t seems rational to regret not having produced a lesser good than one actually did produce. If you were to take the most pleasant holiday you could, it seems rational to regret not having gone to an almost equally pleasant spot, even if one should also be pleased at having taken the best holiday. But this is puzzling. If it is only rational to produce the best situation, and irrational to do less than what is best, then to regret not having done less than what was best is to regret not having done what it would have been irrational to do. And surely it is not rational to regret not having done what is irrational.

The puzzle is solved of course if the forgone lesser good was of a different kind than the greater good produced. Then, as Michael Stocker argues (Hurka, pp. 562-3), there is a way in which the lesser good was better than the greater; it was better of its kind. Hence, one has produced the best situation of one kind, but not of another. And so it is rational to regret not having produced the best situation possible in terms of the kind of forgone lesser good. For instance, it is rational to regret the forgone less pleasant holiday, if the pleasure of that holiday is a good of a different kind from the greater pleasure of the kind in the holiday taken. But if, on the other hand, the forgone lesser good was of the same kind as the greater, then there is no way that the lesser good is better than the greater produced. Then there is nothing lacking in the greater good that is present in the lesser good, and so nothing to rationally regret. It follows that if regret is to be rational, there must be a plurality of kinds of goods. The view that there is only one kind of good, monism, is incompatible with the rationality of regret for a forgone lesser good.

Stocker's argument seems to me to be a good one. But Hurka thinks it is not. Although Hurka is a moderate pluralist, he thinks that a monist can hold that it is rational to love, and so regret, in an appropriate proportion, a forgone lesser good of the same kind as the greater good produced. Why is it important for Hurka to show this? For one thing, the above argument, if sound, threatens to make even moderate pluralism incompatible with rational regret. For if it is rational to regret not listening to music when one wrote a paper instead only if the good of music is distinct from the good of knowledge, surely it is also possible to regret not having written one paper rather than another even if one wrote the best paper possible. But then the good of the paper written must be a distinct good from the good of the paper not written. The rationality of regret thus threatens to pulverize the good, producing ever more fine-grained distinctions in kinds.

Hurka's objection to Stocker is that he wrongly assumes that "all cases of choice between instances of the same good are inclusion cases" (p. 563), that is, cases in which the greater good produced must be regarded as "including" the forgone lesser good. If all such cases are inclusion cases, then the monist must say that it is irrational to regret having produced the lesser good, since it is already "included" in the greater good and nothing that the greater good lacks is present in the lesser good. But, Hurka argues, all such choices are not inclusion cases. For the greater good produced will lack something present in the lesser good of the same kind if the lesser good is a "distinct instance" of that good, namely, it will lack that instance. For example, suppose you produce 10 units of pleasure for B only by failing to produce 5 units of pleasure for B. Then it is irrational to regret having failed to produce the 5 units of pleasure for B since you did produce them. But suppose instead you are failing to produce those 5 units of pleasure for someone else, A. Then you did not produce the forgone 5 units of pleasure in A in producing the 10 units for B. Hence, what one did produce lacks something forgone in what one did not.

Hurka is right that there is something lacking. But it is not clear how the monist can say that it is rational to regret not having produced this something. It all depends on what rationality requires, and Hurka does not give an account of this. Stocker's supposition, I take it, is that it is rational only to produce the best situation possible, and for a hedonistic monist the best situation possible is only the most pleasure possible. As a consequence, the forgone pleasure for A contains nothing rationally relevant that is not present in the pleasure produced for B. So how is the fact that it is A's pleasure that is forgone rationally relevant? Perhaps A deserves pleasure just as much as B. But then we have two kinds of good, pleasure and justice. Or perhaps A is a friend and B a stranger. Again, we then have two kinds of good, pleasure and friendship.

Hurka must think that it is simply less of a particular instance of the same kind of good that is forgone when B's ten units are produced and A's 5 units are not. Hence, one has produced the best situation possible of one instance, but not of another, of the same kind of good, even though one has produced the best situation possible--the best situation tout court. And so he thinks that there is a way in which the lesser good was indeed better than the greater good produce; it was better "of its instance", though not "of its kind". But this way in which the lesser forgone good is better than the greater good produced--its being better for A--cannot be rationally relevant for a monist, so far as I can see. For it is rational only to produce the best situation tout court. I think, therefore, that Hurka owes us an account of what it is about the distinctness of A's pleasure that can make it rationally relevant if it is not that producing A's pleasure would produce a distinct kind of good.