Two fanzines for high-brows

Geoffrey Sampson fails to be bowled over by Chomsky's acolytes.

Noam Chomsky is world famous for contributions to two apparently quite separate intellectual domains - language and human cognition, and radical politics. In 1979 I published a book that argued that there were strong links between Chomsky's linguistics and his politics. The book had a largely hostile reception; many critics seemed to find it absurd to link such different domains. Chomsky himself has often been reluctant to recognise the connections.

But the truth is that they were there all along; and now James McGilvray and Neil Smith have simultaneously brought out surveys of Chomsky's structure of thought that both make the point that his views on language, and his views on politics and current affairs, are all of a piece. Unlike my book of 20 years ago, which argued that Chomsky's linguistics and his politics were both fundamentally misguided, for McGilvray and for Smith, Chomsky is right in all he touches. Indeed, the existence of links between Chomsky's linguistic and political thought is the only detail on which either author takes issue with the master.

Of these two similar books, Smith's is to my mind the better. McGilvray, who is evidently a philosopher rather than a linguist, sometimes misunderstands Chomsky's ideas about innate knowledge of language. For instance, he represents Chomsky as believing that our innate linguistic knowledge is so complete that young children never make any grammatical mistakes (except for over-regularisations, such as "buyed" for "bought"), and suggests that innate knowledge covers not merely universal features but properties that differ from language to language, so that Japanese and English children inherit knowledge of Japanese and of English phonology respectively. Smith, on the other hand, seems a reliable guide to Chomsky's technical thinking. His second chapter strikes me as an unusually lucid brief exposition of 40 years of Chomsky's changing grammatical theories.

Smith does sometimes seem to distort others' work, where this conflicts with Chomsky. For instance, Chomsky's linguistic nativism implies that there should be occasional individuals who have congenital mental difficulties that affect just language structure and leave other cognitive faculties wholly intact, and Smith cites two writers who have discussed this phenomenon of "specific language impairment". One of them, Laurence Leonard, is known for arguing at length that the concept in this strong sense is a myth, but no reader would guess that from Smith's citation.

McGilvray includes more material than Smith on Chomsky's politics; but McGilvray is sometimes naive about political theory. He repeatedly discusses a description by James Buchanan of the conflicts of interest between individuals that political life must somehow resolve, as if it were Buchanan's description of an ideal society - McGilvray takes a statement of the political problem for a statement of a solution. Respected thinkers who have envisaged language differently from Chomsky, such as Ludwig Wittgenstein, Willard Quine, Richard Montague, are scattered like ninepins by both writers: they differed from Chomsky in one respect or another, so they were wrong, end of story.

Each of these authors seems to aim at an almost slavishly exact reflection of Chomsky's work. Frequently they use identical, oddly unpersuasive illustrations. Both of them exemplify an obscure point about word meaning by claiming that, if a town were to filter its water supply through teabags, so that the liquid that came out of people's taps was chemically indistinguishable from tea, it would nevertheless be called "water" rather than "tea". I should have thought it was more likely to be called tea; but evidently Chomsky somewhere states that it would be called water, so for Smith and McGilvray the matter is settled. More significantly, both authors treat the concept of "poverty of the stimulus" as central to the Chomskyan argument for innate knowledge of language: allegedly, children succeed in mastering aspects of their mother tongue for which there is no evidence in the speech they hear. According to Smith, "half a century of research ... has uncovered innumerable such examples". But the truth is that Chomsky's writings through the years have used the same example over and over again, and Smith and McGilvray use it too: namely, the ability of English speakers to form questions (such as: "Is the man who is tall in the room?") that correspond to statements that contain a subordinate clause preceding the main verb. Chomsky has repeatedly said that this ability must be innate, because he feels sure that examples are vanishingly rare in real life. But linguists who have taken the trouble to look at the data have found that such questions are actually quite common. So much for poverty of the stimulus.

McGilvray and Smith are enlightening, perhaps inadvertently, on the thinness of Chomsky's political analysis. Chomsky denounces many actions and policies of his own and other western governments, sometimes justifiably, but (as McGilvray and Smith admit) he never offers any details of how the alternative political system he advocates would work in practice. Surely, every second saloon bar contains a loudmouth who can do as much as Chomsky in that respect?

McGilvray suggests that it may be a good thing for Chomsky to keep his political ideal vague, so that it can be adapted to the circumstances of different times and places. If Chomsky espoused a standard, middleof-the-road brand of politics, McGilvray's line might perhaps be tenable, but Chomsky's ideal - anarcho-syndicalism - is a political system that has never been realised and that seems impossible or meaningless to many people of goodwill. What use is there in putting the system forward if we are given nothing to allay those doubts?

Smith even praises Chomsky for his role in the Faurisson scandal of 1980-81, when he contributed a preface to a book by an antiSemitic French academic who argued that the Holocaust never happened. Smith describes some of Chomsky's critics as "hysterical" for failing to recognise the importance of freedom of expression for all views, however objectionable. But what troubled Chomsky's critics on that occasion was in part the difficulty of reconciling Chomsky's claims that he merely defended this abstract principle, with the fact that the preface as printed also contained material defending Faurisson himself rather than his right to free speech.

The Faurisson affair, though, is an exceptional case where Smith and McGilvray admit that not everyone shares their assessment of Chomsky. In the linguistic domain, neither writer offers the least hint of the extent to which well-informed commentators disagree with Chomsky, though by now this disagreement is great. Hans Aarsleff has described Chomsky as travestying the ideas of Descartes and others whom he claims as predecessors. William Labov showed how Chomsky deals with conflicting grammatical intuitions by treating his own intuitions as authoritative data to be explained by science, while others' intuitions are treated as mere fallible hypotheses. Gerald Gazdar demonstrated that a more sophisticated concept of phrase structure makes Chomsky's resort to transformational grammar unnecessary. Geoffrey Pullum accuses Chomsky of repeatedly falsifying the recent history of linguistics in order to win his arguments. Daniel Dennett describes Chomsky as simply "a bully".

Possibly Chomsky might be defended against some of these criticisms. But Smith and McGilvray steadfastly ignore them all. What they have written might be better described as high-brow fanzines, rather than exercises in critical scholarship.

Geoffrey Sampson is professor of computer science and artificial intelligence, University of Sussex.