Let's say I want to cite a Russian text in my own paper which is written in English. Should I:

  • reference it as is, in the original Cyrillic alphabet, although the readers won't be able even to pronounce it?
  • transliterate the authors' names, while keeping the rest as is?
  • transliterate the names and translate the title of the work, journal name, etc. into English?
  • 2
    Would the answer to this question differ between alphabets which mix fairly well with Latin (Greek, Cyrillic) and those which don't (Arabic, Thai)? – TRiG Sep 30 '12 at 2:06
  • @TRiG: I am not sure I recognize why the former ones would "mix" any better with Latin than the latter ones. – O. R. Mapper Jun 4 '14 at 15:39
  • 2
    Well, @O.R.Mapper. The Latin, Greek, and Cyrillic alphabets (and the Cherokee syllabary) all use letters which "match". They are the same sizes; the shapes are of roughly the same degree of complexity. A word or two in one alphabet on a page mainly in another would not stand out at a glance. Arabic and Thai (also Chinese, which isn't an alphabet) use completely different patterns to form the shapes of the letters. Thai is taller than Latin, for example. Arabic is considerably more calligraphic. Hebrew is squarer than Arabic, and would probably match better (apart from being RTL). – TRiG Jun 4 '14 at 15:47
  • 2
    @TRiG I don't think aesthetics have anything to do with this. It is a question of being able to convey the information. – Davidmh Jun 4 '14 at 15:52
  • @TRiG: Personally, I have always found especially Cyrillic text to stand out notably from Latin text, as well, but anyway, as those considerations are merely aesthetical, I tend to disregard them if there are any more practical reasons for including the original texts (such as unequivocally finding the original publication or authors). – O. R. Mapper Jun 4 '14 at 15:56
up vote 19 down vote accepted

In general, this would depend on the style prescribed by your publisher (journal, conference, etc.) For instance, a blog post cites the Chicago style manual explains how one should go about sources in Chinese and Japanese:

10.108: Inclusion of original characters

Chinese and Japanese characters, immediately following the romanized version of the item they represent, are sometimes necessary to help readers identify references cited or terms used. They are largely confined to bibliographies and glossaries. Where needed in running text, they may be enclosed in parentheses. Computer technology has made it much easier than it used to be to typeset words in non-Latin alphabets.

Hua Linfu 華林甫, “Qingdai yilai Sanxia diqu shuihan zaihai de chubu yanjiu” 清代以來三峽地區水旱災害的初步硏究 [A preliminary study of floods and droughts in the Three Gorges region since the Qing dynasty], Zhongguo shehui kexue 中國社會科學 1 (1999): 168–79 . . .

Harry Harootunian and Sakai Naoki, “Nihon kenkyū to bunka kenkyū 日本研究と文化研究, Shisō 思想 7 (July 1997): 4–53.

That year the first assembly of the national Diet was held and the Imperial Rescript on Education (kyōiku chokugo 敎育勅語) issued.

Harvard style and reference guide requires a translation, followed by the original name:

Milani, F. (2001) The Phantom of the Opera. [Le Fantome De L’Opera] Paris, LeRoux.

  • 2
    FYI: This is rather painful to do in (La)TeX, but relatively straightforward in Xe(La)TeX, which has native Unicode support built-in. BibTeX is Unicode-ignorant, so it also works just fine (as far as I can tell). – JeffE Jun 13 '12 at 15:21

I will just add to good Alexander Serebrenik's answer. The main point of references is to provide traceable sources to information. This means the translation of the title and journal are key. The title provides insights into the content of the paper and the journal makes it traceable. All journals have different "standards" for how to do this in detail. It is also common that articles in French, German, and/or Spanish are not translated (again local "rules"). It is not common to see, for example, cyrillic in references, instead translitteration seems to be the most common. Again, there should be "rules" about this in each journal and so the best appraoch is to contact the journal editor and ask if no explicit information is available.

Here is an example of how journals may wish to see the references:

Author(s), year. Title in original language (if possible) [Title translated into English]. Publication name in original language (if possible) [Publication name translated into English]. Volume/issue/page information (according to type of publication). [In ‘language’]

and as an example:

Krenke, A.N. and Khodakov, V.G., 1966. O svyasi povercknostnogo tayaniya lednikov s temperaturoy vozdukha [On the relationship between melt of glaciers and air temperature]. Materialy Glyatsiologicheskikh Issledovaniy [Data of Glaciological Studies], 12. 153–163. [In Russian]

  • I think I cannot follow your line of reasoning, or maybe I misunderstand your text. Translating the title and journal name to the paper language provides some superficial insight into what the reference is about, but other than that, providing (exclusively) a translation makes traceability worse. A document certainly can best be found if its original title is known, rather than a translation thereof. – O. R. Mapper Jun 4 '14 at 15:44
  • 1
    @O.R.Mapper You of course cannot provide only the translation. – yo' Oct 28 '15 at 15:04
  • @yo': My word, but the statement "the translation of the title and journal are key. The title provides insights into the content of the paper and the journal makes it traceable." in this answers seems to claim otherwise. – O. R. Mapper Oct 28 '15 at 15:07
  • It is not a perfect world. In the Russian example above there is a Russian text with latin letters but if you consider a title in, for example, Japaneese or Chineese, then typesetting the original title with original typography would be quite cumbersome and hence it is rarely, if ever, done. I am not saying it should not be done but in reality the transalted text would typically be the only text appearing in the reference. – Peter Jansson Oct 28 '15 at 15:11
  • @PeterJansson I think "translation" and "transliteration" may be a problem here. I would be inclined to provide at least a transliteration of the title, if possible. Well, another thing is, if you identify the journal really well, then the title is optional, so ... – yo' Oct 28 '15 at 15:17

Your Answer


By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.