Social value learning shifts conceptual representations of faces
Ariana M. Familiar1, Sharon L. Thompson-Schill1; 1University of Pennsylvania
Values drive our behavioral choices. Ample research has explored the neural underpinnings of value-based computations related to decision-making. Moreover, recent neuroimaging work has shown learning monetary values of novel objects influences neural responses in early visual cortex (Persichetti, Aguirre, & Thompson-Schill, 2015), and results in increased functional coupling between visual and valuation systems (Mattar, Thompson-Schill, & Bassett, 2017). However, behaviorally relevant values that we associate with real-world objects are often not monetary. The present study examines social value learning of naturalistic stimuli. Through a multi-day paradigm, participants learned social values (generosity) associated with different people (face images). Generosity was defined as the proportion of a pool of points shared by a given person on average. Additionally, before and after learning, participants completed a task in which they arranged the face images according to similarity, where distances between faces defined their similarity (e.g. closer together are more similar). The difference in pairwise distances before and after learning was used as a measure of the shift in conceptual space due to learning. Shifts were positively correlated with generosity values, but were not correlated with perceptual similarity or point magnitude. In other words, faces of more similar social values became closer together after learning. Furthermore, shifts were positively correlated with a post-learning index of preference to interact with a person in a future cooperative game. These results indicate that learning social values of faces influences their representations in conceptual space, and such representational changes are related to propensities in future behavior.
Topic Area: LONG-TERM MEMORY: Semantic