In Defense of Linguistic Prescriptivism

I've blogged on the subject of linguistic prescriptivism versus descriptivism before, so I wasn’t motivated to do so again until I saw Neal’s response to my Two Things post:

The Two Things about linguistics:

1. You already know more about the grammar of your native language than could ever be taught in a class. (synchronic linguistics)
2. Language change is inevitable, and neither bad nor good. (diachronic linguistics)

Item 2 can be taken in a couple of different ways, one of which strikes me as obviously true, the other of which strikes me as (less obviously) false. The true interpretation is that change is not inherently good or bad; it is just an unavoidable fact. But the false interpretation, which seems pretty common among linguistic descriptivists, is that we cannot reasonably pass judgment on the changes; or more broadly, that we cannot judge some linguistic uses or constructions better than others. That conclusion is thoroughly unjustified. Language serves certain purposes (communication being chief among them), and some uses and constructions serve those purposes better than others.

Language is, of course, a spontaneous order, and its evolution lies beyond the control of any single person or organization. And that’s good, because I would not trust the judgment of any monopolistic entity empowered to control the language (as the French Academy attempts to do with French). Similarly, I favor free markets in large part because I distrust the centralized power of government. Moreover, decentralized orders like language and markets have a great capacity to respond quickly to the real needs of their users. But a general approval of decentralized orders does not imply approval of every specific outcome of such an order. For instance, one could consistently oppose the federal antitrust action against Microsoft while still rooting for some competitor in the OS market (like Apple or Linux) to undermine Microsoft’s dominant market position. One might even go so far as to boycott Microsoft products and encourage others to follow suit. Indeed, I know many libertarians who have done just that.

Similarly, one can recognize the evolutionary character of language while nonetheless resisting particular changes in language. For my part, I oppose linguistic constructions that I think erode logical clarity or valuable distinctions. For instance, the use of “I miss X” to mean both “I dislike no longer having or doing X” and “I regret” creates confusion (see Neal’s earlier post on this topic). If both meanings are valid, and someone says, “I miss not seeing Larry,” I can correctly interpret the statement only if I already know something about (a) how the speaker feels about Larry and (b) whether Larry is still present. Without such information, it’s possible that the speakers likes Larry and wishes he were here; but it’s also possible that the speaker dislikes Larry and is annoyed that he came back!

I don’t mean to claim that the original, unaltered language never has logical ambiguities. It does, and I welcome changes that help to eliminate the especially confusing ones. Simultaneously, I resist changes that create new ones. As I am only one user of the language, I have no illusions: the language will do what it will do, and I’m mostly helpless to resist it. Still, individuals do sometimes make a difference. Teachers and journalists, in particular, are in a position to have a disproportionate influence on the evolution of the language. I contend that teachers and journalists have a special obligation to think about the utility and disutility of changes in the language, rather than throwing up their hands and saying change is neither good nor bad.

One more caveat on a post that’s already too long: I also support the rejection of pointless rules of grammar that (a) do not serve any useful purpose such as avoiding logical ambiguity and (b) were made up by grammarians enamored of Greek and Latin. Neal tells me, for instance, that the prohibition on split infinitives (“to boldly go”) was not an indigenous part of the English language, but the creation of grammarians who observed that it was literally impossible to split infinitives in Latin (where an infinitive is a single word). Linguistic prescriptivism can certainly go too far, as Eugene has emphasized repeatedly on his blog. But there is nothing wrong with a modest dose of prescriptivism administered by thoughtful users of the language.