Monday, November 1, 2010

My bright idea: English is on the up but one day will die out | Technology | The Observer

It's been the received wisdom in language technology that machine translation isn't good enough. But all that's preventing it from being good enough is just a problem of scale. The way that machine translation is now being pushed forward simply involves being able to process more and more data in order to find the significant patterns. The power and cheapness of computers is increasing all the time. There's no way that the little problem of incompatibility between languages is going to stand in the way of it for long.
And because it's being done in a data-based way, the techniques which will solve the problem will solve it for all languages, not just the big important ones. So even remote Aboriginal groups will benefit – maybe a generation later, maybe sooner. And when that happens, people will be able to fulfil themselves through their own language, which is what they always wanted to do anyway.
Well, it's happening gradually. But I want to draw a distinction between a language which is spread through nurture, a mother tongue, and a language that is spread through recruitment, which is a lingua franca. A lingua franca is a language that you consciously learn because you need to, because you want to. A mother tongue is a language that you learn because you can't help it. The reason English is spreading around the world at the moment is because of its utility as a lingua franca. Globish – a simplified version of English that's used around the world – will be there as long as it is needed, but since it's not being picked up as a mother tongue, it's not typically being spoken by people to their children. It is not getting effectively to first base, the most crucial first base for long-term survival of a language.

Link:
My bright idea: English is on the up but one day will die out | Technology | The Observer

No comments: