On Reasons and Epistemic Rationality

In a recent paper, John Hawthorne and Jason Stanley propose an analysis of how knowledge and action relate to each other. According to their Reason-Knowledge Principle (RKP), it is appropriate to treat the proposition that p as a reason for acting iff we know that p, for p-dependent choices; within their account, in addition, knowledge delivers probability 1. Hawthorne and Stanley also note that sometimes it is intuitively rational to act on partial beliefs. What is appropriate to treat as one’s reason for action, in this case, is the epistemic probability of p conditional on the agent’s total knowledge K. RKP then requires that one knows that P (p | K) = r (for some r). Here I am not concerned with the general formulation of RKP, but I shall focus exclusively on the more restricted contention according to which agents should not invoke probability claims as reasons unless they know that such claims are true. I believe there are grounds to think that this is a problematic demand.