Causal Decision Theory for Unratifiable Actions
-----
Recently, some philosophers have argued that causal decision theory (CDT) gives incorrect advice in decisions involving unratifiable acts, acts that seem suboptimal after being chosen. Joyce will elaborate a sophisticated version of CDT that handles unratifiable acts correctly. Along the way, we will explore: (i) the proper way to assess future choices that lie outside one’s current control; (ii) the role of deliberational equilibria in rational decision-making (Skyrms 1990); (iii) the distinction between willful choosing and mere picking, and why it matters when unratifiable acts are in play; (iv) the conditions under which unratifiable acts may be rationally performed; and (v) the right way to incorporate uncertainty about the choices one faces into one’s decision-making. The result will be a better way of thinking about decisions involving unratifiable options and decisions made over time. To test the theory, Joyce will discuss three decision problems that have recently been put forward as counterexamples to CDT. The first is Frustrater (Spencer and Wells, 2019; Oesterheld and Conitzer, 2021), a one-shot decision in which CDT allegedly recommends passing up sure rewards to secure prizes that are almost surely worthless. Two Rooms (Spencer 2021) and Frustration and Delay (Ahmed 202x) are dynamic decisions in which CDT allegedly recommends avoidable sequences of choices that will almost surely produce suboptimal outcomes. Though some philosophers see these examples as existential threats to CDT, we shall see that the theory, properly elaborated and understood, has nothing to fear from them.
-----