BOOK REVIEW / Xxsy tx xndxrstxnd wxrds wxthxxt vxwxls: 'The Language Instinct' - Steven Pinker: Allen Lane, 20 pounds

A FRIEND of mine once came across some unusual cut flowers that were so spectacularly red she thought they must be fake. 'These are amazing,' she said. 'Are they dyed?' The florist shook his head. 'No, no, not at all,' he said. 'Just put 'em in a drop of water and they'll be fine.'

It would be going it a bit to say that drastic punning is a central theme in Steven Pinker's dazzling new book on the nature of language - he has some very serious concerns. But his work is a tribute to the miraculous intricacy of the way humans talk and write, so we can be pretty sure he would have enjoyed the florist's faux pas: he might have written a chapter about it, evaluating the margin for error in the glottal-labial hypertensile expansion of the tongue root, or composing tree-diagrams to indicate that what we hear does not always square with what we have been told.

Pinker's big idea is that language is an instinct. Not a trick, not a talent, not even a slog - just an unavoidable basic instinct, as natural as eating with our hands or walking on our feet, as innate to us as flying is to geese. It is a universal feature of all human societies, he argues: no mute clan has ever been discovered, and the apparent gobbledygook that is spoken by remote tribes in Papua New Guinea or Amazonia turns out to obey an agile and systematic grammar.

Children learn language, he concludes, not because they are clever or well taught, but because they can't help it. If you doubt that language is a mazy process of constant invention, jxst lxxk xt hxw xxsy xt xs tx xndxrstxnd wxrds wxthxxt vxwxls. Pinker watches Mum and Dad's proud gurgling, and compares it to the African tribe that believes you have to teach children how to sit and walk - parents waste months building little piles of sand to encourage their babies to do something they would have done anyway. 'No one has yet located a language organ or a grammar gene,' he declares. 'But the search is on.'

This is all immensely fine and trenchant, and Pinker embarks on his argument with brilliant dash and swagger: 'I want to debauch your mind with learning,' he begins - a chirpy introduction to a work which is going to include words like sesquipedaliaphobia and floccinaucinihilipilificational. But what a wonderful ambition. Not many writers aim this high.

Pinker is an avid fan of Chomsky's theory of universal grammar ('A visiting Martian would surely conclude that aside from their mutually unintelligible vocabularies, Earthlings speak a single language') but the disciple, unlike the master, is a canny writer and a bit of a wag. It is possible he does not realise how elevated is the company he keeps. 'Anyone who goes to cocktail parties,' he writes, 'knows that one of Chomsky's main contributions to intellectual life is the concept of 'deep structure'.' But readers tempted to retort, Not round here they don't, will soon be disarmed by a manner both meticulous and genial. Besides, in his heroic chapter on how language fits with natural selection, Pinker plucks up the nerve to call Chomsky 'flip'.

Implicit in Pinker's thesis is the notion that thought and speech are distinct and separate. He mentions the routine modern theory that words condition what can be said or thought, and dismisses it as 'wrong, all wrong'. Then he explains: 'The idea that thought is the same thing as language is an example of what can be called a conventional absurdity; a statement that goes against all common sense but that everyone believes because they dimly recall having heard it somewhere and because it is so pregnant with implications.'

Much recent thinking about language, he suggests, is motivated by political fantasies. We enjoy hearing that Eskimos have hundreds of different words for snow: it confirms us in our relativist belief that primitive people are in truth more subtle and closer to nature than we bourgeois imperialists. The trouble is - it isn't true. Eskimos have relatively few snow- words (two, according to one dictionary). It is a polar myth, a conventional absurdity, to suggest otherwise. Words, in any case, are an arbitrary product of the reflex to name and share things. We call a dog a dog; in France they call it a chien - big deal; pigs go oink in Britain and boo-boo in Japan - same difference. To Pinker these are just surface swirls: the main thing is the urge to communicate.

Pinker chooses an ugly term - mentalese - to describe the fluctuations of our interior life. 'People do not think in English, or Apache or Chinese,' he argues. 'They think in a language of thought . . . Knowing a language is knowing how to translate mentalese into strings of words.' Yet in his own way Pinker is a relativist, too. He is wonderfully tough on tedious 'language mavens', with their precious little fads posing as absolute truths. And he knows how foolish and pompous it is to see dialects as corrupt versions of the master tongue. 'A language,' Max Weinreich once wrote, 'is a dialect with an army and a navy.'

This feeling prods Pinker towards a lyrical attention to minutiae, which in turn confirms the old saw that God dwells in the details. The Language Instinct vibrates with delicious asides and poignant discoveries. 'There are no little silences between spoken words the way there are white spaces between written ones,' Pinker points out. The improvisatory nature of birdsong he calls 'Charlie Parker with feathers'. He celebrates the 60,000 muscles in an elephant's trunk, and takes us on a marvellous guided tour of our own vocal tract, alert to the acoustic effects of lips, gum ridges, soft palates, resonant cavities and vocal folds in the larynx.

There are times when even Pinker plods. He commits the odd conventional absurdity himself in defiance of common sense - as when he insists that children don't learn by imitation (how many children of English-speaking parents grow up speaking Swahili?) And you do have to worry about a man who can cite Johnson's definition of a lexicographer as a 'harmless drudge' and then say that this is an unfair stereotype - surely we can all see that this is only a falsely modest in-joke. He also seems overstruck by Chomsky's celebrated example of grammatical nonsense: 'Colourless green ideas sleep furiously.'

This was a cunning attempt to show that nonsense could obey grammar, and to show by extension that mere grammatical correctness is no guarantee of good sense. Pinker is too busy celebrating its clever meaninglessness to notice that it does, in a way, make sense. Each of the words in the sentence has a metaphorical twist: colourless means drab, green means raw, and so on. We can easily read it as a rare poetic line about how some as yet unformed, revolutionary energies are about to kick off their bedclothes and leap into the world - or something. At any rate, we sense here that language is even more anarchic than Pinker, with his flow charts and well-organised wit, can see. Like all linguists, he constantly fidgets with Alice in Wonderland, with one of those annoying maybe-this-will-help-explain-it smiles.

But these are minute tremors in an otherwise terrific book. Words can hardly do justice to the superlative range and liveliness of Pinker's investigations. Actually, that's not quite an accurate version of my mentalese. What I meant to say is that The Language Instinct is absolutely nsfluqdfh dsfuyehajxa kodtrwfe sdoiwrbfbn nd-nd] Yes, that about sums it up.