might account for this apparent difference:
Insofar as Kant expresses esteem for the continent person, he seems clearly to be assuming that the person is not responsible for the errant inclinations. One's inclinations act as external obstacles to duty just as much as enemy gunfire or rising floodwaters do. The continent person is heroic. According to Aristotle (NE III, 5), on the other hand, one is responsible, in the long run, for having errant passions, even though one with them cannot immediately be rid of them. The continent person is no more heroic than one who negligently allows the house to catch on fire and then scrambles through the flames to save the child. 6
Or it could be that when Kant imagines the person whose choice aligns with inclination he is imagining that the choice was determined by inclination. Since inclination is not a reliable guide to proper behavior, a character guided by inclination is dangerous: One might worry that inclinations are unsteady, whereas allegiance to duty is firm; or one might worry that inclination is not sufficiently responsive to the nuances of morally relevant features of a situation, whereas allegiance to duty is. But insofar as Aristotle prefers the virtuous person to the continent person, it is important to recall his distinction between natural virtue and virtue in the strict sense (NE VI, 13). The person of mere natural virtue has proper passions without practical wisdom, and the person's actions are guided by, or due to, passions. Such a person might well be unreliable, and to that extent the continent person might be preferable to the person of natural virtue. But the proper comparison is with a person of virtue in the strict sense. Such a person is guided by practical wisdom and not by passion, though passion concurs with, or does not dissent from, action.7
In the famous passage in Section I of the
Groundwork (G 398), Kant does seem to be distinguishing the person of moral worth (whom Aristotle would call continent) from the person of natural rather than strict virtue. Presumably it is counterfactually true of the person of strict virtue that if the person's passions did get out of line, reason (though no longer practical wisdom, since that requires proper passions) would still control the action. So the person would lapse into continence, rather than incontinence. Aristotle is willing to praise the person of continence, but only in relation to incontinence. He wishes to encourage the victory of reason over passion. Yet, he has no sense of praise in which the continent person is more deserving of it than the strictly virtuous person.It might be thought, however, that what Kant esteems is not the capacity to defeat passion, which is indeed shared by the continent and the strictly virtuous person, but the actual defeat of passion--which is true only of the continent person.8 But if esteem, for
Kant, is limited to what is purely under the control of the agent and influenced by no element of external luck, then he surely could not require the actual victory of reason over the passions, for whether the agent happens to have errant passions that need defeating might itself be a matter of luck (bad luck, according to Aristotle). Given Kant's pessimistic view of the human condition, such opportunities might be thought inevitably to present themselves. (Even Aristotle, with his optimistic view, would admit that opportunities inevitably present themselves--during the acquisition of virtue, though not during the exercise of virtue.) But how frequent and how challenging the opportunities are is still a matter of luck. Furthermore, the fact that actual victory, as opposed to the capacity for victory, over the passions is something external to the notion of the good will itself convinces me that it cannot be the exercise of the capacity that Kant esteems--it must be the capacity itself.
Perhaps, finally, Kant is more impressed by the continent person than by the virtuous person because he is bothered by the epistemological difficulty of distinguishing natural virtue from strict virtue. We can only be sure reason, and not inclination, rules in a person when inclination fails to control action--and this is only clear in the continent person.
Aristotle is not bothered by this sort of epistemological worry. One way he might distinguish natural from strict virtue is by the fact that the person of strict virtue will have all the virtues. This follows from his unity of the (possession of the) virtues thesis (NE VI, 13). There would be other signs, too, of the possession of practical wisdom, such as deliberation, and the ability to give a proper account of one's choices. Kant, on the other hand, has to acknowledge (indeed, he does this at great length (G 406-408)) that even in an apparent case of continence we cannot be certain that it was reason that won out over inclination. We cannot be sure that "some secret impulse of self-love, merely appearing as the idea of duty, was not the actual determining cause of the will" (G 407). So epistemological considerations cannot explain why Kant is impressed by the continent person.
In sum, the differences between Aristotle and Kant are not what they appeared to be:
They agree about the ordering of character-types according to what sort of character one would like to have. But they differ about what character-types humans can hope to have (an issue in philosophical psychology), and they differ about how to structure the concept of 'virtue' (an issue in the philosophy of language). And these differences conspire to explain their different labeling of character-types. They agree about which character-types deserve praise. But they differ over whether there is a specifically moral sort of praise, and they differ over whether, or the extent to which, people should be held responsible for their passions (both issues in moral psychology).9
These explain their different attitudes towards continence.