Abstract

An important but neglected difference between modern utilitarian and Kantian ethics on one side and virtues ethics on the other concerns the relation of good and evil. By taking virtues to be ethical primitives, standard versions of virtues ethics entail that some goods are logically evil‐dependent. That is, at least some central virtues cannot be characterised without reference to the possible existence of an evil, and cannot be exercised without the actual existence of that evil. Given this account on the logical relation of good and evil, can a virtues ethic be defended? In particular, does a virtues ethic sanction an inference from a virtue to the evil that is the occasion for its exercise? The paper examines some prima facie examples of such inferences and considers how they might be blocked from within a virtues ethic.