Abstract

In this paper, I propose and defend a new criterion for appropriate choice under conditions of moral uncertainty. According to this new criterion, it is appropriate to follow some moral theory T under conditions of moral uncertainty iff T maximises expected truthlikeness. This approach to moral uncertainty has several important advantages: (1) it does not require any intertheoretic choiceworthiness comparisons; (2) it respects agent-centred prerogatives and supererogation; (3) it does not suffer from any ‘individuation problems’; and (4) it has a prima facie plausible motivating idea.