Topsiderunner Diglot Senior Member United States Joined 6929 days ago 215 posts - 218 votes Speaks: English*, French Studies: Italian, Arabic (Written)
| Message 1 of 9 26 October 2006 at 5:47am | IP Logged |
http://news.bbc.co.uk/2/hi/health/6083994.stm
Just an article about a translation system that allows two people to speak
with each other in different languages. Each person mouths their words
and it is then translated and spoken aloud to the other. I'm just
wondering what this kind of technology will mean for language learning
in the future, say 20, 30 years.
1 person has voted this message useful
|
Felixelus Senior Member United KingdomRegistered users can see my Skype Name Joined 6849 days ago 237 posts - 244 votes Speaks: English* Studies: Spanish
| Message 2 of 9 26 October 2006 at 6:15am | IP Logged |
Yeh I've just read about that! It's slightly disheartening at first but then you think that this sort of technology will only be available to the rich/powerful at first. So I should continue down my learning road.
1 person has voted this message useful
|
patuco Diglot Moderator Gibraltar Joined 7026 days ago 3795 posts - 4268 votes Speaks: Spanish, English* Personal Language Map
| Message 3 of 9 26 October 2006 at 6:30am | IP Logged |
Topsiderunner wrote:
I'm just wondering what this kind of technology will mean for language learning in the future, say 20, 30 years. |
|
|
Assimil, Pimsleur and the others will be out of business since there'll be no need to learn any other languages ;-)
1 person has voted this message useful
|
lady_skywalker Triglot Senior Member Netherlands aspiringpolyglotblog Joined 6901 days ago 909 posts - 942 votes Speaks: Spanish, English*, Mandarin Studies: Japanese, French, Dutch, Italian
| Message 4 of 9 26 October 2006 at 6:33am | IP Logged |
To be honest, I don't know how effective such technology would be. They'd have to program it with a whole lot of languages...just imagine how big the vocabulary databases would have to be to include slang, idiomatic expressions and the like! We've all seen how Babelfish and other translation software mangles sentences. I'm just wondering if this 'universal translator' would be just as useless.
1 person has voted this message useful
|
SamD Triglot Senior Member United States Joined 6670 days ago 823 posts - 987 votes Speaks: English*, Spanish, French Studies: Portuguese, Norwegian
| Message 5 of 9 26 October 2006 at 8:16am | IP Logged |
It's also worth considering the rate at which language changes, particularly slang and idiomatic expressions and jargon.
Factor in accents and people who don't speak clearly, and somehow I don't think Assimil and Pimsleur and the like have much to worry about.
1 person has voted this message useful
|
rkunz Diglot Senior Member United States learnthatlanguagenow Joined 6836 days ago 103 posts - 101 votes Speaks: English*, Spanish Studies: Japanese
| Message 6 of 9 03 November 2006 at 5:07am | IP Logged |
"If the prototypes used a small vocabulary of about 100-200 words they worked with about 80% accuracy, researcher Tanja Schultz said. "
Yeah, it's going to be a while...
1 person has voted this message useful
|
Topsiderunner Diglot Senior Member United States Joined 6929 days ago 215 posts - 218 votes Speaks: English*, French Studies: Italian, Arabic (Written)
| Message 7 of 9 03 November 2006 at 12:53pm | IP Logged |
Well, seeing that computers can already instantly translate to a relative degree of accuracy most major languages in text form, I think that within most of this forum's members' lifetimes, instant face to face translation will become a reality. Would this make language learning obsolete, more valuable, only an antiquated hobby like horseback riding today?
1 person has voted this message useful
|
virgule Senior Member Antarctica Joined 6851 days ago 242 posts - 261 votes Studies: Korean
| Message 8 of 9 03 November 2006 at 5:01pm | IP Logged |
That's going to be like speech recognition for the personal computer ;) I mean in that it's going to be technology around the corner for a very long time to come. I don't know how they want to do a device (I did read the BBC article when it came out) when speech recognition is not accurate enough? Add to this uncertainty when translating by machine...
Merge ViaVoice with BabelFish and TextToSpeech ;)
It's a welcome addition to translation and knowledge of languages, but it's not a replacement. I can see uses for such devices, such as in situations where other options are unavailable.
Why did I mention dictation software? Because for a long time they thought that if only computers were powerful enough to handle really large databases, the problem would be solved. Computers have been powerful enough for quite a while now, and accuracy doesn't increase any more. What has happened is that the researchers have found that they employed the wrong paradigm. It's just that nobody knows what would work.
As with speech recognition, the mentioned device will work quite reliably in situations where the context and thus vocabulary is heavily restricted. Recognizing all the numbers of a languages is no problem for a computer (thus the success of voice activated telephone systems), but if the sounds could be anything...
1 person has voted this message useful
|