ELSC Seminar Series

The cognitive science of learning and decision-making has developed almost entirely in separate from the study of emotions. In this talk, I will argue that these two fields of study investigate the same latent processes. I will thus show how emotions and moods map onto fundamental computations of learning and planning, specified within the framework of reinforcement learning. Crucially, acknowledging this mapping offers two major benefits: first, it enables cognitive science to explain the many real-world behaviors that involve emotions; second, accounting for emotions leads to a revision of the algorithms by which people learn and plan. I will illustrate these benefits in three complementary studies that examine different ways in which emotions are involved in learning and decision-making: (1) the bidirectional relationship between reward learning and mood, demonstrated through an intensive large-scale longitudinal study; (2) the mechanistic role of emotions in translating environmental controllability into decision-making, shown through an experimental manipulation of control; and (3) the impact of emotions on performance optimization, revealed through analysis of professional tennis matches. These findings establish a key role for emotions in how organisms adapt to their environments, and provide a computational framework for integrating emotional processes into cognitive models of humans and animals. This framework opens new avenues for understanding both normal variation in behavior and disrupted learning processes in mood and anxiety disorders.