Authority-Based Moral Responsibility

So far, I have argued that on two influential accounts of shared action the participants involved in shared action will either necessarily or typically take themselves to have authority over one another, where this, in turn, yields what the participants take to be protected reasons in favor of compliance.

30.3.1 Establishing an Agential Division of Labor

Suppose Person A asks Person B to assault a particular innocent. In response, Person B promises to do so. In this case, A and B have established an agential division of labor, in that Person A counts as the ‘decider’ and Person B counts as the ‘doer.’ Person A counts as the decider in the sense that she has the function of normatively determining whether Person B is to commit the assault, by either holding him to his promise or releasing him from it. And to say that Person B counts as the doer is to say that she has the function of acting accordingly. But what does it mean for someone to have a ‘function’? And how does someone attain a function?

All functions have an end — an aim at which the function is normatively directed. We can characterize a function in terms of a protected reason to act in accordance with the function’s end.

Recall, though, that protected reasons also characterize authority. Person A and Person B believe that A has authority over B that B do φ if they both believe that A’s claim that B do ô provides a protected reason for B to do φ. So, one way to characterize the functional rela­tionship between Person A and Person B is in terms of the authority Person A has over Person B. This is because both authority and functional roles are cached out in terms of protected reasons. The upshot, then, is this: the protected reasons resulting from the promise which Person B makes and Person A accepts grounds both the putative authority Person A has over Person B, and the function each of them has with respect to each other.

The idea that Person A’s claim against Person B settles the matter for B — in that B takes himself to have limited deliberative say in the matter — might seem to overstate the role that the promise plays in B’s motivational economy. Suppose that if fulfilling the promise were not beneficial to B, he would renege on it. Or suppose B decides to go through with the promise he made only after deliberating about the merits of committing the assault, in that he re-considers the advantages of doing so, of getting caught, whether he feels like doing it, and so on. In these cases, Person B seems to be violating the deliberative injunctions imposed by his protected reasons, even if he ultimately decides to go through with the assault.

Nonetheless, so long as B believes that the promise provides a protected reason to commit the assault, then by B’s own lights, Person A has the authority to settle the matter for him, and to exclude further deliberation about the matter. If he denies that, then he has not made a sin­cere promise. So long as they both believe that A has a claim against B where that claim yields a protected reason for B to commit the assault, then the functional relationships characterizing the division of agential labor between them remain in place, even if B’s commitment to the protected status of that reason is less than perfect.

In summary: Person A and Person B establish an agential division of labor in which A functions as the decider and Person B functions as the doer. Each has this function in virtue of their belief that A’s claim against B yields protected reasons for B to act accordingly. That is, each has the function that they do in virtue of the putative authority that A has over B. And that authority is grounded in the promise that B makes and A accepts.

Clearly, both A and B are severally and fully morally responsible for the ensuing assault. Person A is responsible partly in virtue of the causal role she plays in convincing or motivating Person B to commit the assault. But in what follows, I will argue that A is also morally respon­sible on different grounds: in virtue of the role she plays as the ‘decider’ and the role Person B plays as the ‘doer.’ This will have consequences for responsibility in the context of shared action in general, and in war specifically.

30.3.2 Responsibility in an Agential Division of Labor

When Person B commits the assault, his victim is entitled to an explanation of what happened. She has an interest in ascertaining the wrongdoer’s practical reasons for acting in the way that he did. Whether and why the wrongdoer believed the harm to be warranted is morally relevant to assessing the wrongful conduct.

But what type of practical reason is morally relevant to assessing the conduct? A fact-relative practical reason is a practical reason that there actually is — a consideration that counts in favor of acting in a particular way.

The relevant belief-relative practical reasons are those that the wrongful conduct in question had the function of enacting. The purpose of practical reasoning, is, after all, to normatively guide conduct by determining what ends to pursue and how to pursue them. Concomitantly, the purpose of conduct (or at least intentional conduct) is to enact the practical reasons the actor takes there to be.

Clearly, B’s conduct had the function of enacting some of his own practical reasons of which there might be a variety. Person B might have committed the assault partly because he promised to do so, and partly because he enjoyed it, and partly because he was paid to do so, and partly because he wanted to earn Person A’s respect, and so on. Suppose, though, that like most people Person B thinks that when we make a promise, we make a commitment to the person to whom we make the promise, and that this gives us a protected reason to fulfill the promise irrespective of whether we benefit from doing so. This isn’t to say that B is motivated solely or even mostly by that sense of commitment. Neither does it suggest that B is so committed that he would follow through with the promise even if he thought doing so wouldn’t benefit him. Rather, Person B registers the commitment he made as a reason — albeit a defeasible one — to fulfill the promise.

But if the relevant belief-relative practical reasons are those that the wrongful conduct in question had the function of enacting, then they will include not only B’s reasons, but A’s as well, since A has the function of specifying whether B is to commit the assault, and B has the function of acting accordingly, given the division of agential labor the two of them established.

This means that, if we evaluate conduct by addressing the belief-relative practical reasons that the conduct has the function of enacting, then we ought to evaluate B’s conduct by adverting to A’spractical reasons. So, if the victim wants to know all the belief-relative practical reasons behind the decision to commit the assault, she needs to address not just Person B, but Person A as well. In this way, the practical reasons A takes there to be serve as a basis by which we evaluate what B does. To the extent that the practical reasons A takes there to be are morally problematic, Person A is responsible for a wrong-making feature of Person B’s conduct. After all, if we evaluate conduct by addressing the belief-relative practical reasons that the conduct has the function of enacting, and if Person A’s belief-relative practical reasons are problematic, then Person A has made it so that Person B’s conduct is morally problematic as well.

To be clear, the claim here is not that Person A is morally responsible in virtue of having influenced Person B’s motivations (though that too might be a basis for moral responsibility). Rather, the basis of Person A’s responsibility for Person B’s conduct is constitutive rather than causal. Person B’s conduct has the function of enacting the practical reasons Person A takes there to be; so, by adopting morally problematic practical reasons, Person A, from afar, constitutively determines the purpose of Person B’s conduct.

IfA is responsible for a wrong-making property of the conduct in this way, then A (in add­ition to B) is a proper object of the victim’s reactive attitudes in response to the wrong she has suffered. The result is that when B fulfills a claim that they both believe A has against B, Person A can be morally responsible for the harm Person B inflicts quite apart from the causal role A plays in that harm. I will call this the ‘authority-based argument for moral responsibility.’

Of course, in the example under consideration, A is also a cause of B’s conduct; so it goes without saying that B’s victim can blame not just B but A for the harm she has suffered. In this case, the authority-based grounds for moral responsibility is largely otiose. But in cases of shared action where each participant’s causal contribution is morally insignificant, the authority-based grounds for moral responsibility will play an important role in inculpating the participants, especially in war. I turn to this next.

30.4