An article by Jim Yardley in today’s NY Times reports on the Dongxiang, an Islamic population in China’s Gansu province whose isolation has preserved their culture and language, which is part of the Mongolian branch of the Altaic family. Yardley writes:

For years, many Chinese scholars assumed that the Dongxiang descended from the Mongol soldiers in Genghis Khan’s army who eventually settled in Gansu during the 13th century when the Mongols ruled China under the Yuan Dynasty. But their exact origins were never fully known, an uncertainty that fed an inferiority complex.
“A man once asked me, ‘Where do the Dongxiang come from?’ ” said Ma Zhiyong, a historian who grew up in the county but moved to the provincial capital, Lanzhou, as a teenager. “I was 18 or 19, and couldn’t answer the question. I was ashamed.”
Mr. Ma decided to look for an answer. Over several years, he scoured research libraries in Gansu, talked to other scholars and studied old maps. He found that some Dongxiang villages shared names with places in Central Asian countries like Uzbekistan.
He also found shared customs. He said peasants in Uzbekistan and Dongxiang both learn to cut a slaughtered chicken into 13 pieces. He found that Dongxiang people described themselves as sarta, a term that once referred to Muslim traders in Central Asia.

There was even a physical similarity, as many Dongxiang look more like people from Central Asia, as opposed to Han Chinese.
Mr. Ma decided that the story about Genghis Khan’s army was only half right. Some of the Dongxiang ancestors were Mongol soldiers. But he concluded that many others were a diverse group of Middle Eastern and Central Asian craftsmen conscripted into the Mongol army during Khan’s famed western campaign. They brought several languages and, in many cases, a strong belief in Islam.
Mr. Ma said that generations of intermarriage, including marriages with local Han Chinese and Tibetans, resulted in a new ethnic group and language.

I presume the “new language” part refers to the fact that (by Ethnologue’s reckoning) 30% of the vocabulary is Chinese; it would be interesting to know how different it is from other Eastern Mongolian languages. I refer the curious reader to Oliver Corff’s nicely done page on “The Dongxiang Mongols and Their Language”; Corff says “This short article can hardly be called more than an appetizer for the interested reader,” but he’s too modest—it’s a lot more than I expected to be available for such an obscure language.


  1. The Mongol word kümün “man” mentioned by Corff is, of course, a derivation of the Anglo-Saxon gumman “man” (adapted for vowel harmony).

  2. i love your blog, language hat. it’s all so very interesting. sometimes you are too technical for me. my aspiration as a child was to be a linguist, which seemed to be too outlandish to the people around me. it’s not their fault, i just gave up, and now, especially when i read your blog, miss it…
    so thanks for filling in some of the details! i really love visiting!

  3. Robert Ramsey’s book The Languages of China says the people prefer to be called Monguor, which is their version of “Mongol”, as -r and -l have become indistinct. You can get more information by googling “Monguor”, including my summary of Ramsey on Recycled Knowledge.

  4. In The Languages of China, Ramsey distinguishes the Santa (Dongxiang) from the Monguor, though he places them both with the Gansu-Qinghai group of Mongolian languages. He also states that “the people call themselves ‘Santa'” (p. 199). See pp. 199-200 for the Santa and pp. 200-1 for the Monguor. I have some of this in my own entry on the recent articles.

  5. Yeah, Monguor and Santa/Dongxiang are two different groups. Ramsey has nice little sections on each.
    kasturi: I see from your blog you translate Danish poetry, so your interest in language has certainly not been wasted!

  6. My bad. Thanks for the corrections, and I shoulda looked it up.

  7. I’m wondering why Hat says Gansu and Wikipedia says Gansu but the link is to Kansu (which Wikipedia automagically redirects to Gansu).

  8. I probably typed “Kansu” into the search box because that’s how this grizzled old pre-pinyin type thinks of it. As to why I wound up putting Gansu in my text, it may be because the article used that spelling and I didn’t want to confuse people, though god knows I don’t usually worry my head about that.

Speak Your Mind