In this paper, three major tasks are undertaken: 1. showing that our traditional understanding of the nature of emotions, which equates them with certain sorts of inherently private affective experiences that are brought about by various causal factors (esp., exciting events, cognitive interpretations, and biological states of affairs), does justice neither to the conceptual nor to the empirical facts; 2. presenting an alternative conception of emotions as a specific class of perceived relationships between oneself and some person, object, event, or state of affairs; 3. demonstrating how this relational conception of emotions heuristically suggests a far greater range of therapeutic options than does the traditional view.