I'm resisting the urge to write about Jack Hitt's article in New York Times Magazine about language extinction. Various other linguistibloggers, all probably better qualified than me, have already discussed it. It's an interesting article, though, and it's got me thinking more about language diversity and death.
Some you may have read my earlier post about interesting languages in the Ethnologue database. In it, I wrote that I think a tree structure doesn't adequately describe the relationships among the world's languages. In particular, a tree implies that once languages have split, they can't merge again. That's an over-simplification. Obvious counterexamples include pidgins and creoles, which often combine two languages that are completely unrelated (er, except that all human languages are probably related, although their common ancestor is lost deep in prehistory). But there are other examples of languages merging with or replacing other languages—languages are learned, after all, not inherited—and they're tied up with communication, travel, modernity, and civilization.
For a language to diverge into two mutually unintelligible variants (which is a pretty good working definition of a "language" as opposed to a "dialect"), you need two populations of speakers of the proto-language who, for one reason or another, stop communicating with each other for long enough that they drift apart. In the bad old days, group of people could fall out of communication with the next group of folks over for many reasons—half a group could get angry and head for the hills to form a new community, say, or a bad harvest or plague could kill off the people in the middle of a large group, leaving only small scattered groups around the margins of their former territory. This divergence has been happening for tens of thousands of years, and has resulted in the 6,000-or-so languages spoken today. However, the communication barriers that led to language diversity are rapidly disappearing.
Take writing, for example. When a group of people start writing down their language, the accumulated literature from previous generations of speakers will slow language change—and the higher the literacy rate, the greater the effect. Sometimes, this is an organized state action—one dialect, usually in an administrative center, is fixed and declared to be True and Correct, and everyone is required to learn it in schools. Even without this kind of coercive dialect leveling, though, a corpus of written documents will slow down language change, because it exposes present-day language learners to input data from the past. In effect, writing allows communication with speakers of past versions of a language, and this makes it harder for old forms to die out.
Greater ease and speed of travel and communication will also put the brakes on language change. Groups that would previously have fallen out of contact no longer do, continuing to exchange information and speakers. For example, in a number of countries in early modern Europe, the process of language divergence was arrested by improvements in communications (I'm thinking of France, but it's not a perfect example—if you know more about this than I do (likely!), feel free to mentally replace it with a better example). If enough people move around quickly enough, in fact, the gradual drift of regional dialects can be continously smoothed out. I doubt American English, for example, will ever split into different languages—we move around too damn fast. It's possible that the various large geographically separated groups of English speakers (US, UK, NZ, Australia, etc.) might drift apart, since the exchange of speakers is more limited, but is it plausible that they'll lose mutual intelligibility in a world that has TV, telephones, and the Internet? I doubt it.
Think of it this way: any given global circumstance of geography, communication, and travel can support a certain amount of language diversity. If there's little communication and travel (as there was 5,000 years ago), just about every village can have it's own language. But when faster travel and better communications come along, the amount of linguistic diversity that the global circumstance will support drops. The result is bound to be language extinction. Historically, of course, these extinctions often happened as side-effects of genocides—nothing speeds a language on its way to extinction like killing off all its speakers. Even without this sort of unpleasantness, though, language diversity will disappear, either through leveling of dialects or when new generations of speakers of minority languages choose to learn the majority language.
Is this bad news? I'm not so sure. Obviously, as a budding linguist (working on a typology project, fer chrissakes!) I think that a great variety of languages is handy to have around, because they're grist for the research mill. Also, we should remember that in many cases moribund minority languages are on the brink of extinction as a result of historical oppression, and in these cases it's only reasonable for an (appropriately apologetic) majority to help keep the languages alive. But apart from cases where languages are dying because people have been (or are being) coerced into not speaking them, I think we'd better resign ourselves to the coming extinction of most of the world's languages. This doesn't mean we should do nothing—there's still time for linguists to document them before they're gone. It's often claimed that the loss of languages means the death of the associated culture, but I'm not sure that's the case. Human languages are extremely flexible: stories, oral history, and traditional medical knowledge that can be expressed in a dying language can also be expressed in a new language. In addition, language communities above some hard-to-determine size threshold (perhaps one medium-size city with at least one TV station and newspaper?) can probably preserve their languages if appropriate steps are taken. The example of Welsh in Hitt's article seems promising.
In short, I think the current level of language diversity in the world is unsustainable given how easy travel and communication are in the modern world. What should be done? Well, speakers of minority languages should try (with the help of the linguistic majority) to preserve them if they can, and linguists should document them if they can't. Other than that, I think we speakers of non-moribund languages should be ready to welcome a lot of new folks with interesting accents and stories to tell into our language communities. That's not such a dismal-sounding future, is it?
[Aside: the NYTime Magazine article mentions Kenan Malik's essay "Let Them Die", which discusses some of these same issues. I just read it, and don't entirely agree with what he has to say or how he says it, but it's another opinion on the same subject, and so I wanted to mention it.]