A Possibly Final Post on Mercy

In previous posts (here, here and here) I've been arguing that the concept of mercy should take into account agent beliefs and attitudes. In other words, somewhere in the analysis of the concept, we should be talking about what the agent believes (and what his attitudes are) concerning the permissibility of his action and the nature of his situation. As I argued, if our analysis doesn't include some such element, certain merciful actions will not be called merciful. For example, a judge who, mercifully, gives a prisoner a sentence smaller than the maximum, when in actual fact it turns out the judge was required to give the maximum sentence, will not be said to have engaged in a merciful action if our account of mercy only takes into account what is actually permitted instead of what the judge believes is permitted.

Yet, as noted in those prior posts, every account of mercy that can be found in the relatively contemporary literature fails to take such considerations into account. All of them turn on questions of what is actually permitted by some standard or other, and none of them turn at all on what the agent believes about his situation or the permissions that apply. 

Is it simple to correct this omission? Should we just take whatever the best account of mercy is that leaves out agent beliefs, and simply add an "agent belief" clause? For example, suppose the (otherwise) best account of mercy says an act is merciful when both it and a more severe option were permitted by some standard or other, and the agent chose it out of compassion. In such a case, can it be as simple as amending this to say an act is merciful when the agent believed a more severe option was available and also believed both the more severe option and less severe option were permitted?

This may seem inevitably to be the kind of thing I must say if I'm going to insist on taking agent beliefs into account. But I don't quite like it. It seems to me that agents can be merciful without believing themselves to be following any single standard. They may, in fact, believe themselves to be breaking with all applicable standards. (It may be they're necessarily wrong to think this, but right now, recall, we're talking about what the agent believes). It doesn't seem to me that mercy rests, conceptually and necessarily, on standards in this way. Permissibility standards are often an important part of the discussion about whether an act was merciful or not. But the agent need not have any particular beliefs about standards in order for an act to be merciful.
And so in my latest post I introduced the Internal Principles Account of Mercy:

IPAM: x is merciful to y at t =df x believes there are two alternative acts, A and B, available to x at t such that (1) x performs A, (ii) x believes x's performing B would have resulted in more harm to y than x's performing A, (iii) x is motivated by compassion for y to choose A over B, (iv) two unselfish principles p1 and p2 move x, such that p1 moves x to choose A and p2 moves x to choose B. (p1 and p2 may be identical, as some principles allow for more than one option in a given situation.)

(Less precisely, an act is merciful when the agent undertaking the act is genuinely psychologically moved by some principle to do something relatively severe, and is also genuinely psychologically moved by some principle (possibly the same principle, but not necessarily or even usually) to do something relatively gentle, and out of compassion, chooses the more gentle act.)

One strength of Markosian's contextualist style account of mercy is that it offers a nice explanation as to why intuitions about mercy often differ. Take the case of the Nazi Doctor, for example:

There is a Nazi doctor whose job is to torture innocent children. But there is something about one of the children that strikes a chord in the doctor. As a result of the compassion that he feels for this one child, the doctor chooses not to torture her.

This is a case where intuitions vary. Many consider this a clear case of mercy, while others consider it just as clearly not to be a case of mercy. Markosian's explanation for the difference in cases like these (though he doesn't specifically explain the difference in this particular case) is that an action's mercifulness will vary according to the context of utterance of the judgment that the act is merciful. Different contexts will pick out different standards of permissibility. So for example, perhaps if a moral saint is talking to several other moral saints, then the standard of permissibility that will apply is moral permissibility—and the doctor's action isn't merciful because torturing the child wasn't permissible in the first place. But if, instead, the context of utterance is a Nazi philosopher talking to other Nazi philosophers, all of whom believe torturing innocent children is often justifiable but that it is okay to make the occasional exception, then their judgment that the act was merciful will be true, because the relevant standard makes both options permissible.

So then, those of us who are not Nazis but who do find this to be a merciful act may be thinking (on Markosian's account) that the relevant permissibility standard is something like "permissibility according to Nazi laws." Well, actually, I'm not too sure any of us who find the act plausibly merciful really do have any single such standard in mind. To my mind this is a weakness in Markosian's attempt to explain differing intuitions—it's not very plausible that there is any single permissibility standard anyone ever has in mind that makes both torturing innocent children and not torturing innocent children permissible. (Not that no such standard could be constructed artificially, rather I'm just making the psychological claim that none of us thinking about the doctor's actions have any such standard in mind.) But anyway, if there are such standards that we do have in mind, this would explain the differing intuitions. Different people having different beliefs about which permissibility standard is picked out by their context of utterance.

But my account, in addition to giving the right result for a certain difficult case where Markosian's fails (LINK), also seems to give a plausible explanation for differing intuitions about mercy. For people have different ideas about what principles other people are, or can be, genuinely psychologically moved by. In the Nazi doctor story, his compassion on the single child seems inexplicable. Why this child and no others? For this reason, we're invited, even if only subconsciously, to imagine some motivation for this compassion. And if IPAM is the correct account of mercy, and when thinking about the Nazi doctor we fail to think of a plausible way to explain how one person could both be genuinely moved to torture innocent children and genuinely moved by some other principle to spare this particular child, and moreover to choose the latter course of action out of compassion, we could very well be tempted (based on IPAM) to reject the categorization of this act as merciful. In other words, if we can't make sense of the doctor's action as an instance of the scheme described in IPAM, then of course we will not judge the doctor to have been merciful. And it's plausible to think someone might sensibly fail to see how to make the doctor's actions make sense in this way. For, really, how could a plausible human being be so motivated? If he's genuinely moved to torture small children, doesn't this practically preclude any genuine psychological motivation, associated with compassion, to spare this particular child?

Meanwhile, if one can come up with such an imagined set of motivations, it surely becomes much more plausible to think of the doctor as merciful. It's a little hard for me to come up with such a thing, since I myself find the doctor's action to be unmerciful simply because it's inexplicable (just as described in the previous paragraph) but I can make a go of it. Perhaps the doctor is genuinely moved by a belief that the torture will lead to important physiological discoveries, something he cares about inherently. And perhaps he's also genuinely moved by the kind of generalized protective feelings we all have for children. And though in most cases he has managed to quash that protectiveness in the name of physiological discovery, in this particular case something about the child reminds him of his protective instincts. Perhaps she subconsciously reminds him of a child he once knew. And so he arranges to have her sent away from the camp somewhere safe. If I think of the doctor as someone experiencing this as a difficult decision, one his mind quivers back and forth over, causing him real emotional tension, because he really is genuinely moved by two opposing principles and it is ultimately his feelings of compassion—usually dormant but briefly awakened in this particular case—which move him toward the less severe act, then I do begin to think of the doctor as merciful. Granted I had to add some details. But the details are there, if anything, simply to make the example explicable as the actions of a human being I can understand. (It may be there were actual Nazi doctor cases in Germany before and during world war two. I don't claim it to be impossible—I'm simply expressing how inexplicable such cases seem to me. But once I supply certain details, and begin making the doctor's actions "explicable" as the product of principles genuinely psychologically moving the doctor in incompatible directions, then his actions start to feel more like merciful actions to me, even as I continue to see him as repugnantly evil.)

So in any case, whether this is convincing or not, I think I have now completed my blog exploration of this topic. I now have a paper to write, to present a new account of mercy and explain how it has all the benefits of Markosian's powerful account, none of the downsides, and accounts for more than what he accounts for.