Computational origins of cortical brain circuits for social cognition.

No domain rivals the importance and complexity of our social lives. Given the principle of exaptation in biology — the repurposing of existing structures for new functions — it is likely that brain regions originally evolved to perform computations in one context have been recruited for related computations in other contexts. From this point of view, brain regions for supporting social cognition should also be active in non-social contexts in which the computational demands mirror those of social situations. In this Perspective, we examine the computations required to navigate the social lives of human and non-human primates and identify brain activity patterns responsible for these functions, assessing the degree to which similar activity carries out similar computations in non-social contexts with analogous computational demands. This approach offers a unifying framework that bridges social and non-social domains and has implications for multiple areas within cognitive neuroscience, as well as emerging fields such as human–artificial agent interactions. Brain activity in regions traditionally linked to social cognition in primates also supports analogous computational demands in non-social contexts. In this Perspective, Mahmoodi and Rushworth examine the computations required to navigate the social lives of human and non-human primates, arguing how shared neural mechanisms carrying out similar computations in non-social contexts indicate computational rather than contextual specialization in the ‘social brain’.