In what ways, and to what extent, is social cognition distinguished from cognition in general? And how do data from cognitive neuroscience speak to this question? I review recent findings that argue social cognition may indeed be specialized, and at multiple levels. One particularly interesting respect in which social cognition differs from the rest of cognition is in its close interaction with the social environment. We actively probe other people in order to make inferences about what is going on in their minds (e.g., by asking them questions, and directing our gaze onto them), and we use the minds of other people as a collective resource. Experiments from our own laboratory point to the amygdala as one structure that is critically involved in such processes.