The cost of losing too many tongues

Each language is a new map of the world, and our knowledge of the world shrinks when a language dies, writes Andrew Dalby.

"Once upon a time," according to the biblical story of Babel, "all the world was of one language and of few words." The implication is that the number of words in the world varies with the number of languages. Logical, and in a sense true. As the number of languages in the world goes into steep decline (5,000 languages now; 2,500 or fewer by 2100) there really will be fewer words.

This is a deeper issue than it might appear. Each language is a new map of the world: a map on a different projection, with different conventions, making different choices of which details to show and which types of detail to name. Maps vary in these ways, and so do languages. The words of any one language have different ranges of meaning and connotations from those of any other.

Academics and journalists have asserted that Eskimos have anything up to 100 words for snow. This shameless exaggeration goes back to a serious study by anthropologist Franz Boas, who identified four English words for snow and four Eskimo ones. His point was that the concept of snow is divided up in a different way in the two languages. English has snow, sleet, drift and blizzard. Inuit has aput , "snow on the ground"; qana , "falling snow"; piqsirpok , "drifting snow"; and qimuqsuq , "drift". It is less interesting to count the words than to know how differently a speaker, when using these words, is thinking about the world.

Another Arctic example involves the Yupik language of Alaska. The Yupik word for "world" carries associations such as "outdoors" and "weather". European words for "world" embrace concepts such as "people", "crowd" and "inhabitants": the French have le monde , "the world"; du monde , "a lot of people, a crowd", literally "some of the world"; tout le monde , "everybody, or everybody who matters just now", literally "all the world". In classical Greek, he oikoumene is translated as "the world" and means literally "the inhabited zone". In old-fashioned English, "all the world" means "everybody we can think of". The difference is unsurprising: if you look outdoors in urban Europe, you see a world of people. If you look outdoors in rural Alaska, you see a world of weather. The interesting point is the different associations that come with words that are translated identically. This mismatch is one of the reasons why translations of Chinese lyric poetry into English are impossible to do well. If you learn Yupik or Chinese or even French, part of the process is learning different associations for what you thought were commonplace ideas.

Each language has its own classification of the world and its contents. English, for example, has a term "nut" that includes walnuts, hazelnuts, almonds, peanuts, cashews and others. French has a superficially similar word, " noix ", but its range of meanings is different. Noix in French means "walnut". French has no precise equivalent for nut; the temptation to translate French noix by English nut must usually be resisted. So it is that you can never go directly from one language to another in translating, nor even from one to another by way of a dictionary. You always have to go from one to another by way of the real world, because it is the real world that each language is a map of.

These differences between languages make life hard for translators, but they guarantee that each language brings different insights, new connections, unforeseen enlightenment. Children learn one way of looking at and classifying the world from their first language. If they grow up bilingual, they learn more ways. The feeling of puzzlement or frustration occasionally expressed by a bilingual child who can't quite make the precise statement intended in the language currently being used is richly compensated by the growing ability to see behind the structure of languages, to identify the mismatches between them, to grasp the concepts that are more easily explored in one language than another.

In each successive generation, there are millions of bilinguals. In the least developed cultures that anthropologists can now reach, the proportion of bilinguals and multilinguals is generally greater than in the developed world. It is likely that this was the case through the tens of thousands of years of human existence before recorded history. It is quite likely, though unprovable, that an absolute majority of humans have been bilingual or multilingual. At present, there is still lots of language knowledge in the world. In many countries, children learn a local language at home and a national language at school. English is the second or the third language almost worldwide. All this means a lot of multilingualism.

If current trends continue, that will change. Across the world, minority languages are being dropped as people realise that their children will be safer and more prosperous if they grow up speaking a national language - or English. If young people are not learning their local language, that language will go. Hence the estimate that 2,500 languages will disappear this century. Why should it stop there? Nationalists aim at one language per nation. We could be down to about 150 languages by 2200. As for multinational corporations, they know how much easier it will be for them when only one language counts; and their global power is rapidly increasing. Not many years after 2200, there could be just one language left. No bilingualism then.

What will our great-grandchildren have lost when there are no bilinguals? What will human beings stand to lose when the world is "of one language and of few words"?

First is the loss of insight that each individual bilingual can give - there will be no more of those flashes of inspiration that are among the rewards for everybody who learns another language. We have needed those flashes: over and over again they have forced a rethink and inspired a new technological approach or a new scientific theory. For example, where did research linking stress and the pituitary gland originate? From a north Frisian phrase for stress, meaning "he burnt his pituitary gland", which in turn derives from whaling. Whalers noticed whales caught after a long chase had swollen pituitary glands.

Second, and just as threatening, is the loss of catalysts for language change that have emerged from the clash between languages. Where did satire come from 2,200 years ago? Out of a fruitful mistranslation: we can blame the Oscan slave Ennius who, almost single-handed, created a world-class literature by mistranslating from his second language, Greek, into his third language, Latin. Where did Zen Buddhism come from? From the untranslatability of Buddhist sutras into early medieval Chinese. Where did the gloriously rich and flexible word-hoard of 21st-century English come from? Most of it borrowed straight out of other languages.

Third, there is the pure loss of knowledge, repeated and cumulated as each language disappears. "One should attend to what is old, and not allow it to be forgotten," a Sumerian poet said 5,000 years ago. We are still ignoring his advice. Eight languages were spoken in Tasmania when English speakers began to settle the island in the 19th century. They are gone: all that remains of them are three Tasmanian loanwords in local Tasmanian English, including two names of edible plants. Yet every indigenous language, if properly recorded, names up to a thousand useful plants - any one of which might be useful to human health. If forgotten, as the medicinal plants of Tasmania are forgotten, they take hundreds of years to rediscover. Do we have that long?

Andrew Dalby's Language in Danger is published by Penguin Press on June 6, and is available to THES readers at the special price of £16.99. Call Bookpost on 01624 836000 with credit/debit card details.