In their revealing and highly important analysis of American public opinion about the ethics of war, Scott Sagan and Benjamin Valentino report three main findings: (1) that most Americans reject the traditional doctrine of the moral equality of combatants; (2) that they tend not to recognize duress as a condition that excuses or mitigates the responsibility of soldiers who fight for unjust aims (“unjust combatants”); and (3) that they tend to think it justifiable to punish soldiers who act in ways that revisionist just war theory judges to be wrong.Footnote 1 The first and third of these results suggest, surprisingly, that Americans have more intuitive sympathy with revisionist just war theory than with traditional just war theory. The second is also more closely aligned with revisionist than with traditional thought; yet it should be worrying for revisionists because, like the third finding concerning punishment, it suggests that Americans who are sympathetic to revisionist views have a harsher, more retributive understanding of those views than any revisionist theorist would find acceptable. The startling nature of these results and the careful methods by which Sagan and Valentino have revealed them demand that we examine them more closely.
The Moral Equality of Combatants and Excusing Conditions
Of the Americans who responded to Sagan and Valentino's survey, 78 percent believe that unjust combatants who are volunteers (“unjust volunteers”) do not act ethically, while 81.5 percent believe that unjust combatants who are conscripts (“unjust conscripts”) do not act ethically. These beliefs, along with those shown in figure 6 of their article about the justifiability of punishing both unjust combatants and “just combatants” (those who fight for just aims), are inconsistent with an acceptance of the moral equality of combatants. Sagan and Valentino also take them to show that the experimental subjects do not accept that the duress involved in conscription “mitigate[s] individual soldiers’ moral responsibility for participation in unjust wars” (p. 414). This is because a higher percentage say that unjust conscripts do not act ethically than say that unjust volunteers do not.
But here I think that Sagan and Valentino's term—“act ethically”—is misleading. To deny that soldiers act ethically is to imply that they act unethically, or wrongly. But to believe that unjust conscripts act wrongly is compatible with believing that they are excused by the duress involved in conscription, or that their responsibility for their wrongdoing is mitigated. Indeed, the belief that acts of war by conscripts are wholly or partially excused presupposes the belief that those acts are wrong; but it includes the additional belief that there are circumstances that exempt the conscripts, either wholly or partially, from blame or punishment. Believing that conscripts do not act ethically is therefore a necessary condition of believing that there are excusing conditions that make them less blameworthy than unjust volunteers, or not blameworthy at all.
There is, moreover, evidence in Sagan and Valentino's data that many of their subjects do recognize the duress involved in conscription as a mitigating condition. In figure 6, for example, while a slightly higher percentage of subjects say that unjust conscripts do not act ethically, significantly fewer say that it would be justifiable to imprison unjust conscripts than say that it would be justifiable to imprison unjust volunteers, and only about half as many say that it would be justifiable to execute unjust conscripts as say that it would be justifiable to execute unjust volunteers. Figure 6 also shows, as the authors comment, “that subjects were more inclined to imprison leaders for initiating unjust wars than they were to imprison soldiers for fighting in such wars” (p. 431). This might be explained by the belief that soldiers often have excuses that political leaders lack. It is these beliefs about the justifiability of punishing those already identified as wrongdoers that may best reveal whether the subjects, and by extension Americans in general, accept that there are “contextual factors that . . . mitigate individual soldiers’ moral responsibility for participation in unjust wars” (p. 414).
There is, however, other evidence not shown in any of the figures. In discussing their second main finding, the authors report that “less than half of the public (45 percent) agreed with this statement: ‘Soldiers in an all-volunteer army are more ethically accountable than soldiers who are drafted into service if they follow orders to fight in a war of aggression’” (p. 428). Being ethically accountable is the antithesis of being excused and both are matters of degree. Thus, the 55 percent who believe that unjust conscripts are no less accountable than unjust volunteers seem committed to believing that the duress involved in conscription is not an excusing or mitigating condition for wrongly fighting in a war of aggression. This seems broadly consistent with the evidence in figure 6; together these results suggest that close to half the subjects believe that duress diminishes the responsibility of unjust conscripts, but slightly more than half believe that it does not. The latter belief is curious, as duress is recognized as a mitigating or excusing condition for offenses, including homicide, in domestic criminal law.
Sagan and Valentino did not probe for whether their subjects accept that unjust combatants might be excused on other grounds also recognized by revisionists, such as nonculpable ignorance or diminished responsibility. Sometimes two or more of the various commonly recognized excusing conditions occur together in war—for example, when a government systematically lies to its citizens both about the facts and about its reasons for going to war, and also conscripts very young people into its armed forces with threats of severe punishment for noncompliance. In these cases, the excusing or mitigating conditions are usually additive, in that the force of each combines with that of the others. Hence, even if each on its own is only somewhat mitigating, together they may be highly excusing. If the authors had explored their subjects’ beliefs about the other excusing conditions in addition to duress, it is possible that they would have found stronger support among the subjects for the idea that unjust combatants are often not highly culpable, or perhaps not culpable at all, and thus might not deserve punishment, or might deserve only some lesser punishment.
There is, in fact, evidence in Sagan and Valentino's figures that perhaps as many as one in five of the subjects recognize that unjust combatants may have excuses other than duress. Figure 2 shows that 22 percent think that unjust volunteers act ethically and thus do nothing for which they could be excused. Figure 6 shows that 56.9 percent believe that it is justifiable to imprison unjust volunteers. I interpret this as indicating that they believe that it is justifiable to punish unjust volunteers after the war has ended, though not necessarily by means other than imprisonment. (It is entirely coherent to believe that it is justifiable to imprison combatants even when it is not justifiable to punish them. This is the virtually universal view of the imprisonment of prisoners of war until the end of a war. But this is irrelevant here.) But if subjects believe that it is justifiable to punish unjust volunteers, this suggests that they believe these volunteers are not excused. There is therefore a total of 78.9 percent who seem to believe that unjust volunteers are not excused, leaving 21.1 percent who believe that they act unethically, or wrongly, but ought not to be punished. There are of course reasons why it may not be justifiable to punish those who do wrong other than that they are excused for their wrongdoing. But if the reason it is unjustifiable to punish unjust volunteers is that there are excusing conditions for their wrongdoing, those excusing conditions are likely to be different from that which might be thought to excuse the action of conscripts—namely, duress. (It is also possible that some of the 56.9 percent who believe that it is justifiable to punish unjust volunteers accept either that they acted ethically or that they are excused, and thus do not deserve punishment, but believe that it is justifiable to punish them anyway—for example, for the purpose of deterrence. This, I concede, weakens the force of my point here.)
Extremism and Confusion
I will return to the subjects’ views about punishment in the next section. In this section I will discuss a few peculiarities of the views that the authors’ questions elicited in order to illustrate what seems to me a high degree of confusion in the general public's thinking about these issues. In particular, I will highlight the surprising percentages of views at the extreme ends of the range of possible views about issues in the ethics of war. For example, 12.5 percent of the subjects, or one in every eight, say that a state is justified in fighting a war of unprovoked aggression (figure 1). And that percentage increases to 14.3, or one in seven, if the war of aggregation includes a massacre of innocent women and children (figure 1). Similarly, 22 percent say that soldiers who fight in a war of unprovoked aggression act ethically (figure 2), while 37.2 percent say that soldiers who fight in a war of defense against aggression do not act ethically (figure 6). Finally, 21.7 percent say that these latter soldiers who fight only in just defense may justifiably be punished, while 13.8 percent say they may justifiably be executed (figure 6). All this is bizarre. And it is incompatible with any view that has been defended in the tradition of thought about just wars.
An Amoral Public?
Sagan and Valentino were sufficiently puzzled by such views that they later conducted a further survey that revealed that 68 percent of the American respondents—almost three quarters—agreed that “in war, the strong do whatever they can and the weak do whatever they must. Ethics just don't apply” (pp. 431–432). The second of these two sentences indicates an unambiguous amoralism about war. But if this amoralism was comparably prevalent among the subjects in the authors’ original survey, most of those subjects ought to have been unable to answer questions about whether soldiers or political leaders “acted ethically” in war; for they believe that the concepts of ethical and unethical action have no application in war. Possibly when they answered questions about whether certain acts were justified, they understood justification in prudential or rational terms, according to which a state or soldier is justified in acting when the action serves the national interest. Yet this supposition is undermined by the evidence in figure 1, which shows that far more subjects (70.2 percent) think a war is justified when it is defensive than think it is justified when it is aggressive (12.5 percent think an aggressive war is justified if it is fought by volunteers, while 5.7 percent think it is justified if it is fought by conscripts). Similarly, far more (73.1 percent) say that a state whose soldiers massacre civilians is justified in fighting when its war is one of defense than say that it is justified when its war is one of aggression (14.3 percent). These views are difficult to explain unless the subjects understood “justified” to mean “ethically justified.”
A Pacifist Public?
Amoralism about war is at one end of the spectrum of possible views about the morality of war. Pacifism is at the other. The views reported in figures 1 through 6 in Sagan and Valentino's article seem, on the whole, anything but pacific. But that appearance is deceptive. Consider, for example, the subjects’ responses to a claim about justification that does not appear in any of the figures—namely, “Eastland's soldiers were justified in killing the Westrian troops at the military base because, if they had not done so, the Westrian troops would have killed them” (p. 426). The authors note that 60.3 percent agreed with this statement when Eastland's war was defensive, or just, but only 31.6 percent agreed with it when the war was aggressive, or unjust. The remaining 68.4 percent of the subjects—more than two-thirds—who disagreed with this statement when the war was aggressive, or unjust, thus “accepted the underlying revisionist logic that soldiers who are engaged in an unjust war cannot claim the right of self-defense as justification for killing their just adversaries” (p. 426).
This is a pleasant surprise for revisionists, but the other percentage (60.3) is rather baffling. It implies that the remaining 39.7 percent who disagreed with the statement when the war was defensive, or just, believe that a just combatant may not kill an unjust combatant even if doing so is necessary to defend his own life. Consider what this means, given the assumption, which is common both in military planning and in just war theory, that any killing of an enemy combatant contributes to making victory more probable. This assumption is of course defeasible in special circumstances, but the killing of enemy combatants is usually a quite general aim of military action in war. It seems natural, therefore, that subjects would assume that when just combatants kill unjust combatants who would otherwise have killed them, they are both preserving their own lives and making the achievement of their war's defensive aims more likely. If 39.7 percent of the subjects believe that such killing would not be justified, they must certainly believe that it would not be justified if it were unnecessary for individual self-defense but would merely contribute to the just goal of national self-defense. So, unless these subjects believe that there is some aim of war other than defense against unprovoked military aggression that can make it justifiable for soldiers to kill enemy combatants, they must believe that killing in war can never be justified—that is, they must be pacifists.
This surprising inference is reinforced by the findings shown in figure 5. In case B, as I noted earlier, 37.2 percent of respondents do not accept that soldiers acted ethically by fighting in a war of defense against unprovoked aggression. Figure 5 further shows that 41.2 percent do not accept that the political leaders acted ethically in ordering the defense of their country against that aggression. These responses, too, are strongly indicative of pacifism. And these percentages are very close to the percentage of subjects (39.7) who, as I noted in the preceding paragraph, deny that it is justifiable for a just combatant fighting in a war of defense to kill an enemy combatant in self-defense. Indeed, as by now one might have come to expect, the percentage of those who say that it is unjustifiable to kill an unjust combatant in the course of a just defensive war when doing so is necessary for individual self-defense is slightly higher than the percentage of those who say it is unjustifiable to do the same act in the same type of war when it is not necessary for individual self-defense. In other words, for a small percentage of respondents, that a killing would save the life of a just combatant counts against its being justifiable.
The responses just noted suggest that roughly 40 percent of the subjects are pacifists. In figure 1, however, only 29.8 percent say that a state is unjustified in fighting defensively against unprovoked military aggression (though, as we have seen, 2.9 percent seem to have overcome their scruples and abandoned their pacifism when they considered a war of defense that included a massacre of civilians). Taken at face value, the evidence from the survey thus suggests that between 30 and 40 percent of the subjects are pacifists.
But we know, both from our general knowledge of Americans and from other evidence in the survey, that the percentage of Americans who are pacifists is substantially smaller than that. In particular, even 30 percent cannot be pacifists if 68 percent are amoralists about war; for that would leave only 2 percent who accept that there can be both just and unjust wars.
A Confused Public
What these various baffling percentages suggest is that many of the subjects who participated in the survey are deeply confused in their beliefs about the ethics of war. This is evident throughout the figures. In figure 1, for example, between 70.2 and 73.1 percent say that Eastland, the state, is justified in fighting a war of defense against aggression. Yet in figure 2, only 62.8 percent say that Eastland's soldiers acted ethically in participating in such a war. Similarly, in figure 5 only 58.8 percent of subjects say that Eastland's political leaders acted ethically in ordering the war that between 70.2 and 73.1 percent believe that Eastland is justified in fighting. One wonders how almost 10 percent could believe that soldiers act unethically in fighting a war that their state is justified in fighting, and how around 12 or 13 percent could believe that political leaders act unethically in ordering that same justified, defensive war to protect their citizens.
Punishment
Let us next consider the American public's views, shown in figure 6, about the punishment of a state's soldiers by the enemy state in the aftermath of a war. There is much that is shocking here, but I will confine my remarks to two topics: (1) the subjects’ views about punishment of those who fight in wars that involve the massacre of civilians; and (2) the subjects’ apparent failure to consider the implications of their expressed views for recent actual wars or wars that might occur in the foreseeable future.
Figure 6 shows the extent to which the subjects agree or disagree, in the five different cases, with the statement “After the war, Westria would be justified in seeking prison terms for/trying to execute the Eastlandic soldiers/leaders who carried out/ordered the attack against its military base” (p. 430). I will discuss only the views expressed about soldiers in cases D and E, which both involve a massacre of civilians (which I will henceforth refer to as “war crimes,” following the terminology of the authors). In case D, soldiers fought in a war of unprovoked aggression during which their side massacred forty-eight civilians. Of those surveyed, 73.1 percent think it justifiable for the state that these soldiers attacked to imprison them, and 49.6 percent think it justifiable for the state to execute them. In case E, soldiers fought in a war of defense during which their side massacred forty-eight civilians. In this case, 50.6 percent say that it is justifiable to imprison these soldiers and 32.5 percent say it is justifiable to execute them. (The percentages in each case are not mutually exclusive; hence those who indicated a belief in the justifiability of execution are presumably also among those who indicated a belief in the justifiability of imprisonment.) In case D, the percentage of those who approve of punishing the soldiers is higher than the percentages of those who approve of punishment in cases A and C, which are also wars of aggression but do not involve war crimes; and the same is true in the comparison between case E and the corresponding case B.
What is notable about these views, in addition to their harshly vengeful character, is that they seem to reveal a high level of acceptance of the justifiability of collective punishment. What the subjects were asked about is the justifiability of punishing all the soldiers who fought in a war. Because these soldiers are described as having occupied 100 square miles of territory, killed five hundred enemy soldiers, and taken thousands more prisoner, and because the cases were intended to be realistic, subjects should have inferred that a large number of soldiers fought in each of the wars described. It would thus have been unrealistic for them to suppose that every soldier who participated in the war also participated in the massacre of the forty-eight civilians. It is therefore reasonable to suppose that, when the subjects were asked about the justifiability of punishing the soldiers, their responses referred to the punishment not only of the perpetrators of the massacre but also of the presumably much greater number of soldiers who were not among the perpetrators. This suggests that those who thought punishment would be justified in the cases involving war crimes but not in the corresponding cases without war crimes had one of two beliefs. They may have believed that the nonperpetrators deserved to be punished solely because of their membership in the group that included the perpetrators, or they may have believed that it is better to punish a larger number of innocent nonperpetrators than to allow a smaller number of perpetrators to go unpunished (the inversion of Blackstone's dictum that “it is better that ten guilty persons escape than that one innocent suffer.”Footnote 2) Both of these beliefs are, in my view, mistaken.
When Sagan and Valentino discuss the cases involving war crimes, they refer to the subjects’ views about “soldiers fighting for an unjust cause who committed war crimes” and “[the soldiers] who had executed forty-eight civilians” (p. 431). These descriptions suggest that the views elicited were about the perpetrators only. But the wording of the statements to which the subjects responded did not allow them to distinguish between perpetrators and nonperpetrators. Perhaps many of them would have distinguished morally between the two groups if they had had that option. But, in the absence of that option, many expressed support for collective punishment (which seems to encompass what the authors refer to as “vicarious retribution” [p. 432]Footnote 3).
Sagan and Valentino write that “we chose to focus [our stories] on a hypothetical conflict scenario, using two imaginary countries to help minimize the possibility that the subjects’ knowledge of either country's previous behavior or their subjective loyalties to one side might bias their ethical judgments” (pp. 420–421). This choice was, I believe, correct and unavoidable, but it may have elicited a different and subtler form of bias—a bias toward severity arising from the subjects’ failure to reflect on the application of their views to actual wars or wars that might occur in which the United States or one of its allies was or would be a belligerent.
Consider, for example, the 21.7 percent of subjects who believe that a state guilty of unprovoked aggression is justified in imprisoning volunteer soldiers who resisted the aggression without committing war crimes, as well as the 13.8 percent who believe that the state is justified in executing them (figure 6). Perhaps these subjects did not know that when Germany invaded France in 1940, the defending French forces were volunteers. But when that fact is taken into account, the belief expressed by 21.7 percent of respondents logically commits them to the view that the Nazis would have been justified in imprisoning French prisoners of war, and the belief expressed by 13.8 percent commits them to the view that the Nazis would have been justified in executing the French prisoners. It is, however, difficult to believe that even the most hardened amoralists about war could really accept these implications of their expressed beliefs.
It is, moreover, even more difficult to believe that many of the subjects would have retained their expressed views about punishment if they had been compelled to consider cases, especially actual cases, in which the captured soldiers were Americans. Consider, for example, the views, shown in figure 6, that the subjects expressed about case D, in which volunteer soldiers participate in a war of unprovoked aggression and some subset of them intentionally kill civilians. Of the subjects, 73.1 percent say that it would be justifiable for the state that was the victim of the aggression to imprison the soldiers they captured and 49.6 percent say that it would be justifiable to execute them—presumably, in both cases, irrespective of whether the soldiers were among the perpetrators of the war crimes.
Next, notice that the description given of the Eastlandic invasion of Westria in case D also applies, in a general way, to the American invasion of Iraq in 2003. Iraq had not attacked the United States or any other state. It did not, as the U.S. government alleged, have weapons of mass destruction that it might have used to attack the United States. Yet the United States invaded Iraq and occupied its territory, including its oil fields, and American soldiers intentionally killed civilians on numerous occasions. The Iraqi government was, of course, a barbarous dictatorship that treated many of its citizens abominably, but it is consistent with the description in case D to suppose that Westria had a similar government; for no information about the Westrian government is supplied in the descriptions of the cases.
It seems, therefore, that 73.1 percent of the subjects have committed themselves to the view that Iraq would have been justified in imprisoning captured American soldiers and that 49.6 percent are committed to the view that Iraq would have been justified in executing them. Had these subjects been forced to confront these implications of their views, it is highly likely that most would have softened their views considerably. On this point, I am in strong agreement with one of the main claims of Michael Walzer's excellent commentary in this symposium.
Implications for Just War Theory and the Law of War
Sagan and Valentino summarize their main conclusions for just war theory in several places. These conclusions are that just war theory should not aim to codify existing moral intuitions about the ethics of war but should instead “serve as a check” on these intuitions, which would thereby help to restrain soldiers from engaging in atrocities, and that the law of war ought not to be revised to accommodate the central tenets of revisionist just war theory. In this final section, I will offer a few comments on these conclusions.
First, Sagan and Valentino are right that moral philosophy generally, and just war theory in particular, should not seek to codify our existing moral intuitions. People's intuitions, both individually and collectively, are to varying degrees confused, inconsistent, and incomplete. Most moral philosophers nevertheless accept that our intuitions are essential elements in moral theorizing, as they represent the ways that things appear to us to be, morally. Moral philosophy should aim, therefore, to school our intuitions, testing them for consistency with one another and with plausible moral principles, and in the process rejecting some, revising or refining others, and affirming those that survive the theoretical winnowing.
This process of sustained moral examination and reflection is necessary if we are to avoid engaging again in the atrocities and “terrible deeds” that, as the authors observe, our instincts or intuitions have prompted us to commit in the past. Yet we should not assume that we can confidently identify which deeds are terrible independently of this process. Pacifists think that any act of war is a terrible deed. Traditional just war theorists, particularly those who are moral absolutists, think that any intentional killing of a civilian is a terrible deed, even, for example, the killing of a scientist who would otherwise provide a demonic leader with a weapon of mass destruction. I think these views are mistaken, but this has to be established by rigorous and scrupulous moral argument.
I and a number of other moral philosophers believe that the principles of revisionist just war theory are better supported by careful philosophical argument than those of traditional just war theory. Sagan and Valentino are right, however, that we should not assume that it would be desirable for these principles to be incorporated into the law, at least in current conditions. Law is an instrument for inducing us, with all our cognitive, moral, and motivational infirmities, to act in morally acceptable ways. We should not expect that the best way to do this is to translate the correct moral principles directly into law and threaten people with punishment for violating them. Indeed, the philosopher Henry Sidgwick argued that what he believed to be the correct moral theory might itself imply that it ought not to be widely promulgated. “A Utilitarian,” he wrote, “may reasonably desire, on Utilitarian principles, that some of his conclusions should be rejected by mankind generally; or even that the vulgar should keep aloof from his system as a whole, insofar as the inevitable indefiniteness and complexity of its calculations render it likely to lead to bad results in their hands.”Footnote 4
Revisionists do not, however, wish that most people should “keep aloof” from their understanding of the ethics of war. Revisionists should, for example, want their view that virtually all acts of war by unjust combatants are morally impermissible to bear on the consciences of soldiers ordered to fight in a war they believe to be unjust. And they may rightly hope that the revisionist understanding will guide the design and development of institutions—such as, in Sagan and Valentino's words, “an authoritative judicial body that could adjudicate rival claims about the justice of war causes” (p. 418)—that might eventually enable the law to be brought into closer congruence with morality.Footnote 5
The one general point that the authors make about revisionist just war theory that I would dispute is the claim in their concluding section that “our results suggest . . . that the logic of revisionism” threatens to “undermine the protection of noncombatants during war” (p. 435). It is true that, unlike traditional theorists, revisionist theorists tend to accept that in some instances civilians can make themselves morally liable to be harmed in war. But this possibility does not arise in any of the cases discussed in the article. In cases A, B, and C, it is explicit that there are no civilian casualties, and in the two cases, D and E, in which there are killings of civilians, the killings are described in a way that makes it clear that they cannot be justified on grounds of liability; rather, they constitute a massacre of the innocent that revisionists and traditionalists alike condemn. So while it may be true that revisionist just war theory rejects the traditional theory's grounds for distinguishing morally between combatants and civilians, I do not think that the results of Sagan and Valentino's survey add anything to the concern that general acceptance of revisionism would erode existing protections for civilians in war.
