When performing an action of a certain kind, an agent typically has several reasons for doing so. I shall borrow Davidson’s term and call these rationalising reasons (Davidson 1963, 3). These are reasons that allow us to understand what the agent regarded as favourable features of such an action. (There will also be reasons against acting, expressing unfavourable features of such an action, from the agent’s point of view.) I shall say that R is a rationalising reason of agent X’s for K-ing iff R consists of (i) a desire of X’s to L and (ii) a belief of X’s that K-ing promotes L-ing (to be discussed shortly). It is frequently said that when an agent X is K-ing and has several rationalising reasons for K-ing, not all of those reasons are reasons for which X is K-ing, that motivate X’s K-ing, or that explain X’s K-ing. In this paper I challenge this view. The notions of a reason for which X is K-ing, a reason that motivates X’s K-ing, and a reason why X is K-ing incorporated in this popular view I assume to be nonpragmatic notions that do not depend on a context of enquiry. It should also be noted that there are pragmatic reason notions that are especially likely to be intended when the definite article is used