The Perennial Difficulty of Defining What ‘Descriptive’ Means in Grammar

Correct summerhouseIncorrect summerhouse

I have a distinguished colleague who unceasingly tries to persuade me to see that my attempts at descriptive grammar really have a prescriptive subtext. We were both participants in a workshop on the history and philosophy of linguistics last week, and the dispute reared its ugly head again.

Being a descriptive linguist has nothing to do with the crazy notion that “anything goes” or the absurd idea that everything anyone says has to be accepted as correct. So I acknowledge that rules of grammar have what philosophers call normative force, not in virtue of containing normative predicates like ought, but in the sense that they are intended to define a distinction (over an indefinitely large range) between two kinds of sentence structure: the kind that the described language has and the kind that it does not have.

But in my colleague’s view, though I may claim that I am trying to make a neutral statement of the right structural principles, my description harbors advisory or recommendatory force like a lurking virus. My humble scribe’s garment does not quite hide the uniform of the authority-wielding storm trooper glimpsed beneath.

The disagreement between us smoulders on without resolution, despite the responses I offer on each point. Let me illustrate a few of them. I will paraphrase my colleague’s positions as fairly as I can. (You may be tempted to doubt my sincerity in this, but you shouldn’t; I’m playing by philosophy rules here.)

Objection 1: So-called descriptive grammarians often slip into calling ungrammatical strings of words “wrong” or “bad” — self-evidently negative evaluative words. So the very terms in which grammarians talk reveal their underlying condemnatory attitude.

Response to Objection 1: No, they reveal nothing. Certainly, in casual talk, grammarians use terminology borrowed from ethics, like “wrong” or “bad” to characterize word sequences that violate the grammar of the language. We use “bad” to characterize putrid eggs, too, without implying any disapprobation or blame. It’s just a façon de parler. Mentally replace “x is bad” by “x is out of compliance with the principles of grammatical organization for the language or dialect under consideration” if you really need to.

Objection 2: Your book, The Cambridge Grammar of the English Language, claims to be a grammar of “standard English.” The very name of the language you treat lacks neutrality, implying a standard that people ought to comply with.

Response to Objection 2: Call it something else if the usual term worries you. Call it Stuck-Up White People’s English if you like (though the late Kofi Annan spoke it too). The name we use for it is a complete red herring as regards the logical character of the grammatical rules it observes.

Objection 3: Even a supposedly simple descriptive rule like “The suffix s occurs on the end of any present-tense verb with a third-person singular subject” in effect suggests a model people should follow. Foreign learners will understand you as saying that they ought to put an s on the end of these verbs.

Response to Objection 3: In describing a structural regularity I am not recommending that it be respected. If I show you the design plan for my summerhouse, and you imagine that I’m saying yours ought to have a tilted flat roof like mine, that would be an insane misinterpretation. There is no covert recommendation that you should build yours just the same, or that you should build one at all.

As it happens, given an accurate description of a language, a learner can misinterpret it as a recommendation to speak in accord with it, and will be well served in that respect (it’s the language they want to learn). This simply reminds us that accuracy of description dissolves the reasons for discord between descriptively and prescriptively motivated grammarians.

Objection 4: But millions of English speakers around the world never put an s on the end of present-tense verbs with third-person singular subjects. Your grammar is defining them as thereby getting something wrong. That too is prescriptive.

Response to Objection 4: But without an account of the normal practice in the dialect that I describe, how can you even frame this objection? How can you identify the interdialectal difference? Dialect A may have a restriction which is completely absent in dialect B; dialect C may operate with a restriction similar to dialect A but different in detail. Describing any one of them is not tantamount to recommending in its favor or deprecating the others.

But the dispute grumbles on regardless. I will never convince my colleague, it seems. My only consolation is the warm feeling in my tummy that comes from knowing I am right. He seems completely intractable. Unless of course it’s me.