Turco-Manjurica Revisited: a Closer Look at Haenisch 1951

Turco-Manjurica Revisited: a Closer Look at Haenisch 1951

Eric T. Schluessel
Ph.D. Candidate
Harvard University

Historical scholarship on Qing Xinjiang (East or Chinese Turkestan) has experienced something of a florescence in the Anglophone world since the publication of Millward’s 1998 Beyond the Pass, and figures importantly in most accounts of the New Qing History, with Perdue’s China Marches West earning particular attention.  However, despite the NQH priority on Manchu-language sources, advances in Manjuristics have not yet made themselves felt in this specialized field.

There are plenty of good reasons for this.  First, an informal standard has arisen in “Xinjiang Studies” that demands fluent reading ability in several other languages (Chinese, Chaghatay, Russian, Persian, etc.), and studying Manchu further increases the burden on the student. Second, there is a sense that the Manchu-language documentary record is scanty, especially for the later period of Qing rule from 1877 through 1911, during which the jurisdiction of the Ili general was formally limited to the Ili-Tarbaghatai Circuit. Moreover, scholarly attention has moved from the grand strategies that produced empire in the territorial sense, which are well-documented in Manchu monuments (though Perdue did not take advantage of these sources), and turned instead towards the infrastructure of empire, which appears as a Chinese enterprise that produced Chinese-language documentation.  While it might offend certain historiographical sensibilities to think that late-Qing Xinjiang was anything but a Chinese colony, something prefiguring Xinjiang’s relationship to the modern Chinese state, the idea that Manchu language, people, and politics retained their earlier importance up to the time of the Xinhai Revolution cannot be overlooked, even if it greatly complicates the genealogy of modern Xinjiang.

Recent scholarship on Qing Xinjiang from the earlier period, before the uprisings of the 1860s, has turned up some fascinating Turco-Manjurica, which is to say texts in Manchu and in Turkic, as well as Mongolian, documenting the interaction between Qing and Muslim authorities. Noda Jin and Onuma Takahiro’s documents and discussion of Kazakh-Qing relations stands out.[1] Laura Newby’s, Onuma’s, and Hua Li’s pieces in the recent Saguchi-Sugawara-Millward volume are not to be missed.[2] Looking towards the future, the publication last year of the 280-volume 清代新疆滿文檔案彙編, including facsimiles of documents produced as late as 1911 dealing with a wide range of topics, will, I believe, demonstrate the necessity of Manchu language for research on Xinjiang history generally. Continuing work on the Sibe, too, ought to be more fully integrated into a regional field that tends to focus on Uyghur history.

This is not, however, the first time Turco-Manjurica has come to light. I think it is valuable to revisit older work that seemed obscure at the time and see what it can offer from today’s perspective. To that end, I turn to Erich Haenisch’s 1951 article “Turco-Manjurica aus Turfan”[3], which presents two multilingual texts: one is a commandment from the Court of Colonial Affairs to the King of Qumul (Hami) in 1905. The primary text is in Manchu, and next to the Manchu words and phrases are their relative equivalents in Turki. (As both languages share a broadly similar syntax, the translation is not greatly garbled.) It bears no seal, and there are some errors in the spelling, so Haenisch concludes that it must be a draft or copy, but I must note that the copyist’s hand is far neater than in the majority of Turki official documents I have read. The other, a Manchu-Mongol-Turki phrasebook, was produced about a century earlier. Haenisch identifies it as the Manju Monggo Hoise ilan hacin-i gisun, which translates readily to the title of a work in Chun-hua’s 2008 catalog, 滿蒙回文三種譯語.[4] The phrasebook, including roughly 1,900 items, provides Mongol and Manchu translations for colloquial Turki phrases.

In the hope that these two works will soon be contextualized and joined by a richer Turco-Manjuristic corpus, I here present short passages from both pieces in a slight reworking of Haenisch’s German translation. I provide interlinear glosses and translations for both Manchu and Turki. I have also adopted Brophy’s transliteration of Turki.[5]

From the command to the King of Qumul:


Manchu: morin deleri feksibume isibu
Gloss: horse top gallop-CAUS-IMP.CONVERB arrive-CAUS-IMPER
Translation: Galloping on a horse, send this.
Turki: aṭ üstüni bilä čaṗturup yätkürsun
Gloss: horse top-3.POSS with gallop-CAUS-PERF.CONVERB arrive-CAUS-OPT
Translation: He shall gallop atop a horse and relay it.



tulergi golo-be dasara jurgan-i bithe
outer land-ACC rule-IMP.PART office-GEN letter
A letter from the Court of Colonial Affairs [lit. the Office that Rules the Outer Land].
līfāngänniŋ ḫaṭi
理藩院-GEN letter-3.POSS
Lifanyuan’s letter.


3-4 Hami-i jasak hoise cin wang sa de

Hami-i jasak hoise cin wang ša-de afabume unggihe afabume yabubure jalin
Qumul-GEN jasak/banner chief Muslim [<回子] 親王 Shāh-DAT “be charged with”-CAUS-IMP.CONV send-PERF.FIN “be charged with”-CAUS-IMP.CONV do-CAUS-IMP.PART reason
A command sent to Qumul’s jasak Muslim Loyal King Shāh. On the matter of: commanding him (to do something).
Qumūlniŋ jāsāk(ni?) yärlik čīn wāŋ šāhğä taṗšurup ibärgänimiz taṗšurup yürütürduğanniŋ učuri
Qumul-GEN jasak-GEN locality-ADJ 親王 Shāh-DAT find-CAUS-PERF.CONV send-PERF.PART-3P.POSS find-CAUS-PERF.CONV run-CAUS-PRES.PART-GEN dispatch-3.POSS
Entrusted by us to Qumul’s jasak local Loyal King Shāh. A dispatch entrusting him to do something.



ishun aniya aniya biya-i juwan uyun-i muduri erinde doron neimbi
next year “first month”-GEN 10 9-GEN dragon/5th branch time-LOC seal open-FIN
Next year, on the 19th day of the first month, at the dragon hour [7-9 AM], the [duties of the] seal will commence.
kelür yil birinči ayniŋ on toqquzi beliq sāʿatidä tāmğa ačadu
come-IMP.PART year 1-ORD month-GEN 10 9-3.POSS fish hour-3.POSS-LOC seal open-FIN
Next year, on the 19th day of the first month, at the fish hour, the [duties of the] seal will commence.


From the Manju Monggo Hoise ilan hacin-i gisun:

From dialogue 11:

siz burun tamākūsi tartamsizmu
2.S.formal nose tobacco-3.POSS “draw in”-INT-2.S.formal-INT
Do you snort snuff?
si oforo dambagu omimbio
2.S nose tobacco smoke[lit. “drink”]-FIN-INT
Do you snort snuff?


Dialogue 12:

siz apīn tartamsizmu
2.S.formal opium “draw in”-INT-2.S.formal-INT
Do you smoke opium?
si yarsi dambagu omimbio
2.S opium tobacco smoke-FIN-INT
Do you smoke opium-tobacco?


bizniŋ yärlik kiši apīn tamākūsi härgiz tartmaydu
3.P-GEN locality-ADJ people opium tobacco-3.POSS never “draw in”-NEG-PRES-3
We/our local people never smoke opium.
meni hoise niyalma yarsi dambagu umai  omire unde
3.PL-GEN Muslim people opium tobacco not at all smoke-IMP.FIN not yet
We/our Muslim people don’t smoke opium-tobacco at all.


Some thoughts: From these brief passages, we can detect something interesting with Manchu hoise “Muslim” and Turki yärlik “local.” Why should one be a Muslim in Manchu, but a local in Turki?

It is unclear from the text whether shāh is meant as a title (P. “king”) or if it is simply a name. The King of Qumul at this time was named Shāh Mahsūd (r. 1882-1930), so it is probably the latter case.

While the Manchu text uses Manchu titles, the Turki sticks to transliterations of Chinese. I am curious to know if there was any point when Turki translations were issued of Manchu titles, or if the Turki was always figured through Chinese, perhaps out of tradition or a sense that the positions bound the King of Qumul to a distinctly Chinese order. That said, Chinese words were at this point not uncommon in colloquial Turki. Later in the command text, the Turki uses a peculiar word, bänlämäk, to translate Manchu icihiyambi “to put in order.” Bänlämäk comes from Ch. 辦 “to manage,” to which has been added a verbal suffix. It could simply be the case that Chinese loanwords were at this time nativized to such a degree that their use even in official documents seemed natural.

Finally, we have different words for tobacco and opium. What are their etymologies? I sense that a research project on leisure, consumption, and the circulation of luxury goods is in order.

[1] Noda Jin and Onuma Takahiro, A Collection of Documents from the Kazakh Sultans to the Qing Dynasty, TIAS Central Eurasian Research Series Special Issue 1, (Tōkyō: The University of Tokyo, 2010).

[2] Laura J. Newby, “A Preliminary Discussion of Sources in Manchu Relating to Xinjiang (c.1760-1912)” in James A. Millward, Shinmen Yasushi, and Sugawara Jun, eds., Studies on Xinjiang Historical Sources in 17-20th Centuries, (Tōkyō: Toyō Bunko, 2012), 165-184; Onuma Takahiro, “A Set of Chaghatay and Manchu Documents Drafted by a Kashgar Hakim Beg in 1801: A Basic Study of a ‘Chaghatay-Turkic Administrative Document’”, 185-217; Hua Li, “Materials in the Manwen lufu Regarding Hui Muslim Migrants to Xinjiang”, 218-238.

[3] Oriens 4:2 (Dec 1951), 256-272.

[4] Chun-hua, Qingdai Man-Mengwen cidian yanjiu (Shenyang: Liaoning Minzu, 2008), entry 178. Chun-hua changes the character 回 “Muslim” in each of the titles to the anachronistic 維 “Uyghur,” emphasizing that the referent is the language of the Turkic Muslims. The compiler is not alone in doing so: I am seeing more and more collections of documents switch the characters 回 or 纏 “Turkic Muslim (pejorative)” to 維.

[5] David Brophy, “The Qumul Rebels’ Appeal to Outer Mongolia”, Turcica 42 (2010), 329-341.


  1. schluessel

    A couple of further thoughts:

    1. Note that the word 理藩院 is written līfāngän. w > g is a common sound change in Qumul dialect in this period, as demonstrated in Katanov’s interviews recorded in the 1890s.

    2. The verb bänlämäk is suggestive of what Bellér-Hann (2008, p. 81), after Jarring, refers to as a “Turco-Chinese jargon” used by officials in regular contact with the Qing. I wonder how extensive and consistent this jargon was…

  2. Really interesting, you made me requestthe whole set of 清代新疆滿文檔案彙編 for my library. Just a question, why do you translate 親王 as loyal king? Doesn’t it just mean imperial prince?

    • hasuran

      I think the author’s translation is sort of reasonable. In contemporaneous administration, 親王 and imperial prince were actually at different hierarchy levels.

  3. schluessel


    I’m really happy you liked it. You’re quite right about 親王, and Brunnert & Hagelstrom confirm your translation. (Entry 873) “Loyal king” is a reflex from taking Qing docs — this is how we translated Zhong Renjie’s adopted title. I admit, I was rather more focused on the Turkic and Manchu and just wasn’t thinking about the Chinese.

  4. Ah, here you are!

    Have missed The New Dominion very much, and look forward to reading more of your, uh, musings.

    Bruce Humes
    Ethnic ChinaLit

  5. xinjiangreview

    Very interesting piece on the use of Manchu and Turkic languages in XJ. I was wondering the definition of Hoise in this context:

    1. Does hoise originate from Huizi (which was used at least from the Ming dynasty referring to Muslims in China proper)? why did the Manchu not use Musliman (or Muslimilar or whatever..).?

    2. If so, does it indicate that the Manchu themselves have no knowledge about Turkcs in Xinjiang before entering into China proper? It seems the Manchu already recruited Muslims in their military before conquering Beijing (think about Huihui ying or Hui camps) ?

    3. If Hoise in Manchu is equal to yarllik in Turki-Uyghur (in Hami), does it mean that other Muslims or non-yarllik Turks/Uyghur in Xinjiang were not considered by Hami Shah as Muslim or Hui?

    4. It is also interesting to think about the Qing legal practices towards the Hami Yarlliks who migrated to Suzhou of Gansu that a Miao statutes were applied to them, does it mean that the Manchu see Hami Hoise as same as Hoise in China proper, or Chinese Muslims?


    • hasuran

      1. Yes, I think the word Hoise is organically from Huizi. Why not use Muslim? I don’t get it either. Theoretically speaking Muslim makes more sense than Hoise—The saying Hoise does not match the vowel harmony. Maybe because Chinese or Mongolian used Hoise more frequently than the word Muslim in those days…

      4. “does it mean that the Manchu see Hami Hoise as same as Hoise in China proper, or Chinese Muslims?”— Just my personal understanding, I think at least Hami Hoise and Hoise iin China were considered to be the same in that period of time. For example, in the Manchu version of the book ,欽定回疆則例 was written as Hesei Toktobuha Hoise Jecen i Kooli i Bithe. You can review the book if you can find it.

      • brotlein

        “Musulman” would make even more sense than “muslim”, since the former is what they call themselves.

        I don’t think it’s that abnormal to refer both Turkic-speaking and Chinese-speaking Muslims as Hoise…As a matter of fact, to regard them as belonging to different “ethnic” groups on the basis of their languages is a quite modern view of what we know as “nationalism”.

    • brotlein

      I don’t see why “Hoise in Manchu is equal to yarlik in Turki-Uyghur (in Hami)”. Hoise is a convenient (if not accurate) cover term for Muslims, since the Manchu rulers (or their Russian counterparts, I suspect, who applied “Tartars” to Central Asian peoples) didn’t really care about who’s who.

      On the other hand, “qumulning yarlik kishi” means “locals of Qumul” without referring to locals of other cities/towns, since the document concerned only the people of Qumul. In my opinion the muslim ruler of Qumul would also see muslims of other oasis cities equally as muslims. It was indeed a time when there was no unified identity called “Uyghur” which covered all Turkic-speaking muslims in the present-day Xinjiang. In a sense this new ethnic identity, just like those of the Kazakhs, the Kirghiz, the Uzbeks and the Turkmens, was forced into being by Russian/Soviet divide-and-conquer policies.

  6. xinjiangreview

    can anyone help answer these questions? thanks

Leave a Reply


Get every new post delivered to your Inbox

Join other followers: