This dissertation defends a cognitivist alternative to the Humean belief-desire theory of motivation against standard philosophical arguments. Moral judgments influence our action. For instance, someone might donate to charity because she believes she has a duty to give back to her community. According to the Humean orthodoxy, some additional state—some passion or desire—is needed to explain her action. She may want to donate the money, to give back to her community, or to fulfill her duty. Yet there must be something (...) she wants, the Humean insists, because only desires are capable of moving us. Even moral judgment is no more than desire’s slave. This dissertation explores the possibility that cognitive states are capable of playing a directly motivational role. I argue that the standard philosophical arguments against this possibility do not survive close scrutiny. Instead of proceeding from assumptions about rationality, morality, and agency that frequently drive motivational cognitivists, my arguments are distinctive in that they are built largely out of Humean materials; these arguments show how cognitivism is compatible with many of the considerations Humeans have used to make their account seem compelling. For instance, agents who are unmoved by their moral judgments are often taken as evidence for the Humean Theory. This is odd, since agents are not uniformly moved by their desires either. Moral beliefs and desires seem as though they may be closely analogous in this respect. I also try to show that desire-based motivation might serve as a useful model for cognitive motivation by arguing that cognitivists can explain motivated action in ways that parallel desire-based explanations. While these cognitivist explanations are committed to the existence of besires, I argue that this is no problem for the view. Humean a priori proclamations that besires would be incoherent or absurd notwithstanding, the arguments of this dissertation suggest that besires are not so bizarre. Indeed, I argue that their existence would follow from plausible empirical hypotheses. (shrink)