In his classic paper “Moral luck,” Thomas Nagel claims that Kant denied the relevance of moral luck (i.e., Kant denied that any factor outside an agent’s control should determine how we morally appraise an agent or her actions), and that the explanation of Kant’s denial is that Kant sought to make virtue possible for everyone.
In other words, if we admit the relevance of moral luck, then perhaps some people, because of bad moral luck, could not fulfill their moral obligations or could not live generally upstanding moral lives. So Nagel in effect thinks that Kant arrived at an ethics of maxims at least in part because such an ethics, by rejecting the significance of moral luck, best enables virtue to be put within the reach of every human being. What I find fascinating about Nagel’s suggestion is that it implies that the question of how demanding morality ought to be, or how much a moral theory ought to demand of us, is itself a moral question. That is, on this interpretation, Kant denies moral luck because to acknowledge it would be unfair. Denying moral luck turns out to be an egalitarian, anti-elitist move.
In contrast, the Greeks (especially Plato, and to a lesser extent, Aristotle) clearly thought moral knowledge, and hence moral virtue proper, were not easily accessible, and were in fact only acquired after very careful and extensive conditioning or education. And as many have observed (especially in connection with Aristotle), this make virtue very much a matter of moral luck.
So I’m interested to know what people think about how we would decide among competing moral theories with respect to the degree to which moral knowledge, virtue, acting rightly, etc., are matters of moral luck according to these theories, and in particular, whether this decision ought itself be guided by theory-specific moral considerations (as Nagel claims it was for Kant), or whether there might be some theoretically-neutral way of proceeding here.
Michael,
Fascinating post, really good question. My original thought was that we should answer this question by appeal to what we want a moral theory / scheme of moral evaluation for. For example, the Greeks wanted to know who you would want to have as a fellow citizen of the polis, and it’s just a fact that who makes a good citizen is to some extent a matter of luck.
But then I thought that this teleological approach (“what do you want the theory for?”) is itself a very Greek-ethical way of proceeding. Maybe Kant would have rejected the question, and said that a moral theory is about displaying the dignity of individual human beings. And that cannot be a matter of luck (given the appropriate concept of human dignity).
So I’m thinking there is no theory-neutral way of going about this. But great, thought-provoking post.
Well, being a sort of contractualist myself I always thought that this theory can cope well with this issue. Of course, I don’t deny that other accounts may do so as well. I admit that the my of dealing with the problem is ‘theory-ladden’, but I am happy with that in this case.
Imagine that there is an act A the doing of which saves someone from having to bear a burden B. Now, very roughly put, contractualism would tell us that it would be my duty to do this act A if I cannot reasonably reject the principle, which requires me to do A. I could reasonably reject the principle, if complying with that principle would in general produce a bigger burden than B1 for me (or for someone else for that matter). What kind of a burden doing A will cause for me of course depends on the the nature of the act and the general circumstances.
It may be that in no circumstances at all being required to do A will cause me a burden bigger than B, and hence my moral luck can have no effect on whether I ought to do A. On the other hand, it may that the required act A is of the kind that the requirement for it may in some general circumstances produce a bigger burden for me than B and in other general circumstances only a smaller burden than B. If this was the case, then the fact in which general circumstances I happened to live would have an effect on whether I am required to A or not. Thus, my moral luck of being in certain circumstances would have an effect on what happened to be my duty.
If that is along the right lines, then it appears to be the case that contractualism provides a neat way to investigate the effects of moral luck for our duties. Anyway, this is just a suggestion (which hopefully makes at least some sense), and for those who are not that keen on contractualism it doesn’t of course have much appeal.
Just to see if I understand Nagel’s suggestion: Is he saying (or is he interpreting Kant as saying) that a theory should say P if it would be better that P?
Ben,
I think that’s basically right: that Kant’s denial of the significance of moral luck is rooted in a moral judgment, and in that broad sense, Kant says ‘moral luck is insignificant’ because it would be better (fairer, etc.) that moral luck be insignificant. I actually think things are little more complicated in the case of Kant than Nagel acknowledges, because one suspects that Kant, because of his metaphysical and epistemological commitments (the phenomenal-noumenal distinction, etc.) might have thought that there is no knowable fact of the matter regarding the influence of moral luck on people’s actions, character, etc., and so to adopt the attitude of denying the significance of moral luck is not so much to affirm some proposition about moral luck but to adopt a practical stance toward rational agents. (This of course doesn’t seem to address the other kinds of moral luck Nagel identifies, such as the luck associated with the consequences of our actions and the circumstances that provide the contexts of action in the first place.)
But perhaps that suggests that the real issue here is whether moral luck is exclusively a normative question. That is, the position Nagel attributes to Kant is that whether moral luck is significant depends solely on moral considerations. In contrast, one could hold that whether moral luck is significant is at least in part a question of whether the various sorts of moral luck in fact constrain how we behave, the results of that behavior, etc.. There might then be room for normative discussion of the significance of moral luck, but only within the confines of the aforementioned facts. So I guess the issue is not whether the significance of moral luck is at all a normative question, but the extent to which it is a normative question. (Again, apparently for Kant, it was an entirely normative question.)
I think that we are wise to tolerate a certain kind of resultant moral luck that arises when there are circumstances in which the act someone should perform is not one the agent could reasonably take to be the one that she ought to perform.
It seems that on nearly all moral theories, one could reasonably but mistakenly judge that some option O1 has the general characteristic in virtue of which right acts count as right when in fact it is O2 that has it. Under these conditions, O1 is a decoy duty and O2 an non-culpably unknown (perhaps unknowable) obligation. There is a sense in which only a certain kind of luck would lead the agent to perform the action she ought to (she could only manage to avoid failing to fulfill her obligations by accident or because as fortune would have it, the agent has the right combination of shabby motives and false beliefs).
If a theory were to deny that such cases were possible, it seems that it would have to significantly limit the sorts of conditions that could figure in our general statement of what makes right acts right. And that’s bad for a host of reasons I don’t want to get into. How’s that for a general argument that our moral theories should tolerate a kind of luck?