Emotion as Information

Our perception of the world is limited to what we can directly experience. However, even from a quick glimpse at the photos on the right, you can recover parts of the scenes that are not explicitly portrayed: an exciting sports game in Scene 1, a gruesome horror movie in Scene 2, and a peacefully sleeping baby in Scene 3. How is this possible? Among many bits of information present in these photos, the most critical clue comes from the emotional expressions of people in the scene. Going beyond using emotional expressions as indicators of others’ internal feelings, we intuitively use these expressions to draw rich inferences about unobservable aspects of the world.

By combining developmental and computational approaches, I study how humans harness others’ emotions as a source of information for learning. Grounded in a computational framework that formalizes how external events and people’s mental states give rise to their emotions, my research demonstrates how infants, children, and adults solve an inverse inference problem: using observed emotional expressions to recover unobservable aspects of the world. The overarching goal of my research program is to provide a more complete understanding of how the human mind works by studying the role of emotion in human learning and reasoning.