Linguists have, in general, done a poor job of articulating why people should care that half of the approximately 6,900 languages spoken on the planet will be extinct in a century. And despite heaping scoops of truism and sentimentality atop exoticism, journalists haven't done much better. As for me, afraid of having to dip into the sentimentality and the fetishizing of Last Things, I've kind of been repulsed by the topic and have never written about it.
Until now, that is.
A new book by K. David Harrison, a linguist at Swarthmore, titled When Languages Die, looks at what we lose when languages disappear. Unusual ways of counting. Unique landscape names and calendars. Specialized vocabularies for the natural and agricultural world. Fantastic rarities of grammar, such as the suffix -sig in the Siberian language Tofa that means "smelling like." Examples come from dozens of languages from all over the world. He even illustrates with his own adventuring among nomads in Siberia and Mongolia, hunting down the last speakers of atrophied cultures.
But what caught my eye was this claim by Harrison: "Languages can package knowledge in radically different ways, thus facilitating different ways of conceptualizing, naming, and discussing the world." Elsewhere he calls languages "packaged information." In systems of kinship terms, for instance, which vary dramatically among different cultures, each one is "the result is a highly compact, highly efficient system of knowledge that packs multiple bits of information into small spaces."
In other words, languages are design objects. And I thought: no one loves extinct or endangered design objects more than designers do.
Harrison stakes a few small poles of his argument with sentiment (e.g., we will be immeasurably impoverished when these languages disappear) and truism (e.g., people have a right to speak their native tongue). The big pole in his tent, though, is the cold hard trope of information packages.
Folk taxonomies have already proven useful for knowledge architects and ethnobotanists from pharmaceutical companies alike. "For efficient communication, the packaging of information is crucial," he writes. "Names for animals and the way these names fit into organizational structures, can transmit (or omit) a great deal of information."
Harrison goes even further: local calendars, such as the lunar calendar of the Natchez, provide evidence of the diffusion of non-native plants like peaches and watermelons to the lower Mississippi, which became the names for months (along with "mulberries," "great corn," and "chestnuts") by the 1750s. No one speaks Natchez anymore. Some languages with words for categories called "classifiers" demonstrate how varied the ways of parsing the world: in Nivikh, a Siberian language with 300 speakers, has 27 classifiers; in Squamish, a Pacific Northwest language with 15 speakers, you use a different number depending on if you're counting humans or animals.
"Of course these languages are endangered!" someone will inevitably crow. "English or the other major languages are informationally more efficient!" This isn't true. For example, what information is encoded in the English "my nephew"? For sure, it's a male person. But (as Harrison writes) "is he related to me by blood or marriage? Unclear. Is he older or younger than me? Unclear. Is he the son of my sister or my brother? Unclear. Is he the son of an older sibling of mine or a younger sibling? Unclear. Is he a boy or a man? Unclear." Taken in absolute terms, English isn't so efficient: we'd need a separate book to list all its major inefficiencies.
One drawback of When Languages Die is that Harrison goes back and forth between "knowledge" and "information;" surely these are part of a folk taxonomy that every 21st century American academic should know cold. He doesn't nail the "so what?" question, either. But it doesn't really matter. This book has enough detail for a sympathetic reader to fill in the blanks.
But let me take the information-language connection to the next logical step: how do you manage a language as intellectual property? This is not as absurd as it sounds: last year the Mapuche, an indigenous group in Chile and Argentina, sued Microsoft for not consulting them when their language, Mapuzugun, was included in a foreign language pack for Windows XP. Approximately 300,000 people speak the language, enough people for Microsoft to justify getting involved. There are numerous nuances to consider here, many of which have been discussed by linguists at Language Log.
At one level, this is absurd. How can you protect a language, and who would own it? Yet if we grant David Harrison his argument, perhaps that's exactly what the speakers of an endangered language should do to protect it: learn how to manage their information packages better.
Michael Erard has written in The New York Times, Wired, Slate, and The New Republic about language at the intersection of technology, policy, law and science. He has an MA in linguistics and a PhD in English from the University of Texas. His book about verbal blundering, titled Um..., will be published by Pantheon later this year.