Verified by 
It is always gratifying to have your theory confirmed, but more satisfying still when it is endorsed using someone else's approach and terminology—particularly if its author has specifically ruled out what your theory proposes!
The extreme male brain theory of autism was originally out-lined by Asperger and later developed by Simon Baron-Cohen. It suggests that autism represents a pathological over-development of typical male cognitive tendencies.
According to Baron-Cohen's systemizing/empathizing model of cognition illustrated in the diagram, males averagely are Type S, with systemizing better than empathizing. Females are generally the other way round: Type E, with empathizing skills superior to their systemizing ability. Autistics take the male tendency even further: Extreme Type S, with major deficits in empathizing and over-development of systemizing. However, the paradigm suggests that an Extreme Type E also exists with the opposite cognitive configuration: a candidate for an extreme female brain equivalent of autism.
Baron-Cohen explicitly rejected the idea that psychosis might be to the female brain what autism is to the male one in his book, The Essential Difference. Nevertheless, a paper soon to be published in Personality and Individual Differences (Volume 49, Issue 7, November 2010, Pages 738-742) by Mark Brosnan, Chris Ashwin, Ian Walker, and Joseph Donaghue asks if an "Extreme Female Brain" (EFB) "can be characterized in terms of psychosis?" These authors equally explicitly answer that it can.
The study investigated the relationship between levels of empathizing and systemizing, as well as self-report measures of psychosis, depression and anxiety, in 70 healthy female undergraduates. The findings showed that what Baron-Cohen calls hyper-empathizing was positively correlated with higher psychosis, but was not associated with measures of depression or anxiety. Furthermore, the two psychotic items that correlated most significantly with hyper-empathizing were mania (feeling elated) and paranoia (feeling others are against you). The researchers point out that "these findings are consistent with Crespi and Badcock's theory linking the EFB to psychosis spectrum disorders, and in particular paranoia and ‘positive symptoms'." Indeed, they add that their findings show that hyper-empathizing in a non-clinical female sample is "associated with higher psychosis scores, providing empirical support for the autism-psychosis model and suggesting hyper-empathising may be consistent with a pathological profile of EFB."
They also comment that "Hyper-empathising may constitute a case where people attribute intentionality when it is not there," and add that such people "might attempt to use social-emotional explanations to explain non-social elements of the world around them." This epitomizes what I would call hyper-mentalism because it clearly goes far beyond mere empathizing, and in my book I devote the first two chapters to explaining why, despite a strong desire to do so, I was unable to subscribe to Baron-Cohen's view of autism as representing excessive systemizing and defective empathizing.
As I argue at length there, autistic savantism—the epitome of autistic cognitive style—is not so much systematic as mechanistic, and autistics cannot be seen as simply deficient in empathy (even if empathic deficits are a part of the disorder). In the next two chapters I argue the case for seeing psychotic symptoms as not just the opposites of those of autism, but as instances of hyper-mentalism.
Finally, I show how the diametric model is explained by genetic conflict and in passing point out that autism is not so much an extreme male brain disorder as an extreme paternal brain one. This immediately explains why females can be autistic because people of both sexes have paternal brains (the limbic system in the main: so called because it is built by paternally-active genes). And an obvious deduction is that, if psychosis is the diametric opposite of autism, it should be associated with an extreme maternal brain (mainly the frontal cerebral cortex and underlying striatum, where maternal genes are predominantly expressed).
Researchers who start out with a particular theory's paradigm and terminology in mind and then find themselves confirming it can sometimes arouse suspicions of having been biased in the theory's favour. Clearly, no such suspicions attach to this particular study. On the contrary, these researchers had little choice but to use existing measures of autistic and psychotic cognition based on Baron-Cohen's model simply because alternatives derived from the diametric mentalistic/mechanistic model have not yet been developed. Indeed, if you hadn't done your homework and read this paper without sufficient care, you might almost get the impression that it was Baron-Cohen and not Badcock and Crespi who had been vindicated by this remarkable research!
(With acknowledgements and thanks to Marco Del Giudice for bringing this to my attention.)
