightness of the Union cause in the American Civil War can justify the burning of Atlanta. Recognizing the independence thesis and the idea that discrimination is a basic moral rule seem not only more accurate as moral theory, but also the best way to constrain the violence of war by insulating the rules of combat from political exigencies at the ad bellum level.
Finally, a bit of speculation on where this discussion leaves the morality of war. Let us assume that discrimination is a basic moral rule, not a mere convention. What should we then say about the difference and the contrast between this rule, which relies on defensive liability-2, and the moral rules applying to interpersonal interactions, which rely mainly on fault liability? I will briefly discuss two possibilities, each of which, in a different way, embodies Walzer’s claim, cited earlier, that “jus in bello represents an adaptation of morality to the circumstances of combat, to the heat of battle.”30 The first possibility is that what appears to be two different sorts or realms of morality, that of war and that of ordinary life, can be united under a rule-utilitarian sort of approach. The second is that war is in a strong sense a different moral realm, one to which different basic moral rules apply.
(1) According to McMahan, the law of war consists in mere conventions, like discrimination, that achieve the best results in reducing human suffering and, in particular, better results than adhering to ordinary morality (the deep morality of war). The conventions and basic deontological moral rules are at odds with each other. But for the rule-consequentialist, or anyone who fashions moral rules on the maximization of moral value, the rules that achieve the best results are in fact the correct moral rules. Given that war involves such a radically different form of human interaction than everyday life, it would not be surprising, from the perspective of the rule-consequentialist, that the moral rules of everyday life are different from the moral rules of war. It is not surprising, for example, that, while fault is morally generally dispositive in everyday social life, it is not in war, since, while it is plausible that rules respecting fault maximize utility in ordinary social life, they clearly do not in warfare.
So, the contrast McMahan points to between deep morality and the law of war is not a contrast between morality and mere convention, but rather between one set of correct moral rules and another, distinguished by the very different circumstances in which they hold. Under a rule-utilitarian perspective, Richard Brandt suggests, the conflict McMahan represents as between the laws of war and the morality of war is rather a conflict between the moral rules that apply to war and the moral rules we are socialized into (“ideal rules of conscience”) because they are the best moral rules for everyday life.31 Henry Shue, though he does not adopt a rule-utilitarian perspective, says something similar. In criticizing McMahan’s distinction between the morality of war and the law of war, Shue observes: “The circumstances of war are so different from the context of ordinary life that even when the same fundamental moral touchstones are the reference, the differences in the circumstances yield different specific guidelines.” The proper conclusion is that “morality is all of a piece;” there are not “two separate moralities, one inside war and one outside.”32
(2) Here I get more speculative. What might be said for the claim that there are two separate moralities, one inside war and one outside? What if the moral rules of war and the moral rules of everyday life are not of a piece, as they are for the rule-consequentialist, but are separate and independent, belonging to distinct moral sites or moral realms? This is the second general approach to understanding the difference between discrimination as a moral convention and the moral rules of everyday life. The general idea that there are separate sites of morality with different and independent moral rules is central to John Rawls’s theory of justice as a political virtue applying to the basic structure of society and not derivable from moral principles applying to individual interactions in social life. Rawls asserts that “the correct regulative principle for a thing depends on the nature of that thing.”33 For Rawls, Thomas Nagel observes, “justice should be understood as a specifically political value, rather than being derived from a comprehensive moral system.”34 Nagel again: “There is no single level of full moral concern, because morality is essentially multilayered.”35 Thomas Pogge refers to the different sites of morality in Rawls’s theory as “morally deeply distinct.”36
Liam Murphy calls Rawls’s view dualist. Dualism, according to Murphy, is the “claim that the two practical problems of institutional design and personal conduct require, at the fundamental level, two different kinds of practical principle.”37 In contrast, monism is the view that there are no separate sites of morality in this sense, or, as Murphy says, no “separate normative realm, requiring separate normative first principles.”38 (I will follow Nagel in using the term “anti-monism” instead of “dualism,” since the number of distinct sites may be more than two.) Nagel applies anti-monism at the international level, using it as a move in the debate over global distributive justice, criticizing a monistic, cosmopolitan approach, countering it with a statist, dualistic approach instead. Rawls applies his anti-monistic view to international relations in a similar way, also adopting an anti-cosmopolitan stance, denying that the moral rules of justice that apply to the domestic basic structure apply as well at the global level.39
I propose that we consider an analogous form of anti-monism as regards the contrast between the moral conventions of war, such as discrimination, and the moral rules of normal interpersonal interactions. The idea is that war is morally sui generis, that discrimination, in particular, is a moral rule specific to war and not applicable to, nor derivable from, moral rules of ordinary social life. In his discussion of global justice, Nagel points out that the moral appeal of cosmopolitanism is that the great difference in average wealth among members of different societies seems from a moral perspective “highly arbitrary.”40 This has a counterpart in our debate. The fact that the principle of discrimination makes liable to attack the innocent combatant as well as the guilty combatant, ignoring the everyday moral concern with fault, also seems, from a moral perspective, highly arbitrary. Such an appearance is doubtless behind the concerns McMahan brings to the development of his version of jus in bello. But, if war is a “morally deeply distinct” site from that of everyday morality, there would be no reason to think that the morality of the former would have to correspond to the latter.
But is this effort to apply anti-monism to the rules of war anything more than a suggestive analogy? Can the idea of anti-monism apply beyond the concerns about distributive justice to which Rawls and Nagel put it? Murphy remarks that “almost all the substantive arguments for and against dualism that I am aware of concern distributive justice.”41 Consider the genus of anti-monism, of which its application to distributive justice is a species. The idea of the social contract in Rawls is that cooperation among the members of a social group yields benefits and burdens, and the moral question of justice arises—how should those burdens and benefits be distributed, given the major impact that that distribution has on the life prospects of the members. It is the institutions of society, what Rawls calls the basic structure, that distribute the burdens and benefits, so justice is a matter of determining the rules by which the distribution should be carried out. These rules, Rawls claims, apply to the institutions, not directly to the behavior of individuals. This is why the site of justice, the basic structure, is a realm of moral rules separate from that of individual interaction.
We could tell the same kind of story about war. War is a set of institutions in which we are, for good or for ill, enmeshed. It likewise involves benefits (such as national self-determination) and burdens (such as death, suffering, and destruction), so rather heavy on the burdens. The impact of these institutions on individual life prospects is great. As such, the institutions of war also raise distributive questions. From the original position, one can argue for discrimination as the key rule of distribution. Under the veil of ignorance, those in the original position would not know the likelihood of their being a combatant or noncombatant, or of their being participants in a just or an unjust war. But they do know general facts, such as that combatants are often conscripted or otherwise pressured into the military and that combatants often, for a variety of reasons, have false beliefs about whether or not the war they fight is just. I suggest that in this choice situation, people would choose discrimination because they would reason on prudential grounds that the best thing for them would be to be immune from attack, were they noncombatants, and to be allowed to fight for their lives and for their nation’s cause, were they combatants. Thus, the notion of fault, which plays such a major role in the morality of individual interactions, would not play this role in war.
1 I would like to thank Seth Lazar, who provided me with helpful comments on an earlier draft of this paper.
2 Quoted at http://www.americanprogress.org/issues/kfiles/b79532.html.
3 Quoted at http://www.americanprogress.org/issues/kfiles/b79532.html
4 George Mavrodes, “Conventions and the Morality of War,” Philosophy & Public Affairs 4, no. 2, pp. 117-131.
5 Robert Fullwinder, “War and Innocence,” Philosophy & Public Affairs 5, no. 1, pp. 90-97.
6 Walzer, Just and Unjust Wars, p. 37.
7 Michael Walzer, Just and Unjust Wars (New York: Basic Books, 1977), pp. 145, 146.
8 Walzer, Just and Unjust Wars, pp. 127, 37.
9 This discussion is drawn from the following works of Jeff McMahan: “Innocence, Self-Defense, and Killing in War,” Journal of Political Philosophy 2, no. 3 (1994), pp. 191-221; “The Ethics of Killing in War,” Philosophia 34 (2006), pp. 23-41; “On the Moral Equality of Combatants,” Journal of Political Philosophy 14, no. 4 (2006), pp. 377-393; “Liability and Collective Identity: A Response to Walzer,” Philosophia 34 (2006), pp. 13-17; and “Killing in War: A Reply to Walzer,” Philosophia 34 (2006), pp. 47-51.
10 Walzer, Just and Unjust Wars, pp. 127-28.
11 Michael Walzer, “Terrorism and Just War,” Philosophia 34 (2006), pp. 3-12, passage at p. 12.
12 McMahan, “Ethics of Killing in War,” p. 38.
13 McMahan, “Killing in War: A Reply to Walzer,” p. 47.
14 McMahan, “Innocence, Self-Defense, and Killing in War,” p. 209.
15 McMahan, “Killing in War: A Reply to Walzer,” p. 48.
16 McMahan, “Ethics of Killing in War,” pp. 32.
17 McMahan, “Innocence, Self-Defense, and Killing in War,” p. 214.
18 McMahan, “Innocence, Self-Defense, and Killing in War,” pp 203-4.
19 McMahan, “Ethics of Killing in War,” pp. 35-36.
20 McMahan, “Ethics of Killing in War,” p. 38.
21 Walzer, Just and Unjust Wars, chapter 13, passages at pp. 210, 215.
22 McMahan, “Killing in War: A Reply to Walzer,” p.49.
23 Walzer, Just and Unjust Wars, p. 133.
24 McMahan, “Innocence, Self-Defense, and Killing in War,” p. 209.
25 McMahan, “Ethics of Killing in War,” p. 39.
26 See Steven P. Lee, “Weapons of Mass Destruction,” in Larry May (ed.), War and Political Philosophy (forthcoming, Cambridge University Press).
27 McMahan, “Killing in War: A Reply to Walzer,” p.47.
28 McMahan, “Liability and Collective Identity,” p. 13.
29 Walzer, Just and Unjust Wars, pp. 127-28.
30 Walzer, “Terrorism and Just War,” p. 12.
31 Richard Brandt, “Utilitarianism and the Rules of War,” Philosophy & Public Affairs 1, no. 2 (1971-72), pp. 145-165.
32 Henry Shue, “Do We Need a ‘Morality of War’?” in Henry Shue and David Rodin (eds.), Just and Unjust Warriors: The Moral and Legal Status of Soldiers (Oxford: Oxford University Press, 2008), pp. 87-111, quotations at pp. 87, 88.
33 John Rawls, A Theory of Justice, revised edition (Cambridge MA: Harvard University Press, 1999), p. 25.
34 Thomas Nagel, “The Problem of Global Justice,” Philosophy & Public Affairs 33, no. 2 (2005), pp. 113-147, at p. 120.
35 Nagel, “Global Justice,” p. 132.
36 Thomas Pogge, “On the Site of Distributive Justice: Reflections on Cohen and Murphy,” Philosophy & Public Affairs 29, no. 2 (Spring, 2000), pp. 137-169, quotation at p. 154.
37 Liam Murphy “Institutions and the Demands of Justice,” Philosophy & Public Affairs 27, no. 4 (1998), pp. 251-291, quotation at 254.
38 Murphy, “Institutions,” p. 253. Murphy himself argues against dualism in favor or monism.
39 In John Rawls, The Law of Peoples (Cambridge, MA: Harvard University Press, 1999).