Learning to talk

One of the great achievements of science is that so many of its branches, from astronomy to zoology, have been blessed by such great popularisers — your Attenboroughs, your Sagans, your Dawkinses. Alas, there is one inglorious exception to this marvellous rule — linguistics. A discipline that has produced enormous and enormously important advances over the last century — but not one linguist who has managed to tell the rest of the world about them. Steven Pinker did have a bestseller with The Language Instinct, but he was moonlighting from his day job in neuropsychology.

Linguistics does have one world-class intellectual celebrity, but Noam Chomsky is celebrated mainly for his radical politics, and he has done his very best to make his work on language as arcane and incomprehensible as string theory.

The world outside linguistics departments remains unaware of it, but Chomsky’s crazed theories — about humans’ innate language-learning devices and the deep structure of a universal grammar that creates all languages — have been comprehensively disproved. The new orthodoxy is the empirical school of cognitive linguistics, and Daniel Everett is its star pupil — and the one thinker with the credentials and ambition to try to reach the general public.

Here, Everett takes on one of Chomsky’s daftest claims — that the innate neural gizmo which makes us able to talk didn’t evolve gradually but just turned up, created by accident, by some genetic mutation that miraculously gave us brains wired for words. Chomsky estimates that this fluke happened about 50,000 years ago… when rock art and cave paintings also began to appear. And cue the Twilight Zone theme tune.

Complete nonsense, of course. There is no innate linguistic machinery in our brains, there was no magical quirk in our DNA that gave us a language-learning machine, and it did not all happen 50,000 years ago.