Science News

Languages more endangered than animals

Wed, 3rd Sep 2014

Chris Smith

Up to a quarter of the world's existing languages are facing the threat of extinction, says UK scientists.  Taiwan Aborigines

Seven thousand languages are spoken globally but many are disappearing and the rate of loss is accelerating. When tongues become extinct, descendents of native speakers talk of the loss of cultural heritage and history, cohesion and tradition.

However, the factors that contribute to language loss, and the parts of the world most at risk, have not been well studied.

But what is clear is that the same factors that give rise to high levels of species diversity also appear to drive language diversity.

In other words, the parts of the world with the broadest range of plants and animals also tend to have the greatest language diversity, because the same factors that drive the evolution of new species also drive the evolution of new languages. These include geographical isolation and highly specialised living conditions.

This being the case, Cambridge scientist William Sutherland and his colleagues applied the same criteria that are used to identify at-risk species to the spoken word.

These include factors such as the speaker population size, how that poulation size is changing, and the scale of geographical range they inhabit.

This has enabled them to identify those languages and those regions most at-risk, including hotspots in northern Australia, northwestern USA and Canada, the tropics and the Himalayas.

In many of these areas a toxic cocktail of small population sizes coupled with rapid economic growth, which inevitably leads to adoption of more common global languages, like English, is driving the loss of indigenous tongues.

"We need to take steps to preserve these languages," says Sutherland, "because wrapped up in them is a huge amount of cultural and historical information. There are also social impacts on the next generation who have lost a language."


Subscribe Free

Related Content


Make a comment

If you're in a position to do so, record as much of a dying language as you can. Artificial intelligence will eventually be able to revive it by bringing up new native speakers whose parents decide they want the language of their ancestors to be revived. There is no reason for any language to be allowed to disappear altogether. David Cooper, Mon, 15th Sep 2014

Do you think that's possible though, David?

In my question to Bill Sutherland in the interview I pointed out that there are nuances and subtleties to a language that might not carry over with a recording. chris, Mon, 15th Sep 2014

Some languages that were actively opposed by the government have had a bit of a resurgence as providing a local identity.
Two examples I have seen are Breton, in North-west France and Catalan on the Mediterranean coast of Spain; in both cases the government attempted to suppress the local language and enforce sole use of the national language.
Perhaps more familiar to UK listeners might be the Welsh language - but it has not been recently suppressed like the other two languages.
What will be preserved here is a regional language; an homogenised and codified version of what was once a group of local dialects.

I think the bulk of languages most in danger of disappearing are those spoken by just one small, isolated tribe - there are a number of examples in New Guinea and outback Australia. evan_au, Mon, 15th Sep 2014

The purpose of language is to facilitate communication. The fewer the languages, the easier the communication. Languages die because nobody needs them (Cornish), or because they evolve (Latin), or as a result of conquest and extermination (not a good thing, but it is the conquest and extermination that are bad, not the loss of a language).

Whatever happened to Middle English? Why does nobody mourn the death of Anglo Saxon? Because modern English is understood by far more people.

Classic cars and vintage aeroplanes are fun to play with, but not for getting from A to B. Why don't we see  Stanleys or Klemms every day? Because Fords and Boeings are better. By all means study the 300 languages of the Papuan Highlands, but if a Papuan wants to trade with anyone outside his immediate village, don't complain if he uses English, and if he decides it would be better for his children to learn English, rejoice!    alancalverd, Mon, 15th Sep 2014

When I say record, I mean store as much detail as you can about the language in terms of its grammar and vocabulary, but also record audio of as many speakers as possible and get full translations of everything, getting them to talk about the past and to tell any traditional stories they might remember. The more of this you can get, the greater the number of idiomatic phrases and other expressions you will be able to collect, and that will make it possible some day to revive a richer version of the language than you otherwise would. In the case of many languages it is urgent and there is no opportunity to teach a next generation, so collecting the data matters now. A lot will likely still be lost, and this is inevitable once the number of speakers is down to low numbers because different people use their language in different ways, some having much more impoverished speech than others. Many subtle things will be lost, but there is no case in which we should just give up and just let the whole thing go.

As for why we should try to preserve languages, they are not only cultural treasures, but they also hold clues to how other languages are related and any one of them may contain clues to the nature of older languages which are long gone, but which may be possible to reconstruct to some degree from those that descended from them. A word, for example, might be similar in several languages in very different places on the planet, and it may just be a survivor of a language that came with us when we first came out of Africa. We don't know how much we may be able to reconstruct, but if we lose the data that exists now, we lose the ancient data in the process.

I find no end of interest in looking at how different languages hold together grammatically and have worked on over fifty now. Each one is fascinating, and the radically different ways which they approach things offers clues into how the underlying language of thought works. This can sometimes impact on the way people think too. There are two main groups of languages, some which I call "outward reading" and others which I call "inward reading". English is outward reading in that you speak the components of relative clauses outwards from the central idea of the sentence. Japanese, Basque and Hindi are inward reading, because you start at the far end of relative clauses and work back in towards the central point of the sentence. For example, "Have you met the man who works in the garage next to the post office" is reading the relative clause outwards, but an inward-reading language would say something more like, "post office by garage-in works man, you met?" It has been shown in experiments with a fish tank that Japanese test subjects will take in more of the details as to the whole environment of the tank, while American test subjects will take in more details of the main occupants of the tank. This is likely because their language forces them to think inwards instead of starting with the main event.

The human mind is fully capable of working with a dozen or more languages (speaking them fluently), but merely having more than one is good for you as it appears to make people better able to cope with dementia for a good few years longer, until their decline speeds up and they catch up with monolinguals. But, there is one particularly good reason for wanting to speak an obscure language that few other people will have bothered to learn. In the future, the main way of working with computers will be to talk to them and to have them talk back. You don't always want everyone around you to be able to listen in and understand what you're saying. It would be well worth choosing a random obscure language for every child to be brought up speaking as a second language, one which machines use to communicate with them. Machines should also refuse to translate anything they overhear if nosey people ask them to. With AGI helping to bring up all children, every child can get an extra language for free, or maybe half a dozen if they take to it easily.

I am disappointed that two of my ancestors' languages have been lost without being recorded. Pictish has some surviving words in placenames, but that's about it. Fortunately it is a dialect of British and was not greatly unique - it was similar to Cumbric, Welsh, Cornish and Breton. While Cumbric is extinct, I have heard that it is extensively recorded (on paper) so it could perhaps be revived. It used to be spoken as far north as Glasgow. The other language spoken by my recent ancestors (several thousand years ago) was the first language into the British Isles, and that was believed to be a relative of Basque.

The language(s) that came out of Africa are thought to have been of the inward-reading variety. Over time there has been a switch away from that to outward reading, with languages evolving in that direction. They do not appear to evolve in the opposite direction, so it may be that the original human language was a long way from having the best word order in sentences. This may go back to homo erectus for all we know, but it would be great if we could find out. Clues to help with that could turn up in any rare language, and that's why they need to be preserved. A lot of people won't care, of course, but they have an impoverished world view which leads them not to value knowledge which is of interest to others. There is no requirement for them to take an interest, but they should not stand in the way and should not encourage others just to throw away irreplaceable knowledge without a thought.

Older versions of English (and other ancient languages) are still highly valued by many people who like to learn them and to study the literature that was written in them. The same applies to other modern languages, and it is much more rewarding to read a book in its original language than to read a translation. But, for those who only care about utility, prepare to ditch English. I've been building the perfect artificial language for a very long time and will release it at some point in the distant future. Its lack of ambiguity, perfect logical form and the naturalness of its sentence order will make it far more successful than Esperanto, Lojban or any other artificial language that has previously been constructed. It actually allows you to use it as an inward and outward reading language, but it sticks to SVO or OVS order everywhere. The most important thing though is to minimise vocabulary by making as many words as possible derivable from more fundamental words through the application of standard rules. I will not show any examples because the most important idea behind it is so radical that I don't want anyone to see it until the language is launched. The result may be a complete dictionary on a single page, though it would take many pages of examples to show how to build the fundamental words together to make the rest. People have attempted to do this before, but without the 100% regularly-derived aspect, so they take a couple of words and fling them together to create a new word which doesn't quite fit the components. I'm trying to do the whole job properly. Maybe I'll get it finished in another twenty years. David Cooper, Mon, 15th Sep 2014

It looks as though you are trying to invent German, though probably order of words the best it has not. Have you considered the advantages of structured case endings? It's easy to be poetic in Latin because the rhythm and the meaning of a sentence are not linked.

Dead languages certainly have their charm and interest, but surely the reason they are dead is because they were less useful than their successors, so people chose to use the new ones. Wherever the old language is recorded, nobody is throwing away knowledge or destroying anything, but it isn't up to everyone to preserve everything: we pay museum curators to keep material things that are of interest rather than utility, and there are still plenty of scholars studying ancient texts and societies.

I'm intrigued by the effect of having multiple languages on dementia. My British aunt and uncle became fluent in French in their middle years, then moved to the USA. When my aunt had a stroke she was unable to transmit  or receive in English but my uncle (and others) could communicate with her in French until she died. I've read that languages learned in later life reside in a completely different area of the brain from your childhood language, but I'd like to know what goes on in the brains of people brought up truly bilingual from birth. 

I hope that in a few years people won't talk to computers, but will go back to talking to people. As for having a secret language for man-machine interaction, it's already here! I earn most of my living nowadays waiting for people to find the password to their medical imaging computers: the record so far was 4 hours on site to do a 20 minute test (I'm paid by the hour). When I design machines I try to make them transparent, not obscure, so anyone can use them. alancalverd, Mon, 15th Sep 2014

German's just as much a mess as any other natural language. Sticking the words "far" and "see" together makes the word "television", for example, but it's all done in a slapdash way. It's difficult to do it properly with most things without ending up with very long names which describe what they are and do, but there are ways round that which make things manageable. There are whole areas of vocabulary which can be built out of fundamental components but which never are in natural languages, and that's what my artificial language seeks to put right.

Many dead languages were not abandoned, but evolved into modern languages. Of those which actually have died, it is nothing to do with them being inadequate, but is caused by political change. In Scotland, Pictish was abandoned in the course of a few generations as people switched to Gaelic which had a higher status due to there being a Gaelic-speaking king. It was not recorded and cannot be recovered. I'm not demanding that anyone should go on speaking dying languages though - I'm simply trying to get across the importance of recording them. I don't want to force languages on people who don't want them either, but I do want them to be available to people who decide they want them back.

The most productive conversations people have in the future will be the ones they have with machines. That doesn't mean that they'll spend less time talking to other people though - quite the opposite, because most arguments and discussions will be quickly settled and proved beyond doubt by AGI, so we'll be free to spend more time socialising with people instead. It will, for example, be possible for every important idea about an issue to be presented to everyone who needs to hear it, and they will no longer just have the loudest voices waste most of their time bombarding them with a tiny selection of biassed and distorted ideas. David Cooper, Tue, 16th Sep 2014

See the whole discussion | Make a comment

Not working please enable javascript
Powered by UKfast
Genetics Society