Abstract

According to Utilitarian Generalization an act is right or wrong depending on what would happen if everyone were to do acts of that kind. One chief difficulty in applying UG is to determine which acts share the same relevant properties and are therefore acts of the same kind. In focusing on this problem I first examine the criteria of relevance proposed by Jonathan Harrison and by David Lyons. I show that each of their proposals is inadequate because each allows us to designate as relevant some properties we noncontroversially take to be irrelevant. I next propose a criterion which not only allows us to designate as irrelevant every property that we noncontroversially take to be irrelevant, but which also expresses the underlying causal and generalization features of UG. Since the acid test of any such proposal will be its response to the highly controversial issue of whether the behavior of others is relevant for applications of UG, I turn to this important issue.