Aristotle on Virtue: A Question About Circularity

I have a question about the Nicomachean Ethics that has been driving me nuts. I'm sure there is an obvious answer, but I still seem to be missing it after my third read. How do we determine the good in action?

Here's the issue I am having: The end of action is set by wish (boulesis). The means are decided on through deliberation (prohairesis). And prudence, it seems, is the virtue of deliberating well. The virtuous agent wishes for what is noble (kalos) and correctly deliberates about the best actions to attain the noble. So we have an account of the end of the virtuous agent, the faculty by which he pursues that end (deliberation) and the virtue of that faculty (prudence). We also have the faculty by which he chooses/recognizes the end: wish. But is there a virtue of wishing? In other words, it does not help us to have an account of how one attains the end if we have no account of how the end is to be found. (Of course Aristotle could just say that the noble is the end that the virtuous person chooses and leave it at that. But he does leave it at that in the case of deliberation, which he fills out at length with an account of prudence. So if this were his strategy, some account would be needed for the difference between deliberation and wish.)

Aristotle does say that the end is grasped through virtue (1151a), so I suppose the doctrine of the mean is supposed to give us the end. That is: someone who has reached the mean will also correctly grasp the end. But how do we figure out what the mean is? Aristotle repeatedly tells us that we find the mean through correct reason, and correct reasoning is the virtue of prudence. But this raises a further problem. Since deliberation is about means rather than ends, we cannot deliberate at all without some end in view. And it seems that we cannot deliberate prudently without having grasped the noble. That makes it seem like no one can be prudent—and thus find the mean—without already possessing the correct wish. But if the correct wish is grasped through virtue, this is circular.

Now circularity is not a huge problem for Aristotle, I think. Or, rather, I think it is a problem, but he does not see it that way. For example, his account of praise and blame in Book III.5 uses a similarly circular account. The problem there is this: We shape our character through our actions, and so we become virtuous by choosing and performing virtuous actions. But we can only choose for the good, and our apparent good is dependent on our state (virtuous or not). So while our character is (partly) up to us, since we choose the actions that shape it, our ability to choose the right actions hangs on our character. Aristotle thinks this resolves the problem of blame, but of course Galen Strawson uses a variant of this very argument (though without reference to ends) for the impossibility of moral responsibility. The circularity implies that no one can be ultimately responsible for their actions, and thus that no one can deserve praise or blame except in a very tempered sense.

But if we accept Aristotle's answer to my earlier question as similarly circular, then I cannot see any reason to insist that virtues are a matter of quality rather than degree. That is: for Aristotle, one cannot be kind of prudent, or sort of brave; one either has the virtue or one doesn't. And it seems like it is in response to something like my concern that Aristotle, in VI.13, brings up the thesis of the unity of the virtues. He argues there not that virtues cannot be partial or admit of degrees, but rather than one cannot have one virtue without having the others. The point is roughly the same, however. If we take Aristotle's argument there to be convincing, however, it doesn't seem like he has proven that the virtues are unified. If anything, by analogy to Strawson's reversal of the III.5 argument, he has proven that virtue is impossible. Here is why:

Aristotle's tendency to circularity can be justified by his constant insistence that he is giving a general rather than a universal account, since ethics does not allow of the same level of precision as mathematics. If so, then it makes sense to allow that someone is blameworthy insofar as his control of his character depends on his character: while he cannot be sui generis (which Strawson argues is necessary for moral responsibility), he can be jointly responsible for the creation of his character. And as he adjusts his ends to his virtues (the products of his choices) and his choices to his ends through something like an internal process of reflective equilibrium, his level of responsibility (and his deservingness of praise and blame) increases. Well and good. But this sort of account cannot allow for absolute judgments. It makes no sense to speak of someone being absolutely blameworthy or not at all blameworthy when the condition for blame is itself a matter of degree, as it must be if the circularity account is to work. But then Aristotle's insistence that virtue is an all-or-nothing affair must be inconsistent with his entire project in the Ethics.

I'm sure there is a massive field of Aristotle scholarship on all these issues, and I'll be happy to get to it eventually. But I am wondering if others have thoughts or references they could share.