Translation Dutch to English?
Thank you for creating duolingo, which allows me to learn Spanish.
Now I have been wondering, is there a way I can help to make translations from Dutch to English so that (dutch) people can learn English from their native language?
Don't forget the about 7 million native speakers of Afrikaans. In Indonesia, there are a lot of people (i.e. several millions) who speak it as their second language and since duolingo already has German working, it should be fairly easy to implement. I guess it's more of an economical question for duolingo: is it worth it to hire somebody who speaks Dutch as his/her native language?
There are several other websites available to learn other languages out of Dutch. E.g. if you want to learn Spanish in Dutch, you could use:
-Course on http://www.kerknetputten.nl/kpbasis/spaans.php?msgid=40&groep=spaans&menusel= -Word learning tool on http://www.spaansewoordjes.nl/
There is a very interesting electronic translation engine offered to the public at no charge by Google (translate.google.com). I have found it very useful as a translation/editing tool, and I am very grateful that Google makes it available to the public at no cost!
This resource is based on statistical methods that generate translation algorithms by comparing large numbers of bilingual texts. The more texts it compares, the better it gets. It is now pretty good for going between English and the Romance and Scandinavian languages.
It is less useful for German and Dutch because the subordinate clauses of these languages deviate from the general SVO (subject, verb, object) structure of the principal clauses of these languages. This makes it very hard for Google's system to translate both these languages accurately.
(It does better with the other languages I have mentioned, which generally preserve the SVO order even in their subordinate clauses.)
I am confident that Google's system will continue to improve as it digests more bilingual texts. Beyond a certain point, however, it will be very difficult to produce really good translations using strictly statistical methods. Good translations have to make use of a large body of general cultural knowledge on the part of the people making them, and such knowledge is very difficult to build into electronic translation engines given the current state of the art.
I believe that future research into artificial intelligence will address these deficiencies more fully. But it will be very hard indeed to develop machine-translation systems that are comparable to the kind of work that can be done by really accomplished human translators.