Abstract

According to Railton: x is good for me iff my Fully Informed Self (FIS) while contemplating my situation would want me to want x. I offer four interpretations of this view. The first three are inadequate. Their inadequacy rests on the following two facts: (a) my FIS cannot want me to want what would be irrational for me to want, (b) when contemplating what is rational for me to want we must specify a particular way in which I could rationally acquire the recommended desire. As a result, what my FIS could reasonably want me to want is limited by what information my FIS could reliably convey to me. And therefore what my FIS could reasonably want me to want cannot be grounded in changes in desires that my FIS cannot publicly justify. The fourth interpretation limits the scope of what my FIS could want me to want in a way that avoids these problems, but conflicts with standard intuitions about what is a non-moral good