Į turinį
Atsakyti

Duolingo - a puzzle

sudanglo, 2015 m. gruodis 2 d.

Žinutės: 9

Kalba: English

sudanglo (Rodyti profilį) 2015 m. gruodis 2 d. 11:59:26

According to a TED lecture I viewed on You-tube, the motivation behind Duolingo is to put to work thousands of language learners in providing translations (whilst they are learning), so that eventually the whole of the web (or at least the English Wikipaedia) can be translated into other languages.

The idea is that in the translation exercises, phrases from actual web documents will be employed. Then, by some sort of voting or averaging principle, the best translation can be arrived at.

However in my ditherings with Duolingo, I don't see anything like that. The phrases in the translation exercises seem to be more of the plume de ma tante variety found in older language learning textbooks.

Where is the advanced stuff, and how does one access it? I certainly don't want to plough through all the elementary exercises to unlock the door to the advanced material.

Incidentally, I am not sure that the pedagogical approach at Duolingo is best suited to a language like Esperanto where, because of its regularity and lack of idiom, an abstract rule can be more efficient.

However it is a very good thing that Duolingo is attracting so many new learners to Esperanto, and if its techniques are motivational, keeping the learners on track to higher levels, then one doesn't want to be too picky.

Vestitor (Rodyti profilį) 2015 m. gruodis 2 d. 16:05:23

There are documents you can translate - building up your 'translation level' - but I've only encountered it on some of the courses (Spanish and German). I think it is only in force on the courses solidly out of Beta.

They are under the 'immersion' tab on the menu with Home, Activity, Discussion...

erinja (Rodyti profilį) 2015 m. gruodis 2 d. 16:54:43

Yep it's in a section for extra practice, it's not within the language learning tree for any language. Even the well-established languages have the same old-fashioned textbook style phrases in their trees.

Talisman (Rodyti profilį) 2015 m. gruodis 3 d. 05:58:29

I saw the ted talk you are referring to, Translation services was it's primary thought on enumeration, although I saw him back down from that idea as he said Translation Services is a race to the bottom. Price wise, Instead for Languages like english you can pay $20 for a Language Certificate.

When He met with the president about DuoLingo he did not even mention Translation and mentioned the Certifications for income.

Alkanadi (Rodyti profilį) 2015 m. gruodis 3 d. 15:25:02

I am just as puzzled as you over this issue. I am really confused and curious about how the translation portion works.

mvk20 (Rodyti profilį) 2015 m. gruodis 3 d. 18:12:44

My understanding is that the translation portion of Duolingo has been abandoned. It's still there and functioning for the original languages that the site offered, but I think they encountered quite a few problems in terms of trying to parlay that part of the site into the funding source for the whole project. I've heard that they are now leaning on some very wealthy donors to the project who really believe in it as something that is good and worth supporting, and that they are further developing the Certificate projects to allow that to become a further source of funding as well.

The course itself doesn't differ much from the ones on offer for other languages. If you want to quickly move to the later parts of the "tree", you can test out of huge portions and open up the whole thing rather quickly. Even after the last checkpoint where you can test out of multiple skills, you can test out of the individual skills rather than doing each individual lesson.

My opinion is that the availability of this course has been an amazing boost for our language...and who knows, perhaps for this site as well. There is often talk in the forums of where to go after "completing the tree" to continue progress, not just in Esperanto, but in all of the languages. I've seen this site referenced a number of times. As more and more complete the lessons at Duolingo, perhaps they'll turn here for more!

Vestitor (Rodyti profilį) 2015 m. gruodis 3 d. 18:36:47

mvk20:My understanding is that the translation portion of Duolingo has been abandoned. It's still there and functioning for the original languages that the site offered, but I think they encountered quite a few problems in terms of trying to parlay that part of the site into the funding source for the whole project. I've heard that they are now leaning on some very wealthy donors to the project who really believe in it as something that is good and worth supporting, and that they are further developing the Certificate projects to allow that to become a further source of funding as well.
That's interesting to know. I've been on Duolingo since it began and I never much liked the translation idea. I did one document and never bothered after that. Clearly many others had the same idea.

rikforto (Rodyti profilį) 2015 m. gruodis 4 d. 03:52:55

Yeah, it seems misguided to me.

Translation requires a certain intimacy that you do not get from a course alone. You have to be able to see the literary nuance and technical aspects of the language. Dare I be a bit elitist---not every bilingual speaker has that. Do we think that Duolingo, whatever its merits, is even producing a lot of bilingual speakers?

Think about all the nuance required by Esperanto, a language made to cut language down to something more manageable to learn. Now think about all the nuance required to read something in say, English, and capture the layers of meaning, even in a straightforward text. It just does not add up to a project I think Duolingo is up for.

uosuaq (Rodyti profilį) 2015 m. gruodis 16 d. 04:30:09

The original idea of using reCAPTCHA as a way to get people to do small word-recognition tasks that are difficult for computers was brilliant, but the original Duolingo business model always seemed like a stretch to me -- although I think we should be *very* grateful to Duolingo for taking on Esperanto, since even if their paid-translation business model worked, there was never going to be any money in Esperanto.
My guess is that if Duolingo is going to turn a profit (and I certainly hope it can continue, and continue to be free for the general user), it will be through a sort of paid subscription for educational institutions -- "use Duolingo to teach your 3rd graders French, and for $X per student, we'll have a team of experts standing by to help them with their homework". Something like that.
But failing that, it would be nice if governments would contribute funding -- surely Norway would like people to learn Norwegian, etc.?

Atgal į pradžią