Entering content area for The Web of Language

showing results for: February, 2008

blog posts

  • Doomsday machine? Engineers devise model to predict when your language is going to die . . .

    Two engineers from Cornell University’s Department of Theoretical and Applied Mechanics have devised a mathematical model to predict when the language that you’re speaking right now is going to die.

    After studying what’s been happening to Gaelic and Welsh, which are succumbing to English, and Quechua, an indigenous language of South America that is being eroded by Spanish, Daniel Abrams and Steven Strogatz used probability theory and some graphs to prove that when languages compete, one of them will go extinct. The winner then moves on to the next round in the game of survival of the fittest language.

    Probability formula tracking language death

    Abrams and Strogatz use the formula above to measure language competition and death. The authors tell us, “Suppose an individual converts from Y to X with a probability, per unit of time, of Pyx(x,s), where x is the fraction of the population speaking X, and 0 ≤ s ≤ 1 is a measure of X’s relative status . . . [and] y = 1 - x is the complementary fraction of the population speaking Y at time t.” If this probability calculation is too opaque for most readers, then the graphs that the authors supply, each showing a language in free fall, should get the point across. Of the graph below, for example, they tell us, “Using the fraction of Catholic masses offered in Quechua in Peru as an indicator, we reconstructed an approximate history of the language’s decline.” (Abrams and Strogatz, “Modelling the dynamics of language death,” Nature 424 [Aug. 21, 2003]: 900).

    Graph of decline of Quechua 

    The “take no prisoners” language wars have been going on as long as languages helped define homo sapiens as human about 200,000 years ago, give or take (or, if you’re a creation scientist who thinks the world is only 6,000 years old, give or take, ever since the Tower of Babel suffered a catastrophic structural failure that any good mechanical engineer could have predicted).

    But the battle of the languages has been heating up, and the bad news is that, for most of the world’s languages, time is running out. According to estimates, there are about 5,000 languages in use today, and up to two of them die each month. So in the four and a half years since the Cornell research was reported in the journal Nature, as many as 125 endangered tongues have gone the way of Latin, Sumerian, Cornish, a Celtic language of England that survived into the 19th and possibly the early 20th century, and Eyak, an Alaskan language whose last speaker died last month.

    In every contest, the higher-status language is the one destined to come out on top, and if you don’t speak one of the world’s prestige tongues, your language may be next. This should be good news for the superpatriots convinced that English is about to be overthrown in the United States by a willing coalition of Spanish and the other immigrant languages: in the status department, none of them stands a chance against English.

    Not only does English remain at the top of the charts in the U.S., with speakers of other languages rushing to adopt it, but the reach of English around the globe far exceeds that of previous world record holders like French and Latin, or any of the other languages that have lots of speakers today, like Chinese, Hindi, Arabic, and Spanish.

    If the assumptions of the Cornell model are correct, English will continue to take on all comers and win. But those assumptions, even though they’re propped up by formulas and graphs, just might be wrong.

    Rejecting the possibility of peaceful co-existence, the researchers insist that languages in contact always compete against one another for speakers. They further assume that no geographical or social boundaries separate speakers. And they assume that everyone uses only one of the two competing languages.

    But while these assumptions might work for a shoot-’em-up video game, where designers can control every element of the play, they don't provide a very good picture of how language functions in the real world, where no one’s really in charge of what we say.

    For example, real people routinely face social barriers that impede their interaction. They often speak more than one language. And the language they hold onto may not always be the one with highest status.

    The engineers do acknowledge the existence of bilingualism, insisting, though, that it’s always a cause of conflict. It’s true that transitional bilingualism signals that a population is abandoning language X for language Y, but bilingualism may also be stable over the long term, and we have only to look at Hellenistic Greece and Moorish Spain to remind us that complex, multilingual cultures have flourished since people started to talk.

    Finally, as far as status goes, even high prestige is no guarantee of linguistic survival: Latin remained the highest-status language of Western Europe for more than a millennium after it passed from living language to dead one.

    Pointing to the increasing health of French in Québec, Abrams and Strogatz are optimistic that their mathematical model of language death offers activists a way to revive moribund languages by taking steps to raise the value of s in the equation: “strategies such as policy-making, education and advertising [can increase] an endangered language’s status.”

    It’s a relief to know that language preservationists don’t have to master probability or theoretical mechanics after all. They can fight to hold onto their languages by doing what they’ve always done: lobby governments, cajole students, and put up the occasional billboard, though to get anyone to read it, the ad would have to be in the dominant language, not the endangered one. And face it, even a TV spot in English at half-time during the Super Bowl would have been too little, too late to save Eyak from going under.

#1
john.bowden@anu.edu.au Feb 17, 2008 6:51 pm

An interesting article, and you're right to point out that multilingualism is an important factor that needs to be taken into account when we look at 'competition' between languages. In parts of the world like much of Melanesia, groups of people with as few as a couple of hundred speakers have managed to maintain their languages for centuries because everyone expects to speak several languages.

There is another problem with the formula, too, though, which you've overlooked  in your otherwise excellent article. Latin never died: it just developed and split off into a whole lot of other languages including Spanish, French, Catalan and a myriad of Italian dialects. Proto-Germanic split into English and German and Dutch and many other languages. And these languages will all do the same eventually (if they're not killed off by the competition in the meantime). The signs are already here that English is doing just that: there are millions of native speakers of Indian English, which is not necessarily comprehensible to people from the US and other 'traditional' English speaking areas. As a speaker of Australian English, I find Apalachian and highland Scots pretty harfd to understand as well. Singapore and Malaysia have their own distinctive English too. Given enough time, English won't necessarily die, but it willhave turned into a lot of new languages, just as a lot of so-called 'dead' languages have done in the past.

additional blog information