V. S. Ramachandran at Edge.org:
What is the self? How does the activity of neurons give rise to the sense of being a conscious human being? Even this most ancient of philosophical problems, I believe, will yield to the methods of empirical science. It now seems increasingly likely that the self is not a holistic property of the entire brain; it arises from the activity of specific sets of interlinked brain circuits. But we need to know which circuits are critically involved and what their functions might be. It is the “turning inward” aspect of the self — its recursiveness — that gives it its peculiar paradoxical quality.
It has been suggested by Horace Barlow, Nick Humphrey, David Premack and Marvin Minsky (among others) that consciousness may have evolved primarily in a social context. Minsky speaks of a second parallel mechanism that has evolved in humans to create representations of earlier representations and Humphrey has argued that our ability to introspect may have evolved specifically to construct meaningful models of other peoples minds in order to predict their behavior. “I feel jealous in order to understand what jealousy feels like in someone else” — a short cut to predicting that persons behavior.
Here I develop these arguments further. If I succeed in seeing any further it is by “standing on the shoulders of these giants”. Specifically, I suggest that “other awareness” may have evolved first and then counterintutively, as often happens in evolution, the same ability was exploited to model ones own mind — what one calls self awareness. I will also suggest that a specific system of neurons called mirror neurons are involved in this ability. Finally I discuss some clinical examples to illustrate these ideas and make some testable predictions.
More here.