Two-Boxing Is Not Rationally Required (Departmental Seminar)

Abstract:

In this paper I argue that any theory implying that two-boxing is the rationally required act in the Newcomb case is mistaken. I do so by building on recent work by Williamson (2021) and Mackie (2022). Williamson defends Causal Decision Theory against the challenge presented by Egan (2007), and in particular the Psychopath Button Case, appealing to a case he calls ‘Two Buttons’. His argument is that if CDT is correct, or any decision theory advocating two-boxing as the rational act in the Newcomb case is correct, pressing the button is the rationally required act in the Psychopath Button case. Furthermore, he argues, this is not just an opportunity for the evidential theorist to apply Modus Tollens since pressing the button in the Psychopath Button case is right. I elaborate on and endorse Williamson’s argument for his first claim – that if two-boxing is rationally required in Newcomb pressing the button is the rational act in Psychopath Button. But I deny his second. I argue, appealing to the ‘Knowledge Principle’ of Mackie (2022), that the agent in the Psychopath Button case can have no aim in pressing, so that pressing cannot be the rationally required act. I conclude that using Modus Tollens one can indeed establish that any decision theory which advocates two-boxing alone in Newcomb is mistaken.

Delivered in English.
All are welcome.