Since then there have been numerous significant books and a plethora of influential articles (some by Walzer himself ), but now the American philosopher Jeff McMahan has written an important, highly intelligent book that challenges, or seems to challenge, many of the basic tenets of contemporary just war thinking. McMahan’s Killing in War does indeed confront received interpretations of the just war tradition, but the scope of his challenge is more restricted than it appears at first blush (or than he is inclined at times to claim). This is because there are elements in the tradition that actually support his critique (and to which he occasionally appeals), and also because he qualifies his challenges in ways that seem to blunt much of their force. On the first point, his critique is best seen as mostly directed at Walzer’s version of just war thinking and the moral underpinnings of current international law related to warfare that Walzer’s theory largely supports.

Although McMahan’s book is densely packed with intricate argument, its most basic claims can be summarised fairly straightforwardly. A key idea in much contemporary just war thinking is that soldiers are entitled to kill enemy soldiers but that they are not entitled intentionally to kill enemy civilians, though there are some complexities involved in these entitlements since, for instance, soldiers cannot kill enemy troops who have surrendered, and it is permissible to kill civilians in some circumstances where their deaths are incidental to attacks upon military targets. The first entitlement is endorsed by Walzer with a doctrine of ‘the moral equality of soldiers’. As for the second entitlement, the civilian deaths are said to be foreseen but not intended, and are permitted when ‘proportional’ and otherwise unavoidable. Some contemporary philosophers reject the significance of the ‘incidental’ permission because they reject the double effect doctrine or similar theories with which it is usually supported, but McMahan does not follow them. His critique of the key idea proceeds by arguing that both the permissions and restrictions on killing in war are based upon the wrong moral premise.

According to McMahan, the standard view relies upon the idea that the justification for killing in war is an entitlement to kill those who are harming or posing a threat of harm. Combatants, as fighting men, fit that description, and non-combatants do not. Hence, combatants are entitled to kill enemy combatants no matter whose cause is just, but are not entitled to kill non-combatants. Against this, McMahan, drawing on the model of individual self-defence, argues that the entitlement to kill applies only against unjust attackers, so that unjust warriors have no moral right to kill just warriors. Furthermore, if some civilians are sufficiently implicated in the unjust attack, then they too lose their immunity from lethal violence. So soldiers are not ‘morally equal’ and not all civilians deserve immunity.

Critics will object in a variety of ways. One way is to challenge the idea that the moral rights and restrictions on killing in war can be modelled on the simple scenario of individual rights to self-defence. War is surely unique and collective in ways that defy such assimilation. McMahan responds by insisting that both involve the killing of human beings in ways that require justification and that the mode of justification surely remains constant at the individual and collective levels, even if there are some differences in treatment to be taken into account. This is a complex issue, but I think McMahan is essentially right, as is indicated by the way that supporters of war frequently meet pacifist objections by citing the validity of individual rights of self-defence against unjust lethal attack. Critics are on safer ground in objecting to the unworkability of McMahan’s position, and I will return to this later.

McMahan frames his account in terms of the concept of liability and argues that ‘the criterion of liability to attack in war is moral responsibility for an objectively unjustified threat of harm’. This criterion is part of his rejection of the interpretation of ‘innocence’ that current just war theorists use. They insist that, in the context of war, innocence does not mean morally innocent but is closer to its Latin root in meaning ‘not harming’, or, as McMahan often puts it, ‘not posing a threat of harm’. Hence they conclude that opposing soldiers are not innocent and can be liable to attack because they are engaged in harming. For McMahan, however, soldiers fighting in an unjust cause are not innocent, because they are morally responsible for an objectively unjustified threat of harm, while those who have a just cause to resist them are innocent and not liable to attack. In McMahan’s support, there is the fact that when I defend myself by violence against a robber who unjustly attacks me with a lethal weapon, my justified violence against the assailant surely gives him no right to defend himself against me; he should just stop his unjustified attack.

But some would argue, like Walzer, that, in the case of war, the unjust warriors are too ignorant or coerced to be regarded as responsible agents. McMahan responds that this is an overgeneralisation, and that the most it could show is that some of the unjust soldiers are partially excused of guilt for their actions, but that they have sufficient moral responsibility to be liable to defensive violence. It is an overgeneralisation in that some soldiers know enough about what they are doing to realise that they shouldn’t be doing it and many are not coerced, especially in the absence of conscription. Even where the factors of ignorance and/or coercion obtain, McMahan claims, it is more likely that they are only partially excused, since at best only extreme coercion can excuse unjustified killing and those who want to fight in a war should realise that good moral reason for fighting is hard to come by: indeed, it is plausible that most wars in history have been unjust on one or both sides.

As for civilians, McMahan argues that civilian status is not enough to clear people of some degree of moral responsibility for an unjustified war and hence, in some cases, of liability to attack. There are some instances where this is persuasive: the political leadership that decides on the war, scientists who devise new weapon technologies for fighting the war, and media bosses who deliberately whip up enthusiasm for the war. Elsewhere, I have argued that such categories of civilian should be viewed as combatants, given the role of strong agency that they adopt, so I do not accept the identification of non-combatant with civilian (as McMahan does). But terminology aside, many of those non-combatants whom McMahan deems to some degree morally responsible and hence potentially liable to lethal attack do not seem to me at all plausible candidates for this status. Those who vote for politicians who run for election on a war platform (amongst other policies), or who write letters of support to relatives at the front, or who simply don’t do anything to oppose the war, may or may not fall short of an ideal of moral behaviour, but spreading the net of moral responsibility for the war and moral liability to harm to cover them seems to me dangerously moralistic. McMahan even thinks that citizens of neutral countries who could do something that might stop the unjust war-making bear some moral responsibility for it. I think it damages the seriousness and utility of our concept of moral responsibility to spread it so thin.

McMahan pursues various arguments and distinctions about this matter, some of which seem more plausible than others. He is at his best when dealing with real examples, such as the status of armed Israeli settlers in occupied territories, but like much current analytic philosophy, his arguments are peppered with artificial examples and appeals to ‘our’ intuitions about them. Sometimes this technique works to his advantage, sometimes not. He claims, for instance, that a motorist whose car unaccountably veers out of control and is about to kill or maim a pedestrian has enough moral responsibility to be liable to being intentionally killed by the pedestrian if killing the driver would prevent the harm to the pedestrian. This is because even a totally non-negligent, careful driver knows that driving the safest of cars carries with it a tiny risk of harming by malfunction. But even if we ignore the improbability of a pedestrian’s being in a position to save herself in this way, it is surely extravagant to impute any degree of moral responsibility to the driver on the basis McMahan chooses. My own intuition, contrary to McMahan, is that the pedestrian has no right to kill the driver. The example is of course quite dissimilar to the unjust soldier who, whatever his excuse, is actually intending to kill the just warrior.

In any case, the complex arguments and distinctions about this issue do not have the radical practical conclusions that appear at first blush. On the liability of civilians to attack, McMahan concedes that there are powerful factors that together entail that ‘civilians are almost never liable to intentional military attack, and even … when some are liable it is still generally impermissible to attack them’. These factors include the generally low level of moral responsibility of most civilians, the great uncertainty of such an attack’s military effectiveness, and the difficulties in telling which civilians are responsible among those subject to attack. And this is the moral situation. When we come to the legal regulation of war, there are further pragmatic considerations that ‘argue decisively for an absolute, exceptionless legal prohibition of intentional military attacks against civilians’.

On the moral equality of soldiers, McMahan’s strong position is maintained at one level, but diluted at another. He insists that the moral equality thesis is false and provides a rich and sustained line of argument against it, but he concedes that there are pragmatic reasons for maintaining international law’s strict version of a legal equality of soldiers. This is partly because of difficulties in determining which side is warring unjustly, partly because of the extent of partially excused combatants on any unjust side, and partly because of the likely bad effects of allowing different legal status and permissions to one side and not the other. Nonetheless, there remain important consequences of maintaining the strong moral line while distinguishing it from the best regulatory régime. A principal one is that it rightly puts more pressure on citizens to make their own judgements about the rightness of any war in which they are called upon to serve, since they should recoil from the prospect of unjust killing, and hence it gives added weight to the demand for selective conscientious objection laws.

This is a good book, well-informed, carefully written and full of insight, scholarship and tough argument. It will certainly stimulate extensive debate amongst philosophers, but its proliferation of distinctions, intricate dissection of arguments, and recourse at times to arcane examples is likely to restrict its appeal to a wider intellectual audience.