Background

In international law, the legal permission to target combatants derives from their status rather than their conduct. That is, what makes an individual a legal target is not whether she poses a threat, but whether she qualifies as a combatant (though a civilian directly participating in hostilities thereby qualifies as a combatant).

On the orthodox view, the morality of war parallels international law with respect to the laws pertaining to the legal permission to target combatants. The most influential defender of the orthodox view is Michael Walzer. He argues that all active-duty combatants — even those whose actual contribution to the war is small or nonexistent — are morally permissible targets since they are all ‘currently engaged in the business of war’ (Walzer, 2000, p. 43). By having made themselves into combatants they have made themselves dangerous.

The orthodox view, though, has been heavily criticised, most prolificacy by Jeff McMahan (McMahan, 2009). He argues that whether you pose a threat is neither a necessary nor a suf­ficient condition for losing your right not to be killed. A combatant does not lose her right to be killed when she poses a morally justified threat, no more than the victim of an attempted mugging loses her right not to be attacked when she engages in necessary and proportionate self-defence. Hence, combatants who are fighting in accordance with the rules of war and in furtherance of morally just aims have done nothing to lose their right against being killed. Put differently, such combatants are not morally liable to be killed.

On this revisionist view, combatants who are acting in furtherance of a war’s unjust aims have indeed forfeited their right against being subjected to necessary and proportionate defensive violence.

So, on McMahan’s revisionist view, only combatants responsible for contributing to an unjust cause (or against a just cause pursed by unjust means) are morally liable to be targeted. It is what you do — not your status — that determines your liability in war. If this is correct, the orthodox view is mistaken in its claim that all combatants are morally entitled to kill enemy combatants.

The revisionist view faces a challenge, though. On the revisionist view, a combatant is mor­ally liable to be killed only if she contributes to an unjust cause (or to a just cause unjustly). Moreover, that contribution must be substantial — so substantial that she loses the right against being subjected to necessary and proportionate lethal violence. However, a significant portion of combatants, such as those who occupy subsidiary support roles, contribute very little to the war being fought. Indeed, most combatants never see combat. If, on the one hand, these combatants are not morally liable to be killed, then how do we distinguish the liable combatants from the non-liable ones? Waging a morally just war would seem to be pragmatically impossible.

Anticipating this problem, McMahan argues that the threshold for liability is quite low; as a result, virtually all combatants acting in furtherance of an unjust aim will likely be morally liable to be killed (McMahan, 2009).¹ Other responses are possible, though.

Recently, a different strategy for defending the revisionist view has emerged. Some have argued that most combatants — including those whose contributions to an unjust aim are insig­nificant — are typically morally liable to be attacked because they are engaged in shared action. For example, Adil Ahmad Haque argues that each participant is morally responsible not just for her own contribution to that shared action, but for the shared action itself. He says, “[i]n gen­eral, persons are morally liable to defensive killing only if they pose unjust threats directly, jointly with others, or indirectly through others they control” (my emphasis) (Haque, 2017, p. 57). So when combatants cooperate in furtherance of a mission, each is typically responsible for what they together do, since “members of a military unit performing coordinated roles in a combat operation are paradigmatic joint perpetrators” (Haque, 2017, p. 59). Haque elaborates further:

several people jointly kill another by together performing a joint action that causes the other’s death, without subsequent intervening agency by others and, typically, without which the victim would not have died. [...] For example, if A restrainsV while B stabs V to death, each performing coordinated roles in a common plan, then A and B together kill V Importantly, joint actions are the actions of each participant, not solely the actions of the last participant to perform her role or only the actions of those participants without whose participation the plan would not have succeeded.

(Haque, 2017, p. 65)

This view has an intuitive appeal; indeed, Haque isn’t the only one to have proffered it.²Yet we need to provide grounds for this view. We cannot just stipulate that when individuals cooperate in furtherance of a shared goal, each cooperator can end up responsible for more than what she

causes. To treat this as a bedrock claim is problematic since the intuition seems to conflict with the equally plausible pronouncement that each of us is morally responsible only for what is within her own causal reach. What we need is an account explaining how and why cooperation dramatically expands the scope of responsibility.³ It’s my aim here to provide such an account and to apply it to war.

My argument, in brief, is this. We share an action in the relevant sense when we coordinate and combine our actions in furtherance of doing something together, as when we walk together or paint a house together (to take canonical examples from Gilbert [1990] and Bratman [1992]). In typical cases, when you and I are engaged in shared action, I will take myself to have authority over you that you do your part (whatever that might be) and you will likely take yourself to have authority over me that I do my part (whatever that might be), where in both cases this authority is cached out in terms of protected reasons. A protected reason to do ô is comprised of a first- order reason to do ô and a second-order reason to exclude certain competing considerations from deliberation pertaining to φ. (See Raz, 1977; Raz, 1990, pp. 35—84)). This authority you and I take ourselves to have over the other is grounded in the agreement we made to partici­pate in the shared action. In arguing for this view, I will focus in section 30.2 on the two most influential accounts of shared action: Margaret Gilbert’s and Michael Bratman’s.

Why is this so? Normally, my intentional conduct has the function of enacting the prac­tical reasons I take there to be; concomitantly, the practical reasons I take there to be have the function of normatively guiding my conduct. However, it is possible for me to attribute to you the role of enacting those practical reasons; should you accept that role, I thereby change the object at which my practical reasoning is directed. Its object is no longer my conduct, but rather yours. That is to say, the practical reasons I take there to be have the function of guiding your conduct, and your conduct has the function of enacting the practical reasons I take there to be. In this way, we establish a division of agential labor, in that I count as the ‘decider,’ and you count as the ‘doer.’

We assess conduct partly by assessing the practical reasons for which it was done. So, to morally assess your conduct, we need to repair to the practical reasons I take there to be insofar as I count as the decider in the division of agential labor we’ve established. So, if my practical reasons are morally problematic, then I am responsible for a wrong-making feature of your con­duct. The result, then, is that I am at least partly morally responsible for what you do (though my responsibility in no way diminishes yours).

To be clear, the claim here is not that I am morally responsible for what you do in virtue of having influenced your motivations (though that too might be a basis for moral responsibility). Rather, the basis of my responsibility for your conduct is constitutive rather than causal. Your conduct has the function of enacting the practical reasons I take there to be; so, by adopting morally prob­lematic practical reasons, I constitutively determine, from afar, the purpose of your conduct. Insofar as that affects the moral assessment of your conduct, I am on the hook for that difference I make. I call this ‘authority-based moral responsibility.’

The upshot is this. The decider is morally responsible for what the doer does qua doer. Such cases might be better described as proxy agency or extended agency rather than shared agency, insofar as the doer acts not to achieve some mutually desired or shared end but rather out of a perceived obligation to the decider. But the basis for inculpation in these cases applies also to shared action as well. Any given participant in typical instances of shared action has authority over the other participants that they do their part, in that she has a claim against them yielding a protected reason for them to do their part. The practical reasons she takes there to be has the function of normatively guiding their conduct in that regard, and their conduct has the function of enacting the practical reasons she takes there to be. So, given the authority-based argument for responsibility, she is morally responsible for what they do should they conform to those reasons. Because this is true of every participant, the result is that each participant is morally responsible for what they together do — irrespective of the degree to which they causally con­tribute to what they together do. Finally, in section 30.4, I argue that the account provides a basis for thinking that an individual soldier in a war is responsible for more than what she causes or the difference she makes.

30.2