Shelly Kagan

Many proposed moral principles are such that it would be difficult or impossible to always correctly identify which act is required by that principle in a given situation. To deal with this problem, theorists typically offer various methods of determining what to do in the face of epistemic limitations, and we are then told that the right thing to do – given these limitations – is to perform the act identified by the given method. But since the method and the underlying principle can diverge, it would seem that in such cases we are being given contradictory advice: some particular act will be both right and not right. Various attempts to resolve this apparent paradox are surveyed, but none are completely satisfactory.

In Rethinking the Good, Larry Temkin argues that the common belief in the transitivity of better than is incompatible with various other value judgments to which many of us are deeply committed; accordingly, we should take seriously the possibility that the better than relation is not, in fact, a transitive one. However, although Temkin is right, I think, about the mutual incompatibility of the beliefs in question, for the most part his examples don’t leave me inclined to deny transitivity. Nonetheless, there is one example, involving infinity, that does seem to me particularly troubling.