How the Mind Works, by Steven Pinker Penguin 660pp, £25 in UK
In a dumbed-down media world, our vague fear of the bloat-brained intellectual has made a star of Steven Pinker, the linguist from the Massachusetts Institute of Technology with the sunny, suburban sense of humour veined with collegiate irony. After the surprise hit of his The Language Instinct (an enormously entertaining popularisation of the ideas of his mentor, Noam Chomsky), it must have seemed logical, for both his publisher and himself, to move on to the grander theme of the Mind, and produce the Great American Doorstopper on the subject.
That may seem cynical, but despite the title (Pinker begins by admitting "we don't actually know how the mind works"), this is less a state-of-the-art round-up than a thesis, driven through a vast, fragmented terrain of anecdotal information culled from papers, colleagues and general reading. Starting from a computational model of the brain as a self-assembling symbol manipulator, Pinker argues that the mind is a product of natural selection. He trots out the old hypothesis that the highly visual brains of tree-dwelling apes produced visual, spatial and (by complex analogy) abstract reasoning which in turn opened up a new "cognitive niche" for a social hunting/foraging hominid on the open savanna. So far, so fair enough; but Pinker is determined at every turn to argue his case, some of which is contentious only within his own habitat, the federally-funded halls of academe. Dismissing the Cartesian homunculus (the oft-derided little angel of consciousness in the neural cockpit), he is less concerned with an autonomous "I" than with the mind as complex software running on neural circuitry - an approach he backs up with "revelations" from cognitive psychology and artificial intelligence research.
He sets about "reverse engineering" the brain to see in theory how it works. He does so in a long, tiresome chapter on "neural networks" and "the production system", a kind of hypothetical, hierarchical Turing machine which crudely models problem-solving with parallel-processing modules, themselves assemblies of sub-programming "demons" competing with and complementing each other. Very much in the vein of Marvin Minsky's "society of mind", this is dry stuff, and Pinker's vocabulary of "thingamabobs" and "getting fuzzy on technicalities" "to keep things simple" makes you wonder if he got his own head around the abstruse theories of the artificial intelligence boffins, still scuppered by frame problems such as "salience".
On the core strand of evolution, Pinker owes much to Stephen Jay Gould and Richard Dawkins (surprisingly, he does little justice to the metaphoric elegance of Dawkins's "memes"). Breezily outlining the scale and pace of evolutionary adaptation, he predicts that the current dating of the archaeological revolution (some 50,000 years ago) and that of the "mitochondrial Eve" (at least 100,000 years) will converge when corrected for African evidence. He is thinner on how modernity affects the gene pool, or why a foraging life-style should have given rise to our mental capacities.
On vision, Pinker cherrypicks a few familiar perceptual illusions: Brewster's pop-outs, anaglyphs (3D movies), autostereograms, or the Perky effect (where volunteers, imagining a banana, wonderfully fail to see a dim projection of same on a screen), etc. But, puttering about with pattern-recognition and fuzzy logic, the furthest he strays into actual physiology is to cite research findings that more optical-brain disruption occurs in young animals who have one eye covered for long periods (rather than both), due to the fierce competitiveness of neurons, once they become stimulated to begin circuit-laying the optic chiasma and the visual brain.
The book gets chattier after that, as the congested thickets of theory give way to Pinker's free-association jumbles of snippets from anthropology and linguistics, with allusions ranging from Darwin, Berkeley and Locke to The Twilight Zone, Star Trek and Asimov's Rules for Robots. Written, it seems, like an Edward de Bono book (by accretion, under headings), it resolves into long chapters about the brain's "scramble for unity"; anomalies such as geniuses, psychotics or romantic love; families, altruism and the incest taboo; and finally, a spin through the arts. His closing remark is that our own "awareness" may be beyond our conceptual grasp, but that such "cognitive closure" is made up for by "combinatorial infinity" (a concept imported from linguistics).
Although he is always thought-provoking, and often fun, Pinker's lack of grounding in neurophysiology, or indeed biology, serves him ill in positing any new, or even unified, macrotheory of mind. Despite the fact that he heads MIT's Centre for Cognitive Neuroscience, his own discipline is psycholinguistics (he cites a co-study on mental rotations of two- and three-dimensional symbols); and he largely ignores data gathered from the open-skull horrors practised on lab monkeys, or human waking-brain PET scans and magnetic resonance imaging.
As a result, this is a highly incomplete work of speculation, and you don't learn a great deal about brain gender differences, adult learning, models of memory, etc. Despite the hype, this is a big, glib, pop text. Pinker might have taken another year or two to digest the field fully. Only the winsome colours of his style may heave this tome up alongside Dennett's Consciousness Explained, Oliver Sacks's case histories, or Douglas Hofstadter's periodic outings - which remain the buzziest popular classics on the subject.
Mic Moroney is a journalist and critic