Social learning through prediction error in the brain

Abstract

Learning about the world is critical to survival and success. In social animals, learning about others is a necessary component of navigating the social world, ultimately contributing to increasing evolutionary fitness. How humans and nonhuman animals represent the internal states and experiences of others has long been a subject of intense interest in the developmental psychology tradition, and, more recently, in studies of learning and decision making involving self and other. In this review, we explore how psychology conceptualizes the process of representing others, and how neuroscience has uncovered correlates of reinforcement learning signals to explore the neural mechanisms underlying social learning from the perspective of representing reward-related information about self and other. In particular, we discuss self-referenced and other-referenced types of reward prediction errors across multiple brain structures that effectively allow reinforcement learning algorithms to mediate social learning. Prediction-based computational principles in the brain may be strikingly conserved between self-referenced and other-referenced information.

Fig. 1

Key brain regions involved in representing information with respect to another individual. These areas are often implicated in mentalizing, detecting the beliefs of others, or signaling decision variables concerning another individual. See texts for how these areas are implicated in representing information with respect to another individual. The insets with coronal magnetic resonance images indicate the sections (red line) that correspond to the Nissl-stained sagittal slices. The dotted outline around an area indicates that this area is projected medially from the lateral surface for the purpose of including the area on a more medial aspect of the brain. Adapted with permission from http://www.brains.rad.msu.edu, http://brainmuseum.org, supported by the US National Science Foundation and the National Institutes of health. ACCg anterior cingulate gyrus, ACCs anterior cingulate sulcus, dlPFC dorsolateral prefrontal cortex, dmPFC dorsomedial prefrontal cortex, IPL inferior parietal lobule, MTG medial temporal gyrus, PCC posterior cingulate cortex, PrCu precuneus, STS superior temporal sulcus, TPJ temporoparietal junction, vmPFC ventromedial prefrontal cortex

Fig. 2

Key brain regions that have been shown correlates of self-referenced prediction errors (in yellow) or other-referenced prediction errors (in purple), or both kinds of prediction errors (in green) in the domain of actions (a) and reward/value outcomes (b). For motor learning-related errors, we only provide representative areas as they are beyond the scope of this review. It is worthwhile to note that the distributions of these areas for their involvements in self-referenced or other-referenced prediction error signaling are naturally constrained by the amount of research examining different types of prediction errors. The insets with coronal magnetic resonance images indicate the sections (red line) that correspond to the Nissl-stained sagittal slices. The dotted outline around an area indicates that this area is projected medially from the lateral surface for the purpose of including the area on a more medial aspect of the brain. Adapted with permission from http://www.brains.rad.msu.edu, http://brainmuseum.org, supported by the US National Science Foundation and the National Institutes of health. ACC anterior cingulate cortex, ACCg anterior cingulate gyrus, ACCs anterior cingulate sulcus, dlPFC dorsolateral prefrontal cortex, dmPFC dorsomedial prefrontal cortex, LHb lateral habenula, LIP lateral intraparietal area, MTG medial temporal gyrus, OFC orbitofrontal cortex, SC superior colliculus, SN substantia nigra, STS superior temporal sulcus, TPJ temporoparietal junction, vmPFC ventromedial prefrontal cortex, VS ventral striatum, VTA ventral tegmental area