Plato long ago suggested that to understand the world, one needed to "carve nature at its joints," that is, one needed to classify the objects and phenomena of the world into deep and meaningful categories that accurately reflect valid distinctions in nature. Since the early 20th century, clinicians and researchers interested in the phenomena of depression have suggested that depression is probably a heterogeneous category, waiting to be "carved at its joints" (e.g., Craighead, 1980; Gillespie, 1929; Kraepelin, 1913). Historically, a wide variety of nosological distinctions have been proposed to classify the depressive disorders into subtypes, such as bipolar-unipolar, endogenous-reactive, psychotic-neurotic, and so on. Most of these subtyping schemes have grouped people together on the basis of similar symptoms (either through clinical observation or quantitative statistical approaches), with the expectation that the resulting subtypes would be associated with distinct causes, courses, or treatments. More recently, investigators from a cognitive perspective have used a theory-based approach to the subdivision of the depressive disorders and have proposed that one or more distinct cognitive or personality subtypes of depression may exist in nature. For example, in the hopelessness theory of depression, Abramson, Metalsky, and Alloy (1989) hypothesized the existence of a hopelessness depression subtype with a particular cognitive vulnerability-stress etiology and distinct symptoms, course, treatment, and prevention. Similarly, Beck (1983; 1987) and Blatt (1974; Blatt & Zuroff, 1992) have hypothesized distinct sociotropic and autonomous or dependent and self-critical subtypes of depression, respectively, each with a unique constellation of associated symptoms and a cognitive or developmental diathesis-stress matching etiology. Each of these proposed cognitive/personality subtypes of depression has generated considerable theoretical interest and a growing body of empirical research over the last 5-10 years.