The dominant model in research on autism is that of a hard-wired "theory of mind" that operates in neurotypical folks but that doesn't operate (maybe the wires get crossed in some more-or-less literal sense) in those developing children who end up "on the spectrum." Whereas the neurotypicals come to understand without reflexion that there is another person behind the various human faces they encounter, this never comes easily to those with the crossed wires. On this theory, (call it the meta-theory of mind) many if not all of the characteristic symptoms of the disorder derive from that point. It is a neat (meta-) theory, but it is worth mentioning now and then that it isn't established fact. Indeed, the more neurologists get to work trying to home in on this wired-in theory of mind, the more evasive it seems. And (assuming the paradigm) there are theory-of-mind deficits that one would not call autistic. OTOH, I'm sure there is something...