In Defence of Killing the Innocent, Deliberately But Not Intentionally

In his generous review of my book, In Defence of War, Christopher Tollefsen objects to my understanding of intention, which he thinks is, at least, unfortunately ambiguous or, more probably, mistaken. Whether or not it is mistaken, I fear that my view is too deliberate to find an excuse in ambiguity.

Tollefsen agrees with my claim that soldiers should not intend to kill their enemy. Since such a view both has the temerity to dissent from Thomas Aquinas and risks appearing empirically implausible, I am reassured to have at least one ally. Nevertheless, Tollefsen takes issue with my account of intention and with its implication for what may be done to innocent civilians. He articulates his objection partly with reference to the case I drew from the film version of Patrick O’Brian’s novel, Master and Commander. Here is the case as I presented it:

… the ship is engulfed in a severe Atlantic storm. One of her masts is broken and crashes into the sea—sails, rigging, and all. The midshipman warns the captain that unless the ship is cut free from the fallen mast, it will drag her over and cause her to sink. But there is a dilemma; for clinging to the wrecked mast is a sailor who was blown over with it. If the ship is cut free from the wrecked mast, the sailor will—with a practical certainty—drown. The captain makes his decision. He calls for axes and orders another sailor standing nearby to help him cut the ropes binding the ship to the fallen wreckage. This sailor happens to be the best friend of the fellow clinging to the wreckage; and as he lifts his axe and brings it down on the ropes, tears stream down his cheeks. The sailor with the axe knew what he was doing. He knew that he was performing an act that would help to cause his friend to drown. He could have disobeyed orders and refused to cut the ropes. But he did not refuse; he chose to obey. Did he therefore intend to kill his friend? To give an affirmative answer would, I think, imply that the sailor was intent upon killing his friend—that that is what he wanted. Such an implication would be entirely inappropriate. What he was intent upon, what he wanted, was to save the ship and its crew by the only means possible. The death of his friend was entirely unwanted, as his streaming tears bear witness. In this case, then, it seems appropriate to say that in choosing to cut the ropes, the sailor intended the effect of saving the ship, but that he only accepted with the deepest reluctance the effect of causing his friend to drown. His friend’s death was quite beside his intention—even though he chose it. What this analysis reveals is that intention is not just about deliberate choice, but also about desire; not just about willing, but also about wanting. An effect that I intend, therefore, is one that I both choose and want; and an effect that I accept is one that I choose but do not want.

If I understand him correctly, Tollefsen wants to say that the death of the sailor’s friend did not “enter into [the sailor’s] deliberations” as “any of the nested means,” and that it was neither chosen nor therefore intended. As he sees it “what one intends is the end one rationally desires, and everything one takes to be needful for the attainment of that end, and thus one sets oneself to achieve for its sake . . . however emotionally repugnant one finds it.” That is to say, whatever one chooses, one intends.

I disagree. I think that it is important to distinguish between intending and choosing, in order to be able to defend double-effect analysis from the charge that it excuses agents from responsibility for their acts. In saying that the sailor did not intend his friend’s death, I do not say that he was not responsible for it. He chose to perform an act that he knew, with a practical certainty, would result in his friend’s death. He did not do it in his sleep or while drugged or under immediate physical coercion; he did it consciously and deliberately. He was therefore responsible—but only in the sense that he was accountable, not in the sense that he was (necessarily) culpable. He was obliged to give a morally justifying account of his deliberate action, and if he could render one, then he would be responsible but not culpable. I believe that he could justify his responsible action partly in terms of his intention—namely, that, while choosing to perform an act that he knew would certainly cause his friend’s death, he did not want that death but only accepted it with proportionate reluctance.

Tollefsen’s comments force me to articulate my view in terms of means as well as ends, something that I am aware of having been shy of doing hitherto. Did the sailor choose his friend’s death as a means toward the end of saving the ship? I could say no, and argue that the means, strictly speaking, comprised the detaching of the wrecked mast from the ship, to which his friend’s death was accidental. I could add that the accidental nature of the death’s relationship to the chosen means is made clear when one considers that, had his friend (by some miracle) survived, the sailor would have regarded his ship-saving action as nonetheless successful. However, since the circumstances dictated that his friend’s death was causally bound up with the cutting loose of the mast, it was in fact a part of the means, albeit accidental, and insofar as the means were chosen, then so was the death. Unlike the detaching of the mast, however, the friend’s death was not intended.

What most worries Tollefsen—and quite rightly so—are the implications of my preferred analysis for the killing of the innocent, for it allows me to say that “the deliberate killing of the innocent is not wrong as such.” He is correct to think that by “deliberate” here I mean both “consciously chosen” and “consciously accepted” (but the former only in terms of the latter and not in terms of “consciously intended”).

This does allow me to justify the deliberate mercy-killing of a moribund, agonized, and desperate soldier. The agent chooses to perform an action that he knows will certainly cause the soldier’s death. He is responsible for that action and must give a justifying account of it. He can give such an account in terms of intention and proportion: he intends or wants only to relieve the soldier’s distress; in these circumstances the only means of pain-relief is in fact bound up with the soldier’s death, which the agent accepts with appropriate reluctance; and since the soldier has no realistic prospect of surviving, the choice to accept his death is proportionate.

It also allows me to justify Vénuste killing his brother Théoneste (with or without his consent), in order to prevent the Hutu Interahamwe from slaughtering the rest of their family. Vénuste chooses to perform an action that he knows will certainly kill his brother. He can justify his responsible action in terms of intention and proportion: he intends or wants only to save his family; in these circumstances the only means of saving them is in fact bound up with Théoneste’s death, which Vénuste accepts with the deepest reluctance; and since Théoneste has no prospect of surviving anyway, and since he is only one and his family are many, the choice to accept his death is proportionate.

In both cases, the lethal means are chosen, but the agent’s will disposes itself in a manner significantly different from each of the two effects realized by them: it positively desires or wants the soldier’s pain-relief and the family’s survival, while it only accepts with appropriately deep reluctance the soldier’s and Théoneste’s death.

It is notable that the two cases under consideration both belong to a state of nature—that is, of war or of civil war. What is permissible in such extreme circumstances need not be permissible in normal, civil ones. Thus, it is my view that, while “mercy killing,” which accepts but does not intend death, is permissible on the battlefield, there are good prudential reasons why it should not become part of legal medical practice in civil society. Where palliative care and sedation are widely available and can be made more readily so, and where large numbers of elderly people would be vulnerable to abuse, it would be disproportionate to take the risk of breaking the taboo against killing by granting physicians even the conditional liberty of taking the lives of their patients. (See my book Aiming to Kill: The Ethics of Suicide and Euthanasia.)

Although I am aware that my view diverges from some versions of just war and, more broadly, natural law reasoning, it is not true, as Tollefsen might be taken to imply, that it diverges from all of them. Francisco de Vitoria, for example, argues that “it is occasionally lawful to kill the innocent not by mistake, but with full knowledge of what one is doing, if this is an accidental effect . . . since it would otherwise be impossible to wage war against the guilty.” And, according to Hugo Grotius, “if one subject, tho’ altogether innocent, be demanded by the enemy to be put to death, he may, no doubt of it, be abandoned, and left to their discretion, if it is manifest, that the state is not able to stand the shock of that enemy.” What is more, if the subject will not voluntarily surrender, Grotius argues, his sovereign may force him to do so.

The permission to kill the innocent for the greater social good is undeniably a very dangerous one, especially when taken up by powerful modern states. For that reason, I support the granting to innocent citizens of positive legal rights against being killed, even on public authority. Such rights, of course, presuppose a civil social context.

I fear that what I have written here in response to Tollefsen’s criticisms will not allay his anxieties. I can only say that I take no pleasure in that and that I, too, am aware of the morally dangerous implications of my position. I am also open to being persuaded otherwise. Unfortunately, I am not, so far, persuaded.