Changing preferences : A study in preferentialism

Abstract: This thesis deals with preferentialism, a family of theories claiming that moral considerationsshould be based entirely on preferences. The emphasis is on preference utilitarianism,according to which preferences determine the values of outcomes and these values determinewhat we should do. The main aim is to find a version of preference utilitarianism that (1) fitsclosely to our intuitions about cases in which preferences change, either across time or acrossavailable outcomes, and (2) can be embedded in a consequentialist framework.In Part One, I define several important versions of preferentialism. In particular, Idistinguish between the satisfaction version, according to which value is ultimately assignedto satisfactions and frustrations, and the object version, according to which value is ultimatelyassigned to the objects of preferences. I also distinguish between outcome-transcendent andoutcome-bound versions of preferentialism. The former theories but not the latter accept thatwhat we should do may depend on factsabout how well an outcome matches preferencesoutside that outcome.Part Two concerns outcome-bound versions of preferentialism. I focus on cases inwhich preferences change across time and cases in which preferences depend, for their veryexistence, on what the agent does. Theories that exclude preferences because they have acertain temporal location or because they are contingent on the agent's choice are considered.I argue that they lack normative appeal and that they cannot be incorporated into aconsequentialist framework.Part Three is devoted to outcome-transcendent theories. I show that these theories haveimplausible implications for cases in which preferences change, and that they cannot beconstructed as consequentialist theories. Since outcome-transcendent theories go hand in handwith the object version, this also shows that a preference utilitarian has good reasons toembrace the satisfaction version.