Hacker News Reader

Translators can end up with considerably different results. The first version of the Tao Te Ching I ever read was Stephen Mitchell's translation. I later picked up another version and just couldn't stand it.

View on HN

THAT is really weird! Because in NLP and Machine Translation the symmetric translation of text is seen as proof that the algorithm works perfectly.

And I had the opinion, that professional linguists or bilingualists, should all come up with almost or mostly identical translations. And when their translations are published, they should easily be able to have consensus, that one translation is better than the other, which reflects the coherent structure of language.

that only works for some languages - as someone else mention chinese is quite a conceptual language. I've read 3 different translations of Tao Te Ching, and there have been some wild differences. One of the copies that i have has the authors translation, and then the remaining 2/3 of the book devoted to explaining the different possible meanings of each and every sentence

in NLP and Machine Translation the symmetric translation of text is seen as proof that the algorithm works perfectly.

Are you serious? That's the dumbest metric ever. Even a translator that does nothing passes that metric.

Reply on news.ycombinator.com