Abstract

In this paper, I start from the observation that there are obvious instances of instrumental irrationality, i.e. cases where subjects act knowingly against their strongest preferences. This observation raises an important question: Which facts determine the ‘strength’ of preferences? I consider a standard answer to this question – ‘revealed preference theory’– which turns out to be unsatisfactory. Then I turn to a more promising alternative: the ‘higher order theory’ of preference strength. But this proposal also faces a major problem, the ‘problem of authority’. I argue that this difficulty can be overcome, but only if we are prepared to make some significant normative assumptions – assumptions which imply, among other things, that instrumentalism – the view that instrumental rationality is all there is to rationality – cannot be maintained