Abstract
Considers and rejects the view that motivating reasons are best specified in psychological form rather than in non‐psychological form. Asks how we should then explain the fact that, if S acts for the reason that p, S must believe that p. It also asks whether the fact that the agent can be mistaken forces us to prefer the psychological form of explanation. It is argued that it does not, because explanation in terms of reasons is not factive; we explain S's action by appeal to what he believed even when he is wrong.