English like she is spoke: an abuser's guide

The new Oxford English Grammar should be the last word on split infinitives and correct usage. But it ducks all the issues. Super-pedant William Hartston confronts its libertarian editor, Sidney Greenbaum

Take a million words of English, comprising 500 texts of around 2,000 words each, all uttered - in speech or writing - by educated native speakers. Add another three million or so words from the Wall Street Journal, then churn them all through a computer programmed to analyse linguistic structure and to tag each word with one of 170 different indicators of its possible grammatical role. Next check by hand every one of the computer's attributions. You will then be in a position to start thinking about writing a descriptive grammar of the English language and, if you are as thorough and meticulous as Professor Sidney Greenbaum, you will have finished some 650 pages later.

"It was originally commissioned as a 350-page book," Greenbaum explained, "and I would have been happy to cut it down to that size, but OUP decided to keep it at the length it was written." The final product, the Oxford English Grammar (pounds 25), is undoubtedly a landmark in linguistic publishing and an imposing monument to scholarship. Never before has our language been given such a thorough overhaul; never has there been, in a single volume, so complete an account of the structure of English as it is used by all of us. Yet despite - or perhaps because of - its thoroughness, the Oxford English Grammar is a curiously unsatisfying and even unsettling work. For this is very much an account of English like she is spoke rather than English as she should be spoken. It is, as Greenbaum explained to me: "a descriptive book on the contemporary English language, restricted largely to British and American English with some reference to non-standard differences and with a reference to historical antecedents." He talks a good deal like that. Long, clear sentences, slightly cumbersome, pausing, in the style of all the best academics and politicians, only at moments when it is most difficult for one to interrupt.

Oh dear. Shouldn't that last sentence have had a verb in it somewhere? If you want advice on that sort of thing, the OEG is not the place to look. It will tell you that such a sentence is an example of ellipsis: "He talks in long, clear sentences...", but, adopting the strictly non-interventionist approach of a modern descriptive grammar, this book does not set out to be a guide on whether to ellipt or not to ellipt. (Ellipt, incidentally, is a word invented by Professor Greenbaum which he hopes will find its way into dictionaries soon.)

So if this Grammar does not tell us whether prepositions may be used to end sentences with, or when to acceptably split infinitives, who (or whom) is it aimed at? "It is written for the generally educated public," the professor maintains. Ah yes, but there is educated and educated: the "educated" potential reader of the book is very clearly a different animal from the "educated" sample who provided the raw material on which it is based. Compare the following:

1) "Noun phrases may be co-ordinated syndetically (with co-ordinators) or asyndetically (without co-ordinators). In polysyndetic co-ordination, co-ordinators are inserted between each pair of noun phrases." 2) "Use it as a mixer for my uhm lemonade and lime lemon lime and someone introduced it to me the other day."

The first is from a summary of a section on noun-phrases; the second is cited as an example of how a yes-no question (in this case: "Do you drink quite a lot of it?") may elicit more than a one-word answer.

The syndetically co-ordinated educated classes of Example 1, for whom this book is intended, are clearly a different bunch from the vodka-and- lime educated classes of Example 2. But (there we go again, starting a sentence with a conjunction) would it have made a difference if the original sample of English-speakers had been more stringently selected? Greenbaum explains that the sample was restricted to educated people, "those who had completed their secondary school education, but we made one or two exceptions. For example, we included examples from John Major and the Queen." A more restrictive sample, he believes, would not have made very much difference. It is difficult to agree, particularly when the approach of the book is so determinedly non-condemnatory of incorrect English. When I accused Greenbaum of being pedantically non-prescriptive, he denied the charge, maintaining that the book clearly differentiates between correct and incorrect. Yet he only very rarely goes beyond "non-standard" or "less formal" as terms of reproach.

"I'd go to the Palmer one if I was you," is cited as an example of how the past indicative was has been replacing the subjunctive were, "particularly in less formal use." A sentence beginning: "Schmidt (1982) explained this phenomena ..." is described as a "controversial" use of treating a plural noun ending in -a as singular. "We don't want none neither", is "non-standard". Worst of all, "They're making a birthday party for their youngest, which I'm invited to it," fails to attract even mild disapproval. It is simply an example of how: "resumptive pronouns are sometimes introduced, echoing the relative pronoun."

That's not controversial, it's not non-standard, it's not less formal, it's wrong, wrong, wrong, and it's a dereliction of duty by grammarians when they refuse to say so. And let me add another thing before Jean Aitchison goes on the radio again to extol the richness and breadth of an English language that tolerates a great variety of non-standard forms and to denigrate those she calls "pedants" who criticise anyone who dares to, if I may be permitted to bring up the topic again, split an infinitive. We pedants have never objected to split infinitives. Splitting them persistently is bad style, not bad English. It's pseudo-pedants who think they should never be split. It's the between-you-and-I merchants, and the singular- phenomena brokers at whom we pedants gnash our teeth, and if grammarians are going to sanction them as non-standard, it just shows that language is far too important to be left to grammarians.

The OEG, however, is not a book about correct English (i.e. English that abides by the rules of usage). It is not even a book about good English (English that fulfils its purpose as a medium of comminication). As Greenbaum explains, correct English need not be good, nor good English correct. The intention of this book, in which it succeeds outstandingly, is to give a picture of English, warts and all. But why is the line drawn to exclude the non-standard greengrocer's apostrophe ("Carrot's 60p lb") but include the occasionally informal "innit" version of "isn't it"? And who really needs such a Martian's-eye view of the language, mixing written and spoken, formal and informal, standard and non-standard, educated and edgerka'ed?

The OEG will be an indispensable reference to anyone working in the fields of computer translation or machine transcription of the spoken language. It will be invaluable to linguistic theorists and teachers of high-level colloquial English as a foreign language. It may even, eventually, be a starting point for a series of grammars of all forms of English, both standard and non-standard. But lumping them together gives the impression that the rules of language are far more flexible than they actually are. Each non-standard form is a language game with its own set of rules. Nobody has the choice between saying "He hasn't eaten anything," and "'E ain't eaten nuffink." It's one or the other. Each is correct within its own domain. Whatever the libertarian grammarians may tell us, we should all be fighting to preserve the unique status of standard English, without which there would be no universally acceptable form of the language for the precise and unambiguous expression of complex ideas.

The main trouble with the OEG, however, is that, with the word "Grammar" in the title, it will be acquired automatically by libraries and institutions, placed on the shelves with the grammar books and consulted in vain by people looking for clear rules of good English. Attracted by the words "ground-breaking new authority" and "ideal for non-specialists" which appear on the cover, they are liable to find that the book is not quite what they expect.

Finally, lest I be (subjunctive) accused of failing (gerund) my pedantry duties, I ought to point out at least one grammatical error in the book. Here's a long sentence from the middle of page 27: "Chomsky's conception of competence in most of his work is restricted to the knowledge that enables a native speaker to produce an indefinitely large number of sentences, some of which are novel in the sense that they do not replicate sentences that the speaker (and perhaps anybody else) has produced before." Look at the bit in the brackets. That "and" should be an "or". That's not non- standard, that's a mistake. Innit?