Language is Power by John Honey Faber & Faber, pounds 8.99; Colin MacCabe attacks the sloppy Sixties liberalism that deprived the poorest children of a command of standard English, and calls for a modern way to teach grammar in our schools
John Honey's book marks another contribution to that dialogue of the deaf which passes for a debate about how English should be taught in schools. Given the importance of the topic, it is unfortunate that almost all discussion of it is couched in the most reductive of ideological terms. Honey sketches a world in which linguistics, ruined by the cardinal error of assuming that all languages are equal, has encouraged teachers to abandon teaching a national standard in favour of encouraging students to develop their own dialects. As these dialects have none of the features of a standard language - extensive vocabulary, explicit codified rules, and so on - any pupil limited to one is denied the possibility of full entry into the public intellectual world defined by standard English.

Much of Honey's book sees an enthusiastic slaughtering of sacred cows. He is particularly good on the "linguistic equality" thesis: the notion that any language has the same potential range of expression as any other. This formulation ignores the extent to which a language cannot be considered independently of its history and institutions. It may be true that a Glaswegian dialect could take on all the features of standard English - but to do that, it would need centuries of codification in relation to an educational system.

For Honey's book to be the kind of clarion call that he wishes, it would need a great deal more theoretical and empirical sophistication. Let's admit that for many uses - scientific, philosophical, legal - standard English has incomparable advantages. But let's also admit that, for some emotional or comic purposes, Glaswegian is richer.

Curiously enough, Honey is aware of that argument but never properly poses the crucial question which follows. How do we compare languages? Nor does he ask the even more important pedagogical question, which is how one teaches a child so that she both acquires the standard but also develops her own form of the language.

It would be a great deal to ask Honey to come up either with the basis for a new comparative study of language, or even a consideration of what range of variation is possible within a standard. However, his failure to address basic pedagogy makes him a fully paid-up participant in the current debate, where political positions pose as beliefs about language and ignore the reality of classrooms and curricula.

The paradox is that we do have a clear ideal of how standard English should be taught (both the Kingman and Cox reports at the end of the 1980s spelt out the local forms), but very little idea of what that ideal means in practice. If the two reports had been taken seriously, then every undergraduate who planned to teach English would have to learn both the history and structure of standard English in courses designed to reflect the position that Cox and Kingman adopted . I am not aware of any such courses.

The reasons for this lack are many and complex. Simplistically, however, one could suggest that the political right were so disappointed that the teaching of "correct grammar" had not been recommended that they lost interest. On the left, meanwhile, any teaching about language remained tarred with an historic brush: the imposition of a class-based norm.

Honey spends much of this book attacking straw men and conducting a debate which might have been more appropriate 10 or even 20 years ago. While his demonstrations of historical ignorance among the class warriors of sociolinguistics are often both pointed and funny, his own belief that a simple category of "the educated" can sweep away all questions of class and oppression in the history of language renders him as ignorant as his opponents.

Above all, Honey ignores the question of what teachers should be taught. I wonder how many teachers would now doubt the usefulness of teaching grammar; but how many teachers with English degrees have ever been taught it themselves? Honey never addresses what the content of the grammar teaching he persistently urges might be. One of the advantages of traditional grammar teaching, which lasted some four centuries and all but disappeared at the beginning of the 1960s, is that it held its latinate categories to be fixed in a way we cannot.

How to teach grammar without the traditional notions of "correctness" is perhaps one of the burning questions of our time - but it is too hot for academics to handle. If Tony Blair is serious about wanting Britain to be the best educated of countries, it now needs to be addressed at the highest level of government.