I have to return the Critical Companion to the Russian Revolution to the library soon, so I was looking through the section “Nationality and Regional Questions,” which I hadn’t yet investigated, and decided to read Martha Brill Olcott’s chapter on “The Revolution in Central Asia” and compare it with the account in my copy of Central Asia: 130 Years of Russian Dominance, edited by Edward Allworth. I immediately hit a snag. The Companion uses Russianized forms of the names of the locals: Muhammad Tynyshpaev, Halel Dos Muhammedov, Ali Khan Bukeikhanov, Ahmed Baitursunov, Mir Jakup Dulatov. Hélène Carrère d’Encausse (who wrote the relevant chapters in the Allworth book) uses forms that I presume are closer to the local-language versions: Muhamedjan Tanishbay-uli, Qalel Dosmahambet-uli, Aliqan Bokeyqan-uli, Aqmet Baytursin-uli, and Mir Jaqib Duwlat-uli, respectively. I can understand both choices, but it’s a shame that people already so marginalized by history are rendered even harder to investigate by such discrepant transliterations.


  1. Central Asian Studies is so marginalized and underfinanced that anything can happen. For example, the same word is transcribed Karakitai, Qarakhitai, and Qaraqitai. Kirghiz is transcribed Qyrqyz and Kirgiz. I don’t even know whether different dialects are being transcribed, or if Russian or Persian forms are being transcribed into Roman script, or if it’s just differences in transliteration systems salted with free-hand unsystematic forms.
    Juzjani’s history of the Mongol invasions was translated ~140 years ago and hasn’t been retranslated. His transliteration system garbles proper names so badly that the book is almost useless.

  2. I’m really pleased to have found this site. I did it quite accidentally – looking for some detailed info on semantic primitives which you discucced some three yrs ago:)
    Anyway, hope I find some free time to take a closer look at other posts as well.
    For the tim being, Merry Xmas.

  3. zuzentzailea says

    I’m a bit sceptical about the supposed Central Asian (Turkic) variants. I don’t know any language where [x] or pharyngeal [h] can come out as [q], and it’s not phonetically natural, and I don’t know of a script that uses that way.

  4. Wow, nice to read soemthing close and familiar from you, languagehat. Thanks for showing interest in Central Asia. Indeed, it has been quite a problem to transliterate our names in Russian and English. But the problem even (dazhe) remains with our country’s name: Kazakhstan (why not Khazakhstan?), Kazakstan, Qazaqstan. Seems like the last one would be the best.

  5. John Emerson says

    Iskrazhan seems to confirm my suspicion that the name of the state in question should be spelled either Kharakhitai, Qaraqitai, or Karakitai — i.e., that the initial consonants are the same. But as far as I know, these are among the less-common versions.
    Zuzent, Q is a common way of writing “Kh”. My argument is just that transliterations should be consistent.

  6. Charles Perry says

    Q is the conventional way (

  7. Charles Perry says

    Sorry, that post got truncated. Q is the conventional way of representing a back velar stop in Turkic languages, which is realized as a uvular stop in some and a velar or uvular fricative in others. So it’s not a case of a fricative becoming a stop but vice versa.

  8. Charles Perry says

    Rest of the post: The Russians spelled the ethnonym of the Qazaqs as Kazakh to distinguish them from the Cossacks. Since independence the Qazaqs have been trying to get us to spell their country as Kazakstan. We can just be glad they aren’t insisting in Qazaq and Qazaqstan.

  9. John Emerson says

    I’ve always thought that Qyrqyz would be a great Scrabble word, though Scrabble players tell me that it really isn’t.

  10. Several central asia countries now use latin alphabet (with diacritics), so transliteration is much simpler for them.

  11. David Marjanović says

    Just about nobody seems to respect those orthographies. Have you ever seen the spelling Türkmenbaşy reportedly concocted for himself?

  12. Speaking of orthography, congratulations on getting your ć back!

  13. Dmitri Minaev says

    Sorry, but you are not quite right. While the authentic, purist, forms do exist, they are disappearingly rare. Look at the government sites of, say, Azerbaijan (http://www.president.az), Kazakhstan (http://www.government.kz) or Uzbekistan (http://www.gov.uz). They give the following names of the presidents: İlham Əliyev, Нұрсұлтан Назарбаев, Islom Abdug’aniyevich Каriмоv. Note that the Uzbeks even use Russian patronyms.

  14. John Emerson says

    Google results: Karakhitai 760, Karakitai 579, Qarakhitai 137, Kharakhitai 41, Qaraqitai 16, Kharakitai 2, Qarakitai 1, Karaqitai 0, Kharakhitai 0.
    The name “Khitai” is well-known in English in that form, and perhaps the Kara- and Qara- prefixes were added to the existing English word without regard for consistently. Though as I’ve said, contamination by Russian, Persian, or even Mongol transcriptions is possible.

  15. David Marjanović says

    For a couple of days I’m back in Vienna, at the PC.

  16. Hasan MARAŞ says

    Don`t yuo worry, it is all Turkish.Yuo can find for every word almost every word a meaining in that languaege, so it is not russian not persian and not mongol either.
    Kıta(Khita-i) – means LAND
    Tay (Khi-tai) – means YOUNG HORSE
    What stuppider is is that some people say that kurdish could be an indo-european language, that is stupid !

  17. David Marjanović says

    Why, then, is it “qıtay” and not “qıtatay” or “tayqıta”? (Turkish has lost the difference between k and q in historical times; almost all other Turkic languages retain it.)
    On Kurdish… what is it, if not an Iranian language (or two, or three…) with a rapidly growing number of Turkish words? What do you suggest?

Speak Your Mind