³ÉÈËVRÊÓƵ

How we learn to speak is a mystery. Myrna Gopnik claims to have found a gene which transmits bad grammar. The language gene is the holy grail of linguistics. Find that, and you are up there with Darwin. Noam Chomsky, following Darwin's belief that language ability is "an instinctive tendency", first articulated the idea that children are innately equipped with a language template - the so-called "universal grammar" - that allows them to distil the syntactic patterns out of parental speech.

But, as his colleague at the Massachusetts Institute of Technology, Steven Pinker puts it, Chomsky is "a paper-and-pencil theoretician", and so he has never proved the point. That has been left to others. One, a great Chomsky fan, is Myrna Gopnik from McGill University in Canada. She thinks she has found evidence for a language gene - one which disrupts grammar rather than controls it - after studying an English family with extraordinary language problems. Three generations of the K family have what linguistics specialists term "an inflectional deficit": they say things like "Carol is cry in the church" and they cannot work out the plural of "wug".

On discovering this, Gopnik telephoned Chomksy to tell him the good news. Since then, several leading specialists have voiced their approval of the findings, including the evolutionist John Maynard Smith and Pinker himself, who backs the "suggestive evidence for grammar genes in the sense of genes whose effects seem most specific to the development of the circuits underlying parts of grammar". Others, however, remain deeply sceptical, including Reading-based Paul Fletcher.

Fletcher points out that the K family suffer other inherited disabilities - notably problems of pronunciation as well as jaw, tongue and lip co-ordination. "These facts," he says, "mean that it is not possible to interpret their problems solely in terms of an inherited deficit in the construction of inflectional rules." By way of emphasis, he adds that their understanding of language is "relatively unimpaired".

So who is right? That is the question bugging academics. Linguistics specialists are now hunting for the chromosomal locus of the putative gene, and the human genome project is expected to guide them in the right direction. Already, the Chomskian idea of a universal grammar is pretty much accepted, as the extraordinary case of the multi-lingual "savant" Christopher suggests.

Edinburgh professor Jim Hurford thinks the answer will be found in the next 20 years, and certainly by the middle of the next century. But the prospect of a Dr Doolittle-type conversation with animals implanted with the language gene is likely to remain science fiction.