Appraisal theories of emotion generally presuppose that emotions are directed at various items. They also hold that emotions have motivational properties. However, although it coheres well with their views, they have yet to seriously develop the idea that the function of emotional direction is to guide those properties. I argue that this guidance hypothesis can open up a promising new field of research in emotion theory. But I also argue that before appraisal theorists can take full advantage of it, they must drop their further assumption that to determine an emotion's direction, one need only retrace the process that caused it. Contrary to this retracing view, I argue for an independence thesis: directed emotions are produced by two functionally independent sub-processes. The first, affect-causation, functions in part to produce a state with certain motivational properties given certain representations. The second, affect-direction, has the function of optimally guiding those motivational properties by associating them with representations that may properly be quite dissimilar from the causal ones. By provisionally adopting the independence thesis and empirically testing the guidance hypothesis, I argue that appraisal theorists stand a good chance of significantly increasing the explanatory power of their theories.