The Oxford English Grammar 
by Sidney Greenbaum.
Oxford, 652 pp., £25, February 1996, 0 19 861250 8
Show More
Show More

Grammar could once seem synonymous with all learning, including magic and astrology; hence French grimoire (book of spells) and English glamour. But as early as the 14th century its OED sense 1.a had emerged:

That department of the study of a language which deals with its inflexional forms, or other means of indicating the relations of words in the sentence, and with the rules for deploying these in accordance with established usage: usually also the department which deals with the phonetic system of the language and the principles of its representation in writing. Often preceded by an adj. designating the language referred to, as in Latin, English, French grammar.

The OED entry proceeds: ‘In early Eng. use grammar meant only Latin grammar, as Latin was the only language that was taught grammatically. In the 16th c. there are some traces of a perception that the word might have an extended application to other languages ... but it was not before the 17th c. that it became so completely a generic term that there was any need to speak explicitly of “Latin grammar”.’ Finally: ‘As above defined, grammar is a body of statements of fact – a “science”; but a large portion of it may be viewed as consisting of rules for practice, and so forming an “art”.’

Rules, rules. And Latin. As Latin slowly passed out of all but liturgical and classroom existence, it became the model of the rule-bound language. Each noun followed one of five declension systems that carried it through nominative, vocative, accusative, genitive, dative, ablative. Each verb was likewise faithful to its conjugation, active and passive voices and all those tenses, not omitting the subtle distinction between the perfect and the pluperfect. When I studied Latin in Canada decades ago, we had two full years’ drill in grammar before we so much as looked at anything any real Roman had written. The sentences we confronted had been contrived to illustrate ‘the rules’ as plainly as possible. Yes, that was grammar: a clean, tight, rule-bound system in an ideal unspoken void.

English grammar by contrast was dismissed as a disappointment. Inflections were illustrated by pronouns – he, she, it; his, hers, its – but by little else. Singulars were distinguished from plurals, but plurals were annoyingly similar to possessives – watch where you put that apostrophe – and as for verbs, there were the ones that change altogether: I go, I went, I have often gone, and the ones that merely wobble, I shiver, I shivered, I have often shivered. For grammarians, it’s not much of a language, clearly. The trouble with English was often hinted at: it’s a language millions of people just keep speaking, whereas the purity of Latin derives from its deadness. I’ve even heard analogies drawn with the way medical students get a fundamental education from the study of corpses.

All this by way of prologue to what’s unique about the late Sidney Greenbaum’s Oxford English Grammar. It does not, to begin with, treat the grammar of English as a withered version of the grammar of Latin. Nor does it set forth some rules and then contrive neat illustrations. No, most of its numerous citations ‘are taken from two sources: ICE-GB (the British million-word component of the International Corpus of English, drawing on language used in the period 1990-93) and the Wall Street Journal (about three million words from this American newspaper for 1989, provided in a CD-ROM by the Association for Computational Linguistics Data Collection Initiative)’. We learn on another page, moreover, that the ICE-GB material isn’t confined to the written and printed; citations marked ‘S’ pertain to spoken texts, with subcategories for ‘Dialogue, Private Conversation, Public Dialogues, Monologue, Unscripted Monologues, Scripted Monologues’. What’s spoken not being subject to revision, it testifies better than writing does to people’s nigh-innate sense of the language they daily navigate through. (And very early in the introductory matter we find a tidy six-page summary of Noam Chomsky’s ‘internalised grammar of native speakers’: what he says we all learnt in childhood before we’d studied grammar. Greenbaum doesn’t necessarily subscribe to Chomskyism, but he thinks it merits our attention. It’s bracing to be reminded that people spoke grammatically before there were classrooms.)

That is what the jacket means by ‘The ground-breaking new authority, based on the evidence of real English.’ Untidy though the utterances on display may be, their underlying grammar persists like a leper’s flawless skeleton. ‘Tea he makes tea he makes phone calls he makes gets me (O) lollipops (O)’ (where O = object). That was spoken, in an office, about five years ago. It exhibits ‘an indirect object (O) followed by a direct object – also (O)’, exactly as in the neat but contrived example, ‘I am sending you (O) an official letter of complaint (O).’ Next we’re told that the indirect object ‘can generally be paraphrased by a phrase introduced by to or for, but that phrase follows the direct object.’ Example: ‘We tell everything (O) to each other.’ And ‘sometimes the direct object is absent and the indirect object alone is the complement of the verb.’ Example: ‘And we shall I promise you (O) bring our own forces back home just as soon as it is safe to do so.’ And: ‘When there is one object, the basic structure is SVO; when there are two objects, it is SVOO, the first object being indirect and the second direct.’

These remarks on word-order help to emphasise the fact that in English it’s an aspect of grammar, whereas in Latin it’s chiefly an aspect of style. When the distinction between ‘Dog bites man’ and ‘Man bites dog’ inheres in the way the nouns are spelt, word-order is free to concentrate on rhythm and emphasis. The closest English will bring us to Sunt lacrimae rerum may be ‘In this fallen world we weep’, and in adducing Christian theology that greatly falsifies Virgil.

Next, ‘A verb taking one object is monotransitive, a verb taking two objects is ditransitive.’ ‘Monotransitive’, ‘ditransitive’: consider yourself present at the birth; neither of those words is to be found in the OED. That is to say that no one before Greenbaum seems to have perceived a need for them. ‘Ground-breaking’ the book is. Unlike previous grammars, it’s no dull rearrangement of what we can remember being bored by, the way grammar bored me when I was being pumped with Latin.

To continue: ‘When we change an active sentence into a passive sentence, the active indirect object can become the passive subject’: ‘I am sending you (O) an official letter of complaint (O) vs. ‘You (S) are being sent an official letter of complaint (O).’ Or ‘the direct object can become the passive subject’: ‘An official letter of complaint (S) is being sent you (O).’

So runs, with several omissions, a summary of what’s on two pages, section 3.17, ‘Indirect Object’. That’s section 17 of Chapter 3, ‘An Out-line of Grammar’, just 40 pages long. It’s preceded by two chapters called ‘The English language’ and ‘The Scope and Nature of Grammar’, and it’s followed by nine more: 4: ‘Word Classes’; 5: ‘The Grammar of Phrases’; 6: ‘Sentences and Clauses’; 7: ‘Text’; 8: ‘Words and their Meanings’; 9: ‘The Formation of Words’; 10: ‘Sounds and Tunes’; 11: ‘Punctuation’; 12: ‘Spelling’. We’re evidently a long way past what ‘Grammar’ normally connotes. A Preface tells us that the book is ‘primarily addressed to native speakers of English and others who use English as their first language ... It serves as a reference work and can be used as a textbook.’ Also, the author confides, ‘I have drawn on my many years of experience in teaching, research and writing. I have taught English language in a range of institutions and to different age-groups: at primary schools, at a secondary (grammar) school, at a college of further education and at universities. My university teaching has encompassed a British university, universities in the United States and a university in a country where English is a foreign language. I have been in English language research for over thirty years.’

The experience does show. There’s a seamless effortlessness about, oh, say, Section 7.3, ‘Different Kinds of Speech’.

In casual conversations the speakers are normally visible to each other. New technologies during this century have introduced various kinds of distanced communication ... The most common kind involves people using telephones [who] often feel impelled to accompany their speech with appropriate facial expressions and gestures ... It is usual for listeners to signal that they are attending to what is being said and can understand it by interjecting yes, mm, and the like. Longer silences are tolerated on television broadcasts than on radio broadcasts.

That takes account of something often missed, that the ‘words’, spoken and transcribable, are only a portion of intelligible discourse. And it returns us to the fact that we’ve normally thought of ‘grammar’ as a system created to govern the written language, where improper grammar leaps out. The telephone offers an interesting special case, speech minus visual presence; it abounds, as Greenbaum suggests, in unwriteable cues.

Or sample Section 7.7, about ‘Connections across Sentences’. The unity of a text, we’re told, depends on cohesion and coherence. The first refers to devices for linking parts of a text. The second refers to ‘continuity of meaning that enables others to make sense of a text’. It’s added that readers tend to assume coherence ‘and make the effort to interpret [texts] as having unity’. An instance of cohesion: ‘John Maynard Keynes, the century’s most influential economist, once said that in his utopia members of his profession would be like dentists – useful but humble people. Utopia may be arriving with the administration of President-elect Bill Clinton.’ What holds that together is simply ‘the repetition of the word utopia’. And coherence? It simply follows from cohesion.

We are a long way from what’s been understood as ‘Grammar’ and with Chapter 8, ‘Words and their Meanings’, we move still farther. Thus, ‘Onomatopoeic words tend to be language-specific: cocks crow cock-a-doodle-doo in English but kikeriki in German ... A bell sounds ding-dong in English but talan-talan in Spanish.’ (We may add that dogs say wow-wow in English, riau-riau in French, wang-wang in Chinese. They utter seemingly any monosyllable at all, but imprint the mark of dogdom by always uttering it twice.)

Language spoken and overheard can yield complex wonders. This is from the BBC’s Gardeners’ Question Time: ‘When you tie a standard rose and this applies to any standard rose whether you do it yourself or whether you buy it, you really need two ties on it.’ Here ‘The and-clause is parenthetic, expressing an elaboration of the point made in the initial subordinate when-clause. The and-clause itself contains two subordinate whether-clauses linked by or.’ That’s from a section on Clause Relationships, where surprises abound. ‘Whether he speaks or not remains to be seen’: there you have what is structurally a subordinate clause, but it’s serving as subject of a verb. Though perfectly idiomatic, that’s something you’d never derive from the usual outlines of the grammar of English. It turned up in an ITV programme called Athletics, and the list of examples classifies it among ‘Spontaneous Commentaries’.

Or here’s the opening of Chapter 6, ‘The Sentence’: ‘The traditional definition of a sentence states that a sentence expresses a complete thought. The trouble with this notional definition is that it requires us to know what a complete thought is. Does God or our home express a complete thought? Is there just one complete thought in [the following example]?’ The example runs: ‘Some 4000 people (most of whom had heard about, but not actually read the book) wrote to Dr Robinson, telling him of their own faith, beliefs, convictions, feelings, or special knowledge concerning matters religious.’ It’s next pointed out that the example can easily be converted into three sequential sentences, each perfectly complete.

So what’s a sentence? We fall back on the principle that ‘The measure of grammatical completeness is the clause,’ and define the canonical sentence as consisting of ‘one or more grammatically complete clauses. That is to say, each clause contains the constituents that must be present according to the general rules for constructing clauses – subject, verb and complements of the verb – except that the understood subject you is generally omitted in imperative sentences.’ Fussy though that may sound, it gets rid of the ‘complete thought’ vagueness.

There are also ‘elliptical sentences’, incomplete but ‘perfectly normal and acceptable’, as in the dialogue ‘You told me at the time.’ ‘Did I?’ And there’s ‘the unfinished sentence’, which is ‘very common in speech’, because speakers change their minds, or correct themselves, or get nervous, or just plain lose the thread of what they were saying. Such sentences ‘are not rule-governed, since speakers may fail to finish their sentences at any point. Grammars, therefore, cannot account for them.’ It’s refreshing to find a grammar saying that.

I trust these random examples convey some idea of how greatly this book departs from the schoolmarm connotations of ‘Grammar’. It’s a thorough and exhilarating effort to define what makes for unjarring English and what doesn’t, once we extend our field from the Cartesian obvious to the sequences people actually write and say and unresistingly understand. Greenbaum doesn’t offer instances you have to puzzle out. The very first example I quoted – the ‘Tea he makes’ utterance – is, I think, the sole exception in this review, and I’ll admit to having led off with it for its shock value. Had you heard it spoken, it’d not have given you a moment’s pause.

The ‘rules’ of grammar constitute a description of what native speakers understand and accept without question. It’s true that ‘native speakers’ is a loose designation; an Alabama black idiom is spoken by natives, and conveys nuances, even as does the idiom I acquired as a native of Canada and now, as a resident of Georgia, an offering to British editors. But Greenbaum’s welcoming arms aren’t spread indiscriminately wide. The book’s very second sentence calls it ‘a comprehensive account of present-day English that is chiefly focused on the standard varieties of American and British English, but it also refers frequently to non-standard varieties and it draws on the history of the language to illuminate and explain features of English of today.’

Read that last sentence again and note how it twice uses the pronoun ‘it’. A book addressed not only to ‘native speakers of English’ but also to ‘others who use English as their first language’ must bear in mind how the latter group can perhaps get lost amid multiple pronoun references. That’s less part of English Grammar than it’s part of the difficulty of deploying English grammar, unambiguously, for the maximum sample of a likely readership.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.


Vol. 18 No. 16 · 22 August 1996

In his review of The Oxford English Grammar (LRB, 18 July) Hugh Kenner betrays an ignorance of at least a century of developments in grammatical description, including Sidney Greenbaum’s own landmark collaboration, A Comprehensive Grammar of the English Language (1985). As important and welcome as Greenbaum’s new grammar is, it is not, as Kenner suggests, the first to depart from what he ingenuously terms ‘the schoolmarm connotations of “grammar" ’. These are, I infer, a preoccupation with normative rules (as opposed to description), ‘accidence’ (as opposed to syntax) and written, as opposed to spoken, forms. Descriptive and syntactic grammars of English have been around for at least as long as the twenty years I have been teaching English (the language). Admittedly, their concern has been mainly with written English. Nevertheless, as long ago as 1761, Priestly premised his rudiments of English Grammar on the principle that ‘the custom of speaking is the original and only just standard of any language.’

Nor is Greenbaum’s grammar ‘unique’ by virtue of the fact that it draws on a corpus of authentic texts for its citations. In Jesperson’s A Modern English Grammar (1909-49) all the citations are taken from published books; The Collins COBUILD English Grammar (1990) draws on an enormous computerised database of both spoken and written English housed at the University of Birmingham. Nor, incidentally, is Greenbaum the first to use the terms ‘monotransitive’ and ‘ditransitive’, as Kenner suggests.

Finally, his contention that coherence ‘simply follows from cohesion’ is demonstrably false. The following text, while notionally cohesive, is clearly incoherent: ‘Hugh Kenner lives in Georgia. And I’ve got Georgia on my mind. Mind the gap. Gap has got three letters, therefore.’ Kenner’s review reminds us that, whereas specialists in English (the language) have written engagingly and with great scholarship about English (the literature) – witness Michael Halliday and Henry Widdowson – the reverse is sadly not the case.

Scott Thornbury

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN

Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences