9

This question is being asked on behalf of @lvc, who had the concern triggered by a question where another user helpfully improved a machine translation.

Back in 2012, I gave this answer to a question: May we use non-English languages on this site?

Quote the original source (in the original language), but please also include an English translation. A rough translation is acceptable to start with. (If an automated translation tool makes a reasonably successful attempt, that is also acceptable.)

If you are bilingual and see a poor translation, please be bold and edit the post to improve the translation.

@lvc points out that this could be considered Original Research which we deprecate.

Should we permit user-translated text, or should we require links to support the translation (e.g. a third-party source, or machine translation), perhaps in addition to user-translated?

Related:

12

Disclosure: I am the user who originally posted the edit which triggered this debate.

My personal point of view is that personal translations from bilingual users, when posted together with the original language (or at least a link to it, if the original is too long/unwieldy), does not result in significantly decreasing the quality of the posts, despite it prima facie being original research.

This is because as long as both the original and translations are provided, anyone with the capability to do so can simply check whether the translation is valid. I personally see my edit as an example of such review, and had the original translation been a poor (or maliciously false) human translation, I believe I would also have pointed it out or edited it.

I see it as no different from an answer which interprets and summarises an English source, and posting the summary in English. If the act of summarising requires technical knowledge, it is no different from the act of translation, which requires linguistical knowledge. Neither should change the overall meaning significantly when done correctly, and both require expertise to do correctly.

Of course, if the question asker doesn't speak the original language, it may in fact be a better idea for the translator to add to (as opposed to overwrite) the original machine translation, a fact that I acknowledge to have possible been slightly overboard when I initially overwrote the machine translation.

  • 1
    I perfectly agree, anyone can normally check whether a translation is valid via Google. It's one of those cases where checking something is much easier than doing it. – Sklivvz Nov 7 '15 at 15:41
  • 3
    I'm good right up to the point of adding versus overwriting. I think overwriting is fine. Three versions of the same quote seems excessive for readability. – Oddthinking Nov 9 '15 at 0:55
  • 1
    +1. As a bilingual user, I have had in the past checked other people's translations (not only on Skeptics) and explicitly noted any deficiencies. The fact that both original and translation are present makes it a valid approach. – user5341 Nov 16 '15 at 12:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .