Terrorism versus Collateral Damage: What is the Scope of Civilian Immunity in War?
Whitley R.P. Kaufman
One of the most firmly established elements of the Just War tradition is that of noncombatant immunity. Those who are not directly participating in the fighting (I will use the term ‘civilian’ interchangeably with ‘noncombatant’) are not to be targeted; this rule prohibits any direct attacks on the person or property of civilians, or indeed any action taken with the intention of causing harm (physical or even psychological) to noncombatants. A fundamental dilemma of war, however, is that civilians will inevitably be hurt, no matter one’s intentions. How then is it possible to fight a just war? The traditional answer, codified in the Doctrine of Double Effect, is that justice requires that one not intend harm against civilians. Harm that is merely accidental, or even that is foreseen but not intended, is permissible, though of course regrettable and to be avoided where possible. The basic dividing line for just conduct in war, then, is that between harm which is intended and harm which is at most foreseen but not intended.
The intend/foresee distinction has long been criticized as morally arbitrary; harm is harm, the argument goes, and it is little consolation to the civilian that it was not intended. To undertake an action knowing that it will harm innocent persons is wrong and should not be sharply distinguished from harm which is strictly intended. Recently, a group of critics of the DDE including Colm McKeogh and Charles Dunlap have proposed an alternative conception of noncombatant immunity which does away with the intend/foresee distinction. For McKeogh, even foreseen killing of civilians violates principles of justice, and the Just War doctrine needs to be tightened to this extent. However, McKeogh et al. propose simultaneously loosening the doctrine at another point, allowing non-lethal harm (specifically, property damage) against civilians. Thus under their proposed revision, the significant moral distinction is not between intended versus foreseen harm, but between lethal and non-lethal harm. In wartime, they argue, it is never legitimate to kill civilians (other than accidentally), but it should be legitimate to target civilian property as a way of pursuing victory in war.
I will argue in this essay against such a revision to Just War doctrine. Let me first state some limitations on my aims in this paper. First, I will not explicitly defend the intend/foresee distinction on moral grounds. Even though I think the distinction is morally legitimate, there is no space here to defend that view; my focus will rather be on whether the proposed alternative is morally satisfactory. Second, I will assume here that the debate takes place within the framework of those accepting Just War constraints; to the Realist I have nothing to say here for the most part (Dunlap’s argument appears to be essentially a Realist one, so for this reason my argument will focus more on McKeogh). Third, I will assume a deontological framework as the foundation of the Just War doctrine; to the consequentialist again I have nothing to say (McKeogh too adopts the deontological view, so our dispute is not over this question). Finally, I will assume that the basis for noncombatant immunity is not the moral “innocence” or guilt of civilians vis-à-vis soldiers, but rather that killing soldiers in wartime is justified only on self-defense grounds. As noncombatants pose no immediate, imminent threat to life or limb, they are not permissible targets during war. This again is a controversial assumption (McKeogh appears to reject it, and Dunlap in fact bases his argument on the fact that civilians are not truly “innocent”), but there is not sufficient space to defend the position fully here.
First, a brief statement of the four provisions of the Doctrine of Double Effect. This doctrine permits an action which has both good and bad effects so long as the following conditions are satisfied:
1) The action must not be wrong in itself. For example, strategic bombing of legitimate military targets in a just cause is not itself wrongful.
2) The agent must not intend the bad effects. Thus, the strategic bomber must not intend the harm to the civilians.
3) The bad effects must not be the means to the good effect. In strategic bombing, the good effect is the destruction of the legitimate targets; harm to civilians is a side effect but not the means by which the legitimate military targets are destroyed.
4) The good sought must be proportionately important relative to the bad effects that are foreseen. The destruction of legitimate military targets must be a sufficiently weighty value to justify harm to civilians; a target of trivial value would not justify significant harm to civilians.
If all four of these conditions are met, then the action is justified despite the foreseen harm to civilians. This position remains, as I have said, the dominant one in Just War theory.
McKeogh claims that the DDE, in allowing foreseen civilian deaths, is morally unacceptable. He argues that it is “callously indifferent to the killing of civilians in war as a foreseen side effect of one’s action” (168). Alternatively, McKeogh claims, if one reads the DDE (especially condition 4) as a consequentialist condition, it is morally unacceptable in that it permits “trading innocent lives off against some other good” (id.). Either way, it fails to treat the individual person as justice demands: that “no one should be punished except for their [sic] own crime, that innocent life should never be taken as a means towards an end, and that the individuality of all should be acknowledged and respected” (165). McKeogh is skeptical about whether there is a moral justification for the principle of noncombatant immunity (149). But he does appear to accept at least the basic moral idea that the “sacrifice of innocent lives to a good cause is wrong,” and that it is wrong to “treat people only as a means to an end” (173), and this principle, he argues, is violated when military forces cause foreseen civilian casualties. He therefore proposes abandoning the intend/foresee distinction and banning any actions which foreseeably cause civilian deaths.
McKeogh goes further, however, and (following Dunlap) appears to offer a sort of trade-off to the military. While he insists on taking away the military’s right to cause foreseen civilian deaths, he at the same time suggests granting the military the right to destroy civilian property: “A strategy of destroying civilian property may be an effective and technologically feasible way of avoiding the unjust taking of human life” (171). International law strictly prohibits any attacks on innocent persons, including their property (as does morality; attacks on innocents or their property even for a good cause would violate the DDE). McKeogh approvingly quotes an author also cited by Dunlap, that “hurting a civilian’s pocketbook is more ethical than bombing him” (171). He proposes, along with Dunlap, that civilian property be a legitimate military target, so long as that property is not indispensable to survival. Taking a life, after all, is a far greater wrong than destroying property.
Let me turn now to evaluating this proposal. First, an initial clarification is necessary. McKeogh and Dunlap both explicitly state their proposals as prohibitions solely against killing civilians. Oddly, they do not specifically exclude non-lethal attacks on civilians, i.e. attacks which cause non-lethal physical harm to noncombatants. Obviously, this cannot be correct, since their proposals as stated would license such crimes as wounding, maiming, even rape and torture of civilians. But even if we modify their proposal to permit attacks only against civilian property and never against civilian persons, there are still troubling questions. Would this mean that one may destroy food and water supplies, shelter, and medicines so long as the harm is non-lethal (e.g. merely pain, suffering, temporary or even permanent disability)? This too seems far too harsh. Their proposal as it stands is far too vague for us to know whether it is wholly morally justified. But for the purposes of this discussion, we will assume that they mean that one may destroy civilian property only if it will not cause either death or permanent and severe physical harm. (Analogously, one presumes that McKeogh means to prohibit not only foreseen civilian deaths, but also foreseen physical harm to civilians, though again he does not address this point). As we will see below, however, the unclarity on these issues raises serious questions about the moral basis for their proposed revisions.
What then could be a convincing moral justification for this radical change in Just War doctrine, which prohibits foreseen lethal harm to persons, but permits intentional destruction of civilian property as a means to victor in war? It is rather difficult to say just what McKeogh’s justification is, but I will try here to identify some of his arguments. The first justification McKeogh attempts is that “death is different to any other harm; it cannot be offset by any good” (165). Lost property can be replaced or compensated, but an individual who is dead cannot be restored to life. The difficulty with this argument is that it proves too much. I mentioned above the odd fact that McKeogh does not mention non-lethal, physical harm to civilians, and here we might have the explanation why. The “death is different” justification of course prohibits only lethal harm, and not non-lethal but appalling direct physical harm to civilians (torture, rape, maiming, etc.), since one can provide compensation for such harm after the fact. Thus the “death is different” is simply implausible on this ground alone. But even more important, the fact of compensability is by itself inadequate as moral justification. For it simply ducks the question as to whether a particular action is right or wrong; my ability to compensate someone does not in general give me moral permission to harm him. The fact that an unjust action is compensable does not make it permissible. This is, after all the Just War doctrine we are talking about, not the Compensable War doctrine.
A second possible justification rests on the claim which both McKeogh and Dunlap quote from Stauffer that harming property is “more ethical” than harming persons. Here I must simply confess to puzzlement as to what he could mean by the phrase “more ethical.” Right and wrong, just and unjust, ethical and unethical are all qualitative, not quantitative, terms. An action is either ethical or it is not; there is no third alternative. Indeed, on McKeogh’s view, killing 10 civilians would be more ethical than killing 20, and maiming would be more ethical than killing. The apparent guiding principle here is that it is okay to do wrong, since one could have done much worse. But this is simply to confuse the amount of harm one causes, with whether one has done wrong. One suspects that the phrase “more ethical” is a rhetorical device which in effect concedes that even causing civilian property damage is a moral wrong.
Indeed, on McKeogh’s own principles, such actions appear to be clear moral violations. For they violate the moral prohibition against using innocent persons as a means. As McKeogh asserts: civilians “have done, and are doing, nothing that warrants attack on them”; hence “the most fundamental principles of justice…require that civilians be immune from targeting in war. Civilians may not be targeted whatever the consequences” (165). But if civilians are immune from being killed because they have done nothing wrong, then it is hard to see why they are subject to having their property destroyed. (Conversely, if it is morally permissible to destroy their property, on what grounds is it prohibited to target their persons?) It is notable that Dunlap offers a straightforwardly consequentialist justification for the permission to harm civilian property. McKeogh has foreclosed this justification for himself by his rejection of consequentialist arguments. Yet it is difficult to see what, other than a consequentialist argument, could suffice to justify this clear harm against those “who have done nothing to deserve it” (173). (The “death is different” argument also founders on this same point).
It is of course true that, all things being equal, it is better always to do less harm than more. But that does not render causing the lesser harm permissible, unless one is in a situation of the “lesser of two evils”, where whatever one does is wrong. Since there is always the option of doing neither of the wrongful actions, one can hardly invoke a “lesser evil” defense. Here, one need not kill civilians, nor need one harm their property. (There are cases where it is permissible to destroy property, e.g. pitching goods overboard during a storm to save a ship. But this is a genuine case of the choice of evils in an emergency situation.)
Indeed, this raises one of the most puzzling points in the argument. Both McKeogh and Dunlap present the issue as if it were one of a choice between either allowing foreseen civilian deaths, or permitting destruction of civilian property. For Dunlap, we should “replace the killing of people with the destruction of objects” (16; emphasis added); for McKeogh, in some conflicts, “genuine civilian immunity could come at the expense of civilian property” (171). But this is surely a fallacy of false alternatives, as if we have to make a choice between foreseen deaths and intentional property damage. In fact, the two have nothing to do with one another.[i] If either one is morally wrong, then Just War doctrine forbids it. The only sense I can make of this argument is that it seems to be a significant concession to Realism or to practical necessity. The idea is perhaps that military commanders are unlikely to give up the right to cause foreseen civilian casualties, unless they are given some alternative that compensates them for this restriction, or perhaps that civilians would be happy to give up rights over their property in return for a guarantee of personal safety. It appears to be a straightforward example of a trade-off , in McKeogh’s view: “attacks on civilian property would be welcome [!] if they were accompanied by a genuine immunity for civilian life in war” (172). For Dunlap, the shift seems to be an instance of the gradual heightening of values in war; destroying objects instead of people “would seem to represent a positive step toward reducing human suffering in armed conflicts” (16). However, Dunlap does not appear to see this new permission to destroy property as merely temporary, but rather a “valuable tool for the forces of the moral against those of the amoral” (17).
McKeogh may well be right that it would be impractical to demand the end of foreseen civilian deaths without giving the military some sort of tradeoff. I also leave aside the issue of whether, as Walzer argues, urgent circumstances justify overriding moral constraints, so that in some circumstances it may be permissible to target civilian property. My concern here, however, is to point out the moral non sequitur in the present argument. Unless we are given a positive argument as to why the destruction of civilian property is morally right, we cannot use it as a bargaining chip to end the practice of allowing foreseen civilian deaths. The two issues are morally distinct, and each has to be evaluated on its own terms.
is perhaps a further reason why both Dunlap and McKeogh insist on seeing this
issue in terms of a moral trade-off. My
guess is that each of them realizes that disallowing even foreseen civilian
deaths (or even, as I have argued, foreseen physical harm to civilians) would
so cripple the war effort that it might even threaten our ability to win in
wartime, and that some alternative effective tool such as the targeting of
civilian property is necessary as replacement.
In fact, it is quite possible that a ban on foreseen harm to civilians
would undermine the possibility of strategic bombing even more than these
critics realize. But even apart from jus
Perhaps McKeogh would respond that he is not referring to speculative possible future deaths, but concrete identifiable immediate harm. I doubt that this response will succeed; civilian casualties in wartime are not merely speculative but quite predictable, and it is hard to see how the temporal gap between the decision to go to war and the occurrence of the deaths is morally relevant. One should not overstate the difference between a decision to go to war, and a decision to engage in a particular act of strategic bombing; in both cases the extent of the civilian harm is speculative and not knowable with any kind of certainty. McKeogh might also say that the possibility of civilian deaths is acceptable so long as one attempts to minimize civilian casualties wherever possible – but that is precisely what the current rules already mandate; the strategic bomber must attempt to minimize civilian harm.
Nor should one underestimate the effect that McKeogh’s new rules would have on the war against terrorism, and the need for a clear distinction between just and unjust uses of violence. This is an issue both from a practical and theoretical standpoint. Practically speaking, laws and statutes regarding terrorism would have to be rewritten, for they currently almost universally ban attacks against property as well as civilian life. Indeed, there are terrorist groups (such as the Environmental Liberation Front) that adopt a policy of only damaging property, never human or animal life. While I am glad that they respect the sanctity of life, I am not sure we want to go on record as condoning the destruction of property even in the name of a good cause. Even more important however is the theoretical issue. The war on terrorism requires moral distinctions to be as clear and comprehensible as we can make them. What is wrong with terrorism is not that the terrorist lacks a just cause (he may well have a just cause), nor that his action takes place in or out of wartime, nor that he deliberately takes human life (soldiers in wartime deliberately take human life). Terrorism is wrong because it involves intentionally harming innocent people – i.e. those that are not an immediate threat to one’s safety. To harm innocent people physically, or to harm them indirectly by attacking their property, is terrorism either way: it is using innocent people as a means to one’s ends. McKeogh’s alternative moral rule would, I fear, confuse this fundamental principle, by allowing that it is morally permissible to deliberately harm the innocent (so long as it is non-lethal harm).
But if McKeogh’s proposal is ultimately unsatisfactory, that begs the question as to whether he still has a valid objection against the DDE and current Just War doctrine. While I cannot enter into this issue in detail, I do want to briefly respond to his charge that the permission to cause foreseen civilian death betrays “callous indifference” to human life. In fact, I think this characterization represents a common misunderstanding of the DDE: that, so long as one does not intend the harm, it is simply not one’s concern or responsibility. Michael Walzer repeats this misunderstanding when he insists that the DDE be supplemented with the principle that “the foreseeable evil be reduced as far as possible” (155). But a more plausible reading of the DDE holds that such a principle is already implicit in (or even explicit in) the second provision of the doctrine. Thus the New Catholic Encyclopedia gives provision two as following: “The agent may not positively will the bad effect but may permit it. If he could attain the good effect without the bad effect he should do so” (Vol. 4). That is to say, the agent always has the duty to minimize the bad effects. The action is permitted only where the good effect cannot be obtained without the bad effects also occurring – and unless the good effect is proportional to the bad effect (condition 4). As many theorists have emphasized, in wartime this means that a combatant must be willing to accept a substantial increase in risk to his own safety in order to protect innocents from the risk of harm; it also requires that one take any possible steps to protect civilians (leaving an escape route from a city under siege, etc.). Suzanne Uniacke has insisted that the DDE does not deny that one is responsible for the foreseen deaths; and there is no reason to think one should not regret these deaths. I do not accept McKeogh’s characterization of this doctrine as expressing callous indifference.
Finally, a word about the meaning and purpose of the DDE. The doctrine is consciously formulated as a response to a radically imperfect world, in which even the best of intentions cannot avoid harm, in which good and evil are tragically mixed and combined. To aspire to moral perfection, wherein one could act only where one was certain of never causing harm, would be to undercut the possibility of any significant action at all. To act in this world is to risk harm, even great harm. Indeed, to undertake an action which one knows will cause the ultimate harm – the death of innocent people – is to take on a terrible responsibility. It is one which should be taken in “fear and trembling” and only adopted if the good being sought is of such great importance that it justifies the harm being caused. The DDE, and Just War theory in general, acknowledge this tragic necessity
[i] A clear example of the false alternatives fallacy is Dunlap’s claim that we should “honor the value of human life over property interests” (16). Current law does not “choose” between lives and property.