SNF Ambizione Grant
The grant runs from 01.09.2024 – 31.08.2028 and focuses on how we learn about and from others based on their affect (i.e., feelings). Science has much to say about learning in objective contexts. From algorithms playing chess to organisms seeking food, the computations involved in adaptive learning and decision making are increasingly well understood. When it comes to affect, however, the corresponding computations are more elusive. Intuitively, this might seem unsurprising: Unlike objective states, affect is inherently intractable and should not be expected to adhere to clear computational principles. However, recent research challenges this intuition by documenting how affect can align with computational principles. This is promising, as a more precise understanding of the subjective holds promise for elucidating contexts that cannot be understood based on objective features alone. Past research has begun to act on this promise, showing that people’s own affect can explain their choice behavior beyond what can be explained by objective features. However, this does not speak to social functions of affect in terms of computations involved in observing other people's affect.The objective of this project is to address this gap and identify systematic computations that tie others’ affective responses to observers’ inferences. To this end, it addresses two questions from a computational perspective: How do we learn about (1) and from (2) others based on their feelings? By revealing systematic computations that link affective responses to social inferences and behavior, this project aims to inform fundamental research along with a broader perspective on how expressions of affect (be it human, artificial, public, or private) may relate to social cohesion.