Interactive representations of power and status signals in learning social hierarchies.

Abstract

Social hierarchies, which organize our social relationships, have heterogeneous effects on choice behavior. Two central dimensions of social hierarchies, status, which is earned through recognition of competence, and power, which is the capacity to control the outcomes of others, may contribute to different behavioral profiles. Yet, the neurocomputational processes by which status- and power-related signals are learned based on social feedback, and influence subsequent decision-making, remain unclear. Here, we combine functional neuroimaging with computational modeling to examine how individuals acquire knowledge about others' status and power, and how these learned hierarchy signals shape social influence on advice-taking during a risky choice task. We found that the medial prefrontal cortex covaried with status estimates, whereas the amygdala covaried with power estimates. Moreover, estimates in one dimension influenced choices in the other during learning. This indicates interactions between status and power signals, with retrosplenial cortex activity associated with hierarchy signals beyond immediate choice demands. Finally, status broadly modulated response biases toward both safe and risky advice, whereas power primarily increased bias toward safe advice. These findings characterize how experimentally instantiated status and power signals are learned and influence risky decision-making, indicating partially separable yet interactive representations of these hierarchy signals.