ABSTRACT
The role of dysfunctional cognitions in maintaining or contributing to symptoms is central to most theories of depression. The nature of the dysfunction however, is controversial. Beck (1967) emphasizes the presence of negative schemas, others a depressogenic attributional style (Abramson, Metalsky, & Alloy, 1989). Paired with an enabling stressor, dysfunctional cognitions encourage negatively biased beliefs about the self, contributing to depressive symptomatology (for detailed reviews see Coyne & Gotlib, 1983; Gotlib, Kurtzman, & Blehar, 1997; Teasdale & Barnard, 1993; Williams, Watts, MacLeod, & Mathews, 1997). However, there is other evidence that, when evaluating certain causal scenarios, depressed people are more accurate than their nondepressed counterparts (Alloy & Abramson, 1979). This depressive realism stands in sharp contrast to a dysfunctional cognitions view of depression. The most compelling evidence for depressive realism comes from studies that examine learning (Ackermann & DeRubeis, 1991). We discuss here how contemporary learning theory provides insight into the depressive realism phenomenon. We also show how an appreciation of the principles of learning, involving associations and basic motivational states, undermines the argument that depressed people have more realistic cognitions than the nondepresssed. We propose that the data from learning tasks are actually consistent with a cognitive bias and reduced motivation account of depression. Moreover, we argue that both depressed and non-depressed people react the same way to traditional associative and behavioural manipulations but that the depressed have different sensitivities to these manipulations. Specifically we examine two aspects of learning that are different in depression, namely context exposure and motivation, and examine how depressed mood relates to these variables.
