Grammar for the 21st century (Professor Rodney Huddleston)
Report: Tim Bugler

Professor Rodney Huddleston came to our April meeting to discuss some of the ideas in The Cambridge Grammar of the English Language (CGEL), of which he is the principal author. The CGEL is a monumental work that aims to build a bridge between traditional grammar (or grammars) and linguistic grammar. In this context, ‘traditional grammar’ refers generally to the body of work produced in the late nineteenth and early twentieth centuries, which is to a large extent the grammar that used to be (and in many cases still is) taught in schools, is assumed in most modern style guides, and underlies the part-of-speech classification of words in dictionaries. ‘Linguistic grammar’ is the result of research by linguists from the early twentieth century on.

Traditional grammar came to us with a number of flaws: it is overly simplistic; it assumes Latin is the model of good language; its definitions are often notional (based on meaning) instead of formal (based on grammatical properties); and it seeks to describe categories that cover both English in particular and all languages in general. Linguistic research in the twentieth century established what now might seem common sense: languages are complicated; English is not much like Latin; the form of a word does not necessarily match its meaning; and different languages use similar constructions in different ways.

It could be said that traditional grammar, particularly as it was taught in schools for much of the twentieth century, was simplistic for the excellent reason that simple rules are easier to teach. However, its simple rules are very often inadequate to describe what actually happens in English. For instance, the statement that 'tense = time' is short and neat, but it does not reflect actuality: past tense does not always indicate past time, nor is past time always indicated by past tense. As an example, if I said 'I thought the course started next week', I would be using a verb in the past tense ('started') as part of a construction referring to the future. Conversely, in a construction such as 'I regret inviting the Smiths', the form 'inviting' is not a past tense but nevertheless refers to past time. (The CGEL does not discard traditional grammar altogether, so much of the terminology such as 'tense' and 'participle' remains the same; the definitions, however, are more precise.)

[Traditional grammars of English were largely modelled on the grammars of Latin and Greek, which is why they analyse parts of speech in ways that make no sense in English. For instance, the gerund and the present participle are different forms of a Latin verb, so traditional grammar listed gerund and present participle forms for English verbs. But in English the equivalent function is filled by a single form — 'taking'. Therefore, the CGEL identifies this as the gerund-participle. By such means, this modern analysis of English on its own terms gives us six forms for a verb where the traditional analysis would give us about thirty.]

Professor Huddleston presented many other examples of the weaknesses of traditional grammar, and we might wonder how traditional grammarians got it so wrong. Part of the answer is that traditional grammar is prescriptive, whereas linguistic grammar is descriptive; that is, traditional grammarians had (and have) their own notions of what was 'good' or 'correct grammar, and wrote books to tell other people how to speak and write 'properly', whereas linguists are concerned with describing the systems that govern how people actually speak and write.

This is perhaps the single most important aspect in which linguistic theory has informed modern grammar: it is based on observation. Traditional grammars were by and large based, on examples from literature, but a modern grammar such as the CGEL is derived from seeing how thousands of real people speak and write in everyday situations.

Informed by the linguistic approach, the CGEL is descriptive rather than prescriptive. Rather than speak in terms of 'correct' and 'incorrect', Professor Huddleston contrasted 'formal' and 'informal', 'standard' and 'non-standard'. He insisted that forms and constructions that are perfectly normal in informal speech should not be regarded as grammatically incorrect. Furthermore, the same speaker may have different rules for different situations; for instance, I might say 'She is older than I' if I was dining with the Governor, but in the office or on the street I would probably say 'She is older than me.' Grammatical correctness, then, becomes more a question of which grammatical variety is appropriate in which circumstances. In other words, as all editors know, it's about the audience.

[The contrast of standard and non-standard varieties is essentially a sociolinguistic issue, not an issue of good or bad in itself. A particular use may be socially stigmatised, but that doesn't mean that use is illogical or otherwise 'bad'. For instance, the double negative — 'He didn't say nothing' — might be looked down upon by many speakers of English, but the same construction is perfectly legitimate in many other languages, including French ('Je n'ai vu personne'). Grammar is not always logical.]

But editors also know that consistency is the basis of coherent communication; we have to agree on some standard. Fortunately, the standard grammar of English as described in the CGEL is basically the grammar we hear on the ABC and read in the newspaper. We can only hope that it is the informed and up-to-date grammar described by Professor Huddleston, rather than traditional grammar, that will be taught in the schools of this century.