Expanding the bounds of rationality: three optimistic pieces

This thesis is composed of three independent chapters on the themes of rationality and rational choice. In chapter 1, I consider the money pump, which is the most prominent argument against the rationality of nontransitive preferences. I argue that the money pump fails to undermine such preferences: instead of condemning an agent's preferences in response to the money pump, we ought to condemn her way of deciding. More specifically, I argue that this should be our default response to the money pump and that a particularly plausible attempt to undermine this default response fails. In chapter 2, I examine resolute choice. An agent chooses in a resolute manner if she makes a plan when first faced with a sequence of decisions and then resolutely sticks to it. It is commonly accepted that resoluteness is a useful characteristic for an agent to possess but also commonly accepted that rationality does not require that agents have the capacity to be resolute. I will dispute this second assumption. I will argue that, given that resoluteness enables agents to avoid undesirable manipulation, we have a prima facie reason to think that rationality requires resoluteness. This argument is well-known but not well-liked. Accordingly, I spend a substantial amount of time defending the strength of this basic argument. In chapter 3, I consider the issue of rational choice given normative uncertainty: that is, I consider how an agent ought to decide if she is uncertain about the normative truths about rationality (rather than uncertain about the empirical truths about the world). In response to this issue, I argue in support of a number of sufficient criteria for choice under such circumstances and show that these criteria provide guidance in a range of interesting cases of normative uncertainty.