That's an excellent read. The author's reply to his critics is just as enjoyable, if not more so.

For anyone who's vaguely interested, but can't be arsed to wade through the text, the gist of it seems to be:

There's this lass Nadia, who showed a savant ability to draw from the age of 3, particularly horses and cattle. Her drawings are stylistically similar to the cave paintings at Chauvet, Peche Merle and Lascaux. Humphrey reckons that Nadia's autism induced lack of language acquisition left space in the wiring of her brain, so that her vision and motor cortices made some kind of connections that are usually unavailble to most people as their brains develop in early childhood.

He then draws a connection between Nadia's lack of language/drawing skill and the drawing skill of the cave painters, arguing that the level of realism depicted in the caves may have been the result of an analogous level of underdeveloped language in paleolithic humans. He acknowledges that it is very unlikely that paleolithic folk had no language, but rather that their language was possibly restricted to social functions, rather than descriptions of the natural world, (In a manner similar to that proposed by Steven Mithen) and that this left space in their brains for a different perception of the natural world in their minds.

In his rebuttal, Humphrey goes on to offer the suggestion that the differences between paleolithic and modern human minds is not so much due to differences in brain structure dictated by genetics, as it is from differences in environmental stimuli from a more 'developed' culture. I can't bring myself to totally agree with this, as I reckon that other non-cultural environmental factors may have been more important, such as the effects of diet and climate during childhood (including neonatal dvelopment). But he's onto something with the idea that the inherent plasticity of the human brain could produce apparently anomalous artefacts. The point at which humans developed joined up, general purpose language is still open to debate, but one thing is for sure, the bits of brain that are currently used for language were there before we developed our current lingustic repetoire, and they may well have served a different purpose in the days before language. Maybe those cave painters were a crucial part of the bridge between social-only language and general-purpose language. The act of representing the bison may have made it possible for future generations to have a language that included a frame of reference for bison, because the bison had been removed from the outside world and placed into the social sphere.

I'll shut up now, before I go off on one about the possible implications of the weak central coherence mentioned in the paper, and how it might slot into the old '3 stages of trance' model and the cave paintings. 'Cos I'm not really sure what I'm on about when it comes to central coherence.