Kangyu Wang
London School of Economics
Incommensurability gives rise to hard choices. A and B are incommensurable when neither option is better than the other, yet they are not equal either (Parfit 1984; Griffin 1986; Raz 1986; Chang 2002, 2010; Sugden 2009). In a hard choice, the agent has no more normative reason to choose either option. Reason internalism (Williams 1981; Broome, 1993) says that normative reasons are based on desires. Granting all these, it follows that in hard choices one desires neither option more than the other yet does not desire them equally. It further follows that if to resolve a hard choice by deliberating requires one to come to have more reason for one option than the other, one must desire one option than the other after deliberating.
I develop an internalist model for this process, drawing on
(1) a model of incommensurability developed by Hajek and Rabinowicz (2021),
(2) a theory of multidimensionality developed by Hedden and Muñoz (2023),
(3) analyses of the objects of desires of Savage (1954) and Jeffrey (1965), and
(4) discussions on desire moderation by Sinhababu (2009) and Yip (2022).
A choice is hard when there are multiple permissible orderings among the options in terms of the desires for those options, resulting from the multidimensionality of desires and from the agent’s taking multiple ways to aggregate those multidimensional desires as permissible. To resolve a hard choice, the agent has to moderate how strong those desires are.
Ruth Chang’s alternative theory of hard choice is combined with a hybrid theory of reason (2022a, 2022b). Chang says that when and only when external reasons presented by the world “run out”, one can exercise one’s normative power to create a will-based internal reason for oneself to resolve this problem and that her theory allows one’s human agency to play an “active role” to play in determining what reasons one have. Chang’s theory of normative reason has a bootstrapping problem: one can sometimes make what they primitively desire to do the thing that they ought to do by committing a mistake. Moreover, The “active role” Chang grants agents is not really so active: there is nothing one can actively do about whether one has an active role to play or whether to play an active role when such a role is available. For these reasons, my internalist theory is superior to Chang’s.
Chair: Kendra Gordillo
Time: September 13th, 12:00 – 12:30
Location: SR 1.007 (online)
SOPhiA 2026 