I have often excoriated the useless traditional definitions of the “parts of speech” presupposed by all usage manuals and dictionaries (see my “Being a Noun,” “Being a Verb,” “Being an Adjective,” etc.). I seldom praise popular grammar books; too many of them are unredeemed horse dung from cover to cover. But I am not implacable: I did have a kind word or two here for Edward D. Johnson’s treatment of the passive. Today I want to make a positive remark about a grammar book by a fine philosopher.
Professor Sir Michael Dummett (1925–2011) was a philosopher and logician of great distinction, an expert on Gottlob Frege’s philosophy of mathematics, logic, and language. In the 1980s he served for a while as an examiner for undergraduate final examinations at Oxford, and was shocked to find essays replete with “erroneous grammar, style, and spelling.” He resolved to write a book to light students in their grammatical darkness. And in 1993, having retired from the Wykeham Professorship of Logic, he published Grammar & Style for Examination Candidates and Others (Duckworth, London, 1993).
Some of it is dreadful, an atavistic compilation of peeves and prejudices. Dummett was famous as a campaigner for racial tolerance and equality in Britain, but was no liberal when it comes to lowlifes like feminists and Americans. He cherishes the purportedly sex-neutral use of the masculine pronoun, resenting the “feminist campaign against this use of he”; he wants men to be the noun of choice for denoting the human race, noting bitterly that it “is under severe attack from the feminist camp"; he even wants to bring back the old British sense of billion (1012 rather than 109) and corn (it used to cover more than just maize, but has become a “victim of American impact on our usage”). He hates impact as a verb, nouns modifying nouns (as in an undecidability proof), and supposedly misplaced instances of only. He spends four pages on trying to revive the prejudice against split infinitives.
In short, he often sounds like a ridiculous old reactionary railing against language change and attempting to revive irrational proscriptions that were fashionable two centuries ago.
But he was highly intelligent. (It’s not impossible for a reactionary to be smart, just a bit unusual.) And in his opening chapter on “Parts of Speech,” when he is talking about grammatical structure rather than style or idiom, he breaks with the traditional practice of basing everything on fuzzy intuitions about meaning (thing words, doing words, describing words, and all that nonsense). That deserves commendation.
This is how he does it. He begins with verbs, noting that the German translation is Zeitwort (“time word”): Verbs can vary in tense — a grammatical criterion. “Doing words” will not cut it as a definition. The claim that natural numbers exist doesn’t assert that they do something; the status of exist as a verb is supported by facts like the tense contrast between They exist now and They existed yesterday.
Next, surprisingly, he introduces the articles, a(n) and the, in order to form noun phrases and introduce the notion “subject of a sentence.” (He notes carefully that “Being a subject is not an enduring trait of any word, like being an adverb; it is a role played within a sentence by a word or phrase, which may play a different role in other sentences.”)
After a couple of pages making the notion of subject fully clear, almost entirely in syntactic terms, he then observes that “If you can recognise the subject of a sentence, you can also identify a noun-phrase that is not the subject of the sentence in which it occurs.” Next he points out that a noun phrase may consist of a single word, either a noun or a pronoun, and explains a bit more about what pronouns are.
So when he finally reaches nouns, he can say that “the noun is the principal word in the noun-phrase,” where “The principal noun in the phrase is ... the word which (supplemented by an article, possessive pronoun like my or their, or indefinite adjective like some or no) serves most to tell the hearer what is being talked about.” The dog which bit the postman contains two nouns, dog and postman, but when you use it as the subject or object in a sentence, you’re referring to a dog.
That’s the right sort of way to do it. His treatment isn’t perfect: There are many things I would correct in the chapter. But at least Dummett understands in outline the way to go about his expository task. You don’t look for some vague metaphysical notion like thinghood to define nouns; you try to explain how nouns fit syntactically into the structure of sentences. There at least, amid all the peevery, Dummett was heading in the right direction, and I salute him.
