For instance, if a paper is written in Russian, and a researcher wants to use this work and cite it, does he have to learn Russian?

What's the typical practice?

Does he try and use translation tools such as Google translate?

Pay a professional translator?

Or, blindly accept the results and blindly cite the paper and move on?

(The last option is problematic, if one wants to write a self-contained piece of work, using the results and proof from the original work.)

  • 4
    Compare: mathoverflow.net/questions/257095/…. Dec 13, 2016 at 6:49
  • 6
    This is at least somewhat field-dependent. An automatic translation of a philosophy paper may be useless. In math, reading the equations and recognizing key terms (by their Greek or Latin roots) may be enough in some cases. Dec 13, 2016 at 10:25
  • It probably depends also on the field, but I tend to read the parts that are relevant with what I need to show, whenever I read a paper. And although I avoid using papers in other languages, it has happened to be able to find only a chinese relevant study, and in that case I used the available information in the abstract. In a similar manner, how do we deal with non-open-access articles?
    – BioGeo
    Dec 13, 2016 at 13:04
  • 4
    Check with your university if they have a translation service that can help you.
    – Ian
    Dec 13, 2016 at 13:47
  • 2
    You can hire a translator. Dec 14, 2016 at 0:39

2 Answers 2


There is no general answer to your general question. It depends on a paper.

Do you have to cite it? We know that a2+b2=c2 and we don't cite Pythagoras manuscript on it. We don't cite the basic results we use.

Do you have to read it first? You probably don't read Disquisitiones Arithmeticae in Latin to quote some of Gauss results, and you don't read Théorie des mécanismes connus sous le nom de parallélogrammes to check if Chebyshev polynomials are correct.

So, if a result is already well-established, many people have seen it, checked it and built their work on it - you don't have to read the original paper, if you don't want to. If you really want to understand how some details were presented originally (for methodological or historical purposes), you can check if some colleagues in your University can read this language. If they don't, you can hire a translator, but it is not easy to find one who can also read understand math arguments.


In some cases, one mentions the relevant paper, stating explicitly that they could not read it for language reason, e.g. "related results where obtained by Author A in paper P [reference], which we were not able to consult since the article is only available in Russian". That last part can sometimes be omitted, if the knowledge that you did not read the paper itself (but, say, only a review or a translated abstract) is not useful to your reader.

This, however, cannot apply to papers you build on in a way that needs you to make sure the cited paper is correct, or in a way that needs precise statements to be understood finely. If there exist (precise) reports or survey on this research, you can use the secondary source (and mention the primary source, but making clear which one you rely on and read) or, ultimately, a translation service (but I don't know where you can find one which can be expected to be reliable for highly technical papers).

You must log in to answer this question.

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