In recent work,[1] David Velleman pursues a version of what Bernard Williams calls the Kantian strategy.  The point of this strategy, Velleman says, is to “account for the objectivity of morality without positing a normative reality of which judgments of right and wrong can be true.”  To this end, Velleman tries to explain how the demands of practical reason can have a kind of objectivity even if there is no normative reality of which reason judgments can be true.  The key is to recognize that while all reasons rest on motives that are present in those for whom they are reasons, some reasons rest only on motives that are necessarily shared by all who are subject to reasons.  He thinks this is particularly clear in the case of reasons for belief, and extends the point to reasons for action.  In this paper I argue that Velleman’s theory of reasons, even if it is true, fails to support the hypothesis that there is some one motive that is necessarily shared by all those who are subject to reasons for belief, and likewise fails to support the hypothesis that there is some one motive that is necessarily shared by all those who are subject to reasons for action.  To make good on his Kantian strategy, Velleman must establish the hypothesis in some other way.