I just read a comment online about a comment about an attack on Strunk and White in which the commenter declared, “[Grammatically] Prescriptive rules are, by their nature, arbitrary.” By their very nature!
Descriptivism in, prescriptivism out. We are to believe that it is impossible to prescribe any grammar in a way that is both accurate and healthy; what anybody says, goes (what one can accurately describe anybody as saying, goes). (Where does it go? Who knows. We have Established Usage dictating the terms here. Ask who; Who knows.)
So, don’t prescribe any prescriptions! That’s a major rule, an exigent prescription.
Of course, there isn’t any necessary chasm between “describing” the grammar of how persons habitually and competently express themselves in a particular language and “prescribing” what the grammatical usage ought to be.
Prescribers of the don’t-prescribe rule may know only English, let’s say. Children growing up in Englishland imbibe the English and know how to utter it before they ever hear of such terms as “noun” or “verb” or “dangling modifier” or anything of the sort. Yet if the child one day writes, “Running on four superbly coordinated limbs, the man had no chance of keeping up with the gazelle,” we may wonder why the man thought he could increase his chances by dropping to all fours. Grammatical rules should spare the reader from having to slow down to engage in special detective work unrelated to the subject of discussion. Some writers are so confusing in their grammatical usage that no amount of description and divination can save us, or them. One must, reluctantly, prescribe. Copy editors do this. They are paid, by people who cater to readers hoping to smoothly decode grammatical usages.
Every language has a grammar, and to teach that language one must describe and prescribe that grammar. One cannot supply the rules for Cantonese when attempting to communicate the workings of English.
Suppose an alleged speaker of English starts throwing around words at random, in any old order, trusting to the inflections of the words alone to inform the listener of the relevant part of speech and the relationship of each word to other words in the sentence. Somebody would then have to kindly explain that English is not Latin or any other super-inflected language, that meaning in English is expressed far more fundamentally by the distribution of words than is true of Latin, which has a billion inflections for each word.
It is true that the rules for speaking a particular language are embodied in the habitual expression of that language, and a new entrant to the linguistic communication game can soak up grammatical rules and principles implicitly before he can begin to learn them explicitly.
To be sure, some grammatical prescriptions are arbitrary. Outmoded usage may be unnecessarily clung to, or a mandate that was never reasonable to begin with may gain currency despite prolific counterexamples in past and present competent standard usage. But this is a different claim from the notion that all grammatical prescriptions are arbitrary. If that were the case, grammar would be pointless, and linguistic communication impossible. I think you know what I’m saying.