Dangerous. Handle with extreme care

This book will make you furious and will unleash "the tygers of wrath"1 on its authors. First, any grammar of this kind will "vex to nightmare"2 people who care deeply about the English language, or who think grammar is about helping people to write better, or who want to know whether to say "different from" or "different than". This is a descriptive grammar, not a guide to "correct" usage or good style: it describes English as it is, not the way some people think it should be. The authors make some anodyne remarks to this effect in their introduction, just as every other descriptive grammar does. Indeed, they are unusually generous, conceding that "it might beI very sensible for a less expert writer to follow the recommendations of an acknowledged master of the writer's craft". This did not protect them from the "fiery vehemence"3 of The Guardian reviewer who accused the authors of "parochialism", and described their concentration on contemporary English as a "tyranny", which mirrors the traditional insistence that old is best.

Another group of people who are going to be "most seriously displeased"4 are students and teachers of grammar, not to mention university librarians. They will simply have to obtain this book because it contains the most carefully argued, consistent, comprehensive and elegantly presented account of English grammar ever written.

Whether they are interested in English tenses, relative clauses, prepositional phrases, word structure, the position of adjectives or a hundred other matters, this book will be indispensable. They will have to deal with the book's huge price-tag and unwieldy size; the fact that the typography, paper and binding are superb will do little to mollify their feelings. Perhaps a smaller and more affordable student's grammar will appear at some point. I hope so, but in the meantime anyone who thinks that this enormous volume will just gather dust on library shelves is making a big mistake.

Something else that will make these people - and the general reader - "burn and rave at close of day"5 is the fact that the book overturns some widely held conventional wisdom about English grammar. The word before in "I came in before you", "I came in before the rain started" and "I've never seen you before" is treated as a preposition in each case. The set of "subordinating conjunctions" is drastically reduced to whether and that .

The authors reject "noun clause" and "adverbial clause", and argue that there is no reason to distinguish between gerunds and present participles.

They keep the traditional "subjunctive" but use it for a type of clause, not a verb inflection. This will trouble readers without a strong background in linguistics, as will the use of technical terms that are commonplace in the field ("lexeme", "wh-cleft", "polarity-sensitive", "endocentric compound", "anaphoric chain") and the coining of new terms ("supplementation", "hollow clause") where no old one will do. In short, this is a specialised work of scholarship that makes serious demands on its readers.

And for many specialists, their "wrath will wax hot"6 against the authors.

The field of English grammar, as taught in schools, is dominated in this country by the work of Randolph Quirk and his associates, particularly the Comprehensive Grammar of the English Language (1985). This book is a flagrant challenge to the Quirk tradition. Quirk divides clauses into "elements" such as "subject" and "adverbial"; Rodney Huddleston and Geoffrey Pullum divide them into a noun phrase and a verb phrase.

Where Quirk uses "complement" in an idiosyncratic way, The Cambridge Grammar defines the term clearly and uses it consistently. Quirk distinguishes lower-case "verb" as a part of speech from "VERB" as a clause element, a confusing step that this book explicitly and rightly rejects.

Quirk has an inconsistent approach to phrase constituents, which students inevitably find puzzling. This book comes out of linguistics departments, not departments of English, and it seems to be all the better for it. As a wise man once wrote, "few things are harder to put up with than the annoyance of a good example"7.

There is another large community of academics who will find this book outrageous, perhaps even "irksome music to their heart"8, because most of the author's examples are not taken from a corpus (a large collection of authentic texts). Corpus enthusiasts will not be impressed. When grammarians invent sentences, we are told, they run the risk of ignoring the facts of actual usage, and of giving undue weight to marginal constructions that are hardly ever used. Furthermore, a corpus can often reveal surprising facts about which words regularly occur near other words, and about differences between formal and informal varieties of the language.

A large corpus-based grammar, The Longman Grammar of Spoken and Written English , appeared recently, and gave vast amounts of information about frequency patterns and variations. Huddleston and Pullum make some use of corpus data, and acknowledge the insights from recent corpus-based grammars, but their analytical framework is grounded in basic or "canonical" structures, rather than the messiness of authentic data. This has already generated some heated exchanges on the Linguist email list (go to linguistlist.org and search for Pullum). Personally, I think that the way to judge a methodology is by the quality of the research that it produces, and it looks to me as if The Cambridge Grammar scores highly. Others will disagree strongly.

That is not to say that theoretical linguists will be happy either. Many of them will find this book "the just object of their ire"9 because it is not tied to any particular theory of grammar. Various of the analyses are inspired by transformational-generative work, but the book explicitly rejects the notion of deriving sentences from abstract underlying structures. That is legitimate for a descriptive grammar, but the thing that will stick in many a generative craw is the way that Huddleston and Pullum think that they can argue for particular analyses outside of a specific framework. For example, they maintain at some length that auxiliary verbs such as must , and have of the perfect tense, are not really auxiliary to a main verb but are heads of verb phrases and take clause complements (grammarians with long memories will recall the authors separately taking this position in the late 1970s). Theoreticians will be tempted to ignore the book - but will find it hard in all conscience to disregard work of this quality.

So far I have spoken only of other people's venom and contumely, but sadly I am not immune. Since I am, of course, wholly free of jealousy towards my colleagues who produce monumental works of scholarship, my wrath is purely intellectual. Scattered through this grammar there are sections on a blue background, separate from the main text, where the authors address themselves to their colleagues and give reasons (sometimes over several pages) for the analysis that they have chosen. Often I am persuaded, but in some cases I beg to differ. Huddleston and Pullum expound at length on their claim that English does not have a future tense - instead, they insist that will is a modal verb with many meanings. In this they are in keeping with most modern grammars. Several studies of actual usage (yes, corpus-based, I declare an interest here) have shown, however, that more than 90 per cent of uses of will refer to future time. In the face of this evidence, I find it perverse to deny that will is essentially a marker of future time.

My advice is to handle this dangerous book with extreme care. If you see a linguist full of sound and fury any time soon, keep away. At least, don't say I didn't warn you.

1 Blake, The Marriage of Heaven and Hell .

2 Yeats, The Second Coming .

3 Scott, Lady of the Lake .

4 Austen, Lady Catherine de Bourgh in Pride and Prejudice .

5 Thomas, Do Not Go Gentle Into That Good Night .

6 Exodus xxxii, v 10.

7 Twain, Pudd'nhead Wilson's Calendar .

8 Shakespeare, Henry VI , part 2, act 2, sc. 1.

9 Milton, Paradise Lost , Bk X.

Raphael Salkie is professor of language studies, University of Brighton.