Uyghur as a "dialect" — NOT

« previous post | next post »

The latest issue of The Atlantic has an article entitled "The Uighurs, China's Embattled Muslim Minority, Are Still Seeking an Identity".

The comments on language usage and policy in Xinjiang will be of particular interest to many Language Log readers, since they reverberate with a number of recent discussions that we've been engaged in.

Language, like so much else, is contentious in Xinjiang, where many Uighur grow up learning, at best, rudimentary Mandarin (putonghua), China’s official language. For most Chinese citizens, mastery of Mandarin is a priority. Local “dialects” are discouraged in the media and in education, and heavy accents turn many employers off.

Yet the language policy of the People’s Republic of China (P.R.C.) was surprisingly flexible from the start when it came to the ethnic minorities, giving minority tongues equal status as official languages in their own region, establishing minority-language schools, and encouraging Han cadres sent to the border regions to learn the local languages. Chinese bank notes throughout the country are written in five different scripts, including Uighur.

Among the Uighur, however, the policy has created two distinct groups: the minkaohan, minorities educated in Mandarin, and the minkaomin, educated in their own language. Minkaomin education is not taken seriously by non-Uighur employers, and not speaking Mandarin shuts minkaomin graduates out of jobs. In turn, they often resent minkaohan students as opportunistic and unfaithful to their own heritage.

It seems that, in a very vague fashion, Uyghur is being thought of as one of China's many "dialects".  At least the author, James Palmer, realizes that this is a problematic usage, so he puts the word within scare quotes.  Yet referring to non-Sinitic languages as fāngyán 方言 (almost universally mistranslated as "dialect") is nothing new, since I have encountered references to Uyghur (and even Italian!) as fāngyán 方言 in premodern Chinese texts.

For several recent Language Log posts that grapple with the problem of "language" vs. "dialect" in China, see:

Since the author cites the key terms mínkǎomín 民考民 and mínkǎohàn 民考汉, but doesn't translate them (the translations given by Google Translate, Bing Translator, and Baidu Fanyi are completely useless), I should note that the first means "ethnic group is tested in the ethnic language" (i.e., Uyghur, in this case) and the second means "ethnic group is tested in Han language" (i.e., Mandarin).

The author states that "the language policy of the People’s Republic of China (P.R.C.) was surprisingly flexible from the start when it came to the ethnic minorities…."  That may have been true "from the start", when the PRC was under the tutelage of the USSR, which did indeed have enlightened policies with regard to languages other than Russian.  But things have changed greatly since the 50s and early 60s, when China broke with the Soviet Union, and so — by now — Uyghur is often no more than window dressing.  See "Uyghur as ornament".

[Hat tip to Mariah Deters]


  1. Giles said,

    October 1, 2013 @ 12:49 pm

    Is "Fanyi" in "Baidu Fanyi" a word related to fāngyán?

  2. tsts said,

    October 1, 2013 @ 1:55 pm

    "It seems that, in a very vague fashion, Uyghur is being thought of as one of China's many "dialects". "

    That is not how I read this quote. It seemed to me that in the first cited paragraph the article talks about dialects of Mandarin (keeping aside issues such as the status of Wu/Yue etc.), and contrasts this with the attitude towards languages of ethnic minorities such as Uyghur in the second paragraph. Though your reading, that it contrasts current status in the first paragraph with earlier ideas in the second one also makes sense.

    I found it surprising to learn that at some point Han were encouraged to learn local languages – the number of Han Chinese living in Xinjiang or Tibet who know anything about the local languages seems minuscule. (I remember some report observing that in the main Chinese bookstore in Lhasa, there were no books available for Chinese who want to study Tibetan.)

  3. Victor Mair said,

    October 1, 2013 @ 2:06 pm


    Fanyi means "translate"; fāngyán means "topolect".

  4. Giles said,

    October 1, 2013 @ 2:08 pm

    @Victor — I see. Is the similarity etymological? Or do they just look similar from my non-sinophone perspective (eg. different tones) and are actually completely unrelated?

  5. Nuno said,

    October 1, 2013 @ 2:45 pm


    方言 (fāngyán) = 方(fāng) "place" + 言(yán) "speech"

    翻译 (fānyì) = 翻(fān)"turn over/flip over/rummage through/decode/translate etc." + 译(yì) "translate, interpret"

    Any similarity is pure coincidence.

    If you wondering why they don't they don't just say yì / 译 for "translating" it's to avoid confusion with homophones. Many character can be pronounced the same or similarly as 译. See also this:

  6. Nuno said,

    October 1, 2013 @ 2:48 pm


  7. J. Otto Pohl said,

    October 1, 2013 @ 3:19 pm

    The title of this post confused me. I thought you were going to argue against a claim that Uygher was a Turkic dialect or maybe a dialect of Chagatay along with Uzbeks to which it is closely related. Obviously linguistically Uygher is completely alien to Chinese. But, it is close to Uzbek and more distantly related to Turkish.

  8. Victor Mair said,

    October 1, 2013 @ 4:06 pm

    @J. Otto Pohl

    Well, if anyone ever made the claim that Uyghur is a dialect of Chagatay or Uzbek (which it indeed closely resembles), I would have argued against that. In this case, I wanted to emphasize that Uyghur should NOT be considered a fāngyán 方言 of Mandarin, even less so than Cantonese, Taiwanese, Shanghainese, etc.

  9. Jonathan Badger said,

    October 1, 2013 @ 4:50 pm

    Well, doesn't the cynical Yiddish saying about the difference between languages and dialects go: "A shprakh iz a diyalekt mit an armey un a flot"? And the Uyghurs don't have their own army or navy, so…

  10. J. W. Brewer said,

    October 1, 2013 @ 7:06 pm

    I have the vague sense that scholars who work on the classification of Turkic languages tend to be splitters rather than lumpers, so it is probably comparatively unlikely to find a reference work in which Uighur and Uzbek are treated as "dialects" of a single language (which I guess could be called "Karluk" or "Chagatayan" or something. But I don't know that that would be an unreasonable position if the history of the subdiscipline had been a bit different. From a political perspective, there was perhaps an incentive to play up rather than play down differences between historically related groups now separated by the Czarist/Manchu -> Soviet/PRC border.

  11. Kahar Barat said,

    October 1, 2013 @ 7:11 pm

    Let us put aside whether Uyghur can be a dialect of Chinese etc.
    Here is a very interesting phenomenon:
    Among 38 Turkic languages Uyghur and Uezbek are very close and categorized as one group. Historically Uzbeks were a branch of Qipchaq group sharing common elements with Kazakh, Kirgiz, Tatar, Bashqirt and others in far nor Volga region. Uzbeks came down and occupied Mavarannahr (Sir Amu rivers) of today's Uzbekistan in 15th century. Those nomads were overwhelmed by rich culture and Chagatai literature. In 2-300 years they totally became a new nation speaking chagatai based modern Uzbek. But in very underneath of Uzbek we still can see a strong Qipchaq foundation. This is a different kind of assimilation, within same language group.

  12. Victor Mair said,

    October 1, 2013 @ 9:17 pm

    @Jonathan Badger

    After all that poor Yiddish saying has been through here on Language Log during the past few months, I'm surprised that anyone would now cite it as evidence of anything.

  13. Levantine said,

    October 2, 2013 @ 2:13 am

    J. W. Brewer, you'd be surprised at how many pan-Turkists there are out there who like to claim that all Turkic languages are mere dialects, all of which are supposedly mutually intelligible. The fact that basic phrases can indeed be understood across these languages only fuels this absurd Turanian ideology.

  14. Giles said,

    October 2, 2013 @ 5:07 am

    @Nuno thanks! So, completely different roots, but in consonants rather than tones — 'ng' instead of 'n'

  15. Victor Mair said,

    October 2, 2013 @ 6:36 am


    The sounds of Mandarin — which, among Sinitic languages, is phonologically the most highly evolved and most heavily influenced by foreign influences — tell you very little about cognation. One always needs to be very careful about resemblances between the sounds of words in Mandarin, because they usually are so far removed from earlier stages of Sinitic that they are more than likely to lead you astray. For meaningful comparisons of the sounds of words that may be cognate, you need to look at the way they are pronounced in Wu, Min, Yue (Cantonese), and so forth, not Mandarin. Consulting reconstructions of earlier stages of Sinitic by reliable scholars is essential for any meaningful work on etymological relationships between Sinitic words.

    I will give a demonstration of one facet of this type of research in my next post (on the Chinese word for "poetry").

  16. Jeff Moore said,

    October 2, 2013 @ 8:15 am

    >方言 (almost universally mistranslated as "dialect")

    In Japanese, the same characters form the word ほうげん [hoːgeɴ]. It's the word Japanese linguists generally use to talk about dialects both in Japanese and in foreign languages, including Chinese. Could that be the source of the mistranslation?

  17. chris y said,

    October 2, 2013 @ 11:43 am

    Palmer has lived in China for many years and I'm quite sure he's as familiar with the implications of the use/misuse of fāngyán as he is certainly aware that Uyghur is a Turkic language entirely distinct from the Sinic family. But he's not writing an essay in linguistics here, he's writing about the politics of language in the PRC. Using "dialect" in scare quotes to indicate that these are not dialects as commonly understood in English speaking countries, but that they are sometimes treated as if they were for political reasons, seems fair enough to me. No doubt if he writes a book rather than a brief article on the subject he will discuss the question in more depth.

  18. J. W. Brewer said,

    October 2, 2013 @ 12:15 pm

    I think I agree with tsts that it's not clear (maybe better editing would have made it clearer one way or another) whether "dialect" is even meant to apply to Uighur, since the main point is that non-Sinitic languages like Uighur have been subject to different policies than Sinitic topolects like Cantonese. Whether being comparatively free of compulsory Mandarin instruction is in the long-term best interests of the students in question (in terms of career opportunities, social mobility etc.) is perhaps a difficult question, but it's presumably not an option available to most Cantonese speakers being educated in the PRC. It would be rather fascinating, however, if minkaomin students were able to outperform minkaohan students in mastery of English by taking advantage of the time saved from not trying to teach them Mandarin.

  19. Victor Mair said,

    October 2, 2013 @ 12:27 pm

    @J. W. Brewer

    Whether minkaomin or minkaohan, I think that they still have to learn Mandarin, even the former.

  20. Victor Mair said,

    October 2, 2013 @ 12:59 pm

    From Peter B. Golden:

    The ethnonym Uyghur really survived only among the Buddhist Shera Yoghur ("Yellow Uyghurs). The name Uyghur was revived quite consciously in 1921 by Turkistanian intellectuals to be used for the various Turkic, basically medieval Uyghur-derived dialects that were each called by local names (e.g. "Turpanliq"). The 16th century Tarikh-i Rashidi already remarks that the name Uyghur "is quite unknown at the present time; it is not understood which country is meant." It was for a long time NOT associated with Muslims. In short, modern Uyghur, comprising a number of dialects (Kashgharliq, Khotanliq et al.) is a revived and slightly altered, if not invented, tradition. The Uyghurs of today are certainly in part descendants of the medieval, post-imperial (post-840) Uyghurs who found East Turkistan more congenial after the Qirghiz drove them out of Mongolia, but they are also descendants of Turkicized Tokharians, East Iranians and most probably bits and pieces of other Turkic groups that came into the area. Oddly enough, the PRC helped to promote this "integration" of the various Turkic peoples, by its attacks on them (solidarity against a common foe) and governmental policy following Soviet guidelines. The Soviets, too, integrated disparate groups into a common ethnic identity. The Uzbeks of today are a very clear example. Interestingly enough, Uzbek is the Turkic language most closely related to Uyghur, both belonging to the southeastern or Turkî subgrouping of Turkic languages.

  21. Victor Mair said,

    October 2, 2013 @ 1:01 pm

    @my Turkologist colleague:

    "I wish we had a better English term than "dialect" for the various sub groupings of Chinese which, as you well know, are more than "dialects."

    We do have a better English term: it is "topolect", an exact translation of fāngyán 方言.

  22. Daniel said,

    October 2, 2013 @ 2:47 pm

    If the English equivalent of fāngyán 方言 is "topolect", what is the Chinese equivalent of "dialect"?

  23. Victor Mair said,

    October 2, 2013 @ 4:54 pm


    Thank you for your good question.

    I don't know of any Chinese words meaning the same thing as "dialect" that are current. In earlier times there were terms such as xiāngtán 鄉談 ("village patois; local dialect") that would have covered some aspects of "dialect", but not all. I once coined the Chinese term xiāngyán 相言 that means exactly the same thing as "dialect", but never tried to promote it because I was too wrapped up in trying to remedy the unfortunate translation of fāngyán 方言 as "dialect". Now that I've partially succeeded in overcoming that obstacle with the widespread acceptance of "topolect", perhaps I should turn my attention to popularizing xiāngyán 相言 as an accurate translation of "dialect"!

    For xiāngtán 鄉談, see, where we find that this term goes back as far as the Song period Classified Conversations of Zhu Xi (a large and highly vernacular text, by the way! — amazing for such an eminent Neo-Confucian philosopher [perhaps we can discuss the reasons for that on another occasion]) and the famous Ming novel Water Margins (or All Men Are Brothers; which is even more colloquial than the Classified Conversations of Zhu Xi, and skillfully translated as such by Pearl Buck). The term xiāngtán 鄉談 is still in use in the oral and performing arts of certain regions in China, but it is not a word that is used by linguists or even the general public.

    For xiāngyán 相言, see p. 8 in Victor H. Mair, "What Is a Chinese 'Dialect/Topolect'? Reflections on Some Key Sino-English Linguistic Terms", Sino-Platonic Papers, 29 (September, 1991), 1-31 (also available as a 2.2 MB PDF).

  24. Belial said,

    October 2, 2013 @ 5:21 pm

    In fairness to Jonathan Badger, the 'dialect with an army and a navy' bit has definitely been through the wringer here lately, but I don't remember it having been cited as a Yiddish saying. I hadn't heard of it before as a Yiddish phrase but it certainly has a greater poignancy in Yiddish than in English.

  25. J. W. Brewer said,

    October 2, 2013 @ 6:21 pm

    This is a small point, but I was curious enough to dig into the google books n-gram viewer and it appears that (in the American English subcorpus, at least), the spelling "Uyghur" only became more common than "Uighur" about a decade ago. "Uighur" apparently remains an established variant in respectable use (i.e., in the Atlantic article linked to above), and if left to my own devices I may well continue to use it myself. But I'm curious as to what if any factional preference may be associated with the choice between the spellings. If I use the once-dominant spelling "Sinkiang" rather than "Xinjiang" (or, even more so, "East Turkestan" or, let's say, "Occupied Kashgaria"), I at least have some sense of who I might appear to be choosing up sides with and what sympathies I might be taken to have, but I'm really not sure what signal I might be taken to be sending by virtue of sticking with "Uighur."

  26. Terry Collmann said,

    October 2, 2013 @ 6:43 pm

    Belial, the saying actually began life as a Yiddish phrase

  27. Victor Mair said,

    October 2, 2013 @ 7:03 pm

    @J. W. Brewer

    People have been shifting over to "Uyghur" because, as a transliteration, it more closely reflects the native orthography. So far as I know, it's purely linguistic and has no political ramifications.

  28. Dave Cragin said,

    October 2, 2013 @ 8:43 pm

    A discussion with a Belgian friend & his wife offered a completely different perspective on “what is a language?” Having read the many posts on the topic, I wanted their perspectives on language (As background, my friends come from a Flemish speaking area of Belgium, but they also lived in the Netherlands and speak English, French & German.)

    When speaking about their language, they always said Dutch. I said “don’t you call it Flemish?” He said we have a word Vlaams, but we almost always say Dutch (or more specifically Nederlands). He said “however, if we go to the Netherlands, they immediately know we are Belgian and they say we speak Vlaams, not Dutch.” Hence, the Belgians say the language they speak is Dutch and the Dutch say the Belgians speak Flemish – quite interesting.

    (He further noted that within Belgium, they can often pinpoint the town someone comes from based on the way they speak Dutch. That his hometown has its own dialect).

    Also interesting from a cultural perspective was the ease of above discussion, i.e., it was fun. There were no sensitive political or cultural overtones. In contrast, as Julie Lee, myself & others have noted, with some Chinese the topic of whether China has languages or dialects is so sensitive that it can’t be discussed – even with good friends.
    If any Belgians or Dutch want to add to the above and have different prespective, it would be interesting to hear your thoughts.

  29. J.W. Brewer said,

    October 2, 2013 @ 9:54 pm

    OK, so in the future if for some reason I want to refer to the Uighur word for "Uighur" without using arabic script I will try to remember to transliterate it as "Uyghur" while probably continuing to use "Uighur" as the spelling for the English word. Just as I will continue to spell the English word "Flemish" while transliterating the Dutch word with the same referent as "Vlaams." The objection that "Flemish" is not the best available transliteration of "Vlaams" is neither here nor there, because English demonyms/glossonyms are under no obligation to be transliterations of whatever the members of the demos in question may use as their autonym.

  30. David Dettmann said,

    October 2, 2013 @ 10:04 pm

    I always thought the reason why there was a persistence of "Uighur" as opposed to "Uyghur" in the U.S. (in library catalogs for instance) had to do with the conversion of Cyrillic into English with standardization systems. Library of Congress standard is like this. I spent a Summer entering in data from Arabic script Uyghur cassette tapes into our library's database, but I was forced to use the standard which was based on Cyrillic, because that was the system required for our local database to share with the LoC.

  31. Bart said,

    October 3, 2013 @ 12:16 am

    @ Dave Cragin
    Your Belgian friends call the language they speak Nederlands, not Hollands, I presume.
    Speakers of standard Dutch living in the Netherlands often call the language they speak Hollands.

  32. ZAT said,

    October 3, 2013 @ 12:44 am

    It is my understanding that Fangyan originally meant "variety of speech", with no distinction between language or dialect. Lots of older sources in Chinese use the term to refer to any non-standard language variety spoken within China, whether than variety is Sinitic or not. Fangyan, then, stands in contrast to Waiyu ("foreign language").

    This meaning of fangyan predates the connection with the European concept of "dialect" by a considerable amount of time, I think. For whatever reason, rather than devising a different translation for Fangyan, the word became associated with "dialect" and is now used with both meanings in China: as a non-Standard language native to China and as "dialect". I agree that translating fangyan as "topolect" sidesteps the problem altogether, but the cat is already out of the bag: linguists within China typically seem to use fangyan to mean dialect.

    Chinese writers calling Uyghur a fangyan are perhaps playing up the ambiguity on purpose. For one thing, it is presumably good for national harmony to emphasize that the Uyghur language is not a "foreign" language. On the other hand, it is on the same footing as the Qiqihaer dialect of Mandarin. If Qiqihaer Han Chinese have to switch over to Standard Mandarin to be civilized, then why should Uyghuers not have to?

  33. Victor Mair said,

    October 3, 2013 @ 7:33 am


    Thank you very much. Yours are some of the most sensible remarks about the recent (last century or so) history of the term fāngyán 方言 in China that I've ever read.

  34. J. Otto Pohl said,

    October 3, 2013 @ 8:13 am

    In Kazakhstan and Kyrgyzstan Uygher is written in Cyrillic and the most straight forward transliteration is Uighur. Interestingly enough a transliteration from the Russian word for Uygher renders it Uigur.

  35. julie lee said,

    October 3, 2013 @ 1:16 pm

    @Jonathan Badger:

    "A shprakh iz a diyalekt mit an armey un a flot".

    Thanks. I've seen this sentence in English but not in Yiddish. This is my first encounter with a sentence in Yiddish, and I'm glad I could understand it because it 's so close to German. (Would Yiddish be a dialect of German? )

    A glimpse of the language picture in Uzbekistan: As regards students' languages and examinations in Uzbekistan, I was at a bank in Palo Alto yesterday, and the 20-something bank officer who helped me had the name Vitaly with a German surname. He was from Uzbekistan, which he left at age 13. It turns out his mother's side was Russian-Jewish and his father's side German-Polish-Hungarian. He himself was born in Uzbekistan and lived there till 13. Since his great-grandfather's generation (in the 1930s), the family had been forced by political changes to migrate back and forth between Russia (Novosibirsk, Moscow, etc.) and Central Asia. His first language in Uzbekistan was Russian, and like many other Europeans in Uzbekistan he never learned Uzbek but learned German and French as second languages. After Uzbekistan's independence, Russians and Russian-speaking Europeans didn't feel welcome anymore, and many left the country. His family came to the U.S. He said in Uzbekistan you typically graduated from high school at 17. His 14-yr-old friends who knew Russian and not Uzbek and who couldn't leave Uzbekistan had a tough time—they had to learn Uzbek in 3 years in order to take the university entrance exams, now given in Uzbek. He has just married a girl from Kazakhstan, also Russophone, who had a background similar to his own. They are very happy to be in the U.S. How did they meet? At Starbuck's.

  36. J. W. Brewer said,

    October 3, 2013 @ 2:20 pm

    There are now many more Bukharan Jews living in a few neighborhoods in Queens than back in Bukhara itself (now located in independent Uzbekistan). Many non-Bukharans like to go to their restaurants, which apparently feature ridiculously huge portions of various roast meats (not pork, of course). I believe their traditional tongue is a dialect of Tadjik (Bukhara being in the corner of Uzbekistan where there are plenty of goyische ethnic Tadjiks as well as Uzbeks), and is therefore IE (Iranian, specifically – feel free to argue whether Tadjik is a "dialect" of Farsi or a separate "language") rather than Turkic.

  37. Levantine said,

    October 3, 2013 @ 7:22 pm

    Most scholars of Persian (and that, not Farsi, is the preferred name among such scholars) consider Tajik to be a variety of Persian, which is how Tajiks themselves traditionally thought of it. The Persian of Iran, Afghanistan, and Tajikistan are really one and the same language, in the same way that British English, American English, and Australian English are, and it's only for political reasons that people try to distinguish them as separate. That said, it is true that their use of the Cyrillic script means that speakers of Tajik Persian generally can't read the two other varieties, and vice versa.

  38. Alexander said,

    October 3, 2013 @ 7:53 pm

    @J.W. Brewer: Seems to me the spelling with "y" is also preferable practically, inasmuch as it forces you to pronounce "Uyghur" with two syllables, correctly. The spelling with "i" leaves open the possibility that the word should be pronounced with three.

  39. Kahar Barat said,

    October 3, 2013 @ 9:22 pm

    If we completely follow the phonemicirules, Uygur should be the accurate one. Unlike English and Chinese diphthongs, the Turkic Y and W are semivowels, they have clear syllabic functions. The Library of Congress and traditional western scholarship use I and U for transcription purpose. Now the Uyghur is more for transliteration of Arabic form. Also the Gh/G have no phonemic contrast, Gh appears in back sounds and G appears in frot sounds and no need to identify them with separate letters

  40. Dave Cragin said,

    October 3, 2013 @ 10:03 pm

    @Bart – Yes, my Belgian friends called the language they speak Nederlands (but they mainly used the term "Dutch" because we were speaking English). They did talk about the common use of the term Hollands in the Netherlands, but they didn't mention the use of Hollands as a term for the language.

    To link this to Chinese, you might know that the Chinese call the Netherlands “Helan” 荷兰 – a fairly close approximation of "Holland."

    [As a native English speaker, it struck me that Dutch can sound more like English than German (obviously Dutch & German are closer to each other than English). While English-Dutch-German are clearly "languages", I find the overlap in sounds across these languages interesting.]

  41. Rodger C said,

    October 4, 2013 @ 8:30 am

    What's the present status of the pronunciation (in English) "Weeg-er," which I used to hear in grad school?

  42. Levantine said,

    October 4, 2013 @ 9:54 am

    Roger C, I was a grad student at an American university until last year, and that was the usual pronunciation I encountered (and myself used). The 'correct' pronunciation strikes me as somewhat ostentatious. But I suspect that things will change the more that 'Uyghur' replaces 'Uighur'.

  43. julie lee said,

    October 4, 2013 @ 12:33 pm

    @ J. W. Brewer:
    My informant about Uzbekistan was from Tashkent.

    @ Dave Craigin
    Coming from English, and before learning German, I was puzzled that "Germany" was translated "De-guo" (literally, "de nation") in Mandarin, while the adjective "German" was translated "ri-er-man" in Mandarin. Also puzzled that Greece was translated "希臘 xi-la" {"xi" pronounced like English "she"] in Mandarin, read "hei-la" in Cantonese. Only later did I learn that Greece was "Hellas".

    @Rodger C:
    "Uyghur" in Mandarin is translated "Hui-u-er". Coming from English, I found one of the greatest puzzles in reading Chinese history was trying to match Chinese translations with their foreign originals, or rather, with their names in English. Many of the puzzles in ancient Chinese history have been solved by Sinologists. "Anxi" , the ancient Chinese name for Persia, for example, was Chinese for "Arsacid", i.e., the Arsacid Empire. I used to wonder where the dickens Anxi was.

  44. J. W. Brewer said,

    October 4, 2013 @ 3:16 pm

    Since there are essentially no native English words starting either "ui-" or "uy-" (let's see, there's uilleann pipes, which is in pertinent part an Irish loanword, and uygenbogaardtite, a mineral named for a Dutch geologist — and I'm not sure I've ever heard either of those said aloud or tried to say them aloud myself), neither spelling is very useful at cuing a given desired pronunciation (because no Anglophone will have a very good native-speaker/reader intuition as to what phoneme(s) is/are likely being represented), and it is, frankly, highly improbable to think that one is going to systematically yield better results than the other. The now-obsolete-in-English "Ouighur" might even be better, since the first syllable is apparently supposed to be pronounced more or less the way Anglophones think French "oui" is pronounced.

    If anything, it's the change itself that's problematic. By way of parallel, changing the English spelling of "Kirghiz" to "Kyrghyz" seemed to suggest that the word ought to somehow be pronounced differently than it had been when the prior spelling obtained, but without providing the Anglophone reader any clue (because the use of "y" does not in this context match up with any native-word orthographic conventions) as to what that new-and-better pronunciation ought to be. (For that one, I think what may have happened is that a change in English transliteration robotically followed a change in the Cyrillic spelling which was in that context meaningful – i.e., the new Cyrillic spelling usefully cued a Russophone: "pronounce it this way, not that other way"; the problem was in uncritically letting that change flow through to English orthography where it was confusing rather than meaningful.)

  45. J. W. Brewer said,

    October 4, 2013 @ 3:22 pm

    Note fwiw that the old 1st ed. OED (i.e. the one I have easy/free access to in hard copy) uses "Uigur" as the preferred English spelling (while acknowledging other variants) but in the etymology uses "uighur" as the transcription/transliteration of the "East Turkish" word (also given in its original spelling in Perso-Arabic script) from which the English word is derived. So they understood the very basic point that seems to be baffling others here that those two spellings constitute the answers to two different questions and thus need not, except by coincidence, be identical.

  46. Victor Mair said,

    October 4, 2013 @ 9:37 pm

    @J. W. Brewer

    From a Persianist colleague who has a special interest in the Tajiks:

    The historical relationship of Tajik to Persian is totally different from the relationship of Uyghur to anything, because it has to do with the history of Persian as a written language.

  47. Victor Mair said,

    October 4, 2013 @ 9:41 pm

    @Rodger C

    "American English pronunciation of Uyghur proper nouns"

  48. Levantine said,

    October 4, 2013 @ 11:57 pm

    I think I misunderstood what the alternative pronunciation of 'Uighur' might be. Based on the earlier post to which Prof. Mair just provided a link, it seems that even the variants pronounce the first syllable as 'wee'. The alternative I was thinking of (and which I described as ostentatious) uses the authentic Turkic diphthong (as in the recording here:

  49. Rodger C said,

    October 6, 2013 @ 12:05 pm

    @Victor Mair: Aha! Thanks.

  50. JamesP said,

    October 29, 2013 @ 8:24 am

    @tsts yours is the correct reading of the paragraph.

RSS feed for comments on this post